Draft not to be quoted

HOW ARE THE POOR AFFECTED BY

INTERNATIONAL TRADE IN INDIA:

AN EMPIRICAL APPROACH

UNCTAD

OCTOBER 2008

(Draft)

Study prepared under UNCTAD-MOIC-DFID project on “Strategies and Preparedness for

Trade and Globalisation in India”

2

3

Acknowledgements

The core team of the study consists of Abhijit Das

(Deputy Project Coordinator-UNCTAD India project),

Rashmi Banga (Senior Economist-UNCTAD India project),

Danish Hashim (Vice President-MCX, Mumbai), Seema

Bathla (Associate Professor, Jawaharlal Nehru

University), CSC Sekhar (Reader, Institute of Economic

Growth) and Ramma Sambamurty (Consultant, UNCTAD-

India project)

We are grateful to B. Goldar, Institute of Economic

Growth, for his guidance and support. We acknowledge

the support provided by UNCTAD headquaters,

particularly of Lakshmi Puri and Bonapas Onguglo.

4

5

TABLE OF CONTENTS INTRODUCTION ............................................................................................................... 7 CHAPTER 1 ..................................................................................................................... 13 CHAPTER 2 ..................................................................................................................... 79 CHAPTER 3 ................................................................................................................... 113 CHAPTER 4 ................................................................................................................... 142 CHAPTER 5 ................................................................................................................... 168 CHAPTER 6 ................................................................................................................... 184 CHAPTER 7 ................................................................................................................... 213 CHAPTER 8 ................................................................................................................... 240

6

7

INTRODUCTION

The growing volumes of international trade and lowering of the tariff barriers have triggered a continuing debate on the impact of trade on poverty. On one hand, there are scholars who argue, using the conventional theories of trade, that trade provides opportunities by expanding markets, infusing new technologies and improving productivity, which leads to overall growth. Further, higher trade benefits low-skilled labor-intensive production, hence increasing demand and wages of low-skilled workers in developing countries; since low-skilled workers are most likely to be in a situation of poverty, there would be a reduction in poverty. Some have also argued that trade helps in poverty reduction as developing countries pursuing an export-promoting strategy will have to maintain macroeconomic stability. This reduces inflation fluctuations to which the poor are most vulnerable. Therefore greater orientation to trade encourages countries to adopt macroeconomic policies, which invariably favour the poor. On the other hand, scholars have pointed to the complexities involved in the mechanism through which trade may impact poverty. The role of specialization, intra-industry trade and perfectly elastic supply of labour has been brought to the forefront in tracing the impact of trade on poverty. It has also been argued that the trickle-down effect from growth to poverty reduction is based on the assumption that economic growth is distribution neutral, if not distribution improving, which may not be true in many cases. One extreme of the debate argues that the potential benefits of economic growth to the poor are undermined or offset by the inadequate redistributive policies and by increases in inequality that accompany economic growth. The second extreme argues that despite increased inequality, liberal economic policies and open markets raise incomes of everyone in the society including the poor. This proportionally reduces the incidence of poverty. Along with the extent of growth effects on inequality, the direction has also been debated, where it is argued that more inequitable distribution of gains lead to inequality and higher growth. Further, this debate on the trade-poverty nexus has become more complex due to the methodological issues and data limitations. In the context of the existing debates, this study “How the poor are affected by

international trade in India: An empirical Approach” takes an alternative approach to the issue of impact of trade on poverty. Instead of estimating the net impact of trade on poverty, an attempt has been made to assess how the poor are affected by trade. The poor constitute the low income group. This approach does not attempt to arrive at the net impact of trade on poverty, since it is argued that trade may produce both winners and losers and it may not be desirable to compare the gains to losses, as losses may occur to relatively poorer section of the society and gains to relatively well-off sections or vice versa. The framework of the study involves tracing the role played by trade in influencing the four facets of human development, namely empowerment, productivity, equity and

sustainability. Since these are integral parts of the millennium development goals, the study traces the role played by trade in the progress made by India in achieving these

8

goals. An extensive exploration is conducted in each of these issues to trace the role played by trade.

Empowerment

♦ The study conducts an impact assessment to examine whether trade has empowered poor in terms of generating additional employment in the economy whereby more people are employed. Chapter 1 of the study estimates the extent of employment generated in 46 sub-sectors of the economy due to increase in exports from 2003-04 to 2006-07. The study

finds that a rise in exports in the period 2003-04 to 2006-07 increased employment by 26 million person years. If exports of each sector rise by 10% from the level of 2006-07, additional employment of 1.81 million will be generated in agriculture, which has the maximum number of poor.

♦ In the unorganized sector, which employs more than 80% of the total Indian workforce, the study estimates the impact of trade on employment and wage rates paid by the enterprises. Results arrived at in chapters 2 and 3 show that in the unorganized sector, enterprises belonging to export-

oriented industries employ more people and pay higher wages. However, it is only the relatively bigger enterprises, i.e., those which employ more than six workers that gain most from the export-orientation of the industry. Rising import competition is found to have adversely affected employment in these enterprises. Location of an enterprise, in terms of state to which it belongs has an important bearing on the impact of trade. Irrespective to the export-orientation of the industry, unorganized enterprises in states with higher estimated export-orientation are found to gain more from exports, while impact of import competition faced by the state does not vary across states. Statistically significant results with

respect to exports increasing employment in the unorganized sector are

found in Andhra Pradesh, Gujarat, Haryana, , Karnataka, Maharashtra, Punjab and Tamil Nadu.

♦ Focusing on the agriculture sector, chapter 5 and 6 of the study estimates the impact of exports and imports on wages of unskilled labour in agriculture and organized manufacturing. The results show that exports of agricultural products have not led to increase in wage rates of unskilled workers but imports of agricultural products have led to lowering of

wage rates of unskilled workers in agriculture.

♦ In organized manufacturing sector, the results indicate that exports have had a favourable impact on wages of unskilled labour. Strict labour laws

and downward rigidity of wages in India has perhaps prevented rising

import competition from displacing unskilled labour or adversely affecting their wage rates.

♦ It has been often argued that gains from trade are not gender-neutral and women tend to gain les than men. An in-depth analysis has been undertaken to estimate the impact of trade in gender employment. The

results show that increase in exports in the period 2003-04 to 2006-07

9

generated 9.38 million employment for women and 16.60 million for

men. The share of females in additional employment generated due to

increase in exports exceeds the share of females in total employment by

nearly 5 percent points. This suggests that increase in exports have reduced the gap between male – female employment in India. This result is corroborated by the estimations carried out for organized manufacturing sector, which show that export intensity has a positive and significant impact on women employment but imports have not led to any displacement of women employment.

♦ To gender sensitise trade policy in India and harness further gains from trade for women, the study has identified gender sensitive products which can be used in trade negotiations.

Productivity

♦ An important aspect of trade liberalization is to induce competition so as to increase productivity levels. Studies have found that as firms are exposed to international competition (through exports) and domestic competition (through imports), labour productivity rises. However most of the studies have been carried out for organised manufacturing sector. This study estimates the impact of trade on labour productivity in both organized as well as unorganized manufacturing sector. Further, impact of trade on labour productivity of both skilled and unskilled labour is carried out separately.

♦ The results of chapter 2 and 3 show that in the unorganized sector, export

intensity of the industry to which an enterprise belongs has a significant

positive impact on its labour productivity but import intensity conversely reduces labour productivity. But these effects are mainly experienced by enterprises with more than six workers. In the organized sector, the study finds that both export and import competition improves labour productivity. This has an important implication, which is, competition can

have productivity enhancement effects only after an enterprise achieves a certain threshold scale of production. Higher competition, whether domestic or international, may in fact lower productivity levels of very small enterprises.

♦ Results of chapter 1 highlight an important impact of trade on productivity, which is enhancement of women skills. Exports of IT enabled services have opened new avenues of employment for women in India which has encouraged more women to upgrade their skills. The

results show that higher exports of computer and communication

services from India have led to higher Female Technical Education.

10

Equity

♦ Whether trade induced growth is accompanied by higher/lower inequality is an important issue. Most of the earlier studies for India have examined this issue by comparing indicators of inequality, e.g. Ginni coefficients in pre and post liberalization period. However, this approach is unable to establish whether trade is the cause for rising/falling inequality. This issue is approached in this study by comparing the gains from trade across different income groups and segmented labour markets.

♦ The study estimates the incomes generated due to increased exports for people in abject poverty and those below poverty line. The results of the study show that the total income generated by increase in exports in the period 2003-04 to 2006-07 has been of Rs 2,364 billion, equivalent to USD 55 billion. However, total income generated for the people in the

lowest income group (i.e., people in abject poverty and those below poverty line) is only around 1.6% of the total income generated in the economy by increase in exports in the period 2003-04 to 2006-07. The poor have benefited from exports but the gains have been unevenly distributed with 70% of the income generated going to the top two income groups.

♦ The results arrived in chapter 5 indicate that though exports have increased wage rates of unskilled labour, it has led to a faster rise in wages of skilled labour. This implies that exports have led to higher wage

inequality between skilled and unskilled labour.

♦ Another dimension of unequal distribution of gains from trade is the gains that accrue to consumers and those which accrue to producers due to import liberalization. Consumers may gain from lower prices while producers may lose market share and their profits may shrink with higher imports. The results of the study conducted in oilseed sector indicate that the consumer gains from imports of edible oil have been high though

producers have suffered a substantial loss.

Sustainability

♦ Sustainability of social gains from trade is an issue of concern. Whether trade has led to social gains needs to be examined first. For this purpose, the study estimates the role played by trade in India’s progress in achieving Millennium Development Goals (MDGs). A Trade Related Development Index has been constructed (TRDI) using those development indicators which may be affected directly or indirectly by trade. The

causality tests show that trade has led to improvement in India’s progress in achieving MDGs. The improvement in some of the development indicators has been caused by trade.

11

♦ Results of chapter 1 also show that increase in exports have caused

improvement in Trade Related Gender Development Index (TRGDI),

which is an index constructed using trade related gender development

indicators that may be affected by trade, e.g. women employment, etc

♦ To assess the sustainability of increased incomes, the study estimates the extent to which domestic prices of agricultural products, which are extensively traded, are integrated to international prices. Contrary to the results arrived at by other studies; this study shows that domestic prices are integrated to world prices in rice, edible oils, tea and coffee. The analysis shows that the markets for groundnut oil, mustard oil and gram are reasonably well-integrated. Therefore, market mechanism can play a

reasonably effective role in case of domestic production shocks.

This study is a pioneering study in four major respects. Firstly, earlier studies on impact of trade on employment and wages have been at a disadvantage in terms of lack of trade data at industry level. Attempts were made to construct industry-level export data by aggregating firm level exports. However, many firms may not be listed firms in which case the firm-level export aggregatation may have a downward bias. In case of quantifying import competition faced by an industry, lowering of tariffs have been used for indicating higher imports by most of the studies. But lowering of tariffs does not necessarily translate into higher imports, especially when domestic supply is sufficient. To address this data concern the study constructs a concordance matrix between six-digit product level data (2002 Harmonised System of Coding) and three-digit industry level data (National Industrial Classification). Using this concordance matrix, the exports and imports of products have been matched to the respective industries to arrive at industry-level trade data. The industry level trade data is then used to estimate the impact of exports and imports on different characteristics of labour market. Secondly, the study undertakes the impact of export intensity of industries and import competition faced by the industries in the organized sector on employment, wage rate and labour productivity of enterprises in the unorganized sector. This traces an important channel through which the effects of trade can percolate to the poor. Further, the study estimates the effectiveness of the found to have a significant positive impact on the employment and wage rate in the unorganized sector. Thirdly, the study uses similar methodology to estimate the impact of trade in different sectors of the economy in the same period. The data at the enterprise level is taken from National Sample Survey (NSS) 62nd Round. The estimations for the unorganized manufacturing sector are carried out for around 83,000 enterprises for the year 2005-06. The trade data at the three digit industry level is matched to the enterprise level data using the industrial classification specified for each enterprise in NSS dataset. The estimations for the organized manufacturing sector are undertaken using a panel data for 78 industries for the period 1998-99 to 2004-05. The industry level data is extracted from Annual Survey of Industries (ASI) to which the trade data is matched using the concordance matrix. Similar labour demand and wage equations across sectors have been estimated.

12

This makes comparison of trade related effects across sectors possible, which may be extremely useful in case of trade policy formulation, where it is imperative to understand the implications of trade policies across sectors. Finally, the study estimates the extent to which exports in the period 2003-04 to 2006-07 have generated employment in 46 sectors of the economy and incomes for the people in abject poverty and those below poverty line. The employment generated is also disaggregated into employment created for men and women separately. Further, impact of exports and imports on gender employment in the organized manufacturing sector has been estimated. The study makes an attempt to identify gender-sensitive products which may form a practical basis for gender sensitization of trade policy. The chapter scheme of the study is as follows: chapter 1 of the study assesses the role played by trade in India’s progress in achieving Millennium Development Goals (MDGs) and undertakes causality tests between trade indicators and trade related development indicators. The chapter further estimates employment generated in 46 sectors of Indian economy due to rise in exports during the period 2003-04 to 2006-07. The increase in incomes across five income groups has been estimated which include people in abject poverty and those below poverty line. Chapter 2 estimates the impact of trade on labour markets in the unorganized sector. The impact on wages and employment of extent of export intensity and import competition faced by the industry to which the enterprise belongs has been highlighted. Chapter 3 undertakes the deeper analysis into trade-orientation of the state and its effect on labour markets in the unorganized sector. Chapter 4 quantifies the effects of trade on unskilled labour in agriculture and organisd manufacturing industries. Chapter 5 estimates the extent to which trade has affected the wage inequality of skilled and unskilled labour and differentially affected their labour productivities. Chapter 6 undertakes a sectoral analysis of trade and estimates the impact of imports of oilseeds and edible oil on producers and consumers. Chapter 7 discusses the extent to which domestic prices of agricultural products are co-integrated with international prices which reflects the ease of transmission effect of trade on domestic markets. Chapter 8 undertakes a detailed analysis of gender effects of trade by estimating the gender employment generated by exports in the period 2003-04 to 2006-07; it identifies the gender sensitive products and estimates the impact of exports and imports on gender employment in the organized manufacturing sector.

Draft not to be quoted

13

CHAPTER 1

Role of Trade in India’s Progress towards MDGs���� 1.1 Introduction The rising volume of world trade and the on-going integration of the world markets have led to an ever-growing debate on the relationship between trade, growth and poverty reduction. Not only has the causality between the three been debated, but the extent and form of relationship has also been questioned, with a yet to be arrived at consensus. There can be little doubt that, historically, trade has acted as an important engine of growth for countries at different stages of development. It has not only contributed to a more efficient allocation of resources within countries, but has also transmitted growth from one part of the world to another. Not all countries, however, necessarily share equally in the growth of trade or its benefits. Literature indicates that this will depend on various factors including- the production and demand characteristics of the goods that a country produces and trades; the domestic economic policies pursued; and the trading regime it adopts. Accordingly, what is becoming more important is not whether to trade but in what to trade; the terms on which trade should take place and how to maximize as well as sustain the gains from trade by protecting the vulnerabilities of the economy.

Trade is increasingly being used as an instrument for achieving some of the eight Millennium Development Goals (MDGs), which reflect a country’s progress towards inclusive growth and human development. However, the role played by trade as an instrument in achieving these goals is beset with several policy challenges both within the country and internationally. It is a well recognized fact that trade liberalisation does not automatically increase trade, let alone growth. The relationship between trade, growth, income distribution, and poverty differs across countries and is likely to evolve as an economy changes in structure. The changes in product and factor prices triggered by opening up of an economy inevitably produce both winners and losers in the short term. But the extent to which openness affects poverty and has a bearing on where and how these gains and losses occur depends on a number of external factors, such as stage of development of the country; the timing, scale, and sequencing of policy reforms; and pre-existing domestic and international conditions. Ultimately, the role of trade’s in a

country’s progress towards MDGs and in particular alleviation of extreme poverty and

hunger will depend upon the extent to which the poor are able to participate gainfully in

the expanding sectors.

In order to arrive at any meaningful direction for trade policy, it becomes important to quantify the role played by trade in a country’s progress towards achieving MDGs. This will not only indicate the gains/losses from trade but also extent to which they percolate

� This chapter has been contributed by Rashmi Banga (Senior Economist, UNCTAD-

India), Sudip Ranjan Basu (Economic Affairs Officer, UNCTAD) and Mahima Jain

(Lecturer, Delhi University) We are grateful to M.R. Saluja (IDF) for his valuable inputs.

14

to the poor. In this context, the chapter attempts to trace the role played by trade in India’s progress in achieving the MDGs. The analysis is undertaken at three levels; firstly, the study traces and contrasts the trends in trade-related factors along with the human development indicators for India; secondly, a Trade Related Development Index (TRDI) based on the trade-related factors that may influence progress towards achieving MDGs is constructed. A Trade-Gender Development Index (TGDI) based on trade-related factors that may influence MDG goal for “promoting gender equality and empower women” has also been constructed. Causality tests are undertaken between TRDI and TGDI with trade-related variables overtime. To corroborate this analysis, causality tests have also been conducted between trade-related variables and components of human development to identify the channels through which trade may have influenced human development; thirdly, an attempt is made to estimate the extent to which trade has affected employment and incomes of poor. Employment multipliers have been generated using input-output matrix for India and the extent to which exports has generated employment and affected incomes of people in different income strata is estimated.

The rest of the chapter is organised as follows: section 2 attempts to provide a theoretical framework for and existing debates on role of trade in progress towards MDGs. The existing literature on this issue with respect to india is also discused; section 3 outlines the trade policy framework for India in different sectors; section 4 traces the trends in India’s progress in Trade and trade related human development indicators; section 5 discusses the concept and methodology of Trade related Development Index (TRDI) and estimates the trend growth in TRDI. The results of causality tests between TRDI, TGDI and trade are presented; section 6 presents the results with respect to the causality between trade related variables and identified trade related development indicators which reflect progress towards MDGs; section 7 provides the estimated results with respect to impact of increase in exports in the period 2003-04 to 2006-07 on employment using input-output multipliers in 42 sectors of Indian economy and incomes of the poor. Finally section 8 summarizes and concludes.

1.2. Role of Trade in Progress towards MDGs: Theoretical Framework and Existing Debates

1.2.1 Role of Trade in Progress towards MDGs: Theoretical Framework

The Millennium Development Goals (MDGs) are eight goals that respond to the world's main development challenges. The MDGs reflect the consensus that development is ultimately about reducing human poverty and protecting human rights. Poverty may not result only from the lack of income and jobs but also arises from a lack of access to basic social services, lack of equity and empowerment. Trade may play an important role in a country’s progress towards MDGs, though it may not be sufficient. The eight goals of MDGs are broken down into 21 quantifiable targets that are measured by 60 indicators.

Trade may influence a country’s progress towards achieving MDGs directly as well as indirectly.

15

MDG goals which may be directly influenced by trade are:

Goal 1: To eradicate extreme poverty and hunger: Increase in export revenues may improve GDP per capita, which is one of the most relevant indicators of MDG progress. Exports may create additional employment and investments generating incomes for the people, especially for the low skilled and low income groups.

Goal 3: Promote Gender Equality and empower women. Trade creates employment opportunities in sectors that have higher women employment multipliers in developing countries1. This may improve the probability of women pursuing primary, secondary and tertiary education. Sectors that expand due to trade may not only create new employment opportunities for women but may lead to higher remuneration to women which affects their social dynamics and results in women empowerment.

Goal 8: Develop a Global Partnership for Development: Progress on global commitments for improved aid, fairer trade and steep debt relief will determine, to a large extent, the successful achievement of the first seven MDGs.

Apart from these three MDGs which are directly related to trade, other MDGs have targets which may be achievable through trade.

MDGs targets which may be indirectly achieved through trade.

Trade may play a crucial role in achievement of other MDGs like: Goal 4 and Goal 5: Reduce Child Mortality and Improve Maternal Health. Availability of better medicines and medical facilities through trade may enable a country to progress towards achieving this goal. Access to these facilities by poor may then be domestic policy challenge. Goal 7: Ensure environmental sustainability. Trade increases the awareness towards environment and a rise in environment friendly goods can change the production pattern in favour of environment sustainability.

Thus, trade may enable a country’s progress directly or indirectly in six out of eight MDGs. Further, if trade enables higher growth which leads to higher revenues, it can be used as a potent instrument to considerably fasten the progress of a country towards achieving MDGs. It therefore becomes important to examine the trade-growth-poverty nexus.

1.2.2. Trade-Growth-Poverty Nexus: Theoretical Framework and Existing Debates 1.2.2.1 Trade-Growth-Poverty Nexus: theoretical Framework

1 UNCTAD-India (2008) show export sectors to have higher women employment multipliers in India

16

The existing literature on trade and poverty has been categorized into the following two broad strands (a) which explains the static relationship between trade and poverty with resources and technology given; and (b) which explains the dynamic relationship between trade and poverty, which exists via growth. The static literature concentrates on two main channels through which trade can directly affect poverty irrespective of growth. These are (i) through employment effect and (ii) through stable macro economic policies, which indirectly influence the poor. The dynamic strand of literature, on the other hand, breaks down the relationship between trade and poverty into trade-growth and growth-poverty relationship. The literature on trade’s effect on poverty via employment draws strength from standard trade theories, e.g., Stolper Samuelson theorem according to which trade results in gains for labor (the primary asset of the poor) since it is the relatively abundant factor in most low-income countries. In this analytical framework one can alternatively assume that there are two types of workers, high-skilled and low-skilled, with the latter being the relatively abundant factor of production in developing countries. Higher trade would benefit low-skilled labor-intensive production, hence increasing demand and wages of low-skilled workers in developing countries; since low-skilled workers are most likely to be in a situation of poverty, there would be a reduction in the number of poor. Thus according to traditional trade theories, the poor will be the largest beneficiaries of trade liberalization in developing countries. However, the restrictive assumptions upon which the theorem is built are not sufficient to provide a viable interpretation of the complexity of the real world. It ignores the effects of complete specialization and intra industry trade (which in many cases bypass the most poor countries). Further, if one recognizes the possibility of different degrees of mobility of some or all factors over time, the income consequences of trade liberalisation get further complicated. This is demonstrated by the fact that studies, which estimate the effect of trade on wages of skilled and unskilled workers, arrive at ambiguous results. Krueger (1983) supports the traditional argument through a multi-country study on effect of trade on wages and employment. However, Feenstra and Hanson (1996, 1999) find that outsourcing from North to South results in a rise in the real wages of skilled labour relative to that of unskilled workers in both sets of countries; but this is consistent with the fact that the real wage of unskilled labour rise as well. Further, although in the longer run trade opportunities can have a major impact in creating more productive and higher paying jobs, this strand of literature tends to take employment as given. A common finding is that much of the shorter run impacts of trade and reforms involve reallocation of labor or wage impacts within sectors. This reflects a pattern of expansion of more productive firms—especially export-oriented or suppliers to exporters—and contraction/adjustment of less productive enterprises in sectors that become subject to greater import competition. According to Winters (2004), wage responses to trade will also depend to a large extent on the elasticity of supply of labour in a country. If the elasticity of supply of labour is zero, wages will increase but not employment, whereas if it is infinite, employment increases but not wages. In the case of supply of labour being perfectly elastic at the prevailing wage rate, which is the case in

17

many poor countries, the effect on poverty will depend heavily on what the additional workers were doing before accepting these new jobs. If they were engaged in subsistence activities, there is no change in their situation. Only if the switch into this labour market were so great as to significantly reduce labour supply to the subsistence sector and hence raise its wages, would there be a poverty impact. Another way in which trade can affect poor in a static framework is through macro economic policies that are followed to encourage trade. According to Bhagwati and Srinivasan (2002), trade helps in poverty reduction in the developing countries since countries going in for an export-promoting (as opposed to an import-substituting) strategy will have to maintain macroeconomic stability. This reduces inflation fluctuations to which the poor are most vulnerable. Therefore greater orientation to trade encourages countries to adopt macroeconomic policies, which invariably favour the poor. A major lacuna in the static approach is that by ignoring the existing dynamism in the economy, it is unable to suggest ways of using trade as a mechanism to alleviate poverty. The approach treats trade not as a means for attaining goals but as an end in itself.

1.2.2.2 Trickle down Effects of Trade: Existing Debates The dynamic approach to trade and poverty, views trade as a vehicle for attaining higher growth and thereby reducing poverty. Two kinds of relationship emerge in this approach, trade and growth; and growth and poverty and along with this has emerged a stream of literature which debate these relationships. Trade has since long been regarded as an “engine of growth” and this role of trade has been supported by both theoretical as well as empirical literature. But the net effect of trade openness on economic growth has been, and remains, a subject of controversy. On the theoretical side, since the time of Smith through Ricardo and Solow, trade has shown to allow a country to reach a higher level of income since it permits a better allocation of resources. The growth effects of trade openness are made much more explicit by the use of the new growth theory led by Romer (1986) and Lucas (1988). Within such framework, trade allows intensification of capacity utilization that increases products produced and consumed. Openness offers a larger market for domestic producers, allowing them, on one hand, to operate at minimum required scale, and on the other hand, to reap benefits from increasing returns to scale. Some of the empirical studies in the last two decades that support the role of trade in boosting growth include David Dollar and Aart Kraay (2001), which using data for 80 countries over 4 decades reiterates the fact that openness boosts economic growth and that incomes of the poor rise one for one with overall growth. Frankel and Romer (1999), use data for 100 countries since 1960 and conclude that openness does have a statistically and economically significant effect on growth. Sachs and Warner (1997) find that developing countries with open economies grew by 4.55% pa in the 1970s and 1980s, while those with closed economies grew by 0.7% pa. According to this study, open economies double in size in 16 years, whereas closed ones take a hundred years.

18

However, with the growing volumes of trade in the world in the last two decades, doubts have been raised not only on whether trade leads to growth or not but also on the direction of the relationship. It has been argued by some that faster growing economies trade more and therefore the relationship between trade and growth may not be one way. Taking account of this argument more recent studies have estimated impact of trade on growth by controlling for the endogeneity bias using instrument variable approach. Studies that question the growth-enhancing role of trade include Harrison (1996), Barro (1999), Rodriguez and Rodrik (1999), Nye, Reddy and Watkins (2002) and Rodrik (2000). These studies criticize the studies, which establish links between trade and growth, for their alleged lack of control for “other” economic policies and use of largely unsatisfactory trade policy indicators. Rodrik (2005) debates the use of instrument variable strategy in regression analysis to arrive at the effects of government policies with respect to trade, on growth. First, in this area of inquiry it is genuinely hard to find credible instruments, which satisfy both the exogeneity and exclusion requirement and second these regressions do not indicate how effective the purposeful policy interventions have been. Easterly (2004) emphasizes that the large policy effects uncovered in growth regressions are typically driven by outliers, which represent instances of extremely "bad" policies. More recently, Rodriguez and Rodrik (2001) observe that most studies make use of complex indices to establish the relation between trade and economic growth. For example, they are skeptical about the index constructed by Sachs and Warner (1995) which includes information on average tariffs, non-tariff barriers, adoption of central planning, state monopolies of exports and the black-market premium. The link of the last two components to trade policy was questionable. On the other hand, Frankel and Romer (1999) have constructed a variable trade caused by geographical factors to use as an instrument for trade/GDP ratios in a regression in which income levels are dependent. Although, using this trade share in the regression gives significant results, it is questionable on the grounds that this trade share may be acting as a proxy for geography’s direct effect on growth, say effect of climate on disease, international technology transmission etc. An interesting result by Warner (2003) shows that the unweighted average tariff rate on capital and intermediate goods did display a simple negative correlation. However, by using different data sets for growth (Barro-Lee and Pen World Tables version 5.6 and 6.1), Rodriguez and Rodrik found that the results of Warner could not be replicated. The results were consistent with the idea that there is a weak, insignificant statistical relationship between growth and tariffs. New developments in growth theory provide an explanation for questioning the growth regression framework for dealing with complex relationships like the openness-growth nexus. The effect of openness on growth depends on a country’s structural and

19

institutional conditions. Papers such as Chang, Katlani and Loayza (2005) and Dejong and Ripoll (2006) have attempted to take into account these contingencies. The existing voluminous literature on trade and growth is complemented by an equally large existing literature on growth and poverty. It is asserted that if (i) growth is distribution neutral, and (ii) trade enhances growth, then it can be argued that (iii) trade is beneficial for poverty. But the evidence, both theoretical and empirical, is much more complex than this. The trickle-down effect from growth to poverty reduction is based on the assumption that economic growth is distribution neutral or, if not, distribution improving. This is in contrast to the classical stylized facts theoretically consistent with Kuznets’ (1955) theory of capital accumulation as an inverted-U shape between level of development and inequality. In contrast to Kuznets hypothesis, Ravallion (2001) uses World Bank data and computation methodology to argue that growth is inequality neutral, spreading equally to the whole distribution, thereby arguing that economic growth is the main engine of poverty reduction. The same position is taken by Dollar and Kraay (2001a) who find a one-to one effect of growth on the income of the poor, so that the income distribution remains stable and, sometimes, improves. Alternatively, it is argued that though growth may be a necessary condition for poverty reduction, it is not a sufficient condition since much will depend on its distributional aspects, which may not be equitable in a liberalized regime. Much of the recent research and debate has focused on the extent to which poor benefit from economic growth (Ravallion and Chen 2003, Ravallion 2001, Ravallion and Datt 2000). One extreme of the debate argues that the potential benefits of economic growth to the poor are undermined or offset by the inadequate redistributive policies and by increases in inequality that accompany economic growth. The second extreme argues that despite increased inequality in the liberal economic policies and open markets raise incomes of everyone in the society inclusive the poor, which proportionally reduce the incidence of poverty. Along with the extent of growth effects on inequality the direction has also been debated, where it is argued that more inequitable distribution of gains lead to higher growth Apart from these debates, studies have traced the channels through which trade can affect poverty. Winters et al (2004) identify four main channels through which trade shocks, including trade liberalisation, are transmitted into poverty impacts. These include impact on wages and employment; prices of the tradable; taxes and spending; shocks, risks and vulnerability; and economic growth and technology. These impacts are transmitted through four groups of institutions, that are, households; enterprises; distribution channels and government. However, this literature lacks the corresponding microanalysis of the impacts of trade liberalisation, which is required to trace the static as well as dynamic effects of trade policy on household/individual welfare.

1.2.2.3. Review of Empirical Literature on Trade-Growth-Poverty Nexus in India

20

With respect to India, we find that empirical literature on this issue is very limited. Topalova (2005) tries to look into the differential impact of liberalization on poverty and inequality across Indian districts. She compares the poverty and inequality measures in the districts with industries that experienced greater liberalization (measured as a reduction in tariff barriers) to those whose industries remained largely protected. In the baseline specification, the district level outcome of measures of poverty and inequality is taken as a function of the district exposure to international trade. The regression equation also includes district fixed effects which takes care of unobserved district-specific heterogeneity and year dummies to control for macroeconomic shocks that affect the whole country equally. She arrives at the conclusion that lower levels of tariffs have been associated with significantly high levels of poverty (measured as the Headcount ratio and Poverty gap) for rural India. For the urban sector however, no such significant relationship between trade liberalization and poverty was found. Similarly, she finds that there isn’t any statistically significant relationship between trade exposure and inequality (measured as the standard deviation of log consumption and the mean logarithmic deviation of consumption) either. As an explanation for the statistically significant relationship between trade liberalization and poverty in rural India, she highlights the fact that the agricultural sector was also subject to trade reform and that even though the removal of restrictions on agricultural imports was gradual, the imports of the agricultural goods that were liberalized increased significantly in the post reform period. She finds that her results are robust even when one takes into account differences in initial industrial composition of various industries, the correlation between pre-existing conditions in districts and subsequent tariff changes, and the correlation between initial conditions besides industrial composition and subsequent changes in poverty. In conclusion, the results are explained in terms of rigid labor markets which prevent the reallocation of factors of production in the period post reform. The fall in the wages of workers who were associated with a lower income distribution and were employed in industries engaged in trade, was responsible for a slower reduction in poverty of a particular district. This effect was reinforced by the slow rate of growth of employment in registered manufacturing. In his study “Reducing Poverty and Inequality in India: Has Liberalization Helped?”, Raghubendra Jha tries to find out the impact of liberalization on poverty and inequality in India by analyzing trends in aggregate inequality and poverty in India and outlining the major characteristics of each at the state-level. He uses the headcount ratio, the poverty gap index and the Foster-Greer-Thorbecke measure as indicators of poverty. He calculates each of these measures for the parametric specifications of the Lorenz Curve given by the Beta model by Kakawani (1980) and the general quadratic model of Villasenor and Arnold. Three sub-periods are considered: 1951-63, 1964-90 and 1991 onwards. He finds that in the period post 1991, inequality was severely exacerbated, and

21

that poverty also rose due to the economic crisis of 1990-91 and subsequently declined albeit by a marginal amount. Jha uses changes in real wage as a proxy for movements in inequality and rural poverty. A regression equation with the real agricultural wage as the dependent variable and a time trend, inflation in the consumer price index for agricultural laborers, and a dummy for a good/bad monsoon year are taken as dependent variables. He finds that although variations in monsoons account for some fluctuations, the real wages in agriculture have been increasing over time. Additionally he finds that urban poverty has been higher than rural poverty, with urban poverty being highly associated (positively) with industrial growth. In a subsequent exercise he tries to find out the movements in poverty and inequality at the state-level. Calculation of the Gini coefficient for different states reveals that inequality has risen for most states in the post reform period. To find out whether there has been a long term convergence in the real mean consumption, the Gini coefficient and the Headcount ratio, he carries out the rank and level convergence tests. He arrives at the result that there is no convergence between states in ranks and there is a weak level convergence in poverty, inequality and mean consumption in the urban and rural sectors. Chand (1999) analyzes the impact of liberalization on four important crops in India: rice, maize, chickpea and rapeseed-mustard at a national and a farm level. In order to gauge the impact of trade liberalization on the abovementioned crops, he calculates the Net Protection Coefficient (NPC), which is the ratio of the domestic wholesale price and the border price of a commodity, for each of them from 1987/88 to 1996/97. A value of NPC that’s greater than one indicates that the commodity receives protection and would experience a decrease in domestic price if trade is (further) liberalized. Similarly, a commodity with a value of NPC lower than one would experience an increase in its domestic price with greater trade liberalization. He treats rice and maize as exportable and chickpea and rapeseed-mustard as importable. He arrives at the result that trade liberalization would lead to a surge in exports of rice and maize while rape-seed mustard oil would experience a substantial increase in imports. He finds domestic prices and production for chickpeas would not undergo any significant change with trade liberalization. He further conducts an exercise wherein he estimates the impact of liberalization on producer and consumer surplus with the aid of demand elasticity, supply elasticity and price linkage functions for rice, maize and rapeseed-mustard (since these are the only crops that were found to experience a reasonable change in domestic prices and production). He finds that for rice, trade liberalization would lead to an increase in producer surplus by Rs. 7, 237 million while it would reduce consumer surplus by Rs. 7, 545 million leading to a net decline in social welfare. For maize on the other hand, the gain in producer surplus has been estimated to be double the loss in consumer surplus leading to a substantial increase in social welfare. Rapeseed mustard could have either lead to a substantial increase in social welfare or a modest improvement in the same

22

(there are two possible scenarios that have been discussed in the chapter). Thus the impact of liberalization on social welfare would vary across commodities. A study by Gulati and Narayanan (2002) looks at the impact of only rice trade liberalization on poverty in various developing countries. This study gauges the impact of liberalization on domestic production of and prices of rice using an NPC analysis (the NPC values have been taken from Neilson (2002) and the Arkansas Global Rice Model; and finds that India along with Thailand and Vietnam would be the main exporter of rice in the medium-long run in the period post liberalization. They however, stress on the importance of looking at the lagged effect of trade liberalization on agricultural wages and employment rather than merely depending on the immediate impact effect that it has through changes in prices and production of rice, in order to study its impact on poverty. In addition to this, the long term impact the change in relative prices in an economy has on private and public investment is another factor that needs to be taken into consideration. In India for instance, an increase in the price of paddy for the period 1990/91 to 1999/2000 has been accompanied by an increase in the wages of unskilled agricultural labour, an increase in private investment and a decline in public investment (which was more than compensated by the increase in private investment). In conclusion the study stresses on the need for “behind-the-border reforms” (which enhance agricultural productivity and reduce the risk involved with production) to ensure that liberalization of any crop even though it leads to an increase in exports, wages and investment has a poverty alleviating impact in the economy.

1.3. Trade Policy Framework in India After independence, India adopted a mixed economy strategy with self-reliance being the principal objective. Import substitution and export pessimism was an underlying strategy assumption. However, doubts about the effectiveness of this policy regime arose as early as the mid-1970s and since then a series of reforms have been undertaken towards opening of the economy, though effective reforms have taken place only since the early 1990s. These reforms have furthered the globalisation process with respect to cross-border movement of capital, goods and services. Compared to many other developing countries, India has been a slow globaliser. Benchmarking the extent of liberalisation of India with respect to other developing countries, we find that the effective import duties2 in India in 1990 were on an average 45% while it did not exceed 25% in any of the Asian developing countries3. Consequently, the growth rate of merchandise imports remained low in India, i.e., 5% in 1980s and 9% in 1990s. In UNCTAD’s Trade and Development Index (2007), India climbed up the TDI ladder between 2005 and 2006. Over this period, its overall TDI score rose from 413 in 2005 to 433 in 2006, its InputMI score rose from 549 to 576 and its OutputMI rose from 277 to 290. India’s occupies 86th place in TDI 2007 report. Its relatively lower rank of 86 reflects both the problems which it must still confront (that of and its still unrealized potential

2 Effective import duties = import duty revenue / value of imports (source: World Bank 2005) 3 China-4%, Malysia-6%, Viet Nam-14%, Pakistan-25% and Sri Lanka-14%

23

India began to adopt open trade and investment policies from 1991. Some of the liberalisation measures taken were reduction of import tariff rates (e.g. the simple average of all rates were brought down from 71% in 1993/4 to 35% in 1997/98 to about 18.3% in 2004-054), simplification of FDI norms (reduction of FDI caps, permitting FDI in sectors previously not open), reforms in the infrastructure sector, deregulation and restructuring of the financial and telecom services sectors, relaxation of import and industrial licensing etc. Consequently, in 1990s, total imports of goods and services in India grew at a rate of 10% as compared to 5% in 1980s. Exports of goods and services also witnessed a rise from 7% in 1980s to 11% in 1990s with exports of services rising from 5% to 15% in this period. After more that fifteen years of liberalisation, India is the second highest growing economy in the world. The growth rate of the economy in 2007-08 has been 9.0 %. It ranks almost at the top in almost all the indices and surveys that rank economies around the world as future business destinations or market potential (e.g. AT Kearney, Goldman Sachs).5 There has also been a spurt in the growth of trade and FDI. However the impact of globalisation has been felt in different degrees and extent in different sectors of the economy depending mainly on the kind of policies that were adopted for the sector concerned as described below. 1.3.1 Agriculture The Indian agriculture sector continues to remain the mainstay for a large majority of the population (about 600 million people in India depend directly or indirectly on this sector). Given the large livelihood implications of trade measures exports, imports and foreign investment and related restructuring measures have not been adopted for agriculture. Agricultural policy in India has been guided by domestic supply and self-sufficiency considerations. In addition, the government provides support prices to farmers and supplies to the general population at low cost through the public distribution system6. However some measures of decontrol have been adopted in the last fifteen years. For example, - The removal of state controls of the inter state movement of certain grains and of

administered prices - Making the PDS more targeted, while continuing procurement by government

agencies (which in fact had increased- in part due to a rise in minimum support prices)

4 WTO Trade Policy Reviews of 1998 and 2002 5

According to a study by Goldman Sachs, Indian economy is expected to continue growing at the rate of 5% or more till 2050 and is

slated to become the fourth largest economy by 2050. According to ATKEARNEY 2004 Business Confidence Index India is the 3rd most attractive destination, ATKEARNEY Offshoring Index;2004 India is the Best off shoring destination and UNCTAD and Corporate Location Survey April 2004 India is among the top three investment hotspots. (Investing in India: A Comprehensive Manual for Foreign Direct Investment- Policy and Procedures, DIPP, MOCI, GOI) 6 India Trade Policy Review 1998

24

- Freeing of trade in agricultural commodities from controls (under the Essential Commodities Act 1955)

- The removal of licensing and stocking requirements and movement restrictions enabling free and unrestricted stocking and trading in wheat, rice, coarse grains, edible oils, oil seeds, and sugar

The sector still has a range of measures such as import and export controls, including tariffs, state trading, export and import restrictions and licensing etc. Some of the controls are imposed or relaxed in times of shortages, overproduction or price fluctuations, which are not very infrequent as the recent decision to import wheat and reduction of tariffs on wheat to zero percent to encourage imports, indicate. India has protected its agriculture sector from the import threats by keeping bound tariffs on agricultural products much higher than the average applied tariffs. Average bound agriculture tariff is 114 percent: 100% for primary products, 150% for processed products and 300% for edible oils. Certain items (comprising about 119 tariff lines) were historically bound at a lower level in the earlier rounds of negotiations. Out of these low bound tariff lines, bindings on 15 tariff lines (included skimmed milk powder, spelt wheat, corn, paddy, rice, maize, millet, sorghum, rape, colza and mustard oil, fresh grapes etc.) were successfully negotiated to revise the binding levels upward to provide adequate protection to domestic producers (under GATT Article XXVIII) in December 19997. Another important liberalisation measure has been the removal of quantitative restrictions in 2000 and 2001, which India had maintained for balance of payments reason (using GATT Art XVIII (B)), on 1429 items. Since by late 1990s India’s current account, capital inflows and foreign exchange reserves had been strengthened, the US brought a case in the WTO dispute settlement for the elimination of QRs and won the case. Almost half of the agriculture tariff lines in HS 2002 codes at six-digit level were protected by QRs. Agriculture has also benefited from price realignments resulting from the manufacturing sector reforms. However the sector is largely untouched by reforms. The government came out with a New Agricultural Policy in 2000 with an aim of attaining 4% growth. However, this growth target has mostly not been met. In fact the sector is plagued with low productivity and volatile growth. The average annual growth rate of value added in the sector has declined from 4.7% during the Eighth Plan (1992-97) to 2.1% during the Ninth Plan (2.1%)8. The share of agriculture as percentage of GDP has also steadily declined from 23% in 1999-2000 to about 19% in 2004-059. Indian agriculture has been plagued by low productivity and declining public investments. Two/thirds of area under cultivation is still un-irrigated consequently agricultural activities in India are still heavily dependent on monsoon. Among the other problems in agriculture, poor management of resources and, weak organisation of production and marketing structures stand out. The manifestation of these problems has

7 MOCI 8 The Economic Survey 2005-06 9 CSO

25

happened in different ways. Lately, high indebtedness of farmers and a growing number of farmer suicide cases have brought the condition of the agriculture sector in sharp focus in public discussions. While extensive discussions have taken place on whether globalisation can be blamed for such suicides, a closer look at the situation reveals many pitfalls in agricultural sector like failure of credit delivery and other institutional mechanisms, poor infrastructure, low yield, etc which may have contributed to this crisis. 1.3.2. Manufacturing Policies related to the Indian manufacturing sector have undergone considerable change since 1991 and thus the sector itself has experienced a rapid transformation. Various measures have been adopted to reduce trade and investment barriers, restructure, privatise and decontrol the sector. This sector has in fact witnessed maximum policy changes since 1991 compared to the other two sectors. One of the significant liberalisation measures that has been undertaken to boost trade is the reduction of peak as well as average customs tariff rates for manufactures. The peak rate of customs tariffs was 150% in 1991-92. Since then the rate has been reduced in the successive Union budgets with the aim of bringing India’s tariff rates in line with the rates prevailing in the South East Asian countries (which is about 5%). The simple average tariff rate was 81.8% in 1990, which declined to 29% in 2002. The peak rate for non-agricultural goods was brought down to 12.5% in 2006-07. Comparing the applied rate in 2001-02, with bound rate at the WTO, out of 3298 tariff lines bound by India at the WTO (mostly at 40% or 25%), 1040 lines had applied rates equal to the bound rates. Another significant measure that has been undertaken is removal of FDI caps and permitting FDI in the closed sectors. Almost all the sectors are now eligible for the automatic approval route through the Reserve Bank of India except for four cases: 1) activities requiring industrial licence 2) when foreign investor has a prior joint venture/technical collaboration/ trademark agreement in the same sector 3) proposals for acquisitions of shares in an Indian company in the financial services sector and where the SEBI regulations are applicable; and 4) all proposals falling outside sectoral policy/caps or under sectors in which FDI is not permitted. Some of the important areas that have been opened up for 100% FDI are the infrastructure sector (airports, power except atomic energy, manufacture of telcom equipment), mining, construction development projects, floriculture and horticulture, petroleum and natural gas except refining, SEZs and Free Trade Warehousing Zones etc. A few other sectors that have been opened up for FDI are cable network and direct-to-home (49%), defence production (26%), print media (newspapers and periodicals 26%, scientific magazines/speciality journals/periodicals 100%).

26

In addition to these liberalisation measures, there were a few other significant measures, e.g., - The abolition of industrial licensing in 1991 except for 18 industries, which was

reduced to 6 by 2002. - Adoption of a more export oriented foreign trade policy - Disinvestment in public sector units though in a relatively limited and phased manner - Reforms in excise duty regime and the introduction of the VAT - Opening up of a number of sectors which were exclusively reserved for the Small

Scale Industries The impact of liberalisation measures on the Indian manufacturing sector has been positive in terms of growth, productivity and competitiveness. The sector responded to reforms positively though there was some slowing down of growth in 1996/7 (mainly due to infrastructural bottlenecks), but there has been a major upturn in the industrial activity since 2002/3. Between 2000/1 to 2002/3 the industrial growth rate was 5.2% which rose to 7.6%10 in 2005/6. The share of industry in GDP in 2005/6 is 25.8%. Within the industrial sector, the share of registered manufacturing has improved from an average of 58.8% during 1970-82 to 65.7% during 1992-2004. Further, there has been an improvement in the relative share of textile products, chemical and chemical products, rubber and petroleum products and electric machinery11. Further, lowering of tariffs and removal of quantitative restrictions have not led to any significant increase in India’s imports12 and there is no evidence that the impact of globalisation and liberalisation measures on the Indian industry has been negative. In fact many studies have shown that liberalisation, both in terms of trade and FDI, has had a positive effect on industrial productivity and growth [e.g., Goldar (2004), Banga (2005) and Virmani et al (2004)]. FDI has also been found to have led to technology transfer to domestic firms and have had significant productivity and wage spillovers. However even though India has a liberalised FDI regime, its share in world FDI inflows is still quite low. It rose from 0.48% in 2002 (US$3.45 billion) to 0.82% in 2004 (US$ 5.34 billion). Compared to China, whose share in world FDI inflows is around 9.35%, India’s share it still insignificant13. Similarly the share of India’s exports in world exports has been quite low. In 2002, the share was 0.7% and between 2002 and 2005 it remained fixed at 0.8%, though India’s exports grew at 25.7% in 2004 and 21% in 2005. Again compared to China, which had a share of 7.2% in world exports, India’s share and volume of exports have been low14. However, India’s exports have a higher value-added content compared to the Chinese, which have a higher content of imported inputs. The industrial sector has been the major beneficiary of India’s exports boom. In 2004/5

10 Reserve Bank of India Bulletin, August 2006 11 Economic Survey 2005-06 12 ibid 13 ibid 14 ibid

27

India’s manufacturing exports constituted about 74% of its total merchandise exports15. Moreover, India has been able to move into more dynamic and new products since mid-1990. Clearly India has benefited from adopting more open policies for the industrial sector but it requires more liberalisation and restructuring measures, more effective implementation of its policies and an improvement in infrastructure to improve its share in the world market.

1.3.3 Services The services sector in India has undergone many changes in its structure and has grown faster than the other two sectors since early 1990s. The sector has had a high growth in the last 15 years in both absolute and relative terms (with respect to manufacturing and agriculture sectors) thereby increasing its share in India’s GDP, FDI and trade. The trend of growing importance of services in India is consistent with the similar trend in the rest of the world. However it appears that India’s services sector has grown faster than the global services sector. Between 1990 and 2003, the international trade in services has increased at the rate of 6% per annum. In India between 1994 and 2004, the services sector grew at the average rate of 7.9 percent per annum whereas agriculture grew at the rate of 3% per annum and manufacturing at the rate of 5.3% per annum. Within services, business services grew the fastest in this period followed by banking and insurance16. The contribution of services to India’s GDP is also higher than that of the other two sectors. The average contribution to GDP, between 2000-04 services contributed 50% to India’s GDP compared to 27% contribution of manufacturing and 23% of agriculture17. In recent years services sector has grown even faster and has increased its share in India’s GDP. Recording a double-digit growth in 2005-06, it now constitutes more than 60% of GDP18. The growth in the services sector has been driven by an increase in various activities such as - higher revenue generated by railway freight - increase in cargo handled by airlines - rise in domestic and international air travellers - greater number of new cell phone connections - growth in bank deposits and non food credit - expansion in business process outsourcing - larger number of tourist arrivals - more people arriving in India for the purpose of medical tourism Along with the growing importance of services in the Indian economy, the importance of services in India’s exports as well as imports has risen. India’s exports of commercial

15 ibid 16 Banga 2005 17 CSO 18 RBI Bulletin August 2006

28

services increased exponentially between 1994 to 2006 (from US$ 6031 mn to US$ 75056 mn) and its imports of the same in the same period increased by about 7 times (from US$ 8030 mn to US$ 63053 mn)19. As a result, the share of trade in commercial services in India’s GDP has been rising: in 2004 it was about 11.66%, while merchandise as a percentage of GDP in the same year was 24.9% 20. Services are different from goods as they are intangible and at the border measures (e.g. tariffs) are not applicable to them. Regulatory, licensing, recognition and other non-tariff measures are the significant barriers faced by services and thus the approach required for services is different from that required for industry of agriculture. Till date very few deliberate policy measures have been adopted to boost growth in the services sector. Most of the growth in services sector has been fuelled by the growth in information and technology (IT) sector and information and technology enabled services (ITES) sectors. Notably there has not been any integrated policy approach towards the services sector in India21. A variety of approaches have been adopted for different sectors. For example, while there was no systematic policy or even incentives for IT/ITES and other knowledge based services (e.g. R&D), liberalisation measures were adopted for sectors such as banking, insurance and other financial services, and telecom. Various measures that have been adopted for liberalizing the services sectors/sub-sectors are - Permitting setting up of private banks and new foreign banks - Allowing foreign insurance companies in the Indian market through joint ventures

with an equity cap of 26% - Passing of the Insurance Act and setting up of the Insurance Regulatory &

Development Authority (IRDA) - Foreign equity permitted in telecom services (74% for some services, 100% for

others) - Formation of the Telecom Regulatory Authority of India (TRAI) - 51% FDI permitted in retailing of single brands in January 2006 - 49% FDI in air transport services is permitted India undertook commitments in the GATS in 1995 covering a number of sectors such as financial, telecom, professional, R&D, construction, health and tourism services. However India’s commitments were much less liberalising than its unilateral regime even in 1995. India did not even commit fully to the Telecom Reference Paper of the GATS. In the subsequent years India liberalised its services sector by a large measure. Thus the gap between India’s commitments under the GATS 1995 and its present regime has widened even more. In the revised services offer, which is a part of the new round of negotiations launched in 2001 under the GATS. India improved upon its commitments and offered commitment on a few previously untouched sectors (e.g. accountancy, architectural, higher education, environmental, life and non-life insurance, transport,

19 www.wto.org, International Trade Statistics 2005 20 WDI 2005 21 Banga (2005)

29

wholesale trade) and improved upon its commitment on a few others (e.g. 100% foreign equity permitted in computer and related services and R&D services, FDI in banking permitted at 49% and number of licences increased from 12 to 20, increase FDI in most telecom services from 25% to 49%). As mentioned in the first section, an important aspect of the GATS is temporary movement of natural persons (the so-called Mode 4). India has a strong interest in export of both skilled and unskilled workers. One indication of the importance of the movement of people for the Indian economy is the remittances received from overseas Indians. In 2004, India received highest remittances (US$21.7 billion) ahead of China (US$21.3 billion) and Mexico (US$ 18.1 billion). India’s share in global remittance flows was about 10%22. There is a shift in the skill-endowment and direction of migration from India. Firstly, while earlier, Gulf was the favoured destination, more Indian workers are now going to the US and Canada. Secondly, Indians who are now migrating to the US and other developed countries tend to be more high skilled than the ones who used to migrate to the Gulf. The difference in skill levels is one of the reasons for higher average earnings of the workers, which in turn is partly responsible for increased remittances. India has requested its negotiating partners to make more liberal commitments in the current round of GATS negotiations in areas of its interest. However the offers India has received from its negotiating partners are not to its satisfaction. India’s interest is to get more liberal commitment on cross border supply of services (e.g. business process outsourcing) and temporary movement of natural persons. As noted earlier, recent security concerns have made the prospect of developed countries permitting movement of workers from developing countries much more bleak. The sectors of its interest are IT and IT enabled services and professional services (such as for super specialty hospitals, satellite mapping, accountancy and auditing, printing and publishing23). Other than gaining greater market access, India may need to get into agreements on accreditation, recognition of qualifications and licensing procedures to benefit from exports of these services. The GATS negotiations are partly tied to negotiations in other areas (agriculture, non-agriculture market access). India’s benefit from the current round of negotiations partly depends on how much market access India acquires in services and in other areas as well. India has a high potential in the services sector in the knowledge-based industry (IT, R&D), business services, movement of semi and skilled professionals and outsourcing/offshoring. The future growth prospects of the Indian services sector depend as much on international markets (e.g. for business process outsourcing, movement of people) as on the domestic market. The sector has benefited from liberal domestic policies and managed to take advantage of the globalisation process. To sustain and provide further push to the growth of the sector the current reforms process must be maintained.

22 Economic Survey 2005-06 23 Economic Survey 2005/06

30

1.3.4 India’s Bilateral, multilateral and regional trade agreements: current agreements and agreements under negotiation. India was a signatory of the GATT in 1948 and participated in the successive rounds of trade negotiations since then. Eventually India was a signatory and member of the WTO in 1995 and undertook commitments on agriculture, non-agriculture and services sectors as part of it. In the current round of negotiations (the Doha Round of negotiations), India has assumed an important role as a leader of developing countries and a significant member of the G-20. In the recent past, India has entered into various bilateral trade agreements, regional trading agreements, Preferential trading agreements, and Comprehensive Economic Cooperation Agreements. India is implementing an FTA with Sri Lanka and a comprehensive Economic Cooperation Agreement with Singapore. India along with Pakistan, Nepal, Sri-Lanka, Bangladesh, Bhutan and the Maldives signed the South Asian Free Trade Agreement on 6 January, 2005 and the agreement is being implemented in various stages. It has led to a significant increase in exports and imports within the region, although the rate of growth of imports has been higher than that of exports. Agreements that are currently under negotiation are the Indo-EU Trade and Investment Agreement, the India-Japan Economic Partnership Agreement/Comprehensive Economic Cooperation Agreement and the India-Korea Comprehensive Economic and Cooperation Partnership Agreement (CECPA). The Framework Agreement on the BIMSTEC (Bay of Bengal Initiative for Multi-sectoral Technical and Economic Cooperation) Free Trade Agreement which was signed in February 2004 by Bangladesh, Bhutan, India, Myanmar, Nepal, Sri Lanka and Thailand, as well as the Global System of Trade Preferences (GSTP) which is an agreement signed by 44 developing countries (including India) in April 1998 are also under negotiations. India also signed the Asia Pacific Trade Agreement (APTA) along with Bangladesh, Republic of Korea, Sri Lanka, China, and Lao People’s Democratic Republic on September 1, 2006. The third round of concessions has been implemented and the launching of the fourth round of trade negotiations has already been declared. A bilateral Agreement between India and Chile- India-Chile Framework Agreement on Economic Cooperation which was signed on January 20, 2005 has already been implemented. Agreements that have been implemented at some stages and are further being negotiated are the Preferential Trade Agreement between India and Mauritius, the Framework Agreement with South Africa Customs Union (SACU), India-Israel Preferential Trade Agreement and the Framework Agreement on CECA between ASEAN and India. 1.4. Trends in India’s Progress in Trade and Human Development India produces nearly 6.4% of world output and is home to nearly 16.9% of the World’s total population in 2006. Importantly, India’s share of population living under extreme poverty i.e. on income of less than 1$ a day is 34.3% (1990-2005, HDR 2007-08) These numbers more than double if a broader definition of poverty is used i.e. the number of

31

people living on less than 2$ a day which is 80.4% (1990-2005, HDR 2007-08) of the entire population. 1.4.1 Trends in Exports and Imports

India’s total trade, both in terms of exports and imports has been rising steadily with imports rising faster than exports. Exports as a percentage of GDP increased from 5.8% to 14% in 2006-07, while imports increased from 8.8% to 20.9%. Indlia’s merchandise trade has grown at more than 20% since 2002-03, posted 22.6% growth rate in 2006-07. The value of merchandise exports reached USD 111 billion in April-December 2007 (Economic Survey 2007-08) India’s share in world merchandise exports is around 1%. Merchandise imports, on the other hand, has grown faster, i.e., 24.7% in 2006-07. In 2006-07 it reached USD 185.7 billion A large part of the growth in imports has been due to both volume growth and rise in import price of crude oil. In the period post liberalization, the composition of India’s exports has experienced some change. The share of Agriculture and Allied activities in India’s exports has been fluctuating in the period 1994-95 to 2004-05. It was 16 per cent in 1994-95, increased to 19 percent in 1995-96 and peaked in 1996-97 to 20.5 percent. Subsequently it fell to 18.8 percent in 1997-98. The downward trend continued in 1999-00 and the share reduced considerably to 15.2 percent and further to 13.5 percent in 2000-01 and 14 percent in 2001-02. The share of Agriculture and allied activities in 2006-07 is 10.3%, while share of primary products in exports is 15.1% (Table 1.1). The share of manufacturing products in exports has also declined overtime. In 2000-01 it accounted for 78.8 % of total exports while in 2006-7 the share has declined to 68.6%. Within manufacturing, the share of traditional exports like textiles including RMG, gems and jewellery, leather and handicrafts has declined, while share of engineering goods and chemical products has been rising. This marks a shift in India’s export pattern. Interestingly, share of petroleum, crude & products (including coal) has risen significantly from 4.3% in 2000-01 to 15% in 2006-07.

32

Table 1.1: Commodity Composition of Exports

* Growth Rate in dollar terms Source: Economic Survey 2007-08 Unlike export pattern, India’s import pattern does not show too much of a change over past six years. POL continues to have a share of around 30-32%, while capital goods imports have increased 10.5 to 13%. (Table 1.2) Table1.2: Commodity composition of Imports

* Growth Rate in dollar terms Source: Economic Survey 2007-08

India’s tariff levels have experienced a significant decline ever since it embraced liberalization in 1991 (Figure 1.1). Considering Simple Average Tariffs, the tariff level dropped from 81.84 percent in 1990 to 56.34 percent in 1992. In 1997, the tariff level dropped even more sharply to 30.09 percent, but then increased in 1999 to 32.95 percent.

33

There was a marginal decline in 2001 to 32.32 percent, which further decreased to 29.12 percent in 2004 and 16.48 percent in 2007. As far as Weighted Average Tariffs go, the tariff level was down to 27.86 in 1992 percent from 49.58 percent in 1990. In 1997 this level was 20.14 percent and it increased in 1999 to 32.95 percent. In 2001 Weighted Average Tariffs declined to 26.50 percent and continued the downward trend in 2005 (13.42 percent) and 2007 (10.42 percent).

Figure 1.1. India’s Tariff 1990-2007

0.00

10.00

20.00

30.00

40.00

50.00

60.00

70.00

80.00

90.00

1990 1992 1997 1999 2001 2004 2005 2007

Simple Average

Weighted Average

1.4.2 Trends in Income Growth and Poverty Alleviation 1.4.2.1 Trends in Income Growth and Employment

There has been a steady growth in GDP per capita in India since 1980 and this has improved considerably since 1990s. Per capita GDP grew at an average annual growth rate of 2.9% in the period 1980 to 1990, while it grew at 3.9% in the period 1991 to 2000. From 2000-2005 the average annual growth of GDP per capita has been 5.24%. The growth of trade/GDP ratio in the corresponding periods were 0.75%, 5.5 % and 14.2%. The most informative statistics available on unemployment is the daily status of unemployment provided by NSSO24. Employment has almost doubled in the year 2004-05 as compared to 1983 but the growth rate of employment is much higher in the period 1999 to 2004-05 as compared to 1993 to 1999-2000, in which it actually fell below the level of 1983-84 to 1993-94.

24 NSSO gives the average level of unemployment on a given day, during the survey year and thereby

capture the unemployed days of the chronically unemployed; the unemployed days of usually employed who become intermittently unemployed during the reference week and unemployed days of those classified as employed according to the criterion of current weekly status.

34

Table 1.3: Employment and Unemployment Rate in India

Source: Economic Survey 2007-08

The unemployment rate has increased despite an expansion in work opportunities, from about 20.27 million unemployed during 1999-00 to about 34.74 million in 2004-05. This was because the labour force grew at a rate of 2.84%, higher than the increase in workforce. Unemployment rates are over two percentage points higher than ten years ago (at 6.07 % in 1993 compared with 8.28% in 2004) and the employment-to population ratio is lower than it was ten years ago. Whereas a part reason for decline in this ratio could be an extended time spent on education, it can also be indicative of low employment opportunities. This decline in growth of employment during 1993-94 to 1999-00 can also be attributed to lower absorption in agriculture. The share of agriculture in total employment dropped from 61 % to 57 % and further along the same trend fell to 52% in 2004-05. While the share of manufacturing sector increased only marginally during this period, Trade, hotel and restaurant sector contributed significantly to the overall employment. Table 1.4: Employment in India by Sector:

Industry 1983 1993-94 1999-00 2004-05 Agriculture 65.42 61.03 56.64 52.06

Mining & Quarrying 0.66 0.78 0.67 0.63

Manufacturing 11.27 11.10 12.13 12.90

Electricity, Water etc. 0.34 0.41 0.34 0.35

Construction 2.56 3.63 4.44 5.57

Trade, hotel & restaurant 6.98 8.26 11.20 12.62

Transport, storage & communication 2.88 3.22 4.06 4.61

Fin., Insur., Real est., & busi. Services 0.78 1.08 1.36 2.00

Comty., Social & personal Services 9.10 10.50 9.16 9.24

Total 100.0 100.0 100.0 100.0 Source: Economic Survey 2007-08

As per the National Commission for Enterprises in the Unorganized Sector (NCEUS), the organized sector employment has increased from 54.12 million in 1999-00 to 62.57 million in 2004-05. However, the increase has been accounted for by increase in

1983-

84 1993-

94 1999-

00 2004-

05

1983 to

1993-94

1993-94 to

1999-00

1999-00 to

2004-05

million Growth p.a. (%)

Population 718.10 893.68 1005.05 1092.83 2.11 1.98 1.69

Labour Force 263.82 334.20 364.88 419.65 2.28 1.47 2.84

Workforce 239.49 313.93 338.19 384.91 2.61 1.25 2.62

Unemployment Rate (per cent)

9.22 6.06 7.31 8.28

No. of unemployed 24.34 20.27 26.68 34.74

35

unorganized workers in organized enterprises, from 20.46 million in 1999-00 to 29.14 million in 2004-05. Thus, increase in employment in organized sector has been on account of informal employment to workers. 1.4.2.2. Trends in Poverty

In India, the estimates of the incidence of poverty by the Planning Commission on the basis of the head-count ratios given by various rounds of the NSS show that poverty has been consistently declining for the country as a whole. The percentage of people below the poverty line declined from 54.88 % in 1973-74 to 38.36 % in 1987-88. In the period post liberalization, this further fell to 35.97 % in 1993-94 and then to 26.1 % in 1999-00. These figures however are not comparable as they have been calculated for different recall periods- the uniform recall period (URP) for 1993-94, and the mixed recall period (MRP) for 1999-00. In 2004-05, the estimated poverty ratio by the URP method was 27.5 (making it comparable with the 1993-94 ratios) whereas by the MRP method it was 21.8 (making it comparable with the 1999-00 ratio). The Uniform Recall Period (URP) consumption data of NSS 61st Round yields a poverty ratio of 28.3 per cent in rural areas, 25.7 per cent in urban areas and 27.5 per cent for the country as a whole in 2004- 05. The corresponding all India figures are 36 per cent for 1993-94 and 27.5% in 2004-05.

Table 1.5: Poverty Ratios by URP (per cent)

Source: Economic Survey 2007-08

Nevertheless, the percentage of population living below the poverty line in India has been steadily falling. The graph below shows a slight increase in poverty over the last 6 years, but that could be due to differences in the method of poverty calculation.

36

Figure 1.2:Percentage of Population below Poverty Line

Source: Reserve Bank of India

1.4.2.3. Trends in Human Development Index (HDI) and Gender Development Index

(GDI)

The United Nations Human Development Report (2007) shows that India’s HDI in 2005 stands at 0.619, up from 0.611 in 2004. It ranks at 128 out of 177 countries of the world. In terms of Gender Development Index, India ranks at 113 out of 157 countries. This shows an almost zero count in terms of HDI-GDI rank indicating that both move together. But a negative count of (-11) for GDP per capita (PPP US$) – HDI rank indicates that the country as a whole is amassing wealth at a rate higher than which social inequalities are getting corrected. For 1995 to 1999, GDI grew by 0.129 percentage points, while 1999-2005 it witnessed an increase of 0.047 percentage points. Table 1.6: Trends in Human Development Index and Gender Development Index in

India

Source: HDR 2007

Year HDI GDI 1980 0.450

1985 0.487

1990 0.521

1995 0.551 0.424

1999 0.553

2000 0.578

2003 0.586

2005 0.619 0.600

0

10

20

30

40

50

60

1973-74 1983-84 1993-94 1999-00 2004-05

years

perc

en

tag

e o

f p

op

ula

tio

n

belo

w p

overt

y l

ine

0

10

20

30

40

50

60

rural

urban

37

1.4.2.4 Gender Employment and Literacy rates in India

Women’s employment in India following reform and liberalization has increased to the extent of about 5-10 per cent. During 1993-94, about 29 per cent of rural women and 42 per cent of urban women in India were classified as usually engaged in household duties only, while 33 and 16 per cent of rural and urban women respectively, usually carried out some (principal and subsidiary) economic activities ( source: ILO). In 2004-05, women within the ‘15 years and above’ age group were usually engaged in domestic duties, 33 per cent in rural India and 27 per cent in urban India had reported availability for ‘work’ on the household premises. Of them, about 72 per cent in rural areas and 68 per cent in urban areas preferred only ‘part-time’ work on a regular basis, while the percentages of such women preferring regular ‘full-time’ work were 23 and 28 percent in rural and urban India, respectively (source: NSSO)

Presently, for urban females, services sector accounts for the highest proportion (36 per cent) of the total usually employed, followed by manufacturing (28 per cent) and agriculture (18 per cent). Work opportunities for women in urban services and manufacturing sector exists but there is a need to facilitate and improve their Work Participation Rates through better education, skill development and removal of gender associated hurdles like lack of crèches, etc.

Table 1.7: Labour Force and Work Force Participation (in percent)

1983 1993-94 1999-00 2004-05 Labour force participation rates (LFPR) Rural male 52.7 53.4 51.5 53.1

Rural female 21.9 23.2 22.0 23.7

Urban male 52.7 53.2 52.8 56.1

Urban female 12.1 13.2 12.3 15.0

Work force participation rates (WFPR) Rural male 48.2 50.4 47.8 48.8

Rural female 19.8 21.9 20.4 21.6

Urban male 47.3 49.6 49.0 51.9

Urban female 10.6 12.0 11.1 13.3 Source: Economic Survey 2007-08

Coming to literacy, a huge gap exists between the male and female literacy rates. This gap continues to persist even after huge increases in female literacy. There has been no discernible attitudinal change and male education is still considered more essential than females. Males earn and stay with old parents while women tend to household needs in the house of her in-laws. There is thus no return on investment made in educating a girl child.

38

Table 1.8: Percentage of Literacy Rates by sex in India

Year Age Group Persons Males Females Rural Urban

Male-Female Gap in Literacy Rates

1971 5 and above 34.45 45.95 21.97 27.89 60.22 23.98

1981 7 and above 43.56 (41.42)

56.36 (53.45)

29.75 (28.46) 36.09 67.34 26.65

1991 7 and above 52.21 63.86 39.42 44.69 73.09 24.84

2001 - 64.8 75.3 53.7 - - 21.59 Source: Department of Secondary and Higher Education, Ministry of Human Resource Development, Govt.

of India.

By 2001, around 75% males and 50% females were able to read and write simple sentences. But there is a marked difference in rural and urban literacy rates. The 1991 census revealed that while literacy levels in urban areas were around 75%, those in rural areas were only 45%. This shows the level of lopsidedness that exists in human development in India.

1.4.2.5 Trends in Inequality

A more equitable distribution of resources ensures higher human development in a country. A skewed distribution on the other hand means that a significant proportion of population is getting a disproportionately lower share of the total pie. The conventional measure of inequality in income is Gini coefficient, which ranges from zero (absolute equality) to 1 (one person receives all the income). Growth is inevitably followed by some increase in inequality but the worst scenario arises when slow growth is accompanied by increasing inequality. Trends in Gini coefficient for India are similar to those seen for most developing countries i.e growth accompanied with rising inequality. Table 1.9: Trends in the Gini coefficient

Note: (c): Trends in consumption data The rural and urban Gini coefficients calculated for India give us an indication of the level of inequality in these areas. The rural Gini for India was 30.10 in 1983. It increased marginally in 1986-87 and 1987-88 and then dropped to a low of 27.71 in 1990-91. In the subsequent period it continued to increase and was back to the 1983 level in 1997

1980s average Early 1990s Late 1990s 2004-05

India 0.293 (c) 0.315 (c) 0.378 (c) 0.368

39

(marginally higher-30.11). The urban Gini on the other hand was 33.40 in 1983, increased to 33.95 in 1990-91 and further to 36.12 in 1997. Inequality therefore, has been on the rise in both the rural and urban sectors in India. But reveal that inequality is worsening more rapidly in rural areas than in the urban areas but the urban –rural spending gap that had widened in the 1990s, has started to close in the past five years

1.4.2.6 Trends in Environment Sustainability

Deforestation and pollution from industry, agriculture, domestic wood burning and human waste create increasingly hazardous living conditions. Many countries are losing their forest cover, fauna, glaciers and consequently the natural balance that is inherently maintained by the environment. Although temporal comparisons aren’t entirely correct as there has been a change in technology and scale of interpretation, however, it may still be observed that the forest cover of in India has remained between 19.5% to 20.5% in the last two decades. (Forest Survey of India) Figure 1.3: Forest Cover in India

Intensive Deforestation can lead to concentration of CO2 in the atmosphere, which leads to excessive warming at the surface of the earth through the Green House Effect. During the period 1990 to 2002, CO2 emission per capita has increased only marginally in Asia (from 2.2 to 2.5 tons per person) but it has increased rapidly in countries that have experienced energy-intensive economic growth.

The Indian economy today is the second fastest growing economy in the World and the sixth largest Green House Gases emitter. It now contributes to around 4% of the world’s total CO2 flow, up from only 1.86 % in 1990 and 1.13% in 1980.

Forest Cover in India (percent)

18.5

19

19.5

20

20.5

21

1987 1999 1991 1993 1995 1997 1999 2001 2003

years

Fo

rest

co

ver(

%)

40

Table 1.10: India’s share in world emission of CO2

Total emissions(MtCO2) CO2 emissions annual change(%)

CO2 emissions share of world total(%)

Population share(%)

CO2 emissions per capita(tCO2)

CO2 emitters

1990 2004 1990-2004 1990 2004 2004 1990 2004

United States

4,818.3 6,045.8 1.8 21.2 20.9 4.6 19.3 20.6

China 2,398.9 5,007.1 7.8 10.6 17.3 20.2 2.1 3.8 India 681.7 1,342.1 6.9 3.0 4.6 17.4 0.8 1.2

Source: HDR 2007

1.4.2.7 Trends in Energy Security

Data show that sudden spikes in oil prices can have an immediate adverse impact on the growth of booming economies. High oil prices hurt both, production and consumption. Although the developed countries are working towards decreasing the energy intensity, it may be very difficult for the developing economies. The oil exporting countries benefit from increased revenues and better terms of trade. Oil-importing countries on the other hand experience higher production costs, lower consumption and a net transfer of income to the oil-exporting countries. Between developed and developing countries, even if both are significant net importers, the effects are markedly different: the former use roughly half as much oil per unit of output as developing countries. In addition, terms of trade losses can be recouped more quickly by developed countries than by developing countries, given the different price elasticities of their tradable goods. In other words, there would be a variety of impacts depending upon the status of the particular country as an oil exporter or importer, its level of development, its resource situation and the policy measures that the Government takes to deal with the high oil prices. India’s own oil production has been increasing steadily and it has proven oil reserves of 5.625 billion barrels. Despite this increase, its own production accounts for only 30% of its energy needs and it remains a major oil importer.

41

Figure 1.4: India’s oil production and consumption 1990-2006

India's Oil Production and Consumption, 1990-2006*

0

500

1,000

1,500

2,000

2,500

3,000

1990 1992 1994 1996 1998 2000 2002 2004 2006

Year

Th

ou

san

d B

arr

els

P

er

Day

Consumpti

Productio

Net

Source: EIA International Energy Annual

2004 ;

Short-Term Energy Outlook (Jan. 2007)

*2006 is

estimate

In addition to oil, India’s consumption of natural gas has also been growing steadily and despite many new discoveries in recent years, its production of natural gas fell short of consumption by around 300 billion cubic feet. This is also the case with coal. Figure 1.5: India’s net Exports of Coal 1990-2006

This situation has arisen primarily due to the ever increasing demand in India for various sources of energy riding on its high growth ambitions. Though India has improved its GDP per unit of energy used (constant 2000 PPP dollars per kg of oil equivalent), demand remains sky-high.

42

Figure1.6: GDP per unit of energy use (constant 2000 PPP $ per kg of oil equivalent)

GDP per unit of energy used

0

1

2

3

4

5

6

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

Source: World Development Indicators 2005

1.5. Trade Related Development Index

1.5.1 Trade Related Development Index: The Concept The trends in different aspects of human development indicate the extent and direction in which a country is progressing vis-à-vis different aspects of human development namely, income, education and health. In order to have some quantification of the role played by trade in a country’s progress in achieving the targets set for MDGs, we construct a Trade Related Development Index (TRDI) for India. This index is constructed using time series trends in identified components of human development which reflect progress towards MDGs and are affected by trade. The change in the index overtime is a quantitative indication indicative of the direction and speed of a country’s progress towards those MDGs which are influenced by trade. The methodology adopted to construct TRDI provides the possibility to construct a composite measure to account for interactions and interdependence of various development indicators of the proposed development index.25 This analysis helps to identify the year-to-year change in trade-related development, and provides an estimate of growth rate of trade-related development in India. While constructing TRDI as a weighted average of several trade- related development indicators, it is crucial to determine weights to be assigned to each of these indicators. We propose a method of determining the weight that takes into account the variation in development-related indicators over the entire period of observations, viz. 1990 to 2006. The indicators used for constructing the index are listed with data in Appendix I.

25 The methodology is discussed in Nagar and Basu (2002) and Basu, Klein and Nagar (2005). See UNCTAD (2005, 2007) for application of this methodology.

43

1.5.2 Methodology for Constructing TRDI We propose TRDI as a latent variable which is linearly determined by various development-related indicators. Supposing that we have an adequate set of development-related (determinants TRDI), we can assume that the total variation in TRDI over years is accounted for by the variation in various development-related indicators and error variance is negligible. We exploit the total variation in all trade-related development indicators to arrive at the TRDI. For this purpose, we construct principal components (defined as normalized linear combinations) of various trade related development indicators, which have the property that the first principal component (P1) accounts for the largest proportion of total variation in all proposed development indicators, the second principal component (P2) accounts for the second largest proportion of total variation in all development indicators, and so on. If we compute as many principal components as the number of trade related development indicators, the total variation in all these development indicators is accounted for by all principal components together. However, as corr (Pi, Pj) =0, the trade related development indicators are not necessarily mutually uncorrelated. It is also true that the corr (Pi, Pj) = 0, i.e., the principal components are mutually orthogonal. A weighted average of the principal components

k

kk PPTRDI

λλλλ

++

++=

L

L

1

11

defines the trade-related development index.

In the present case K = 6 and 621 λλλ >>> L are successive eigenvalues of the 6x6

correlation matrix of observations on various trade related development indicators (and 5 indicators for TGDI). We have arranged them in descending order of magnitude. It can be shown that

6611 ,, λλ == PVarPVar L .

We assign largest weight ∑λλ i1 / to P1 because it accounts for the largest proportion of

total variation in all trade related development indicators. Similarly P2 has been assigned

the second largest weight ∑λλ i2 / because it accounts for the second largest proportion

of the total variation in all trade related development indicators, and so on. The detailed methodology is explained in Appendix II.

44

1.5.3 Trends in trade Related Development Index for India The estimated UN MDG based trade-related development index (TRDI) for India over the period 1990-2006 is presented in Table 12 . By fixing the base year to 1990 (i.e., 1990=100), the TRDI estimates the annual average percentage change in trade-related development indicators that are indicative of India’s progress in achieving MDGs. The indicators are transformed into their normalized form by subtracting the respective arithmetic mean from each observations and dividing by the corresponding standard deviation. This procedure helps to look into the speed of improvements of TRDI over specific time. Table 1.11: Trends in development index and trade-related development index in India (1990=100)

year

Trade-related Development

Index

Trade-related Development

Index (1990=100)

1990 4.605 100

1991 4.617 101.2

1992 4.634 102.9

1993 4.644 103.9

1994 4.661 105.8

1995 4.683 108.1

1996 4.692 109.1

1997 4.712 111.3

1998 4.715 111.6

1999 4.725 112.7

2000 4.755 116.1

2001 4.758 116.5

2002 4.765 117.4

2003 4.773 118.3

2004 4.787 119.9

2005 4.792 120.5

2006 4.797 121.1

The estimated indices show that there has been a steady rise in the India’s TRDI. The trend coefficient in the regressions estimate gives annual average rate of growth of DI/TRDI, which take the following form: εβα ++= )(*)log( timeTRDI

Now by running equation for TRDI, the coefficient of βTRDI =0.00036 which implies

that the growth rate of gTRDI = 1.26% for the time period 1990-2006.

45

Figure 1.7: Trends in Trade-Related Development Index in India (1990=100) Trade-related Development Index

(1990=100)

100.0

105.0

110.0

115.0

120.0

125.0

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

These results indicate that trade has played an important role in India’s progress in achieving MDGs. However, this analysis assumes that identified trade related MDG indicators are all affected by trade. To test whether the TRDI and the development indicators on which it is based are affected by trade or not, we undertake causality tests.

1.5.4 Trade Related Development Indicators and Trade: Causality tests Many studies have questioned the direction of trade-development nexus. The question that arises is whether trade leads to development or whether development leads to more trade. There is no consensus in this issue in the literature. It is possible that trade and development reinforce each other overtime. But the stronger effect of the two, i.e., trade causes development or development causes trade, may be taken as indicative of in which way the direction goes. To assess the direction of causation between trade and trade-related development indicators based on MDGs, we undertake the standard causality tests in the literature. The trends in Trade Related Development Index and the Trade Gender Development Index have shown a definite increase during the time period 1990-2006. This exercise is useful in analyzing whether increased trade is causing these improvements or does the causation flows the other way. 1.5.4.1 Methodology, Data and Varaibles

Methodology The relationship between the indices and trade has been studied empirically using econometric techniques. Since data is available in time series from 1990 to 2006, the time series techniques of Co-integration and Granger-Causality test have been used. Even though two variables may seem correlated, it may not necessarily imply causation in any meaningful sense of that word. But before testing for causality between variables, it is important to ensure that the two time series variables are stationary and they are co integrated. Broadly speaking, a stochastic process is said to be stationary if its mean and variance are constant over time and the value of the covariance between the two time

46

periods depends only on the distance or lag between the two time periods and not the actual time at which the covariance is computed. Say Yt is a stochastic time series, then Yt is said to be stationary if the following hold:

Mean: E(Yt) = µ

Variance: var (Yt) = E(Yt - µ)2 = σ2

Covariance: γk = E[(Yt - µ)(Yt+k - µ)]

where γk is the covariance between two values of Yt and Yt+k , i.e., between two Y values k periods apart. Stationarity of time series is essential to study the behaviour of the variable for time periods other than the period under consideration. This is useful for forecasting and studying the causality between variables. There are various ways to test whether a series is stationary or not. If it is found that a series is non-stationary, then we can subject the first differences in the time series to the stationarity tests to see whether they are stationary. If the first difference of the series comes out stationary, we say that the series is integrated of order 1, denoted by I(1). Similarly, if a time series has to be differenced twice to make it stationary, we call such a time series integrated of order 2 [I(2)]. But if the series are found to be non stationary, we use the Bounds testing Approach, developed by Pesaran and Shin (2001) also called the autoregressive distributed lag (ARDL) approach. The detailed methodology is given in Appendix III. This is the only methodology that permits the test of co-integration between variables integrated of different orders. This methodology is most suitable for a small sample, such as ours as the power of this test lies in the span of the data, rather than the number of observations. Further, if two variables are non-stationary, the regression of one variable on the other may produce a spurious regression. However, if the errors generated from the regression are stationary, then the two variables are said to be co-integrated. That is, if two variables are non-stationary, but the linear combination between them is stationary, then the series are said to be co-integrated. Ordinarily, regressions reflect only correlations but Clive Granger argued that there is an interpretation of a set of tests as revealing something about causality. The Granger (1969) approach to the question of whether x causes y is to see how much of the current y can be explained by past values of y and then to see whether adding lagged values of x can improve the explanation. y is said to be Granger-caused by x if x helps in the prediction of y, or equivalently if the coefficients on the lagged x's are statistically significant. It is important to note that the statement "x Granger causes y" does not imply that y is the effect or the result of x. Granger causality measures precedence and information content but does not by itself indicate causality in the more common use of the term. If an event

47

A happens before event B, then it is possible that A is causing B. However, it is not possible that B is causing A. In other words, events in the past can cause events to happen today. Thus, it is important to study whether the past experience of India in terms of trade growth has caused the present improvement in development quality measured by the development indices. Following section presents the results of the various tests explained above that have been undertaken to study the causal relationships. Data The data for the various trade indicators have been taken from WITS, Handbook of Statistics on Indian Economy published by RBI, and World Development Indicators. Data for the MDG indicators have been taken from World Development Indicators, Indiastat.com, Annual Survey of Industries and the Handbook of Statistics on Indian Economy. Variables The following is a description of the development variables used in the causality analysis.

SNo Variable name Units Time period

MDG Indicators

1 Number of industrial workers in 1000s 1980-2003

2 Salaries of industrial workers Million Rupees 1980-2003

3 Total women employment in public and private sector

in 1000s 1985-2005,

4 Rate of growth of number of females per 100 males in University Education in Technical and Engineering Disciplines

%ages 1981-2001

5 Rate of Growth of Computer, communications and other services (% of commercial service exports

%ages 1981-2001

6 Infant Mortality Rate ratio 1980-2005

7 Immunization, DPT (% of children ages 12-23 months)

%ages 1980-2005

Trade Indicators

1 Rate of growth of exports %ages 1980-2005

2 Total Industrial exports Constant 2000 Million dollars

1988-2005

3 Total Industrial imports Constant 2000 Million dollars

1988-2005

4 Exports/GDP ratio ratio 1980-2005 5 Imports/GDP ratio ratio 1980-2005 6 Imports of medicinal and pharmaceuticals Constant 2000 1980-2005

48

products Million Dollars

7 Rate of growth of total textiles exports %ages 1981-2005

1.5.4.2 Empirical Results: Trade causes TRDI and TGDI

The trade variables included for testing the causality are total trade as percentage of GDP and total exports as a percentage of GDP in India for time period 1990-2006. The data on these variables was extracted from World Development Indicators, World Bank. The indices, viz., Trade Related Development Index (TRDI) and Trade Related Gender Development Index (TRGDI) have been used to test for causality with the trade variables. TRGDI is constructed using those gender development indicators that may be influenced by trade, e.g., women employment to total employment in organized manufacturing; girls to boys school enrollment ratios, etc. Higher job opportunities made available by trade for women may encourage girls’ education. The trend growth rate of TRGDI over the period 1990-2005 is found to be 0.76%. However, the cause-effect relationship between trade and TRGDI remains to be tested. The stationarity tests were conducted for each time series variable. It was found that all the series were integrated of order 2, that is, their second differences were found to be stationary. Each of the indices was then subjected to co-integration test with the trade variables. The co-integration test that checks whether the linear combination of the two variables is stationary or not revealed that all the development indices are well co-integrated with the trade variables. Except that TGDI was not found to be co-integrated with trade as a proportion of GDP. This may be because in developing the Gender Development Index is based on the indicators such as school enrollment at various levels of education, proportion of seats held by women in parliament and total women employment in the organised sector. Most of these may not be influenced much by the total trade which comprises both imports and exports. However, exports as a percentage of GDP is found to be well co-integrated with TGDI possibly because growth in export sector, especially in sectors like textile, wearing apparel etc. has boosted female employment. The results of the Granger Causality tests are presented in Table 13 (a&b). These results confirm that trade has played an important role in India’s progress in TRDI. Testing the causation between trade as a percentage of GDP and TRDI, we find that trade causes TRDI which implies that the progress achieved in these development indicators is better explained by trade. Using exports to GDP ratio does not reflect causation, which indicates that imports also play an important role along with exports in progressing on development goals. On the other hand, exports as a percentage of GDP significantly cause the Trade-Gender Development Index but trade as a percentage of GDP is not found to have a significant causation effect. This indicates that exports may be empowering women in terms of additional employment created and/or better returns to labour.

49

Table 1.12 (a) Granger Causality Tests: TRDI and GDI are Granger caused by Trade

Trade Variables ↓↓↓↓/Indices →→→→ TRDI TGDI

Trade as a percentage of GDP Yes No

Export as a percentage of GDP No Yes

Table 1.12 (b): F-Statistic and P-Value for Granger Causality Test between Different Variables

Variables F-Stat P-value

TRADEGDP � TRDI 4.16299 0.3552

EXPORT/GDP �TGDI 8.26111 0.01062

The above results shows that trade is one of the important factors that has helped India to progress on achieving the Millennium Development Goals, the progress towards which has been consistent over the years. 1.6. Empirical Estimates: Trade and MDG based Development Indicators: Causality

The results in the previous section show the direction of causality between trade and TRDI and TGDI. However, it is important to identify the MDGs-based indicators that are affected by trade. GOAL 1: To eradicate extreme poverty and hunger.

Standard economic theory (Hecksher-Ohlin model) suggests that gains to trade should flow to abundant factors, which suggests that in developing countries, unskilled labor would benefit most from globalization. If labor is in elastic supply to the growing areas, as in the Arthur Lewis models, then growth will pull more of the reserve army of labor into gainful employment. However, more contemporary theories suggest that trade liberalization could reduce the wages of unskilled labor even in a labor abundant country, thereby widening the gap between the rich and the poor. 26 To test the impact of trade on employment, which is one of the crucial indicators for achieving the goal of eradicating poverty and hunger, we conduct the co-integration and causality tests between trade variables with employment and wages.

26 See Banerjee and Newman, 2004.

50

Table 1.13: Causality Test between Export/GDP and Total Wages to Total Industrial Exports:

We find that export/GDP ratio is found to Granger cause the employment of industrial workers as a proportion of total labour force implying that exports have led to a rise in industrial employment (Table 1.13). However, with respect to wages, though different variables were used like wage rate, etc exports are not found to have increased wages in this period, i.e., 1980-2005.

For India, though the growth of real wages has been slow overall, growth in the decade of 90s has been slower than the decade of 80s in the aggregate organized manufacturing sector, with the only exception being the import competing industries, which experienced a higher aggregate growth in the 90s than in the 80s for real wages (Goldar 2003)

Goldar (2003) has found that the deceleration in the growth of real wages in Indian manufacturing in the 1990s can mostly be attributed to

(1) Substantial reduction of economic rents in the process of moving over from a highly restrictive trade (and industrial control) regime to a much more liberal regime, (2) Weakening of trade union strength in manufacturing industries in the post-reform period. The adverse effects of trade reforms on growth of real wages through the elimination of rents could have been countered by an upward pressure on wages emanating from a tightening of industrial labour market, if trade reforms had led to a large increase in the demand for labour. This as mentioned above has not happened so far. There was a marked acceleration in employment growth in organised manufacturing in the 1990s (Goldar 2000, 2002; Tendulkar 2000), but this was not sufficient to cause a tightening of the labour market, because there was no such acceleration in employment growth in unorganised manufacturing which accounts for bulk of the industrial employment and thus influences wage determination. (Goldar 2003).

Dependent Variable

Explanatory variables

Significance of long run

relation

ARDL model

ECM coefficie

nt

Granger Causality

Number of industrial

workers/Total labour force

Exports/GDP ratio

1% (1,0,4) -0.07218 YES

Wages in local currency units

Total industrial exports 1% (1,3) -0.46462 NO

51

GOAL 3: Promote gender equality and empower women

(2008) have identified “gender sensitive products” which are those products where the corresponding industries have women employment more than 40% of total employment. These products have been identified using a Concordance matrix between six-digit DGCI&S trade data and three-digit level industry data from Annual Survey of Industries. The Gender Sensitive Products belong to the following industries:

1. Tobacco 2. Wearing Apparel 3. Manufacture of food products 4. Agriculture and animal husbandry

UNCTAD-India (2008) have estimated gender employment multipliers for 15 disaggregated services sectors and have identified IT services to have the highest employment multipliers. Using these two studies for India and given the availability of data we estimate the relationship between exports of textiles and export growth with total women employment in the economy in the period 1980-2005. The results presented in Table 1.14 do not show Granger causality between the two indicators. However, we do find that total women employment is co-integrated with both total export growth and rate of growth of textile exports. There is a stable long run relation between the two. Table 1.14: Total Women Employment; Technical Education and Exports: Causality Test

Dependent Variable

Explanatory variables

Significance of long run relation

ARDL model

ECM coefficient

Granger Causality

Total women employment

Rate of growth of textiles exports

1% (1,1,0) -0.0658 NO

Total women employment

Rate of growth of total exports

NO

Rate of growth of number of

females per 100 males in

University Education in

Technical and Engineering Disciplines

Rate of Growth of Computer,

communications and other services as % of

commercial service exports

5% (1,0) -1.3803 YES

Interestingly, as was argued earlier, trade may have an indirect role in achievement of other MDGs, we find that exports of IT services exports as a proportion of total export of services is found to have Granger caused growth in the number of females per 100 males in University Education and Technical and Engineering Disciplines. This indicates that

52

higher employment opportunities in export-oriented high-skill industries have encouraged women to acquire the requisite skills and be a significant part of the workforce. UNCTAD-India (2008) also finds that in Indian industries it is high tech industries which employ more women as a proportion of men. GOAL 4: Reduce Child Mortality and improve Maternal Health Child mortality and infant mortality are very different from adult mortality. Child mortality is directly related to non-availability of adequate nutrition, health care facilities and clean living environment. A young child can succumb to an early death due to the fact that he did not get the necessary vaccinations and nutritional requirements since birth. This is a particularly large problem in case of developing countries, where typically clean drinking water and adequate sanitation facilities are not available. Trade can help reduce child mortality and infant mortality indirectly i.e., to the extent that with imports any shortage in pharmaceutical and medicinal provisions in a country can be overcome. Granger causality tests have been conducted for infant mortality rate and immunization with imports of pharmaceutical products. We do find a long term stable relationship between these variables but the Granger causality test does not indicate causality. Table 1.15: Relationship between Imports of Pharmaceutical products and infant mortality rate and immunization.

GOAL 7: Ensure environmental sustainability

To identify the role played by trade in India;s progress towards this MDG, we trace the trends in some of the indicators which reflect environment sustainability and are related to tarde.

Dependent Variable

Explanatory variables

Significance of long run

relation ARDL model

ECM coefficient

Granger Causality

Infant Mortality Rate

Imports of medicinal

and pharmaceutical products

5% (1,0)

-0.03527

NO

Immunization,

DPT (% of children ages

12-23 months)

Imports of medicinal

and pharma products

1% (1,0) -0.14073 NO

53

Trade in Environmental Friendly Goods Trade may increase the awareness of environment sustainability. Contribution of trade in achieving this MDG is quantified by the trend in the merchandise trade of environment friendly products by India. We have looked at the trade data for three clean technology products. These products are typically more expensive than their more commercial counterparts but their use is on the rise owing to growing environment concerns. These are: Solar Collectors and parts of Alluminium, Equipment Or Gadgets Based On Solar Energy and Solar Lanterns/Lamp etc. Table 17 presents the trend in trade of these products. Table 1.16: Trade in Environmental Friendly Products

Solar Collectors and parts of Aluminum

Equipment Or Gadgets Based On Solar Energy

Solar Lanterns/Lamp

Exports (qty Kgs (1000s)

Imports (Qty) kgs (1000s)

Exports (qty) no.s (1000s)

Imports (Qty) no.s (1000s)

Exports (qty) no.s (1000s)

Imports (Qty) no.s (1000s)

2003 72 0.01 0.02 0.05 40.44 1.42

2004 199.9 0.71 0.01 121.95 21.23

2005 285.97 2.16 2.64 96.24 13.26

2006 114.85 6.48 0.03 25.69 20.39

Source: Ministry of commerce, Government of India

Carbon trading India Inc earned Rs 1,500 crore last year just by selling carbon credits to developed-country clients. This is a fraction of the RS 18,000 crore that experts estimate will be India’s share in global carbon trading by 2012. In the pipeline are projects that would create up to 306 million tradable Certified Emission Reductions (CERs). Analysts claim if more companies absorb clean technologies, total CERs with India could touch 500 million. Trading in carbon credits encourages countries to use clean technologies, which is good for the environment and helps earn a substantial income on a regular basis for the companies. One carbon credit is equivalent to one tonne of CO2 emission reduced. Carbon credits are available for companies engaged in developing renewable energy projects that offset the use of fossil fuels. The Multi-Commodity Exchange of India (MCX), the country’s leading commodity exchange, may soon become the third exchange in the world with a license to trade in carbon credits. The next generation global commodities exchanges have an important role to play in price discovery of all renewable and non-renewable natural resources for most optimal resource allocation. With the increasing environment related concerns, this role assumes significance for environmental related resources.

54

Figure 1.8: India’s Position in the global market as a seller of carbon credits

Source: Ambrosi and Capoor, State and Trends of the Carbon Market 2007, World Bank

India has a relatively low market share at 12%. However, India is second (at 17%) only to China in the CDM pipeline by the number of expected CERs by 2012 and first by volumes

of issued CERs to date at about 18 million. This is partially as a consequence of the relatively small size of projects (70% of projects with deliveries below 50 MtCO2e per annum). A concerted effort to increase the participation of banks and appropriate intermediaries or bundling agents to increase the average project size could help attracting private carbon buyers to India. Others have pointed out difficult negotiations, with high price expectations from the seller-side that may have driven buyers to other countries. There is evidence, however, of this issue is becoming less of a constraint (Ambrosi and Capoor, 2007)

GOAL 8: Develop a global partnership for development Indian economy is increasingly becoming integrated with the world economy. Over The recent past, government has taken distinct policy measures with a view to promote exports. This is evident in the decreasing tariff lines and increasing expenditure on export promotion.

55

Table 1.17: Proportion of India’s Exports admitted duty free by Developed Countries.

Source:www.mdg-trade.org

The following graph maps the trend in total disbursement (nominal) from the Indian Governments budget for foreign trade and export promotion. Figure 1.9: Total disbursement for foreign trade and export promotion

Total disbursement for Foreign trade and export promotion

0200400600800

1000120014001600

1996

-97

1997

-98

1998

-99

1999

-00

2000

-01

2001

-02

2002

-03

2003

-04

2004

-05

2005

-06

2006

-07

Cro

res

of

Ru

pee

s

Source: GOI Annual Financial Statements

These measures have led to an increase in volume of trade as a proportion of GDP overtime in India. This has more than doubled in 2005-06 from 1990-91. The growth has been faster if trade in services is included.

Year Proportion of India’s Exports admitted duty free by Developed Countries

1996 45.12%

1997 47.29%

1998 48.21%

1999 52.72%

2000 52.77%

2001 50.85%

2002 59.53%

2003 61.55%

2004 63.10%

2005 62.85%

2006 65.90%

56

Table 1.19: Trade as a proportion of GDP

Source: RBI

We can also juxtapose India against other countries to make an international comparison. Today, India’s restrictiveness of imports index stands at 21.7% which is higher than other selected countries (Figure 1.10) and restrictiveness of exports index at 8.9%, which is lower than some of the other countries. Figure1.10: Trade Restrictiveness Indices

Restrictiveness Index

0.00

5.00

10.00

15.00

20.00

25.00

Brazil

China

India

Rus

sia

Turke

y

Singa

pore

Thaila

nd

Uni

ted

State

s

Europ

ean

Union

in percent

Import restrictiveness Index Export Restrictiveness index

Source: Global Monitoring Report

The number of trade agreements signed or negotiated is useful to assess the progress of India is achieving this MDG. By March 2007, there were 197 trade agreements as compared with only 24 agreements in 1986 and 66 agreements in 1996. The East and the South East Asian economies in particular have achieved significant integration due to liberalization of foreign trade and foreign direct investment regimes and several bilateral regional cooperation agreements.

Overall the results of this analysis to estimate the role of trade in India’s progress in achieving MDGs show that:

♦ There has been a steady rise in Trade related Development Index (TRDI) and Trade-Gender Development Index overtime

♦ Trade has Granger caused TRDI

Trade as a percentage of

GDP

Trade in Goods and Services as a percentage

of GDP 1990-91 14.6 17.2

1995-96 21.5 25.7

2000-01 22.5 29.2

2005-06 32.5 44.8

57

♦ Exports have Granger caused Trade Related Gender Development Index (TGD)

♦ Exports have Granger caused rise in Industrial Employment

♦ Exports of Computer and Communication services has Granger caused Female Technical Education

♦ Long run stable relationship was found between imports of pharmaceutical products and decline in infant mortality and increase in immunization

♦ Trade in environmental friendly goods has increased

♦ India’s share in global carbon trading has increased and India is improving its position as carbon credit seller.

♦ There is an increasing trend in percentage of India’s exports entering duty free in developed country

♦ Government disbursement for foreign trade and export promotion has increased overtime

♦ But India is found to have high import restrictions.

♦ There has been a rise in India’s engagement with globally as the number of free trade agreements signed and negotiated is increasing considerably.

1.7. Impact of Exports on Employment and Incomes of Poor in India

The causality analysis reflects India’s progress towards MDGs and establishes that trade has contributed to this progress. One of the identified channels through which trade may have contributed to this progress is industrial employment. However, the net impact of trade may have been positive but it may not be desirable to compare the gains to losses; as losses may have occurred to relatively poorer section of the society and gains to relatively well-off sections or vice versa. In an attempt to trace the impact of exports on poor the extent to which exports have created employment and consequently incomes for the low income group is estimated.

1.7.1 Methodology for Estimating Employment and Incomes Generated by Exports

Methodology for Estimating Impact of Exports on Employment

Using the latest available input-output matrix for India for the years 2003-04, employment multipliers have been generated for 46 sub-sectors of the Indian economy. The increase in exports from 2003-04 to 2006-07 across these sub-sectors has been recorded. Exports in each sector in 2006-07 have been deflated so as to remove any price effects. After arriving at the actual increase in exports in different sub-sectors the employment multipliers have been applied to exports to arrive at consequent changes in the output in each sector. These changes for any particular sector will include both direct increase in output of the sector caused by exports as well as indirect increase in output which is generated because of the rise in demand due to exports of any other good in which the sector’s output is used as an input. For example, rise in exports of food products will generate a demand for food crops. In other words, the rise in output due to

58

increased exports of any sector will be due to increase in its own demand as well as increase in demand for goods which may use this sector’s output as intermediary good. The employment multipliers are then applied to the rise in output in each sector to give the extent to which additional employment will be generated due to rise in exports. As in the case of output increase, employment increase in a sector will include both direct as well indirect increases in employment generated by exports.

Methodology for estimating Impact of Increase in Exports on Incomes

To estimate the impact of increase in exports on incomes of low income groups, the increase in value added has been estimated using the corresponding increase in employment. The share of labour and capital in value added has been taken from Input-Output Matrix. It is assumed, based on economic theory, that the increase in value added by labour will be same as increase in share of labour in total income generated. The distribution of the increase in income across different income groups is arrived at by using National Sample Survey (NSS) data on incomes across rural and urban households in different income strata. Income generated by exports has been estimated for five income groups across rural and urban areas which include the lowest income group, i.e., people in abject poverty.

To arrive at these categories, the population is divided into 5 rural and 5 urban expenditure classes. For 1999-2000, detailed item wise expenditures by expenditure classes are given in NSS Report No. 461. NSS also gives the population in each expenditure class. Using these expenditure data, the relative expenditure of each expenditure class is obtained for most of the sectors. For a few sectors the distribution of related or broad groups is used. The total sector wise private final consumption expenditure (PFCE) is divided into expenditure classes by applying the relative sector wise expenditure of each class.

The NSS gives the consumption expenditure by twelve classes for rural and urban areas. We club these classes in five each. The first two classes refer to population below abject poverty and poverty line and the next class is taken up to the expenditure class covering the average per capita expenditure for rural as well as the urban areas. The upper four classes have been clubbed into two classes as mentioned below. Expenditure classes into which PFCE is divided

Rural Expenditure class (Rs. per month)

Urban Expenditure class (Rs. per month)

RH1 000-255 UH1 000-350

RH2 255-340 UH2 350-500

RH3 340-525 UH3 500-915

RH4 525-775 UH4 915-1500

RH5 775- above UH5 1500- above

59

1.7.2 Output Generated by Increase in Exports: 2003-04 to 2006-07

To estimate the output generated by increase in exports across different sub sectors we use input-output matrix for the year 2003-04. Rise in real exports (after deflating by their price indices) and rise in output which includes both direct and indirect rise caused by exports in 46 sectors is presented in Table 1.20.

Table 1.20 Exports and Increase in Output across Sectors: 2003-04 to 2006-07

S.No Exports 2003-04 Exports 2006-07

Increased output from 2003-04 to 2006-07

Rs billion Rs billion Rs billion

1 Food crops 61.2 61.2 42

2 Cash Crops 17.5 32.9 83

3 Plantation Crop 1.8 2.5 21

4 Other crops 35.4 81.6 165

5 Animal Husbandry 17.8 20.9 50

6 Forestry and logging 11 17.2 22

7 Fishing 39.7 42 7

8 Coal and lignite 1.6 1.6 96

9 Crude petroleum, natural gas 1.9 4.4 375

10 Iron ore 27.5 50.5 35

11 Other Minerals 231.5 231.5 42

12 Food Products 164.3 302.2 178

13 Beverages, tobacco, etc. 3.4 4.4 6

14 Cotton textiles 78 116.7 91

15 Wool Silk and Synthetic fiber 58.7 80 55

16 Jute, hemp, mesta textiles 3.9 6 12

17 Textiles products including wearing apparel 358.6 501.4 163

18 Wood, furniture, ect. 4.3 8.1 31

19 Paper & printing, etc. 13.9 23.8 68

20 Leather and leather products 58.8 74.4 26

21 Rubber, petroleum, plastic, cola. 204.1 584.2 627

22 Chemicals, etc. 300 505 598

23 Non-metallic products 31.9 36.7 22

24 Metals 182.7 364.2 753

25 Metal products except mach. And tpt. Equipment 53.4 91.8 128

26

Tractors, agri. Implements, industrial machinery, other machinery 129.5 225.6 158

27 Electrical, electronic machinery and applications 96.3 178.5 137

28 Transport Equipments 76.3 177.3 137

60

29 Miscellaneous manufacturing industries 389.8 760.7 483

30 Construction 0 0 55

31 Electricity 0 0 277

32 Gas and water supply 0 0 8

33 Railway transport services 42.9 77 100

34 Other transport services 249.4 457.3 429

35 Storage and warehousing 0 0 3

36 Communication 0.9 2.2 51

37 Trade 298.4 492.3 561

38 Hotels and restaurants 90.4 213 139

39 Banking 13.7 146.9 388

40 Insurance 19.3 63.3 81

41 Ownership of dwellings 0 0 0

42 Education and research 0 0 0

43 Medical and health 0 0 1

44 Other services 510 1453.8 1186

45 Public administration 0 0 0

46 Tourism 230.5 357.6 127

Total 4110.3 7850.7 8017

Source: Input Output Tables (2003-04)

It is interesting to note that exports have increased substantially for food products (almost double) and crude petroleum and natural gas (more than double). Exports generate both direct as well as indirect demand. Change in output therefore reflects change in demand for the product for final consumption as well as for intermediate consumption. Across sectors we find that the output generated due to rise in exports has been highest for industry. Industry’s share in total output generated is 53%, followed by services (42%) and then agriculture (5%). Within industry, metals (17%); rubber, petroleum and chemicals (14%) had high shares. Within services, maximum output has been generated for other services (34%); followed by trade (16%) and other transport services (12%).

Using the change in output due to exports in each sector, we derive output and employment multipliers. We define the output multiplier of a sector as the amount by which the total output increases for a unit increase in the output of that sector. It is usual to measure the unit change in INR lakhs (00,000) or 0.1 million. Thus, if the output multiplier for a sector is 4, it implies that for every INR 1 lakh (100,000) increase in the sectoral output, total output (of the entire economy) increases by INR 4 lakhs (400,000). That is an increase in total output by 0.4 million for every increase in sectoral output by 0.1 million rupees. Similarly, the employment multiplier of a sector gives an estimate of the aggregate direct and indirect employment changes, in person years, resulting from the increase in INR 100,000 of output of that sector. Table 1.21 reports the output and employment multipliers based on 2003-04 across sectors.

61

Table 1.21: Output and employment multipliers based on Input-Output Matrix of 2003-04

S No Sectors Output Multipliers

Employment multipliers

1 Food crops 1.64 8.56

2 Cash crops 1.38 2.15

3 Plantation crop 1.34 2.34

4 Other crops 1.31 0.40

5 Animal husbandry 1.43 0.68

6 Forestry and logging 1.19 0.97

7 Fishing 1.27 0.56

8 Coal and lignite 1.55 0.35

9 Crude petroleum, natural gas 1.25 0.09

10 Iron ore 1.55 0.26

11 Other minerals 1.35 2.04

12 Food products 2.36 1.72

13 Beverages, tobacco, etc. 2.07 0.62

14 Cotton textiles 2.30 1.40

15 Wool Silk and synthetic fiber 2.52 1.06

16 Jute, hemp, mesta textiles 2.11 1.28

17 Textiles products including wearing apparel 2.37 1.22

18 Wood, furniture etc. 1.82 3.14

19 Paper & printing, etc. 2.35 0.58

20 Leather and leather products 2.40 1.11

21 Rubber, petroleum, plastic, cola. 2.22 0.26

22 Chemicals, etc. 2.31 0.36

23 Non-metallic products 2.14 1.01

24 Metals 2.67 0.53

25 Metal products 2.61 0.64

26 Tractors, agriculture Implements, industrial machinery, other machinery 2.53 0.60

27 Electrical, electronic machinery and applications 2.55 0.38

28 Transport equipments 2.51 0.40

29 Miscellaneous manufacturing industries 2.63 0.65

30 Construction 2.08 0.97

31 Electricity 2.28 0.36

32 Gas and water supply 1.59 0.41

33 Railway transport services 1.94 0.49

34 Other transport services 2.07 0.67

35 Storage and warehousing 1.81 0.67

36 Communication 1.48 0.64

37 Trade 1.37 0.83

38 Hotels and restaurants 2.12 1.79

39 Banking 1.37 0.22

40 Insurance 1.53 0.28

41 Ownership of dwellings 1.15 0.36

42 Education and research 1.23 0.80

43 Medical and health 2.36 0.69

62

44 Other services 1.97 0.69

45 Public administration 1.00 0.64

46 Tourism 2.13 0.83

Given high employment in agriculture sector, we find that employment multipliers are highest for food crops (8.56), followed by wood & furniture (3.14) and plantation crops (2.34). Within manufacturing sector, high employment multipliers are found to be for cotton textiles, jute, hemp & mesta textiles and other textile products. Within services sector, construction, domestic trade and tourism are found to have high employment multipliers.

1.7.3 Employment generated by Increase in Exports: 2003-04 to 2006-07

Applying the employment multipliers to actual increase in output due to exports across sectors, we generate actual rise in employment due to change in exports from 2003-04 to 2006-07. The results are presented in Table 1.22.

Table 1.22: Increase in Employment due to Increase in Exports from 2003-04 to 2006-07

Sectors

Increased Employment from 2003-04 to 2006-07 (in person years)

Increased employment at 10 % increase from 2006-07 (in person years)

Million No's Million No's

Food crops 3.23 1.16

Cash Crops 1.65 0.4

Plantation Crop 0.47 0.09

Other crops 0.27 0.05

Animal Husbandry 0.19 0.05

Forestry and logging 0.2 0.04

Fishing 0.03 0.02

Coal and lignite 0.17 0.04

Crude petroleum, natural gas 0.08 0.01

Iron ore 0.04 0.01

Other Minerals 0.83 0.62

Food Products 0.45 0.1

Beverages, tobacco, etc. 0 0

Cotton textiles 0.55 0.16

Wool Silk and Synthetic fiber 0.26 0.08

Jute, hemp, mesta textiles 0.06 0.02

Textiles products including wearing apparel 0.9 0.3

Wood, furniture, ect. 0.82 0.17

Paper & printing, etc. 0.11 0.02

63

Leather and leather products 0.14 0.06

Rubber, petroleum, plastic, cola. 0.18 0.03

Chemicals, etc. 0.22 0.05

Non-metallic products 0.1 0.03

Metals 0.59 0.12

Metal products except mach. And tpt. Equipment 0.29 0.06

Tractors, agri. Implements, industrial machinery, other machinery 0.34 0.08

Electrical, electronic machinery and applications 0.02 0

Transport Equipments 0.06 0.01

Miscellaneous manufacturing industries 1.18 0.23

Construction 0.27 0.06

Electricity 0.21 0.04

Gas and water supply 0.02 0

Railway transport services 0.28 0.06

Other transport services 1.68 0.37

Storage and warehousing 0.01 0

Communication 0.25 0.05

Trade 4.06 0.92

Hotels and restaurants 0.93 0.17

Banking 0.37 0.06

Insurance 0.09 0.01

Ownership of dwellings 0 0

Education and research 0 0

Medical and health 0 0

Other services 3.95 0.65

Public administration 0 0

Tourism 0.43 0.12

Total 25.97 6.56

The total increase in employment generated by rise in exports in the period 2003-04 to 2006-07 has been around 26 million person years. Applying the same employment multipliers across sectors, it is estimated that a further rise in 10% of exports in each sector will be able to generate employment for an additional 6 million.

It has been argued that that services sector growth in the 1990s was a “jobless growth”. However, using this methodology for the period 2003-04 to 2006-07, we find that the maximum employment generated by exports in this period is in services sector, which is 12 million; followed by industrial sector (7 million) and then agriculture (6 million). Maximum increase in employment has been generated in domestic trade, followed by other services, food crops and other transport services. It needs to be noted that there has been no change in exports of food crops, as seen in table 20, but given the inter-sectoral linkages (particularly with food products) and high employment multiplier in food crops, increase in exports generates high level of employment in this sector. If exports of each

64

sector rise by 10% from the level of 2006-07, additional employment of 1.81 million will be generated in agriculture, which has the maximum number of poor.

1.7.4 Impact of Rise in Exports in 2003-04 to 2007-08 on Incomes of the Poor:

To estimate the impact of increased exports on incomes of different income categories, we use the shares of capital and labour in total value added and expenditures of different income groups to arrive at shares in total incomes of different income groups. Table 1.23 reports the results. Table 1.23: Increased value added and household wise increased income due to increase in exports.

Increased Income

from 2003-04 to 2006-07

Increased Income for 10 % increase in Exports from 2006-07

Rs billion Rs billion

Labour 1366.22 284.34

Capital 1559.19 325.81

RH1 24.00 5.00

RH2 92.09 19.19

RH3 289.08 60.26

RH4 360.34 75.20

RH5 504.27 105.22

UH1 15.41 3.21

UH2 64.58 13.45

UH3 262.24 54.62

UH4 311.78 64.95

UH5 440.51 91.80

Total increase 2364.3 1103.05

RH1 and RH2 -people in abject poverty and below poverty line in rural areas UH1 and UH2 -people in abject poverty and below poverty line in urban areas The results show that the total income generated by increase in exports in the period 2003-04 to 2006-07 has been of Rs 2,364 billion, equivalent to USD 55 billion. However, within rural areas, we find that the distribution of income has not been in favour of people in abject poverty, i.e., in income groups RH1 and RH2. Out of the total income generated in the rural areas due to exports, only 2% reaches the RH1 income group (poorest of the poor); while 7% of the total income generated is for the income group RH2. Together, the low income groups in rural areas get less than 10% of the total

income generated in the rural sectors because of exports. In case of urban sectors, we

find that the situation is not much different. In fact, income generated by exports for the

65

UH1 income group is only 1.4% of the total income group while UH1 and UH2 together have around 7% of the total income generated as in the case of rural areas.

Total income generated for the people in the lowest income group (RH1 and UH1) is around 1.6% of the total income generated in the economy by exports in the period 2003-04 to 2006-07. The highest income group (RH5 and UH5) gets around 40% of the total income generated by exports, while 70% of the total income generate goes to the top two income groups (rural and urban taken together).

The results therefore indicate that though exports from India have been able to increase employment across sectors, mainly because of direct and indirect linkages in the economy and the incomes of the poor have increased. But proportion of the gains that has reached the poor from exports in terms of rise in incomes has been very low. The incomes of the poor and poorest of the poor have increased but it is a small proportion of total increase in incomes generated by exports.

1.8. Summary and Conclusion

To assess the role played by trade in India’s progress towards United Nations Millennium Development Goals (MDGs) the chapter undertakes analysis at three levels; firstly, the study traces and contrasts the trends in trade and along with the selected human development indicators that may be directly or indirectly influenced by trade; secondly, a Trade Related Development Index (TRDI) for India is constructed based on the trade-related factors that may influence India’s progress towards achieving MDGs A Trade-Gender Development Index (TRGDI) has also been constructed based on selected trade-related factors that may influence MDG goal for “promoting gender equality and empower women”. Causality tests are undertaken for TRDI and TRGDI with trade-related variables for the period 1990-2005. To corroborate this analysis, causality tests have also been conducted between trade-related variables and selected indicators of human development to identify the channels through which trade may have influenced human development; thirdly, an attempt is made to estimate the extent to which exports have affected employment and incomes of poor. Employment multipliers have been generated using input-output matrix for India and the extent to which exports have generated employment and incomes of people in different income strata is estimated

The Overall Results derived from the analysis to assess the role of trade in India’s progress in achieving MDGs are as follows:

♦ There has been a steady rise in India’s Trade Related Development Index (TRDI) and Trade Related Gender Development Index overtime, which indicates that the identified human development indicators which are may be affected by trade have shown a steady rise in the trade liberalized regime.

♦ Trade has Granger caused TRDI, which indicates that the progress in trade related development indicators, can be explained by the rise in trade.

♦ Exports have Granger caused Trade Related Gender Development Index (TRGDI) implying that women have gained from exports have to some extent.

66

♦ Exports have Granger caused rise in Industrial Employment

♦ Exports of Computer and Communication services has Granger caused Female Technical Education which indicates that by creating new job opportunities for women, exports of IT services has indirectly improved the skills of women.

♦ Long run stable relationship was found between imports of pharmaceutical products and decline in infant mortality and increase in immunization

♦ Trade in environmental friendly goods has increased

♦ India’s share in global carbon trading has increased and India is improving its position as carbon credit seller.

♦ There is an increasing trend in percentage of India’s exports entering duty free in developed country

♦ Government disbursement for foreign trade and export promotion has increased overtime

♦ But India is found to have high import restrictions

♦ There has been a rise in India’s engagement with globally as the number of free trade agreements signed and negotiated is increasing considerably.

In terms of exports generating employment and incomes the results show that:

♦ Across sectors the output generated due to rise in exports has been highest for industry. Industry’s share in total output generated is 53%, followed by services (42%) and then agriculture (5%).

♦ Given high employment in agriculture sector, we find that employment multipliers are highest for food crops; within manufacturing sector, high employment multipliers are found for cotton textiles, jute, hemp & mesta textiles and other textile products. Within services sector, construction, domestic trade and tourism are found to have high employment multipliers.

♦ The total increase in employment generated by rise in exports in the period 2003-04 to 2006-07 has been around 26 million person years.

♦ Maximum employment generated by exports in 2003-04 TO 2006-07 is in services sector, which is 12 million; followed by industrial sector (7 million) and then agriculture (6 million).

♦ The results show that the total income generated by increase in exports in the period 2003-04 to 2006-07 has been of Rs 2,364 billion, equivalent to USD 55bn.

♦ However, the gains from exports in terms of total income generated have not been lopsided.

♦ The low income groups in rural areas get less than 10% of the total income generated

in the rural sectors because of exports.

♦ In case of urban sectors, we find that the situation is not much different. In fact, income generated by exports for the people in abject poverty is only 1.4% of the total income generated while people below poverty line get around only 7% of the total income generated as in the case of rural areas.

♦ Total income generated for the people in the lowest income group i.e., taking rural

and urban households together (RH1 and UH1) is around 1.6% of the total income

generated in the economy by exports in the period 2003-04 to 2006-07.

67

♦ The highest income group (RH5 and UH5) got around 40% of the total income generated by exports, while 70% of the total income generated went to the top two income groups (rural and urban taken together).

The chapter benchmarks the role played by trade in India’s progress towards MDGs and identifies the channels through which some of the effects have been transmitted in the economy. The broad conclusion that emerges from the results is that trade has played a significant role by favorably influencing some of the human development indicators. Trade has generated employment and incomes for the poor but the share of poor in total incomes generated is marginal. Trade policies are rarely formulated with the objective of reducing poverty. However, in the view of the importance of the role that trade can play in terms of a country’s progress towards human development, it may be imperative for the trade policy makers to use trade policy as an instrument for generating employment and incomes for the poor. Given the fact that there will always be gainers and losers in the process of liberalization, it is not the net positive impact of trade on poverty that should be the goal of trade policy as it may not be desirable to compare the gains to losses; as losses may occur to relatively poorer section of the society and gains to relatively well-off sections or vice versa. What needs to be focused more is e how the poor and dis-advantaged sections of the economy are affected by trade policies.

***

68

References: Acharyya, Rajat 2006. Trade Liberalization, Poverty, and Income Inequality in India.

Asian Development Bank. Ambrosi, Philippe and Karan Capoor. 2007. State and Trends of the Carbon Market.

World Bank Aurora A, C. Teixeira and Natércia Fortuna. Human capital, trade and long-run

productivity. Testing the technological absorption hypothesis for the Portuguese economy, 1960-2001.

Banerjee and Newman, 2004. Inequality, Growth and Trade Policy. Massachusetts

Institute of Technology and University College London. Unpublished Manuscript. Banga, Rashmi 2005. Trade-Growth-Poverty Nexus in Asia Pacific: Empirical Evidence. Barro, Robert and Jhong-Wha Lee (1994). Data set for a panel of 138 countries. National

Bureau of Economic Research, Cambridge, Massachusetts. Basu SR, LR Klein and AL Nagar (2005). Quality of Life: Comparing India and China.

Paper presented at the Project LINK meeting, 1 November, United Nations O� ce, Geneva.

Bhagwati,Jagdish and T.N.Srinivasan, 2002. Trade and Poverty in the Poor Countries.

American Economic Review 92/2, (May 2002), 180-83. Cerra, Valerie and Sweta Saxena. 2000. What Caused the 1991 Currency Crisis in

India?” International Monetary Fund Working Paper No. 00/157 Chang, Katlani and Loayza, 2005. Openness can be good for growth: The role of policy

complementarities. Processed, World Bank, Washington DC. Choong, C.K., Zulkornain Yusop and Venus Liew Khim Sen 2005. Export led Growth

Hypothesis in Malaysia: An investigation using bounds test” Sunway Academic Journal Climate Analysis Tools Indicators, World Resources institute

David Dollar and Aart Kraay, 2001. Trade, Growth and Poverty. World Bank Policy

Research Working Paper No. 2615 Dejong and Ripoll, 2006. Tariffs and Growth: An empirical exploration of contingwent relationships. Review of Economics and Statistics, forthcoming. Easterly W, 2005. National policies and economic growth.In Philippe Aghion and Steven

Durlauf, editors, Handbook of Economic Growth, Elsevier, 2005.

69

Economic Surveys, various issues. Energy Information Administration, U.S. Department for energy Feenstra, Robert and Hanson, Gordon 1996. Foreign Investment, Outsourcing and

Relative Wages. In Robert Feenstra, Gene Grossman and Douglas Irwin (eds), The Political Economy of Trade Policy: Papers in Honor of Jagdish Bhagwati, MIT Press: Cambridge, Mass., pp.89-127.

Feenstra, Robert C. and Gordon H. Hanson 1999. Productivity Measurement and the

Impact of Trade and Technology on Wages: Estimates for the U.S., 1972-1990. Quarterly Journal of Economics, August , 114(3), 907-940.

Frankel and Romer, 1999. Does Trade Cause Growth? American Economic Review 89

(3): 379-399. Frankel, Jeffrey and Romer, David (1999). Does Trade Cause Growth? American

Economic Review, 89 (3), 379-399. Ghosh,Jayati “Globalisation, Export-oriented Employment for women and social policy:

A Case Study of India” UNRISD. Goldar, B.N 2002. Trade Liberalization and Manufacturing Employment: The Case of

India. Employment Paper no. 2002/34, International Labour Office, Geneva. Goldar, B.N, 2004. Productivity trends in Indian manufacturing in the pre- and post-

reform periods. Indian Council for Research on International Economic Relations, New Delhi Working Papers 137, Indian Council for Research on International Economic Relations, New Delhi, India.

Goldar, B.N. 2000. Employment Growth in Organised Manufacturing in India. Economic

and Political Weekly , Vol. 35, No. 14, April 1-7. Goldar, B.N. 2003. Trade Liberalisation and Manufacturing Employment: The Case of

India. Employment Paper 2002/34, Geneva, International Labour Office. Goldar, Bishwanath 2003. Trade Liberalisation and Real Wages in Organised

Manufacturing Industries in India. IEG Gulati, A. and Narayanan, S., 2002. Rice Trade Liberalisation and Poverty. Markets and

Structural Studies Division Discussion Paper No. 51, Washington DC. International Food Policy Research Institute.

Handbook of Statistics on Indian Economy (2007) RBI.

70

Harrison, Ann, 1996. Openness and growth: A time-series, cross-country analysis for developing countries. Journal of Development Economics, Elsevier, vol. 48(2), pages 419-447, March.

Hasan, R., D. Mitra and K. V. Ramaswamy, 2003. Trade Reforms, Labor Regulations

and Labor-Demand Elasticities: Evidence from India. NBER Working Paper 9879, July

Human Development Reports, various issues. Jeffrey D. Sachs and Andrew M. Warner, (1997). Fundamental Source of Long-Run

Growth. American Economic Review, Papers and Proceedings, May 1997. Jena, Pradyot Ranjan and Ulrike Grote, 2008. Growth–trade–environment nexus in India.

Economics Bulletin.

Jha, V. et al (2005), Trade Liberalisation and Poverty in India, Macmillan India Limited. Joshi, Vijay Oxford University, www.rediff.com Kakawani, 1980. On a Class of Poverty Measures. Econometrica. Krueger, Anne 1983. Trade and Employment in Developing Countries, 3 Synthesis and

Conclusions, Chicago, University of Chicago Press. Kuznets, Simon, 1955. Toward a Theory of Economic Growth. In R. Lekachman (ed.),

National Policy for Economic Welfare at Homa and Abroad. Garden City, NY: Doubleday, pp. 12-85.

Lucas, 1988. On the Mechanics of Economic Development. Journal of Monetary

Economics, Vol. 22, pp. 3-42, (1988). Nagar, A.L. and S. R. Basu. (2002). “Weighting Socio-Economic Indicators of Human

Development: A Latent Variable Approach,” in Ullah, A. et al (eds.) Handbook of Applied Econometrics and Statistical Inference, Marcel Dekker, New York, USA.

Nielsen, Chantal Pohl, 2002. Vietnam in the International Rice Market: A Review and

Evaluation of Domestic and Foreign Rice Policies. Report No. 132, Danish Research Institute of Food Economics, Copenhagen.

Nye, Reddy and Watkins, 2002. Dollar and Kraay on trade, growth and poverty: A

critique. Processed, Columbia University, New York. Purushothaman, Roopa, 2008. Is Urban Growth good for rural India. United Nations

Expert Group Meeting on Population Distribution, Urbanization, Internal Migration and Development

71

Raghubendra Jha ,2000. Rising Income Inequality and Poverty Reduction: Are they

Compatible? Paper has been prepared for the WIDER project. Ramesh Chand, 1999. Effects of Trade Liberalization on Agriculture in India:

Commodity Aspects. In The Regional Co-ordination Centre for Research and Development of Coarse Grains, Pulses, Roots and Tuber Crops in the Humid Tropics of Asia and the Pacific (CGPRT Centre).

Ravallion 2001. Growth, Poverty and Inequality: Looking beyond Averages. World

Development. Ravallion and Datt, 2000. When is Growth Pro-Poor? Evidence from the Diverse

Experience. Policy Research Working Paper Series 2263, The World Bank. Ravallion, Martin & Chen, Shaohua, 2003. Measuring pro-poor growth. Economics

Letters, Elsevier, vol. 78(1), pages 93-99, January. Ravallion, Martin, 2001. Inequality convergence. Policy Research Working Paper Series

2645, World Bank. Robert J Barro, 1999. Determinants of Economic Growth: Implications of the Global

Evidence for Chile. Latin American Journal of Economics, Instituto de Economía. Pontificia Universidad Católica de Chile., vol. 36(107), pages 443-478.

Rodriguez and Rodrik, 2001. Trade Policy and Economic Growth: A skeptic’s guide to

the cross-sectional evidence. In Ben S. Bernanke and Kenneth Rogoff (eds.). NBER Macroeconomics Annual 2000. National Bureau of Economic Research, Cambridge, MA.

Rodriguez and Rodrik. 1999. Trade Policy and Economic Growth: A skeptic’s guide to

the cross-sectional evidence. NBER working paper, #7081. National Bureau of Economic Research, Cambridge, MA.

Rodrik, Dani, 2000. Comments on trade, growth and poverty by Dollar and Kraay.

Processed, Harvard University, Cambridge, MA. Romer, Paul M, 1986. Increasing Returns and Long-run Growth. Journal of Political Economy, University of Chicago Press, vol. 94(5), pages 1002-37, October. Sachs, Jeffrey D. and Warner, Andrew (1995), “Economic Reform and the Process of

Global Intergration”, in William C. Brainad and George L. Perry, (eds.), Brookings Papers on Economic Activity, 1, 1-118.

Tendulkar, S. 2000. Employment Growth in Factory Manufacturing Sector during Pre-

and Post-Reform Periods. Paper presented at a Conference to honour Professor

72

K.L. Krishna on the theme ‘Industrialization in a Reforming Economy: A Quantitative Assessment’, Delhi School of Economics, Delhi, December 20-22, 2000.

The Global Employment Challenge (2008) ILO Topalova, Petia 2004. Trade Liberalization and Firm Productivity: The Case of India.

IMF Working Paper, WP/04/28 UNCTAD-India (2008) How are the Poor Affected by International Trade in India: An

empirical Approach Villasenor, J.A. and Arnold, B.C. 1989. Elliptical Lorenz Curves. Journal of

Econometrics, Vol. 40, pp. 327-338. Virmani, A., B. Goldar, C. Veeramani, and V. Bhatt (2004), Impact of Tariff Reforms

on Indian industry: Assessment Based on a Multisector Econometric Model, ICRIER Working Paper no. 135.

Warner, Andrew, 2003. Once more into the breach: Economic reform and global

integration. Processed, Center for Global Development, Washington DC. Winters, L. A., N. McCulloch, and A. McKay (2004) “Trade Liberalization and Poverty:

The Evidence So Far.” Journal of Economic Literature 42 (March): 72–115. Websites and Databases Annual Survey of Industries FAO Stats indiastat.com KILM 2005 (ILO) LABORSTA (ILO) Labour bureau of India Ministry of labour- Various annual reports, www.labour.nic nationmaster.com Population Foundation of India Rbi.org.in saarc-sec.org un.org/esa/population/publications/countryprofile/india.pdf World Integrated Trade Solution (WITS) World Development Indicators

73

APPENDIX I The following is a description of the trade related development index (TRDI) and trade gender development index (TGDI) indicators used in the analysis. SNo Variable name Units Time

period

Trade related development Indicators

1 Population below national poverty line, total, Source: Planning Commission

%ages 1990-2006

2 Employment-to-population ratio, both sexes, Source: ILO KILM

%ages 1990-2006

3 School enrollment, gross, girls to boys, higher education Source: Statistical Abstract of India, Ministry of Education

%ages 1990-2006

5 Women employment to total employment , organized sector Source: Statistical Abstract of India, Ministry of Labour, Manpower Statistics

%ages 1990-2006

6 Forest area in total geographical area Source: Statistical Abstract of India, Manpower Statistics, IndiaStat

%ages 1990-2006

CO2 emissions per capita) Source: World Development Indicators 2007

metric tons 1990-2006

Trade gender development indicators

1 School enrollment, gross, girls to boys , primary education Source: Statistical Abstract of India, Ministry of Education

%ages 1990-2006

2 School enrollment, gross, girls to boys , middle education Source: Statistical Abstract of India, Ministry of Education

%ages 1990-2006

3 School enrollment, gross, girls to boys , higher education Source: Statistical Abstract of India, Ministry of Education

%ages 1990-2006

4 Number of seats held by women of national parliament Source: Statistical Abstract of India, Ministry of Education, UN MDGs database

%ages 1990-2006

5 Women employment to total emplyment , organized sector Source: Statistical Abstract of India, Ministry of Education, UN MDGs database

%ages 1990-2006

74

APPENDIX II Construction of TRDI Index

Construction of TRDI

Given the nature of the analysis, we can draw an analogy with the classical regression model. If the dependent variable (TRDI or TGDI) could be measured, we would simply run a least squares regression of TRDI or TGDI on various trade-related development indicators and determine optimal coefficient estimates. The total variation in TRDI or TGDI is expressed as a sum of (i) variation due to trade-related development indicators and (ii) error variance.

Let us suppose x1, …, xK are K development indicators on whom T observations (xit’s) are available for i=1, …, K and t=1, …, T. We transform the indicators into their standardized form:

ix

iitit

s

xxX

−=

where (arithmetic mean) ∑==

T

titi x

Tx

1

1and

(square of standard deviation) ∑ −==

T

tiitx xx

Ts

i1

22 .)(1

The covariance matrix of the standardized variables, Xit’s, is, in fact, the correlation matrix R of the development indicators.

Let us arrange Kλλλ >>> L21 the eigen values of R in descending order of magnitude,

and the corresponding eigen vectors as

=

=

KK

K

K

K α

α

α

α

α

α MLM

1

1

11

1 ,,

such that

0and1 jiii =αα′=αα′

for i ≠ j = 1, …, K. Then the successive principal components are

KtKKtKKt

KtKtt

XXP

XXP

αα

αα

++=

++=

L

L

L

11

11111

with var Pit = λi for i=1, …, K. The estimator of the development index or trade-related development index is obtained as the weighted average of successive principal components: thus

K

KtKtt

PPTGDITRDI

λλλλ

++

++=

L

L

1

11/ ;

75

maximum weight ( ∑λλ i1 / ) has been assigned to the first principal component as it

describes the largest proportion of total variation in all x’s. The second principal

component has the second highest weight ( ∑λλ i2 / ), and so on.

APPENDIX III ARDL Methodology Essentially understood from a paper by Choong et al. (2005), the methodology is as follows: First a vector autoregression (VAR) of order p, denoted by VAR(p), is formulated for any dependent variable, which is any Human Development Indicator in our case. p

Zt = µ + ∑βi Zt-i + ε t i=1 where Zt is the vector of both Xt and Yt , Yt is the dependent variable and Xt is the vector of explanatory variables. Further, a vector error correction model (VECM) is developed as follows: p-1 p-1

∆ Zt = µ + αt + λ Zt-1 +∑ γ i∆γ i-t + ∑γ i∆ xt-I + εt i=1 i=0

where ∆ is the first-difference operator. Partition the long-run multiplier matrix

λ as follows:

λ xx λ xy

λ=

λ yx λyy The diagonal elements of the matrix are unrestricted, so the selected series can be either

I(0) or I(1). If λ yy =0, then y is I(1). In contrast, if λ yy<0 , then y is I(0). The VECM procedure described above is important in the testing of at most one cointegrating vector between dependent variable y t and a set of regressors x t . To derive a model most suited to our requirements, we followed the assumptions made by Pesaran et al. (2001) in Case III, that is, unrestricted intercepts and no trends. After

imposing the restrictions λ xy = 0 , µ ≠ 0 and α= 0 , the function can be stated as the following unrestricted error correction model (UECM):

76

P q

∆Yt = β0+ β1Yt -1 + β 2 X2t-1 + β3X3t-1 +………+βnX nt-1 +∑β n+1 ∆Yt -1 +∑β n+2 ∆X2t-1

i=1 i=0

r z

+∑β n+3∆X3t-1 +……… +∑β n+n∆X nt-1 + u t

i=0 i=0

where ∆ is the first-difference operator, u t is a white-noise disturbance term. The above equation. can be viewed as an ARDL of order ( p, q , r,….z). From the estimation of UECMs, the long-run elasticities can be estimated as being the coefficient of the one lagged explanatory variable (multiplied by a negative sign) divided by the coefficient of the one lagged dependent variable (Bardsen, 1989). For example, the long-run

elasticities, using the above equation, for the first two variables are (β2 /β1) and (β3

/β1) , respectively. The short-run effects are captured by the coefficients of the first-differenced variables. After regression of the UCEM equation, a Wald test (F-statistic) is calculated to test for the long-run relationship between the concerned variables. A Wald test is conducted by imposing restrictions on the estimated long-run coefficients of explanatory variables. The null and alternative hypotheses are as follows:

H0 : β1 = β 2 = β3……=βn = 0 (no long-run relationship)

Ha: β1 ≠β2 ≠ β3 ≠ ...... β n.≠0 β (a long-run relationship exists)

The computed F-statistic value should then be compared with the critical values tabulated in Table CI(iii) of Pesaran et al. (2001). According to these authors, the lower bound critical values are calculated assuming that the explanatory variables are integrated of order zero, or I(0), while the upper bound critical values have been calculated assuming that they are integrated of order one, or I(1). Therefore, if the computed F-statistic is smaller than the lower bound value, then the null hypothesis is not rejected and we should conclude that there is no long-run relationship between HDI indicators and its determinants. Conversely, if the computed F-statistic is greater than the upper bound value, then the dependent variable and its determinants share a long-run level relationship. On the other hand, if the computed F-statistic falls between the lower and upper bound values, then the test results are inconclusive. Once a long run relation is established, then the ARDL model is estimated using a simple OLS. The Granger Causality effect can be obtained by restricting the coefficient of the explanatory variables with its lags equal to zero (using Wald test). If the null hypothesis of no causality is rejected, then we can conclude that a relevant variable Granger-caused the dependent variable.

77

APPENDIX IV Output & Employment at 2003-04 & 2006-07 exports.

Sectors

Output at 2003-04 exports

Output at 2006-07 exports

Employment at 2003-04 output

Employment at 2006-07 output

Rs Billion

Rs Billion Millions Millions

Food crops 110.23 152.66 8.39 11.62

Cash Crops 119.77 202.54 2.38 4.03

Plantation Crop 21.41 42.27 0.48 0.95

Other crops 166.39 331.19 0.27 0.53

Animal Husbandry 75.02 125.22 0.29 0.48

Forestry and logging 27.25 49.16 0.25 0.44

Fishing 45.04 52.08 0.20 0.24

Coal and lignite 104.19 199.96 0.18 0.35

Crude petroleum, natural gas 308.07 683.53 0.07 0.15

Iron ore 39.43 74.24 0.05 0.10

Other Minerals 278.18 320.62 5.41 6.24

Food Products 206.13 384.52 0.52 0.96

Beverages, tobacco, etc. 7.99 14.07 0.01 0.01

Cotton textiles 169.24 259.79 1.03 1.58

Wool Silk and Synthetic fiber 115.83 170.56 0.55 0.81

Jute, hemp, mesta textiles 21.78 33.83 0.11 0.17

Textiles products including wearing apparel 383.98 547.23 2.11 3.00

Wood, furnuture,ect. 32.08 63.38 0.84 1.67

Paper & printing, etc. 77.08 145.53 0.13 0.24

Leather and leather products 81.77 107.69 0.45 0.59

Rubber, petroleum, plastic, cola. 487.78 1115.04 0.14 0.32

Chemicals, etc. 762.30 1359.99 0.28 0.49

Non-metallic products 49.89 72.09 0.23 0.33

Metals 755.62 1508.65 0.59 1.17 Metal products except mach. And tpt. Equipment 145.93 274.08 0.33 0.62

Tractors, agri. Implements, industrial machinery, other machinery 200.22 358.70 0.43 0.77

Electrical, electronic machinery and applications 150.80 287.57 0.02 0.04

Transport Equipments 114.60 251.33 0.05 0.10 Miscellaneous manufacturing industries 476.11 959.59 1.17 2.35

Construction 56.93 111.53 0.29 0.56

Electricity 320.80 598.07 0.24 0.44

Gas and water supply 9.50 17.94 0.02 0.03

Railway transport services 114.46 214.77 0.32 0.60

78

Other transport services 510.92 940.22 2.00 3.68

Storage and warehousing 4.30 7.67 0.02 0.03

Communication 48.02 98.56 0.24 0.49

Trade 714.67 1275.32 5.18 9.24

Hotels and restaurants 108.17 247.62 0.72 1.66

Banking 284.73 672.46 0.27 0.65

Insurance 59.89 141.11 0.06 0.15

Ownership of dwellings 0.00 0.00 0.00 0.00

Education and research 0.23 0.43 0.00 0.00

Medical and health 1.73 3.09 0.00 0.01

Other services 766.25 1952.55 2.55 6.49

Public administration 0.00 0.00 0.00 0.00

Tourism 230.54 357.63 0.77 1.20

Total 8765.22 16786.07 39.61 65.59

79

CHAPTER 2

Impact of Trade on Employment, Wages and Labour Productivity in Unorganized Manufacturing in India����

2.1 Issues and Objectives: The impact of economic reforms on the organized manufacturing sector in India is a well researched area. By delineating a series of trade and industrial policy reforms initiated from 1991, many studies have empirically analyzed their impact on firms’ productivity, efficiency, growth and competitiveness. The findings based on cross section as well as panel data across a large number of organized manufacturing units have confirmed a favourable effect of reduction in trade barriers and tariffs, increase in exports, industrial delicensing, deregulation and other liberal policies in improving firm’s competitiveness, total factor productivity and growth. Besides, there is extensive literature discussing the impact of foreign direct investment and technological advancements on organized sector’s employment, wage rate and labour productivity, pointing towards mixed results at best. However, there is meager evidence to corroborate whether trade liberalization has brought about any effect on employment, productivity and growth in the unorganized manufacturing sector as well. Lack of research on trade-related effects on unoganised sector is mainly due to lack of data with respect to trade orientation of the industry to which enterprises in the unorganized sector belong. Further, data on unorganized manufacturing is estimated by the NSSO with five years of lag, which rules out empirical analysis based on consistent data series over a longer period of time. However, given the characteristics of the unorganized sector i.e., (i) low levels of productivity, capital and technology and hence lack of competitiveness, (ii) dominance by unskilled workers, and (iii) greater presence in rural areas characterized by weak infrastructure, inappropriate environment to grow and compete, any analysis on the development impacts of trade needs to incorporate the impact of trade on this sector.. Considering the fact that the unorganized manufacturing is closely interlinked with the organized sector due to its backward and forward linkages27, it may not have remained insulated from the reform process that has taken place in the industrial sector during the

� This Chapter has been contributed by Seema Bathla (Associate Professor), R.K.

Sharma (Professor) in Centre for the Study of Regional Development, Jawaharlal Nehru

University and Rashmi Banga, Senior Economist, UNCTAD-India project.

27 The interdependence as highlighted in Mehta (1985), Samal (1990), Shaw (1990) is established through forward linkages by sale of output, subcontracting and marketing of products and through backward linkages by purchase of inputs and raw material, acquisition of skills and technology and credit. The forward linkages are said to be relatively weak compared to backward linkages, which are fairly strong.

80

nineties and early 2000. Further, recent statistics on unorganized manufacturing reveal a decrease in the number of workers during 2000/01 to 2005/06 accompanied by trivial improvement in the rate of growth of enterprises, output, productivity, and capital, particularly in the rural areas. Recognizing that this sector holds considerable importance due to its significant contribution in absorbing burgeoning labour force and in reducing poverty28, sluggishness in employment, and deceleration in output growth raises several issues as follows.

• Is a downfall in the rate of growth in enterprises, workers and output in unorganized manufacturing caused by a growing competition due to greater openness of the economy, perhaps due to surge in industrial imports?

• Is it related to inability of the industrial units to compete and be viable due to low capital and skill base or inadequate funds to invest?

• Do locational factors such as rural and urban settings influence employment and wages in unorganized manufacturing?

• Is the performance affected due to increasing work avenues in other economic sectors, due to higher wages or otherwise?

• If all the above factors assume significance, is it possible to measure the magnitude of effect of imports and exports in influencing employment, wage rate and productivity in the unorganized manufacturing sector?

With the above backdrop, this chapter makes an attempt to analyze the performance of the unorganized manufacturing sector in the context of trade liberalization, in particular, the impact of trade on labour markets in India in the post reform period. The focus is on analyzing inter-industry trends and growth patterns at an all India level as well as separately in the rural and urban sector, and various factors including exports and imports that influence employment, wage rate and labour productivity. This would provide insights into the scope and feasibility of various measures that aid in augmenting of productive employment and accelerating the pace of industrialization. The chapter provides a descriptive analysis based on 2 digit level data at two points of time viz. 2000-01 and 2005-06 relating to 56th and 62nd rounds of the NSS respectively. The empirical analysis of impact of exports and imports on employment, wage rate and labour productivity is carried out using the latest available unit level data for the year

28 The main reason behind this line of thought for the unorganized service sector as a whole are (a)

continuous decline in the share of agriculture workforce in total from 80 percent to 60 percent and a nearly constant share of manufacturing in total workforce hovering between 7-8 percent during the last two decades, (b) growing potential of this sector visible through a phenomenal increase in the number of unorganized enterprises and workers, accounting for three fourth of manufacturing workforce and 17 % of total NDP of the unorganized non-farm sector. This sector is often taken as a sponge for absorbing otherwise unemployed or people displaced from agriculture and poverty alleviation (See among others Kundu and Sharma, 2001; Bhalla, 2005; Sharma and Nayyar, 2005).

81

2005-06. Since trade data is available at the product level, a concordance matrix is prepared matching six digit HS product codes to three digit level National Industrial Classification (UNCTAD-India 2008). The NSS survey data of the unorganized sector at enterprise level reports three digit level industrial classification of the enterprises, using which the exports and imports data for the corresponding industries to which an enterprise belongs is constructed. Data is constructed for two years 2000-01 and 2005-06 at three-digit level of National Industrial Classification for industrial codes. It is important to mention at the outset that the manufacturing sector’s exports may be derived from formal (organized) sector as well as informal small scale manufacturing units, like handicrafts, metals, small scale carpet weaving units etc. Therefore, their impact on unorganized manufacturing sector’s employment and productivity may or may not be direct. It is also possible that unorganized manufacturing units may not be export intensive, rather they might be aiding the organized industrial units which may be export-oriented in their production process, due to existence of strong forward and backward linkages between the two as identified in the literature.

The study is divided into five sections. Following issues and objectives, section II briefly elucidates theoretical background on the effect of economic reforms and trade liberalization on manufacturing and presents literature survey in the Indian context. Section III examines inter-industry trends and growth patterns in the unorganized manufacturing sector based on two digit level data. Section IV explains empirical results of the impact of trade (exports and imports) on unorganized manufacturing employment, wage rate and labour productivity. Section V summarizes and draws implications. 2.2 Economic Reforms and their Impact on Manufacturing Sector in India:

Literature Survey

Literature reveals two divergent views on the impact of economic reforms and trade liberalization on Indian manufacturing. The proponents of trade liberalization have argued that economic reforms opened domestic markets to competition, which along with foreign direct investment and technological advancements would bring improvements in the productivity of domestic industries. Increased productivity, in turn, would lead to more efficient allocation of resources and higher rates of economic growth in the long run. Further, under the Heckscher–Ohlin–Samuelson framework, the economic theory suggests that trade would lead to a redistribution of employment from the import sector towards the export sector, i.e., increase in imports reduce employment and increase in exports adds to employment. Going by the principle of ‘international comparative advantage’ and also the fact that export enterprises are more labour intensive, it is maintained that trade would significantly contribute towards more employment generation in India. Those not in favour of a greater integration of domestic economy with world markets have argued that the envisaged positive effect of trade liberalization may not get realized. This is because the domestic firms would be unable to adopt foreign technologies effectively into their traditional methods of production and may also get restrained due to inadequate investment and credit. Besides, in many countries, extensive subsidies are given via lower average rates of interest on production and high quality low cost infrastructure, which may negate the perceived gains. Further, as envisaged in the

82

theory, capital and labour may not move freely from one activity to another to facilitate structural transformation through trade (Joshi and Little, 1996; Topalova, 2003, 2005). On a somewhat different but interrelated note, Rao (1994), Datt (1999), Papola (1999) and many others have contended that the policy shift towards greater openness is inherently biased towards organized industry and better skilled people in the urban sector. As a result, reforms would have a more substantial impact on the urban economy than the rural economy in India. Notably, rural areas are inhabited by a large proportion of poor having meager source of livelihood who are engaged primarily in the unorganized sector, which is characterized by inadequate social security benefits, job insecurity, poor working conditions, weak asset and resource base. This would imply that the rural non farm enterprises may not be able to compete and share the gains expected from the reform process. Nonetheless, in a recent study by Haggblade, Hazell and Reaerdon (2007), it is emphasized that in rural areas agricultural issues dominate the policy debates rather than the non-farm enterprises and in most situations, the policy environment in which these firms operate, emerge by default. The authors thus, argue that as a result, these policies may some times open up opportunities and in other instances destroy the whole industry. The issue necessarily hinges on an empirical analysis of impact of trade liberalization on the performance of Indian manufacturing sector, the evidence for which is well documented in the case of organized sector. Broadly, the studies have tried to empirically examine improvement in the total factor productivity, profits and price cost margins, employment and wages in the post reform period in response to industrial deregulation, delicencing, reduction in tariffs, non-tariff barriers, inflow of foreign direct investment (FDI) and technology. A brief survey of literature shows that on TFP growth in the Indian organized manufacturing sector, Topalova (2003) established a causal link between trade liberalization and firm productivity during 1989-1996. Based on a panel of firm level data and industry specific tariff reduction, he indicated that a tariff reduction by 10% and hence lesser protectionism leads to growth of firm productivity by about 0.5%. Notably, the public sector companies are less productive than privately held firms and their productivity levels do not change with trade liberalization. This aspect is explained by the fact that these companies do not face hard budget constraints and continue to operate at a loss in the face of competition. In another study, Goldar and Kumari (2003), while emphasizing on a positive and favourable effect of reforms on industrial productivity, found a slowdown in total factor productivity growth in Indian manufacturing in the post-reform period. A downfall in TFP growth is attributable, perhaps to a large extent, to deterioration in capacity utilization in the industrial sector and may be due to decline in the growth rate of agriculture. The study quoted results of Uchikawa (2001, 2002) that there was an investment boom in the Indian industry in the mid-1990s. While the investment boom raised production capacities substantially, demand did not rise which led to capacity under-utilization.

83

Further, Goldar and Aggarwal (2004) examined the effect of tariffs and non-tariff barriers on manufactured imports on price-cost margins in Indian industries. The analysis, based on panel data for 137 three-digit industries for the period 1980-81 to 1997-98, indicated that lowering of tariff and removal of quantitative restrictions on manufactured imports had a significant pro-competitive effect on domestic industries, tending to reduce markups or price-cost margins. However, price-cost margins did not fall in the post-reform period in most industry groups. Rather, there has been a marked fall in growth rate of real wages and a significant reduction in labor’s income share in value added, reflecting perhaps a weakening of industrial labor’s bargaining power. This seems to have neutralized, to a large extent, the depressing effect of trade liberalization on price-cost margins. Literature on impact of trade on employment reveals mixed findings. Ghose (2000) found that in the case of developing countries that have emerged as important exporters of manufactures to industrialized countries, trade has a large positive effect on employment and wages in the manufacturing sector. Also, trade tends to lead to a decline in wage inequality by increasing demand for unskilled workers. In the Indian context, Goldar (2002) found that employment elasticity for aggregate manufacturing increased from 0.26 in the pre reform period (1973-74 to 1989-90) to 0.33 in the post reform period (1990-91 to 1997-98). But a significant increase in employment elasticity is observed only in export-oriented industries group, as import competing industries revealed a fall in employment elasticity from 0.425 in the pre reform period to 0.264 in the post-reform period. As regards trend in real wages, results showed that growth in real wages per worker declined appreciably from 3.29 per cent per annum during the pre reform period to 1.16 per cent per annum in the post-reform period. The author attributed a decline in wage growth to government interventions in determination of wages in the organized sector in India. In a similar vein, Tendulkar (2003) analyzed organized industrial growth over three distinct policy regimes i.e., 1973-74 to 1980-81, 1980-81 to 1990-91 and 1990-91 to 1997-98. He found that 1973-74 to 1980-81 marked as the period of restrictive industrial and trade policies, the trend growth rate of output was 4.65 per cent and employment grew by 3.83 per cent. Product wage per worker increased at 3.2 per cent and implicit growth of productivity per worker grew at a negligible 0.8 per cent. The subsequent period from 1980-81 to 1990-91 was a period of somewhat liberal trade and industrial policies combined with an aggregate demand push provided by rising fiscal deficits and good agricultural harvests. It experienced jobless growth in manufacturing indicating only output growth at 7.1 percent. Real product wage grew by 4.5 per cent compared to implicit growth of 7.3 per cent in productivity per worker. The last period from 1990-91 to 1997-98, i.e. the period when economic reforms were initiated, witnessed considerable improvement in both output and employment growth at 9.0 and 2.9 percent respectively, with moderate product wage growth of 2.6 per cent. In another study, Goldar (2004), while stressing on slow down in productivity trends in post-reform periods in organized manufacturing, reiterated that trend growth rate in employment after 1997-98 i.e. from 1997-98 to 2001-02 was significantly negative, at about –3.3 per cent per annum. Further, trend growth rate in real value added during the same period was also very low at about

84

0.5 per cent per annum, which was much lower than trend growth rates in real value of output and index number of industrial manufacturing production in this period, both exceeding 5 per cent per annum. Using more updated data, Banga and Sharma (2008) have tried to empirically examine the impact of trade on employment and wages of unskilled workers at state level in organized manufacturing and agricultural sector. Wages of unskilled workers is taken as an indirect indicator of poverty to explain the trade-poverty nexus. Analysis carried out at three digit level for 54 industries from 1998-99 to 2005-06 shows favourable impact of exports on wages of unskilled worker in organized manufacturing. Import competition doesn’t seem to have displaced labour or adversely affected wages, which is explained by strict labour laws and downward rigidity of wages in the country. Broadly, the study concludes that positive impact of trade on employment and wages depends on the extent to which the poor are able to gainfully participate in the expanding sectors. Apart from exploring the effect of economic reforms and trade liberalization, many other studies have attempted to analyze impact of technology and FDI on productivity and employment growth in organized manufacturing sector. Banga (2005a) undertook analysis for 78 three-digit level industries to examine impact of technology, FDI and trade on employment and wage rate. Impact of technology is captured through three components, viz. research and development intensity, import intensity of capital goods and import intensity of soft technology. An index for technology acquisition is also constructed using principal component analysis to estimate impact of technological progress. Results based on dynamic panel data (DPD) model using generalised method of moments (GMM), show, while higher extent of FDI in an industry leads to higher wage rate in the industry, it has no impact on its employment. On the other hand, higher export intensity of an industry increases employment in the industry but has no effect on its wage rate. Technological progress is found to be labour saving but does not influence wage rate. Further, domestic innovation in terms of research and development intensity has been labour utilizing in nature but import of technology has unfavourably affected employment. In another study, Banga (2005b) examined impact of liberalization on labour productivity and wage inequality between skilled and unskilled labour in Indian manufacturing. The cross industry analysis carried out from 1991-92 to 1997-98, revealed that FDI, trade and technology have improved labour productivity but along with higher labour productivity, wage inequality between skilled and unskilled workers has also increased. All the three viz. FDI, trade and technology have differential impact on wage inequality. Higher FDI in an industry raises wage inequality, while higher export intensity of an industry is associated with lower wage inequality. Further, technological progress is found to be skill biased in this period and accordingly, higher extent of technology acquisition in an industry is found to be associated with higher wage inequality. It, thus, appears that impact of trade on organized manufacturing is diverse across industries and the time period chosen. As also reiterated by Gera and Massé (1996), exports are a dominant factor in employment growth in high-technology and skill-intensive industries, while

85

import penetration adversely affect employment growth in low-technology, labour-intensive industries. In contrast to a large number of studies on impact of trade reforms in organized manufacturing, evidence of the same in unorganized manufacturing is limited. Unni, Lalitha and Rani (2000) compared trends in growth and efficiency in utilization of resources in manufacturing at all India level and Gujarat before and after the reform periods. They found that both organized and unorganized manufacturing sectors in Gujarat have done better in terms of growth in value added. In another study, Rani and Unni (2004) found initial economic reform policies to have adversely affected employment in organized and unorganized manufacturing sectors, which got improved in the subsequent years. Also, the reform measures initiated had differential impact on various industry groups, in particular, growth in automobiles and infrastructure enabled growth in the unorganized segment. Raj and Duraisamy (2006) analyzed efficiency and productivity performance of unorganized manufacturing in 13 major Indian states using a large scale National Sample Survey data for five periods, broadly representing the pre (1978-79, 1984-85, 1989-90) and post reform (1994-95 and 2000-01) periods. The analysis based on Malmquist productivity indexes showed that in all states but Rajasthan, annual rate of total factor productivity growth has been higher in the reform period than in the pre-reform years, on an average. A better performance of unorganized manufacturing was due to good progress made in technical efficiency rather than due to technological progress, which has been a major factor in achieving high levels of total factor productivity during the reference period. Bathla and Sharma (2008) analyzed performance of rural unorganized manufacturing at disaggregate enterprise level for three rounds viz. 1995-96, 2000-01 and 2005-06. This was done for OAME, NDME and DME29. An increase in employment growth in rural sector during 1994-95 to 2000-01 is followed by a negative growth during 2000-01 to 2005-06 but only in self/family operated OAMEs in a majority of Indian states. While a downtrend in OAME workforce is explained by low capital, income and hence closure of enterprises, an increase in workers in NDME and DME is related to their steady performance over the years. However, the most disturbing feature is that in all three enterprises, while rate of growth of real wage rate and labour productivity is positive, the pace of growth has got decelerated in the post reform period, raising concerns about future prospects of this sector. A unit level analysis for 2000-01 and 2005-06 indicates labour productivity to be influenced by size of firm, capital intensity and loans advanced. Based on these findings, one may infer that trade liberalization has had a favourable effect on Indian manufacturing (both organized and unorganized), towards improving competitive strength by enhancing TFP and labour productivity. Concomitantly, there has

29 As per NSS, Own Account Manufacturing Enterprise (OAME) is one, which runs without any hired

worker employed on a fairly regular basis and is engaged in manufacturing and/or repairing activities (with family labour only). An establishment employing less than six workers (household and hired workers taken together) and engaged in manufacturing activities is termed as Non-directory Manufacturing Establishment (NDME). Finally, a Directory Manufacturing Establishment (DME) is one which has employed six or more workers (household and hired workers taken together) and is engaged in manufacturing activities.

86

been a decrease, and not an increase in rate of growth of TFP in the post reform period. Two, results are diverse across organized industrial units as far as response of employment and wage rate to economic reforms and liberal trade is concerned. Discrepancies in estimates could be due to different time period taken in the analysis and also not enough studies have been done on the subject to arrive at a consensus. Three, FDI, trade and technology have different costs and benefits associated with them. While FDI has improved labour productivity, it has also led to higher wage inequality. Technological progress has been skill biased in Indian manufacturing sector and is perceived to have led to an increase in wage inequality between skilled and unskilled labour. It is important to mention here that these findings mainly pertain to the organized manufacturing sector as there are not many studies done in the context of trade liberalization in unorganized manufacturing. The available literature has judged the impact of reforms merely by comparing trends and growth performance during the pre and post reform periods. In doing so, the studies have not tried to capture influence of actual exports, imports, tariffs and technological changes that have taken place over a period of time on employment and wage rate. Also, since unorganized manufacturing is heterogeneous across industries, states and rural-urban location, impact of trade at such disaggregate level is not much known. For sure, literature abounds on an exploration of rural and urban labour markets at district and state level, focusing on diverse aspects viz. employment, wage patterns and their linkages with poverty and growth, separately in farm and non farm sectors. But, in most of the studies, non-farm sector is studied as a whole, of which unorganized manufacturing constitutes just one of the economic sectors30. This paper fills this research gap in our understanding by empirically analyzing the impact of exports and imports on unorganized manufacturing sector by taking into account variations across industries, states and location (rural-urban). Recognizing the enormity of the subject, scope of this study is confined only to unorganized manufacturing based on two NSS rounds viz. 56th and 62nd. The focus is on analyzing growth performance across rural and urban industries and magnitude of impact of trade on three foremost components of labour market viz. employment, wage rate and labor productivity. Notably, impact of trade on unorganized manufacturing may not be straightforward as data available on this sector is at industry/enterprise level and export-import data is available at product level. Therefore, as mentioned above, a concordance matrix is constructed to match six digit HS 2002 codes to three-digit National Industrial Classification so as to arrive at export-import figures at industry level. Also, since export data is at product level, it is possible that it could be sourced from organized or small scale industrial sectors or both and imports may influence any of these sectors. Nonetheless, knowing that the Indian unorganized manufacturing is growing and absorbs

30 See among others, Chadha (2003); Bhalla (2005); Srivastava and Singh (2005); Sundaram (2007); Singh (2008) for a detailed exposition on the subject. The analysis is largely based on various rounds of quinquennial NSS employment-unemployment surveys.

87

vast labour force having low levels of skills, wages and productivity than the organized, this exploratory exercise would enable to provide some insights on some of the pertinent issues being raised in the literature in the context of trade liberalization. Before that, it would be useful to understand the basic structure and composition of unorganized manufacturing sector and its growth patterns at industry level in the rural and urban areas. This would set the stage for carrying out an empirical exercise. 2.3 Inter-industry Trends and Growth Patterns in Unorganized Manufacturing Sector in India

The unorganized manufacturing sector in India comprises a large number of small enterprises, often unregistered, mostly under proprietorship. The composition shows three main types of enterprises viz. OAME, NDME and DME. Table 2.1 provides summary of key variables in the unorganized manufacturing sector at all India and separately for rural and urban unorganized manufacturing during 56th round (2000-01) and 62nd round (2005-06). The information is obtained from quinquenial NSS surveys and is available at industry and state levels. The all India picture shows that during 2000-01 there were 17.02 million unorganized manufacturing enterprises (11.9 million in rural and 5.08 in the urban areas) providing employment to 37 million people, of which nearly 23.98 million fall in rural areas and 13.09 million in urban areas respectively. While there has been a marginal increase in number of enterprises to 12.13 million in rural and decrease to 4.94 million in urban during 2005-06, number of workers has fallen in both, in rural to 23.45 million and in urban to 12.98 million respectively. There exist 24 odd groups at two-digit level of industries (based on NIC 2004) engaged in a wide range of activities starting from manufacturing of cotton ginning-cleaning (code 1405), food and food products (codes 15), tobacco (code 16), textiles (code 17), wearing apparels (code 18), wood and wood products (codes 20), paper and paper products (code 21) to manufacturing of basic metals (code 27), electrical machinery and transport equipments (code 31), radio-television (code 32), furniture (code 36) and recycling (code 37). At three digit level, nearly 67 industries are identified. Apart from rural-urban bifurcation, data at industry and state levels is given as per OAME, NDME and DME, of which OAME occupies an overwhelming share (80-90%) in both enterprises and work force, particularly in rural areas. While share of OAME, NDME and DME in total rural enterprises has nearly been the same during 2000 to 2005, their share in total workers has changed. In rural manufacturing, it has declined from 79.83% to 76.82 % in OAME i.e. 3% point decline, and has improved in NDME and DME by 2 percentage points from 8.06 to 10.16 percent and 12.11 to 13.01 percent respectively. Whereas urban manufacturing shows OAME, NDME and DME having 70.9, 21.3 and 7.9 percent shares in total enterprises in 2000/01 which remained almost same during 2005-06, there has been a nearly 2-3 percentage point decline in the share of workers in OAME and an increase in DME over a period of five years. Further, 79.83 percent of workers in rural OAME contributed 63.05 percent to value added during 2000, the share of which declined to nearly 50% during 2005/06. In contrast, NDME and DMEs, with 10 and 13 percent of total workforce respectively contributed 17 and 33 percent to gross value added, and their share in total GVA has shown a consistent improvement over time. In urban manufacturing too, DME has progressed as its share in

88

workers and GVA accelerated from 27.13 to 30.21 and 40.34 to 51.85 percent respectively during 2000 to 2005. Apparently, in rural unorganized manufacturing, it is the OAME that contributes the maximum in terms of workforce and value addition but lags behind NDME and DME as GVA per worker in OAME has remained less than half of that in other two enterprises. Possibly, the OAMEs absorb people who might have been displaced from agriculture and have low capital base and productivity. Also, due to low levels of technology and capital, these small family run enterprises are not able to compete with others and hence likely to become unviable. On the other hand, in urban areas, NDME and DME contribute the maximum in GVA, and together employ around 54.8 percent of workforce and add 74.2 percent of value added. Further, NDME in urban areas is growing rapidly compared to rural areas which could be due to ‘ancillarisation’ of enterprises, easy subcontracting due to greater presence of organized manufacturing, inadequate employment opportunities outside this sector, better infrastructure and so on (Kundu, Lalitha and Arora, 2001; Sharma and Dash, 2006). Which industries have witnessed a high/low concentration of enterprises and workers in the rural and urban unorganized manufacturing and an increase/decrease in the workforce? Following is an inter-industry analysis for all the three enterprises together for the years 2000-01 and 2005-06. The analysis is carried out at 2 digit level separately for rural and urban manufacturing focusing mainly on the key variables that broadly indicate the performance of this sector. The variables considered are, number of enterprises and workers and employment per enterprise to represent size; gross value added (GVA) per enterprise to represent both size as well as viability of the unit; GVA per worker also called labour productivity to represent state of health of a unit; market value of fixed assets (owned+hired) per worker and per enterprise and loan outstanding per enterprise to reflect capital or asset base of an enterprise; wages/emoluments per hired worker to assess the status of workers. These variables broadly determine growth of the units as well as point towards relationship between value added and various inputs like credit and assets. Table 2.2 presents share of each industrial unit at two-digit level of industries in total enterprises, employment, GVA, market value of fixed assets (owned+hired) and loan outstanding (LO) during 2000/01 to 2005/06. It is evident that in both rural-urban locations, 8 sectors viz. food, tobacco, textiles, wearing apparels, wood, other non-metals, fabricated metals and furniture are prominent in terms of maximum number of enterprises and workers. Barring the last three i.e. metals and furniture, all above mentioned units are mostly agro processing or agro based units accounting for nearly 80% share in total in rural areas and 65% in urban areas. The maximum share in enterprises in total (rural-urban) during 2005-06 was occupied by wearing apparel at 18.83%, food, tobacco and textiles units at around 16% followed by wood at 12% and furniture at 6.75%. Except in wearing apparel and furniture, share of these units in total units is more in rural areas compared to urban areas during both the years. Further, share of most of the units in total enterprises has increased over a period of five years except in food and wood products units.

89

As shown in Table 2.3, rate of growth in all the enterprises during 2000/01 to 2005/06 is positive and low at 0.32 in rural areas and negative at 0.58 in urban areas. At a disaggregate industry level, of all the rural units, cotton ginning, tobacco, textiles, wearing apparels, paper, chemicals, electrical machinery, radio-television, motor vehicles and other transport have witnessed a positive rate of growth. In urban units, a positive growth is observed in tobacco, textiles, wearing apparel, leather, chemicals, machinery, office equipment, electrical machinery, medical instruments, and other transport equipment. Trends in employment, wages and GVA per worker are presented below. Employment: As observed from data, out of 24 sectors, 8 sectors which hold the highest share in total enterprises in both rural and urban manufacturing apparently provide employment to the maximum number of workers. Among these, the highest share in employment (nearly 75% in rural and 60% in urban) is captured by food and food products-21.16%, textiles-16.11%, wood-15.03%, tobacco-14.58, and wearing apparel-11.02% respectively. This could be explained by (i) presence of surplus labour force in agriculture, (ii) stagnant agricultural growth forcing people towards occupational diversification, (iii) easy entry and exit in agro based units due to low skill and capital requirements, and (iv) proximity to markets for procuring raw material. From 2000 to 2005, the share of rural workers in total employment has increased only in tobacco and wearing apparels and has remained unchanged in food, textiles and furniture. Wood and non-metals have observed a decline in workforce by 2-3 percentage points. In urban manufacturing, an increase in the share of workers is again seen in textiles and wearing apparels by 2%age points, decline in food and wood and nearly the same in tobacco, leather, chemical and furniture. In terms of rate of growth in the rural workforce during 2000 to 2005, employment has increased by 21.48% in cotton ginning, 4.53%in tobacco, 2.32% in wearing apparels, 24.97% in paper products, 14 and 35% in chemicals and furniture. A negative growth rate in rural employment is the maximum in recycling at 20.82%, followed by leather tanning at 7%, wood and non metals at around 5% each. A positive and high rate of growth for urban workers is identified mainly in office accounting at 58%, transport at 14% and tanning of leather at 7.49%. A high and negative growth rate (more than 10%) is observed in coke, radio-television, motor vehicles and recycling, followed by 6.41% in agro based wood, and 4% each in cotton, food and paper manufacturing. Wages/Emoluments per Worker: Data on wages is not given at the 2 digit level. Rather it is given for emoluments per hired worker which comprise wages/salaries and group benefits contributed by the employer. The share of each industry in total absolute value of emoluments in 2005 indicates the maximum, nearly 23%, is cornered by food and food products and non metals enterprise, followed by textiles and furniture at 14.30 and 11.9% in rural areas. While food products and furniture have shown an increase in their share over a period of five years, the other two have shown a decline. In urban manufacturing, the maximum share of emoluments is found only in three viz. textiles and fabricated metals at 17% each and furniture at 14.92%. All other enterprises hold a relatively low share (less than 5%) excepting food at 6.94 and machinery at 8.73%. Though in both rural and urban manufacturing, the share of enterprises in wages has been more or less stable during 2000-2005, but food products, textiles and wearing apparels have shown a

90

decline in their share in urban areas. The wage rate (emoluments per hired worker) growth is found to be positive and high for cotton ginning, tobacco, food products, metals and electrical machinery in both rural and urban areas, averaging at 9.11 and 7.18 percent together for all the industries. A higher rate of growth could be due to the fact that the analysis has been done at nominal prices. GVA and Labour Productivity: Gross value added is the most important variable as it reflects an overall performance of the enterprise. The absolute value of GVA, again at nominal price shows food products, textiles, wearing apparels, wood products, non metals and furniture to hold the maximum share ranging between 8 to 26% in rural areas and between 10-16% in urban areas. From 2000 to 2005, share of manufacturing of food products in rural GVA has increased from 23 to 26% and has remained the same for other. In the urban GVA, share of food and wearing apparels has declined from 12.74% to 10.67% and from 14.9% to 11.9% and that of fabricated metals and furniture has increased from 8.44 to 14.99% and 14.37 to 16.44% respectively. It is evident that in rural manufacturing, the share of food, tobacco, textiles, wearing apparels and wood products constitutes nearly 65% in total GVA and the same in urban manufacturing is nearly 40%, suggesting a relatively greater dominance of these units in rural areas. The rate of growth in labour productivity is almost same at 8% in both rural and urban manufacturing, but shows large variations across industries. While labour productivity in rural manufacturing is positive in all units except tobacco and other transport, it is negative only in coke and refined products in urban. The highest rate of growth is observed in cotton ginning (28.43%), food (10.64%), textiles (8%), paper products (7.88%), coke (13.22%), metals (13%) and radio-television and furniture (33%) in rural units during 2000-2005. In urban manufacturing, labour productivity is positive and high in cotton ginning (31%), food, publishing-printing and chemicals (8% each), metals (20%), motor vehicles and furniture (11% each) during the same period. Other Performance Indicators: Under this, we have considered two key variables in unorganized manufacturing sector viz. market value of fixed assets (owned and hired) per worker and loan outstandings (LO) per enterprise. Both variables broadly reflect the capital intensive nature of the units. A closer look at the composition of industrial units again reveals that like GVA, food and food products have the maximum share in absolute value of fixed assets at 31% in rural areas, followed by textiles and wearing apparels at 13% each, wood and non metals at 8 % each, totaling around 70% share. The same in urban manufacturing is cornered by food, textiles, wearing apparels and furniture between 12-14% each followed by fabricated metals at 10.52%. As far as share of these units in absolute loan advanced is concerned, it is again high in food, textiles, and then in chemicals, metal products, basic and fabricated metals enterprises. While in rural areas, the share of LO has increased over time in nearly all units except textiles and chemicals, in urban it has increased in fabricated metal and furniture and has declined in almost every unit or at best has remained the same. The growth in capital intensity defined as the market value of fixed assets per worker (FAW) is found to be negative only in two rural units viz. paper products and motor vehicles and in one urban unit viz. medical

91

equipments. Similarly, LO per enterprise has grown over a period of five years in rural and urban manufacturing at 9.83% and 25.91% respectively. Clearly, urban manufacturing has an edge over rural as the rate of growth in LO per enterprise is higher in the former compared to the latter in almost all the units. Only two sectors viz. electrical machinery and transport equipment have shown a negative rate of growth in urban manufacturing. Whereas, in rural manufacturing five enterprises have witnessed a negative growth, of which chemicals and tobacco have the highest at 31.8% and 25.66% respectively followed by cotton ginning at 19.14%, paper products at 12.86% and textiles at 7.68%. Besides looking at inter industry variations, another important aspect that is worth highlighting is a considerable difference in most of the performance indicators, when bifurcated by location. The rural-urban ratios presented in Table 4 make evident that rural manufacturing has almost an equal and in some cases 2-4 times more number of enterprises and workers, such as in cotton ginning, food and food products, tobacco, textiles, wood and wood products, coke, chemicals, and non metals. There has also been a perceptible increase (more than double) in the number of rural workers and enterprise in two time periods in cotton ginning, paper products and coke. Though the rural enterprises employ maximum number of workers and also hold a relatively higher share in absolute values of GVA and fixed assets, these lag behind the urban enterprises as far as labour productivity, GVA per enterprise, capital intensity, LO per enterprise and wage rates are concerned. This phenomenon is pertinent in almost all rural units and the situation has not improved much over a period of time. Across various enterprises, GVA per enterprise and labour productivity in rural units is nearly half of that in urban in most of the units-almost two- thirds in cotton, tobacco, rubber, non-metals, medical equipments and 25-40% more in coke and radio-television manufacturing units. Same is the case with capital intensity and LO per enterprise where rural share is half of that of urban during 2005/06. Only in one sector viz. office equipment, capital intensity is high in rural manufacturing, and in 5 sectors loan outstanding per enterprise is more in rural areas viz. rubber, metals, radio-television, medical and motor equipments. At the aggregate level, rural urban ratio for LO per enterprise was 0.29 in 2000/01 which got reduced to 0.14 in the subsequent round, indicating low amount of loans advanced to rural manufacturing. Again, LO per enterprise was higher in agro based units in rural during 2000-01 due to high loans taken by cotton ginning and food units, which became equal to that in urban in the subsequent round. Clearly, the rural urban gap has narrowed down for wage rate, which in most rural units is nearly equal or at best 2/3rd of urban wage rate. Food, textiles, and printing are a few selected sectors that have hardly witnessed any improvement in wages, possibly due to the fact that these are family operated. 2.4 Main Findings of Descriptive Analysis: 2001-02 to 2005-06 From the preceding analysis, two observations can be made. First, the unorganized manufacturing sector is marked by an uneven pattern of distribution of enterprises and

92

workers across various enterprises in both rural and urban areas. Over a period of time, this sector has certainly grown in terms of number of enterprises and output. However, a positive growth in enterprises, and also labour productivity, capital intensity and institutional finance is accompanied by a negative growth in employment, which is a cause of concern. A little more deliberation into this issue reveals that in both rural and urban manufacturing, it is the OAME that has an overwhelming share in enterprises (and simultaneously workers) and value addition. The performance of OAME at both locations has relatively been sluggish compared to NDME and DME as it suffers from low levels of output, labour productivity and capital assets. This suggests that growth of the unorganized manufacturing sector as observed during a span of five years is attributable primarily to DME and at best NDME. Another disquieting feature is that percentage decline in OAME share in total GVA is more in rural areas (from 63 to 49.5%) compared to urban areas (from 25.75 to 18.43%). This would also indicate that the low and negative rate of growth in enterprise and workers may be accompanied by a deceleration or a negative growth in labour productivity. The growth rate in GVA per worker during 2000-2005, estimated at real prices using GDP deflator for unregistered manufacturing, is -0.88% in OAME, 2.5% in NDME and 9.7% in DME in rural manufacturing. It corroborates the argument given in the literature that OAMEs are possibly opted due to stagnancy in agriculture growth and/or other better job avenues and are hence distress driven. And the scenario has remained unchanged over the years from 1978/79 to 2005/06, reinforcing that these small enterprises continue to be in a poor state of low productivity trap in either situation of an increase or decrease in the number of workers (Kundu, Lalitha and Arora, 2001; Bhalla, 2005). Conversely, the growth performance of DMEs and in some cases of NDMEs, though decelerating over the years, has been positive, implying a move into these as a matter of choice and opportunities. Second, looking at a large share of enterprises and workers, rural manufacturing appears to be at an advantageous position. But employment per enterprise in the rural sector is 2.41, which is far below than that in urban manufacturing at 3.48, indicating that the latter has more potential in absorbing people. By and large, rural manufacturing lags behind the urban in terms of GVA per enterprise and GVA per worker and is also relatively less capital intensive. A similar picture emerges for agro and non agro based sectors revealing food, tobacco and textiles units to hold maximum number of enterprises and workers in both rural and urban manufacturing but lagging behind others (mainly non agro enterprises) in labour productivity and capital intensity. Also, non-agro units have experienced a relatively higher absolute increase as well as growth rate in value added per worker, capital intensity, wage rate and institutional finance compared to agro units, especially in the rural areas. A categorization of rural and urban industries is done in Table 5 as per the value of one important performance variable viz. GVA per worker. It shows that in rural manufacturing tobacco (code 16), textiles (code 17), wood (code 20) and paper products (21) had labour productivity less than Rs. 10,000 during 2000/01. Over a period of five years, labour productivity in tobacco remained low at Rs. 6,900 in 2000 and 6,500 in

93

2005 respectively whereas other units including wearing apparels and chemicals moved to the next category ranging between Rs.10-15 thousand. Nonetheless, these units were again at their lowest absolute value during 2005/06. The only exception agro based units are cotton ginning and food products, which have experienced a phenomenal increase in labour productivity from Rs. 12,000 during 2000/01 to Rs. 20,000 and Rs. 42,000 respectively during 2005/06. Similarly, in urban manufacturing, tobacco units have experienced the lowest value of output per worker during both survey rounds indicating their adverse condition over the years. No doubt labour productivity in urban textiles, wearing apparels, wood and paper units is higher than that in rural, but it continued to fall in the same range of Rs. 20-30 thousand, implying not so significant improvement. Like in the rural sector, cotton ginning and food products in urban have also shown tremendous improvement. Overall, labour productivity in both rural and urban sectors is higher in the non agro based units during 2000-01 and 2005-06. Similarly, large divergences can also be seen in wage rate, which is estimated to be as low as Rs. 17,814 in food units to a higher value at Rs. 46,690 in machinery, and loan outstanding per enterprise at its lowest value at Rs. 77 in tobacco to its highest value at Rs. 13 lakh in office-accounting machinery during 2005-06. Higher loan outstandings in non-agro units is understandable due to capital intensive nature of these units necessitating high capital and credit requirements. Importantly, wage gaps are found to have narrowed down considerably over a period of five years, indicating some improvement in the status of workers. A higher concentration and heterogeneity of enterprises and workers in unorganized manufacturing along with large productivity gaps between rural and urban as well as between agro-non agro units may be attributable to (i) inadequate opportunities, forcing people to open up small units having low capital intensity and technological base, that result in low levels of output per worker and income (ii) most of the agro units open up due to easy backward linkages with agriculture for procurement of raw material and also entry/exit of workers into these units may be easier due to low skill and capital requirements, and (iii) weak infrastructure and difficult access to institutional loans. All these factors may negatively impinge on the efficiency of rural sector, particularly agro based units. Units other than agro are relatively more efficient and have performed better. This may be due to the fact that these have high level of capital intensity and technology, and also workers might have acquired some skills or training to sustain in that work. The analysis certainly underscores the need to find the extent to which each of these factors may influence labour market in the unorganized manufacturing sector. More so, when creation of employment in meaningful and productive ways31 is considered to be the key to higher income growth, increase in wage earnings and reduction in poverty32.

31 See among others, NCEUS (2007), Thakur and Venkata Ratnam (2007), Kashyap and Mehta (2007) on

improving the efficiency of labour. 32 Sharma and Bathla (2008) elicited high correlation between labour productivity and poverty during 2005-06, in rural manufacturing at -0.66; urban manufacturing at -0.51; together in rural-urban at -0.62. The

94

Further, recognizing the fact that all these factors are internal to the working of the industrial units and bear strong inter-linkages33, it is pertinent to explore whether external factors such as trade have any effect on employment and labour productivity. The issue, which is the main focus in this chapter, stands important as India has opened its markets to external trade from the early nineties after initiation of macro policy and sector specific reforms. Both industrial exports and imports have witnessed an increase over the years showing ratio of trade in GDP to have accelerated from 16% in the early nineties to 27% in the early 2000. As shown in Table 2, share of industrial exports (either sourced from organized or unorganized sectors) across industries shows higher share of textiles, wearing apparels, chemicals, metals and furniture. There has been a substantial decline in the share of exports of textiles in total from 24% to 10.26%; chemicals from 23% to 13.3%, metals from 10.16 to 7.4 and nearly the same in wearing apparels during 2000-01 to 2005-06. Share of exports of two enterprises viz. coke & refined petroleum products and furniture exports has picked up from 0.29% in 2000/01 to 12.51% in 2005/06. Compared to exports, share of industrial imports in total is the maximum for coke-refined petroleum products, followed by chemicals, and furniture, accounting for almost 60% of total share in imports. Again, while share of imports of petroleum products in total has fallen from 47.41% to 34.66% and that of furniture from 29.9% to 16.8%, it increased for chemicals from 0.68% to 10.3%. Exports from the agro based units account for 20-25% of the share in total exports and their share in total imports is negligible. Both exports and imports figures reveal a high rate of growth over time, particularly in tobacco, leather, wearing apparels, chemicals, office accounting machinery, furniture and refined petroleum products. For all enterprises taken together, exports have grown at a much higher rate at 70% compared to that of imports at 30% during the five years time. Results of various studies presented in section II, support that trade liberalization, and the resultant increase in exports and imports has influenced organized manufacturing in many ways. However, no consensus has been reached as to whether trade has had any significant impact on unorganized manufacturing as well. In what follows, we present empirical results of the effects of trade and other variables on unorganized manufacturing sector’s employment, labour productivity and wage rate.

2.5 Impact of Trade on Employment, Wage Rate and Labour productivity The impact of trade (exports and imports) on labour demand, wage rate and labour productivity in unorganised manufacturing is empirically examined within a framework

estimate was high and statistically significant in rural manufacturing in both the survey years viz. 2000/01 and 2005/06, but was found to be significant in urban manufacturing only in 2005/06. 33 Correlations worked out between labour productivity, GVA per enterprise, capital intensity, wage rate and institutional loans in the unorganized manufacturing sector are found to be positive and statistically significant in both rural and urban sectors (Kundu, Lalitha and Arora, 2001; Sharma and Dash, 2006).

95

of economic and trade theories. The analysis is carried out for 2005-06 based on enterprise level NSS data at three digit level. Following are the three equations that are generated and the specification of variables.

1. Labour demand (workers)=f(size, wage rate, export intensity, import intensity, rural-urban dummy, industry dummies, state dummies)

2. Wage Rate =f(size, labour productivity, capital intensity, export intensity,

import intensity, rural-urban dummy, industry dummies, state dummies) 3. Labour productivity=f(fsize, capital intensity, export intensity, import

intensity, rural-urban dummy, industry dummies, state dummies) Where, Enterprise (firm) size is captured by GVA, Capital intensity denotes market value of fixed assets per worker Labour productivity is GVA per worker, Wage rate is wage per hired worker, Export intensity is the ratio of exports of the industry (to which the firm belongs) to the total output of the industry. Similarly, import intensity is the ratio of imports of the industry to its output. For a given year, i.e., 2005-06,we estimate the impact of export intensity and import intensity of the industry to which the firm belongs on labour demand, wage rate and labour productivity of the firm (also referred as enterprise). The estimates are generated using OLS. After trying for various functional forms in linear and log linear forms, the analysis is restricted to double log form only. The estimates are checked for consistency and robustness using different techniques. As mentioned earlier, the unorganized manufacture sector in India comprises of three segments-OAME, own account manufacturing enterprises employing no hired worker i.e. either the owner or a helper is self-employed in these tiny units. Second component is NDME ,employing at least one hired worker and having an employment size of 2-5 workers per enterprise and finally the DME, directory manufacturing establishments employing 6-10 workers. Recognizing a diverse pattern across these three types of enterprises, which are dominated by OAMEs, the regression analysis is carried out separately for all three enterprises. OAME is finally dropped from the analysis as the estimates generated lacked robustness and also, by definition, employment and number of enterprises at unit level are closely related. That is, growth in these enterprises will automatically generate employment of almost equal number of workers. Again, by definition there are no wages or zero wages in OAMEs as these units are generally self operated and log of zero is not defined. Results for all three enterprises together, and for NDME and DME are presented in Tables 6-8.

96

Empirical Results on Employment Table 6 presents the OLS estimates of the impact of trade on employment in unorganized manufacturing sector during 2005-06. The labour demand (dependent variable is log of employment per unit) is explained by explanatory variables viz. log GVA used as a proxy for the ‘size’ of the unit variable. Literature in the context of export market has used firm size as a proxy for essential firm resources required to venture into the market. The size of a firm is expected to have positive impact on employment. Other explanatory variables used are log of wage rate, export intensity defined as exports of the industry to which the firm belongs as a ratio of the industry’s output and log import intensity defined as total imports of the industry as a ratio of its output. The latter is taken to explore the extent to which imports as a proportion to domestic output can influence labour demand. The rural urban dummy is used to capture the impact of urbanization, the unit located in rural areas is assigned zero and one otherwise. Industry dummies have also been used to control inter-industry variations. In all, there are 67 industries at three digit level, out of which recycling sector is excluded in the analysis and hence 65 dummies are specified. Similarly, state level dummies, numbering 34, have also been used to control state specific effects, as there are 35 states and union territories. Table 2.6: Impact of Trade on Employment in Unorganized Manufacturing :2005-06

Dependent Variable: Log No. of workers

All Enterprises NDME DME

Independent Variables in Log

Coefficient (t value)

Coefficient (t value)

Coefficient (t value)

GVA 0.464 (75.17)* 0.26 (26.49)* 0.38 (34.21) *

Wage Rate -0.246 (-31.83)* -0.10 (-8.51)* -0.30 (-22.90) *

Export Intensity 0.0403 (2.84) * 0.055 (2.25)* 0.16 (4.22) *

Import Intensity -0.0127 (-0.92) -0.055 (-2.71)* -0.099 (-3.24) *

Rural-Urban Dummy

0.09 (15.78) * 0.008 (1.76)** 0.10 (10.08) *

Industry Dummies Yes Yes Yes

State Dummies Yes Yes Yes

Constant -1.34 (-27.67) -0.67 (-10.94) -0.33 (-2.76)

R-squared 0.68 0.32 0.47

Number of observations

28461 17449 9574

Note:* indicates significance at 1%, ** indicates significance at 5%, ***indicates significance at 10%.

The results are explained for NDME and DME. They reveal that size of the unit has a positive and significant impact on employment and wage rate, as expected, has a negative and significant impact on labour demand,

97

The results show that higher the export intensity of the industry to which the enterprise belongs, higher will be the employment of the enterprise. It has been pointed out that exporting industrial units in developing countries are expected to have favorable impact on labour market as their choice of technology and wage policy tends to be much in line with the comparative advantage in the international production of the host country. In India, exporting industries continue to be mainly labour intensive, like garment manufacturing and, this is likely to have a positive impact on employment as exports increase. The results show labour demand in an enterprise belonging to export-oriented industry will be positively influenced by export intensity which indicates that. increase in exports has a favourable impact on employment in unorganized manufacturing. There has been an increase in the number of firms working on contractual basis from 27.6% during 2000/01 to 31.7 % during 2005/06, which, in a way, corroborates this finding. The import competition faced by the industry to which the enterprise belongs is found to have a significant negative impact on employment. It may be that because of increased domestic competition, unorganised manufacturing loose out and get eradicated as has happened in the case of OAME where enterprise/employment growth rate has declined over the years in the post-reform period. A positive and significant impact of rural urban dummy suggests that urban areas are doing better as compared to rural areas, which has already been elicited in section III. These results appear to be in accord with inferences drawn from Heckscher-Ohlin model that higher exports may lead to more employment of the enterprises in the unorganized sector. In India, exports generally take place from labour intensive industries and it is evident from the results that exports have influenced labour demand in NDME and DME, which are dominated by unskilled workers. Also, the results, though not comparable with other studies due to organized and unorganized nature of operations and varying time periods for time series/cross section analyses, are in line with Goldar (2002), who found higher employment elasticity of demand in export oriented industries in the post-reform period; and also with Banga (2005a), who estimated positive impact of exports and negative impact of imports on organized employment. The findings also appear to substantiate the argument given in Gera and Massé (1996) that import penetration adversely affect employment growth in low-technology, labour-intensive industries and also in Haggblade, Hazell and Reaerdon (2007) that liberalized trade and exchange rate policies generally hurt rural firms that compete with imported goods, while helping enterprises that serve export markets or use imported equipments and inputs. Empirical Results on Wage Rate Table 2.7 furnishes results of the wage equation. The dependent variable is log wage per worker. The equation has been estimated for NDME and DME as OAMEs do not employ any hired worker and, therefore no wages are paid per se.

98

Table 2.7: Impact of Trade on Wage Rate in Unorganized Manufacturing during 2005-06

NDME DME

Independent Variables in Log Coefficient (t value)

Coefficient (t value)

GVA 0.47 (19.40) * 0.198 (14.18) *

Labour Productivity 0.32 (10.53) * 0.61 (34.43) *

Capital Intensity 0.034 (7.11) * -0.004 (-0.71)

Export Intensity 0.147 (7.66) * 0.074 (2.63) *

Import Intensity -0.18 (-9.65) * -0.099 (-3.00) *

Rural-Urban Dummy -0.023 (-2.87) * -0.005 (-0.32)

Industry Dummies Yes Yes

State Dummies Yes Yes

Constant -0.520 (-5.15*) 0.057 (0.29)

R-squared 0.66 0.87

Number of observations 17426 9570

Note:* indicates significance at 1%, ** indicates significance at 5%, ***indicates significance at 10%.

The results show that after controlling for the size of industry and irrespective of labour productivity, higher level of export intensity of the industry to which the enterprise belongs leads to higher wage rate, which may be because enterprises in the exporting industries are able to pay more. The variable export intensity bears a positive and significant impact on wage rate in the industry, whereas higher the import completion faced by the industry to which the enterprise belongs, lower is the wage rate paid by the enterprise. Banga (2005a) and Banga and Sharma (2008) find no adverse effect of import competition on wages of skilled/unskilled workers in the organized manufacturing. One plausible explanation of this could be that unlike government’s role in determination of wage rate and strict labour laws in the organized sector, wages in the unorganised sector are more flexible, which may be due to absence of government’s intervention through any legislation or otherwise. Finally, as expected, labour productivity favorably influences wage rate for NDME & DME. Further, log of capital intensity has a positive and significant impact on wage rates only in the case of NDME. A negative and significant coefficient of rural urban dummy suggests that wage rates are lower in rural areas in the case of NDME’s whereas there are no significant rural urban differences in wage rate in the DME’s.

99

Empirical Results on Labour Productivity Liberalization is often associated with increasing labour productivity, which is confirmed empirically by Francisco Alcalá and Antonio Ciccone (2001) who analyzed the effect of international trade on average labour productivity across countries. The authors, while reiterating that the causal effect of trade on productivity across countries is large and highly significant, found that trade works through labour efficiency. In the developing countries, an increase in labour productivity is often associated with higher wage rate and in some cases, with increase in wage inequality between skilled and unskilled labour. In the Indian context, sources of growth in labour productivity could also be related to increase in capital intensity or elimination of inefficient units, like OAME. It may have adverse impact on employment but would augment productivity at an aggregate level. The results are presented in Table 2.8. Table 2.8: Impact of Trade on Labour Productivity in Unorganized Manufacturing during 2005-06

All Enterprises NDME DME

Independent Variables in Log

Coefficient (t value)

Coefficient (t value)

Coefficient (t value)

GVA 0.65 (451.11) * 0.79 (213.84) * 0.80 (209.69) *

Capital Intensity 0.091 (56.80) * 0.059 (15.46) * 0.086 (17.94) *

Export Intensity 0.074 (17.29) * 0.031 (0.37) 0.089 (3.06) *

Import Intensity -0.091 (-6.46) * 0.0008 (0.01) -0.077 (-2.87) *

Rural-Urban Dummy 0.01 (3.66) * 0.01 (2.82) * -0.079 (-7.20) *

Industry Dummies Yes Yes Yes

State Dummies Yes Yes Yes

Constant 0.10 (16.92) 0.022 (0.04) -0.87 (-4.69)

R-squared 0.88 0.87 0.94

Number of Observations 81253 17760 9615

Note:* indicates significance at 1%, ** indicates significance at 5%, ***indicates significance at 10%.

As is clear from Table 2.8, labour productivity in the unorganised manufacturing sector is positively influenced by the size, captured by log GVA and capital intensity. There is a direct relationship between labour productivity and firm size in both DME’s and NDME’s showing that larger units tend to have higher labour productivity. The log of capital intensity also has a positive impact on labour productivity, both in NDME and DME. As far as trade is concerned, labour productivity in DME is more affected by trade as compared to that in NDME. Clearly, export intensity has no significant impact on labour productivity of NDME whereas a positive and significant impact is registered in the case of DME, indicating that firms in unorganized sector need to be of a certain minimum size in terms of employment (at least more than five workers) to improve their productivity

100

gains from trade.Similarly, log of import intensity has a counter veiling impact in case of DME’s and no impact on NDME’s. It indicates that an increase in imports may adversely affect labour productivity in DMEs. The results at the aggregate level suggest that enterprises belonging to export-oriented industries have higher labour productivity, while import competition reduces the labour productivity in the unorganized sector. Further, a positive coefficient for rural urban dummy suggests that urban areas have a positive impact on labour productivity in case of NDME’s. A negative value estimated for DMEs indicates that DME’s are doing relatively better in rural areas. The industry dummies and state dummies are used to take care of the inter industry and inter state variations. Overall, as found in case of organized manufacturing in Banga (2005b), results obtained in this chapter for unorganised manufacturing also show export intensity to have a significant and positive effect on labour productivity. This would also indicate that enterprises engaged in exports, mainly DMEs, will have to maintain standards and adopt innovation to be in competition at the international level. 2.6 Conclusions and Implications This chapter makes an attempt to analyze performance of the unorganized manufacturing sector in India, in particular, the impact of trade on labour market in the post reform period. The analysis begins with inter-industry trends and growth patterns at 2 digit level during 2000/01 and 2005/06 followed by an empirical estimation of the impact of various factors including exports and imports on employment, wage rate and labour productivity using unit data at 3 digit level for 2005/06. The findings reveal an uneven distribution of enterprises and workers across three types of enterprises in both rural and urban unorganized manufacturing in the post reform period. Despite the fact that OAME contributes substantially in terms of number of enterprises, workers and value addition, it continues to suffer from low levels of output, labour productivity and capital assets at both locations, relatively poor in the rural areas. Further, of various industrial enterprises, the performance of agro based units, which are large in number as well as in terms of number of workers, lags much behind that of the non-agro based units. A heterogeneous pattern, along with large productivity gaps across various enterprises in the rural and urban sectors may be attributable to several factors such as, (a) inadequate opportunities, forcing people to open up small units having inadequate capital, lower labour productivity and hence, lower incomes, (b) easy entry/exit of workers due to low skill and capital requirements, and (c) weak infrastructure. This suggests that growth of the unorganized manufacturing sector, in terms of increase in number of enterprises, workers and output as observed during a span of five years, is attributable primarily to DME and at best NDME. It is also evident that while concentration of enterprises/workers in OAME appears to be born of desperation as a survival strategy, a steady growth in NDME and DME could emerge as a matter of choice and opportunities.

101

The empirical analysis covers 66 industries at three digit level, and thirty five states and union territories to capture the labour market rigidities at all India, and separately for NDME and DME. The results, based on ordinary least square method, reveal both internal factors (operating within the system) and external factors (trade in this paper) to be at work in labour market, and having differential impact on employment, wages and labour productivity. While higher industrial exports lead to higher employment, higher labour productivity and higher wages, higher imports in the country tend to have unfavorable impact on these three components. The results have been arrived at by controlling industry specific and state specific effects with the help of dummies. The important finding that emerges from the analysis is favourable impact of trade on the unorganised manufacturing sector, which may be due to the fact that exporting units are labour intensive in nature and are able to pay higher wages as well. Import of goods tends to have a negative impact on employment, labour productivity and wages. This could be due to increasing competition as imported items have better technology and are relatively less priced. It may be emphasized that the unorganised manufacturing sector in the country, with its low capital base, is very large and growing. It is responsible for absorbing the growing labour force of the country, mainly from rural areas. Although this sector is not covered by any policy regime but it appears to be sensitive from employment and wages point of view, as this would determine implications of liberalization of the economy for labour absorption and poverty reduction. In the face of this and also due to growing importance of this sector, the government of India has instituted a National Commission for Enterprises in the Unorganised Sector (NCEUS) whose recommendation may be put into practice in the near future. Unlike the organized sector, where wage rigidity is an important feature, this sector has been able to work in a wage flexibility scenario to an extent possible. In fact, results obtained do show that growth in real wages in unorganised manufacturing has decelerated in the post-liberalization period, which in a way has facilitated the cost adjustments (Bathla and Sharma, 2008). Increasing capital intensity may be able to increase the wage rate. To sum up, in order to expand employment, this sector should continue to augment by finding its own space wherever available. To minimize the social costs involved due to liberalization of the economy in a developing country like India, one needs to take precaution that labour displaced due to increasing imports of goods and technology should be absorbed by simultaneously expanding exports of labour intensive goods. In this regard, lessons may be learnt from the experiences and strategies adopted by China in promoting township and village enterprise or TVEs (Mukherjee and Zhang, 2007). Initiating liberal policies, creating physical and social infrastructure, developing institutions that extend loans and export credits for investment, imparting vocational training and skills to people are some of the ways which can go a long way in accelerating and strengthening growth in the unorganized manufacturing sector in India.

102

APPENDIX

Table 2.1: Summary Chart on Key Variables in Unorganised Manufacturing during 2000-01 and 2005-06

2000 Rural Urban Combined 200 5 Rural Urban Combined

Enterprises (million no.) 11.93 5.08 17.02 12.13 4.94 17.07

Employment (million no.) 23.98 13.09 37.08 23.46 12.98 36.44

Employment Per Enterprise 2.01 2.57 2.18 1.93 2.63 2.13

% share of Type of Enterprises in All Enterprises

OAME 92.66 70.88 86.14 91.59 70.90 85.60

NDME 5.27 21.26 10.05 6.14 20.74 10.37

DME 2.07 7.86 3.80 2.26 8.36 4.03

% share of Type of Enterprises in All Workers

OAME 79.83 45.16 67.59 76.82 43.64 65.00

NDME 8.06 27.71 15.00 10.16 26.15 15.86

DME 12.11 27.13 17.42 13.01 30.21 19.14

% share of Type of Enterprises in Total GVA

OAME 63.05 25.75 42.28 49.50 18.43 31.89

NDME 13.80 33.91 25.02 16.90 29.66 24.20

DME 23.10 40.34 32.70 33.60 51.85 44.15

GVA Per Enterprise (Rs.) 22348 65863 35357 31355 100267 51307

Labour Productivity (Rs.) 11120 25598 16233 16211 38167 24034

Capital Intensity (Rs.) 27600 137500 60500 38732 194321 83780 Loan Outstandings per Enterprise (Rs.) 2900 10100 5100 4634 31966 12548

Emoluments per Hired Worker (Rs) 13082 22133 18488 20233 31302 26682

103

Table 2.2: Share of Each Industrial Unit in Total in the Unorganised Manufacturing Sector in India-2000-01 and 2005-06

Enterprises 2000-01 Enterprises 2005-06 Workers 2000-01

Workers 2005-06

NIC 2 Digit Code Rural Urban All Rural Urban All Rural Urban All Rural Urban All

1405 0.02 0.09 0.04 0.05 0.05 0.05 0.02 0.06 0.04 0.05 0.05 0.05

15 19.78 12.78 17.69 17.11 10.67 15.25 21.56 12.76 18.45 21.16 10.64 17.41

16 13.81 8.95 12.36 19.32 9.62 16.51 11.43 5.09 9.19 14.58 5.99 11.52

17 14.16 14.22 14.17 14.56 16.23 15.04 16.15 17.63 16.67 16.11 19.75 17.40

18 14.00 22.37 16.50 16.03 25.70 18.83 9.61 17.32 12.33 11.02 18.05 13.53

19 0.77 1.66 1.03 0.32 2.14 0.84 0.56 2.03 1.08 0.40 2.93 1.30

20 20.67 6.79 16.52 15.79 4.53 12.53 18.67 5.67 14.08 15.03 4.11 11.14

21 0.22 1.23 0.52 0.88 1.25 0.98 0.25 1.46 0.68 0.79 1.30 0.97

22 0.19 2.39 0.85 0.15 2.03 0.69 0.24 3.23 1.29 0.23 2.75 1.13

23 0.04 0.04 0.04 0.04 0.02 0.03 0.06 0.05 0.06 0.08 0.03 0.06

24 0.90 2.22 1.29 2.45 2.45 2.45 1.27 2.02 1.53 2.55 2.04 2.37

25 0.29 1.20 0.56 0.21 0.95 0.42 0.38 1.85 0.90 0.37 1.47 0.76

26 5.75 2.62 4.82 4.42 2.14 3.76 10.69 3.71 8.23 8.37 2.87 6.41

27 0.12 0.48 0.23 0.11 0.44 0.21 0.17 0.70 0.36 0.14 0.61 0.31

28 3.11 5.34 3.77 3.05 5.06 3.63 2.82 6.95 4.28 3.03 7.30 4.55

29 0.68 1.67 0.98 0.56 2.15 1.02 0.64 2.58 1.33 0.55 3.48 1.59

30 -- -- -- -- 0.02 0.01 -- 0.01 -- -- 0.08 0.03

31 0.19 0.85 0.38 0.43 1.18 0.65 0.21 1.55 0.68 0.38 1.39 0.74

32 0.01 0.11 0.04 0.02 0.07 0.03 0.01 0.29 0.11 0.03 0.12 0.06

33 0.01 0.14 0.05 0.01 0.17 0.06 0.02 0.20 0.09 0.02 0.21 0.09

34 0.03 0.37 0.13 0.05 0.18 0.09 0.05 0.74 0.29 0.21 0.33 0.25

35 0.04 0.22 0.10 0.05 0.39 0.15 0.06 0.39 0.17 0.05 0.75 0.30

36 5.17 14.09 7.83 4.40 12.50 6.75 5.08 13.53 8.06 4.82 13.68 7.97

37 0.05 0.17 0.09 0.01 0.07 0.02 0.05 0.19 0.10 0.02 0.08 0.04

All 100 100 100 100 100 100 100 100 100 100 100 100 Note: Industries falling under each of these codes is given before the references.

104

Table 2.2 contd…

Gross Value Added 2000 Gross Value Added 2005 Market Value of Fixed Assets 2000-01

Market Value of Fixed Assets 2005-06

NIC 2 Digit Code Rural Urban All Rural Urban All Rural Urban All Rural Urban All

1405 0.02 0.04 0.03 0.14 0.08 0.11 0.06 0.04 0.05 0.18 0.11 0.13

15 23.83 12.74 17.66 26.60 10.67 17.59 31.18 12.72 18.61 30.86 12.23 18.35

16 7.11 1.46 3.96 5.89 1.25 3.27 3.89 0.84 1.81 3.85 1.57 2.32

17 13.39 15.12 14.35 13.46 14.35 13.97 12.84 13.11 13.02 13.75 14.60 14.32

18 9.82 14.93 12.67 9.16 11.80 10.65 12.28 14.23 13.59 12.82 12.99 12.94

19 0.79 2.19 1.57 0.45 2.40 1.55 0.60 1.62 1.29 0.34 1.96 1.43

20 14.22 5.17 9.18 10.24 3.16 6.24 8.55 4.83 6.02 7.87 3.51 4.94

21 0.20 1.40 0.87 0.64 0.94 0.81 0.40 1.70 1.29 0.61 1.42 1.16

22 0.61 4.37 2.70 0.46 3.80 2.35 1.36 6.34 4.74 1.30 5.78 4.31

23 0.12 0.06 0.09 0.19 0.02 0.09 0.19 0.06 0.10 0.30 0.03 0.12

24 1.24 1.78 1.54 1.90 1.84 1.86 1.84 1.86 1.85 3.00 1.47 1.97

25 0.94 3.13 2.16 0.90 2.16 1.61 1.39 3.32 2.70 1.60 2.84 2.44

26 14.48 3.10 8.14 14.09 2.52 7.55 13.58 2.98 6.37 7.89 3.15 4.71

27 0.88 1.12 1.01 0.95 1.63 1.33 0.43 1.43 1.11 1.07 1.46 1.33

28 3.71 8.44 6.35 4.15 14.99 10.28 3.82 9.64 7.77 4.95 10.52 8.69

29 0.93 4.71 3.04 0.78 7.41 4.53 1.29 5.89 4.42 1.45 6.14 4.60

30 -- 0.01 0.01 -- 0.16 0.09 -- 0.01 0.01 -- 0.08 0.06

31 0.40 2.78 1.72 0.88 2.01 1.52 0.55 2.24 1.70 1.26 2.20 1.89

32 0.03 0.47 0.28 0.24 0.26 0.25 0.04 0.61 0.43 0.20 0.31 0.27

33 0.06 0.39 0.24 0.07 0.38 0.24 0.06 0.33 0.24 0.13 0.42 0.33

34 0.12 1.29 0.77 0.55 0.68 0.62 0.24 2.11 1.51 0.46 0.49 0.48

35 0.22 0.73 0.50 0.10 0.97 0.59 0.06 0.86 0.61 0.12 1.62 1.13

36 6.82 14.37 11.02 8.10 16.44 12.82 5.35 12.97 10.52 5.94 14.98 12.01

37 0.05 0.20 0.14 0.05 0.08 0.07 0.07 0.24 0.19 0.04 0.12 0.09

All 100 100 100 100 100 100 100 100 100 100 100 100

105

Table 2.2 contd…

Loan Outstandings 2000-01 Loan Outstandings 2005-06

Emoluments-Hired Workers 2000-01

Emoluments-Hired Workers 2005-06 2000-01 2005-06

NIC Code Rural Urban All Rural Urban All Rural Urban All Rural Urban All Exports Imports Exports Imports

1405 0.21 0.02 0.10 0.14 0.10 0.11 0.02 0.02 0.02 0.10 0.04 0.06 1.05 0.03 0.48 0.06

15 24.83 9.67 15.63 38.11 8.75 16.45 14.37 8.36 10.07 22.16 6.94 11.76 2.28 1.37 2.59 2.20

16 0.99 0.11 0.46 0.20 0.07 0.10 2.90 0.22 0.98 1.45 0.27 0.64 0.00 0.00 0.35 0.01

17 20.11 13.24 15.86 8.68 6.42 7.01 16.94 19.18 18.54 14.30 17.80 16.69 24.27 0.88 10.26 1.84

18 2.03 5.03 3.79 2.59 5.42 4.68 4.63 14.26 11.51 4.60 11.13 9.06 9.09 0.01 10.44 0.10

19 0.18 1.14 0.75 0.22 0.86 0.69 0.39 2.15 1.65 0.69 2.79 2.13 1.14 0.03 2.69 0.28

20 4.44 4.37 4.36 3.08 1.90 2.21 5.17 4.20 4.48 5.01 2.78 3.49 0.17 0.97 0.10 0.70

21 1.25 2.28 1.85 1.57 1.22 1.31 0.35 1.56 1.22 0.38 0.99 0.79 0.46 1.00 0.46 1.14

22 0.64 7.98 4.98 0.67 7.00 5.34 0.69 4.89 3.70 0.98 4.24 3.21 0.07 0.32 0.16 0.36

23 0.05 0.09 0.07 0.71 0.06 0.23 0.25 0.06 0.11 0.30 0.02 0.11 0.29 47.41 12.51 34.64

24 18.60 4.07 9.82 4.68 2.46 3.04 3.05 1.86 2.20 2.87 1.50 1.93 23.81 5.68 13.35 10.32

25 3.63 8.49 6.48 5.48 2.71 3.43 1.92 3.28 2.89 1.45 2.39 2.09 1.06 0.26 2.20 1.01

26 9.71 4.56 6.57 11.65 2.19 4.67 33.50 3.17 11.81 23.61 2.39 9.11 2.17 0.27 1.39 0.51

27 5.19 3.97 4.42 6.31 1.41 2.69 1.17 1.20 1.19 1.20 1.02 1.08 10.16 2.03 7.47 6.47

28 3.95 6.42 5.37 4.22 10.64 8.95 3.90 9.48 7.89 4.50 17.65 13.49 4.70 0.52 3.71 1.28

29 1.22 9.63 6.19 1.20 4.75 3.82 1.17 5.52 4.28 0.81 8.73 6.23 5.08 3.96 3.73 7.53

30 -- -- -- -- 0.77 0.57 -- 0.02 0.01 -- 0.24 0.16 0.11 0.42 0.52 2.56

31 0.69 8.09 5.07 2.72 1.89 2.11 0.65 3.62 2.77 1.37 1.61 1.53 1.78 0.44 1.56 1.82

32 0.08 0.48 0.31 1.55 0.16 0.53 0.05 0.66 0.49 0.32 0.19 0.23 2.29 1.17 0.79 1.63

33 0.03 0.63 0.39 0.42 0.34 0.36 0.12 0.41 0.32 0.13 0.41 0.32 2.03 1.03 0.84 2.03

34 0.25 1.16 0.79 1.09 0.51 0.66 0.32 1.73 1.32 1.61 0.64 0.95 4.46 0.88 3.18 0.71

35 0.04 1.13 0.69 0.11 0.51 0.40 0.29 0.79 0.65 0.15 1.22 0.88 1.51 1.42 1.72 6.01

36 3.07 7.71 5.79 4.61 39.74 30.52 8.13 13.18 11.74 11.90 14.92 13.96 2.03 29.91 19.50 16.80

37 0.01 0.04 0.03 0.01 0.13 0.10 0.03 0.18 0.13 0.10 0.09 0.09 0.00 0.00 0.00 0.00

All 100 100 100 100 100 100 100 100 100 100 100 100 100.00 100.00 100.00 100.00

106

Table 2.3: Growth Rate of Key Variables in the Unorganised manufacturing Sector in India: 2000/01 to 2005/06

Enterprises Workers Labour Productivity GVA/Enterprise

NIC Code Rural Urban All Rural Urban All Rural Urban All Rural Urban All

1405 25.75 -10.11 7.24 21.48 -3.66 8.30 28.43 31.09 28.21 24.17 40.39 29.47

15 -2.54 -4.11 -2.87 -0.82 -3.72 -1.49 10.64 8.41 9.34 12.61 8.84 10.90

16 7.29 0.85 6.02 4.53 3.11 4.26 -1.08 1.78 -0.51 -3.62 4.05 -2.16

17 0.89 2.08 1.25 -0.49 2.12 0.52 8.00 4.79 6.66 6.53 4.83 5.88

18 3.07 2.22 2.73 2.32 0.66 1.51 3.47 2.48 2.57 2.73 0.92 1.36

19 -15.96 4.58 -3.93 -7.05 7.49 3.48 3.08 2.43 3.90 14.00 5.28 11.91

20 -4.94 -8.32 -5.33 -4.67 -6.41 -4.91 5.46 4.73 4.93 5.76 6.91 5.39

21 32.27 -0.31 13.56 24.97 -2.49 7.06 7.88 2.50 -0.72 1.91 0.26 -6.41

22 -4.68 -3.78 -3.91 -1.17 -3.34 -3.07 2.62 8.77 8.11 6.40 9.26 9.06

23 -0.86 -10.33 -2.99 3.52 -10.35 0.12 13.22 -0.87 9.05 18.17 -0.77 12.56

24 22.61 1.37 13.69 14.54 0.06 8.75 2.13 8.78 3.00 -4.59 7.37 -1.47

25 -5.87 -5.11 -5.38 -1.23 -4.59 -3.62 7.70 5.26 5.51 13.02 5.84 7.48

26 -4.84 -4.52 -4.78 -5.20 -5.19 -5.20 12.63 9.42 11.98 12.21 8.64 11.50

27 -1.13 -2.55 -2.02 -3.48 -2.95 -3.11 13.08 19.96 17.51 10.39 19.47 16.20

28 -0.05 -1.65 -0.71 1.02 0.80 0.90 8.66 20.31 17.65 9.82 23.31 19.56

29 -3.45 4.51 0.93 -3.61 6.01 3.36 7.48 11.68 12.98 7.30 13.28 15.70

30 -- 38.54 40.36 -- 58.71 58.89 -- 9.39 9.33 -- 26.02 24.44

31 18.70 6.18 11.09 11.84 -2.37 1.16 12.63 3.83 3.93 6.13 -4.53 -5.36

32 15.63 -8.90 -3.42 19.55 -16.15 -10.57 35.09 14.57 18.17 39.52 5.45 9.41

33 -2.74 2.88 1.98 -0.73 0.98 0.69 10.08 6.50 7.00 12.56 4.50 5.65

34 12.05 -13.83 -7.30 35.09 -14.89 -3.00 7.15 11.78 6.49 29.28 10.40 11.43

35 2.59 11.59 9.12 -1.13 14.15 11.59 -7.65 0.26 -0.24 -11.04 2.56 2.00

36 -2.83 -2.93 -2.89 -1.49 0.04 -0.57 12.79 11.05 11.74 14.35 14.45 14.40

37 -36.80 -18.30 -24.08 -20.82 -15.84 -17.25 33.02 6.26 12.42 66.54 9.51 22.56

All 0.32 -0.58 0.05 -0.44 -0.17 -0.35 7.83 8.32 8.16 7.01 8.77 7.73

107

Table 2.3 contd… Capital Intensity Loan Outstanding/Enterprise Emoluments /hired worker NIC Code Rural Urban All Rural Urban All Rural Urban All Exports Imports

1405 9.35 32.67 19.95 -19.14 83.26 14.22 33.06 16.52 21.75 45.3 56.2

15 8.02 9.80 8.10 23.17 27.93 24.61 10.62 5.54 6.42 74.5 46.7

16 2.52 17.13 7.61 -25.66 13.22 -16.44 50.10 15.04 42.01 388.6 290.5

17 9.37 6.60 8.29 -7.68 6.10 0.51 3.66 3.68 3.57 43.2 54.5

18 5.83 3.92 4.15 12.21 24.31 21.64 10.11 4.69 5.16 74.9 103.3

19 3.05 3.00 5.31 37.17 13.09 22.82 0.65 3.40 2.89 101.8 107.1

20 10.77 6.76 7.96 7.71 15.52 10.42 6.50 4.44 4.70 54.3 25.0

21 -6.33 5.35 -2.35 -12.86 10.84 -1.49 -5.69 4.82 2.68 70.6 36.8

22 7.61 8.20 8.07 16.58 26.72 26.41 9.46 9.15 8.74 101.4 36.7

23 14.08 4.80 10.20 88.66 31.46 56.52 15.43 6.19 13.37 262.1 25.2

24 3.32 1.62 -0.56 -31.80 11.66 -16.64 10.96 5.22 7.47 51.5 50.3

25 11.83 8.22 8.51 27.09 4.95 11.52 6.05 6.09 5.94 97.0 75.4

26 1.60 13.64 6.04 20.08 13.20 17.51 10.22 7.98 9.81 55.6 50.8

27 33.67 10.16 14.29 15.89 4.35 10.72 16.93 7.32 10.18 59.9 68.1

28 11.91 7.55 8.23 11.70 40.80 33.62 7.72 18.06 16.57 62.2 60.0

29 13.91 1.32 4.15 13.66 4.00 7.77 3.57 10.71 10.44 59.9 51.7

30 -- 0.19 0.20 -- 133.24 130.19 -- 7.20 7.20 133.5 91.6

31 13.24 8.71 7.84 22.18 -11.88 -9.53 22.99 0.45 3.63 65.6 77.3

32 24.76 10.65 8.98 71.09 11.06 37.56 26.59 2.39 6.57 37.5 42.5

33 26.64 10.90 12.55 94.45 7.39 15.86 0.07 6.47 5.15 42.5 52.7

34 -9.61 -6.60 -12.51 32.01 23.32 24.86 5.72 4.46 3.56 59.0 27.6

35 21.88 5.90 8.33 27.99 -4.47 -1.41 0.33 2.59 2.44 74.6 78.1

36 11.32 9.61 10.29 23.01 79.03 72.03 9.11 5.13 5.86 167.4 18.8

37 22.36 10.22 12.75 66.28 94.52 101.54 3.18 3.95 5.22 -- --

All 7.84 6.72 7.16 9.83 25.91 19.73 9.11 7.18 7.61 70.1 33.4

108

Table 2.4: Patterns of Inter Industry Variations in Key Variables in Unorganized Manufacturing in India: Rural Urban Ratios during 2000-01 and 2005-06

Enterprise Workers Labour Productivity GVA/Enterprise

Capital Intensity

Loan Outstanding/Enterprise

Emoluments/Hired Worker

2000-01 2005-06 2000-01 2005-06 2000-01

2005-06

2000-01

2005-06

2000-01

2005-06 2000-01 2005-06 2000-01

2005-06

1405 0.48 2.58 0.57 1.82 0.84 0.76 0.99 0.53 1.15 0.44 11.66 0.19 0.47 0.92

15 3.63 3.93 3.09 3.59 0.48 0.53 0.41 0.49 0.37 0.34 0.48 0.39 0.49 0.62

16 3.62 4.93 4.11 4.40 0.94 0.82 1.07 0.73 0.53 0.27 1.70 0.21 0.45 1.72

17 2.33 2.20 1.68 1.47 0.42 0.49 0.30 0.33 0.27 0.31 0.44 0.22 0.65 0.65

18 1.47 1.53 1.02 1.10 0.51 0.54 0.36 0.39 0.40 0.44 0.19 0.11 0.48 0.62

19 1.08 0.36 0.50 0.24 0.57 0.59 0.27 0.40 0.34 0.34 0.10 0.25 0.83 0.73

20 7.14 8.56 6.03 6.61 0.36 0.38 0.31 0.29 0.14 0.17 0.10 0.07 0.74 0.82

21 0.42 1.72 0.32 1.10 0.36 0.47 0.28 0.30 0.34 0.19 0.89 0.27 0.97 0.57

22 0.18 0.18 0.14 0.15 0.81 0.61 0.60 0.52 0.75 0.73 0.29 0.19 0.67 0.68

23 2.83 4.68 2.33 4.79 0.64 1.25 0.54 1.28 0.58 0.89 0.14 0.84 0.88 1.34

24 0.95 2.46 1.15 2.26 0.48 0.35 0.58 0.32 0.40 0.44 3.24 0.28 0.39 0.50

25 0.56 0.54 0.38 0.45 0.63 0.71 0.42 0.59 0.52 0.61 0.51 1.33 0.82 0.82

26 5.15 5.07 5.27 5.27 0.70 0.81 0.72 0.85 0.41 0.23 0.28 0.37 0.86 0.95

27 0.58 0.63 0.44 0.43 1.41 1.05 1.07 0.72 0.32 0.84 1.51 2.56 1.02 1.56

28 1.36 1.48 0.74 0.75 0.47 0.28 0.26 0.14 0.25 0.31 0.30 0.10 0.81 0.51

29 0.96 0.64 0.46 0.28 0.34 0.28 0.16 0.13 0.23 0.41 0.09 0.14 0.84 0.60

30 -- 0.07 -- 0.01 -- 0.52 -- 0.04 0.00 1.19 -- -- -- --

31 0.52 0.90 0.25 0.49 0.45 0.68 0.22 0.37 0.46 0.57 0.11 0.57 0.56 1.55

32 0.17 0.57 0.08 0.50 0.61 1.40 0.30 1.21 0.35 0.64 0.68 5.89 0.55 1.60

33 0.21 0.16 0.21 0.20 0.61 0.72 0.60 0.87 0.40 0.77 0.14 2.73 0.80 0.58

34 0.19 0.72 0.11 1.14 0.68 0.55 0.39 0.87 0.48 0.40 0.75 1.05 0.85 0.90

35 0.45 0.29 0.27 0.13 0.89 0.59 0.53 0.26 0.13 0.27 0.06 0.26 1.01 0.90

36 0.86 0.86 0.69 0.64 0.55 0.59 0.44 0.44 0.28 0.31 0.31 0.05 0.74 0.89

37 0.74 0.21 0.45 0.33 0.46 1.42 0.28 2.29 0.30 0.51 0.33 0.15 1.33 1.28

All 2.34 2.45 1.83 1.81 0.43 0.42 0.34 0.31 0.26 0.27 0.29 0.14 0.59 0.65

109

Table 2.5: Inter-Industry Variations in Labour Productivity in Rural and Urban Unorganized Manufacturing Sectors in India

NSS Rounds Less than Rs. 10000 - Rs.15000 - Rs 20000- Rs 30000- Rs. Above

Rs.10000 Rs.15000 Rs.20000 Rs.30000 Rs. 40000 40000

Rural Sectors

2000-01 16, 17, 1405,15,18, 19,26,29 22, 23,25, 27,33,34,35 --

20, 21 24, 37, 28, 36 31,32

2005-06 16 17, 18, 20, 19 15,26,28,29, 22,23,25, 1405,27,32,

21, 24 35, 36 30, 31 33, 34,37

Urban Sectors

2000-01 16 1405 -- 15,17,18,19,20, 22,23,28 25,27,29,30,31,32,

21,24, 26,26,27 33,34,35

2005-06 16 -- -- 17,18,20,21 15,19,23, 1405,22,25,27,28,29,

24,26,37 30,31,32,33,34,35,36

NIC 2 digit Code Description

1405 Cotton ginning, cleaning and baling

15 Manufacture of food products and beverages

16 Manufacture of tobacco products

17 Manufacture of textiles

18 Manufacture of wearing apparel; dressing and dyeing of fur

19 Tanning and dressing of leather; manufacture of luggage,handbags,saddler,harness and footwear

20 Manufacture of wood and of products of wood and cork, except furniture; manufacture of articles of straw and plating materials

21 Manufacture of paper and paper products

22 Publishing, printing and reproduction of recorded media

23 Manufacture of coke, refined petroleum products and nuclear fuel

24 Manufacture of chemicals and chemical products

25 Manufacture of rubber and plastic products

26 Manufacture of other non-metallic mineral products

27 Manufacture of basic metals

28 Manufacture of fabricated metal products, except machinery and equipment

29 Manufacture of machinery and equipment

30 Manufacture of office, accounting and computing machinery

31 Manufacture of electrical machinery and apparatus n.e.c.

32 Manufacture of radio, television and communication equipment and apparatus

33 Manufacture of medical, precision and optical instruments, watches and clocks

34 Manufacture of motor vehicles, trailers and semi-trailers

35 Manufacture of other transport equipment

36 Manufacture of furniture

37 Recycling

110

References Alcala Francisco and Antonio Ciccone (2001), Trade and Productivity, CEPR DP3095. Banga, Rashmi (2005), Impact of Liberalization on Wages And Employment in Indian

Manufacturing Industries, Working Paper No. 153, Indian Council for Research on International Economic Relations (ICRIER).

Banga, Rashmi (2005), Liberalization and Wage Inequality in India, Working Paper No. 156, Indian Council for Research on International Economic Relations (ICRIER).

Banga, Rashmi and Sharma (2008), “ Trade-Poverty Nexus in India”, presented in International Conference on “Trade-Poverty Nexus in South Asia”, Colombo.

Bathla Seema and R K Sharma (2008), Employment and Labour Productivity in Rural Unorganized Manufacturing: Growth, Disparities and Determinants, Unpublished Paper.

Bhalla Sheela (2005), Rural Workforce Diversification and Performance of Unorganized Sector Enterprises, in Rohini Nayyar and Alakh N. Sharma (ed.), Rural Transformation in India: The Role of Non-farm Sector, Institute for Human development, New Delhi, 2005.

Chadha G.K. (2003), Rural Non-Farm Sector in Indian Economy: Growth, Challenges and Future Direction, paper presented at JNU-IFPRI WORKSHOP on “The Dragon and the Elephant:A Comparative Study of Economic and Agricultural Reforms in China and India March 25-26, 2003.

Datt, Gaurav (1999), Has Poverty Declined since Economic Reforms: Statistical Analysis, Economic and Political Weekly, Vol … No 8.pp. 3514-18. August.

Gera, S. and Masse, P. (1996), Employment Performance in the Knowledge Based Economy, Industry Canada Working Paper No. 14, Human resource Development, Canada W-97-9E/F, Canada.

Ghose, A.K. (2000), Trade liberalization and manufacturing employment, Employment Paper no. 2000/3 (Geneva, International Labour Office).

Goldar, Biswanath (2002),‘Trade Liberalization and Manufacturing Employment: The Case India’, ILO 2002/34. Goldar, Bishwanath (2004), Productivity Trends in Indian Manufacturing in the Pre- and Post-Reform Periods, Working Paper No. 137, Indian Council for Research on International Economic Relations (ICRIER).

Goldar, Bishwanath and Anita Kumari (2003), Import Liberalization and Productivity Growth in Indian Manufacturing Industries in the 1990s, Developing Economies, December, 41(4): 436-60.

Goldar, Bishwanath and Suresh Chand Aggarwal (2004), Trade Liberalization and Price-Cost Margin in Indian Industries, Working Paper no. 130, Indian Council for Research on International Economic Relations, New Delhi.

GOI-NCEUS (2007), Report on Conditions of Work and Promotion of Livelihoods in the Unorganised Sector, National Commission for Enterprises in the Unorganised Sector, Government of India. August.

Government of India, Unorganised Manufacturing Sector in India, NSS 51st round July 1994 - June 1995, NSS 56th Round (July 2000-June 2001), NSS 62nd Round (July 2005-June 2006),National Sample Survey Organization, Department of Statistics, New Delhi.

111

Haggblade, Steven, Petr B.R. Hazell and Thomas Reardon (2007) (ed.), Transforming the Rural Nonfarm Economy: Opportunities and Threats in the Developing World, Oxford University Press.

Joshi, Vijay and I.M.D. Little (1996), Indian Economic Reforms 1991-2001, Oxford: Clarendon Press.

Kashyap S.P. and Niti Mehta (2007), Non-farm Sector in India: Temporal and Spatial Aspects, The Indian Journal of Labour Economics, (Conference Issue), 50 (4), October-December: 611-632.

Kundu, Amitabh and Alakh N. Sharma (ed.) (2001), Informal Sector in India-Perspectives and Policies, Institute for Human Development, and Institute of Applied Manpower Research, New Delhi.

Kundu, Amitabh, N. Lalitha and Sant Lal Arora (2001), Growth Dynamics of Informal Sector in Urban India: An Analysis of Interdependencies, in Kundu, Amitabh and Alakh N. Sharma (ed.), Informal Sector in India-Perspectives and Policies, Institute for Human Development, and Institute of Applied Manpower Research, New Delhi.

Kundu Amitabh, Niranjan Sarangi and Bal Paritosh Das (2005), Economic Growth, Poverty and Non-farm Employment: An Analysis of Rural-Urban Linkages, in Rohini Nayyar and Alakh N. Sharma (ed), Rural Transformation in India: The Role of Non-farm Sector, Institute for Human development, New Delhi, 2005.

Mehra Meeta (1985), Urban Informal Sector, Economic and Political Weekly, Vol XX, No 8. pp. 326-332.

Mukherjee Anit and Xiaobo Zhang 2007, Rural Industrialization in China and India: Role of Policies and Institutions, World Development, 35 (10), October: 1621-1634.

Papola, T.S. (1999), Labour Studies in India: Some Reflections on Developments in the Post Independence Period, The Indian Journal of Labour Economics, Vol.42, No.2, pp. 175-185.

Raj S N Rajesh and Malathy Duraisamy (2006), Economic Reforms, Efficiency Change and Productivity Growth: An Inter-State Analysis of Indian Unorganised Manufacturing Sector.

Rani, Uma and Jeemol Unni (2004), Unorganised and Organized Manufacturing in India: Potential for Employment Generating Growth, Economic and Political Weekly, Vol.XXXIX, No.41, pp. 4568-4580. October 9.

Rao, C.H. Hanumantha (1994), Economic Reforms and the Prospects for Rural Labour, The Indian Journal of Labour Economics, Vol. 37, No.1.

Samal Kishor, C. (1990), Linkages between Informal and formal Sector, Economic an Political Weekly, Vol XXV, No 8.pp. 1287.

Sharma R.K. and Abinash Dash (2006), Labour productivity in Small Scale Industries in India: A State-Wise Analysis, The Indian Journal of Labour Economics, 49 (3):407-427.

Sharma R.K. and Seema Bathla (2008), Regional Dimensions of Rural Industrialization in India: Patterns and Challenges, Paper Presented at a National Seminar on ‘Rural Industrialization in Punjab: Challenges and Strategies’ during March 4-5 2008 at Guru Nanak Dev University, Amritsar.

112

Shaw, A. (1990), Linkages of Large Scale, Small Scale and Informal Sector Industries: A Study of Thana Belapur, Economic and Political Weekly, Vol. XXV, No 8. pp. M17-M22. Feb. 17-24.

Singh Richa (2008), Economic Reforms and Well being of Rural Lbaour in India: An Inter Regional Analysis, Unpublished Ph.D. Thesis, Submitted to Jawaharlal Nehru University, Centre for the Study of Regional Development.

Srivastava, Ravi and Richa Sing (2005), Economic Reforms and Agricultural Wages in India, The Indian Journal of Labour Economics, Vol. 48, No.2, April June, pp. 407-424.

Sundaram, K. (2007), Employment and Poverty in India, 2000-2005, Economic and Political Weekly, Vol. XXXXII, No 30. pp.3121-3131. July 28-August 3.

Tendulkar, S. D. (2003), Organized Labour Market in India: Pre and Post Reform. Paper prepared for the Conference on Anti Poverty and Social Policy on January 2-4 at Alwar, India.

Thakur C.P. and C.S. Venkata Ratnam (2007), Conditions of Work in the Unorganized Sector in India, The Indian Journal of Labour Economics, (Conference Issue), Vol. 50 No. 4, pp. 749-774. October-December.

Topalova, Petia (2003), Trade Liberalization and Firm Productivity: the Case of India, Yale university, available on line at: www.econ.yale.edu/seminars/NEUDC03/topalova1.pdf.

Topalova, Petia (2005), Trade liberalization, poverty and inequality: Evidence from Indian Districts”’, NBER Working Paper 11614.

Uchikawa, S. (2001), Investment Boom and Underutilisation of Capacity in the 1990s, Economic and Political Weekly, August 25, 34: 3247-54.

Uchikawa, S. (2002), Investment Boom and the Capital Goods Industry, in S. Uchikawa (ed.), Economic Reforms and Industrial Structure in India, New Delhi: Manohar Publishers.

Unel, Bulent (2003), Productivity Trends in India’s Manufacturing Sectors in the last Two Decades, IMF Working Paper no. WP/03/22.

UNCTAD-India (2008), “How are the Poor Affected by Trade in India: An Empirical Approach”

Unni, Jeemol, Lalitha, N. and Rani, Uma (2000), “Economic Reforms and Productivity Trends in Indian Manufacturing,” Working Paper No.119, Gujarat Institute of Development Research (GIDR), Ahmedabad, India.

***

113

CHAPTER 3

Impact of Trade on Labour Markets in Unorganized Manufacturing: A State Level Analysis����

3.1 Introduction: Trade liberalization may lead to differential regional impacts in an economy depending upon the extent and nature of participation by different sections of the region and overall orientation towards trade of the region. India being a diverse economy, consisting of 28 states and 7 Union Territories, differs markedly with respect to the state level production patterns and state government policies. Further, a close examination of the existing unorganized manufacturing sector in different states shows that, there exist wide disparities in the type of enterprises, level of their activity and the labour market characteristics across states. The impact of trade liberalization of the economy on labour market in unorganized sector may therefore differ significantly across states. In this context, the chapter estimates the impact of export orientation of the state on employment, wages and labour productivity in unorganized manufacturing across states. There exists very limited literature on trade impacts at the state level for India and fewer on the unorganized sector. One major limitation to any such analysis is the non availability of trade data at the state level. Most of the studies that have attempted to undertake such an analysis for unorganized sector at the state level have compared the impacts in the pre and post trade liberalization period (Rani and Unni 2004, Raj and Duraisamy 2006). This chapter adds to the existing literature on impact of trade on labour markets by estimating the impact of trade openness of the location on labour markets. The analysis is undertaken at the enterprise level and it is analysed whether, irrespective of the trade-orientation of the industry to which the enterprise belongs, trade openness of the state in which the enterprise is located affects its employment, wages and productivity. A state level export orientation variable is constructed using the concordance matrix between six digit HS 2002 DGCI&S product level data and three-digit National Industrial Classification. The state wise export orientation is estimated by first constructing industry-level shares of exports in a state. The share of each three-digit level industry in a state in total exports is estimated by applying the ratio of industry’s output in the state to total output of the industry at All India level. This assumes that the inter-industry export pattern in a state will be similar to the existing All-India industrial export pattern. By

� This Chapter has been contributed by Rashmi Banga, Senior Economist, UNCTAD-

India project and Seema Bathla, Associate Professor, Center for Study on Regional

Development (CSRD), Jawaharlal Nehru University

114

summing the estimated exports of each industry in a state we arrive at the state level export orientation. The import orientation, on the other hand, may not differ significantly across states, though the production pattern of the state may influence the import pattern. Hence, state level import-orientation is estimated by applying the ratio of output of the industry in a state in total output of the industry at All India level to total imports of the industry. The import competition faced by an industry in a state will be higher if it has a larger share in production of the industry. To identify the states which where higher export orientation has had a significant impact on enterprises employment, wages and labour productivity, we undertake the estimations separately for each state. The literature on effects of trade initiated competition, either in domestic market or international markets, supports the view that export competition by creating new markets for labour intensive products in developing countries increases employment and productivity in the manufacturing sectors, while import competition may displace labour but may increase productivity. At the state level, this would translate to employment inducing effect of export-orientation of the state. Studies have shown that high backward and forward linkages exist between organized and unorganized sector. However, whether this effect percolates to the unorganized sector is an important empirical question. Further, whether these linkages differ regionally and if they do what are the factors that determine higher percolation of effects at the organized sector to unorganized sector may give valuable policy insights. Thus, apart from the trade-orientation of the industry to which an enterprise belongs, impact of the location of the enterprise needs further exploration. Given the unique characteristics of labour markets in India, i.e., downward wage rigidity and difficult firing policies, impacts of trade on labour markets have been found to differ with respect to wage rates in the organized manufacturing sector. Contrasting results have been arrived at by studies arguing from exports leading to lowering of labour share to export intensive industries paying more to labour34. For the unorganized sector, such labour market rigidities may not exist. An issue which needs more exploration is: what effects does flexible labour markets have on wage rates in the unorganized sector. This may have far reaching implications for assessing trade-poverty nexus in India at a more disaggregated level. The rest of the chapter is organized as follows: section 2 briefly reviews the limited literature on trade impacts on unorganized sector at the regional level; section 3 analysis the trends in labour markets across states in the unorganized sector; section 4 discusses the methodology, data sources and variables used; section 5 presents the empirical results with respect trade impacts on labour markets in the unorganized sector at the state level; section 6 concludes and draws policy directions.

34 For a detailed survey see Banga 2005 and Hashim and Banga 2008.

115

3.2. Review of Literature: Regional Impact of Trade Liberalisation

State level analysis of the impact of trade is severely limited by the availability of data on exports at the state level. For the first time, Economic Survey 2007-08, has brought out export shares of different shares for the years 2005-06 and 2006-07. Lack of earlier data on export orientation of the state has led to many studies making a comparative analysis of pre and post liberalization period. Unni, Lalitha and Rani (2000) compared the trends in growth and efficiency in the utilization of resources in both organized and unorganized manufacturing sectors at all India and Gujarat before and after the reform periods. They found the two sectors have done better in terms of growth in value added over a period of time in Gujarat. Rani and Unni (2004) found initial economic reform policies to have adversely affected employment in the organized and unorganized manufacturing sectors, which got improved in the subsequent years. However, the authors maintained that the reform measures had differential impact across various industry groups, in particular growth in automobiles and infrastructure enabled growth in their unorganized segment. Raj and Duraisamy (2006) analyzed productivity performance of the unorganized manufacturing in 13 major Indian states using a large scale National Sample Survey data for five periods, broadly representing the pre (1978-79, 1984-85, 1989-90) and post reform (1994-95 and 2000-01) periods. The analysis based on Malmquist productivity indexes showed tha, on an average, in all states except Rajasthan, the annual rate of total factor productivity growth has been higher in the reform period than in the pre-reform years, on an average. A better performance of unorganized manufacturing was due to progress made in technical efficiency rather than due to technological progress, which has been a major factor in achieving high levels of total factor productivity during the reference period. Topalova (2005) looks into the differential impact of liberalization on poverty and inequality across Indian districts. She compares the poverty and inequality measures in the districts with industries that experienced greater liberalization (measured as a reduction in tariff barriers) to those whose industries remained largely protected. In the baseline specification, the district level outcome of measures of poverty and inequality is taken as a function of the district exposure to international trade. The regression equation also includes district fixed effects which takes care of unobserved district-specific heterogeneity and year dummies to control for macroeconomic shocks that affect the whole country equally. She arrives at the conclusion that lower levels of tariffs have been associated with significantly high levels of poverty (measured as the Headcount ratio and Poverty gap) for rural India. For the urban sector however, no such significant relationship between trade liberalization and poverty was found. Similarly, she finds that there isn’t any statistically significant relationship between trade exposure and inequality (measured as the standard deviation of log consumption and the mean logarithmic deviation of consumption) either.

116

In a more recent study, Banga and Sharma (2008) examined trade-poverty nexus for organized manufacturing and for the unorganized agricultural sector at the state level in India. The impact of trade is seen on employment and wages of unskilled workers, which is taken as an indirect indicator of poverty in the paper. The analysis carried out at three digit industry level for 54 industries from 1998-99 to 2005-06 and for all products and separately for cereals, fruits-nuts, vegetables-roots-tubers and oilseeds from 1990-91 to 1999-2000 show a favourable impact of exports on wages in manufacturing. Import competition doesn’t seem to have displaced labour or adversely affected wages, which is explained by strict labour laws and downward rigidity of wages in India. For the agricultural sector, the results find no evidence of rise in wages of unskilled labour in the states having high shares of exports but imports have caused lowering of wages. The paper concludes that the positive impact of trade on wages and employment depends on the extent to which the poor are able to gainfully participate in the expanding sectors. Marjit and Kar (2008) have analyzed the impact of trade reforms on income and poverty at the regional level. The authors devised a regional openness index as a measure of country’s trade liberalization and develop a ranking of the Indian states according to their exposure to international trade. Based on ASI data on gross value added of each industry including agro based industries from 1980-81 to 2000-02, the authors have estimated share of value added contributed by each industrial group for major 15 states. The study found that states with relatively high levels of income are also the states with greater exposure to trade and such relationship has grown stronger over time. In the process, the authors reconfirm general theoretical intuition that exporting states are getting richer over the years and the import competing states are falling behind. Also, a state generates higher per capita income (net state domestic product) by switching production from import competing sectors to the export sectors. Secondly, across various industries trade openness has led to significant decline in the rate of growth of employees and workers, where the negative impact on workers is stronger compared to their white collared counterparts. The skill based technological changes and the overwhelming contribution of service sector in the country’s GDP have unambiguously led to an increased demand for skilled workforce, unlike in the industrial sector in general, where exposure to foreign competition has reduced (slow/negative growth) employment opportunities significantly. Thirdly, there has been a lower incidence and depth of poverty for urban locations as compared to rural areas, much to the conformity of other independent studies on economic reforms and poverty. Furthermore, there is clear evidence in support of increasing inequality in the rural areas when the level of inequality is correlated with the state level openness index. In short, three observations can be made. One, the impact of trade on this sector is judged mainly by comparing the growth performance in the pre and post reform period without looking at actual data on changes in volume of exports and imports as has been done in the case of organized sector. Two, unlike organized manufacturing, little work has been done at the disaggregate level on unorganized manufacturing in the context of trade

117

liberalization. Recognizing large heterogeneity in employment, enterprises and output in this sector, it may be more meaningful to undertake analysis at the state or industry levels or by rural-urban location. Three, more studies need to be initiated on this sector in view of (a) dearth of studies on the response of unorganized manufacturing sector to trade and industrial reforms (b) an increasing potential of this sector in absorbing the burgeoning workforce (accounting for nearly three fourth of manufacturing workforce) compared to organized manufacturing, which is marked by an almost stagnancy in employment growth, (c) greater presence in the rural areas of some states characterized by low productivity levels, capital and technology and the effect of trade in enhancing/attenuating its competitiveness, (d) a positive rate of growth in workers, enterprises, capital and output witnessed upto the year 2000-01 is followed by sluggishness during 2005-06, (e) to investigate the probable factors (could be internal or related to inflow of imports) behind deceleration in the growth of workers, enterprises and output and its implications for the performance of this sector. In the face of these, an attempt is made in this paper to examine the impact of exports and imports on employment, wage rate and labour productivity in unorganized sector at the state level. The trend analysis is undertaken for two NSSO rounds viz. 56th (2000/01) and 62nd (2005/06) so as to decipher the status and performance of unorganized manufacturing in each of the Indian states and explore the potential therein to realize gains from trade liberalisation or ways to mitigate its adverse impact, if any. The empirical estimations are based on NSSO 62nd for the year 2005-06 using data for altogether a total of 82,897 unorganised manufacturing enterprises. 3.3 Trends in Labour Markets in Unorganised Sector at the State Level 1. the NSS survey further validates the applicability of both ‘pull and push factors’

across the Indian states in the unorganized manufacturing. As shown in Table 5, states like Assam, Bihar, Madhya Pradesh, Jharkhand, Chattisgarh, Orissa and Uttar Pradesh, have shown improvement in the value of GVA per worker over a period of time. But these states continue to fall in the lowest category of labour productivity i.e. less than Rs. 10 thousand and between Rs. 10-15 thousand per annum respectively.

Of all the states, an improvement in labour productivity in rural manufacturing is discernible mainly in Haryana, J&K, Karnataka, Maharashtra, Tamil Nadu and Uttranchal. States like Assam, Bihar, Jharkhand, Himachal Pradesh, Madhya 3.1

3.3.1Description of Data

Enterprise level data on unorganized manufacturing is extracted from 56th and 62nd rounds of NSS surveys, which are based on industry/enterprise level. The term

‘unorganised manufacture’ under the coverage of 62nd

round basically referred to all

manufacturing enterprises, which were not covered by ASI. All government and public sector undertakings were also outside the coverage of the survey. The quinquenial NSS surveys provide information at both industry level as well as the state level. In terms of

118

National Industrial Classification (NIC) 2004 codes, the 62nd round survey on unorganized manufacture covered the NIC 2-digit codes 15-37. In addition, enterprises engaged in cotton ginning, cleaning and baling (NIC 2004 code 01405) were also covered under the survey. The survey covered the whole of the Indian Union35. A total of 82,897 enterprises were covered by the 62nd Round of NSS. About 51% of these enterprises were in rural India and the remaining 49% were in urban India. Maximum number of enterprises belonged to Uttar Pradesh (9.95%) followed by West Bengal (7.97%), Tamil Nadu (7.96%), Andhra Pradesh (7.93%) and Maharashtra (7.76%). State wise number of different kinds of enterprises (OAME, NDME and DME)

36 in rural / urban sectors are reported in Appendix I. Table I of Appendix II provides summary of key variables in the unorganized manufacturing sector at all India and separately for rural and urban areas during 56th round (2000-01) and 62nd round (2005-06). During 2000-01, there were 17.02 million unorganized manufacturing enterprises (11.9 million in rural and 5.08 in the urban areas) providing employment to 37 million people, of which nearly 23.98 million fall in the rural areas and 13.09 million in the urban areas respectively. While there has been a marginal increase in the number of enterprises to 12.13 million in rural and decrease to 4.94 million in urban during 2005-06, the number of workers has fallen in both to 23.45 million in rural and 12.98 in urban manufacturing units. Clearly, OAME occupies an overwhelming share (80-90%) in both enterprises and work force, particularly in the rural areas. While the share of OAME, NDME and DME in total rural enterprises has nearly remained unchanged during 2000 to 2005, their share in total workers has changed. In rural manufacturing, it has declined from 79.83% to 76.82 % in the OAME and has improved in NDME and DME by 2 percentage points from 8.06 to 10.16 percent and 12.11 to 13.01 percent respectively. The same in the urban manufacturing shows OAME, NDME and DME having 70.9, 21.3 and 7.9 percent shares in total enterprises in 2000/01 which remained almost same during 2005-06. But, there has been a nearly 2-3 percentage point decrease in the share of workers in the OAME and an increase in DME over a period of five years. Further, 79.83 percent of workers in rural OAME contributed 63.05 percent to value added during 2000, the share of which declined to nearly 50% during 2005/06. In

35 except (i) Leh and Kargil districts of Jammu & Kashmir, (ii) interior village of Nagaland situated beyond

five kilometres of bus route and (iii) villages of Andaman and Nicobar Islands which remain inaccessible throughout the year. All the sample FSUs of the districts Poonch and Rajouri of the state of Jammu and Kashmir became casualty. Thus, the estimates for Jammu and Kashmir as well as for all-India do not include these areas. 36 As per NSS, Own Account Manufacturing Enterprise (OAME) is one, which runs without any hired

worker employed on a fairly regular basis and is engaged in manufacturing and/or repairing activities. An establishment employing less than six workers (household and hired workers taken together) and engaged in manufacturing activities is termed as Non-directory Manufacturing Establishment (NDME). Finally, a Directory Manufacturing Establishment (DME) is one which has employed six or more workers (household and hired workers taken together) and is engaged in manufacturing activities.

119

contrast, NDME and DMEs with 10 and 13 percent of total workforce contributed 17 and 33 percent to gross value added, and their share in total GVA has shown a consistent improvement over time. In urban manufacturing too, DME has progressed as its share in workers and GVA accelerated from 27.13 to 30.21 and 40.34 to 51.85 percent respectively during 2000-05. The state level analysis is carried out at 2 digit level separately for the rural and urban manufacturing focusing mainly on the key performance indicators. In all, data is analyzed for 20 major states, of which Andhra Pradesh, Bihar, Gujarat, Karnataka, Madhya Pradesh, Maharashtra, Orissa, Tamilnadu, Uttar Pradesh and West Bengal occupy the highest share in total enterprises in the unorganized manufacturing sector during the two rounds of NSSO survey under study. Within the rural sector, the highest share in total enterprises during 2005/06 is held by West Bengal and Uttar Pradesh at 18.34% and 14.05%. This is followed by Andhra Pradesh, Orissa and Tamilnadu at 8.95, 7.18 and 7.01% respectively. Bihar, Jharkhand, Karnataka, Kerala, Madhya Pradesh and Maharashtra’s share in enterprises hovers between 4-5%. In the urban manufacturing, Andhra Pradesh, Maharashtra, Uttar Pradesh, West Bengal, Tamilnadu have the highest share in total enterprises ranging between 9-13%, aggregating 55% share. Other states that hold a significant share include Gujarat and Karnataka at nearly 7% each, Madhya Pradesh and Rajasthan at nearly 5% each. These states occupy a high share in total workers. From 2000 to 2005, there is hardly any change visible in the share of these states in enterprises as well as workers. Data further shows that with some exceptions, the larger the state higher is its share in enterprises and employment. Therefore, to take care of size effect, the percentage share of each state in rural-urban population is also presented for the year 2001. It reveals that Andhra Pradesh, Bihar, Gujarat, Karnataka, Madhya Pradesh, Maharashtra, Orissa, Rajasthan, Tamilnadu, Uttar Pradesh and West Bengal, together have a nearly 75% share in enterprises in rural and urban manufacturing as compared to their respective shares in total population. In many states, the percentage of enterprises set up in rural and urban areas is more than the percentage share of population these states inhabit. This indicates that some of these states are performing better than others in terms of putting up enterprises and employing people more than proportionately to their population. However, not all the states have experienced a positive rate of growth of enterprises during 2000/01 to 20005/06, especially in their rural manufacturing. A negative growth in rural enterprises is observed in Andhra Pradesh, Bihar, Chattisgarh, Karnataka, Maharashtra, Orissa, Punjab and Uttranchal and in urban enterprises in Assam, Bihar, Chattisgarh. Orissa, Himachal Pradesh, J&K, Karnataka, Maharashtra, Punjab, Tamilnadu, Uttranchal, Uttar Pradesh and West Bengal. Following is a detailed analysis of inter-state trends in this sector. For simplicity, inter state trends are analyzed only for the key variables, which broadly indicate the performance of this sector. The variables considered are, number of enterprises and workers and employment per

120

enterprise to represent size; wages/emoluments per hired worker to assess the status of workers; gross value added (GVA) per enterprise to represent both size as well as viability of the unit; GVA per worker also called labour productivity to represent state of health of a unit; market value of fixed assets (owned+hired) per worker and per enterprise and loan outstanding per enterprise to reflect capital or asset base of an enterprise;. Results are presented in Appendix Tables II and V. 3.3.2 Trends in Employment, Wages, Labour Productivity and other performance Indicators at State Level 1. Large inter-state discrepancies in the growth patterns in the unorganized

manufacturing sector are observed. Most of the states not only provide employment more than proportionately to their respective share of population in total but also have more enterprises in proportion to their population.

2. A larger number of unorganized enterprises and workers proportionately to their share in population are identified mainly in Andhra Pradesh, Bihar, Jharkhand, Karnataka, Kerala, Madhya Pradesh, Maharashtra, Orissa, Tamilnadu, Uttar Pradesh and West Bengal. Most of these states have (a) agriculture as their main occupation (b) low economic growth and per capita income except in Andhra Pradesh, Karnataka, Gujarat and Maharashtra, and (c) high incidence of rural poverty.

3. A fall in the labour force or to say low labour demand witnessed during 2000/01-2005/06 may be attributable to (i) closure of enterprises as discernible from negative growth in OAME or their upgrading to NDME and DME (ii) lowering of income/profits due to negative growth in real output and hence their likely liquidation (iii) low labour productivity in OAME and its negative rate of growth in rural manufacturing (iv) selective withdrawal of workers more so from non-agro having better skills to NDME/DME or other economic sectors37 for relatively higher wages, and (v) improvement in output per worker as visible in NDME and DME and hence an unchanged or lesser labour demand. There could also be other reasons behind a decline in employment growth as pointed out by Kundu, Lalitha and Arora (2001) such as subcontracting of jobs to smaller units and underreporting of employment, classification of units under the tertiary sector.

4. Of all the states, Gujarat, Maharashtra, TamilNadu and Uttar Pradesh hold the highest share in total emoluments given to the hired workers to the tune of 11, 18.98, 11 and 13% respectively. This is followed by West Bengal at 8.81%, Karnataka, Kerala and Rajasthan at around 5% each. With only exception of Karnataka, Kerala, Tamilnadu and West Bengal, in all other states, urban share in emoluments is higher than the respective rural shares.

5. Overall, in nearly all the states, the performance of rural manufacturing has consistently been way behind the urban manufacturing, which apart from greater dominance of OAME may be linked to inadequate capital, infrastructure, technology and skills.

6. Apparently, the hypothesis of an inverse relationship between economic growth and non-farm jobs as argued in the literature applies to the unorganized manufacturing

37 The latest NSS data on employment and unemployment shows acceleration in workforce in tertiary and construction activities during 2004-05.

121

sector in these states except for Andhra Pradesh, Karnataka, Gujarat and Maharashtra. In these four states, agriculture may generate surplus which is diverted to non-farm activities, could be in the unorganized manufacturing sector, hence pointing towards the occurrence of ‘push factors’.

7. A closer look at absolute value of labour productivity during the last two rounds of Pradesh, Orissa and West Bengal have failed to keep pace in their value added per worker and these are the states which have low per capita income and high incidence of poverty.

8. In the urban sector, again, Chattisgarh, Andhra Pradesh, Madhya Pradesh, Bihar, Orissa have failed to keep pace while significant improvements in labour productivity Of the total 20 states, Uttar Pradesh, Uttaranchal, Jharkhand and Assam are exceptions which have experienced not only a relatively higher incidence of urban poverty but also a higher level of labour productivity in their urban unorganized manufacturing.

9. The above analysis also indicates that a fall in employment in this sector as observed from 2000/01 to 2005/06 may not be that grave in high per capita income states having relatively high levels of labour productivity and alternative job avenues. But it could be a catastrophe in low income states where labour productivity has hardly improved, meager income opportunities exist and incidence of poverty is also high.

3.4. Methodology and Variables 3.4.1 Impact on Labour Demand To undertake the analysis of impact of trade on labour markets in the unorganized sector, estimates have been undertaken using NSS 62nd Round data for the year 2005-06. The estimation is carried out at enterprise level which belongs to 3-digit industries. The impact of export-orientation of the industry to which the enterprise belongs on wages and employment of the enterprise is estimated. Further, the impact of trade-orientation of the state to which the enterprise belongs on its wages and employment has been tested. To identify the states where the impact of trade orientation is higher, state specific estimations have been carried out To identify the factors that may impact on employment and wage-rate a CES production function that embodies labour augmenting technological progress is considered

Q = χ [ s(k)-ρ + (1-s) (L eλt)- e] -1/e

Where Q is output, K is capital, χ > 0 & 0< s< 1,

Elasticity of substitution is given by [e/(1+e)] & technical progress is labour augmenting

at rate of λt. Marginal productivity conditions gives labours demands function (first order conditions , MPL = real wage rate, assuming mark – up is constant ) taking logs and rearranging we get: Ln ( L) = ln (Q) – 1 / (1+e) ln (w/p) - ρ / (1+ρ) λt + [ 1/1+ e ln [ 1- s/β ] - ρ / (1+ ρ) ln (χ) ]

= ln(Q) - σ ln (w/p) – (1- σ) λt + [σ ln [ 1- s/β] – (1- σ) ln (χ) ]

( where σ = 1/ 1+e )

λt is taken as exogenous technical change which may occur because of trade i.e.,

122

λt = f (Exports, Imports) the labour demand equation therefore becomes:

Ln (L) it = β0 + β1 ln (Q) it + β2 ln (W/P) it - β3 EXPORTS it + β4 IMPORTS it + αi + e it

However, apart from these variables that may affect employment of an enterprise, the export-orientation of the industry to which the enterprise belong may impact the employment of the enterprise. Further, it is argued that the location of the enterprise may also affect the employment. After controlling for the size of the enterprise and the wage rate, enterprises located in states which are trade oriented may have higher demand for labour because of the technical progress which is labour augmenting. Higher linkages between organised and unorgainsed sectors in export-oriented states may also be a plausible reason for this. Import competition, on the other hand, may adversely impact the employment demand in the unorganized sector depending on the state’s share in total domestic production. The equation estimated is therefore:

ln ln( ) ln( ) ln( ) ln(expint int )

................(1)

f f f f f f f i f i i

f f

emp GVA wrate stateor or imp

StateDum

θ α β γ δ

ι ε

= + + + + +

+

where, empf = total employment in enterprise f GVA f = Gross Value Added by enterprise f wrate f = wage rate in enterprise f stateor f = states’ export orientation to which the enterprise belongs expinti = export intensity of the industry to which the enterprise belongs impinti = import intensity of the industry to which the enterprise belongs StateDum = state dummies to capture the other state specific effects

3.4.2 Impact on Wage Rate In a competitive labour market, firms will hire workers until MCL (which is wage rate) equals MR (which is Price x MPL). Labour demand is given by MPL. Assuming that labour supply is a function of the average wage rate we get:

Ls = β (w/p) r

Log (Ls) = b0 + r ln(w/p) Equating labour supply to labour demand we get: wage rate therefore as a function of: (w/p) it = F [ Q it , LP it , EXPORTS it , IMPORTS it , Time, Fixed Effects] However, in the empirical framework along with the above factors we need to include other potential demand shifters, which may also control for industry-specific effects. This is justified by arguing that merely including the factors derived from theory may not capture other influences, which could effect an industry’s demand function (Driffield and Taylor 2000). We therefore control for inter-industry variations by including capital-labour ratios.

123

At the enterprise level, we estimate the following equation which incorporates the export intensity of the industry to which the enterprise belongs and the trade orientation of the states.

' '

' '

ln ln( ) ln( ) ln( ) ln( )

ln(expint int ) ................(2)

f f f f f f f f f i

f i i f f

wrate GVA j lp kl stateor

or imp StateDum

ϕ α ρ γ

δ ι ε

= + + + + +

+ +

where, empf = total employment in enterprise f GVA f = Gross Value Added by enterprise f wrate f = wage rate in enterprise f stateor f = states’ export orientation to which the enterprise belongs expinti = export intensity of the industry to which the enterprise belongs impinti = import intensity of the industry to which the enterprise belongs StateDum = state dummies to capture the other state specific effects lp = labour productivity

kl = capital-labour ratio

3.4.3 Impact on labour Productivity For modeling labour productivity we consider CES production function with CRS

Q = χ [ s(k)-ρ + (1-s) (L eλt)- e] -1/e

Kmenta provides an approximation of CES function by expanding ln Q in Taylor’s series around

ρ = 0, i.e.,

ln (Q) = ln (χ ) + eλt ln (k/L)+ ln (L) + ( ρ eλt (1-s)/ 2) [ ln (k/L) ]2

ln (Q) - ln(L) = ln (χ ) + eλt ln (k/L) + ( ρ eλt (1-s)/ 2) [ ln (k/L) ]2

ln (Q/L) it = ln (χ ) it + eλt ln (k/L) it + ( ρ eλt (1-s)/ 2) [ ln (k/L) ]2 it

(Cobb- Douglas function assumes e = 0)

Labour productivity can therefore be a function of

(Q/L) it = F [(K/L) it , Q it , ,EXPORTS it ,IMPORTS it, Time, Fixed Effects]…(1)

The equation estimated at the enterprise level is as follows:

'' ' '' ''

''

ln n( ) ln( ) ln( ) ln(expint int )

................(3)

f f f f f f f i f i i

f f

lp l GVA kl stateor or imp

StateDum

θ α ρ γ δ

ι ξ

= + + + + +

+

where, empf = total employment in enterprise f GVA f = Gross Value Added by enterprise f wrate f = wage rate in enterprise f stateor f = states’ export orientation to which the enterprise belongs expinti = export intensity of the industry to which the enterprise belongs

124

impinti = import intensity of the industry to which the enterprise belongs StateDum = state dummies to capture the other state specific effects lp f = labour productivity

kl f = capital-labour ratio Variables Estimated

1. Export Intensity of the Industry to which the enterprise belongs: Export intensity is calculated as the percentage of the exports of each industry in the total output of that industry. To arrive at the industry level data on exports and imports, a concordance matrix between six digit product level data on HS 2002 codes from DGI&S and three-digit NIC 98 has been constructed. Industrial output is derived from Annual Survey of Industries for the year 2005-06 and industry level export and import intensities are estimated.

2. State Export Orientation:

The state wise export orientation is estimated by first constructing industry-level shares of exports in a state. The share of each three-digit level industry in a state in total exports is estimated by multiplying exports of the industry with the ratio of industry’s output in the state to total output of the industry at All India level. This assumes that the inter-industry export pattern in a state will be similar to the existing All-India industrial export pattern. By summing the estimated exports of each industry in a state we arrive at the state level export orientation.

The variable has been calculated as follows: State Export Orientation =

∑'

exp of '

State s output in Industry iorts Industry i

India s Total output in Industry i×

The import orientation, on the other hand, may not differ significantly across states, though the production pattern of the state may influence the import pattern. Hence, state level import-orientation is estimated by applying the ratio of output of the industry in a state in total output of the industry at All India level to total imports of the industry. The import competition faced by an industry in a state will be higher if it has a larger share in production of the industry. Using the above formulas, we estimate the export and import orientation of the states for the year 2005-06. Data from Annual Survey of Industries is used to estimate the output of the industries and concordance matrix, as discussed above, has been used to arrive at corresponding export figures for each three digit level industry in a state Table 1 column (1) reports the shares of top ten states in descending order according to our estimates of export orientation of the states. Column (2) of the Table reports the shares of reported by

125

Economic Survey 2007-08 using the actual export data at the state level, which is compiled for the first time. According to the Economic Survey 2007-08, states like Maharashtra, Gujarat, Tamil Nadu, Karnataka, Andhra Pradesh, Haryana and Uttar Pradesh are some of the states with the highest share of exports in total. From our calculations, these are also the states which are more export oriented, measured by the aggregate of the industry shares in total exports of the industries in each state (Table 1). The estimated shares and ranking of the top ten states arrived by estimating the state export orientation is quite close.

Table 3.1 Comparing Export Orientation with Shares in Total Exports of States

Export Orientation of States (1)-2005-06

Share in Total Exports (2006-07) (2)

STATES (our estimates) STATES (Economic Survey)

Gujarat 23.69 Maharashtra 28.6

Tamil Nadu 12.31 Gujarat 19.2

Maharashtra 11.62 Tamil Nadu 10.4

Karnataka 11.42 Karnataka 10

Uttar Pradesh 6.16 Andhra Pradesh 4.3

Haryana 4.24 West Bengal 3.2

Andhra Pradesh 4.2 Haryana 3

West Bengal 3.97 Uttar Pradesh 2.9

Rajasthan 3.75 Rajasthan 2.7

3. Size of the enterprise:

The data on the Gross Value Added by enterprises has been taken as a proxy for the size of the enterprise.

4. State and Industry Dummies:

These are included to control for other factors specific to each state and industry.

3.5. Empirical Results

The effect of trade on employment, wage rates and labour productivity has been estimated using the 62nd Round of NSSO. Since the data is for the year 2005-06 OLS estimates are carried out using state and industry dummies to control for unobservable state and industry specific effects. These results are discussed below.

126

3.5.1 Empirical Results of Impact of Trade Orientation of States on Employment

Table 2 presents the results of the regression of log of employment on the various independent variables explained in Equation 1. Model 1 takes export intensity of the industry to which the enterprise belongs as an independent variable along with GVA, wage rate, state export and import orientation and dummies.

Table 3.2: Regression Results for Employment as Dependent Variable

Model 1 Model 2

Independent Variables Coefficient t-stat Coefficient t-stat

Constant -1.25*** -39.93 -1.26*** -37.81

Log GVA 0.48*** 77.43 0.48*** 77.31

Ln wage rate -0.26*** -33.86 -0.26*** -33.9

Log state export orientation 0.067*** 5.24 0.06*** 5.16

Log industry export intensity 0.002*** 2.17

Log state import intensity 0.0007 0.6

State Dummy Yes Yes

Industry dummy Yes Yes

R-Squared 0.64 0.65

No. of observations 28587 28413

Note *** indicate significant at 1% level of significance. The results show that the coefficient of export intensity of the industry is positive and significant implying that enterprises belonging to the industries which have higher export intensity have experienced a rise in employment. A more striking result is that the coefficient of state’s export orientation is positive and highly significant. It signifies that the location of the enterprise in terms of state in which it operates has a strong influence, irrespective of the fact whether the enterprise belongs to the export oriented industry or not. Enterprises belonging to the state with higher export orientation have higher employment. The size of the firm is positively related to employment implying that the higher the firm’s GVA, higher is the employment level. Wage rate has a negative relation with employment, consistent with the theory. As wages rise, the cost to the firm goes up, therefore, employment is lowered. The goodness of the fit of the regression denoted by the R-Squared is quite high at 0.64, thus, the explanatory power of the independent variables included in the regression is high. Model 2 includes import intensity of the state as explanatory variables. The coefficient of import intensity is positive but not statistically significant. This implies that import competition faced by the state has not affected the employment in the enterprises in the unorganized sector. However, the inclusion of this variable has not changed the results for the other variables.

127

3.5.2 Impact on Wage Rates

Table 3.3 presents the results for regression with wage rate as the independent variable (Equation 2). The results show that size of the enterprises, labour productivity and capital-labour ratio has a significant positive effect on the wage rate in conformity with economic theory.

Table 3.3: Regression Results for Wage Rate as the Dependent Variable

Enterprise Type 2 Enterprise Type 3

Independent Variables Coefficient t-stat Coefficient t-stat

Constant -0.35 -0.73 0.04 0.21

Log GVA 0.45*** 18.13 0.19*** 14.08

Log Labour productivity 0.34*** 11.13 0.61*** 34.96

Log Export Intensity of the Industry 0.25 1.04 0.14*** 2.73

Log of Import Intensity of the State -0.24 -1.06 -0.12*** -2.46

Log of State Export Orientation 0.0004 0.16 0.01*** 2.88

State Dummy Yes Yes

Industry Dummy Yes Yes

R-Squared 0.65 0.86

No. of Observations 16877 9193 Enterprise 2 is NDME Enterprise 3 is DME

The wage rate equation has been estimated for NDME and DME enterprises only as the OAME enterprises do not really pay wages. Interestingly, export intensity of the industry to which an enterprise belongs has a significant positive effect on wages of DME. This implies that enterprises belonging to export oriented industries are paying higher wage rates. Export orientation o the state significantly influences the wage rates in the unorganized sector. Enterprises in states with higher export orientation pay highr age rates in the unorganized sector. Import competition, on the other hand, lowers the wage rate paid by the DMEs enterprises. The flexible wage norms, with minimum wage laws not strictly being implemented in the unorganized sector may be a reason for the lowering of wage rates with rising import competition in case of DMEs.

3.5.3 Impact on Labour Productivity

Table 3.4 contains the results of the regression with Labour Productivity as the dependent variable as in Equation 3.

Table 3.4: Regression Results for Labour Productivity as a Dependent Variable

Dependent Variable: lnlp Model 1 Model 2

Independent Variables Coefficient t-stat Coefficient t-stat

128

Constant 1.32*** 63.49 0.99*** 35.26

Log GVA 0.63*** 462.55 0.65*** 453.81

Log capital labour ratio 0.09*** 64.82 0.09*** 56.81

Log export intensity of industry 0.01*** 20.13

Log import intensity of state -0.03*** -12.58

Log state export orientation -0.01 -0.74 0.002 0.21

State Dummy Yes Yes

Industry Dummy Yes Yes

R-Squared 0.87 0.89

No. of Observations 81680 81136

It is found that the size of the firm and its capital-labour ratio have positive and very significant (Model 1) effect on labour productivity in unorganised manufacturing firms. The export intensity of the industry to which the enterprise belongs has a significant impact on labour productivity. However, the export orientation of the state to which the enterprise belongs does not have any significant impact. In other words, what affects the labour productivity of enterprises is whether they belong to the export oriented industries or not. The location of the enterprises does not affect its labour productivity levels. Model 2 includes import intensity of the state as an explanatory variable. Interestingly, the results show that higher the import competition in the state to which an enterprise belongs lowers will be its labour productivity. The results have not changed much for the other variables. The size of the firm and its capital-labour ratio continue to have a positive and very significant effect and state orientation measure is insignificant. It is seen from the above discussion of the results, that export intensity of the industry to which the enterprise belongs has a very significant positive effect on employment, wages and labour productivity of the firm. However, apart from the export intensity of the industry, what also matters is the export orientation of the state in which the enterprise is located. This indicates that two enterprises belonging to the same industry may have differential employment, wage rate and labour productivity if they are located in different states. In other words, enterprises located in states which have higher trade exposure are likely to be affected more by trade. To identify the states where trade has higher effects and corroborate the results we undertake state specific estimates. 3.5.4 State Specific Empirical Results The impact on employment of export and import intensity of industry to which the enterprise belongs is undertaken separately for each state. The result of the states where export intensity of the industry has a significant impact on employment in the unorganized sector is reported. States for which the results are not reported are the states where higher export-orientation does not have any significant impact on employment in the unorganized sector.

129

Employment is regressed on GVA, wage rates and export intensity of the industry for different states individually. Table 3.5 shows the result for the states where export intensity of the industry raises employment of the enterprises.

Maharashtra Andhra Pradesh Karnataka Tamil Nadu

Coefficient t-stat Coefficient t-stat Coefficient t-stat Coefficient t-stat

Variables

constant -0.77*** -15.44 -2.02*** -19.66 -1.37*** -10.03 -1.48*** -18.93

lngva 0.4*** 18.1 0.46*** 21.03 0.48*** 12.57 0.61*** 39.98

lnwrate -0.23*** -8.57 -0.16*** -7.51 -0.28*** -6.96 -0.39*** -20.81

expint 0.009*** 2.91 0.02*** 3.15 0.02*** 3.63 0.003** 1.99

R-Squared 0.60 0.58 0.60 0.67

No. of Variables 2606 1811 962 2781

Table 3.5 shows that states like Punjab, Haryana, Gujarat, Maharashtra, Andhra Pradesh, Karnataka and Tamil Nadu are the states where export intensity of the industry to which the enterprise belongs has increased the employment of the enterprises in the unorganized sector. The other variables, viz., GVA and Wage Rate also have the expected sign with a high level of significance. According to the Economic Survey 2007-08, states like Maharashtra, Gujarat, Tamil Nadu, Karnataka, Andhra Pradesh, Haryana and Uttar Pradesh are some of the states with the highest share of exports in total. 3.6. Summary and Conclusions The unorganised sector has emerged as an important part of the Indian economy. This sector is bigger than the organised sector in many key respects in spite of the larger control over resources and socio-economic power enjoyed by the latter. Bulk of India’s workforce is unorganised in nature. While almost the entire farm sector can be characterized as informal, roughly 80 per cent of the workforce in the non-farm sector is informal. The unorganised sector, consisting mainly of small economic entities, is not only larger than the organised sector but also fairly diversified and differentiated

Table 3.5: State-wise Regression Results with Employment as the Dependent Variable continued.

Punjab Haryana Gujarat

Coefficient t-stat Coefficient t-stat Coefficient t-stat

Variables

constant -1.29*** -13.94 -1.32*** -13.42 -0.99*** -9.44

lngva 0.36*** 14.43 0.51*** 18.94 0.67*** 21.24

lnwrate -0.11*** -3.73 -0.29*** -8.85 -0.5*** -13.66

expint 0.01** 2.24 0.01** 2.03 0.04*** 5.39

R-Squared 0.60 0.68 0.57

No. of Observations 1181 1103 1190

130

sector of the Indian economy in terms of relative share in GDP and work force. The Task Force constituted by the National Commission of Enterprises in Unorganised Sector (NCEUS) in its report on the Contribution of the Unorganised sector to GDP (June 2008) reported that the unorganised sector contributed as much as 50% to GDP in 2004-05. Given the significance of the unorganized sector, especially in terms of its share in total employment, any analysis of trade impacts on an economy without assessing the impact on this sector gives an incomplete picture. This chapter estimates the impact of trade on employment, wage rate and labour productivity in the unorganized manufacturing industries using enterprise level data from 62nd Round of NSS for the year 2005-06. The analysis emphasizes the significance of the locational effects of trade by estimating the impact of export and import orientation of states and industries to which the enterprise belongs. Overall, the results show that:

♦ Export intensity of the industry in the organized sector increases employment of the enterprises in the unorganized sector.

♦ Enterprises in the unorganized sector belonging to export-oriented industries pay higher wage rates

♦ Labour productivity of enterprises in the unorganized sector increases if the industry to which they belong is export-oriented.

♦ Location of the enterprise is an important determinant of whether trade impacts get percolated to the unorganized sector.

♦ States where exports have favourable affected the employment in the unorganized sector are Punjab, Haryana, Gujarat, Maharashtra, Andhra Pradesh, Karnataka and Tamil Nadu.

♦ Import competition faced by the state to which an enterprise belongs has no significant effect on employment or wage rate of the enterprise but lowers the labour productivity.

. An important conclusion that emerges from the analysis is that there exist inter-regional disparities with respect to impact of trade on employment, wage rate and labour productivity in unorganized sector. While higher export orientation of the industry to which the enterprise belong leads to higher employment but this may not be so in all states. Trade orientation of the state may play a vital role in distributing the gains of overall trade in the economy. Further, in order to percolate the gains of trade to the poor, through forward and backward linkages of unorganized sector to organized sector may play an important role.

131

References Banga, Rashmi (2005a), Impact of Liberalization on Wages And Employment in Indian

Manufacturing Industries, Working Paper No. 153, Indian Council for Research on International Economic Relations (ICRIER).

Banga, Rashmi (2005b), Liberalization and Wage Inequality in India, Working Paper No.

156, Indian Council for Research on International Economic Relations (ICRIER). Gera, S. and Masse, P. (1996), Employment Performance in the Knowledge Based Economy, Industry Canada Working Paper No. 14, Human resource Development, Canada W-97-9E/F, Canada. Goldar, Biswanath (2002),‘Trade Liberalization and Manufacturing Employment: The

Case India’, ILO 2002/34. Goldar, Bishwanath and Anita Kumari (2003), Import Liberalization and Productivity

Growth in Indian Manufacturing Industries in the 1990s, Developing Economies, December, 41(4): 436-60.

Goldar, Bishwanath and Suresh Chand Aggarwal (2004), Trade Liberalization and Price-

Cost Margin in Indian Industries, Working Paper no. 130, Indian Council for Research on International Economic Relations, New Delhi.

Government of India, Unorganised Manufacturing Sector in India, NSS 51st round July

1994 - June 1995, NSS 56th Round (July 2000-June 2001), NSS 62nd Round (July 2005-June 2006),National Sample Survey Organization, Department of Statistics, New Delhi.

Kundu, Amitabh, N. Lalitha and Sant Lal Arora (2001), Growth Dynamics of Informal

Sector in Urban India: An Analysis of Interdependencies, in Kundu, Amitabh and Alakh N. Sharma (ed.), Informal Sector in India-Perspectives and Policies, Institute for Human Development, and Institute of Applied Manpower Research, New Delhi.

Kundu Amitabh, Niranjan Sarangi and Bal Paritosh Das (2005), Economic Growth,

Poverty and Non-farm Employment: An Analysis of Rural-Urban Linkages, in Rohini Nayyar and Alakh N. Sharma (ed), Rural Transformation in India: The Role of Non-farm Sector, Institute for Human development, New Delhi, 2005.

Marjit Sugata and Saibal Kar (2008), Regional Trade Openness Index, Income Disparity

and Poverty- An Experiment with Indian Data, CUTS International, 2008, Jaipur.

132

Raj S N Rajesh and Malathy Duraisamy (2006), Economic Reforms, Efficiency Change and Productivity Growth: An Inter-State Analysis of Indian Unorganised Manufacturing Sector.

Rani, Uma and Jeemol Unni (2004), Unorganised and Organized Manufacturing in India:

Potential for Employment Generating Growth, Economic and Political Weekly, Vol.XXXIX, No.41, pp. 4568-4580. October 9.

Topalova, Petia (2003), Trade Liberalization and Firm Productivity: the Case of India,

Yale university, available on line at: www.econ.yale.edu/seminars/NEUDC03/topalova1.pdf.

Topalova, Petia (2005), Trade liberalization, poverty and inequality: Evidence from

Indian Districts”’, NBER Working Paper 11614. Uchikawa, S. (2001), Investment Boom and Underutilisation of Capacity in the 1990s,

Economic and Political Weekly, August 25, 34: 3247-54. Uchikawa, S. (2002), Investment Boom and the Capital Goods Industry, in S. Uchikawa

(ed.), Economic Reforms and Industrial Structure in India, New Delhi: Manohar Publishers.

UNCTAD-India (2008) “How are the Poor Affected by Trade in India: Empirical Approach”

Unni, Jeemol, Lalitha, N. and Rani, Uma (2000), “Economic Reforms and Productivity

Trends in Indian Manufacturing,” Working Paper No.119, Gujarat Institute of Development Research (GIDR), Ahmedabad, India.

***

133

APPENDIX 1 Table 1: Number of Enterprises Surveyed in 62nd Round by State

134

Table 2

135

APPENDIX II

Table 1: Summary Chart on Key Variables in Unorganised Manufacturing during 2000-01 and 2005-06

Rural Urban Combined Rural Urban Combined

2000-01 2005-06

Enterprises (no.) 11934585 5089519 17024101 12128266 4942554 17070820

Employment (no.) 23985700 13095200 37080900 23458286 12984514 36442799

Employment Per Enterprise 2.01 2.57 2.18 1.93 2.63 2.13 % share of Type of Enterprises in All Enterprises

OAME 92.66 70.88 86.14 91.59 70.90 85.60

NDME 5.27 21.26 10.05 6.14 20.74 10.37

DME 2.07 7.86 3.80 2.26 8.36 4.03 % share of Type of Enterprises in All Workers

OAME 79.83 45.16 67.59 76.82 43.64 65.00

NDME 8.06 27.71 15.00 10.16 26.15 15.86

DME 12.11 27.13 17.42 13.01 30.21 19.14 % share of Type of Enterprises in Total GVA

OAME 63.05 25.75 42.28 49.50 18.43 31.89

NDME 13.80 33.91 25.02 16.90 29.66 24.20

DME 23.10 40.34 32.70 33.60 51.85 44.15

GVA Per Enterprise (Rs.) 22348 65863 35357 31355 100267 51307

Labour Productivity (Rs.) 11120 25598 16233 16211 38167 24034

Capital Intensity (Rs.) 27600 137500 60500 38732 194321 83780 Loan Outstandings Per Ent (Rs.) 2900 10100 5100 4634 31966 12548 Emoluments per Hired Worker (Rs) 13082 22133 18488 20233 31302 26682

136

Table 2: Percentage Share of each State in Total in the Unorganised Manufacturing Sector in India: 2000-01 and 2005-06 States Enterprises 2000-01 Enterprises 2005-06 Workers 2000-01 Workers 2005-06 Fixed Assets 2000

Rural Urban Combined Rural Urban Combined Rural Urban Combined Rural Urban Combined Rural Urban

Andhra Pradesh

10.08 7.93 9.44 8.95 9.06 8.98 9.94 6.98 8.90 8.64 7.03 8.06 8.33 4.01

Assam 2.01 0.75 1.64 2.75 0.76 2.17 1.71 0.68 1.35 2.28 0.75 1.74 1.01 0.28

Bihar 5.73 2.46 4.75 5.47 2.21 4.52 5.15 2.01 4.04 5.14 1.91 3.99 3.88 0.96

Chhatisgarh 1.57 1.14 1.44 1.42 0.71 1.22 1.49 1.03 1.33 1.57 0.70 1.26 0.88 0.54

Gujarat 2.06 5.82 3.19 2.48 7.15 3.83 2.35 7.07 4.01 2.83 9.15 5.08 3.32 8.44

Haryana 0.84 1.82 1.14 0.99 2.23 1.35 0.74 1.85 1.13 0.96 2.46 1.49 2.36 3.58

Himachal Pradesh

0.76 0.15 0.58 0.83 0.14 0.63 0.57 0.13 0.42 0.62 0.15 0.45 1.67 0.25

Jammu & Kashmir

1.17 1.35 1.22 1.16 0.67 1.02 1.27 1.28 1.27 1.08 0.51 0.87 1.78 0.98

Jharkhand 3.39 0.88 2.64 4.45 0.92 3.43 3.46 0.67 2.48 3.59 0.82 2.61 2.89 0.25

Karnataka 5.75 6.84 6.08 5.47 6.04 5.63 5.31 5.79 5.48 5.52 5.24 5.42 4.81 4.64

Kerala 3.53 1.73 2.99 4.06 3.36 3.86 3.54 1.70 2.89 4.30 2.94 3.82 7.55 1.59

Madhya Pradesh

4.43 4.17 4.35 4.65 5.87 5.01 4.06 3.59 3.90 4.76 4.80 4.78 2.91 2.60

Maharashtra 5.60 11.22 7.28 4.59 11.54 6.60 5.17 13.22 8.02 4.29 14.60 7.96 8.72 18.03

Orissa 7.71 1.26 5.78 7.18 1.75 5.61 8.60 1.05 5.93 7.79 1.52 5.55 2.60 0.49

Punjab 1.56 3.03 2.00 1.24 2.89 1.72 1.39 3.15 2.01 0.99 2.84 1.65 3.95 6.23

Rajasthan 3.28 4.56 3.66 3.31 4.77 3.73 2.73 3.76 3.09 3.10 4.38 3.55 4.95 3.70

Tamilnadu 7.09 13.40 8.98 7.01 12.78 8.68 6.97 13.53 9.29 7.60 12.22 9.25 9.45 12.56

Uttaranchal 0.83 0.45 0.72 0.44 0.31 0.40 0.69 0.40 0.59 0.46 0.31 0.41 1.05 0.40

Uttar Pradesh

13.67 12.94 13.45 14.05 13.25 13.82 15.35 13.14 14.57 15.19 13.28 14.51 16.34 8.27

West Bengal

17.79 12.72 16.28 18.34 10.70 16.13 18.41 11.09 15.83 17.80 10.15 15.08 9.54 5.18

All India 100 100 100 100 100 100 100 100 100 100 100 100 100 100

137

Table 2 contd…

States Gross Value Added 2000 Gross Value Added 2005 Loan Outstandings 2000-01 Loan Outstandings 2005-06 EmolumentsWorkers 2000

Rural Urban Combined Rural Urban Combined Rural Urban Combined Rural Urban Combined Rural Andhra Pradesh 8.34 5.08 6.53 7.37 3.64 5.26 5.67 2.74 3.93 4.66 1.73 2.50 6.46

Assam 2.24 0.66 1.36 2.57 0.77 1.55 0.40 0.11 0.23 0.47 0.15 0.24 0.83

Bihar 5.96 1.53 3.49 4.05 0.99 2.32 0.60 0.25 0.39 0.21 0.08 0.11 3.02

Chhatisgarh 0.98 0.55 0.74 1.50 0.54 0.96 0.50 0.79 0.67 6.37 0.43 1.99 0.40

Gujarat 3.58 9.56 6.91 3.72 9.90 7.22 2.23 10.37 7.08 1.39 6.45 5.12 6.19

Haryana 1.26 2.37 1.87 2.26 3.80 3.13 0.96 2.71 2.00 8.84 4.85 5.90 1.13 Himachal Pradesh 0.97 0.21 0.55 0.64 0.60 0.62 17.25 0.18 7.08 1.30 0.29 0.55 1.39 Jammu & Kashmir 2.27 1.12 1.63 2.36 0.65 1.39 0.58 0.68 0.64 0.51 0.54 0.53 1.08

Jharkhand 2.93 0.51 1.58 2.63 0.74 1.56 0.47 0.05 0.22 0.19 0.12 0.14 1.32

Karnataka 4.97 4.76 4.85 7.86 5.23 6.37 3.55 6.44 5.27 2.91 1.79 2.08 4.25

Kerala 5.24 1.89 3.37 6.22 2.13 3.90 14.90 3.29 7.98 26.53 4.63 10.37 10.03 Madhya Pradesh 2.80 2.59 2.68 3.56 2.38 2.89 1.06 2.59 1.97 1.83 2.15 2.06 0.70

Maharashtra 7.04 17.06 12.62 6.12 23.01 15.68 14.02 26.16 21.25 10.90 52.29 41.43 5.85

Orissa 3.72 0.62 1.99 3.78 1.04 2.23 0.88 0.63 0.73 1.55 0.59 0.84 1.35

Punjab 2.56 4.62 3.70 1.46 3.47 2.60 2.25 6.41 4.73 1.64 2.63 2.37 3.43

Rajasthan 4.23 3.60 3.88 4.76 4.00 4.33 2.22 2.03 2.11 4.31 3.28 3.55 2.62

Tamilnadu 7.51 12.05 10.04 9.82 9.09 9.41 7.64 15.36 12.23 7.91 4.87 5.67 11.70

Uttaranchal 0.78 0.33 0.53 0.61 0.24 0.40 0.41 0.14 0.25 1.87 0.10 0.56 0.55 Uttar Pradesh 13.92 9.72 11.58 12.79 14.96 14.02 15.48 4.64 9.03 4.47 6.50 5.97 20.97 West Bengal 16.91 8.46 12.21 12.39 7.51 9.63 7.55 4.17 5.54 9.68 3.75 5.31 13.91

all India 100 100 100 100 100 100 100 100 100 100 100 100 100

Table 2 contd…

States Population -2000-01

%age below Poverty Line 2000/01 %age below Poverty Line 2004/05 Per Capita Income Current Pr.

Rural Urban Combined Rural Urban Combined Rural Urban Combined 2000/01

Andhra Pradesh 7.46 7.27 7.41 11.05 26.63 15.77 11.2 28 15.8

Assam 3 1 2.5 40.04 7.47 36.09 22.3 3.3 19.7

Bihar 12.83 5.12 10.69 44.3 32.91 42.6 42.1 34.6 41.4

Chhatisgarh 40.8 40.8 41.2 40.9

Gujarat 4.27 6.62 4.93 13.17 15.59 14.07 19.1 13 16.8

Haryana 2.02 2.14 2.06 8.27 9.99 8.74 13.6 15.1 14 Himachal Pradesh 0.74 0.2 0.59 7.94 4.63 7.63 10.7 3.4 10 Jammu & Kashmir 1.03 0.88 0.99 3.97 1.98 3.48 4.6 7.9 5.4

Jharkhand 46.3 46.3 20.2 40.3

138

Karnataka 4.7 6.28 5.14 17.38 25.25 20.04 20.8 32.6 25

Kerala 3.18 2.89 3.1 9.38 20.27 12.72 13.2 20.2 15 Madhya Pradesh 8.22 7.04 7.9 37.06 38.44 37.43 36.9 42.1 38.3

Maharashtra 7.51 14.36 9.42 23.72 26.81 25.02 29.6 32.2 30.7

Orissa 4.21 1.93 3.58 48.01 42.83 47.15 46.8 44.3 46.4

Punjab 2.17 2.89 2.37 6.35 5.75 6.16 9.1 7.1 8.4

Rajasthan 5.83 4.62 5.49 13.74 19.85 15.28 18.7 32.9 22.1

Tamilnadu 4.7 9.61 6.07 20.55 22.11 21.12 22.8 22.2 22.5

Uttaranchal 40.8 40.8 36.5 39.6

Uttar Pradesh 18.58 12.83 16.99 31.22 30.89 31.15 33.4 30.6 32.8

West Bengal 7.78 7.84 7.79 31.85 14.86 27.02 28.6 14.8 24.7

all India 98.32 93.75 97.15 27.09 23.62 26.1 28.3 25.7 27.5

Table 3: Growth Rate of Key Variables in the Unorganised Manufacturing Sector in India: 2000-01 and 2005-06 at real (19992000) prices

Enterprises Workers Labour Productivity

GVA Per Enterprise

rural urban combined rural urban combined rural urban combined rural

Andhra Pradesh -2.04 2.11 -0.93 -3.20 -0.05 -2.29 4.39 -2.35 1.94 3.15

Assam 6.73 -0.12 5.90 5.46 1.89 4.86 1.49 6.32 2.41 0.28

Bihar -0.60 -2.71 -0.91 -0.49 -1.27 -0.62 -4.51 -4.01 -4.44 -4.41

Chhatisgarh -1.66 -9.64 -3.32 0.59 -7.63 -1.41 10.77 11.34 9.76 13.30

Gujarat 4.09 3.61 3.83 3.36 5.13 4.47 0.78 -0.25 0.22 0.07

Haryana 3.52 3.58 3.55 4.71 5.71 5.29 10.12 7.45 8.37 11.39 Himachal Pradesh 2.06 -1.36 1.84 1.28 1.71 1.34 -4.57 28.49 6.76 -5.31 Jammu & Kashmir 0.14

-13.64 -3.58 -3.73

-16.85 -7.57 7.67 11.66 8.25 3.51

Jharkhand 5.95 0.35 5.44 0.30 3.88 0.66 -2.02 4.93 -0.10 -7.24

Karnataka -0.68 -3.01 -1.44 0.31 -2.13 -0.57 13.17 8.65 10.46 14.30

Kerala 3.21 13.44 5.29 3.52 11.43 5.37 1.91 -5.62 0.02 2.22 Madhya Pradesh 1.31 6.43 2.88 2.78 5.78 3.80 4.72 -3.89 0.77 6.24

Maharashtra -3.60 -0.02 -1.88 -4.13 1.84 -0.48 4.77 8.48 8.85 4.20

Orissa -1.08 6.21 -0.54 -2.40 7.49 -1.65 4.59 5.68 6.23 3.20

Punjab -4.25 -1.47 -2.95 -6.93 -2.21 -4.20 -1.87 -0.59 -0.24 -4.62

Rajasthan 0.49 0.28 0.41 2.13 2.93 2.47 2.55 2.23 2.45 4.22

Tamilnadu 0.11 -1.53 -0.61 1.30 -2.19 -0.43 6.99 0.00 2.26 8.27

Uttaranchal -11.53 -7.94 -10.81 -8.28 -5.61 -7.61 6.00 2.46 5.04 9.89 Uttar Pradesh 0.88 -0.12 0.60 -0.65 0.03 -0.43 2.12 13.27 8.11 0.57 West Bengal 0.93 -3.96 -0.13 -1.11 -1.92 -1.31 -1.84 3.60 0.24 -3.82

all India 0.32 -0.58 0.05 -0.44 -0.17 -0.35 3.23 3.70 3.55 2.44

139

Table 3 contd….

Fixed Assets Per Enterprise Loan Outstanding per Enterprise Emoluments per Hired Worker Capital Intensity

rural urban combined rural urban combined rural urban combined rural urban combined

Andhra Pradesh 4.26 6.15 6.43 4.10 7.78 6.63 2.09 -1.22 -1.25 5.51 8.45 7.91

Assam 4.24 12.71 5.44 3.16 30.56 10.87 5.94 -2.41 1.92 5.49 10.49 6.48

Bihar 1.64 -2.11 0.05 -

14.26 -2.12 -9.56 -16.03 -2.10 -8.16 1.54 -3.55 -0.24

Chhatisgarh 15.69 30.45 20.09 76.93 17.11 46.80 33.82 1.05 13.15 13.10 27.62 17.77

Gujarat -2.72 0.61 -0.04 -7.40 5.73 4.21 3.46 1.57 2.26 -2.02 -0.85 -0.66

Haryana 8.91 13.16 12.23 58.21 29.63 37.27 12.55 1.88 4.02 7.66 10.89 10.37Himachal Pradesh 0.36 13.14 3.07

-37.12 36.21 -30.81 -3.24 7.41 0.10 1.14 9.73 3.58

Jammu & Kashmir 4.06 20.61 8.05 2.32 32.29 14.60 3.39 5.21 4.42 8.23 25.27 12.71

Jharkhand -16.44 11.53 -10.96 -

19.06 39.71 -3.11 -1.16 -2.41 -1.11 -11.73 7.74 -6.73

Karnataka 9.18 9.86 8.77 2.61 -3.68 -2.55 13.76 3.83 6.19 8.10 8.87 7.82

Kerala 4.65 -2.15 3.48 13.51 12.13 13.92 4.48 0.91 3.45 4.33 -0.38 3.40

Madhya Pradesh 1.51 -5.59 -1.52 15.65 8.24 12.40 29.19 -4.74 2.98 0.06 -5.01 -2.39

Maharashtra 4.62 5.47 6.68 4.34 38.27 34.33 15.29 2.74 4.56 5.19 3.55 5.17

Orissa -1.12 -0.10 1.05 17.87 10.35 17.54 6.55 -1.32 2.95 0.21 -1.29 2.19

Punjab -0.14 -1.82 -0.58 2.53 1.12 2.39 0.05 0.14 0.72 2.75 -1.07 0.72

Rajasthan -0.66 4.66 2.67 18.98 30.77 26.20 1.31 1.54 1.02 -2.25 1.97 0.60

Tamilnadu 8.60 0.09 2.24 5.80 -3.39 -1.07 1.93 0.98 0.59 7.31 0.76 2.06

Uttaranchal 16.73 19.25 19.26 60.12 20.25 50.47 6.04 -3.23 2.22 12.60 16.31 15.13

Uttar Pradesh -3.22 7.07 2.26 -

18.33 28.79 5.42 3.07 14.91 12.87 -1.73 6.91 3.32

West Bengal 2.23 6.33 2.69 10.16 22.71 14.51 -0.35 3.63 1.87 4.33 4.13 3.92

all India 2.45 2.59 2.18 4.89 20.46 14.68 4.46 2.61 3.02 3.24 2.17 2.59

Note: Current/constant for unregistered manufacturing is used as deflator;Source NAS, CSO

Table 4: Rural Urban Ratios in the Unorganised Manufacturing in India: 2000-01 and 2005-06

Enterprise Workers GVA Per Worker

GVA Per Enterprise

Loan Outstanding per Enterprise

Emoluments per Hired Worker

2000-01 2005-06 2000-01

2005-06

2000-01

2005-06

2000-01

2005-06 2000-01

2005-06 2000-01

Andhra Pradesh 2.98 2.42 2.61 2.22 0.50 0.70 0.44 0.64 0.47 0.40 0.52

Assam 6.33 8.82 4.64 5.52 0.58 0.46 0.43 0.29 0.40 0.12 0.79

Bihar 5.46 6.09 4.69 4.87 0.66 0.65 0.57 0.52 0.30 0.15 1.18

Chhatisgarh 3.24 4.95 2.64 4.04 0.54 0.53 0.44 0.43 0.13 1.06 0.40

Gujarat 0.83 0.85 0.61 0.56 0.49 0.52 0.36 0.34 0.18 0.09 0.64

Haryana 1.09 1.08 0.74 0.70 0.57 0.65 0.39 0.42 0.22 0.60 0.60 Himachal Pradesh 12.09 14.34 7.85 7.69 0.47 0.11 0.30 0.06 5.37 0.11 1.23 Jammu & Kashmir 2.03 4.26 1.82 3.79 0.89 0.74 0.80 0.66 0.28 0.08 0.85

140

Jharkhand 9.01 11.82 9.47 7.95 0.48 0.34 0.51 0.23 0.74 0.05 0.87

Karnataka 1.97 2.22 1.68 1.90 0.49 0.61 0.42 0.52 0.19 0.26 0.33

Kerala 4.77 2.97 3.81 2.64 0.58 0.85 0.46 0.76 0.65 0.69 0.76 Madhya Pradesh 2.49 1.95 2.07 1.79 0.42 0.64 0.35 0.59 0.11 0.16 0.23

Maharashtra 1.17 0.98 0.72 0.53 0.46 0.38 0.28 0.21 0.31 0.08 0.47

Orissa 14.39 10.09 15.01 9.26 0.32 0.30 0.33 0.28 0.07 0.09 0.63

Punjab 1.21 1.05 0.81 0.63 0.55 0.51 0.36 0.31 0.20 0.21 0.80

Rajasthan 1.68 1.70 1.33 1.28 0.70 0.71 0.55 0.54 0.44 0.27 0.88

Tamilnadu 1.24 1.35 0.94 1.12 0.53 0.74 0.40 0.62 0.27 0.43 0.74

Uttaranchal 4.34 3.56 3.15 2.73 0.59 0.70 0.43 0.54 0.46 1.92 0.92 Uttar Pradesh 2.48 2.60 2.14 2.07 0.53 0.32 0.46 0.25 0.92 0.09 0.66 West Bengal 3.28 4.20 3.04 3.17 0.52 0.40 0.48 0.30 0.37 0.22 0.87

All India 2.34 2.45 1.83 1.81 0.43 0.42 0.34 0.31 0.29 0.14 0.59

141

142

CHAPTER 4

Have Unskilled Labour in India Benefited from International Trade?����

4.1 Introduction

One of the key arguments for promoting international trade in developing countries is the favourable impact trade on employment and returns to unskilled labour. This argument finds its origin in standard trade theories, e.g., Heckscher-Ohlin theorem and Stolper- Samuelson theorem, according to which trade results in a country exporting more of the product which uses its relatively abundant factor increasing the relative returs to this factor. Labour being the relatively abundant factor in the developing countries; it is hypothesized that trade will favourable affect labour by increasing its demand and consequently its returns. In this analytical framework one can alternatively assume that there are two types of workers, high-skilled and low-skilled, with the latter being the relatively abundant factor of production in low income countries. Accordingly, trade will increase returns to low skilled labour (or unskilled labour). However, the assumptions upon which these theorems are built are not sufficient to mirror the complexities of the real world in low income countries, in which there exist unlimited supply of unskilled labour, disguised unemployment and restrictive trade regimes, to say the least. Empirical estimations on whether trade benefits low-skilled labor-intensive production, hence increasing demand and wages of low-skilled workers in developing countries have resulted in mixed results. Krueger’s (1983) supports the traditional argument through a multi-country study on effect of trade on wages and employment. However, Feenstra and Hanson (1996, 1999) find that outsourcing from North to South results in a rise in the real wages of skilled labour relative to that of unskilled workers in both sets of countries; but this is consistent with the fact that the real wage of unskilled labour rise as well. In an attempt to resolve the debate, Winters (2004) argues that wage responses to trade will also depend to a large extent on the elasticity of supply of labour in a country. If the elasticity of supply of labour is zero, wages will increase but not employment, whereas if it is infinite, which is the case in many poor countries, employment increases but not wages. In the context of India, this debate has an additional dimension. In particular, one of the unique characteristics of the Indian labour markets is its dualistic nature where a large unorganised sector coexists with the organised sector. There are many regulations in India that apply only to the "organised sector."38 Some of these regulations lead to

� This Chapter has been contributed by Rashmi Banga, Senior Economist, UNCTAD-

India project and Shruti Sharma, former Consultant, UNCTAD-India project 38 The "organised sector" in India is defined by the size of establishment in terms of number of workers.

143

rigidities in labour markets which may interfere in the straight forward application of trade theories. Three such types of regulations are: first, fairly stringent rules exist that relate to firing workers and also of closing down of enterprises, along with the requirements of reasonable compensation for retrenchment; second, laws governing the use of temporary or casual labour enforce permanence of contract after a specified time of employment; and third, minimum wage legislation exists, which raises the cost of hiring workers and leads to downward inflexibility in wages. Further, it has been argued that given a large unorganized sector, which contributes around one third of manufacturing output and employs 86% of total workers,39 the arguments for trade increasing gains for unskilled labour are at best relevant for organized industries, which employ only a small percentage of the total labour force. Though, inflexibilities in Indian labour markets may lead to differential impact of trade on labour markets as compared to other countries, but existence of large unorganized sector, by it self may not render the argument concerning gains of trade going to unskilled labour irrelevant for the economy. Gains from trade may directly accrue to labour in the organized sector but given high backward and forward linkages40 they may percolate down to the unorganized sector. Banga and Bathla (2008)41 find that enterprises belonging to export-oriented industries and to states with high trade-orientation ratio are favourably affected by exports. According to Economic Survey (2007-08) most of the increase in organized sector employment in the period 1999-00 to 2004-05 has been on account of increase in employment of informal workers from the unorganized sector. Thus, though organized sector may be a small part of the whole economy, it is the most dynamic sector with respect to distributing the gains/losses of trade. Given the complexities of the Indian labour market and the inter-linkages between organised sector and unorganised sectors, it becomes important to estimate the extent to which the trade has affected unskilled labour in both organised sector as well as unorganised sectors. For this purpose, the paper undertakes an analysis to estimate the impact of trade on wages and employment for the organized manufacturing sector at three-digit industry level for the period 1998-99 to 2005-0642. Analysis is also undertaken for the agriculture sector, which forms a large part of the unorganized sector. The impact of trade on wages of unskilled agricultural labour is estimated at the state level for total agricultural products and separately for three agricultural products, i.e., cereals; fruits and nuts; and vegetable, roots and tubers. The period of analysis is 1990-91 to 1999-2000 for which the data on wages to unskilled agricultural workers is available.

One of the concerns that arise in undertaking such an exercise for Indian organised manufacturing sector is the lack of data on trade at the industry-level. To undertake this exercise, we construct a concordance matrix to match six-digit level HS 2002 product codes with three-digit level National

39 See National Commission of Enterprises in the Unorganised Sector (2008) 40 Mehta (1985), Samal (1990), Shaw (1990) establish strong backward and forward linkages between the two sectors. 41 Banga and Bathla (2008) is a part of the Report by UNCTAD (2008) on “How are the Poor affected by International Trade in India: An Empirical Approach” 42 The choice of the period has been restricted as ASI changed the industrial classification in 1998-99.

144

Industrial Classification (NIC). Using this concordance matrix the trade data at the product level is matched to the corresponding industries and trade data at the industry level is arrived at.

The chapter is organized as follows. Section 2 discusses the theoretical framework on and reviews the literature on impact of trade on wages and employment; section 3 outlines the trends in trends in trade, employment and wages in manufacturing and agriculture sectors; section 4 discusses the methodology adopted to estimate the impact of trade on wages in organized manufacturing and agriculture; section 5 presents the empirical results in manufacturing and agriculture sectors; finally section 6 summarizes and concludes.

4.2. Theoretical Framework and Review of Literature: Labour market Impacts of Trade

The impact of trade on employment and wages is a much disputed area. There are essentially two broad types of models that investigate the relationship between trade and labour markets – general equilibrium or trade models and partial equilibrium or labour and firm-theoretic models. In addition, trade models are of two basic types – competitive models (e.g. the HOS model) and the ‘new’ trade or industrial organisation models.

Standard trade theory, based on Heckscher-Ohlin model, suggests that in developing countries trade will have a favourable effect on manufacturing employment since it is expected that most of the exports originate from labour intensive industries and therefore higher exports may lead to higher employment. Under the H-O-S framework globalisation suggests a redistribution of employment from the import sector towards the export sector, thus increased imports reduce employment and increased exports increase employment. The theory predicts that, in their trade with developed countries, developing countries would export unskilled-labour-intensive goods and services and import skilled-labour-intensive goods and services. According to the theory therefore trade would tend to increase demand for unskilled labour as it moves towards specialization of unskilled labour intensive products and therefore would reduce the wage inequality, all other things being equal. For example, the labour- abundant country specialises in production of the labour-intensive good. Trade for this country will induce a rise in the wage-rental ratio as the labour- intensive sector expands (the Stolper-Samuelson theorem). Thus, the real reward of the country’s abundant factor rises. In turn, this leads to factor price convergence for the two countries. However, the magnitude of the impact on relative wages would depend on their responsiveness to changes in demand. The so-called "new trade theories" were developed to explain trade between countries with similar factor endowments that is characterized by intra-industry trade of similar (but differentiated) products. Input-output analysis (I/O) conducted by Gera and Massé (1996) suggests that the importance of trade in 1980s and 1990s accounts for a larger share of employment variations relative to other factors such as domestic demand and productivity than was the case during the 1970s. In the manufacturing sector, the study found that exports have become a dominant factor in employment growth in high-technology and skill-intensive industries, while import penetration adversely affected employment growth in low-technology, labour-intensive industries. This is consistent

145

with international evidence which shows that trade has been associated with job losses in labour-intensive sectors such as textiles, clothing, wood and leather (OECD, 1994, 1996). The available evidence also suggests that the net impact of trade on employment has been positive. There is also evidence that shows that the growing import content of exports has led to lower growth in export-related employment than might have been expected given the growth in exports as a share of output (Dunganand Murphy, 1999). However, in the context of growing international specialization, it may no longer be appropriate to view imports, from this narrow labour market perspective, as necessarily destructive of jobs. According to OECD (1996, 1997, 1998, 2000) trade is not the main driving force behind increased demand for skilled workers in developed countries. There is evidence that trade did contribute to the declining fortunes of low-skilled workers and increased income inequality in developed countries, but the effect was limited. With respect to developing countries, limited literature is available.

4.3. Trade and Indian Labour Markets: Review of Literature

Most of the studies assessing the relationship between trade and labour markets in India have been undertaken for the organised sector. The "organised sector" in India is defined by size of establishment in terms of number of workers. There are many regulations in India that ply only to "organised sector" and some of these regulations are considered to be especially constraining to the employers leading to rigidities in labour markets. In particular, as mentioned above, three types of regulations are seen as constraining are: first, fairly stringent rules relating to firing workers and also of closing down of enterprises, along with requirements of reasonable compensation for retrenchment; second, laws governing the use of temporary or casual labour which enforce permanence of contract after a specified time of employment; and third, minimum wage legislation which raise the cost of hiring workers.

Literature on impact of trade on employment and wages in the organized sector reveals mixed findings. Goldar (2000) found acceleration in employment growth in the 1990s both at the aggregate manufacturing level and for most two-digit industries. Tendulkar (2000) also finds similar result. According to Goldar (2002) the employment elasticity for aggregate manufacturing increased from 0.26 in the pre reform period (1973-74 to 1989-90) to 0.33 in the post reform period (1990-91 to 1997-98). He also finds a significant increase in the export-oriented industries group. However, in the import-competing industries there was a fall in employment elasticity from 0.425 in the pre-reforms period 0.264 in the post-reforms period.

Goldar and Aggarwal (2004) examined the effect of tariffs and non-tariff barriers on manufactured imports on price-cost margins in Indian industries. The analysis, based on panel data for 137 three-digit industries for the period 1980-81 to 1997-98, indicated that lowering of tariff and removal of quantitative restrictions on manufactured imports had a significant pro-competitive effect on domestic industries, tending to reduce markups or

146

price-cost margins. However, price-cost margins did not fall in the post-reform period in most industry groups. Rather, there has been a marked fall in growth rate of real wages and a significant reduction in labor’s income share in value added, reflecting perhaps a weakening of industrial labor’s bargaining power. This seems to have neutralized, to a large extent, the depressing effect of trade liberalization on price-cost margins. Ghose (2000) found that in the case of developing countries that have emerged as important exporters of manufactures to industrialized countries, trade has a large positive effect on employment and wages in the manufacturing sector. Also, trade tends to lead to a decline in wage inequality by increasing demand for unskilled workers. In the Indian context, Goldar (2002) found that employment elasticity for aggregate manufacturing increased from 0.26 in the pre reform period (1973-74 to 1989-90) to 0.33 in the post reform period (1990-91 to 1997-98). But a significant increase in employment elasticity is observed only in export-oriented industries group, as import competing industries revealed a fall in employment elasticity from 0.425 in the pre reform period to 0.264 in the post-reform period. As regards trend in real wages, results showed that growth in real wages per worker declined appreciably from 3.29 per cent per annum during the pre reform period to 1.16 per cent per annum in the post-reform period. The author attributed a decline in wage growth to government interventions in determination of wages in the organized sector in India. In a similar vein, Tendulkar (2003) analyzed organized industrial growth over three distinct policy regimes i.e., 1973-74 to 1980-81, 1980-81 to 1990-91 and 1990-91 to 1997-98. He found that during 1973-74 to 1980-81 marked as the period of restrictive industrial and trade policies, the exponential growth rate of output was 4.65 per cent and employment grew by 3.83 per cent. Product wage per worker increased at 3.2 per cent and implicit growth of productivity per worker grew at a negligible 0.8 per cent. The subsequent period from 1980-81 to 1990-91 was a period of somewhat liberal trade and industrial policies combined with an aggregate demand push provided by rising fiscal deficits and good agricultural harvests. It experienced jobless growth in manufacturing indicating only output growth at 7.1 percent. Real product wage grew by 4.5 per cent compared to implicit growth of 7.3 per cent in productivity per worker. The last period from 1990-91 to 1997-98, i.e. the period when economic reforms were initiated, witnessed considerable improvement in both output and employment growth at 9.0 and 2.9 percent respectively, with moderate product wage growth of 2.6 per cent. In another study, Goldar (2004), while stressing on slow down in productivity trends in post-reform periods in organized manufacturing, reiterated that trend growth rate in employment after 1997-98 i.e. from 1997-98 to 2001-02 was significantly negative, at about –3.3 per cent per annum. Further, trend growth rate in real value added during the same period was also very low at about 0.5 per cent per annum, which was much lower than trend growth rates in real value of output and index number of industrial manufacturing production in this period, both exceeding 5 per cent per annum. In another study, Banga (2005) examined impact of liberalization on labour productivity and wage inequality between skilled and unskilled labour in Indian manufacturing. The

147

cross industry analysis carried out from 1991-92 to 1997-98, revealed that FDI, trade and technology have improved labour productivity but along with higher labour productivity, wage inequality between skilled and unskilled workers has also increased. All the three viz. FDI, trade and technology have differential impact on wage inequality. Higher FDI in an industry raises wage inequality, while higher export intensity of an industry is associated with lower wage inequality. Further, technological progress is found to be skill biased in this period and accordingly, higher extent of technology acquisition in an industry is found to be associated with higher wage inequality. Mishra and Kumar (2005) in their study for India, in contrast have found a strong, negative and robust relationship between changes in trade policy and changes in industry premiums over time. As per their study trade liberalization has led to decreased wage inequalities between skilled and unskilled workers in India. According to them as tariff reduction were relatively large in sectors with higher proportion of unskilled workers and these sectors experience an increase in relative wages, these unskilled workers experience an increase in income relative to skilled workers. Thus, the findings in this chapter suggest that trade liberalization has led to decreased wage inequalities in India. It, thus, appears that impact of trade on organized manufacturing is diverse across industries and the time period chosen.

4.4. Trends in Trade and Labour Markets in India

4.4.1 Trends in Trade Trends in Exports and Imports

India’s total trade, both in terms of exports and imports has been rising steadily with imports rising faster than exports. Exports as a percentage of GDP increased from 5.8% to 14% in 2006-07, while imports increased from 8.8% to 20.9%. Indlia’s merchandise trade has grown at more than 20% since 2002-03, posted 22.6% growth rate in 2006-07. The value of merchandise exports reached USD 111 billion in April-December 2007 (Economic Survey 2007-08) India’s share in world merchandise exports is around 1%. Merchandise imports, on the other hand, has grown faster, i.e., 24.7% in 2006-07. In 2006-07 it reached USD 185.7 billion A large part of the growth in imports has been due to both volume growth and rise in import price of crude oil. In the period post liberalization, the composition of India’s exports has experienced some change. The share of Agriculture and Allied activities in India’s exports has been fluctuating in the period 1994-95 to 2004-05. It was 16 per cent in 1994-95, increased to 19 percent in 1995-96 and peaked in 1996-97 to 20.5 percent. Subsequently it fell to 18.8 percent in 1997-98. The downward trend continued in 1999-00 and the share reduced considerably to 15.2 percent and further to 13.5 percent in 2000-01 and 14 percent in 2001-02. The share of Agriculture and allied activities in 2006-07 is 10.3%, while share of primary products in exports is 15.1% (Table 4.1).

148

The share of manufacturing products in exports has also declined overtime. In 2000-01 it accounted for 78.8 % of total exports while in 2006-7 the share has declined to 68.6%. Within manufacturing, the share of traditional exports like textiles including RMG, gems and jewellery, leather and handicrafts has declined, while share of engineering goods and chemical products has been rising. This marks a shift in India’s export pattern. Interestingly, share of petroleum, crude & products (including coal) has risen significantly from 4.3% in 2000-01 to 15% in 2006-07. Table 4.1: Commodity Composition of Exports

* Growth Rate in dollar terms Source: Economic Survey 2007-08 Unlike export pattern, India’s import pattern does not show too much of a change over past six years. POL continues to have a share of around 30-32%, while capital goods imports have increased 10.5 to 13%. (Table 4.2)

149

Table 4.2: Commodity Composition of Imports

* Growth Rate in dollar terms Source: Economic Survey 2007-08

India’s tariff levels have experienced a significant decline ever since it embraced liberalization in 1991 (Figure 4.1). Considering Simple Average Tariffs, the tariff level dropped from 81.84 percent in 1990 to 56.34 percent in 1992. In 1997, the tariff level dropped even more sharply to 30.09 percent, but then increased in 1999 to 32.95 percent. There was a marginal decline in 2001 to 32.32 percent, which further decreased to 29.12 percent in 2004 and 18.30 percent in 2005. As far as Weighted Average Tariffs go, the tariff level was down to 27.86 in 1992 percent from 49.58 percent in 1990. In 1997 this level was 20.14 percent and it increased in 1999 to 32.95 percent. In 2001 Weighted Average Tariffs declined to 26.50 percent and continued the downward trend in 2004 (22.84 percent) and 2005 (13.42 percent).

Figure 4.1. India’s Tariff 1990-2005

Simple Average and Weighted Average Tariff

0.00

20.00

40.00

60.00

80.00

100.00

1990 1992 1997 1999 2001 2004 2005

Year

Tari

ff

Simple Average

Weighted Average

150

4.4.2 Trends in Employment and Wages 4.2.1 Trends in Employment

There has been a steady growth in GDP per capita in India since 1980 and this has improved considerably since 1990s. Per capita GDP grew at an average annual growth rate of 2.9% in the period 1980 to 1990, while it grew at 3.9% in the period 1991 to 2000. From 2000-2005 the average annual growth of GDP per capita has been 5.24%. The growth of trade/GDP ratio in the corresponding periods were 0.75%, 5.5 % and 14.2%. The most informative statistics available on unemployment is the daily status of unemployment provided by NSSO43. Employment has almost doubled in the year 2004-05 as compared to 1983 but the growth rate of employment is much higher in the period 1999 to 2004-05 as compared to 1993 to 1999-2000, in which it actually fell below the level of 1983-84 to 1993-94. Table 4.3: Employment and Unemployment Rate in India

Source: Economic Survey 2007-08

The unemployment rate has increased despite an expansion in work opportunities, from about 20.27 million unemployed during 1999-00 to about 34.74 million in 2004-05. This was because the labour force grew at a rate of 2.84%, higher than the increase in workforce. Unemployment rates are over two percentage points higher than ten years ago (at 6.07 % in 1993 compared with 8.28% in 2004) and the employment-to population ratio is lower than it was ten years ago. Whereas a part reason for decline in this ratio could be an extended time spent on education, it can also be indicative of low employment opportunities. This decline in growth of employment during 1993-94 to 1999-00 can also be attributed to lower absorption in agriculture. The share of agriculture in total employment dropped

43 NSSO gives the average level of unemployment on a given day, during the survey year and thereby

capture the unemployed days of the chronically unemployed; the unemployed days of usually employed who become intermittently unemployed during the reference week and unemployed days of those classified as employed according to the criterion of current weekly status.

1983-

84 1993-

94 1999-

00 2004-

05

1983 to

1993-94

1993-94 to

1999-00

1999-00 to

2004-05

million Growth p.a. (%)

Population 718.10 893.68 1005.05 1092.83 2.11 1.98 1.69

Labour Force 263.82 334.20 364.88 419.65 2.28 1.47 2.84

Workforce 239.49 313.93 338.19 384.91 2.61 1.25 2.62

Unemployment Rate (per cent)

9.22 6.06 7.31 8.28

No. of unemployed 24.34 20.27 26.68 34.74

151

from 61 % to 57 % and further along the same trend fell to 52% in 2004-05. While the share of manufacturing sector increased only marginally during this period, Trade, hotel and restaurant sector contributed significantly to the overall employment. Table 4.4: Employment in India by Sector:

Sector 1983 1993-94 1999-00 2004-05 Agriculture 65.42 61.03 56.64 52.06

Mining & Quarrying 0.66 0.78 0.67 0.63

Manufacturing 11.27 11.10 12.13 12.90

Electricity, Water etc. 0.34 0.41 0.34 0.35

Construction 2.56 3.63 4.44 5.57

Trade, hotel & restaurant 6.98 8.26 11.20 12.62

Transport, storage & communication 2.88 3.22 4.06 4.61

Fin., Insur., Real est., & busi. Services 0.78 1.08 1.36 2.00

Comty., Social & personal Services 9.10 10.50 9.16 9.24

Total 100.0 100.0 100.0 100.0 Source: Economic Survey 2007-08

As per the National Commission for Enterprises in the Unorganized Sector (NCEUS), the organized sector employment has increased from 54.12 million in 1999-00 to 62.57 million in 2004-05. However, the increase has been accounted for by increase in unorganized workers in organized enterprises, from 20.46 million in 1999-00 to 29.14 million in 2004-05. Thus, increase in employment in organized sector has been on account of informal employment to workers. 4.2.2. Trends in Wages

Wages in Organised Manufacturing Sector The most reliable source for wages paid in the organized manufacturing sector is Annual Survey of Industries, which give at three digit level a detailed data on wages paid to workers, supervisory staff and other employees. If the definition of unskilled labour is taken as the ‘blue collar’ workers and skilled labour as ‘white collar’ workers in the organized manufacturing sector, then we find that there has been a rising trend in the wage rates of both skilled and unskilled workers, though the rise has been much higher for the skilled workers as compared to the unskilled workers.

Agricultural Wages in India Agricultural Wages in India (AWI) is the oldest series of wage data available in the country. This wage data which provides data for various agricultural operations is collected by the Directorate of Economics and Statistics, Ministry Of Agriculture. In spite

152

of its limitations as an indicator of wages for rural labour, this source remains the major source for wage data used by researchers as well as the government for policy formulations and analysis regarding level of living of rural labourers. Table 4.6 shows the real wage rates and growth in wage rates for different states. The figures suggest a high growth rate growth for the period 1983 and 1987-88 for All India, i.e., 6.12%. In this period most of the states show better performance in increase in wage rates for agricultural labourers. Haryana and Rajasthan turn out to be the exceptions not showing any significant increase in wage rates. Using triennium averages, even the traditionally lag states otherwise classified as BIMARU states show the best performance in growth of real wages with real wages growing at the rate of 6-7 percent for Orissa, Assam, Bihar, Karnataka, Madhya Pradesh, Maharashtra, and Uttar Pradesh.West Bengal is the best performer showing the highest growth rate of around 10% for the same period. For the period 1987-88 to 1993-94, which also included the years of fiscal crisis and the consequent economic reforms, wage rates show a deceleration in almost all the states and also at the all India level where the growth rate drops down to around 2% from around 6% in the previous period considered here. This decline in growth of wage rates is sharper for Andhra Pradesh, Assam, Bihar, Gujarat and Karnataka. For Uttar Pradesh and West Bengal also there is sharp decline compared to the previous period. For the next time period considered here, that is, 1993-94 to 1999-00 most of these states continue to show a slower growth rate compared to the 1983 to 1987-88 periods. However, at the all India level wage rate growth improves to 2.3 percent compared to 2% in the previous sub-period. But this increase is mostly accounted for by the increased growth performance of the southern states of Kerala and Tamilnadu which grew at a growth rate higher than 7%. Even at the all India level, growth rate during 1993-94 and 1999-00 turns out to be lower than the entire period of 1983 to 1993-94.

153

Table 4.6: Agriculture Wage Rate State Wise: 1983-1999-2000

154

4.5. Impact of Trade on Employment and Wages: Methodology and Data

4.5.1 Labour demand Equation in Organised Manufacturing

To identify the factors that may affect employment in an industry we derive labour demand equation. For this purpose, we assume two inputs constant elasticity of substitution (CES) production function, which allows for non-constant returns to scale provided the function remains homogenous of degree µ i.e.,

Q = χ [ s(k)-ρ + (1-s) (Leλt)- ρ] -µ/ρ ………………………………(a)

Where χ > 0 & 0< s< 1

Q is the output, k is the capital, s is the share parameter and �ρ determines the degree of substitutability of the inputs. The elasticity of substitution� ��can take any non-negative constant value (including unity as in the Cobb-Douglas case) & technical

progress is labour augmenting at rate of λ. �χ is the efficiency parameter as it changes output in the same proportion for any given set of input levels and the parameter s�can be interpreted as a distribution parameter since it determines the distribution of income through the factor payments.

To examine the factors that affect the demand for labour and consequently the employment in an industry we use marginal productivity theory, and equate marginal product of labour (MPL) to the real wage (W/P) (first order conditions, assuming mark – up is constant). Taking logs and rearranging (a) we get:

λt is taken as exogenous technical change which may occur through different channels

e.g., exports and imports may affect the technology and productivity which has an effect on demand for labour, i.e.,

λt = f (λ1 Time + λ2 FDI + λ3 Exports + λ4 Imports )…..(c) Allowing for persistence in labour demand and adding fixed effects we have

Ln (L) it = α + β0 ln (L) it-1 + β1 ln (Q) it + β2 ln (W/P) it +

( )ρσ

+=

1

1where

( ))ln(

)1(1ln

)1(

1

1ln

)1(

1)ln()( χ

ρρ

βλ

ρρ

+−

−

++

+−

+−=

s

ep

w

eQLLn t

)(..........)ln()1(1ln()1(ln)ln( bs

p

wQ t

−−

−+−−

−= χσ

βσλσσ

155

β5 EXPORTS it + β6 IMPORTS it + + αi + e it

Thus, demand for labour in the manufacturing sector is a function of: L it = F [L it-1 , Q it , (w/p) it , EXPORTS it , IMPORTS it, , Time, Fixed

Effects]……………………………………………………………………………..(1).

4.5. 2 Wage-Rate Equation in Organised Manufacturing

In a competitive labour market, firms will hire workers until MCL (which is wage rate) equals MR (which is Price x MPL). However, in the developing economies wage rate is influenced by a number of additional factors, e.g., minimum wage rates set up by the government and the relative bargaining power of the labour unions. Following Greenway et al (1999) we adopt a dynamic specification in order to allow for the possibility of sticky wage adjustment through time:

Ln Wit = β1 ln W i, t-1 +β0 Xit + λi + u it

where W= wage rate, X = explanatory variables, λi = industry specific fixed effects X it = f (SIZE it , K/L it , LP it , EXPORTS it , IMPORTS it) Where: SIZE = Size of the industry (Log of Output); K/L is capital-labour ratio; LP is labour Productivity; EXPORTS is exports of the industry; IMPORTS is the import intensity, i.e., imports of the finished goods produced by the industry divided by the total output of the industry. i represents the industry and t represents the time period. The equation to be estimated for manufacturing sector therefore is: Ln(w/p) it = F [ Ln(w/p) it-1 , LnSIZE it , LnK/L it , LnLP it , LnEXPORTS it , LnIMPORTS; Fixed

Effects]……………………………………………………………………..(2)

4.5.3 Wage Rate Equation in Agriculture

For the agriculture sector, the wages of unskilled agricultural workers differ across states, though not across crops. We therefore, undertake state-level analysis where wages of unskilled workers are influenced by the following factors:

Ln Wit = β1 ln W i, t-1 +β0 Xit + λi + u it

where W= wage rate, X = explanatory variables, λi = state specific fixed effects X it = f (SDP it , RAINFALL it , SHAREAGRI it , IRRIGATEDAREA it , NOTRACTORS

it , FERT it , MINWAGES it, EXPORTS it , IMPORTS it , State-specific fixed effects)

156

Where SDP = State Domestic Product, RAINFALL = Average annual rainfall received, SHAREAGRI = Share of agriculture in total SDP, IRRIGATEDAREA = Extent of gross irrigated area in the state, NOTRACTORS = Number of tractors used, FERT = Amount of fertilizers used, MINWAGES = Minimum wages of unskilled labour in the state, EXPORTS = Share of state in total exports of the product, IMPORTS = Imports of the product.

The equation to be estimated for agriculture sector therefore is:

Ln(w/p) it = F [ Ln(w/p) it-1 , LnSDP it ,LnRAINFALLit; LnSHAREAGRI it , Ln

IRRIGATEDAREA it , Ln NOTRACTORS it ,Ln FERT it ,Ln MINWAGES it ,Ln EXPORTS it , Ln

IMPORTS it , Fixed Effects]……………………………………………………………….(3)

4.5.4 Empirical Methodology

Keeping in mind the unique characteristics of Indian labour markets and the important role played by the government in the wage-setting we arrive at the wage equation. We assume that labour is available in perfectly elastic supply at any given wage rate. In this case, wages will be fixed exogeneously depending on the minimum wages fixed by the government and the bargaining power of labour unions. For the purpose of estimating the impact on employment and wage we need to take into account rigidities in the Indian labour market. We therefore construct dynamic panel data (DPD) models, which are estimated using Generalised Method of Moments (GMM) following Arellano and Bond (1991). GMM has become an important tool in the empirical analyses of panels with a large number of individual units and relatively short time series. This model can be written as

yit = αyi,t-1 + η i + νit

where i=1,…,N; t=2,…,T; T ≥ 3 and α < 1 For such models the within group estimator (for the fixed effects models) and the GLS estimator (for the random effects model) are not applicable. Therefore GMM estimator is applied. Adopting standard assumptions concerning the error components and initial conditions (i.e. error terms are not autocorrelated) Arellano and Bond (1991) propose moment conditions44. The validity of moment conditions implied by DPD models is commonly tested using conventional GMM test of overidentifying restrictions associated with Sargan (1958).

44 For details see Blundell and Bond (1998) pp: 118

157

4.5.3 Database Construction

For the manufacturing sector, no single source of data exists for the Indian economy that provides data required by this study. The study therefore draws data from two different sources, i.e., The Annual Survey of industries (ASI), which is published by the Central Statistical Organisation, Government of India and DGCI&S for trade data. ASI provides a reasonably comprehensive and reliable dissaggregated estimates for the manufacturing industries. It covers all the production units registered under the Factories Act, 194845, ‘large ones’ on a census basis (with definition of ‘large’ changing over time) and the remaining on a sample basis. DGCI&S provides data at eight digit level on HS 2002 codes. A concordance matrix is constructed to arrive at trade data at ASI three-digit industry level NIC codes. The data used in the study is constructed for 54 industries at three-digit level of industrial classification (National industrial Classification) for the period 1998-99 to 2005-0646. There are considerable problems in obtaining good quality time series data on wages by skill level. For our purpose, we use the available information on wage contained in the ASI database. This source provides the average number of full time production-process ‘workers’ and ‘employees’ (which includes in addition to ‘workers’, non-production workers like supervisors, clerks etc.) employed per day after taking account of reported multiple shift working. Wages of production workers is taken as wages of unskilled workers.

For the agriculture sector, the data is collected from the INDIAIHARVEST, which is compiled database provided by C.M.I.E. The state level wages of unskilled agricultural labour have been collected from Department of Agriculture, Ministry of Agriculture. The data on trade at the state level is not available. To estimate the state’s share in total exports of an agricultural product, we apply the share of the state in total production of the product to its exports. This assumes that the production pattern of product at the state is similar to its export pattern. Similar estimates to construct share of state in total exports have been used by other studies, e.g., Marjit and Kar (2008). Imports of a particular agricultural product, on the other hand, may affect labour more in those states which produce that product. We therefore take state’s share in the total production of the product and apply the ratio to total imports to arrive at states’ share in imports. The analysis is carried out for four categories of agricultural products, i.e., fruits and nuts; vegetables, roots and tuber; cereals; and total agricultural products. The analysis is undertaken for the period 1990-91 to 1999- 2000 (for which the data on wages of unskilled agricultural labour was available) for 14 states of India.

45 The Factories Act, it may be noted, applies to those units employing 10 or more workers and using power / 20 or more workers not using power. 46 The period chosen has been constrained by the availability of comparable data since 1998 onwards. The reason being ASI changed its industrial classification from 1998-99.

158

4.6. Empirical Results: Impact of Trade on Wages and Employment of Unskilled Labour in Organised Manufacturing 4.6.1 Impact on Wages of Unskilled Labour To estimate the impact of exports and imports on wages of unskilled labour (blue collar workers or ‘workers’ in ASI definiion), we undertake an inter-industry analysis for 54 industries for the period 1997-98 to 2005-06. Since wages and employment in Indian industries are characterised by downward rigidities (as discussed earlier), we use Generalised Method of Moments (GMM-IV) one step estimators, following Arellano and Bond (1991)47. Table 4.7 presents the results of the DPD (dynamic panel data) estimation of wage equation i.e., equations (1) and estimation of labour demand equation (2). The dependent variables are Log of wages of unskilled workers and Log of number of unskilled workers. All estimates are based upon heteroscedastic robust standard errors. Consistency of the GMM estimates requires that there is no second order correlation of the residuals of the first-differenced equation. Our results of the AR(2) test on the residuals as developed by Arellano and Bond (1991) do not allow us to reject the hypothesis of the validity of instruments used. We also use industry dummies at two-digit level to control for industry-specific effects. The results with respect to the impact of trade on wages of unskilled workers show that after controlling for inter-industry differences, export intensity of the industry has a positive and significant impact on wage rates of unskilled workers; however, impact of import intensity, which reflects the imports of the products produced by the industry as a proportion of its total output i.e., import competition does not have any significant impact on wages of unskilled workers. A plausible explanation for the result is the downward rigidity in wage rate, given the minimum wage norms being applicable in the organized manufacturing sector. Export orientation of the industry will create pressure on the industry to retain labour and improve their skills, which may raise their returns. Output of the industry has a positive and significant coefficient and so does labour productivity. Output expansion and labour productivity will increase the wages. We control for differences in the technology used by the industry. While low tech industries may pay higher wages to unskilled workers, the variable is not found to be statistically significant. 4.6.2 Impact on Employment of Unskilled Labour In terms of employment, the results are not very encouraging as we do not get any evidence that industries which are export-oriented lead to higher employment of

47 The coefficients and standard errors reported are those of the one-step estimation since, as Arellano and Bond (1991) argue, inference based on standard errors obtained from the two-step estimates can be unreliable. The Sargan test of over identifying restrictions and the test for second order autocorrelation are, however, based on two-step estimates (see Arellano and Bond 1991).

159

unskilled labour. This results appears to be contrary to the results arrived at other studies like Banga (2005), Goldar (2002) who find higher employment elasticity of demand in export oriented industries in the post reforms period. However, the dependent variable is employment of unskilled workers and not total employment. The results indicate that the export intensity of an industry does not lead to higher employment of unskilled labour after controlling for the output produced. One of the plausible reasons for this could be strict labour laws, which as argued by many, discourage firms from employing more labour. Employment of informal workers has been increasing over the years in organized manufacturing (Economic Survey 2007-08) which is also indicative of the fact that export-oriented industries may be outsourcing more or employing more of informal labour, which is not captured by the data used. Import competition also does not have any significant impact on employment of unskilled labour. This is not very surprising given the strict labour firing policy in India. Other variables are with the expected signs. Table 4.7: Impact of Trade on Wages and Employment of Unskilled Workers in Indian Manufacturing Sector: 1997-98 to 2005-06 Explanatory Variables Dependent

Variable: Log Wages of Unskilled Labour

(1)

Dependent Variable: Log number of Unskilled Labour

(2)

Log Wages to unskilled workers Lagged ( L1)

-0.13*** (-7.19)

Log Number of Unskilled Workers Lagged (L1)

- -0.02** (-14.99)

Log wage rate (Predicted) - -0.01* (-0.74)

Log Output 0.32*** (8.14)

0.04* (1.97)

Log Labour Productivity of Unskilled Labour

0.02* (1.77)

Log TECH 0.01 (1.32)

Log Export Intensity 0.03*** (2.43)

0.01 (1.39)

Log Import Intensity -0.01 (-0.34)

-0.02 (-1.53)

Cons 7.31*** (17.06)

4.64*** (4.42)

Industry Dummies Yes Yes

Wald chi2(6) 355.4* 821.3*

Auto correlation(z) -1.20 0.91

N 302 302

160

Note: 1. *** indicates significance at 1%, ** indicates significance at 5%, * indicates significance

at 10%. 2. The predicted values of wage rate arrived from the wage equation is used as an

instruments for wage rate in the employment equation. 3. The estimations are carried out for 54 industries for the period 1998-98 to 2005-06 4. The figures reported are the coefficients and the figures in bracket are the t-values.

4.6.3 Impact of Trade on Wages of Unskilled Labour in Agriculture Sector

Table 4.8 present the results of impact of exports and imports on wages of unskilled labour in three different agricultural products, i.e., fruits and nuts; cereals; vegetables, roots and tubers; and for all agricultural products. The estimation is undertaken for 14 states of India for which comparable data was available for the period 1991-92 to 2000-2001.

4.6.3.1. Results for All Agricultural Products

With respect to agricultural products as a whole, we find that exports have not had any significant impact on wages of unskilled labour. In other words states with higher exports of agricultural products do not have corresponding higher wages for unskilled workers but higher imports of agricultural products have led to lower wages of unskilled workers which implies that higher imports of products is affecting significantly the wages given to unskilled labour in states where the production of the corresponding product is higher. After controlling for the size of state we find that states using better technology, in terms of, more fertilizers and have larger irrigated areas pay more to unskilled labour. It is interesting to note that states that have higher minimum wage legislations for agriculture sector pay higher wages to unskilled labour. Appendix 1 reports the minimum wages applicable at the state level in different states. This indicates that social security nets may be desirable and effective in improving the share of unskilled labour in total income.

Fruits and Nuts

The results show that after controlling for state specific variables that may impact wages of unskilled labour in agriculture, for fruits and nuts, exports has increased the wages of unskilled labour but it is statistically significant only at a low level, however higher imports of fruits and nuts has led to a decline in wages of unskilled labour in agriculture. Size of the state affects the wages given to unskilled workers. The wages paid are higher if the state has larger gross irrigated area. Cereals

161

In case of cereals, the results indicate that exports of cereals have led to a significant rise in wages of unskilled workers and imports have not had any significant impact. Other variables that positively affect the wages of unskilled labour in production of cereals are better rainfall, higher gross irrigated area and more use of fertilizers. Vegetables

For vegetables, we find that the results indicate imports of vegetables have led to a fall in wages of unskilled labour but impact of exports is not significant. Number of tractors used in as state reflects its technological advancement. The results show that states that use better technology compared to others pay more to their unskilled labour. On the whole, we find that only in case of cereals has exports led to a favourable impact on wages of unskilled labour. For all agricultural products and at disggregated level for fruits and nuts and vegetables, exports have had no impact but it is the imports that have had significant adverse impact on the wages of unskilled labour. Other variables used indicate that wages of unskilled labour are positively affected if state domestic product is higher, rainfall is higher and minimum wages of unskilled labour are fixed at higher levels.

162

Table 4.8: Impact of Trade on Wages of unskilled Workers in Agriculture in India: 1991-92 to 2000-2001 Dependent variables: Log real wages of unskilled labour in Agriculture

Note: 1. *** indicates significance at 1%, ** indicates significance at 5%, * indicates significance

at 10%. 2. The figures reported are the coefficients and the figures in bracket are the t-values. 3. The estimations are carried out for 14 States for the period 1991-92 to 2000-2001

4.7. Conclusions and Policy Implications The chapter examines the impact of trade on wages and employment of unskilled labour in organized manufacturing and agriculture sectors. Results show that differential impact of trade may be transmitted through exports and imports in different sectors. The analysis with respect to the manufacturing sector has been undertaken by estimating the impact on wages and employment of unskilled labour in 54 industries of India for the

Explanatory Variables Fruits and Nuts Coefficient (t value) (1)

Cereals (2)

Vegetables, Roots and Tubers

(3)

All Agricultural Products

(4)

Lag of Real Wages -0.26 (-00.4)

0.51** (2.07)

1.13*** (4.91)

0.64 (3.00)

Log Exports 0.30 (1.53)

0.05*** (2.43)

-0.004 (-0.24)

-0.03 (-0.70)

Log Imports -0.25* (-1.83)

0.01 (1.58)

-0.009*** (-2.62)

-0.11*** (-2.44)

Log State Domestic Product

0.80*** (2.91)

-0.07 (-0.73)

0.25 (1.20)

0.52*** (2.92)

Log Rainfall 0.64*** (2.46)

0.13* (1.85)

-0.04 (-0.50)

0.001 (0.03)

Log Share of Agriculture

-0.97* (-1.57)

-0.20 (-1.12)

-0.20 (-1.10)

-0.16 (-0.69)

Log Gross Irrigated Area

-0.4*** (-2.16)

0.28* (1.92)

0.28 (1.05)

0.34* (1.76)

Log Number of Tractors

0.92 (1.64)

-0.05 (-1.60)

0.08* (1.81)

-0.01 (-0.32)

Log Fertilizers

0.06 (1.30)

-0.13** (-2.23)

-0.29*** (-2.49)

-0.23*** (-2.42)

Log Minimum Wages

-0.06 (-0.85)

0.05 (1.93)

-0.04 (-0.85)

0.06*** (3.06)

Constant 0.21 (1.1.3)

0.39 (0.45)

-1.56 (-0.82)

-1.31* (-1.78)

No. of observations 83 92 88 89

Sargan test Chi2 9.19 5.31 5.96 5.0

Auto correlation(z) 1.03 0.98 0.86 0.32

163

period 1998-99 to 2005-06. The results indicate that exports have had a favourable impact on wages of unskilled labour. Import competition does not seem to have displaced labour or adversely affected the wages. Strict labour laws and downward rigidity of wages in India may be a plausible reason for this result. Overall the results of agriculture sector indicate that for states which produce higher proportion of the agricultural products that are exported, there is no evidence of corresponding rise in wages of unskilled labour but states which produce higher proportion of agricultural products that are imported do witness lower wages of unskilled labour. The importance of existing regulations in protecting the wages and employment of unskilled labour in manufacturing and agriculture sectors is highlighted by the results of the study. Existence of downward rigidity of wages due to minimum wage regulations and its enforcement, especially in low tech industries, is required to mitigate the adverse impact that imports may have on wages. The study also indicates that minimum wages of unskilled labour at the states levels has led to relatively higher wages of unskilled agriculture labour in states. The results indicate that upward revision of level of minimum wages and their enforcement may ensure appropriate returns to unskilled labour.

164

References

Ahluwalia Montek Singh (2002) “State Level Performance Under Economic Reforms in India,” published in Economic Policy Reforms and the Indian Economy, edited by Anne O. Krueger, University of Chicago Press.

Acharyya Rajat(2006) “Trade Liberalization, Poverty, and Income Inequality in India”, INRM Policy Brief No. 10, Asian Development Bank.

Banga Rashmi (2006) “Impact of Liberalisation on Wages and Employment in Indian Manufacturing Industries”, Journal of South Asian Studies, Vol 11-2.

Deaton Angus & Jean Dreze (2002). "Poverty and Inequality in India: A Re-Examination," Working Papers 184, Princeton University, Woodrow Wilson School of Public and International Affairs, Research Program in Development Studies.

Goldar, Biswanath. (2002), “Trade Liberalization and Manufacturing Employment: The Case of India”, Employment Paper no. 2002/34, International Labour Office, Geneva. Greeenaway, D., Hine, B. and Wright, P. (2000). ‘Further evidence on the effect of foreign competition on industry level wages’, Weltwirtliftshaftliches Archiv, Vol. 136, pp. 41–51. Gulati and Narayanan (2002) Rice Trade Liberalization and Poverty, MSSD Discussion Paper No. 51, International Food Policy Research Institute Petia Topalova (2005) Trade Liberalization, Poverty and Inequality: Evidence From Indian Districts, Working Paper 11614 Raghubendra Jha (2000) Paper has been prepared for the WIDER project “Rising Income Inequality and Poverty Reduction: Are they Compatible? Ramesh Chand (1999) “Effects of Trade Liberalization on Agriculture in India: Commodity Aspects” in The Regional Co-ordination Centre for Research and Development of Coarse Grains, Pulses, Roots and Tuber Crops in the Humid Tropics of Asia and the Pacific (CGPRT Centre). Winters, L. A., N. McCulloch, and A. McKay (2004) “Trade Liberalization and Poverty: The Evidence So Far.” Journal of Economic Literature 42 (March): 72–115.

***

165

APPENDIX 1

State-wise Daily Rates of Minimum Wages for Agricultural

Workers fixed under Minimum Wages Act, 1948@

(As on 1.10.2001, 30.06.2003, 31.12.2004 and 31.03.2006)

Minimum Wages for unskilled Agricultural

Workers (in Rs. per day)

States/UTs As on 1.10.2001 As on 30.06.2003 As on 31.12.2004 As on 31.03.2006

Rs. 52.00 to Rs. 55.50 p.d.

Andhra Pradesh

(According to Zones)

Rs. 52.00 to Rs. 55.50 p.d. (According to Zones)

52.00# 64.00 to 84.00 (as per zone)

55.00 (Area-I)

Arunachal Pradesh

Rs. 39.87 to Rs. 42.11 p.d. (According to Areas)

Rs. 39.87 to Rs. 42.11 p.d. (According to Areas)

39.87# (Area-I) 42.11 (Area-I) (According to zones)

57.00 (Area- II)

Assam Rs. 45.00 p.d. * without food, Shelter and clothing Rs. 38.60 p.d. plus food, shelter and clothing

Rs. 60.00 p.d. without food, Shelter and clothing Rs.50.00 p.d. plus food, shelter and clothing

50.00 with food, Shelter and clothing 60.00 without food, Shelter and clothing

69

Bihar Rs. 37.88 p.d. * Rs. 37.88 p.d. * 50 66

Chhatisgarh - 52.87 52.87 52.87

Goa Rs. 58.00 p.d. 58# 94.00# 94

Gujarat Rs. 75.80 p.d. 50 50 50

84.29 with meal

Haryana Rs. 74.61 p.d. * Rs. 79.31 with Meal 84.29 with Meal 88.29 Without Meal

88.29 without meal

Himachal Pradesh

Rs. 51.00 p.d. 83.31without Meal# 65.00# 70

Jammu & Kashmir

Rs. 45.80 p.d. 45.00# 45.00# 66

Jharkhand - 45 -$ -

Karnataka Rs. 51.63 p.d. 56.3 56.3 56.48

72.00 for Lighting Work

Kerala Rs. 30.00 p.d. for light work Rs. 40.20 p.d. for Hard work

Rs. 100.00 p.d. for light work Rs. 150 for Hard work

100.00 for light work 150.00 for Hard work#

125.00 for Hard work

Madhya Pradesh Rs. 51.80 p.d. * 54.56 56.96 56.96

Zone-I 51.00

Zone- II 49.00

Zone- III 47.00

Maharashtra Not Available Not Available Zone-I 51.00 Zone-II 49.00 Zone-III 47.00 Zone-IV 45.00

Zone- IV

166

45.00

Manipur Rs. 62.15 p.d.* for Valley Areas Rs. 65.15 p.d. for Hill Areas

66 66 72.4

Meghalaya Rs. 50.00 p.d. * 50.00# 70 70

Mizoram Rs. 70.00 p.d. 84.00# 84.00# 91

Nagaland Rs. 45.00 p.d. 50.00# 50.00# 66

Orissa Rs. 42.50 p.d. * 52.5 52.5 55

Punjab Rs. 72.38 p.d.* with Meal Rs. 82.08 without meal

82.65 87.59 90.58

Rajasthan Rs. 60.00 p.d. 60 673.00# 73

Sikkim The Minimum Wages Act, 1948 yet to be extended.

The Minimum Wages Act, 1948 yet to be extended.

The Minimum Wages Act, 1948 has been extended w.e.f.1.10.2004

-

Tamil Nadu Rs. 54.00 p.d. 54 54.00# 70.00-80.00

Tripura Rs. 45.00 p.d. 50# 50.00# 50

Uttar Pradesh Rs. 58.00 p.d.* 58 58 58

Uttaranchal - 58 58 73

62 with meal

West Bengal Rs. 58.90 p.d.* (with Meal) Rs. 62.10 (without meal)

Rs.108.57with Meal Rs. 111.77 (without meal)

107.99 with Meal 110.97 without meal

65 without meal

100.00 (Andaman)

Andaman & Nicobar Islands

Rs. 70.00 p.d. (Andaman) Rs. 75.00 (Nicobar)

Rs.100.00 p.d. (Andaman) Rs. 107.00 (Nicobar)

100.00# (Andaman) 107.00# (Nicobar)

107 (Nicobar)

Chandigarh Rs. 81.65 p.d. 100 100 114

Dadra & Nagar Haveli

Rs. 60.00 p.d.* 60 84 89

Daman & Diu$ - - - -

Delhi Rs. 99.70 p.d.* 50 to 60# 110.1 125.8

Lakshadweep Rs. 46.80 p.d.* 107.1 -$ -

Pondicherry 52 -

Pondicherry Region

Rs. 20.00 to Rs. 22.00 p.d.

Rs. 45.00 to Rs. 119.00

45.00 to 100.00 45.00 for 5 hours (women)

Rs. 30.00 p.d. for light work

Mahe Region

Rs. 40.20 p.d. for Hard work

Rs. 30.00 for light work Rs. 40.20 for Hard work

30.00 for light work 40.20 for Hard work

-

55.00 for 5 hours (women)

Yanam Region Rs. 19.25 to Rs. 26.25 p.d.

Rs. 19.25 to Rs. 26.25 55.00 to 75.00

65.00 for 6 hours (Men)

Karaikal Rs. 20.00 to Rs. 22.00 p.d.

Rs. 45.00 to Rs. 100.00

45.00 to 100.00 54.00 for 6hours (Men)

167

Central Sphere Rs. 83.02 to* Rs. 92.71 p.d.

Rs. 90.05 to Rs.100.48 94.04 to 104.89 102.78 to114.78

Note : The Minimum Wages also include the variable dearness allowance, wherever

provided.

* : Indicate the Provision of variable dearness allowance with the minimum rates of wage.

# : No Provision of variable dearness allowance with the State.

$ : Not applicable.

@ : The Minimum wages also include the variable dearness allowance,

wherever provided.

Source : Ministry of Labour & Ministry of Agriculture, Govt. of India.

168

CHAPTER 5

Impact of Trade on Labour Productivity and Wage Inequality: A Case Study of Indian Organized Manufacturing����

5.1 Introduction A number of developing countries over last two decades have embarked on a program of trade liberalization in an attempt to integrate with the rest of the world. The proponents of liberal trade regimes argue that one of the main beneficiaries of greater trade openness are the workers in these countries. It is felt that, given abundant labour supplies, trade liberalization encourages the producers to relocate output in favour of labor intensive goods. The resulting increase in the demand for labour leads to a combination of an increase in employment and/or wages. This argument is supported by the experiences of the newly industrialized countries of East Asia which were the early liberalizers.48 At the same time, the opponents feel that as the developing countries liberalize their policies, one of the critical issues that has to be addressed is the social cost of liberalization, in particular, the impact on the labour markets through its negative effect on wages and employment. In fact recent experiences of trade liberalization have not been associated with large improvements in prospects of typical worker (Robbins, 1996 and Wood, 1997). Several factors can be cited for these divergences between expectation of liberal trade regimes and actual experiences. For example, if trade liberalization causes a flow of new technologies from abroad, then it will lead to an increase in demand for small number of workers with relatively high skill. It is also possible that available results may not be complete in different ways. For instance, the short term impacts may be adverse for labour because the reallocation of resources to new industries takes time and available data may not be sufficient to trace out the effects of trade fully. In the same way the sample countries that have been examined may not be appropriate (Hasan, 2001). Large number studies which found limited benefits from trade have focused on the experience of Latin American countries. Still other empirical literature on the impact of trade on labour market has mainly been focused on developed countries. It is mainly concerned

� This Chapter is contributed by Danish Hashim, Vice -President in the department of

Economic Policy and Publication, at Multi Commodity Exchange of India Limited

(MCX), Mumbai and Rashmi Banga, Senior Economist, UNCTAD-India project.

48 Although this paper intends mainly to explore the link between international trade and labour market

outcomes, the impact of trade reforms cannot be ignored. For, trade liberalisation itself impacts international trade. Further, this study pertains to post reform period, therefore, the policy of trade liberalisation needs to be considered.

169

with how trade liberalization has affected trade in U.S labour market (Freeman and Katz, 1995 and Krugman and Lawrence, 1994). It is also argued by some that trade does have the power to benefit labour at large, but this is constrained by the nature of labour market regulation (Edwards and Edwards, 1994). This is especially true for Indian economy, where a large unorganized sector co-exists with the organized sector and many regulations apply only to the organized sector. Some of these regulations such as stringent rules relating to firing of workers, closing of enterprises and minimum wage legislation are considered to be constraining for the employers leading to rigidities in labour market (Banga, 2005). Thus, where labour market regulations come in the way of adjustment to changes in the demand and supply conditions of labour, the potential for trade liberalization to benefit workers will remain limited. It is against this background that the present study makes an attempt to contribute to the existing literature by estimating the impact of trade on the labour market of organized Indian manufacturing sector in the post-reform period. This has been necessitated by the fact that despite considerable debate concerning the possible impact of reforms, very few systematic works. There has been undertaken to study the effects of international trade on Indian labour market, especially for post 1997 years. This paper attempts to fill the gap by capturing the differential effects of trade on skilled and unskilled labour in terms of wages and labour productivity. The next section provides a theoretical framework underlying this analysis. It is illustrative and guiding rather than comprehensive. This section also provides a carefully selected survey of existing literature on the subject in order to justify the undertaking of this study. The next section explains the methodology and the data used. Section 4 presents the results. Section 5 concludes the study.

5.2. Theoretical Framework and Review of Literature 5.2.1 Theoretical Framework There are mainly two theoretical approaches or analytical framework for understanding the channels through which international trade might impact on labour market. The first is neoclassical Hecksher-Oklin (H-O) model which provides predictions about the impact of trade between countries with different resource endowments as is the case of trade between developed and developing countries. The second approach is subsumed into what is called “new trade theories” which describes trade between countries with similar resource endowment as in the case of trade between developed countries.

The H-O model predicts that comparative advantage arises from the differences in relative endowments of factors of production. Nations will therefore specialize in the production of goods that employs more of their relatively abundant factor. For instance, developed countries which have relatively more abundant capital would export capital-intensive goods and services and import labour-intensive goods and services from developing countries where labour is relatively more abundant. Such prediction is reflected by actual pattern of trade between developed and developing countries (OECD,

170

1994). Thus, under the assumption of two factors and two goods version of the model, the movement from autarky to trade is associated in both countries with an increase in the relative price of good which uses the relatively abundant factor more intensively. Assuming each country produces both the two goods, the relative price of the two goods will increase in the labour abundant country making the firms there to change to production towards labour intensive good. The opposite will happen in the capital abundant country. Such a change will lead to increase in demand for labour in the labour-abundant country. Under the model’s assumption of full-employment, this will entail an increase in wages. If this assumption is relaxed, then the increase in the demand for labour may translate into higher employment as well as an increase in wages. The precise magnitude would depend upon the labour market conditions. However, in predicting the effect of trade on labour on the basis of H-O model, several caveat needs to be kept in mind. For, this model relies on a series of other restrictive assumption: Constant returns to scale in production, competitive labour and goods market, full mobility of factors within each country, and an inelastic supply of labour. Especially, the last assumption may not hold true in many developing countries. Thus trade may result in higher employment but not increase in wages. Further, H-O model can best be interpreted as a long-run rather than a short-run prediction. In the short-run, even labour may be regarded as immobile as they may have to acquire skill, undergo training and search for jobs before they move to the expanding labour-intensive sector. In such a scenario, international trade will be counter productive in the sense that it will serve to reduce the real return to labour. At another level, unionization of the labour force, minimum wages legislation and other government mandated labour regulation may also dilute the benefits of trade and impede the frictionless clearing of labour market and contribute to the stickiness of wages. Nevertheless, caveats such as mentioned above do not deny the potential of international trade to benefit labour – at the most, they may postpone such benefits. Indeed, it is argued by some that labour market intervention can even facilitate adjustment by protecting the well-being of workers. The resolution of these debates is essentially an empirical issue. This study can therefore be viewed as an attempt in this direction. New trade theories were developed to explain trade between countries with similar factor endowment such as the trade that takes place between the developed economies. The gains from trade occur because production cost falls as the scale of output increases translating in lower prices. Although, society as a whole benefits from the lower prices and higher consumption possibilities, the displacement of resources may invariably create losers in a situation where more recent waves of technological change appear to have favoured skilled workers more than the unskilled workers. These theories also assert that the relationship between labour skills and structure of commodity trade. It is felt that the quantity of labour, measured by the levels of skill and technical education embodied in human beings varies significantly across nations. Similarly skill requirements differ among different traded commodities. This being so, countries which is abundant in skilled labour, export goods which require large amounts of highly skilled labour.

171

Whereas, the exports of countries with lack of skilled labour displayed low need for skilled labour and a high requirements of unskilled labour (Keesing, 1966). Here it is pertinent to note that foreign direct investment (FDI) and international trade are mutually related components. In fact a large part of international trade is the outcome of FDI flows.49 There is a voluminous body of literature concerning the relationship between FDI and trade flows. Theories predict FDI and trade are substitutes. However, empirical works have achieved different results. A two way positive feedbacks between FDI and trade are found not only in country level research (Brainard, 1997, Noyi and Aizennan ,2005) but also in industry level research (Head and Ries, 2001, Faciane, 2005). Consequently, the impact of trade on labour market can also be traced through FDI’s impact. Economic theory suggests that FDI has a positive effect on wages in the industries of the host country. It is argued that foreign firms pay higher wages to domestic labour in comparison to local firms, not entirely due to productivity of labour. One explanation has been provided by the “efficiency wage hypothesis”. This hypothesis states that if work effort depends directly on the wage level, it will be profitable for a profit maximizing firm to pay above the market level. So far as productivity is concerned it is propounded that foreign firms are associated with higher labour productivity as compared to local firms. Theoretically, the reasoning is based on the argument that foreign firms adopt new techniques based on the latest R&D and will therefore have higher labour productivity than domestic firms. Besides, FDI may also cause higher productivity in local firms through spillover effects. It is in the context of the above theoretical framework that a review of earlier studies will be now attempted, so as to provide a justification for the present study. Since the study attempts to analyze the impact of trade on labour markets, the survey of literature is undertaken accordingly. 5.2.2. Review of Literature a. Trade and Productivity In the analysis of the potential link between trade and economic growth, one of the directions in which research has proceeded it to investigate the microeconomic link between countries trade and their firm’s productivity. It asks whether firm’s higher productivity growth is by becoming exporters or being forced to improve as a result of more intense competition with foreign rivals.

49 As is well documented the MNC’s are the driving force behind FDI. These firms have large internal

market, access to which is available only to affiliate. They also control large markets in unrelated parties having exclusive brand names and distribution channels spread over several national location. They can, thus, influence granting trade privileges not only in home markets but also elsewhere. As per one estimate MNC accounted for more than 50 percent of the total manufactured exports from developing countries (Mallery, 2005).

172

The impact of trade on labour productivity requires to be analysed with care for it is possible that higher labour productivity may lead to higher trade. Frankel and Romer’s (1999) pioneering work on the casual effect of trade on average productivity across nations was based on the perception that trade is partly determined by the location characteristics of countries that are unrelated to productivity. They had examined this idea empirically for a large set of countries in 1985 and concluded that trade has positive effect on average labour productivity. Other studies have also examined the link between trade and labour productivity. Kraay (1997) found that controlling for past firm’s performance and unobserved firm’s characteristics, past exports are a significant indicators of an enterprises current performance. The estimated co-efficient indicate that a ten percentage point increase in a firm’s export to output ratio in a given year causes a thirteen percent increase in labour productivity.50 Sometimes trade exposes firms to the latest available technology. Exposure to international markets may provide a network for sources of new knowledge and new technique. These may have positive effect on labour productivity. Bloch and McDonald (2000) have analysed this issue. They found that the labour productivity in the manufacturing firms in Australia had increased with increased exposure to exports. Alcala and Ciccone (2001) had undertaken a study to ascertain the effect of trade on average labour productivity across countries. Their findings show that the casual effect of trade on labour productivity is large, highly significant and very robust. They have examined the channels through which trade affects average labour productivity and found that trade works through total factor productivity. They also found that average labour productivity is influenced in a statistically significantly way by the size of a country’s work force once international trade is taken into account. A study by Douglas (2003) on the impact of USA-Canada FTA on Canadian manufacturing suggests that tariff reductions helped boost labour productivity by a compounded rate of 0.6 percent p.a in manufacturing as a whole and by 2.1 percent p.a in the most affected (i.e high-tariff) industries. These productivity effects were achieved by a mix of plant turnover and rising technical efficiency within plants. By increasing productivity the FTA also helped to increase the annual earnings of workers. Another study by Doyle and Zarzoso (2005) used the real openness measure as a determinant of labour productivity in a cross country setting over 1980-2000 periods. This study suggests that a 1 percent point increase in real openness increases labour productivity only by 0.55 percent. This differs in the findings of earlier studies mentioned above.51 Banga (2005a) has found that exports have raised labour productivity in Indian manufacturing industries. Higher competitive pressures have driven firms to improve their productivity. With regards to import’s impact it was found that import intensity of an industry to have a strong negative effect indicating that higher the extent of import freedom higher will be the labour productivity.

50

Cited in Banga (2005a) 51 Real openness is national imports plus exports (in US$) divided by national GDP in PPP US $.

173

The survey of literature attempted above regarding the effect of trade on different labour outcomes generally provide conflicting views about the impacts. That is, these studies are ambiguous in their conclusions. Further, these studies have viewed the different impacts in isolation and in different time period. Moreover, very few studies have been undertaken for India in the recent past to explore the linkages between trade and labour market. Consequently, this research represents one of the first attempts at analyzing econometrically the link between trade and different labour market outcomes in India. b. Trade and Wage Inequalities

Very few studies have been conducted to examine the relationship between trade and wage inequalities. One of the first attempts of trade based hypothesis to explain increased differentials between skilled and unskilled worker was made by Bhagwati and Dehejia (1994). The authors argued that increased economic integration have increased the volatility of comparative advantage. This has, in turn, led to increased labour turnover reducing the relative wages of less skilled, either because they have skill that are less transferable than those possessed by skilled workers or because they are less likely than high-skilled individuals to invest in skill improvement during jobless spells . Some empirical support for this hypothesis for Canada was found by Zalkiwal (2000). Durevall and Munshi (2006) have explored the relationship between trade liberalization and skilled-unskilled wage inequalities in Bangladesh cotton Textile industry. Their major finding is that opening up to international trade has affected unskilled and skilled wages in the same way; there is no reduction or increase in wage inequality. Moreover, trade opening seems to have increased real wages across the board, possibly because of trade-induced increases in productivity. However, Mishra and Kumar (2005) in their study for India, in contrast have found a strong, negative and robust relationship between changes in trade policy and changes in industry premiums over time. As per their study trade liberalization has led to decreased wage inequalities between skilled and unskilled workers in India. According to them as tariff reduction were relatively large in sectors with higher proportion of unskilled workers and these sectors experience an increase in relative wages, these unskilled workers experience an increase in income relative to skilled workers. Thus, the findings in this paper suggest that trade liberalization has led to decreased wage inequalities in India. Similarly Banga (2005a) found that higher export intensity of an industry is associated with lower wage inequalities. Higher exports in an industry have reduced wage inequality because most of the exports take place from low skill and labour intensive industry, as a result higher exports by raising the demand of these labour increases their return so as to subsequently reduce the wage gap. However, as technological progress is skilled biased, a higher extent of technology acquisition is found to be associated with higher wage inequalities.

174

5. 3. Methodology and Data 5.3.1. Methodology

Labour Productivity Equation Impact of trade on labour productivity has been estimated in the framework of a production function. A Constant Elasticity of Substitution (CES) production function has been chosen for the purpose. The CES production function, unlike the Cobb-Douglas production function, allows for non-unit value of elasticity of substitution between the inputs. Further, the production function is assumed to be of the type of constant returns to scale (CRS), i.e., µ = 1.

eetLesKsQ

/1]))(1()([ −−− −+= λρχ

Kmenta (1967) has shown that labour productivity can be estimated by an approximation

of CES production function by expanding ln Q in Taylor’s series around ρ = 0, i.e.,

2)]/)][ln(2/)1([)ln()/ln()ln()ln( LKseLLKeQ tt −+++= λλ ρχ

2)]/][ln(2/)1([)/ln()ln()ln()ln( LKseLKeLQ tt −++=− λλ ρχ

Labour productivity is therefore found to be a function of:

],,,)/()/[()/( 2

, itititititit TechTradeQLKLKFLQ =

Wage Inequality Equation

To analyse the effect of trade on the market for skills, we use demand and supply framework.52 Following Katz and Murphy (1992), we take two-factor CES production function with low-skilled labour (U) and skilled labour (S) as follows:

ρρρ δχδχ /1]))(1()([),( tsttuttt SUSUF −+=

where, δut and δst are functions of labour efficiency units and the parameter ρ <1.

52 FDI can have a composite effect on market for skills. Traditional trade theory (the Heckscher-Ohlin

model) suggests that FDI in developing countries with abundant low-skilled workers is located in low-skill sectors14. New trade models also based on Heckscher-Ohlin foundations consider cases where transnational corporations (TNCs) transfer activities abroad, which are less-skilled compared to the home average but more skilled compared to the host-country average (Feenstra and Hanson, 1995). In addition, new trade models have been developed where TNCs locate abroad because of firm-specific assets (Markusen and Venables, 1997) and foreign firms may have different skill intensities from domestic firms, pushing up the demand on an average for skill labour.

175

The labour efficiency index can be interpreted as accumulated human capital or the skill-

specific technology level. Elasticity of substitution between U and S is σ = 1/ (1-ρ). In neo-classical theory, technological change happens exogenously. However, trade can also shift the pattern of technological change. We let the labour efficiency indices (skill-specific technological progress) depend on trade-intensity (TRDI) [export-intensity (EXPI) and import intensity (IMPI)] and technology (TECH). Using the first-order condition that factor productivity equals the real factor price, the wage of skilled labour (WSK) relative to un-skilled workers (WUSK) can be represented as:

( ) TechTRDIUSWW tttUSKSK 431 )/1()/1()/1()/ln(1

]/)1ln[(/ln λσσλσσλσσσ

χχ −+−+−+−−=

5.3.2 Data Sources and Construction of Variables For the manufacturing sector, no single source of data exists for the Indian economy that provides data required by this study. The study therefore draws data from two different sources, i.e., The Annual Survey of industries (ASI), which is published by the Central Statistical Organisation, Government of India and DGCI&S for trade data. ASI provides a reasonably comprehensive and reliable dissaggregated estimates for the manufacturing industries. It covers all the production units registered under the Factories Act, 194853, ‘large ones’ on a census basis (with definition of ‘large’ changing over time) and the remaining on a sample basis. DGCI&S provides data at eight digit level on HS 2002 codes. A concordance matrix is constructed to arrive at trade data at ASI three-digit industry level NIC codes. There are considerable problems in obtaining good quality time series data on wages by skill level. For our purpose, we use the available information on wage contained in the ASI database. This source provides the average number of full time production-process ‘workers’ and ‘employees’ (which includes in addition to ‘workers’, non-production workers like supervisors, clerks etc.) employed per day after taking account of reported multiple shift working. Wages of production workers is taken as wages of unskilled workers. The data on employment, wages, and capital formation are extracted from Annual Survey of industries (ASI). ASI, published by the Central Statistical Organization (CSO), Government of India, provides a reasonably comprehensive and reliable disaggregated estimates for the manufacturing industries, covering all production units registered under the Factories Act, 1948, ‘large ones’ on a census basis (with definition of ‘large’ changing over time) and the remaining on a sample basis. Data is drawn for 58 industries at three-digit level of industrial classification (National Industrial Classification, 1998) for a period between 1998-99 to 2004-05

53 The Factories Act, it may be noted, applies to those units employing 10 or more workers and using power / 20 or more workers not using power.

176

For the purpose of estimating the regression equations, a set of variables have been constructed using various data series. The wage series of both skilled and unskilled labour have been deflated with WPI index to reflect the wage rates in real terms. Gross Value Added (GVA) of industries is defatted with respective WPI series to reflect the values at constant (1993) prices. To capture the technological level across industries, a ratio of capital formation to the labour is arrived at. Exports and imports intensities have been calculated by divining the share of each of them in total output produced. 5. 4. Empirical Findings 5.4.1 Trade and Labour Productivity Table 5.1 presents the results of the dynamic panel-data estimation of labour productivity equations. Since wages and employment in Indian industries are characterized by downward rigidities, we use Generalized Method of Moments (GMM-IV) two-step estimators, following Arellano and Bond (1991). Estimates are done for aggregate as well as disaggregate level of skilled and unskilled labour. Equation (1) presents the results for labour productivity, irrespective of the skills of labour. Equation (2) presents results with respect to skilled labour and equation (3) presents results with respect to unskilled labour. Each of the equations is estimated separately for export intensity and import intensity as independent variables. In view of high degree of correlation between (K/L) and (K/L)2,

the second term was dropped from the model estimated. A note of caution is that skilled labour imply ‘white collar workers’ while unskilled labour imply ‘blue collar workers’. For aggregate labour, as reported in equation (1), the relationship between both export intensity and import intensity are found to be statistically significant at 5% level. This implies that both export intensity and import intensity of the industries lead to higher labour productivity. This supports the view that competition, whether in international market or domestic market will raise labour productivity. Trade has, however, differently impacted the productivity of skilled and unskilled labour. When the relationship between trade and labour productivity is decomposed into skilled and unskilled labour, it is found that while in case of former (as reported in equation 2) the relationship is statistically not significant whereas in case of later, i.e., unskilled labour, the impact is positive and significant. This means that industries with higher export intensity, productivity of unskilled labour is higher, but impact of export intensity on productivity of skilled labour is inconclusive. Import intensity, on the other hand, improves labour productivity of both skilled as well as unskilled labour. Domestic competition may therefore raise productivity levels of labour more than competition in international markets.

177

Table: 5.1 Estimates of the Determinants of Labour Productivity

(1) (2) (3) Variables / Stats Skilled + Unskilled Skilled Unskilled

Constant -6.7899 (-43.64)*

-6.7067 (-

28.35)*

-2.7560 (-

13.65)*

-2.6673 (-

11.15)*

-6.8530 (-

43.11)*

-6.8389 (-

30.63)* Log (LPTY)-1 -0.2593

(-4.52)* -0.3108 (-3.70)*

0.2335 (1.78)

0.03512 (0.24)

-0.2686 (-4.84)*

-0.3576 (-4.29)*

Log (GVA) 0.5030 (25.98)*

0.4830 (22.01)*

0.3512 (13.32)*

0.3398 (11.32)*

0.5271 (23.34)*

0.5051 (21.31)*

Log (exports/output)

0.0138 (2.32)*

- -0.0011 (-0.13)

- 0.0121 (2.08)*

-

Log (imports/output)

- 0.0258 (3.26)*

- 0.0192 (1.77)*

- 0.0224 (2.96)*

log (K/l) 0.03495 (2.84)*

0.0399 (3.22)*

0.0661 (4.32)*

0.0514 (3.26)*

0.0269 (2.32)*

0.0340 (2.91)*

Wald Chi Sq 1373.2 623.1 187.2 204.4 1411.4 648.5 Note: (i) Estimations are done using Arellano-Bond dynamic panel-data technique. (ii) * signifies statistical significance at 5% level. (iii) Due to high degree of correlation between K/L and (K/L)2, (K/L)2 was dropped from estimation.

Higher capital intensity improves labour productivity is an expected result. Productivity of both skilled and unskilled labour is positively related with the capital labour ratio. However, the impact in case of skilled labour is much larger than that in case of unskilled labour. Modernization of industry, reflected in terms of rise in capital labour ratio, hence, increases the productivity of skilled labour more than that of unskilled labour. The results control for the size of the industries. Larger the size of the industry higher is the labour productivity. The relationship between labour productivity and output (GVA) is found to be statistically significant in all the cases. Trade on Wage Inequality Table 5.2 presents the results of the dynamic panel-data estimation of wage inequality equations. Since wages and employment in Indian industries are characterized by downward wage rigidities, we use Generalised Method of Moments (GMM-IV) two-step estimators, following Arellano and Bond (1991). Two specifications have been tested, one controls for technology differences across industries by using K/L ratio while other equation is estimated dropping k/l ratio. The results are reported in Equations (1) to (4).

178

Table: 5. 2 Estimates of the Determinants of Wage Inequality Variables /

Stats (1) (2) (3) (4)

Constant -4.289 (-7.04)*

-5.0263 (-7.75)*

-5.0967 (-10.03)*

-5.4317 (-9.15)*

Log (Wd) 0.3503 (20.39)*

0.3877 (22.42)*

0.2988 (22.06)*

0.3456 (30.21)*

Log (GVA) 0.1101 (1.41)

0.2281 (3.48)*

0.1295 (1.88)

0.2111 (3.40)*

Log (unsk/Sk)

-0.2193 (-2.22)*

-0.2831 (-2.19)*

-0.3873 (-4.54)*

-0.4070 (-3.77)*

Log (exports-output)

0.0254 (3.55)*

0.0317 (4.06)*

- -

Log (imports-output) -

- 0.0178 (2.32)*

0.0369 (5.15)*

log (K/l) 0.0853 (4.02)*

- 0.0994 (5.61)*

-

Wald Chi Sq 3460.7 2087.0 2616.0 2708.5 Note: (i) Estimations are done using Arellano-Bond dynamic panel-data technique. (ii) * signifies statistical significance at 5% level.

Wage inequality is defined as the difference between wage rate of skilled and unskilled labour. Higher the wage inequality, lower is the wage rate of unskilled labour as compared to skilled labour. It is interesting to note that the lagged wage- inequality has a significant positive relationship with wage inequality in the current period, which indicates that industries with higher wage differential continue to have larger wage inequalities. Keeping other factors constant, results indicate a positive relationship between exports and imports with wage inequality. With the increase in share of import and export intensity, the wage differential between skill and unskilled worker has tendency to widen. This is in line with other studies including Bhagwati and Dehejia (1994). Banga and Sharma (2008) using the same methodology show that exports has led to higher wages of unskilled labour in the same period. This indicates that though the real wage rate of unskilled workers have increased due to exports, the rise in wage rate of skilled workers has been higher than that of unskilled workers. Buoyancy in economic activities as witnessed in last few years has fast augmented the demand for skilled labour. But as industries have become more labour intensive, the demand for unskilled labour has not increased in the same proportion. The fact that increase in trade in India is positively related with the widening wage inequality is also reinforced by the direct relationship between scale of output and wage inequality across the sectors. The GVA as a determinant of the output is found to be directly related with the inequality in the wage rate, though this relationship is not powerfully supported by the statistical significance in all the equations estimated. But

179

since the relations are statistically significant in one of the two equations in each export and import, it can be argued that scale of output has tendency to widen the inequality. Employment differential between unskilled to skilled labour reflects that as proportion of unskilled labour to skilled labour rises, the wage inequality lowers. This indicates that in industries which employ more unskilled labour as a proportion of skilled labour, the wage inequality between skilled and unskilled labour is lower. High tech industries are found to have larger wage inequality, which is on the expected lienes as they would hire more expensive skilled labour. The impact of exports and imports on wage inequality is also tested by dropping k/l ratio, we find the results still hold indicating the robustness of the arrived result. 5. 5. Conclusion This paper takes a step forward in examining the linkages between international trade and labour market outcomes in India. The theory predicts that with greater openness to trade and resulting expansion of trade, India’s labour-intensive manufacturing sector will benefit in terms of improvement in productivity. It is also forecasted that the abundant factor (unskilled labour) should benefit more from the movement towards free trade. This will result in the reduction in the wage differential between the unskilled and skilled labour. In the light of this, the present paper examines whether such a predictions holds true for India also, by considering the case of organized manufacturing sector for the post-1997 period. The paper begins with the introduction where the need for the study is stated. Next, it provides the theoretical framework along with a review of existing literature in the subject. As the main focus of the paper is the impact of trade on productivity and wage inequality, the literature is surveyed accordingly. The literature so surveyed provides conflicting views about the impacts. In the subsequent section methodology and data sources has been discussed. The data for exports and imports at the industry level for India is not available. Using the concordance matrix between six digit HS 2002 codes at product level with three-digit National Industrial Classification (NIC) the trade data is constructed at the three digit industry level. This is an important contribution to the existing literature as it enables estimation of impact of export and import competition on industry characteristics. The impact of trade on labour productivity has been estimated on the basis of a constant elasticity of substitution (CES) production function, which allows for non-unit value of elasticity of substitution between the inputs. For analyzing the effect of trade on the market of skill, the two-factor CES production function with low skilled and skilled labour has been considered. The necessary data is culled from Annual Survey of industries, published by the Central Statistical Organization, Government of India. Data is drawn for 54 industries at three-digit level of industrial classification for the period 1998-99 to 2004-05.

180

Focusing on the registered manufacturing industry it is found that trade is positively related with labour productivity. Higher is the export intensity and import intensity of an industry, higher is the labour productivity. But, although, trade has benefited labour in general and unskilled labour in particular, it has in no way reduced the differences in the wage earnings between the two classes of labour. In fact trade has led to higher wage inequality. This suggests that higher skills enable higher gains from trade. But it needs to be noted here that though the organized Indian manufacturing sector provides employment to a small fraction of the total labour force but given the backward and forward linkages between organized and unorganized sectors in India, effects of trade on labour markets in the organized sector may percolate to the large unorganized sector.

***

181

References: Alcala, F and A. Ciccone (2001). “Trade and Productivity”, CEPR Discussion paper 3095. Arellano, M. and Bond, S. (1991). ‘Some tests of speci.cation for panel data: Monte Carlo evidence and application to employment equations’, Review of Economics Studies, Vol. 58, pp. 277–297. Banga, R (2005). “Impact of liberalisation on wages and employment in Indian manufacturing industries”, working paper No. 153, ICRIER working paper series, New Delhi. Banga, R (2005a). “Liberalisation and wage inequality in India, working paper No. 156, ICRIER working paper series, New Delhi. Bernard, A. B, J. Bradford Jensen and Peter K. Scholt, (2006). “Trade cost, Firms and Productivity, Journal of Monetary Economics, 53 Bhagwati, Jagdish and Vivek H. Dehejia (1994) “Freer Trade and Wages of the Unskilled: Is Marx Striking Again?”. In J. Bhagwati and M. Kosters, eds, Trade and Wages: Leveling Wages Down? Washington, DC: AEI Press. Bhalotra, S (2002). “The impact of economic liberalisation on employment and wages in India, International policy group, ILO, Geneva. Bloch. H and J. McDonald (1999). “The Spillover Effects of Industrial Action on Firm Profitability”, Review of Industrial Organization, September, 15(2). Brainard, S. L (1997). “ An Empirical Assessment of the Proximity-concentration Trade-off between Multinational Sales and Ttrade”, American Economic Review, Vol.87. Doyle, E and I. M Zarzoso (2005).”The Productivity, Trade and Institutional Quality Nexus: A panel Analysis, www.aeaweb. Org/annual_mtg_papers. Douglas,A. I (2003). Free Trade under Fire, Princeton University Press. Durevall, P and F. Munshi (2006). “Trade liberalisation and wage inequality: Empirical evidence from Bangladesh, Department of Economics, School of Business and Law, Gutenberg University, Sweden. Dutt, P (2003). “Labour market outcomes and trade reforms, The case of India”, in Hasan, R and D. Mitra, (ed.), The impact of trade on labour: Issues, perspectives and experience from developing Asia, Haarlem: North-Holland/Elsevier Science B.V. Dutta, V. P (2004). “Trade protection and inter-industry wages in India”, PRUS working paper No. 27, University of Sussex, Sussex.

182

Edwards, A.C and S. Edwards (1994). “Labour market distortions and structural adjustment in developing countries”, in H. Susan, et al , (ed.), Labour market in an era of Adjustment, The World Bank, Washington D.C. Faciane, K. A (2005). “Analyzing the Effect of transitory Components on Prices on Volatility Patterns with Feedback Traders”,Economics Working Paper Archive at WUSTI. Feenstra, C.Robert and Gordon H. Hanson (1995), ‘Foreign Direct Investment and Relative wages: Evidence from Mexico’s Maquiladoras’, NBER WP 5122. Frankel, J and David Romer (1999). “Does Trade Cause Growth”? American Economic Review, 89 (3). Freeman, R. B and L. Katz (1993). Difference and changes in wage structure, University of Chicago press, Chicago. Gera, S and P. Masse (1996). “Employment performance in the knowledge based economy”, Industry Canada, working paper No. 14, Canada. Ghose, D. K (2000). “Trade liberalisation and manufacturing employment”, ILO Employment paper No. 2000/3, Geneva. Hasan, R (2001). “The impact of trade and labour market regulations on employment and wages: Evidences from developing countries”, East-West Center working papers, August, No. 32. Hasan, R, D. Mitra and. V. Ramaswamy (2007). “Trade reforms, labour regulations and labour-demand elasticities: Empirical evidence from India”, The Review of Economic and Statistics, August, Vol. 89, NO. 3. Head. K and J. Ries (2001). “Overseas Investment and Firm Exports, Review of International Economics, Vol. 9. Hill, W. L (2003). International Business, TMH Publishing Ltd., New Delhi. Hirway. I (2008). Trade and gender inequalities in labour market: Case of textiles and garment industry in India, paper presented at International seminar on moving towards gender sensitization of trade policy, organized by UNCTAD, New Delhi, 25-27, February. Katz, L. F., and K. M. Murphy (1992), ‘Changes in Relative Wages, 1963-1987: Supply and Demand Factors’, Quarterly Journal of Economics, 107, 35—78.

183

Krishna. P and D. Mitra (1998). “Trade liberalisation, market discipline and productivity growth: New evidence from India, Journal of development economics, 56(2). Kmenta J. (1967), ‘On Estimation of the CES Production Function’, International

Economic Review, 8(2), 180-189. Krugman, P and L. Robert (1994). “Trade, jobs and wages,” Scientific American, April. Keesing, D.B !966). “Labour Skills and Comparative Advantage”, American Economic Review, Vol. 56. Marjit, S and R. Acharya (2003). “International trade, wage inequality and developing country: A general equilibrium approach, Heidelberg, Physica/Springer Verlag. Markusen, James R. and Anthony J. Venables (1997), ‘The Role of Multinational Firms in the wage-gap Debate’, Review of International Economics, Vol. 5, pp. 435- 451. Menon, N and Y V Rodgers (2007), “Trade competition, market power and gender wage differentials in India’s manufacturing sector, www.aeaweb.org/annual-mtg_pp. OECD, (1994). “Jobs study: Evidence and explanations” Chapter – 3, Paris. Robbins, D (1996). “HOS hits facts: Facts win; Evidence on trade and wages in the developing world”, Development discussion paper No. 557, HIID. Topalava, P (2004). “Trade liberalisation and firm productivity: The case of India, IMF working paper 04/28. Wood, A (1997). “Openness and wage inequality in developing countries: The Latin American challenge to East Asian conventional wisdom”, The World Bank Economic Review, Vol.11, No. 1. Zalkilwhal, O (2000). “Trade and wages: A Non-Stolper-Samuelson explanation: Evidence from Canada”, Phd. Dissertation, Department of Economics, Carleton university, Attowa.

184

CHAPTER 6

Effects of Trade Liberalization on Oilseed and Edible Oil Sector in India����

6.1 Introduction

India is the leading importer and consumer of edible oils in the world. Out of the 11-12 million tonnes of edible oil consumed in the country, around 5.5 million tonnes are imported annually, which includes about 2 million tonnes of soya bean oil from Argentina and Brazil, and 3.5 million tonnes of palm oil from Malaysia and Indonesia. Despite the importance of the edible oil sector in the national income and international trade, production is unable to cope with domestic demand. Over the last two decades, domestic consumption has grown at an annual rate of 6%, which has resulted in an increase from around five million tonnes in the early 1990s to around 12 million tonnes in recent years. Non-packaged oils account for nearly 50% of the consumption in both urban and rural markets. The remaining is contributed by packaged oils, with branded oils constituting a small portion of approximately 10-15%. The major factors responsible for the high consumption and imports of edible oil in India are increase in per capita income and population, change in tastes and preferences, low productivity in the domestic oilseed sector, and liberal policies for edible oil imports. With a constant rise in demand, imports are slated to rise even further in the coming years. So far, however, the tariff liberalization policies pursued in India have been the prime movers for the increase in imports since the late 1990s [Dohlman et al (2003)]. On the input front, oilseed yields in India are among the lowest in the world. The demand for oilseeds is a derived demand that emerges from edible oil and oil meal production. However, oilseed cultivation became increasingly unattractive due to low and unstable yields, and due to decreasing edible oil prices. The area under oilseed cultivation has stagnated at 26-27 million hectares over the last several years. Serious concerns are emerging over the future of oilseed production in view of this stagnation and in the likelihood of grains enjoying priority in government policies because of the food grain shortage. As a result, imports—which have been helped by tariff liberalization—have surged to meet the excess demand. However, the situation was different up to the early 1990s with a near self-sufficiency in edible oils [Lahiri Committee Report (2006)]. But as the domestic production of edible oils started falling short of demand, in April 1994 edible vegetable palmolein was placed under Open General License (OGL) with an import duty of 65%, much in conformity

� This Chapter has been contributed by Nilanjan Ghosh, Senior Vice President,

Takshashila Academia of Economic Research, Mumbai

185

with WTO principles. This move paved the way for a substantial part of domestic demand being met by imports. Prior to that, edible oil was on the negative list of imports. Beginning in 1994, tariff liberalization in edible oils happened in a phased manner. Subsequently, imports of other edible oils were also placed under OGL. The period from 1994 can broadly be divided into two distinct phases: the first between 1994 and 1998, when customs duty on edible oils progressively came down to reach a low of 15% in July 1998, and the second after 1999, when such duties witnessed a general upward trend to reach a high of 92.2% for refined palm oil in April 2005. Between 1994 and 2005, import duties were revised 17 times (Table 6.1), creating apprehension and uncertainty for both edible oil producers and importers.

Table 6.1:

Tariff and Trade Policy on Edible Oils (1994 to 2005) April 1994 Import of RBD Palmolein placed on OGL with 65% import duty.

March 1995 Import of all edible oils (except coconut oil, palm kernel oil, RBD palm oil, RBD palm stearin) placed on OGL with 30% import duty.

1996-97 (in regular Budget)

Reduction in import duty to 20%. With 2% special customs duty, the total duty was 22%. Another special duty of custom of 3% was later imposed bringing the total duty to 25%.

July 1998 Import duty further reduced to 15%.

1999-2000 (Budget)

Import duty raised to 15% (basic) plus 10% (surcharge), bringing the total import duty to 16.5%.

December 1999

Import duty on refined oils raised to 25% (basic) plus 10% (surcharge), that is 27.5%. In addition, a levy of 4% of Special Additional Duty (SAD) was imposed on refined oils.

June 2000

Import duty on crude oils raised to 25% (basic) plus 10% (surcharge), that is 27.5%, and on refined oils to 35% (basic) plus 10% (surcharge) plus 4% (SAD), that is 44.04%. Import duty on Crude Palm Oil (CPO) for manufacture of vanaspati retained at 15% (basic) plus 10% (surcharge) that is 16.5%.

November 2000

Import duty on CPO for manufacture of vanaspati raised to 25% and on crude vegetable oils to 35%. Import duty on CPO for manufacture, other than of vanaspati, raised to 55%. Import duty on refined vegetable oils raised to 45% (basic) plus 4% SAD, that is 50.8%. Import duty on refined palm oil and RBD palmolein raised to 65% basic plus 4% SAD, that is 71.6%.

March 2001 As amended on April 26, 2001

Import duty on crude oils for manufacture of vanaspati/refined oils by importers registered with Directorate of VVO&F raised to 75% (for others, duty at 85%) except on soya bean oil, rapeseed oil, and CPO, at 45%, 75% and 75%, respectively. Import duty on refined oils including RBD Palmolein raised to 85% (basic) except in the cases of soya bean and mustard oil where it was placed at 45% (basic) and 75%(basic) respectively due to the WTO binding. A 4% SAD also levied on refined oils.

October 2001

Import duty on CPO and its fractions, of edible grade, in loose or bulk form reduced from 75% to 65%.

186

November 2001 Import duty on crude sunflower oil or safflower oil reduced to 50% up to an aggregate of 1,50,000 tonnes Tariff Rate Quota (TRQ) of total imports of such goods in a financial year subject to certain conditions. Import duty on refined rape, colza, or mustard oil reduced to 45% up to an aggregate of 1,50,000 tonnes TRQ of total imports of such goods in a financial year subject to certain conditions.

March, 2002

Status quo on import duty structure maintained. Import of vanaspati from Nepal brought under SAD of 4%.

August 2002 SAD made non-applicable on vanaspati imported from Nepal under TRQ.

March 2003

Status quo on import duty structure of vegetable oils/edible oils maintained.

April 2003 Import duty on Refined Palm Oil and RBD Palmolein reduced from 85% to 70% and SAD made non-applicable on edible oils.

July 2004 Import duty on Refined Palm Oil and RBD Palmolein raised from 70% to 75%

February 2005

Import duty on crude Palm Oil and RBD Palmolein raised from 65% to 80%, and that on Refined Palm Oil and RBD Palmolein from 75% to 90%.

Source: Lahiri Committee Report (2006)

It was in January 2006 that the Report of the Committee on Rationalisation of Customs and Excise Duties on Edible Oils and Oilseeds headed by Ashok Lahiri came to the fore. Based on the recommendations of the Report, the government announced a cut in the import duties of palm oil and sunflower oil. The customs duty on refined palm oil was cut by 12.5 percentage points and on crude palm oil by 10 percentage points. Import duty on soya bean oil remained unchanged at 45%. The cut in import duty, however, has been accompanied by a freeze in base import prices on which the duty is being calculated. In the process, edible oil tariff liberalization seems to have played a significant role in determining the fate of the domestic industry, consumption patterns, and the oilseed producers [Persaud and Landes (2006)]. Eventually in 2002-03, India was the destination for over 15% of global vegetable oil imports. In the same year, imports represented about 55% of the country’s edible oil consumption and about half the value of its total agricultural imports. 2002-03 also recorded India’s import-export trade in vegetable oils and oilseed products at around Rs 17,000 crore, of which Rs 12,000 crore were attributed to import of edible oils and the remaining to export of oil meals, oilseeds, and oils [Pavaskar and Kshirsagar (2008)]. In fact, the annual per capita consumption of edible oils in the country has almost doubled in the past three decades—from 6 kg in the mid-1970s to a little over 11 kg in 2002-03. Given this background, there is a critical concern about the potential impact of tariff on the domestic edible oil economy. The conservative left wing as well as the domestic extraction industry group is against tariff liberalization on the plea that such a move will destroy the domestic industry and negatively impact farmers. Even objective agricultural

187

economists have subscribed to this view [e.g. Chand (2002), Chand et al (2004), and Hashim (2008)]. On the other hand, practitioners and exponents of free trade, as well as the multilaterals, are convinced of the lack of competitive cost advantage in such production processes and opine that import liberalization in edible oils will be beneficial for the economy [e.g. World Bank (1997), Gulati and Kelley (1999)]. There is no doubt that the edible oil and oilseed sector currently faces several challenges. Oilseed cultivation has become increasingly unattractive to the farmer due to low and unstable yields. The so-claimed success of the technology mission on oilseeds has been questioned from various corners. Production of oilseeds has eventually stagnated. On the other hand, while trade liberalization has led to low demand for oilseeds, the resultant low prices have led to poor supply response. High import tariffs and non-tariff barriers such as sanitary and phyto-sanitary restrictions have made oilseed imports unattractive [Srinivasan (2005)]. Low domestic output of raw material combined with restricted import of oilseeds has also been responsible for high underutilization of the processing capacity. On the other hand, as explained by Srinivasan (2005), the government has attempted to help oilseed growers by providing Minimum Support Prices (MSP) through its stocking policy and by imposing customs duties on imports of edible oils and oilseeds. MSP does not appear to have worked as well in the case of oilseeds as it has in the case of wheat and rice. On the other hand, it is also claimed that import tariffs on edible oils tend to impose a large burden on consumers and help processors more than oilseed farmers [Srinivasan (2005)]. The question that arises is: Is tariff liberalization necessarily a bad process that will only harm the Indian economy? Or, does it have the potential of opening up several other options by way of which the economy can concentrate on production of other commodities in which it has a comparative advantage? An objective analysis is of utmost necessity at this juncture to answer these vital concerns, and the same has been attempted in this study. While answering these concerns, this paper propounds a partial equilibrium framework that links the edible oil and oilseed sector, and examines the impact of tariff reduction on consumer surplus, processing margins, farmer surplus, as well as on the production of edible oils and oilseeds, area, imports of edible oils, and yield of oilseeds. The period of analysis for the study is 1986-87 oil year to 2005-06 oil year, where oil year ’86-’87 spans from October ’86 to September ’87. For convenience of analysis and considering the focus of the study, the 20-year period has been divided into three phases—1986-87 to 1993-94, 1994-95 to 1998-99, and 1999-2000 to 2005-06. The first phase is defined as the pre-liberalization era when edible oils were in the negative list of imports; the second phase is the liberalization era when edible oil imports were placed under OGL and tariffs were reduced in a phased manner; the third phase again witnesses a rise in tariff rates and marks an era of deviation from the second phase. The paper has three major sections. In section 6.2, the paper propounds a theoretical model in the partial equilibrium framework. In the process, section 6.2 provides the theoretical explanation of welfare changes—as far as the edible oil and the oilseed sector

188

are concerned—when tariffs change. The framework states the linkage from the consumer demand for edible oil till oilseed production and attempts to cover the major nodes of the edible oil value chain, including the equilibrium conditions. Section 6.3 presents an econometric simultaneous equation system, based on which the tariff elasticity of the various variables is estimated. Based on these estimates, a counter-factual framework or a scenario analysis is presented. The scenarios are devised on the basis of several hypothetical tariff regimes, and thus present the sensitivity of the various variables to tariff changes. Recommendations are presented in section 6.4.

6.2 Impact of Tariff on Edible Oils and Oilseeds: A Theoretical Framework

The assumptions in the model are as follows: 1. Both product and factor markets are competitive. Hence, in both markets the

consumers and suppliers are price-takers. The product (edible oils) and factor (oilseeds) markets are cleared.

2. Domestic source of supply of edible oils is identical as the domestic production of edible oils.

3. There are two sources of supply of edible oils: domestic production and imports.

4. As a restrictive assumption, demand is considered as a function of its own price and income only.

5. Import is a supply-side phenomenon, and need not reflect the import demand. Rather, supply from imported sources is an inverse function of the level of protection given to domestic producers in the form of tariffs.

6. Domestic prices exceed import prices by the proportion of the tariff rates. 7. There is a representative processing firm that produces edible oil by using two

inputs, labour and oilseeds. [Of course, there are many other inputs used, but for the sake of simplicity, they are considered to be absent.]

8. Typical of a competitive system, the processing industry maximizes its profits on its input use.

9. Labour supply in the edible oil sector is perfectly elastic, and hence wages are fixed. Labour demand is a derived demand from the quantity produced.

10. Oilseed demand is a derived demand from edible oil production, like labour demand. However, oilseed prices are determined in the oilseed market by the continuous interaction of demand and supply forces.

11. There is instantaneous information dissemination between product and factor markets the framework is devoid of information asymmetries.

12. The two inputs are independent of each other. This implies that an increase in use of one input will not necessarily entail an increase in the use of the other input. However, to increase total production, both inputs have to be increased.

13. Oilseed production and supply functions are identical. Oilseed supply and production are dependent on area under production and oilseed prices.

14. The farmer takes a decision on oilseed production from the market prices of oilseeds only. [The other possible information categories that can affect the decision on acreage are not considered.]

189

6.2.1. Symbols Used

dQ ≡≡≡≡ Quantity of edible oil demand

P ≡≡≡≡ Domestic price of edible oil

Y ≡≡≡≡ Consumer income

L ≡ Labour input in the edible oil processing industry

S ≡ Demand for oilseeds or usage of oilseeds in the production function

EQ ≡ Domestic supply/ production of edible oil

MP ≡ Import price of edible oil

M ≡ Quantity of edible oil imported

T ≡ Tariff rates

π ≡ Profit of the processor

Gw ≡ Wage rates fixed at the labour market

Sp ≡ Unit price of oilseeds

SS ≡ Supply/ production of oilseeds

A ≡ Area under oilseeds production

AC ≡ Area under competing crops

A ≡ Total available area

W ≡ Water use

ω ≡ Crop-water requirement for oilseeds

Cω ≡ Crop-water requirement for competing crops

Cp ≡ Price of competing crop

r ≡ Relative price of oilseeds In each case, adding a “*” to the symbol will indicate the equilibrium. 6.2.2. Equations and Identities in the Model

As stated earlier, the model entails the entire value chain from edible oil consumption to oilseed production, and eventually impact of a tariff change on the entire value chain. The relation and the linkages between the various variables in this static model have been expressed through several equations and identities.

Quantity demanded has been assumed to be a function of own price P and the

given income of the individual, Y . The demand function is well behaved, being drawn

from a well-behaved utility function, and diminishes with P and increases withY .

),( YPQd φ= , P∂

∂φ< 0,

Y∂

∂φ>0 … (1)

Quantity supplied is quantity produced, which depends on the two inputs of

production, namely, labour L, and oilseeds S, as well as the prices P. Of course, the supply curve is also well behaved, being derived from a well-behaved cost function.

),,( PSLfQE = , L

f

∂

∂ > 0,

S

f

∂

∂ > 0,

P

f

∂

∂ > 0. … (2)

190

Domestic price P, is higher than the imported price MP , in the proportion of the

tariff rate T. This gives us the following identity:

P = )1( TP M + … (3)

Net protection coefficient (NPC) is defined as the ratio of domestic prices and the

import prices. Hence, MP

PNPC = … (4).

NPC reflects on the level of protection or insulation of the economy from international trade. A higher protection will always lead to lower imports. Thus,

)1()( TP

PM

M

+== θθ , 0<dT

dθ [Qfrom (2.7), )]1( T

P

P

M

+= … (5)

i.e. )(TM γ= 0<dT

dγ… (6)

The equilibrium condition of the economy states that the quantity demand will be

equal to the total edible oil availability, which is the sum of domestic production and

imports. This is given in (7).

MQQ Ed += … (7)

The representative processor has a profit function, π , which is the difference

between its total revenue, EQP. , and its total cost, SpLw S .. + . Therefore,

SpLwQP SE ... −−=π … (8)

Now, (7) implies the following:

SpLwPSLfP S ..),,(. −−=π … (8A)

The input demand functions for labour and oilseeds will be obtained by differentiating (8A) with respect to L and S respectively, and setting each equal to zero, which is the first order maximizing condition. In other words,

L∂

∂π = 0 and

S∂

∂π= 0, respectively, will give us the following input demand

functions:

),( PwLL = … (9)

),( PpSS S= … (10)

Now, (9) signifies the labour demand function, while (10) indicates the oilseed demand function, both derived from the profit function of the processing industry.

On the input supply side, labour supply has already been taken as perfectly elastic, thereby making wages fixed and exogenous in the framework. However, prices of oilseeds are determined by the interaction between the demand and supply/production

forces. Oilseed supply has been taken as a function of area A, and oilseeds prices Sp , as

given in (11). It is also assumed that as oilseed prices increase, the farmer will tend to bring more area under oilseed cultivation. This is of course on the assumption that all other agricultural crop prices remain constant. Hence, a rise in oilseed prices will only bring about an increase in the relative prices of oilseeds as well.

191

),( SS pAS ρ= , where, A

SS

∂

∂>0,

S

S

p

S

∂

∂>0, as also

Sp

A

∂

∂>0. … (11)

Finally, the oilseed market equilibrium in which the equilibrium price and quantity are determined is given by:

SSS = … (12)

The equations and identities presented from (1) to (12) present the edible oil value chain from edible oil consumption to the price determination in the oilseed sector. In section 2.3, the impact of tariff change on the edible oil and oilseed sector is analysed.

6.2.3. Impact of Tariff Change: The Comparative Statics in the Edible Oil Market Here, the equilibrium condition in the edible oil market, as given by (7), is considered, and equilibrium prices are obtained. The equilibrium price P* can be expressed as a function of equilibrium labour input L*, oilseeds input in the production process S*, and tariff T, as shown in (13):

)*,*,(* TSLgP = … (13)

Here, one needs to keep in mind that L* and S* are also determined by their own market dynamics where edible oil prices play an important role, as shown by (9) and (10). For

the oilseed market, *Sp is the equilibrium price reached through the interactive demand

and supply forces. Hence:

*),(* PwL ψ= … (14)

*)*,(* PpS Sχ= … (15)

Again, the equilibrium quantity should be *)*,*,(* PSLfQ = … (16)

The system of equations given by (13), (14), (15), and (16) provide us with the equilibrium condition of the entire value chain of edible oils and oilseeds. This is depicted in Figure 1 in its five schedules 1A–1E.

Figure 6.1

1A 1B 1C P QE P P L S

TS

P*

Qd

L S

O Q* M Qd, QS L* L S* S

w pS SS

pS*

Gw LS Ld Sd

192

L* L S* S 1D 1E One has to appreciate the fact that the equilibrium is reached here with a continuous interaction between the various markets. In this framework, therefore, the tariff rate is assumed to be exogenous and is fixed by the government. As a result, imports are also exogenous in Figure 6.1, and in 1A the 90° line from M is the import schedule. The horizontal sum of QS and M gives the TS schedule, which reflects the total availability of edible oils as a function of P. The equilibrium price and quantity, namely P* and Q* respectively, are obtained by the intersection of TS and Qd. Now, this information about the equilibrium P* flows instantaneously in the factor markets. In the labour market, where the wage levels are assumed to be fixed because of a perfectly elastic labour supply curve (LS in Figure 1D), the equilibrium labour employment L* is dependent on P*, and hence there is no change in nominal wages w. LL in 1B shows the locus of the equilibrium levels of P and L. In a similar mode, as shown in the interactive schedules of 1C and 1E, given P*, the oilseed demand and supply interactions provide the equilibrium

price and quantity of oilseeds, given by *Sp and S* respectively.

Now, we consider a situation where tariffs T change. The change can be observed by total differentiation of P* in (13) with respect to T. While doing this, it needs to be noted that even L* and S* happen to be functions of P*, as elaborated in (14) and (15). Hence, the total differentiation of P* with respect to T yields the following:

MPT

P

P

S

S

g

dT

dP

P

L

L

g

dT

dP+

∂

∂

∂

∂

∂

∂+

∂

∂

∂

∂=

*.

*

*.

*

*.

*

*.

*

* … (17)

From (17), we obtain:

)*

*.

**

*.

*1(

*

P

S

S

g

P

L

L

g

P

dT

dP M

∂

∂

∂

∂−

∂

∂

∂

∂−

= … (18)

Equation (18) is an extremely significant result in this context, as it clearly states the extent of increase (decline) in equilibrium domestic prices due to an increase (decline) thereby linking the same to the factor markets. Had the product and factor markets linkage been not there, the equilibrium prices would have increased by the extent of the

import prices, i.e. MPdT

dP=

*. However, that does not happen. But whether the

equilibrium prices increase more or less than PM is the critical question. This solely

depends on the signs of *

,*

*,

* S

g

P

L

L

g

∂

∂

∂

∂

∂

∂ and

*

*

P

S

∂

∂.

We argue that *L

g

∂

∂< 0, as a rise in nominal wages (or an upward shift in LS

schedule), will bring about a decline in equilibrium labour input L*, thereby causing the

193

edible oil supply/production to decline, and eventually causing edible oil prices to rise.

On the other hand, *

*

P

L

∂

∂ >0, as a rise in prices, will induce the processing firm to produce

more, thereby demanding more labour inputs. In a similar vein, we argue that 0*

<∂

∂

S

g

and *

*

P

S

∂

∂>0. Therefore, we infer that:

M

M P

P

S

S

g

P

L

L

g

P

dT

dP<

∂

∂

∂

∂−

∂

∂

∂

∂−

=)

*

*.

**

*.

*1(

* since )

*

*.

**

*.

*1(

P

S

S

g

P

L

L

g

∂

∂

∂

∂−

∂

∂

∂

∂− > 1 …

(19)

The question is why the increase in tariff brings about an increase in edible oil prices, which is less than import prices. Let us have a re-look at the linkages. As tariffs increase, the imports contract and there is a decline in total availability of edible oils. This brings about a rise in edible oil prices. However, there remains a counteractive force that can again bring down the prices. This counteractive force emerges from domestic edible oil supply. As prices increase, there is greater incentive for the processor to produce more. To do that, the processor needs more labour and oilseed inputs. Hence, the demand for labour increases, though the nominal wage in our model does not increase because of perfect elasticity of labour supply. On the other hand, oilseed input increases; as with rise in edible oil prices, there is higher demand for oilseeds, which also pushes up oilseed prices. Figure 2 brings out the comparative statics of the entire system.

Figure 6.2

2 A 2 B 2 C P QE P P TS’ L S

P** TS P**

P*

Qd

L S

O Q** Q* M’ M Q O L* L** L S* S** S

pS

w pS** SS

pS*

w LS Sd’ Ld Ld’ Sd

L* L** L S* S** S 2 D 2 E

As stated earlier, an increase in tariff will result in a decline in imports, and the M schedule shifts to the left to M’ in 2A. This is associated with TS-schedule shifting left to

194

TS’. A new price-quantity equilibrium is reached (P**, Q**). With the new equilibrium price P**, that induces domestic producers to produce more, the equilibrium labour input increases to L**, with labour demand schedule shifting right from Ld to Ld’. On a similar note, the demand schedule for oilseeds also shifts right to Sd’ because of a rise in edible oil prices. The excess demand in the oilseed market also causes oilseed prices to increase from pS* to pS**. Hence, while the economics of the edible oil and oilseed sector are being explained from this analysis, it will be interesting to note the extent of change in oilseed prices and production dynamics.

For doing so, we assume the following equilibrium condition in the oilseed market, obtained by equating (18) and (19), and as given in (20):

),(),( SS pASPpS = … (20)

This gives us the following equation under the equilibrium condition:

*)(* PpS κ= … (21)

Now, the equilibrium price pS* (and quantity) in the oilseed market changes with the changes in the equilibrium prices in the edible oil market P* (which dictates the

demand side of oilseeds) and the equilibrium area A*. We already know that *

*

P

pS

∂

∂>0, as

shown in Figure 2. Ideally, we could have also taken the area under oilseeds as another determinant of oilseed prices and considered that as the supply-side phenomenon. The farmer can obtain a range of other information like subsidies, support prices, changes in relative prices of other crops vis-à-vis oilseeds, etc, and all these information can affect his acreage decision for oilseeds. If any of this information would have increased acreage, oilseed prices ideally would have fallen. However, we are assuming away the role of any such information from the supply side in our model.

6.2.4. Implication on Water Use

To explain the implication of tariff increase on water use, we incorporate the following equations in the model, and assume that these occur under equilibrium condition of the economy.

*.*.* CC AAW ωω += … (22)

*

**

C

S

p

pr = … (23)

AAA C =+ ** … (24)

Equation (22) states the total water use at equilibrium, when area under oilseeds is at equilibrium A* and the area under the competing crop is also at an equilibrium AC*. Equation (23) states the relative price of edible oil vis-à-vis the competing crop at equilibrium. Equation (24) states the identity that the sum of the area under oilseeds and that under competing crops is equal to the total available area. Now, area of each is the function of relative prices. We are assuming that even the competing crop market is in equilibrium. Hence, the equilibria can be stated as:

195

*)(* 1 rA χ= where *

1

dr

dχ> 0; … (25)

*)(* 2 rAC χ= where *

2

dr

dχ< 0… (26)

The implication here is that an increase in the relative price of oilseeds will increase the equilibrium area under oilseeds, and diminish the same under the competing crops. As far

as (23) is concerned, a rise in *Sp will increase r*, if *Cp does not change. So, under

such circumstances, what might be the impact of a unit change in tariff rates on water use? If we differentiate W* with respect to T, we obtain the following:

dT

dP

dP

dp

dp

dr

dr

dW

dT

dW S

S

*

*

*.

*

*.

*

**= … (27)

Equation (27) reveals the entire linkage through which tariff affects the equilibrium water use. By incorporating (24), (25), and (26), (27) can be rewritten as:

*)](.[*)(.* 11 rArW C χωχω −+= … (28)

Differentiating (28) with respect to r*, we obtain:

*.

*.

*

* 11

dr

d

dr

d

dr

dWC

χω

χω −= … (29)

Incorporating (26) in (24), we obtain:

dT

dP

dP

dp

dp

dr

dr

d

dT

dW S

S

C

*.

*

*.

*

*.

*].[

* 1χωω −= … (30)

The symbol of dT

dW * entirely depends on ][ Cωω − , as every other component term in

dT

dP

dP

dp

dp

dr

dr

d S

S

*.

*

*.

*

*.

*

1χ is positive. If ][ Cωω − >0, then

dT

dW *>0, otherwise if

][ Cωω − <0, then dT

dW *<0.

The implication is quite simple. An increase in tariff will increase the relative prices of oilseeds, and thereby bring more acreage under oilseed production from the competing crop. If the competing crop is more water-consuming than oilseeds, the total water consumption will increase, and if otherwise, the total water consumption will decrease. In case the competing crop is wheat (which requires marginally less water than groundnut, the oilseed in contention), an increase in tariff will simply increase water consumption and increase pressure on water resources as more area is brought under the oilseed. However, the same need not be true for a less water consuming crop like rapeseed or mustard. Rather, in such a scenario, an increase in tariff will diminish the pressure on water resources. 6.2.5. Welfare Impacts in the Partial Equilibrium

The theoretical framework brings about a few things to the fore. From a closer look at the sector, the increase in tariff, inevitably, brings about an increase in consumer prices, as has already been argued. While the total availability of edible oils has shrunk, there is a

196

consequent decline in consumer surplus, which is proportional to the increase in prices. Hence, a decline in tariff will diminish the prices and result in an increased consumer surplus. While the consumption of edible oil will increase under free trade in India (due to substitution and income effects), the consumption of other consumables in the consumers’ baskets too can increase if the positive income effect outweighs the negative substitution effect. In other words, the fall in the relative price of edible oil along with rise in real income leads to an increase in the consumption of edible oil and in the consumption of other commodities. This is despite the increase in the relative price of the latter if the income effect outweighs the substitution effect. With a tariff rise, therefore, the price rise has been detrimental for the consumer from all angles. However, it can prove to be more revenue generating for the processing industry, which can receive adequate protection under a rise in tariff. While domestic price increases, thereby leading to higher revenues, the impact on producer surplus will depend on the impact on input prices. A higher protection to the edible oil sector can also make the sector more amenable to employment generation; moreover, it can give a boost to farmers to produce more oilseeds at better prices. Yet, the impact on resource use like water solely depends on the water requirement of the competing crop vis-à-vis the oilseed in contention. We will estimate all these impacts in section 3 in the counterfactual analysis.

6.3 The Econometric Framework

The econometric framework attempts to indicate the linkage between the edible oil and oilseed sector. The framework borrows its theoretical underpinning from what has been presented in the previous section. A simultaneous equation system, which attempts to trace the value chain from the consumption of edible oils to the area and yield of oilseeds, has been estimated. The symbols are as follows:

tDQ = Edible oil consumption in time t

tP = Price of edible oil in time t

w

tP = Import price of the edible oil in time t

τ = Tariff per unit of price of edible oil in time t

ty = per capita income in time t

tSQ = Edible oil supply in time t

tM = Imports of edible oil

tD

OS = Oilseed demand in time t

tS

OS = Oilseed supply in time t os

tP = Price of oilseed in time t

tA = Area under oilseed crop in time t

tY = Yield of oilseed crop in time t

tI = Percentage of irrigated area under the oilseed crop in time t

tR = Annual rainfall in time t

tOMQ = Quantity of oil meal produced in time t

tOM

P = Export price of oil meal in time t

197

PC = price of competing crop (wheat) in time t

St = Ending Stock of edible oils at time t.

tσ = Ending stock of oilseeds at time t.

Z = proxy of factors affecting domestic prices, other than imports at time t,

itε are the random disturbance terms.

The simultaneous equation system presenting the partial equilibrium model is as follows:

ttttD

yPQ 1310 . εααα +++= -------------- Demand Function for edible oils (31)

ttS

ttS

OSPQ 2765 . εααα +++= -------- Domestic Supply Function for edible oils (32)

ttW

tt PPM 311109 .. εααα +++= --------------------- Import Function (33)

ZPP tW

t ++= τ -------------------- Domestic price and tariff identity (34)

ttttS

tM

SSMQQ −++= −1 ------------------ Identity stating equilibrium in edible oils (35)

ttttOS

APP 414.1312 . εααα +++= ---------- Price Function for oilseeds (36)

tttS

YAOS *= -------------------- Production of oilseeds (37)

ttC

tOS

tt PPAA 5)1(18)1(1711615 ... εαααα ++++= −−− ----------Area function for oilseeds (38)

tttOS

t RPY 621)1(2019 .. εααα +++= − ---------Yield Function for oilseeds (39)

tttS

tD

OSOS σσ −+= −1 -------------------- Equilibrium in oilseeds economy (40)

tsttomt QEPQ 7242322 .. εααα +++= ---------------- Production of oil meals (41)

Equation (31) presents edible oil demand (or consumption) as a function of edible oil prices (or WPI of edible oils) and per capita income. There are many other important factors that can affect demand, such as population. Population and per capita income have often been taken in demand models to explain increasing consumption [e.g. Ahluwalia (1976), IBRD]. However, in our case, we found the existence of a spurious positive relation between the two variables over time, and have thus removed population from the explanatory variables list. Equation (32) presents edible oil supply as a function of edible oil prices, and production of oilseeds. Here, as in section 2, we have taken domestic supply and domestic production of edible oils to be identical. Hence, the forces of supply and production should be the same. That is precisely why oilseed production, which is an input in edible oil production, has been taken as an explanatory variable, along with price of edible oils that drives the oil supply. Equation (33) presents imports as a function of domestic prices as well as import prices, while equation (34) reflects on the identity between domestic prices, import prices, and tariff. Equation (34) is extremely significant for our counterfactual analysis. The identity states that apart from the import price and tariff, there might be other factors that could affect domestic prices. This is given by Z. In that sense, Z represents a “slack” variable. More importantly, what we have intended to present here is that a unit change in tariff will bring about a change in edible oil prices, all other things remaining constant, and this price rise, in the process, affects the entire economic system presented here. Equation (35) states the equilibrium condition of the edible oil economy, where demand equals the sum of domestic production, imports, and change in the stocks. In our analysis,

198

the price of oilseeds is viewed as a function of oilseed supply, and price of edible oils as given in (36). Oilseed prices are determined by the supply and demand side factors, and the same is being reflected in this equation. This owes its logic to equation (40), where the equilibrium in the oilseed economy is depicted. On the other hand, equation (37) depicts the production identity for oilseeds as the product of area and yield of oilseeds. Equations (38) and (39) present the area and the yield functions respectively. In the yield function, the one-year price lag of oilseeds captures several input factors that include irrigation and use of fertilizers. We assume here that a higher price of the crop in the previous year might induce the farmer to opt for more intensive cropping and input use. On the other hand, rainfall has been taken as a natural input in the production process. In equation (41), oil meal production has been taken as a function of its own price and edible oil prices. Oil meal is a compulsory by-product of edible oil production, and a rise in edible oil prices might induce an increase in edible oil and oil meal production. We have run the simultaneous equation by taking the natural logarithm of all variables, so that the elasticities are easily obtainable from the regression coefficients. The results are given in Table 6.2. The slope coefficient of P in the various regression equations gives the elasticities of the concerned dependent variable with respect to price. In certain cases where edible oil price has not featured in the regression equation, elasticities have been derived from a variable that is linked to price in some other regression equation. The equation system has been estimated in STATA 10 by using a three-stage-least-squares method (Table 6.2).

199

Table 6.2: Results of the Simultaneous Equation System

.

os_pr_lag wht_pr_lag rainfall Expt_Oilmeal Exogenous variables: PCI WPI_Oil prod_seeds Imp_price_Oil lag_area Yield Prod_meal Endogenous variables: Consump_Oil Production_Oil Import_Oil OS_price area _cons 5555....33330000666677778888 ....222200008888666633332222 22225555....44444444 0000....000000000000 4444....888899997777888866669999 5555....777711115555666699992222 WPI_Oil ....222255559999555588881111 ....0000444477774444444488881111 5555....44447777 0000....000000000000 ....1111666666665555888844444444 ....3333555522225555777777776666Expt_Oilmeal ....3333444455556666777799997777 ....0000111199999999888855557777 11117777....33330000 0000....000000000000 ....3333000066665555000088883333 ....333388884444888855551111PPPPrrrroooodddd____mmmmeeeeaaaallll _cons ----....1111000033330000444499993333 ....8888333300005555555566669999 ----0000....11112222 0000....999900001111 ----1111....777733330000999911111111 1111....555522224444888811112222 os_pr_lag ....1111777766662222888833333333 ....0000333355554444444488882222 4444....99997777 0000....000000000000 ....1111000066668888000066661111 ....2222444455557777666600006666 rainfall ....8888555588889999777755555555 ....1111222233331111111111112222 6666....99998888 0000....000000000000 ....666611117777666688882222 1111....111100000000222266669999YYYYiiiieeeelllldddd _cons 1111....000066667777222299996666 ....2222444477774444888888887777 4444....33331111 0000....000000000000 ....5555888822222222222266669999 1111....555555552222333366665555 wht_pr_lag ----....1111555544440000222277771111 ....0000555522226666000011118888 ----2222....99993333 0000....000000003333 ----....2222555577771111222244447777 ----....0000555500009999222299995555 os_pr_lag ....2222000077773333444422222222 ....0000666611110000999900002222 3333....33339999 0000....000000001111 ....0000888877776666000077776666 ....3333222277770000777766669999 lag_area ....555599993333777744445555 ....0000888833333333777788881111 7777....11112222 0000....000000000000 ....4444333300003333222266669999 ....7777555577771111666633331111aaaarrrreeeeaaaa _cons ----....7777000011119999777777773333 ....6666999955555555777722229999 ----1111....00001111 0000....333311113333 ----2222....000066665555222277775555 ....6666666611113333222200005555 area ----....2222555577777777333311111111 ....2222000033334444000066663333 ----1111....22227777 0000....222200005555 ----....6666555566664444000000001111 ....1111444400009999333377779999 WPI_Oil 1111....33332222888866666666 ....1111111177779999666644446666 11111111....22226666 0000....000000000000 1111....000099997777444455554444 1111....555555559999888866667777OOOOSSSS____pppprrrriiiicccceeee _cons ----1111....777777776666222277771111 5555....111133335555444411114444 ----0000....33335555 0000....777722229999 ----11111111....8888444411115555 8888....222288888888999955556666 WPI_Oil 3333....111122229999888822229999 1111....00000000999999998888 3333....11110000 0000....000000002222 1111....111155550000333300004444 5555....111100009999333355555555Imp_price_~l ----1111....222277775555222299996666 ....6666222244443333666677777777 ----2222....00004444 0000....000044441111 ----2222....444499999999000033334444 ----....0000555511115555555577778888IIIImmmmppppoooorrrrtttt____OOOOiiiillll _cons 5555....555500008888888844446666 ....1111999944449999222277777777 22228888....22226666 0000....000000000000 5555....111122226666777799994444 5555....888899990000888899997777 WPI_Oil ....2222888822229999111166662222 ....0000555500007777888800002222 5555....55557777 0000....000000000000 ....1111888833333333888888889999 ....3333888822224444444433336666 prod_seeds ....5555444477777777555500005555 ....0000666633330000111177776666 8888....66669999 0000....000000000000 ....4444222244442222333388882222 ....6666777711112222666622228888PPPPrrrroooodddduuuuccccttttiiiioooonnnn~~~~llll _cons 4444....555555551111000099997777 ....2222444433336666888833336666 11118888....66668888 0000....000000000000 4444....000077773333444488886666 5555....000022228888777700008888 WPI_Oil ----....5555444400008888666600005555 ....0000999900009999000099997777 ----5555....99995555 0000....000000000000 ----....7777111199990000444400002222 ----....3333666622226666888800007777 PCI ....7777444477778888666633339999 ....0000333366663333777755553333 22220000....55556666 0000....000000000000 ....6666777766665555666699996666 ....8888111199991111555588882222CCCCoooonnnnssssuuuummmmpppp____OOOOiiiillll Coef. Std. Err. z P>|z| [95% Conf. Interval]

PPPPrrrroooodddd____mmmmeeeeaaaallll 11119999 2222 ....0000444466668888666633335555 0000....9999444444448888 555511117777....11110000 0000....0000000000000000YYYYiiiieeeelllldddd 11119999 2222 ....0000555544446666888866666666 0000....8888000077770000 99999999....33333333 0000....0000000000000000aaaarrrreeeeaaaa 11119999 3333 ....0000444444441111999988885555 0000....7777444488887777 88880000....55551111 0000....0000000000000000OOOOSSSS____pppprrrriiiicccceeee 11119999 2222 ....1111222211119999000055554444 0000....8888666666662222 111133334444....66665555 0000....0000000000000000IIIImmmmppppoooorrrrtttt____OOOOiiiillll 11119999 2222 1111....000088881111555533334444 0000....3333111133336666 11112222....00008888 0000....0000000022224444PPPPrrrroooodddduuuuccccttttiiiioooonnnn~~~~llll 11119999 2222 ....0000444444443333333300008888 0000....9999222288888888 222266664444....11115555 0000....0000000000000000CCCCoooonnnnssssuuuummmmpppp____OOOOiiiillll 11119999 2222 ....0000555566661111222222229999 0000....9999777722221111 888800000000....11119999 0000....0000000000000000 Equation Obs Parms RMSE "R-sq" chi2 P Three-stage least-squares regression

200

6.3.1. Estimates of Elasticities Table 6.3 gives the various elasticities that have been estimated from the regression results presented in Table 6.2.

Table 6.3: Elasticities of various variables on price (tariff) of edible oils

Consumption of edible oils -0.5409

Production of edible oils 0.2829

Import of edible oils -1.2752

Oilseed prices 1.3286

Area 0.2736

Yield 0.2327

Production of oilseeds 0.5063

Production of oilmeal 0.2596

The critical element, in this context, is that the sensitivity analysis is conducted on the assumption that the tariff elasticity of edible oil price is 1, which is evident from the identity (34) in the system of equations. As shown in Table 6.3, there are two variables that are highly sensitive to tariff, namely import of edible oils and oilseed prices, with elasticities being greater than unity. The associated signs for each variable are as expected.

6.3.2. Tariff Values and Effective Tariff Duty

To check the under-invoicing of edible oil imports, the government started the practice of fixing tariff values on import of certain edible oils from August 3, 2001. The tariff values were revised from time to time according to variations in international prices. The import duty is imposed on the tariff values. In practice, the tariff value has often diverged from the actual CIF India prices. Hence, the effective duty needs to be estimated. If the tariff value is higher than the CIF India prices, the effective duty will be lower than the import duty. Therefore, in that case, there is a lower protection to the economy than that apparently shown by the actual import duty. If otherwise, there is higher protection to the system. The same can be made out from Table 6.4 where the effective duty has been estimated by taking into account the import duty on crude palm oil (CPO) and crude soya bean oil (CSO) as in March 2007.

Table 6.4: Effective import duty as in March 2007

Source: Mehta (2008)

Import duty Tariff CIF ($/MT) Effective duty CPO 46.35 447 920 22.52

CSO 40 580 1010 23

201

The tariff price has not been changed since August 2006 at $447 per tonne for CPO and $580 for CSO, against the CIF price of $920 per tonne for CPO, and $1010 per tonne for CSO. This has reduced the effective duty to 22.5% for CPO and around 23% for CSO. These changes have been made under exceptional circumstances to fight food inflation in an era when international prices of edible oil have been increasing.

6.3.3. The Scenario Analyses For the scenario analyses, we have considered 2005–06, the last year in our study period, as the base case. The following hypothetical scenarios have been conceived with changes in tariff regimes. Base case (2005–06 levels): In the base case, we have considered that tariffs and all other associated variables are at 2005–06 levels. In 2005–06, with the exception of soya bean oil (whose tariff was fixed at the WTO bound rate of 45%), all edible oils were subjected to high import duties between 75% and 90%. An important feature of this tariff regime relates to a differential duty structure for crude and refined palm oil since December 1999. With CPO constituting 80% of edible oil imports, we assume the tariff rate prevalent for palm oil to be at 80%, which is almost equal to the weighted average of the tariff rates across all oils, with import quantities being the weights.

� WTO binding tariff rates (300% tariff): This is the scenario where we have conceived the overall average at 300% tariff, which is the WTO bound rate.

� Lahiri Committee Recommendation (65% tariff): The Committee on Rationalisation of Customs and Excise Duties on Edible Oils and Oilseeds headed by Ashok K. Lahiri submitted its report on January 2006 and suggested reduction of duty on most crude edible oils to 65%.

� Zero tariff: This is the situation when the tariff or customs duty is zero. � 50% decline in tariff duty: This is a scenario where the existing duty is

reduced by 50%. � Zero import: In the sensitivity analysis and from the import equation, we

have estimated the tariff regime under which imports are zero and the import duty is around 143%. 6.3.3.1. Identicality of “WTO binding tariff rates (300% tariff)” and “Zero Import”

Scenarios

It is assumed that the relation between customs duty and domestic variables like consumption cannot be linear all through. What we assume here is that after reaching the critical zero-import point, the consumption and production variables will remain the same (all other things remaining constant) and will no more be sensitive to imports. The likely situation is shown in Fig. 6.3.

202

Fig. 6.3: Consumption of edible oils and tariff relationship

Hence, once the import reaches zero, tariffs cannot have any impact on the domestic price of edible oils, and the equilibria will be dictated by domestic demand and supply factors only. Therefore, all other things remaining constant, there will be no difference in the scenarios of WTO binding tariff rates (300% tariff) and Zero import.

6.3.3.2. The Effective Customs Duty in Various Scenarios

As suggested earlier, there is a difference between import duty and effective duty due to the difference in tariff value and CIF prices. Table 6.5 presents the effective duty that has been used in the various scenarios, considering the CIF prices and the average tariff value across the various edible oils in 2005–06.

Table 6.5: Effective duty in various scenarios

Scenario Import

duty (%) Average tariff value ($/MT) CIF ($/MT)

Effective duty (%)

Base Case (2005-06 June-July) 80 450 444.00 81.15

WTO Binding (300%) 300 450 444.00 304.32

Lahiri Report Recommendation

(65%) 65 450 444.00 65.94

Zero Tariff 0 450 444.00 0.00

50% tariff reduction 40 450 444.00 40.58

Zero Import

Customs Duty

Consumption of

Edible Oils

WTO Bound Rate

203

Zero Import Situation 143 450 444.00 145.06

In fact, in 2005–06, as the average tariff value was higher than CIF, the rate of

protection (in terms of effective duty) was higher than the import duty. Hence, considering that the average tariff value and CIF pertaining to 2005–06 prevail across all the scenarios, we have estimated the effective duty for each scenario. The effective duty will be put to use in sensitivity analysis by using the various elasticity values that we have already obtained.

6.3.3.3. The Various Variables

The variables that have been estimated for each scenario in the sensitivity analysis are:

� Consumption of edible oils

� Production of edible oils

� Import of edible oils

� Oilseed prices � Area � Yield

� Production of oilseeds

� Production of oil meal � Production of wheat (which has been assumed as the competing crop)

In the five scenarios we have examined the changes in the four critical variables

with the base case. The variables are: 1. Change in producer surplus for the processors 2. Change in consumer surplus 3. Change in farmer surplus 4. Change in government revenue

Change in producer surplus: For each scenario we have estimated edible oil production and the associated price, oil meal production and its export price, and oilseed production and its price. Assuming that all the oil meal is exported and that the oilseed produced is going to the processor as input, we have estimated the profit for the processor. Eventually, on the base case we have estimated the change in the producer surplus in each scenario. Similar to the import price of edible oils, the export price of oil meal has been taken as an exogenous variable in our model and fixed at 2005–06 levels (when the export price of oil meal was Rs 6.70 per kg). Change in consumer surplus: Edible oil consumption and price have been estimated for each scenario with the help of the tariff elasticity of consumption. The total consumer expenditure in each scenario has been obtained as the product of consumption and price of edible oils. The difference between the expenditure in a scenario and the expenditure in the base case signifies the change in consumer surplus in relation to the base case.

204

Change in farmer surplus: As tariff increases and edible oil prices rise, oilseed prices also rise—the reason for this has already been stated in section 6.2. Here we have run two separate exercises. First, we have shown the change in the net surplus for the farmer, and eventually for the economy as a whole, by considering that the farmer does not shift to any other crop. In the second case, we have shown a farmer, who without keeping his land barren, actually shifts to a separate competing crop. Wheat has been taken as the representative competing crop. A higher oilseed price, in relation to the competing crop (wheat), induces the farmer to bring in more land under oilseeds, diverting the same away from wheat. Hence, while there is more production of oilseeds, there is an associated opportunity cost of loss of production of wheat at the all-India level. We have estimated the regression coefficient of wheat production on area and used the same estimates for obtaining wheat production across various scenarios. Now, we estimate the change in production value of oilseeds and wheat in each scenario from the base case, and then add them up. Assuming that all costs of oilseed production and wheat production remain constant, the sum of the changes in the value of production gives the change in farmer surplus in the concerned scenario as compared with the base case. Change in government revenue: The government revenue from the total tariff duties in each case has been estimated by multiplying the total value of imports by the effective tariff rates. The difference between the revenue in each scenario and the revenue in the base case gives us the estimate of the change in government revenue.

As far as farmers are concerned, what they receive is farm-gate prices and not wholesale prices, while we assume that the processors procure oilseeds at wholesale prices. We have taken wholesale prices as weighted averages of wholesale prices of individual oils and oilseeds. Farm-gate prices have been estimated in a similar manner. We have further assumed that farm-gate prices are moving in the same direction as the wholesale price index.

The notable issue here is that we assume that farm-gate prices of oilseeds cannot go below the minimum support prices (MSP). Hence, in this model of perfect information across the various markets, whenever farm-gate prices fall below the MSP, the procurement will be at MSP only. Under such circumstances, by taking the proportional mark-up of wholesale prices over farm-gate prices in the base case, we have estimated the wholesale prices of edible oils for these scenarios.

The estimates of the various economic variables in the various scenarios are shown in Table 6.6.

Table 6.6. Estimates of various variables in various scenarios

Base case (2005-06 Levels)

WTO binding (300%)

Lahiri Report recommendation

(65%)

Zero tariff

50% tariff reduction (from base

case)

Zero import

Consumption (000 tonnes)

12031 6906.29 13251.17 18538.57 15284.78 6906.29

Production (000 tonnes) 6318 7725.55 5982.87 4530.64 5424.32 7725.55

Import (000 tonnes) 4905 0.00 6077.79 11159.86 8032.43 0.00

Area (million hectares) 27.86 33.86 26.43 20.24 24.05 33.86

205

Yield (kg./ hectare) 1004 1187.98 960.19 770.37 887.18 1187.98

Production of oilseeds (million tonnes)

27.98 39.14 25.32 13.81 20.90 39.14

Production of oil meal (000 tonnes)

14111 16995.78 13424.15 10447.78 12279.39 16995.7

8 Wholesale price of edible oils (Rs/kg)

40.45 54.72 37.05 22.33 31.39 54.72

Wholesale price of oilseeds (Rs/kg)

19.25 39.26 14.49 14.25** 14.25** 35.10

Farm-gate price of oilseeds (Rs/kg)

16.21 33.06 12.20 12.00* 12.00* 33.06

*MSP ** Wholesale prices when farmers get MSP.

Hence, under WTO binding and zero import, consumption of oils declines,

production of oils increases, and area and yield of oilseeds also increase (thereby marking a rise in production of oilseeds). An increase in production of oil is also associated with an increase in production of oil meal. Wholesale prices of oils and oilseeds, as also farm-gate prices of oilseeds, increase. The other three scenarios that reveal tariff liberalization compared with the base case, in contrast to the above two scenarios, reveal diametrically opposite movements of the variables. The reasons have already been explained in section 6.2.

The changes in the critical variables in monetary terms have been estimated in Tables 6.7 and 6.8.

Table 6.7: Change in surplus generated in the economy in various scenarios when

cultivation does not shift ( in Rs billion) WTO binding

(300%) Lahiri Report

recommendation (65%)

Zero tariff 50% tariff

reduction

Zero import

Change in producer surplus for processors

-811.37 133.30 162.82 143.26 -811.37

Change in consumer surplus

-108.74 4.33 72.70 6.87 -108.74

Change in government revenue

-1767.37 11.96 -1767.37 -320.25 -1767.37

Change in farmer surplus 152.99 -29.44 -48.05 -37.32 152.99 System-wide surplus change (Rs billion)

-2534.48 120.16 -1579.89 -207.44 -2534.48

206

Table 6.8: Change in surplus generated in the economy in various scenarios when cultivation shifts to wheat (Rs billion)

WTO binding (300%)

Lahiri Report recommendation

(65%)

Zero tariff 50% tariff reduction

Zero import

Change in producer surplus for processors (Rs billion)

-811.37 133.30 162.82 143.26 -811.37

Change in consumer surplus (Rs billion)

-108.74 4.33 72.70 6.87 -108.74

Change in government revenue (Rs billion)

-1767.37 11.96 -1767.37 -320.25 -1767.37

Change in farmer surplus (Rs billion)

-129.58 72.45 394.56 188.91 -129.58

System-wide surplus change (Rs billion)

-2817.06 222.05 -1137.29 18.79 -2817.06

As can be seen in Tables 6.7 and 6.8, the economy tends to lose the maximum

when there is no import of oil, which is prevalent in both “zero import” and “WTO binding” scenarios. In the “zero tariff” scenario, while the economy tends to gain the maximum in the context of producer surplus for processors, consumer surplus, and farmer surplus under the shifting cultivation case (and not in the non-shifting cultivation), there is a substantial loss in government revenue from tariff. Now, one might get intrigued by the maximum rise in producer surplus for processors even when the industry is not protected under a “zero-tariff” regime. This is because of a more than proportional decline in input (oilseed) prices due to a decline in tariffs, as evidenced by a greater unity elasticity shown in Table 6.3. As a result, though the total revenue for the processing industry might diminish, the profits increase. Under the condition of the Lahiri Committee recommendation, where the tariffs were recommended to be brought down to around 65% on an average, the partial equilibrium system reveals the maximum increase in the total surplus of the economy for both shifting and non-shifting cultivation. When there is a 50% decline in tariffs from 80% in 2005–06 to 40%, the system can generate an increased surplus of Rs 18.79 billion, while the system tends to lose if the farmer does not shift to other profitable crops like wheat. This is obviously lower than the Lahiri Committee recommendation scenario; yet, it is clear that the lower increase in surplus is solely due to lower governmental revenue. There are a few things that are amply clear from this exercise. First, a lowered tariff regime will always help increase the processing margin of edible oil producing firms. Second, greater the tariff liberalization, higher is the consumer surplus. Third, a greater tariff liberalization associated with farmer movement toward more profitable crops will increase farmer surplus and help increase overall economic surplus. Fourth, government revenue takes a cut due to tariff reduction, and hence it is important for the government to choose an optimum level of tariff that will be beneficial for the economy as a whole. Given the scenarios, the best case arises for the Lahiri Committee recommendation, where the total potential surplus generated in the economy is the maximum.

207

6.3.4. The Context of Individual Oils The obvious question is: What will be the situation in each case for individual oils? To analyze that, we have considered individual edible oils on their consumption, import, and production proportions, and analyzed the price behaviour and the conditions of the associated oilseeds. However, in some extreme situations, even these proportions change. For example, under a zero import situation, there cannot be consumption of imported oils, and consumption will be confined to domestic production only. In that case, all consumption will have to be derived from domestically produced oils and get distributed among the oils in the existing proportion. On the other hand, the identity that has been assumed here is that the total consumption is equal to domestic production and import of edible oils. Let us first look at the consumption of various edible oils as given in Table 6.9. In the cases of “zero import” and “WTO binding,” we assume that import of palm oil will be zero, and hence the proportion of consumption of imported oil (palm oil and soya bean oil make up 62% and 35% of imports respectively) will be distributed among the other oils. On the other hand, peanut oil (or groundnut oil) will be negatively affected by the zero-tariff condition and confined to production only, which will witness a major decline.

Table 6.9: Consumption of various edible oils in various scenarios (in ’000 tonnes)

Tariff regime Base case (2005-06 levels)

WTO binding (300%)

Lahiri Report recommendation

(65%)

Zero tariff

50% tariff reduction (from base

case)

Zero import

Palm oil (a) 3248.37 47.65 3812.80 6956.80 5022.03 47.65

Peanut oil (b) 1564.03 1609.17 1385.18 1048.95 1255.86 1609.17

Rapeseed oil (c) 2285.89 2375.76 2044.41 1548.16 1853.54 2375.76

Soya bean oil (d) 2887.44 1084.29 3074.12 4636.77 3675.14 1084.29

Total of four oils (e) = (a)+(b)+(c)+(d)

9985.73 5116.87 10316.50 14190.68 11806.57 5116.87

Others (f) 2045.27 1789.42 2934.67 4347.88 3478.21 1789.42

Consumption (g) = (e) +(f) 12031.00 6906.29 13251.17 18538.57 15284.78 6906.29

As can be seen in Table 6.9, in the WTO binding and zero import situations,

the consumption of palm oil records a major decline and gets confined to domestic production only. The decline is also noted in soya bean oil to the extent of the loss in its imports. The consumption pattern shifts substantially toward domestically produced oils, and hence the consumption of peanut and rapeseed/mustard oils increases. As has already been noted earlier, these scenarios will be characterized by a decline in overall edible oil consumption.

On the other hand, consumption of both imported and domestic oils increases in the other three scenarios. However, under a zero-tariff regime, we firmly believe that the consumption of peanut oil cannot be zero, and hence there will be production due to the existing demand. Of course, production will decline and be the least in all the scenarios. This is shown in Table 6.10.

208

Table 6.10: Production of various edible oils in various scenarios ( in ’000 tonnes)

Base case (2005-06 levels)

WTO binding (300%)

Lahiri Report recommendation

(65%)

Zero tariff

50% tariff reduction (from base

case)

Zero import

Cottonseed oil 803.26 982.21 760.65 576.01 689.63 982.21 Peanut oil 1462.77 1788.65 1385.18 1048.95 1255.86 1788.65

Rapeseed oil 2158.92 2639.90 2044.41 1548.16 1853.54 2639.90 Soya bean oil 986.45 1206.22 934.13 707.39 846.92 1206.22 Coconut oil 382.37 467.55 362.09 274.20 328.28 467.55

Sunflower seed oil

481.01 588.18 455.50 344.93 412.97 588.18

Palm oil 43.22 52.84 40.92 30.99 37.10 52.84 Total

production 6318 7725.55 5982.87 4530.64 5424.32 7725.55

Note: Production shares in 2005–06 are applied

Imports will be as given in Table 6.11. In this case, there will be no imports

under WTO binding and zero-import situations. Palm oil and soya bean oil imports will clearly double during the zero-tariff case, as expected.

Table 6.11: Import of various edible oils in various scenarios (’000 tonnes)

Base case (2005-06 levels)

WTO binding (300%)

Lahiri Report recommendation

(65%)

Zero tariff

50% tariff reduction (from base

case)

Zero import

Palm oil 3044.04 0.00 3771.87 6925.81 4984.92 0.00 Soya bean oil 1727.05 0.00 2139.99 3929.39 2828.22 0.00 Sunflower oil 112.82 0.00 139.79 256.68 184.75 0.00 Coconut oil 21.09 0.00 26.13 47.99 34.54 0.00 Total import 4905.00 0.00 6077.79 11159.86 8032.43 0

6.3.5. The Context of Water Resources On an average, around 22% of the area under oilseeds has been irrigated. We have estimated the average water requirement of oilseeds, considering the four major oilseeds and taking the proportion of irrigated area. Hence, as oilseed areas change, the water use due to irrigation also changes, and the same is given in Table 6.12. The same has been obtained by assuming the average crop water requirement estimated from FAO data and multiplying the same with the area under irrigation.

209

Table 6.12: Irrigation water use for oilseeds in various scenarios (million cubic metres)

Water use Base case (2005-06 levels)

WTO binding (300%)

Lahiri Report recommendation

(65%)

Zero tariff 50% tariff reduction (from

base case)

Zero import

Soya bean 443.33 620.08 401.24 218.87 331.10 620.08 Groundnut 6874.80 9615.86 6222.17 3394.09 5134.44 9615.86 Rapeseed 36108.80 50505.78 32680.95 17826.91 26967.86 50505.78 Sunflower 3369.60 4713.10 3049.72 1663.57 2516.59 4713.10 Total 46796.53 65454.83 42354.07 23103.44 34949.98 65454.83 Water savings

(from base

case)

0 -18658.3 4442.46 23693.1 11846.6 -18658.3

Quite naturally, the maximum water use happens under the WTO regime, and a more liberalized scenario will be associated with lower amount of water consumption as the area diminishes. Eventually, there is water savings as the regime becomes more liberalized, as is evident from Table 6.12. However, if at all there is a movement toward a competing crop in this scenario, and if we assume that the competing crop is wheat, which is an irrigated crop, the situation will be different. In that case, as 88% of the area under wheat is irrigated, the pattern of water use and water savings will change. This is shown in Table 6.13.

Table 6.13: Total irrigation water use for oilseed and wheat and water savings in

each scenario with the base case (Million Cubic Metres) Base case

(2005-06 Levels)

WTO binding (300%)

Lahiri Report recommendation

(65%)

Zero tariff

50% tariff reduction (from base

case)

Zero import

Total Water Use 198262.13 182584.89 201994.80 218169.72 208215.92 182584.89 Water Savings 0 15677.23 -3732.67 -19907.60 -9953.80 15677.23

Because wheat consumes more irrigation water, any process of liberalization that brings about a drop in oilseed area, and a subsequent increase in wheat area, will lead to more water consumption as compared with the base case. Hence, if the shift is to a crop consuming less water (say, ragi), water consumption will be lowered and there will be water savings.

6.3.6. Limitations of the Framework and the Estimates

The estimates obtained here for various scenarios are, indeed, suggestive enough for policy implications. Yet, it is wise to put in place the limitations of the model and of the associated estimates. Most of these limitations have arisen because of a few assumptions that have been considered for the convenience of analysis, while the rest have arisen for the convenience of data availability. Some of these are stated here.

210

First, we assume in the model that export price of oil meal is an exogenous variable and determined by the dynamics of international trade. In the “zero-tariff” case in our scenario analysis, we find that production of soya bean oil has diminished. However, in reality in 2008, we find that the area under soya bean has increased as import tariffs have come down drastically. Now, the increase in area under soya bean in the present circumstances can be attributed to an increase in export demand for soya meal, and not really of soya oil. In our scenario analyses, we are only concerned with a change in tariff, keeping all other variables constant. While our econometric model can capture the reasons of the recent phenomenon of soya bean area increase, the sensitivity analysis has not taken into account the change in the export price of oil meals, but has retained it at 2005–06 levels as the focus of the entire analysis was to solely look at the impact of tariff change, controlling all other impacts. Second, we have assumed that wheat is the competing crop for oilseeds. This is a rather simplistic assumption, and we are aware that this might not be the case in reality for several factors. The first concern is that the agro-climatic and topographical conditions conducive for oilseeds are not always conducive for wheat. Hence, it is not necessary that the farmer can shift his production from oilseeds to wheat. The second issue is related to irrigation and water availability. Oilseeds are usually rain-fed with the exception of rapeseed/mustard, which is primarily irrigated. On the whole, 22% of the area under oilseeds is irrigated. In most cases, therefore, oilseeds are kharif crops. On the other hand, wheat is essentially an irrigated crop, with 88% of the area being irrigated. Hence, shifting from oilseeds to wheat based on these factors might not be a feasible case. Yet, the justification of considering wheat as a competing crop emerges from the fact that at a macro scale, wheat occupies the maximum area under cultivation. Significantly, wheat receives a favourable MSP in relation to many crops, and it has worked successfully. Moreover, wheat is a staple for consumption and generates large-scale demand. More critically, one should not consider wheat as the crop to which the oilseed producer should shift under unfavourable conditions. Rather, in our exercise wheat is merely a representative crop that acts as a proxy for all other crops that could have competed with oilseeds. In that sense, it represents the case for other competing crops like cottonseed which is prominent in the west, and even the case for sugarcane which is prominent in various parts of India.

6.4. Policy Recommendations of the Study The recommendations from this study are simple. We are neither in a position to recommend extreme scenarios like high tariffs where imports will be zero, nor in a position to advise a situation like zero tariff. It is clear that none of these extreme cases will prove beneficial to the economy as a whole, though the situation of zero tariffs is definitely more advantageous than a situation of zero imports. This is because the consumer surplus and processing margins will definitely be better in a zero-tariff situation as compared with zero import or the WTO binding. The impact on government revenue will be severe in both cases.

211

In this context, any movement toward liberalization of tariffs will be more beneficial to edible oil producers and consumers. It might even prove beneficial to farmers if they choose to cultivate the right competing crop substituting oilseeds, and the government also creates supportive mechanisms for such competing crops by allowing for a more favourable minimum support price regime that can help in making a favourable market price for the competing crop. This is what is exhibited in section 3 by taking wheat as the competing crop. We would rather recommend the tariff rates to be line with the Lahiri Committee recommendation, which is emerging as the best case in our analyses. An average tariff rate of 65% marks a moderate liberalization of tariffs from the base of 2005–06 at an average rate of 80%. This shows an improvement in all the variables—even in government revenue from customs duty that is based on the nature of tariff elasticity of imports. Moreover, in all cases, an average rate of 65% customs duty reveals the maximum increase in surplus of all the scenarios. This might also be true for water use, if the farmers shift to crops having lower crop-water requirements. In the process, the study has brought about new dimensions to the policy debate on liberalisation of tariffs on edible oils. First, the analyses show that tariff liberalization has the potential of making the edible oil industry more cost-efficient as the input costs of oilseeds diminish. Second, there is no doubt that there should be tariff liberalization for edible oils. Yet, to obtain the best results, one need not take the extreme route of zero tariffs. Of course, if the government intends to forgo its revenue completely, the zero tariff scenario is a favoured policy route. Third, the impact on water use is still a bit ambiguous at this juncture and solely depends on how the government promotes a competing crop under a liberalized tariff regime. Ghosh and Bandyopadhyay (2009 forthcoming) have shown how the movements of relative prices in favour of water consuming crops have caused excess pressure on water resources. They have further exhibited the important role of minimum support prices in this context. Hence, it is important that the competing crop could be determined from market prices in which the minimum-support-price regime can play an important role. Therefore, just fixing up the tariff rate is not always sufficient. It has to be complemented by monitoring other instruments that should move in the right direction in conjunction with the tariffs to get the best-possible results.

212

References

Ahluwalia, M.S. (1976): “Income Distribution and Development”, American Economic

Review, 66 (2), 128-35. Chand, R. (2002): ‘Trade Liberalization, WTO and Indian Agriculture’, (New Delhi

Mittal Publications). Chand, R., D. Jha and S. Mittal (2004): ‘WTO and Oilseeds Sector: Challenges of Trade

Liberalization’, Economic and Political Weekly, February 7, 2004, pp 533-537. Dohlman, E., S. C. Persaud, and M. R. Landes (2003): “India’ Edible Oil Sector: Imports

Fill Rising Demand”, OCS090301, United States Department of Agriculture, Economic Research Service.

Ghosh, N. and J. Bandyopadhyay (2009-forthcoming): “A Scarcity Value based Explanation of Transboundary Water Disputes in Cauvery Basin”, Water Policy.

Gulati, A., and T. Kelley (1999): Trade Liberalization and India Agriculture, (New Delhi: Oxford University Press).

Hashim, D. A. (2008): “Indian Oilseed Products Industry: Impact of Tariff Reforms on Sources of Output Growth”, Commodity Vision, 1(4), 86-96.

IBRD (1969): Estimation of Long-run Changes in Domestic Demand for Food in LDC’s.

(International Bank for Reconstruction and Development). Kiresur, V.R., S.V.R. Rao and D.M. Hedge (2001): “Improved Technologies in Oilseeds

Production – An Assessment of their Economic Potentials in India”, Agricultural

Economic Research Review, 14(2), 95-108. Lahiri Committee Report (2006): Report of the Committee on Rationalization of Customs

and Excise Duties on Edible Oils and Oilseed, Ministry of Finance, Govt. of India. Mehta, B. V. (2008): “India’s Vegetable Oil Market: Fundamentals and Projection”,

Commodity Vision, 1(3), 56-61. Murthy, K. N. (2001): “Inter-Regional Variation in Consumption Patterns in Semi-arid

Tropical India: A Demand System Approach”, Indian Statistical Institute, Delhi, Discussion Paper.

Pavaskar, M and A. Kshirsagar (2008): “Paradox in India’s Trade in Oilseed Products”, Commodity Vision, 1(3), 48-54.

Persaud, S. and M. R. Landes (2006): The Role of Policy and Industry Structure in

India’s Oilseed Markets, United States Department of Agriculture, Economic Research Report, Number 17, April.

Solvent Extractors’ Association of India: Various Annual Report, SEA, Mumbai. Srinivasan, P. V. (2005): Impact of Trade Liberalisation on India’s Oilseed and Edible

oils Sector, Report prepared for IGIDR-ERS/USDA Project: Indian Agricultural Markets and Policy, Indira Gandhi Institute of Development Research, February 2, Mumbai.

World Bank (1997): India-The Indian Oilseeds Complex: Capturing Market

Opportunities, Report No. 15677-IN, Rural Development Sector Unit, South Asia Region, 31 July.

213

CHAPTER 7

Agricultural Market Integration in India – An Analysis of Select Commodities*

7.1 Introduction The issue of market integration is central to many contemporary debates relating to trade liberalization, price policy and reform of the state trading practices in the food markets of the developing countries. Markets aggregate demand and supply of individual economic actors distributed across space. In the absence of spatial market integration, price signals are not transmitted from the deficit to surplus regions and as a consequence, prices may become more volatile. In such a scenario, agricultural producers do not specialize according to long-term comparative advantage and the gains from trade will not be realized. At the national level, well integrated markets ensure that the macro level economic policies change the incentive structure faced by economic agents at the micro level. For example, in a well integrated domestic food market, a producer in the surplus region can increase investment in response to price signals emanating from deficit in other region(s). On the other hand, poorly integrated markets stifle the prospective gains from investment and technological change in response to market stimuli. Market integration can also play an important role in risk management associated with demand and supply shocks, through adjustment of net export flows across space, thereby reducing price variability faced by consumers and producers. At the international level, effects of monetary policy, exchange rate adjustment and trade policy depend crucially upon price equilibrating mechanism across countries. A study of dynamic market integration should inform the policymakers about the length of the persistence of localized shortages which assume critical importance during droughts and famines. A non-interventionist policy in the food markets is advocated generally based on the assumption that the grain traders’ response to spatial price differences will quickly eliminate a local scarcity. However, there are some important counter arguments to this – the incompatibility between the profit-maximizing behaviour of the private trade vis-à-vis the social objectives (Sen, 1981), that the markets are too slow to respond and (Ambirajan, 1978) and that the localized scarcity sometimes may not manifest in price increases particularly if it is accompanied by large-scale demand contraction (Sen, 1981). The last argument merits more serious attention since it is not uncommon to find instances of widespread decreases in income during severe droughts and a consequent decline in effective demand, which in turn, results in a situation where the supply scarcity does not lead to a price rise thereby precluding the market response as envisaged.

* This Chapter has been contributed by CSC Sekhar, Associate Professor, Institute of

Economic Growth, University of Delhi, Delhi

214

Summing up, integrated markets ensure transmission of macro policy change to the micro level, facilitate increased investment in response to price signals from deficit markets and also play an important role in risk management through adjustment of net export flows across space. In the absence of spatial market integration, price signals are not transmitted from the deficit to surplus regions and as a result, prices become more volatile, producers do not specialize according to long-term comparative advantage and gains from trade will not be realized. However, integrated markets are beneficial only in cases of price differentials induced by supply shocks. When scarcities do not result in price increases (possibly due to concomitant demand contraction), role of markets becomes limited. However, market integration still remains an important issue to investigate in cases where supply shocks induce spatial price differentials. An empirical assessment of the nature and speed of market adjustment to spatial price differentials is therefore imperative. We have attempted to undertake this exercise in the present study with special focus on agricultural markets. The objective of the study is not to assess the effect of trade liberalization on market integration. Rather the objective is to assess the potential role of market in mitigating the adverse effects of supply shocks. Therefore, we have not attempted to test for market integration before and after the reforms, as is the practice in most of the studies. The main idea of the study is to test the transmission of international price movements into domestic markets at the micro level. This is done in two stages.

i) Testing integration among the important domestic markets and also their integration with international market

ii) Assess the degree of market integration iii) Draw policy implications for trade liberalization and parastatal reforms

A competitive general equilibrium exists in an economy consisting of a set of N regions among which trade occurs at fixed transport costs (Takayama and Judge, 1971). In this model, if trade takes place between any two regions, then the price in the importing region equals the price in the exporting region plus the unit transportation cost. If this condition holds, then the markets are said to be ‘integrated’. However, market integration is not a sufficient condition for the Pareto optimality of the competitive equilibrium, in the sense that, it does not by itself, imply efficient allocation of resources (Newbery and Stiglitz 1984). Nonetheless, market integration remains a necessary condition for Pareto optimality of the spatial competitive equilibrium and hence the interest in empirical testing of market integration. Much of the problem in market integration analysis revolves around the concept of ‘market integration’ itself and the empirical evidence needed to demonstrate that condition. There are essentially two strands of thought. In the first, mainly in the macroeconomics and international economics literature, the conceptualization of market

215

integration is centered around tradability i.e. transfer of excess demand from one market to another through actual or potential physical flows. Positive trade flows are sufficient to demonstrate that markets are integrated and prices need not equilibrate across markets. It is therefore clear that this concept implies a Pareto inefficient distribution. Therefore, we propose to follow the broad framework of Enke-Samuelson-Takayama-Judge (ESTJ) spatial equilibrium model (Enke 1951, Samuelson 1952, Takayama and Judge 1971). In this model, the dispersion of price in two locations for an otherwise identical good is bound from above by cost of arbitrage when trade is unrestricted and, from below when trade quotas exist. If trade occurs and is unrestricted, then marginal trader earns zero

profits and prices in the two markets co-move perfectly. This is the condition we attempt to test in this study. 7.2 Literature Review Testing for market integration has evolved over time from the early stages of using bivariate correlation coefficients to the more recent techniques that take into account non-stationarity, common trends and endogeneity of food prices. The literature on market integration can be divided into three broad categories.

i) Early methods based on bivariate correlation coefficients (Lele 1967, Jones 1968) and their critiques (Blyn 1973, Harris 1979)

ii) Methods accounting for non-stationarity of food prices – Granger causality

(Gupta and Mueller, 1982), Error-correction models (Ravallion 1986, Goletti 1994), Engle-Granger cointegration (Dercon 1995, Jha et.al 2003), Johansen cointegration and common trends (Gonzalez and Helfand 2001, Wilson 2001, Jha et.al 2005) and panel unit roots (Jayasuriya, 2008)

iii) Methods emphasizing on transfer costs instead of comovement of prices

(Barett 1996, Baulch 1997) The early work on market integration applied measures of static price correlations or regression coefficients between time series of spot prices at different markets (Lele 1967, Jones 1968). Another less common method was of calculating the spatial variance of spot prices and test for its long-run convergence to zero or close to zero (Hurd, 1975). Ravallion (1986) highlighted the inferential problems with these methods and proposed an alternate multivariate model of market integration within which various market integration conditions can be formulated and tested as nested hypotheses. This model is the first major attempt to study market integration in a systematic manner. The model is a dynamic formulation of the following static price formation model among N markets where market 1 is the central market.

1 1 2 1( , ,........, , )N

p f p p p X= ---------------- (1)

1( , )i i

p f p X= ---------------- (2)

where i

X (i=1……N) is a vector of other exogenous variables influencing the i th market.

216

The dynamic reformulation of the model for T periods and N variables is as follows.

1 1, 1 , 1 1 1

1 2 0

n N nk

t ij t j j k t j t t

j k j

p a p b p X c e− −= = =

= + + +∑ ∑∑ ---------------- (3)

, 1,

1 0

n n

it ij i t j ij t j it i it

j j

p a p b p X c e− −= =

= + + +∑ ∑ ---------------- (4)

where a’s, b’s and c’s are the parameters to be estimated and e’s are the error terms. A wide range of possible hypotheses can be formulated and tested as parameter restrictions on eqn (4).

i) Market Segmentation: 0ijb = for j=0,..…………,n

ii) Short-Run Market Integration: A price increase in the central market is

immediately transmitted to the i th market if 0 1i

b =

a) Weak Short-Run Integration : when lagged effects are present, that

is, if either 1

0n

ij

j

a=

≠∑ or 1

0n

ij

j

b=

≠∑

b) Strong Short-Run Integration: No lagged effects if 1 1

0n n

ij ij

j j

a b= =

= =∑ ∑ .

In such cases the i th market is integrated with the central market within one time period.

iii) Long-Run Market Integration: A long-run equilibrium is one in which

market prices are constant over time, undisturbed by the local stochastic

effects. In this case eqn (4) takes the form *

it ip p= , *

1 1tp p= and all

0it

e = . Then eqn (4) reduces to

* *

1

0

1

1

n

i ij it i

j

n

ij

j

p p b X c

a

=

=

= +

−

∑

∑---------------- (5)

Long-run market integration now requires that the coefficient of

*

1p =1⇒ 0

1

1

1

n

ij

j

n

ij

j

b

a

=

=

=−

∑

∑

217

1 0

1n n

ij ij

j j

a b= =

⇒ + =∑ ∑ ---------------- (6). This is the condition for long-run market

integration. This implies that if the conditions for short-run integration described by (ii b) which is,

0 1i

b = and 1 1

0n n

ij ij

j j

a b= =

= =∑ ∑ hold, then the above long-run integration condition (eqn 6) is

satisfied but the reverse is not true. After this influential work by Ravallion (1986), with the advances in time-series techniques, methods using Johansen cointegration gained popularity (Gonzalez and Helfand 2001, Wilson 2001, Yavapolkul et.al 2004). In this framework, spatial price analysis is carried out by testing for long term cointegration and short-run error correction mechanism by using price data at different geographic locations. Some of the underlying assumptions of the model are the linear pricing and stationary transaction costs. This methodology has drawn criticism on the grounds that the tests for market integration hypothesis become indistinguishable from tests of these abovementioned assumptions (Barett 1996, Baulch 1997, Fackler and Goodwin 2001). Alternate methods were proposed in the literature to overcome these limitations using mixture distribution estimation methods, integrating price data with transactions cost data and trade volume data (Baulch 1997, Barett and Li 2002). However, there are major limitations to these methods also since they rely on arbitrary distributional assumptions in estimation and typically ignore the time series properties of the data. Also, they do not permit the analysis of the dynamic adjustment of short-run deviation from long-run equilibrium and potentially important distinction between short-run and long-run market integration (Ravallion 1986). One recent strand of literature uses panel unit root methods in combination with half life-cycle estimation of convergence to assess the nature and degree of market integration (Jayasuriya et.al. 2008). The panel unit root methods are expected to exploit the larger (and richer) panel datasets. However, there are some limitations to this approach. Firstly, panel unit root tests can not disentangle the speed of convergence of individual time series (markets), which is a very important policy issue in market integration. Secondly, half-life estimation is based on Impulse Response Functions (IRF) and IRF s are not uniquely distinguishable when shocks to the system are correlated, which is precisely the case in spatially separated markets. A shock to one market is expected to be transmitted to other markets as well, depending on the degree of integration. Also, IRF s are not invariant to the ordering of the variables. In view of these limitations, we propose an alternate empirical methodology which is described in the next section

218

7.3 Methodology The proposed methodology is broadly in the Johansen framework of cointegration and error-correction with the following additional features.

a) Identifying a common integrating factor based on the Gonzalo and Granger (1995) methodology. This methodology is particularly attractive because it allows identification of the location(s) that contribute to the long-run behaviour of the prices through the common integrating factor, which is based on observable variables.

b) Measuring the degree of market integration, defined as the reaction time to

remove the disequilibrium or shock, by means of ‘Persistence Profiles’ (Pesaran and Shin 1996). A persistence profile characterizes the response of a cointegrating relation to a system-wide, rather than, to an individual shock. It measures the reaction time of each of the long-run equilibrium relations to absorb a system-wide shock.

7.3.1. Testing for Market Integration: Gonzalo-Granger Methodology A common integrating factor is identified using the Gonzalo and Granger (1995) methodology. This methodology is particularly attractive because it allows identification of the location(s) that contribute to the long-run behaviour of the prices through the common integrating factor, which is based on observable variables.

Let t

X be a vector of order ( 1)P × and each element of t

X be I(1) with mean 0. Let the

elements of t

X be cointegrated with rank r ⇒ there exists a matrix p r

α × of rank r s.t.

'

tXα is I(0). There are two features of

tX .

i) t

X has an error-correction (ECM) representation of the following form -

'

1

1r pt p r t i t i t

i

X X Xγ α ε×

∞

× − −=

∆ = + Γ ∆ +∑ where ∆ =I-L with L being the lag operator.

ii) The elements of t

X can be expressed in terms of a smaller number (p-r) of I(1)

variables, called the common factors and denoted by t

f , plus some stationary I(0)

components.

1 1 11p p k k pt t tX A f X

× × × ×= + % where k = p-r.

This is very similar to standard factor analysis. In factor analysis the main objective is to

estimate the loading matrix 1A and the number of common factors k. In the ECM

219

representation above, these two are already known - 1A is any basis of the null space of

the cointegrating vector α such that '

1 0Aα = and k is equal to p-r where r is the

cointegrating rank. The goal of the G-G methodology is to estimatet

f . The standard

restrictions used in factor analysis are not adequate in this case since the data are non-

stationary. The following additional conditions are imposed to identifyt

f .

a) t

f is a linear combination of elements of t

X

b) 1 tA f and

tX% form the permanent and transitory components respectively of

tX .

According to this condition the only shocks that can affect the long-run forecast of t

X

are those coming from pt

u - the error term of the permanent component 1t tP A f= . From

these conditions, after substantial manipulation, we obtain 1 1

'

k k p pt L tf Xγ

× × ×= such that

' 0L

γ γ = and k=p-r

The t

f s are the linear combinations of t

X∆ that have the common feature of not

containing the levels of the error-correction term 1tX − in them. Once

tf s are identified,

inverting the matrix '( , )L

γ α we obtain the permanent-transitory decomposition of t

X .

1 1 1

' '

1 2p p k k p p p r r p pt L t tX A X A Xγ α

× × × × × × ×= +

where ' 1

1 ( )p k p k k p p kL L L

A α γ α× × × ×

−= , ' 1

2 ( )p r p r r p p r

A γ α γ× × × ×

−=

Therefore the decomposition of t

X into permanent and transitory components is complete

1 1 11 2p p k k p r rt t tX A f A z

× × × × × = + in which the first term is the permanent component and the

second is the transitory component. 1 2 and A A are as described above.

1 1

'

k k p pt L tf Xγ

× × ×= = I(1) Common Factor

1 1

'

r r p pt L tz Xα

× × ×= = I(0) Stationary Component

p rγ

× = Weights (speed of adjustment) of the cointegrating equations

p kLγ

×= Orthogonal matrices of γ

p rα

×= Cointegrating vectors

p kLα

×= Orthogonal matrices of α

p= order of t

X (number of elements in vector t

X ), r= number of cointegrating vectors,

k=p-r

220

As may be seen, the order of t

f is 1k × where k=p-r. If r=p-1, then there is a single

common factor i.e. t

f is a single element vector. In case of integrated markets, we expect

to find a single common factor (of prices) driving the entire system. Here it is important to note that a single factor does not imply a single market but a single linear combination

of a subset of markets oft

X . Therefore for a set of n markets to be integrated, we should

have n-1 cointegrating vectors. We make use of this property to identify the set of

integrated markets in this study.

7.3.2 Degree of Market Integration The degree of market integration is traditionally assessed by looking at the adjustment

coefficient (γ ) or the statistical significance of the lag structure i

Γ . To jointly evaluate

the effects of innovations in one equation on the entire system represented by the VECM, impulse response functions (IRF) are normally used. IRF s trace the impact of a shock in location j on the price of i over time. The main drawback of the IRF s is that they are not unique when errors in different equations are correlated. In a study of spatial market integration, it is highly probable that the errors at different locations are correlated since the movements in prices at these locations are generally correlated. The solution attempted in the literature to overcome this problem is the Cholesky decomposition, but this solution is crucially dependent on the ordering of the variables that enter the VECM. Changing the order of the variables can significantly alter the IRFs. Therefore, we attempted the persistence profile (PF) approach of Pesaran and Shin (1996). A PF characterizes the response of a cointegrating relation to a system-wide shock. PF s are unique functions which allow us to quantify the degree of integration among the locations that belong to the same economic market.

A persistence profile characterizes the response of the cointegrating relation '

t tZ Xα= to

a system-wide, rather than to an individual shock, where the response is measured in units of variance. The system-wide shocks may be understood as shocks drawn from a

multivariate distribution of t

ε s without any need to orthogonalize the shocks.

The persistent profile (unscaled) is defined as follows

( ) ( ) ( )1 1 1X t n t t n tH n V X I V X I+ − + − −= − for n= 0,1,2,……., where H(.) denotes the

variance upto time horizon n, conditional on the information available upto t-1 ( )1tI − .

Since ( ) ( ){ } ( )1 1 1 1t n t t n t n t t t n tV X I V X E X I I V Iη+ − + + − − + − = − = where

t nη + is the n-step

ahead forecast error oft

X , PF can also be viewed as the change in variance of forecasting

t nX + as compared to that of forecasting 1t n

X + − given the information 1tI − . A scaled

persistent profile ( )Xh n is defined as ( ) ( ) ( )/ 0X X Xh n H n H= which has a value of unity

on impact i.e. at n=0.

221

The concept of PF of a variable t

X can be easily extended to that of cointegrated system

tZ The information content of the whole time profile of ( )zh n is of much greater interest

in economic analysis than the mere knowledge that 't t

Z Xα= forms a cointegrating

relation. The profile of ( )zh n , as n allowed to increase, provides important information

on the speed with which the effect of a system-wide shock on the cointegrating relation

't t

Z Xα= disappears (even through these shocks generally have lasting effects on each

of the individual variables in the vector t

X ). Also looking at ( )zh n for n=0,1,2,… allows

us to discern the short-run dynamic properties of the cointegrating vector as well as their long-run properties. 7.3.3 Commodity Coverage and Period of Analysis The commodities showing major share in trade volumes are shown below. Out of these, we have selected rice, castor oil, oilcakes (castor, groundnut and mustard), tea and coffee on the export side for our analysis. On the imports side we have selected gram, groundnut oil and mustard oil. The selection of commodities is based not only on their importance in the trade basket but also on the availability of long time series of official published data on monthly domestic prices. Therefore, although palm oil and soybean oil are prominent edible oils on the import side, they could not be selected due to non-availability of monthly price data. However, this need not be a major limitation because of the high degree of substitutability on the consumption side. This is proved by the fact that although palm oil and soybean oil are the oils that are imported the most, the crops that suffered the maximum decline in area after edible oil import liberalization are groundnut and rapeseed-mustard.

Import and Export Share of Commodities (in %)

Commodity Average

(2000/01-2006/07)

Imports

Cereals and preparations 1.66

Wheat 1.08

Pulses 12.23

Edible oils 44.70

Total 59.67

Exports

Tea 4.37

Coffee 3.18

Rice 12.80

Nuts and seeds 8.27

Oil meals 8.04

Marine products 16.18

Cotton raw incl. Waste 3.78

Total 56.63

222

Source: Agricultural; Statistics at a Glance 2007, DES, Ministry of Agriculture, GOI

The period of analysis is from 1986/87-2006/07, subject to data availability for individual commodities. The study period starts from 1986/87 since some major changes started occurring in edible oil markets around that time. The data used is monthly market price data for the major domestic and international markets. This domestic data is not readily available in machine-readable form and has been collected manually from various publications, mainly Agricultural Situation in India and Agricultural Prices in India. The international price data has been collected from the International Financial Statistics (IFS) of the International Monetary Fund (IMF). 7.4 Results The correlation coefficient matrices of the market prices of rice, gram, groundnut oil and mustard oil are presented in Appendix A. As can be seen, the correlation coefficients are quite high, and mostly above 0.8. However, high correlation coefficients are not sufficient to conclude that the markets are integrated. Hence, the need for a more refined method of testing for market integration. The crucial insight provided by the Gonzalo-Granger model (GG) is that, in an integrated market, there is a single common factor (of prices) driving the entire system. Therefore, for a set of n markets to be integrated, we should have n-1 cointegrating vectors. Using this result we start with an appropriately large number (n1) markets out of a total number of n markets such that n1<n. We then test for the number of cointegrating vectors in this set of markets. If the number of cointegrating vectors is n1-1, then we add one more market to the set and again test for the number of cointegrating vectors. If the number of cointegrating vectors is still n1-1(instead of n1) then we drop the newly included market and try another market. This process is continued till we have the largest set (say n2) of integrated markets. If all the n markets are integrated, then n2=n. We used the following sequential procedure to test for market integration. We tried to test in the following order.

i) Markets within a state ii) Markets within a region (south, west, north, east) iii) Markets in the entire country

We have used the following vector error-correction (VECM) model for each commodity -

'

1

1r pt p r t i t i t

i

p p pγ α ε×

∞

× − −=

∆ = + Γ ∆ +∑ where ∆ =I-L with L being the lag operator and t

p is a

vector of prices at n markets (therefore of order 1n × ). Other parameters are as explained in section 7.3.1. We have also included WPI of the commodity to account for general price movements in the commodity and monthly dummies to account for seasonal fluctuations in market prices.

223

The commodity-wise results are as follows. First commodities on the export side are discussed after which we turn to those on the import side In case of rice, the markets within each of the major states are all well integrated. However regionally, south and west show better integration than north and east (Table 7.1). In the south, out of a total of 13 markets tested, 11 are integrated (12 cointegrating vectors) and the corresponding numbers for the western region are 8 and 6 respectively. However, in Eastern India, out of a total number of 15 markets only 11 are integrated and in the northern region, out of a total of 8 markets only 5 are integrated. In all, out of a total of 44 markets all over India, 33 (75%) are integrated. We have also tested for integration of each of the regional markets with international rice market, namely Bangkok market. On this criterion again, southern and western regions perform the best. In the southern region all the 11 domestically integrated markets are also integrated with the international market. In the western region, all the 6 domestically integrated markets are integrated with international market as well (Table 7.2). North follows next with 4 of the 5 markets being integrated with the international market. Eastern region shows relatively poor integration with international market. In this region, out of a total of 11 domestically integrated markets, only 8 are integrated with the international market. On the side of imports, gram appears to be completely integrated, both regionally and nationally (Table 7.3). In the western, northern and eastern regions there are in all 5, 7 and 5 markets respectively and all of them are integrated. Similar is the case with the two edible oils. All the 8 markets of groundnut oil are integrated and all the 5 markets of mustard oil are also integrated (Table 7.4). As regards integration with the international market, in case of groundnut oil although all the 8 domestic markets are integrated, the integration with international market is not complete. The Calcutta market of West Bengal appears to be non-integrated with the international market. In case of mustard oil, the data on the international price of mustard oil is not available. Therefore, we have used the price of soybean oil in our analysis. The results show that all the 5 domestically integrated mustard oil markets are integrated with the international market as well (Table 7.4). On the side of exportable commodities, castor oil and castor oilcake, with 4 and 3 markets respectively are well integrated (Table 7.5). Similarly, the mustard oilcake and groundnut oilcake markets, with 4 markets each, are well integrated. In the case of coffee and tea, since there is only one market (Coimbatore) for which a long time series data on wholesale prices is available, domestic market integration is not relevant. Therefore, we have tested only for integration with the international market (Table 7.6). For coffee, we tested with four international prices and the Indian coffee price is found to be integrated with three of them. Only the Ugandan coffee price shows movements different from that of the Indian market. In case of Tea, we have tested with Sri Lankan and London markets and the Indian market is found to be integrated with both these international markets.

224

Table 7.1: RICE - Domestic Market Integration

Region λ trace No of C.V

s

No of Integrated Domestic Markets

Total No of Domestic Markets

No of Non-Integrated

Domestic Markets

South 15.63* 10 11 13 2

West 18.54* 5 6 8 2

North 18.58* 4 5 8 3

East 5.83* 10 11 15 4

All-India 33 44 11

Table 7.2: RICE - Integration with International Market

Region λ trace No of C.V

s

No of Integrated Markets with Int. Market

No of Integrated Domestic Markets

No of Non-Integrated

Domestic Markets

South 75.86* 11 11 11 0

West 19.89* 6 6 6 0

North 31.27* 4 4 5 1

East 75.86* 8 8 11 3

All-India 26 33 7

Table 7.3: Gram - Domestic Market Integration

Region λ trace No of C.V

s

No of Integrated Domestic Markets

Total No of Domestic Markets

No of Non-Integrated

Domestic Markets

West 60.14** 4 5 5 0

North 47.07** 6 7 7 0

East 19.15** 4 5 5 0

All-India 15.89** 16 17 17 0

Table 7.4: Edible Oils - Domestic and International Market Integration

Commodity λ trace No of C.V s

No of Integrated Domestic Markets

Total No of Domestic Markets

No of Non-Integrated

Domestic Markets

Domestic Integration Groundnut

Oil 23.53** 7 8 8 0

Mustard Oil 37.78** 4 5 5 0

Integration with International Market

Groundnut 30.02* 7 7 8 1

225

Oil

Mustard Oil 21.86* 5 5 5 0

Table 7.5: Oilcakes - Domestic Market Integration

Commodity λ trace No of C.V

s

No of Integrated Domestic Markets

Total No of Domestic Markets

No of Non-Integrated

Domestic Markets

Castor Oil 23.78** 3 4 4 0

Groundnut Oilcake 25.88

** 3 4 4 0

Mustard Oilcake 27.49

** 3 4 4 0

Castor Oilcake 43.90

** 2 3 3 0

Table 7.6: Tea & Coffee - Integration with International Markets

Commodity λ trace No of C.V s

No of Integrated

International Markets

Total No of International

Markets

No of Non-Integrated

International Markets

Coffee 3 3 4 1

Brazil 21.85** 1 1

Brazil (New York) 15.41

* 1 1

Other Milds 15.41** 1 1

Uganda 9.93 0

Tea 2 2 0

Sri Lanka 15.41** 1 1

London 15.41** 1 1

7.5 Degree of Integration The plots of the persistence profiles are provided in figures 7.1(a) to 7.10(b). Two graphs are given for each commodity – the first one illustrating the response of the first cointegrating vector that corresponds to the maximum eigen value and second one shows the response of all the r cointegrating vectors. The PF plot of rice shows that the adjustment for the 1 st cointegrating vector from 1 unit shock is complete within a time period of one month in the southern and eastern regions of the country (Figs 7.1a, 7.4a). In western India, the period is slightly longer with 3 months (7.2a) followed by northern India (7.3a). However, the adjustment of the entire

226

system of cointegrating vectors seems to take much longer – with the average period of adjustment of about 10 months. In the case of gram, the adjustment of the first cointegrating vector appears to take 1-3 months (7.5a, 7.6a, 7.7a) in the west, north and east. At the all-India level also the pattern is similar (fig 7.8). The PF plots for the entire system of cointegrating vectors shows that the gram markets are reasonably integrated (7.5b, 7.6b, 7.7b). It takes about 5 to 15 months for any price shock to dissipate at the regional level. The western and northern regions show a maximum of 5 months for any disequilibrium to dissipate while the period is longer in the eastern region (15 months). Overall, the maximum period at the all-India level is about 10 months. In the case of groundnut oil, the period of adjustment varies between 1 to 20 months with the average being 8 months (7.9a, 7.9b). The period of adjustment for the first cointegrating vector is about one month. The degree of market integration for the mustard oil market is much higher with the average period of adjustment being 3 months (7.10a, 7.10b). The maximum period of adjustment is for the first cointegrating vector (2 months).

shock to CV'(s)

CV1

Horizon

0.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 3030

Figure 7.1a: Persistence Profile of the effect of a system-wide shock - Rice (South India)

CV1

CV2

CV3

CV4

CV5

Horizon

0.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 30 35 40 45 48

Figure 7.1b: Persistence Profile of the effect of a system-wide shock- Rice (South India)

227

shock to CV'(s)

CV1

Horizon

0.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 3030

Figure 7.2a: Persistence Profile of the effect of a system-wide shock - Rice (West India)

shock to CV'(s)

CV1

CV2

CV3

CV4

CV5

Horizon

0.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 3030

Figure 7.2b: Persistence Profile of the effect of a system-wide shock - Rice (West India)

shock to CV'(s)

CV1

Horizon

0.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 3030

Figure 7. 3a: Persistence Profile of the effect of a system-wide shock - Rice (North India)

shock to CV'(s)

CV1

CV2

CV3

CV4

Horizon

0.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 3030

Figure 7.3b: Persistence Profile of the effect of a system-wide shock - Rice (North India)

228

Figure 7.5a: Persistence Profile of the effect of a system-wide shock - Gram (Western Region) shock to CV'(s)

CV1

Horizon

0.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 3030

CV1

CV2

CV3

CV4

Horizon

0.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 3030

Figure 7.5b: Persistence Profile of the effect of a system-wide shock

CV1

Horizon

0.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 3030

Figure 7.4a: Persistence Profile of the effect of a system-wide shock - Rice (East India)

shock to CV'(s)

CV1

CV2

CV3

CV4

CV5

Horizon

0.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 3030

Figure 7.4b: Persistence Profile of the effect of a system-wide shock - Rice (East India)

229

Persistence Profile of the effect of a system-wideshock to CV'(s)

CV1

Horizon

0.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 3030

Figure 7.6a: Persistence Profile of the effect of a system-wide shock - Gram (Northern Region)

Persistence Profile of the effect of a system-wideshock to CV'(s)

CV1

CV2

CV3

CV4

CV5

CV6

Horizon

0.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 3030

Figure 7.6b: Persistence Profile of the effect of a system-wide shock

- Gram (Northern Region)

Persistence Profile of the effect of a system-wideshock to CV'(s)

CV1

Horizon

0.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 3030

Figure 7.7a: Persistence Profile of the effect of a system-wide shock - Gram (Eastern Region)

Figure 7.7b: Persistence Profile of the effect of a system-wide shock - Gram (Eastern Region)

CV1

CV2

CV3

CV4

Horizon

0.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 3030

230

CV1

Horizon

0.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 3030

Figure 7.9a: Persistence Profile of the effect of a system-wide shock - Groundnut (All India)

shock to CV'(s)

CV1

CV2

CV3

CV4

CV5

CV6

CV7

Horizon

0.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 3030

Figure 7.9b: Persistence Profile of the effect of a system-wide shock

shock to CV'(s)

CV1

Horizon

0.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 3030

Figure 7.10a: Persistence Profile of the effect of a system-wide shock - Mustard Oil (All India)

CV1

CV11

Horizon

0.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 3030

Figure 7.8: Persistence Profile of the effect of a system-wide shock - Gram (All India)

231

7.6 Conclusions Our results indicate that the rice markets are well integrated within the states and also regionally but the integration at the national level is not satisfactory. This result is also in agreement with Jha et.al (2005) but is not in correspondence with Jayasuriya (2008), possibly due to the methodological differences. Out of the domestically integrated markets, majority are also integrated with the international market. In case of gram, edible oils and oilcakes, the markets are well integrated domestically. Edible oils, tea and coffee, for which integration with international market could be tested, have been found to be integrated with international markets. The assessment of the degree of integration, using the persistence profile approach, shows that in rice markets the speed of adjustment is relatively longer. In the other markets, the period of dissipation of a price shock is much shorter. The analysis shows that the markets for groundnut oil, mustard oil and gram are reasonably well-integrated and for rice a large number of markets are integrated although the speed of adjustment is longer. Therefore, the policy implication is that the market

mechanism can play a reasonably effective role in case of domestic production shocks. However, sometimes the production shocks may not lead to price increases if such shocks are accompanied by demand contraction, as is common during severe droughts and also like the one witnessed in 2002/03. In such cases the desired response may not be forthcoming from the private trade and the role of the state becomes critical.

CV1

CV2

CV3

CV4

Horizon

0.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 25 3030

Figure 7.10b: Persistence Profile of the effect of a system-wide shock - Mustard Oil (All India)

232

References Ambirajan, S (1978), Classical Political Economy and British Policy in India, Cambridge: Cambridge University Press, 1978. Barrett, C. (1996), Market analysis methods: are our enriched toolkits well suited to enlivened markets? American Journal of Agricultural Economics, 78, 825–9. Barrett, C. (2001), Measuring integration and efficiency in international agricultural markets. Review of Agricultural Economics 23, 19–32. Barrett, C. and Li, J. (2002), ‘Distinguishing between equilibrium and integration in spatial price analysis’, American Journal of Agricultural Economics 84, 292–307. Baulch, B. (1997), Transfer costs, spatial arbitrage, and testing for food market integration. American Journal of Agricultural Economics 79, 477–87. Blyn, G (1973). ‘Price Series Correlation as a Measure of Mar- ket Integration’, Indian

Journal of Agricultural Economics, 28(1973):56- 59. Dercon, S (1995), ‘On Market Integration and Liberalization: Method and Application to Ethiopia’, Journal of Development Studies, 32 (0ctober 1995). Enke, S. (1951), Equilibrium among spatially separated markets: solution by electrical analogue. Econometrica 19, 40–7. Fackler, P. and Goodwin, B. (2001), Spatial price analysis. In Handbook of Agricultural

Economics, vol. 1B, ed. B. Gardner and G. Rausser. Amsterdam: Elsevier. 7

Goletti, F. (1994) “The Changing Public Role in a Rice Economy Approaching Self-Sufficiency: The case of Bangladesh”, Research Report no.98, International Food Policy Research Institute, Washington, DC.

Gonzalez-Riviera, G., Helfand S.M. (2001), ‘The Extent, Pattern and Degree of Market Integration: A Multivariate Approach for the Brazilian Rice Market’, American Journal of Agricultural Economics, 83(3), 576-592 Gonzalo J., and C.W.J. Granger (1995), ‘Estimation of Common Long Memory Components in Cointegrated Systems’, Journal of Business & Economic Statistics, Vol 13, No. 1, Jan 1995, pp 27-35 Gupta, S., and R. Mueller (1982), ‘Analysing the Pricing Efficiency in Spatial Markets; Concept and Application’, European Review of Agricultural Economics, 9(1982):24-40. Harriss, B (1979). ‘There Is Method in My Madness: Or Is It Vice Versa? Measuring Agricultural Market Performance’, Food Research Institute Studies, 17(1979):197-218.

233

Hurd, J (1975). ‘Railways and the Expansion of Markets in India, 1861-1921’, Explorations in Economic Hisory , 12(1975):263-88. Jayasuriya, Sisira, Jae K. Him and Parmod Kumar (2008), ‘Opening to World Markets: A Study of Indian Rice Market Integration’, paper presented at the NCAER-ACIAR Conference, New Delhi, )6 June 2008. Jha, R., Murthy, K., Nagarajan, H. and A. Seth (2003) “Market Integration in Indian Agriculture” Economic Systems, 21(3): 217–34. Jha, R., Murthy, K.V. Bhanumurthy and Anurag Sharma (2005), ‘Market Integration in Wholesale Rice Markets in India’, ASARC Working Paper 2005/03 Jones, W.O (1968), ‘The Structure of Staple Food Marketing in Nigeria as Revealed by Price Analysis’, Food Research Institute Studies, 8 (1968):95-124. Lele, U (1967). ‘Market Integration: A Study of Sorghum Prices in Western India’, Journal of Farm Economics, 49 (February 1967): 149-59. Newbery, D., and J. Stiglitz (1984), Pareto Inferior Trade, Review of Economic Studies. 51(1984):1-12. Pesaran M.H., and Y. Shin (1996), ‘Cointegration and Speed of Convergence to Equilibrium’, Journal of Econometrics, 71, pp. 117-43 Ravallion, M. (1986), Testing market integration. American Journal of Agricultural

Economics 68, 102–9. Samuelson, P. (1952), Spatial price equilibrium and linear programming. American

Economic Review 42, 283–303. Sen, A. K (1981), Poverty and Famines: An Essay on Entitlement and Deprivation. Oxford: Oxford University Press, 1981. Takayama, T. and Judge, G. (1971), Spatial and Temporal Price Allocation Models. Amsterdam: North-Holland. Wilson, Edgar J., (2001), ‘Testing Agricultural Market Integration: Further Conceptual and Empirical Considerations Using Indian Wholesale Prices’ in Indian Agricultural

Policy at the Cross-Roads (ed) S.S. Acharya and D.P.Chaudhari, Rawat Publishers, New Delhi Yavapolkul, Navin., Munisamy Gopinath and Ashok Gulati (2004), ‘Post Uruguay Round Price Linkages Between Developed and Developing Countries: The Case of Rice and Wheat Markets’, MTID Discussion Paper No 76, International Food Policy Research Institute, Washington, D.C.

234

APPENDIX A RICE-ALL INDIA

HYDERABAD KOZHIKODE TRIVANDRUM SHIMOGA KUMBKONAM MADRAS TIRACHIRAPALLI TIRUNALVELI KAKINADA NIZAMABAD NAGPUR DURG RAIPUR BULSAR JAMSHEDPUR ARR

HYDERABAD 1 0.98 0.90 0.98 0.92 0.82 0.83 0.89 0.96 0.94 0.97 0.97 0.95 0.98 0.97 0.93

KOZHIKODE 0.98 1 0.90 0.98 0.92 0.83 0.85 0.90 0.97 0.95 0.98 0.98 0.97 0.98 0.97 0.93

TRIVANDRUM 0.90 0.90 1 0.90 0.87 0.96 0.81 0.89 0.92 0.95 0.90 0.89 0.93 0.93 0.93 0.80

SHIMOGA 0.98 0.98 0.90 1 0.90 0.83 0.86 0.90 0.97 0.94 0.98 0.98 0.96 0.98 0.97 0.94

KUMBKONAM 0.92 0.92 0.87 0.90 1 0.81 0.83 0.85 0.91 0.91 0.90 0.91 0.92 0.91 0.92 0.84

MADRAS 0.82 0.83 0.96 0.83 0.81 1 0.73 0.83 0.86 0.88 0.82 0.84 0.89 0.86 0.86 0.71

TIRACHIRAPALLI 0.83 0.85 0.81 0.86 0.83 0.73 1 0.91 0.86 0.85 0.87 0.83 0.83 0.84 0.85 0.77

TIRUNALVELI 0.89 0.90 0.89 0.90 0.85 0.83 0.91 1 0.90 0.89 0.91 0.88 0.88 0.89 0.89 0.84

KAKINADA 0.96 0.97 0.92 0.97 0.91 0.86 0.86 0.90 1 0.97 0.97 0.95 0.95 0.97 0.97 0.91

NIZAMABAD 0.94 0.95 0.95 0.94 0.91 0.88 0.85 0.89 0.97 1 0.95 0.93 0.95 0.97 0.96 0.87

NAGPUR 0.97 0.98 0.90 0.98 0.90 0.82 0.87 0.91 0.97 0.95 1 0.97 0.96 0.98 0.98 0.93

DURG 0.97 0.98 0.89 0.98 0.91 0.84 0.83 0.88 0.95 0.93 0.97 1 0.97 0.97 0.97 0.92

RAIPUR 0.95 0.97 0.93 0.96 0.92 0.89 0.83 0.88 0.95 0.95 0.96 0.97 1 0.97 0.97 0.88

BULSAR 0.98 0.98 0.93 0.98 0.91 0.86 0.84 0.89 0.97 0.97 0.98 0.97 0.97 1 0.98 0.93

JAMSHEDPUR 0.97 0.97 0.93 0.97 0.92 0.86 0.85 0.89 0.97 0.96 0.98 0.97 0.97 0.98 1 0.92

ARRAH 0.93 0.93 0.80 0.94 0.84 0.71 0.77 0.84 0.91 0.87 0.93 0.92 0.88 0.93 0.92 1

JEYPORE 0.96 0.97 0.91 0.98 0.90 0.86 0.85 0.90 0.97 0.94 0.98 0.97 0.96 0.97 0.97 0.94

CONTAI 0.95 0.96 0.86 0.96 0.90 0.78 0.86 0.89 0.95 0.92 0.97 0.95 0.94 0.96 0.95 0.94

CALCUTTA 0.93 0.93 0.96 0.93 0.90 0.93 0.80 0.89 0.94 0.95 0.93 0.93 0.95 0.95 0.94 0.85

SAINTIA 0.95 0.96 0.86 0.96 0.89 0.77 0.84 0.88 0.95 0.93 0.96 0.93 0.93 0.96 0.95 0.94

NOWGARH 0.96 0.96 0.87 0.96 0.87 0.79 0.80 0.87 0.94 0.92 0.96 0.94 0.92 0.96 0.95 0.96

ALLAHABAD 0.93 0.94 0.86 0.93 0.87 0.80 0.75 0.83 0.91 0.89 0.93 0.94 0.92 0.94 0.95 0.93

VARANASI 0.97 0.97 0.88 0.98 0.87 0.80 0.86 0.89 0.96 0.94 0.97 0.97 0.94 0.97 0.96 0.95

DEHRADUN 0.98 0.99 0.92 0.98 0.92 0.84 0.87 0.92 0.97 0.96 0.99 0.98 0.96 0.98 0.98 0.94

TADEPALLIGUDEM 0.96 0.97 0.91 0.96 0.91 0.82 0.83 0.88 0.97 0.96 0.96 0.94 0.94 0.97 0.97 0.92

C.V 7.30 6.81 8.15 7.58 6.47 10.08 7.52 6.54 7.01 7.75 6.78 6.62 7.10 6.97 6.41 5.53

235

GRAM-ALL INDIA

ARRA

H PATN

A DELH

I ROHTA

K AURANGABA

D JABALPU

R SAGA

R ABOHA

R JAIPU

R SRIGANGANAGA

R COIMBATOR

E BAHRAIC

H LNUPBAN

D HAPU

R KANPU

R CALCUTT

A SAINTHI

A

ARRAH 1 0.97 0.96 0.95 0.95 0.94 0.94 0.94 0.95 0.96 0.93 0.95 0.93 0.96 0.95 0.94 0.95

PATNA 0.97 1 0.98 0.97 0.97 0.96 0.96 0.96 0.97 0.98 0.95 0.97 0.96 0.98 0.98 0.97 0.97

DELHI 0.96 0.98 1 0.98 0.98 0.97 0.97 0.97 0.98 0.99 0.96 0.98 0.97 0.99 0.99 0.98 0.97

ROHTAK 0.95 0.97 0.98 1 0.97 0.96 0.96 0.97 0.97 0.98 0.93 0.96 0.97 0.98 0.98 0.97 0.96

AURANGABAD 0.95 0.97 0.98 0.97 1 0.95 0.97 0.96 0.98 0.98 0.95 0.95 0.97 0.98 0.98 0.97 0.96

JABALPUR 0.94 0.96 0.97 0.96 0.95 1 0.96 0.95 0.96 0.96 0.94 0.94 0.95 0.96 0.96 0.96 0.96

SAGAR 0.94 0.96 0.97 0.96 0.97 0.96 1 0.95 0.97 0.97 0.94 0.95 0.97 0.97 0.97 0.96 0.95

ABOHAR 0.94 0.96 0.97 0.97 0.96 0.95 0.95 1 0.98 0.98 0.92 0.94 0.96 0.97 0.97 0.96 0.95

JAIPUR 0.95 0.97 0.98 0.97 0.98 0.96 0.97 0.98 1 0.99 0.94 0.95 0.97 0.98 0.98 0.97 0.96

SRIGANGANAGAR 0.96 0.98 0.99 0.98 0.98 0.96 0.97 0.98 0.99 1 0.94 0.96 0.98 0.99 0.99 0.98 0.96

COIMBATORE 0.93 0.95 0.96 0.93 0.95 0.94 0.94 0.92 0.94 0.94 1 0.94 0.93 0.95 0.95 0.95 0.95

BAHRAICH 0.95 0.97 0.98 0.96 0.95 0.94 0.95 0.94 0.95 0.96 0.94 1 0.95 0.97 0.97 0.96 0.97

LNUPBAND 0.93 0.96 0.97 0.97 0.97 0.95 0.97 0.96 0.97 0.98 0.93 0.95 1 0.97 0.97 0.97 0.95

HAPUR 0.96 0.98 0.99 0.98 0.98 0.96 0.97 0.97 0.98 0.99 0.95 0.97 0.97 1 0.99 0.97 0.97

KANPUR 0.95 0.98 0.99 0.98 0.98 0.96 0.97 0.97 0.98 0.99 0.95 0.97 0.97 0.99 1 0.98 0.97

CALCUTTA 0.94 0.97 0.98 0.97 0.97 0.96 0.96 0.96 0.97 0.98 0.95 0.96 0.97 0.97 0.98 1 0.96

SAINTHIA 0.95 0.97 0.97 0.96 0.96 0.96 0.95 0.95 0.96 0.96 0.95 0.97 0.95 0.97 0.97 0.96 1

C.V 7.12 6.76 7.04 7.05 6.84 6.82 7.03 7.08 6.90 6.98 7.01 7.82 7.13 7.12 7.01 5.81 6.55

236

GROUNDNUT OIL-ALL INDIA

RAJKOT HYDERABAD NANDYAL MADRAS BANGALORE CALCUTTA DELHI MUMBAI

RAJKOT 1 0.98 0.93 0.97 0.98 0.93 0.90 0.98

HYDERABAD 0.98 1 0.93 0.99 0.98 0.94 0.90 0.97

NANDYAL 0.93 0.93 1 0.93 0.92 0.85 0.83 0.91

MADRAS 0.97 0.99 0.93 1 0.97 0.93 0.89 0.96

BANGALORE 0.98 0.98 0.92 0.97 1 0.93 0.89 0.97

CALCUTTA 0.93 0.94 0.85 0.93 0.93 1 0.91 0.94

DELHI 0.90 0.90 0.83 0.89 0.89 0.91 1 0.90

MUMBAI 0.98 0.97 0.91 0.96 0.97 0.94 0.90 1

C.V 4.08 4.23 4.37 3.93 4.15 5.09 4.98 4.06

MUSTARD OIL-ALL INDIA

DELHI MOGA HAPUR KANPUR CALCUTTA

DELHI 1 0.98 0.95 0.99 0.99

MOGA 0.98 1 0.95 0.98 0.98

HAPUR 0.95 0.95 1 0.94 0.94

KANPUR 0.99 0.98 0.94 1 0.99

CALCUTTA 0.99 0.98 0.94 0.99 1

C.V 4.62 4.47 5.02 4.36 4.38

237

Appendix B - List of Markets Analyzed.

Rice

State Market (Variety) Integrated

Andhra Pradesh Kakinada(akkulu) √

Nizamabad(coarse) √

Bhimavaram(akkulu hansa/I.R.-8) Х

Tadepalligudem(coarse akkulu) √

Hyderabad(coarse) √

Assam Guahati(coarse Sali) √

Hailkandi(Sali) Х

Bihar Rachi(coarse mota) √

Dumka(coarse) √

Jamshedpur(coarse) √

Arrah(coarse) √

Gaya(coarse) √

Patna(coarse) Х

Gujarat Bulsar(coarse) √

Rajkot(deshi medium) Х

Haryana Karnal(coarse begmi) Х

Karnataka Shimoga(coarse) √

Kerala Trivandrum (coarse) √

Alleppey(Basundra Punja) Х

Kozhikode (Coarse (IR8) √

Madhya Pradesh Raipur(coarse gurmatia) √

Jabalpur(coarse) Х

Durg(coarse) √

Indore(coarse kallimooch)) √

Maharashtra Nagpur(coarse), √

Gondia(coarse), √

Orissa Balasore(coarse) Х

Jeypore(coarse) √

Tamilnadu Kumbakonam(coarse), √

Madras(coarse), √

Tirunelveli(coarse), √

Tirachirapalli(coarse) √

Tripura Agartala (Coarse) Х

Uttar Pradesh Azamgarh(3rd arwa) Х

Kanpur(coarse 3rd Gr.) √

Nowgarh(coarse 3rd Gr.) √

Varanasi(coarse 3rd Gr.) √

Dehradun(coarse 3rd) √

Allahabad(3rd arwa). √

West Bengal Santhia(coarse) √

Bankura(coarse) √

Contai(coarse fine) √

Calcutta(common) √

238

Delhi Delhi (Begmi) Х

Gram

State Market (Variety) Integrated

Bihar Patna (deshi) √

Arrah (deshi) √

Haryana Rohtak (local) √

Madhya Pradesh Jabalpur (deshi) √

sagar (deshi) √

Maharashtra Aurangabad (yellow) √

Punjab Abohar (black) √

Rajasthan Sriganganagar (FAQ) √

Jaipur (local) √

Tamilnadu Coimbatore (punjab) √

Uttar Pradesh Hapur (punjab) √

Bahraich (deshi) √

Kanpur (deshi) √

Banda (deshi) √

West Bengal Sainthia (big/punjab) √

Calcutta (big) √

Delhi Delhi (garara) √

MUSTARD OIL

State Market (Variety) Integrated

Punjab Moga √

Uttar Pradesh Kanpur (FAQ) √

Hapur (local mills) √

West Bengal Calcutta (local/ordinary) √

Delhi Delhi (ghani) √

GROUNDNUT OIL

State Market (Variety) Integrated

Andhra Pradesh Nandyal (expeler) √

Hyderabad √

Gujarat Rajkot √

Karnataka Bangalore (local) √

Maharashtra Bombay (ready) √

Tamilnadu Madras (expeler) √

West Bengal Calcutta (Madras) √

Delhi Delhi √

239

CASTOR OIL

State Market (Variety) Integrated

Andhra Pradesh Hyderabad √

Tamilnadu Madras (FAQ) √

Uttar Pradesh Kanpur √

West Bengal Calcutta (local) √

MUSTARD OILCAKE

State Market (Variety) Integrated

Punjab Moga (fair/desi) √

Uttar Pradesh Kanpur √

West Bengal Calcutta √

Delhi Delhi (papri) √

GROUNDNUT OILCAKE

State Market (Variety) Integrated

Andhra Pradesh Hyderabad √

Gujarat Rajkot √

Maharashtra Bombay (decorticated) √

Tamilnadu Madras (Decorticated expeller) √

CASTOR OILCAKE

State Market (Variety) Integrated

Andhra Pradesh Hyderabad √

Tamilnadu Salem (local) √

Uttar Pradesh Kanpur √

State Market (Variety) Integrated with

International Market

TEA

Tamilnadu Coimbatore (Roses/Brookbond) √

COFFEE

Tamilnadu Coimbatore (Plantation) √

240

CHAPTER 8

Has Trade Enhanced Gender Employment in India?����

8.1 Introduction

Gender- neutral impact of trade on employment has been questioned by many studies. Theoretical literature points out that trade liberalization may not be a gender-neutral process. With the change in production structure of India, consequent to trade liberalization, new avenues of employment may favor one gender as compared to the other. Increased exports may lead to higher employment of women, only if intensity of women employment is high in export oriented units. Similarly, a rise in imports may adversely affect the employment of women, if imports are mainly in those sectors where women have higher probability of employment. The issue becomes even more relevant in case of developing countries where women are at a disadvantage in terms of access to resources right from their births. In case of India, we find that the percentage share of female population in total population in India is around 48%, while the work participation rate of females in only 26% as compared to 52% in males. In 2004-05, women within the ‘15 years and above’ age group were usually engaged in domestic duties, only 33 % in rural India and 27 % in urban India had reported availability for ‘work’ on the household premises. In the organised sector, only 18.7% out of the total employees are women. This clearly indicates a gender-lag where women start with a lower access and participation in economically gainful activities. Not only the access to resources but the returns also differ significantly with respect to gender. According to NSSO Report (2005-06)54 the average wage rate for regular wage/salaried women employees is only 67% of the men employees in the urban areas while it is only 55% in case of rural areas. Trade may impact gender returns favourably as well as unfavourably. The gender wage gap may be reduced because trade can increase competition among firms resulting in pressure to cut costs. Lower wages of women with comparable skills to men may result in increase in demand of their labour, ultimately leading to wage equality. However, trade often results in a premium on skills. Thus, the resulting increase in the wage gap between skilled and unskilled55 workers may increase the gender wage gap given that in most countries the average man has a higher level of labor market skills than the average woman.

� This Chapter has been contributed by Rashmi Banga, Senior Economist, UNCTAD-

India project, Danish Hashim, Vice President, MCX, Mumba, and M.R. Saluja, IDF.We

are thankful for the research assistance provided by Manini Ojha, JNU 54 Employment and Unemployment Situation in India, NSS 62nd round 55 As found by many studies, e.g., Banga (2005), Hashim and Banga (2008)

241

Trade therefore may have vital influences on gender equality and if gender-sensitized trade policy may be used to enhance gender-neutral impact of trade. But in order to arrive at directions for trade policy, it becomes important to identify and estimate the trade- gender impacts in an economy. However, limited literature exists that estimates gender-sensitive impact of trade. For this purpose the chapter undertakes the following analyses:

♦ Trade Related Gender Development is traced to examine the role played by trade in promoting gender neutral growth.

♦ Gender Employment multipliers have been estimated across 46 sub-sectors in the economy using the input-output tables. This helps in identifying the sub-sectors across agriculture, industry and services where employment of women is relatively high.

♦ Impact of exports and imports on employment of women in organized manufacturing sector is estimated through a labour demand equation for 54 industries over the 1999-00 and 2004-05.

♦ Finally, “Gender sensitive Products”, defined as the products where women employment is relatively high are identified. This is done by using a concordance matrix between six-digit product level data (HS 2002 codes) and three digit industries level data (NIC). Using this concordance matrix the trade data at the product level is matched to the industrial data. Using the gender employment at industry level, industries with more than 40% of women employment have been identified. The corresponding products of these industries have been termed as gender sensitive products. The list of these products can be used by the trade policy makers as a tool to enhance gains from trade to women and protect women vulnerabilities from trade.

The chapter is organized as follows: section 2 provides a brief review of literature on trade-gender issue; section 3 discusses the trends in Trade Related Gender Development; section 4 presents the gender employment multipliers for 46 sub sectors of Indian economy; section 5 discusses the methodology adopted for estimating the impact of trade on employment in organized manufacturing sector and presents the empirical results; section 6 details the gender sensitive products for India; section 7 summarizes and concludes. 8.2. Review of Literature

Most of the studies which estimate the impact of trade on gender employment conclude that trade liberalization does not have a gender-neutral impact. Depending upon the intensity of employment of women in export and imports sectors of the economy, trade may favour the employment of one gender more than the other. However, the extent of the impact may differ considerably across sectors and countries. In this context, we briefly review the existing literature and highlight gender-neutralizing trade policy implications of the studies. For India, only limited empirical literature is available, given the lack of availability of comparable gender and trade data.

242

Gender Impacts of Trade: Review of Literature

Ramya Vijaya (2003) in her report “Trade, Skills and Persistence of Gender Gap: A Theoretical Framework for Policy Discussion” explores the impact of trade openness on gender divergence in skills in developing economies. In economic theory, derivations from the Heckscher-Ohlin theory of international trade suggest that greater trade openness can cause inter-country differences in skills to expand. This study combines the Heckscher-Ohlin trade perspective with the evidence of the increasing feminization of the low-skilled export labor force. The resulting theoretical and empirical analysis provides a platform for reviewing the gender dimensions of education and technology transfer policy reform in the context of global trade negotiations. This paper introduces the gender differences in labor demand into the H-O based theoretical framework and also look at the data trends to test the presence of a trade related gender biased skill divergence. The results highlight that a combination of higher trade and a larger initial gender gap in education, leads to a larger increase in the gender gap in school enrollments. Therefore in countries that start out with a gap between adult male and female skills, trade has a positive impact on the change in the enrollment gap i.e. the enrollment gap increases further. The regression analysis supports causality between trade participation and gender divergence in education investments. Larger trade volumes cause the secondary school enrollment gender gap to increase further in countries that have larger gaps between the average levels of male and female adult education. Menon & Rodgers (2006) in their study “The Impact of Trade Liberalization on Gender Wage Differentials in India’s Manufacturing Sector” address the question of whether the increasing competitive forces from India’s trade liberalization affects the wages of male and female workers differently. (Becker 1971). This idea has been incorporated into a theoretical model of competition and industry concentration. The study demonstrates that although an increase in trade still has a mitigating effect on the gender wage gap (neoclassical), under certain conditions, the net effect may be a widening of the wage gap between male and female workers (non-neoclassical). The theory is tested by estimating the impact of trade reforms on gender wage differentials using four cross sections of household survey data from the National Sample Survey Organization between 1983 and 2004. These data have been aggregated to the industry level and merged with several other industry-level data sets for international trade, output, and industry structure. The relationship between the male-female wage gap and variations across industry and time in the exposure to competition from international trade have been examined, while controlling for changes in worker characteristics and domestic concentration. The OLS and Fixed Effects techniques at the industry level have been employed to estimate the relationship between the male-female residual wage gap and measures of domestic concentration and international trade competition. Results indicate that increasing openness to trade is associated with a widening in the wage gap in India’s

243

concentrated manufacturing industries. Results show that groups of workers who have weak bargaining power and lower workplace status may be less able to negotiate for favorable working conditions and higher pay, a situation that places them in a vulnerable position as firms compete in the global market place. Rather than competition from international trade putting pressure on firms to eliminate costly discrimination against women, pressures to cut costs due to international competition appear to be hurting women’s relative pay. Raihan, Khatoon, Husain & Rahman (2007) in their research “Modelling Gender Impacts of Policy Reforms in Bangladesh: A Study in a Sequential Dynamic CGE Framework” explore the gender aspects of policy reforms in Bangladesh in a sequential dynamic computational general equilibrium (CGE) framework. The general objective of this research is to explore how gender interests are affected by greater exposure to trade and other policy reforms in the short and long run. This research uses the most updated SAM of Bangladesh and is the first attempt to build a gendered sequential dynamic CGE model for the Bangladesh economy. The research performs two simulations to examine the impact of: (1) domestic trade liberalisation in Bangladesh, and (2) the phasing-out of Multi-fibre Agreement (MFA) on textile and garments. It further builds a gendered social accounting matrix (SAM) for the year 2000 and uses it in a sequential dynamic computable general equilibrium framework. It is found that domestic trade liberalisation leads to a significant expansion of the ready-made garment sectors in the economy as a result of which the share of market labour supply of unskilled female labour increases. But, this results in a fall in the shares of domestic labour supply and leisure of unskilled female members of the households. The fall in the share of leisure time may have significant negative implications for the time spent on education by this labour category. It is also observed that the long run impacts are different from the short run impacts with respect to the magnitude of the effects. In the case of second simulation, it is noted that the phasing out of the MFA works in completely the opposite direction. The share of market labour supply of unskilled female members of the households decrease, and the shares of domestic work and leisure increase for most of the households both in the short and long run. STADD Development Consulting Pvt. Ltd (2008) in the report on “Impact of Trade and Globalization on Women Workers in the Handicraft Sector: Evidence from the Carpet and Embroidery Sectors” for UNCTAD- India, tracks impact of globalization on women’s employment and empowerment, using an analytical framework suited to the conditions prevailing in the handicrafts sector of India characterized by a stagnant production structure. The analysis adopts a multivariate regression framework. The share of women’s employment was estimated from ASI unit level data for the intervals 1981, 1986, 1993 and 1997. The multivariate model is used to test the hypotheses linked to responsiveness of female share of employment to wage income, technology and other control, as well as explanatory factors.

244

The study includes a primary survey of stakeholders, apart from a multivariate least square model to decipher functional specification among the key variables involved in the production process. The multivariate and cross-sectional regressions show the decision-making power of women is linked to women’s employment and income, which in turn is a critical element of the crusade against poverty and illiteracy. The empirical analytics of the sector indicate that strength lies in openness, openness undoubtedly being a positive factor for income as well as productivity growth. The study therefore, observed a positive impact of globalization on productivity using both secondary as well as primary sample data from the carpet and embroidery sectors of India.

CSR in its report on “Impact of Trade and Globalisation on Gender in India: A Case Study of Women Workers in the Fisheries Sector” for UNCTAD-India looks at the correlation between the performance of fisheries industry, particularly the export component, and its impact on the women labourers engaged in the fish processing industry.

Using the secondary and primary data, a two-tier investigation is undertaken. An econometric model, based on the Thirlwall analysis of trade, is carried out using secondary data; and primary data analysed using simple statistical tools. Contrary to the accepted wisdom that trade in general stimulates growth in the concerned sector, the study concludes that the general perception of the fishing community has been depressing. While at one hand the opening of the trade has created competitiveness, it has exposed the Indian fish export sector to world market movements, without adequate support in terms of technology and accumulation of excess capacity in the sector. The overall impact of liberalisation has been adverse, though increasing unskilled employment opportunities in specific cases are in evidence along with the mobility of labour demand across states. National Productivity Council (2008) in “Impact of Trade and Globalisation on Gender” in their Report to UNCTAD-India studies the relationship between trade liberalisation in India and its gender specific impact at the sectoral and regional level. Trade and gender links are studied in the Plantation sector, Food Processing, Textiles and clothing, Handicrafts, and Fisheries & other marine products. One major finding of the econometric estimation is that female workers earn less than male workers irrespective of the industry, region or location. Trade and globalisation have significant impact on the labour market and daily earning of the workers. Income for both male and female workers has improved wherever trade and globalisation have positively affected the labour market; however, workers who are indifferent to trade and globalisation have a mixed experience. Moreover, benefits reaped by the male workers are higher than those of females in terms of gain in income in a globalised environment. USAID (2006) studied how the trade liberalization impacted the growth, employment and poverty in South Africa. More specifically, by using the dynamic general

245

equilibrium and micro simulation model, the study attempts to show how trade influenced the process of growth and reducing inequality in job opportunities to men and women. Study found that trade liberalization has contributed positively to the growth by inducing trade related technological improvement. At the same time, however, it increased the income inequality among men and women. It is argued that while men and women both benefit from trade induced growth; it is male-headed households who have befitted more from rising factor incomes. Study concludes that trade liberalization should not be relied to generate pro poor growth or to address the issue of income inequalities between men and women. Gender Impacts of Trade in Services: Review of Literature

Riddle (2004) undertakes a detailed analysis of gender impacts of trade in services across 74 developing countries, including 20 of the least developed countries. The study examines the potential links between liberalization of trade in services and development focusing on the central role of services in all economies, with many of the service suppliers being women. It has been noted in the study that any growth in services, whether domestic or through trade, will not in itself ensure equity or an improved quality of life for girls and women. The study concludes that to maximize development benefits of trade in services, the focus needs to be on strengthening the ability of developing economies to ensure and implement gender sensitive employment, pay equity legislations and effective domestic regulatory reforms prior to further liberalization of trade in services. Puri (UNCTAD, 2004) analyses the issues related to women’s participation in two modes of service delivery which are Mode 1 and Mode 4. Mode 1 involves cross border trade in services like ICT (Information and Communication Technology), Professional services such as Accountancy, Engineering and Management e.t.c. and Mode 4 involves the temporary movement of service suppliers of country to supply services in the territory of another country i.e. national movement of people in services like care services, doctors, nurses, teachers, health workers and social workers. Both modes have been analyzed in depth with respect to their importance in trade and development-- trends, sources and destinations of service providers, remittances, benefits, costs, and status of commitments. The study finds that multilateral liberalization of Modes 1 and 4 is key to increased and beneficial participation of developing countries not only in world services trade, but also in trade led-growth and development in general. Both Modes provide the greatest opportunity to developing countries to foster gender equity, welfare and women’s social and economic empowerment. Achievement of broader liberalization commitments through GATS negotiations would renew developing countries faith in the ability of globalization and liberalization of trade and investment and the application of modern technologies to deliver major development gains. Williams (2002) examines tourism and development from the perspective of social and gender equity and finds that the issue is multi-dimensional. The study argues that tourism growth may increase competition with other sectors such as domestic agriculture and other export areas. Most of these sectors provide wages for women and therefore it might be possible that tourist development may not be in line with social

246

and sustainable development. Also, it has been argued that there are significant gender biases and inequality that may predispose women to greater vulnerabilities and constraints in enjoying the presumed benefits of tourism development and disproportionately shouldering the adjustments to its negative consequences. Anh-Nga Tran-Nguyen (2004) highlights that international trade influences growth process and gender equality in positive as well as in negative ways. Positives of trade are enlargement of markets and an exchange of technology and information, contributing thereby to growth and development. Trade benefits all—men and women. But within the same country benefits are distributed differently between men and women because the society assigns them different roles. Implementation of multilateral trading rules should, therefore, provide governments with enough policy and regulatory space for the pursuit of the gender equality objective. Korinek’s (2005) examines ways in which greater integration through trade impacts women and men differently ensuing implications for growth. The paper finds that trade creates jobs for women in export oriented sectors. Jobs that bring more household resources under women’s control lead to greater investments in the health and education of future generations. Women also have less access to productive resources, time, and particularly in many developing countries to education. Professional women continue to encounter discrimination in hiring and promotion, including in OECD countries. Once different impacts are ascertained, well designed policy responses may aid women in taking advantage of greater openness to trade. Grown (2005) explores the linkages between trade liberalization and the provision of and access to sexual and reproductive health services. The study finds that trade liberalization can possibly create new opportunities for improving reproductive health, but at the same time it can also make it more difficult to advance reproductive/sexual health and rights’ objectives in policies, programmes, and services. There are two ways in which trade effects health. Direct pathways through trade policies like GATS and TRIPS affect the supply of reproductive health services by possibly interfering with the national health policies and increasing costs of reproductive drugs, supplies, and vaccines. Second, trade policies and movements in goods and services affect women’s demand for services indirectly through changes in their labour force participation. Coche (2004) examines the issues related to vulnerability of women to trade liberalization in terms of economic roles, sectors and industries intensively employing one gender; asymmetrical impacts on employment and working conditions of men and women; unequal impacts of loss of tariff revenue on men and women; and tools to address this. The study draws three general conclusions. First, women play particular economic roles that impact how they are affected by trade liberalization as women allocate their work time differently than men, are intensely employed in different sectors than men, and receive lower wages than men for the same work. Second, trade liberalization tends to have asymmetrical impacts on men’s and women’s employment and working conditions. Some of these impacts are positive for women while others can be negative. The balance of the different impacts and mechanisms can only be determined in specific contexts and country circumstances. Lastly, trade liberalization policies may have positive impact on aggregate economic performance but these

247

policies may hurt the welfare of certain groups. So for trade liberalizing policy to succeed, both in social and political terms, policymakers must recognize which groups may lose from liberalization and therefore take steps to mitigate the negative consequences of liberalization on these groups and sectors. It is important to pay attention to the issues of gender in relation to trade liberalization and international economic integration. On the whole, studies show that the extent of women participation in different sectors, social empowerment and cultural and economic equality in terms of access to resources are some of the factors that may lead to differential gender impacts. Gender inequalities which are visible often persist in terms of:

• Employment opportunities/job segregation

• Returns from labour

• Conditions of work and quality of employment

• Access to basic services and resources

• Access to technology and training

• Distribution of income inside and outside the household However, the extent of the impact may differ considerably across sectors and countries. Research and monitoring of gender aspects of trade expansion and liberalisation is thus a high priority with respect to various sectors. This chapter makes an important contribution to the existing literature on gender and trade liberalisation in India. It estimates the impact of exports and imports on gender employment in organized manufacturing sector and estimates the gender employment created by exports in 46 disaggregated sectors of the economy for the period 2003-04 to 2006-07, which include 15 services sectors. Further, it identifies Gender Sensitive

Products (GSP) which may be very relevant in trade policy formulation. 8.3. Trends in Trade Related Gender Development in India 8.3.1 Trends in Gender Development Index (GDI)

Trends in Gender Development Index (GDI), which is based on life expectancy, education, (the adult literacy rate and the combined primary to tertiary gross enrollment ratio) and estimated earned income, indicates that there has been some improvement in gender related devlopment in India (Table 8.1).

248

Table 8.1: Trends in Human Development Index and Gender Development Index in India

Source: HDR 2007 For 1995 to 1999, GDI grew by 0.129 percentage points, while 1999-2005 it witnessed an increase of 0.047 percentage points The United Nations Human Development Report (2007) shows that India’s HDI in 2005 stands at 0.619, up from 0.611 in 2004. HDR (1007) India ranks at 128 out of 177 countries of the world. In terms of Gender Development Index, India ranks at 113 out of 157 countries. This shows an almost zero count in terms of HDI-GDI rank indicating that both move together. But a negative count of (-11) for GDP per capita (PPP US$) – HDI rank indicates that the country as a whole is amassing wealth at a rate higher than which social inequalities are getting corrected. 8.3.2 Gender Literacy rates in India

Women’s employment in India following reform and liberalization has increased to the extent of about 5-10 per cent. During 1993-94, about 29 per cent of rural women and 42 per cent of urban women in India were classified as usually engaged in household duties only, while 33 and 16 per cent of rural and urban women respectively, usually carried out some (principal and subsidiary) economic activities ( source: ILO). In 2004-05, women within the ‘15 years and above’ age group were usually engaged in domestic duties, 33 per cent in rural India and 27 per cent in urban India had reported availability for ‘work’ on the household premises. Of them, about 72 per cent in rural areas and 68 per cent in urban areas preferred only ‘part-time’ work on a regular basis, while the percentages of such women preferring regular ‘full-time’ work were 23 and 28 percent in rural and urban India, respectively (source: NSSO)

Presently, for urban females, services sector accounts for the highest proportion (36 per cent) of the total usually employed, followed by manufacturing (28 per cent) and agriculture (18 per cent) Table 8.2. Work opportunities for women in urban services and manufacturing sector exists but there is a need to facilitate and improve their Work Participation Rates through better education, skill development and removal of gender associated hurdles like lack of crèches, etc.

Year HDI GDI 1980 0.450

1985 0.487

1990 0.521

1995 0.551 0.424

1999 0.553

2000 0.578

2003 0.586

2005 0.619 0.600

249

Table 8.2: Labour Force and Work Force Participation (in percent)

1983 1993-94 1999-00 2004-05 Labour force participation rates (LFPR) Rural male 52.7 53.4 51.5 53.1

Rural female 21.9 23.2 22.0 23.7

Urban male 52.7 53.2 52.8 56.1

Urban female 12.1 13.2 12.3 15.0

Work force participation rates (WFPR) Rural male 48.2 50.4 47.8 48.8

Rural female 19.8 21.9 20.4 21.6

Urban male 47.3 49.6 49.0 51.9

Urban female 10.6 12.0 11.1 13.3 Source: Economic Survey 2007-08

Coming to literacy, a huge gap exists between the male and female literacy rates (Table 8.3). This gap continues to persist even after huge increases in female literacy. There has been no discernible attitudinal change and male education is still considered more essential than females as there is less return on investment made in educating a girl child. Table 8.3: Percentage of Literacy Rates by sex in India

Year Age Group Persons Males Females Rural Urban

Male-Female Gap in Literacy Rates

1971 5 and above 34.45 45.95 21.97 27.89 60.22 23.98

1981 7 and above 43.56 (41.42)

56.36 (53.45)

29.75 (28.46) 36.09 67.34 26.65

1991 7 and above 52.21 63.86 39.42 44.69 73.09 24.84

2001 - 64.8 75.3 53.7 - - 21.59 Source: Department of Secondary and Higher Education, Ministry of Human Resource Development,

Govt. of India.

By 2001, around 75% males and 50% females were able to read and write simple sentences. But there is a marked difference in rural and urban literacy rates. The 1991 census revealed that while literacy levels in urban areas were around 75%, those in rural areas were only 45%. This shows the level of lopsidedness that exists in human development in India.

8.3.3 Trends in Gender Employment in India

The percentage share of female population in total population in India is around 48%, while the work participation rate of females in only 26% as compared to 52% in males.

250

There exists a wide urban-rural divide in the participation of women and men in the economy. About 24.9% of women in rural areas and about 14.8% of women in urban areas were in the workforce in India during 2004-05, whereas about 54.6% of men in rural areas and 56.6% of men in urban areas are in work force. In the organised sector, out of the total employees in 2004, about 18.7% were women (Table 8.4). Table 8.4: Employment by industry:[per cent of employment according to Usual Status (PS+SS)]

1993-94 1999-2000 2004-05

Agriculture

Rural males 74.1 53.4 66.5

Rural females 86.2 85.4 83.3

Urban males 9 6.6 6.1

Urban females 24.7 17.7 18.1

Manufacturing

Rural males 7 7.3 7.9

Rural females 7 7.6 8.4

Urban males 23.5 22.4 23.5

Urban females 24.1 24 28.3

Construction

Rural males 3.2 4.5 6.8

Rural females 0.9 1.1 1.5

Urban males 6.9 8.7 9.2

Urban females 4.1 4.8 3.8

Trade, hotels & restaurants

Rural males 5.5 6.8 8.3

Rural females 2.1 2 2.5

Urban males 21.9 29.4 28

Urban females 10 16.9 12.2

Transport, storage & communications

Rural males 2.2 3.2 3.9

Rural females 0.1 0.1 2

Urban males 9.7 10.4 10.7

Urban females 1.3 1.8 1.4

Other services

Rural males 7 6.1 5.9

Rural females 3.4 3.7 3.9

Urban males 26.4 21 20.8

Urban females 35 34.2 35.9

251

Within services, share of different services sectors in total female employment in services is computed. It is found that percentage of female employment in total employment in the services sector, comprising other transport services, communication, banking, insurance, tourism and other services, is around 16%. Share of other services in total female employment in services sector is found to be highest (79%), followed by communication services (11%) and Banking services (5%). Female employment in labour force is found to low in other transport services, insurance services and tourism 8.3.4 Trends in Gender Wage/Salary

To assess to what extent have the benefits of growth of services sector in India been shared between the two genders, it is important to examine to what extent do females participate in different services sectors and gain positively in terms of wages and salaries. With respect to the wages and salaries, studies indicate large gender disparities. Very limited information exists for on wage/salary across gender in for disaggregated services sectors. NSSO (2005) estimates average wage/salary received per day by regular employees for two broad categories of services (Table 8.5).

252

Table 8.5: Average Wage/ Salary (in Rs) received per day by regular wage/salaried employees of age 15-59 years by Industry of work, sex, sector and broad education level for India.

Very interesting insights emerge from this information. In urban areas, on an average wage/salary paid to females is only 67% of that paid to males, while in rural areas females are paid 55% of what is paid to the males(according to the 61st round, NSSO). This wage disparity differs across sectors and education levels. In urban areas, the highest gender wage disparity exists in mining and quarrying and manufacturing sectors. Female wage/salary are around 80% of male wage/salary on an average. In rural areas, the gender wage disparity is higher in almost all the sectors as compared to the urban areas. Better access to resources like education, technology and knowledge, may be a reason for the rural-urban differences in the trends. But this clearly reflects existence of gender wage disparity in all sectors of the economy. Across services sectors, the wage/salary trends show that as the literacy levels of females increase the wage disparity declines. However, it is interesting to note that

253

even for graduate and above, the salaries earned by females on an average is only 70-75% of that earned by males with similar education level. This implies that with higher growth of services, even if employment opportunities for women grow at the same rate, the benefits of the growth goes more to males as compared as females. 8.4. Gender Employment 8.4.1 Methodology for Estimating Gender Employment

In order to estimate the impact of exports on Gender Employment across different sector, we use the latest available input-output matrix for India for the years 2003-04. Using the employment coefficients and change in output due to increased exports we derive the output and employment multipliers for each sector over the period 2003-04 to 2006-07. Output multipliers indicate the total increase in output of the sector due to direct as well as indirect demand created because of exports in the economy. Employment multiplier of a sector indicates the increase in employment required in the sector to produce the increase in output demand.

We define the output multiplier of a sector as the amount by which the total output increases for a unit increase in the output of that sector. It is usual to measure the unit change in INR lakhs (00,000) or 0.1 million. Thus, if the output multiplier for a sector is 4, it implies that for every INR 1 lakh (100,000) increase in the sectoral output, total output (of the entire economy) increases by INR 4 lakhs (400,000). That is an increase in total output by 0.4 million for every increase in sectoral output by 0.1 million rupees. Similarly, the employment multiplier of a sector gives an estimate of the aggregate direct and indirect employment changes, in person years, resulting from the increase in INR 100,000 of output of that sector. Exports in each sector in 2006-07 have been deflated so as to remove any price effects.

We generate gender employment by applying gender employment ratios to the increase in employment generated by the exports across sectors. The following are the sources used for gender wise employment coefficients.

1) For plantation crops the ratio of female to total employees is taken from Indian labour statistics 2003-04.

2) For other crops the estimates are based on gender wise work force from the 2001 census. The same coefficients are used for the food crops, cash crops and “other crops”

3) For minerals the gender wise estimates are based on “Statistics of mines in India, vol-I-Coal and vol-II- other minerals. Indian Bureau of Mines

4) For different sectors under manufacturing the estimates are based on report no 515 of the NSSO on “ Employment and Unemployment situation in India” 2004-05.This source is also used for getting the estimates in case of animal husbandry, forestry, fishery, and other service sectors.

254

8.4.2 Gender Employment Generated by Increase in Exports during 2003-04 to 2006-

07

Table 8.6 presents the increase in male and female employment across 46 sectors due to rise in exports during the period 2003-04 to 2006-07.

Table 8.6: Gender Employment generated by increase in Exports from 2003-04 to 2006-07

Increased Female Employment from 2003-04 to 2006-07

Increased Male Employment from 2003-04 to 2006-07

Sectors (in person years) (in person years)

Million No's Million No's

Food crops 1.27 1.96

Cash Crops 0.67 0.98

Plantation Crop 0.22 0.25

Other crops 0.11 0.16

Animal Husbandry 0.12 0.07

Forestry and logging 0.09 0.11

Fishing 0.01 0.02

Coal and lignite 0.06 0.11

Crude petroleum, natural gas 0.03 0.05

Iron ore 0.02 0.02

Other Minerals 0.13 0.70

Food Products 0.18 0.27

Beverages, tobacco, etc. 0.00 0.00

Cotton textiles 0.21 0.34

Wool Silk and Synthetic fiber 0.10 0.16

Jute, hemp, mesta textiles 0.02 0.04

Textiles products including wearing apparel 0.36 0.54

Wood, furniture, ect. 0.29 0.53

Paper & printing, etc. 0.04 0.07

Leather and leather products 0.05 0.09

Rubber, petroleum, plastic, cola. 0.07 0.11

Chemicals, etc. 0.08 0.14

Non-metallic products 0.03 0.07

Metals 0.18 0.41

255

Metal products except mach. And tpt. Equipment 0.09 0.20

Tractors, agri. Implements, industrial machinery, other machinery 0.11 0.23

Electrical, electronic machinery and applications 0.01 0.01

Transport Equipments 0.02 0.04

Miscellaneous manufacturing industries 0.38 0.80

Construction 0.09 0.18

Electricity 0.08 0.13

Gas and water supply 0.01 0.01

Railway transport services 0.10 0.18

Other transport services 0.55 1.13

Storage and warehousing 0.00 0.01

Communication 0.09 0.16

Trade 1.30 2.76

Hotels and restaurants 0.36 0.57

Banking 0.14 0.23

Insurance 0.03 0.06

Ownership of dwellings 0.00 0.00

Education and research 0.00 0.00

Medical and health 0.00 0.00

Other services 1.53 2.42

Public administration 0.00 0.00

Tourism 0.16 0.27

Total 9.38 16.60

The results show that exports in the period 2003-04 to 2006-07 generated 9.38 million employment for women and 16.60 million for men. This implies that though exports generated additional employment for women in India in this period, it was only 36% of the total additional employment generated. However, the share of females in additional employment generated due to exports exceeds the share of females in total employment by nearly 5 percent points. This suggests that exports may have led reducing the gap between male – female employment in India. It is interesting to note that the generated female employment is found to be high in agriculture sector, mainly in food crops, plantation crops and cash crops; in manufacturing sector the female employment generated is found to be high in cotton textiles, textile products, wood furniture and miscellaneous manufacturing products. Among services sectors female employment generated is found to be high in domestic

256

trade, hotels and restaurants, other transport services and tourism. Animal husbandry is the only sector where female employment is found to be higher than those of males. On the whole, it can be said that rise in exports in the period 2003-04 to 2006-07 increased female employment in almost all sectors. 8.5. Impact of Trade on Employment in Organized Manufacturing In order to estimate the impact of exports and imports on gender employment in the Indian organized manufacturing industries trade data is needed at the industry level. However, Indian industry level data does not report trade data as the trade data is at the product level. We therefore construct a concordance matrix matching six-digit HS codes (Harmonised system of tariffs 2002) and three-digit level NIC codes (National Industrial Classification). This concordance matrix has been used to create trade data at three-digit ASI classification of industries. The industrial data reports the employment by gender. Using ASI data we estimate the labour demand equation for men and women separately. Applying dynamic panel data techniques (Arellano bond Model) we estimate the impact of export intensity and import intensity of industries on gender employment. 8.5.1. Methodology and Data Sources

8.5.1.1 Methodology

In order to identify the factors that influence the demand for labour force (women and men) in an industry, a specific demand function for labour force is estimated. It has been derived from a two inputs case of constant elasticity of substitution function (CES) that allows for non-constant returns to scale.

[ ] ρµρλρχ/

))(1()(−−− −+= tLesksQ ----------- (1)

Where, 10&0 <<> sχ ; Q is the output, k is the capital, L is labour, s is the share

parameter and ρ determines the degree of substitutability of the inputs. The elasticity of substitution can take any non-negative constant value (including unity as in the Cobb-

Douglas case) & technical progress is labour augmenting at rate of λ. χ is the efficiency parameter as it changes output in the same proportion for any given set of input levels and the parameter s can be interpreted as a distribution parameter since it determines the distribution of income through the factor payments. To examine the factors that affect the demand for labour and consequently the employment in an industry we use marginal productivity theory, and equate marginal product of labour (MPL) to the real wage (W/P) (first order conditions, assuming mark – up is constant). Taking logs and rearranging (1) we get:

257

( )χρ

ρβ

λρ

ρln

)1(1ln

)1(

1

)1(ln

)1(

1)ln()ln(

+−

−

++

+−

+−=

s

ep

w

eQL t

−−

−+−−

−= )ln()1(1ln()1(ln)ln( χσ

βσλσσ

s

p

wQ t ----- (2)

)1(

1,

ρσ

+=where , λt is taken as exogenous technical change which may occur

through export, imports and technology.

( , , log )t f Exports imports Techno yλ = ------------- (3)

However, in the empirical framework along with the above factors we need to include other potential demand shifters, which also control for industry-specific effects. This is justified by arguing that merely including the factors derived from theory may not capture other influences, which could effect an industry’s demand function [Driffield and Taylor (2000)]. We therefore control for inter-industry variations by including capital/labour ratio. Allowing for persistence in labour demand and adding fixed effects we arrive at:

ititititititit eortlkpwQlL ++++++= − )ln(exp)/ln()/ln()ln(}ln()ln( 5432110 ββββββ --(4)

ititititititit eimportlkpwQlL ++++++= − )ln()/ln()/ln()ln(}ln()ln( 5432110 ββββββ --(5)

8.5.1.2 Dynamic Panel Data Estimation

For the purpose of estimating the impact of trade on employment we need to take into account rigidities in the Indian labour market. We therefore construct dynamic panel data (DPD) models, which are estimated using Generalised Method of Moments (GMM) following Arellano and Bond (1991). GMM has become an important tool in empirical analyses of panels with a large number of individual units and relatively short time series. This model can be written as:

it

i

tiit cayyι

ι

ι

11, ++= − ----------------

(6)

1;3;,......,2;,.....,1, <≥==ι

aandTTtNiwhre

For such models the within group estimator (for the fixed effects models) and the GLS estimator (for the random effects model) are not applicable. Therefore GMM estimator is applied. Adopting standard assumptions concerning the error components and initial

258

condition (i.e. error terms are not autocorrelated) Arellano and Bond (1991) propose moment conditions6. The validity of moment conditions implied by DPD models is commonly tested using conventional GMM test of overidentifying restrictions associated with Sargan (1958). CES production function will be utilized to derive the following labour demand equation:

)ln()1/()]/(1ln[)]1/(1[)1/()/ln()1/(1)ln()ln( γρρβλρρ +−−+++−+−= sepweQL t

-----(7) From the above equation the impact of trade on employment will be captured through

λt, assuming the following:

)(),(Im),((, 321 TechprotsExprotsFt λλλλ =

The estimations are carried out using the Generalized Method of Moments (GMM).

8.5.1.3 Data and Variables

To estimate the impact of exports and imports on gender employment, we first use the concordance matrix (HS-NIC) to arrive at three-digit industry level export and import data. The data on exports and imports has been taken from World Integrated Solutions (WITS). The data on total persons engaged, total emoluments, Gender employment, and capital formation is extracted from various issues of Annual Survey of industries (ASI) 2005-06. ASI, published by the Central Statistical Organization (CSO), Government of India, provides a reasonably comprehensive and reliable disaggregated estimates for the manufacturing industries, covering all production units registered under the Factories Act, 1948, ‘large ones’ on a census basis (with definition of ‘large’ changing over time) and the remaining on a sample basis. Data is drawn for 54 industries at three-digit level of industrial classification (National industrial Classification) for a period between 1999-00 and 2004-05. For the purpose of estimating the regression equations, a set of variables have been constructed using various data series. Assuming that the industry wage rate is common to both the genders, the same has been calculated by dividing the total emoluments by the total persons engaged in an industry. The wage series is obtained same has been deflated with WPI index to reflect the wage rates in real terms. Intensity of employment of women in an industry is calculated by dividing the women employed in an industry by the total employment in that industry. Gross Value Added (GVA) of industries is deflated with respective WPI series to reflect the values at constant (1993) prices. To capture the technological level across industries, a ratio of capital formation to the labour is arrived at. Exports and imports intensities have been calculated to represent their respective shares in total output of the industry concerned.

259

8.5. 2 Empirical Findings of Impact of Trade on Gender Employment in Organised Manufacturing In order to estimate the impact of trade on gender employment, two specifications have been reported, i.e., impact of export intensity and import intensity of industries on proportion of women employment in total employment; and on total employment in the industries. The results of the two specifications are reported in Table 8.7 and Table 8.8. Table 8.7 shows that export intensity has a positive and significant impact on women employment. But imports have not led to any displacement of women employment. Contrary to this, results for total employment show that though export intensity has positively affected employment in the industry, imports have led to a fall in employment intensity (Table 8.9). As expected, wage rate is negatively related to women and total employment. Interestingly, the results show that technology is positively related to women employment ratio while it is negatively related to total employment. This indicates that low tech industries have higher employment as compared to high-tech industries but in case of women employment, better the technology higher is the proportion of women employment. Further, we find that when real wage rate rises, the fall in women employment would be higher than that of men. In other word, at a given wage rate, the chances of women getting employed is lesser as compared to men. At aggregate level, as seen in Table 8.9, the wage and employment relationship is inconclusive due lack of statistical significance. What it indicates is that wage rate may be important for determining the employment of women, but at aggregate level it does not have a very strong relationship with the employment. Weak relationship between wage rate and employment at aggregate level could be attributable to rigid labour laws. The positive relationship between technology and the intensity of women employment in both export as well as import equations indicates that as capital intensity improves, women have higher chances of employment than the men. Women may find themselves less preferred for jobs in relatively labour intensive sector but they are preferred to men in relatively capital intensive sector. Improvement in technology reduces the need of physical strength, and enables women to go for jobs which earlier could have been held by men. This is evident from the fact that electric machinery saw an increase in the female share of its regular salaried workforce going up from 4.6 percent in 1983 to 19.5 percent in 2004 (Menon, 2007). With the rapid increase in modernization of the business, one would expect intensity of women employment in job to increase. The positive relationship between intensity of women employment and technology is reinforced by the result at aggregate level. The latter shows that improvement in technology would have adverse effect on employment, though this relationship is not strongly conclusive. Thus improvement in technology might reduce overall employment, but this would increase women employment. Export is found to be positively related with the intensity of women employment unlike the imports whose relationship is not as strongly conclusive. Examining the gender-sensitive products, arrived at in the next section of the chapter, we do find that these

260

products are largely those which have high share in India’s export basket. In terms of industries, there are number of export oriented sectors such as textiles where the concentration of women is not only high but increasing over the years. In case of wearing apparel for instance the female share of its regular salaried workforce went up from 16.1% percent to 23.1 percent between 1983 and 2004 (Menon, 2007). Export not only increases the intensity of women employment in Indian organized manufacturing but at the same time also leads to increase in overall employment, as is evident from Table 8.8.

The gross value added in both export as well import equations is negatively related with the intensity of employment, though the relationship is conclusive only at 6% and 11% level of statistical significance. If GVA is taken to mean the scale of employment, the result can be interpreted that with the increase in scale women intensity of employment in an industry falls. In other words, larger the industry, lesser is the intensity of employment. At the aggregate level, evidence of GVA affecting the employment is almost completely inconclusive in view very poor level of statistical significance. This is not surprising as the growth in employment stands far below the growth of output in the Indian manufacturing. It may therefore be argued that the scale of operation of an industry is weakly related with the growth in output but at the same time it is not in favour of the women employment. Table 8.7: Dependent Variable is ln (Women / Total Employment) N=199

Independent Variables Coefficients Equation (1) Equation (2) Lag L(1) 0.131

(1.76) 0.125 (1.66)

Ln GVA -0.083 (-1.59)

-0.098 (-1.89)

Ln(W/P) -0.510* (-2.07)

-0.534* (-2.27)

Ln (K/L) 0.050* (2.24)

0.050* (2.17)

Ln(export/output) 0.018* (3.21)

-

Ln (imports/output) - 0.003 (0.70)

Constant 1.051 (1.17)

1.147 (1.33)

Note: (i) * indicates statistical significance at 5% level; (ii) T-values in parenthesis; (iii) Equation (1) and Equation (2) are with export and imports as independent variables respectively.

261

Table 8.8: Dependent Variable is ln (Total Employment)

Independent Variables Coefficients Equation (1) Equation (2) Lag L(1) -1.56*

(-4.26) -1.54* (-4.08)

Ln GVA -0.001 (-0.02)

0.032 (0.49)

Ln(W/P) -0.030 (-0.12)

-0.311 (-1.14)

Ln (K/L) -0.044 (-1.66)

-0.046 (-1.69)

Ln(export/output) 0.026* (2.28)

Ln (imports/output) -0.040* (-3.61)

Constant 28.83* (5.85)

30.01* (5.89)

Note: (i) * indicates statistical significance at 5% level; (ii) T-values in parenthesis; (iii) Equation (1) and Equation (2) are with export and imports as independent variables respectively.

8.6. Identifying Gender Sensitive Products

Trade negotiations, at this juncture, are hardly cognizant of the gender effects which they may lead to, whether positive or negative. For trade gains to be favourably distributed among gender, it is important to gender sensitise the trade policy. The trade negotiations should attempt to address the vulnerabilities of the economy in terms of high dependency of women for employment in few sectors. An attempt is made to identify at eight-digit level tariff lines that correspond to products which have high engagement of women, i.e., gender sensitive products. In order to identify gender sensitive products, three steps are undertaken. Firstly, the extent of women employment in three-digit level manufacturing industries in the organized sector is estimated. The data source for this is Annual Survey of Industries (2005-06). Secondly, the concordance matrix constructed between six-digit HS codes (Harmonised system of tariffs 2002) and three-digit level NIC codes (National Industrial Classification) has been used. Finally, industries where women employment is greater than three times the total industrial average over the period (2003-2005) are identified and products produced by these industries are categorized as gender sensitive

products.

With respect to the Indian manufacturing, we find that women employment as a proportion of total employment has not increased much over the years. In fact, it was

262

only around 15% in 2000-01 and it declined to around 14% in 2004-05. Over the years, on an average we can say that organized manufacturing industries employed around 13% of women employees (Table 8.9). Table 8.9 Average Employment of Women in Indian Organised Manufacturing Industries

Year Ratio of Women Employed in Total Employment (Average across industries in Organised Manufacturing)

1999-2000 0.143

2000-01 0.148

2001-02 0.135

2002-03 0.130

2003-04 0.133

2004-05 0.138

Average over the Years 0.138

Source: Annual Survey of Industries 2005-06 Averaging over the last two years, we find that the industries which had women employment greater than 39% are mainly four, which are manufacture of tobacco products, manufacturing of wearing apparels, manufacture of other food products and agriculture and other animal husbandry service activities, except veterinary (Table 8.10).

Table 8.10: Industries with High Proportion of Women Employment

NIC Codes Description Ratio of Women Employment to Total Employment (%)

160 Manufacture of the tobacco products

0.64

181 Manufacture of wearing apparel, except fur apparel.

0.59

154 Manufacture of other food products

0.39

014 Agricultural and animal husbandry service activities, except veterinary

0.39

Source: Annual Survey of Industries 2005-06

There are 1,379 products at eight digit level, which are produced by these industries and categorized as gender sensitive products.

263

8.7. Conclusions and Policy Directions

Some broad conclusions that emerge and policy directions which may be inferred from the results are as follows: The results of the chapter show that with respect to India, trade has provided more opportunities of employ yment to women and this is particularly true for export-oriented industries. Higher participation of women in industries which are expanding because of exports indicates that export-oriented policies can be instrumental in gender empowerment in case of India. Trade policies can therefore be designed to provide special incentives to export-oriented units favouring higher women employment. For example, a threshold can be decided for share of women in total employment, above which the incentives can be provided. The results also highlight that women’s education and skill accumulation are the most important factors determining the impact of trade on women’s employment and the gender wage gap. As long as women remain less qualified than men, they are likely to remain in lower paying, less secure jobs, even if better-paying jobs become available through trade expansion. Education and skills also provide greater flexibility and power to negotiate wages and other work conditions. It can therefore be said that the process of globalisation is no longer an enclave process which may influence only those that participate in trade. The impact of globalisation reaches all sectors and sections of the societies irrespective of their level of participation in the process. This has necessitated the need to formulate trade policies to incorporate concerns of all affected. Gender is one area which has remained out of the orbit of trade policy formulation in many countries, especially developing countries. One of the major challenges faced by trade policy makers is to ensure gender equity in distribution of gains of trade. For this purpose,

• it is important to assess the impact of trade on gender employment and wages in different sectors;

• identify the sectors where gender inequality is high which would imply that any growth of trade in the sector will further increase gender inequality;

• identify the sectors which provide potential for improving gender equality; and then formulate sector-specific policies.

*********