The Contribution of OFW Remittances to Income …...II. Review of Related Literature Much...
Transcript of The Contribution of OFW Remittances to Income …...II. Review of Related Literature Much...
The Contribution of OFW Remittances to Income Inequality: A Decomposition Analysis
In fulfillment of the requirement of the course Economics 199
Submitted by
Nico M. Ravanilla Eduard Joseph P. Robleza
Submitted to the School of Economics
March 2003
ii
WAIVER
We, Nico M. Ravanilla and Eduard Joseph P. Robleza, allow any part of this thesis to be used or
reproduced in any form and for any purpose whatsoever, provided that the proper citation is given to the
authors.
Nico M. Ravanilla Eduard Joseph P. Robleza
iii
To the overseas Filipino worker.
iv
Acknowledgment For the greater glory of God. Ad majorem dei gloriam.
The authors would also like to thank the following:
- our families, for obvious reasons;
- Dr. Gwendolyn R. Tecson, for her guidance;
- Ms. Sharon Faye A. Piza, for her tolerance (we still owe you that free lunch);
- the kind people at the UPSE Library and Computer Laboratory;
- the people at the ETC tambayan for their love (Robert says hi!);
- Ms. Anne Imelda Pascua, for desperately wanting to be acknowledged;
- and to the people whom we may have forgotten, insert your name here:
“We also thank ____________________.”
Thanks to the people who are reading this. Thanks to those who aren’t.
v
Table of Contents
ABSTRACT ..................................................................................................................................... ix
I. INTRODUCTION...........................................................................................................................1
WORKING HYPOTHESIS 2 SIGNIFICANCE AND OBJECTIVES OF THE STUDY 3 OUTLINE OF THE STUDY 4
II. REVIEW OF RELATED LITERATURE........................................................................................5
FOREIGN WORKS 5 PHILIPPINE LITERATURE 8
III. THEORETICAL FRAMEWORK................................................................................................11
MIGRATION AND MOTIVATIONS 11 MIGRATION AND INCOME DISTRIBUTION 12
IV. METHODOLOGY .....................................................................................................................15
DECOMPOSITION ANALYSIS 15 THE DATA 17
V. PRESENTATION OF RESULTS AND ANALYSIS .....................................................................19
THE MIGRATION CHAIN 19 THE COMPOSITION OF INCOME INEQUALITY 23 THE CONTRIBUTION OF REMITTANCES TO TOTAL INCOME INEQUALITY 31
VI. CONCLUSIONS AND RECOMMENDATIONS.........................................................................40
POLICY RECOMMENDATIONS 40 RECOMMENDATIONS FOR FURTHER STUDIES 42
vi
SELECTED REFERENCES.............................................................................................................44
APPENDICES .................................................................................................................................47
APPENDIX A: CHOOSING A MEASURE OF INEQUALITY: THE FIVE PROPERTIES OF DECOMPOSABILITY 47
APPENDIX B: COMPARING THE DECOMPOSITION RESULTS OF USING
THE GINI COEFFICIENT AND THE GENERALIZED ENTROPY (GE) CLASS OF INEQUALITY MEASURES 49
APPENDIX C: BACKGROUND INFORMATION ON PHILIPPINE MIGRATION 51 APPENDIX D: TABLES A1 – A6: COMPOSITION OF HOUSEHOLD INCOME
INEQUALITY, BY YEAR 55
vii
List of Tables
Table Number
1 2
3A
3B 4 5 6 7 8 9
10
Title
Deployed Overseas Filipino Workers and Total Remittances from Overseas Filipinos, 1984 – 2000 Number of Households Receiving Income from Abroad Average Family Income (in Pesos), Urban – Rural Average Family Expenditure (in Pesos), Urban – Rural Shares of Income Sources in Total Income Gini Coefficients Percent of Households Receiving Income from Abroad Gini Correlation Ratios Contributions of Income Sources to the Gini (in Percent) Relative Concentration Coefficients Distribution of Overseas Filipino Workers, By Major Occupation, 1995 - 2000 (in Percent)
Page
20
20
21
21
24
27
28
29
33
36
38
viii
List of Figures
Figure Number
1 2 3 4 5
Title
Share of Remittances in Total Income Gini Coefficient of Remittances Gini Correlation Ratio of Remittances Contribution of Remittances to Total Income Inequality (in Percent) Relative Concentration Coefficients of Remittances
Page
24
27
31
33
37
ix
Abstract The paper aimed to determine whether remittances from overseas Filipino workers
improved or worsened income inequality. Using decomposition equations, the paper
divided total inequality into its four components, namely wages, entrepreneurial incomes,
other income, and remittances from migrants. The decomposition exercise revealed that
the contribution of remittances to overall income inequality depended on their share in
total income, their distribution among the population, and their correlation with total
income. Remittances were found to accrue mostly to higher- income classes, but they
were found to be gradually becoming less inequality- increasing over time. Therefore,
policies that would aim to reduce income inequality should consider making migration-
facilitating factors more accessible to those in the lower ends of the distribution, because
remittances would only tend to contribute less to income inequality if the lower- income
brackets were also able to migrate. Finally, further studies should look into the
implications of the changes in inequality, especially with regards to welfare.
1
I. Introduction
The Philippines has historically been one of Asia’s largest labor exporters. In the 1960s,
Filipino doctors and nurses took the opportunity to work abroad when the United States
liberalized its immigration law. This was followed by the migration of seamen, engineers and
construction workers to the Middle East during the oil shock in the 1970s and 1980s. The trend
has continued in the nineties and up to the present, as domestic workers, nurses and teachers
respond to increasing job openings all over the world.
One strong impetus that compels Filipino workers to migrate is the Philippines’ persistent
problem with unemployment and low economic growth rates. In fact, the Philippines’ GNP per
capita has grown at an average annual rate of just a little over 1% from 1965 to 1990, a rate that
is below those enjoyed by the country’s Southeast Asian neighbors (World Bank 1995). Faced
with little or no opportunities at home, Filipinos migrate to where the jobs are; migration, after
all, is a decision in response to opportunities available elsewhere.
The decision to migrate is also based on the expected benefits and costs associated with
taking a job abroad. One of such benefits that have strongly attracted Filipinos to overseas
employment is the higher wage being offered in foreign countries. With higher wages, overseas
Filipino workers (OFWs) obviously earn higher incomes that can be saved for future
consumption. Income from overseas employment also affords households the opportunities to
invest in other profitable pursuits such as in small family businesses or in their children’s
education—the benefits of which can be reaped later on. Indeed, this fact leads many to claim
that overseas employment lessens income inequality, because households with income from
abroad are believed to be able to afford a higher standard of living than otherwise possible.
2
The Philippine government has always acknowledged this positive effect of overseas
employment on the welfare of its citizens—and with good reason. OFW remittances have
contributed much to the country’s GNP and foreign exchange earnings. In fact, remittances
comprised, on the average, about 2% of GNP in the eighties, and that contribution has more than
doubled to 4.8% in the nineties (Tan 2000). These remittances are used to service the country’s
foreign debt, and the export of labor is thus seen as a means to achieve or sustain economic
development. In addition, the Philippine government has established secure safety nets—the
Philippine Overseas Employment Administration (POEA) and the Overseas Workers’ Welfare
Administration (OWWA)—to protect the interests of the millions of migrant Filipinos. These
agencies have also lowered the transaction costs associated with overseas employment. In effect,
not only has the government given everyone almost equal opportunities to migrate, but it has
made migration an even more attractive and worthwhile option for its workforce as well.
Working Hypothesis
Despite these obvious and substantial economic benefits, however, there is still reason to
believe that the effects of overseas employment and remittances may not always be positive.
Because migration entails both monetary and psychological costs, some workers in certain
classes are already precluded from migrating. Even if costs are made lower through government
assistance, these workers may still not be able to find jobs abroad, especially if they do not meet
the required qualifications of age, education and experience, because overseas employment is
basically demand-driven.
In other words, the chances for migration may not always be equally distributed among
the population. Studies indicate that Filipino migrants, whether temporary or permanent, are
3
most likely to come from only certain groups of people. More specifically, households with
remitting migrants predominantly come from urban areas and wealthier families, implying that
only these households can more readily assume the costs and risks of migration (Rodriguez and
Horton 1995; Rodriguez 1998). In addition, migrants are also seen to be more educated than non-
migrants. In 1990, OFWs had an average of 10.6 schooling years, while the entire labor force
had only 7.9 years (Tan 2000). This implies that only those with the required skills, training or
experience—which can be too costly for others to acquire—can more easily take jobs abroad and
hence benefit from higher incomes.
There is therefore reason to suspect that the benefits associated with migration,
particularly earnings and remittances, may not be distributed evenly as well. Remittances may
not always improve the distribution of income, and contrary to what many believe, they may
even contribute to income inequality within the country.
Significance and Objectives of the Study
The question of whether remittances from abroad serve to alleviate or worsen income
inequality is of great interest, particularly because of the high level of inequality in the country.
In addition, the country has faced massive outflows of workers in recent years. Considering that
these workers send in an amount almost equal to 5% of GNP in the 1990s, the overall effect of
these remittances on household income distribution appears to be substantial and thus worth
looking into.
There is also a dearth of empirical Philippine studies about the topic. Moreover, the few
studies that exist focused only on the short-term effects of international remittances on income
distribution, covering only one specific year, or up to three years at most. And because their
4
conclusions naturally varied, the relationship between remittances and income distribution
seemed unclear. At best, there is still no general consensus about the effect of migrants’
remittances on income inequality.
In this regard, this paper recognizes the importance of analyzing the said relationship
over a much longer period of time. Migration is a dynamic phenomenon; its flows and structure
change, as other variables, such as the demand for and the skill levels of migrant workers,
fluctuate over time. Hence, by looking at a longer time span, the paper aims to arrive at an
observable trend that allows for a firmer, more realistic statement about the relationship between
remittances and income distribution. This, in turn, will be of help in making policy
recommendations. Furthermore, the paper hopes to contribute to the existing literature on
remittances and income inequality in the Philippines.
Outline of the Study
The paper is divided into six sections. Section II reviews the related literature on the
topic. Sections III and IV outline the theoretical framework and the methodology used,
respectively. Section V presents and analyzes the results. Finally, Section VI summarizes the
main conclusions and offers policy recommendations.
5
II. Review of Related Literature
Much speculation has been made about the sources of income inequality in the
developing world ever since Kuznets (1955) wrote about economic development and income
inequality. Several empirical studies have decomposed the sources of income inequality by
economic sector (e.g., urban vs. rural), family characteristics (e.g., education and occupation of
workers), as well as by income source (e.g., income from labor vs. capital vs. land) (Adams and
Alderman 1992).
Foreign Works
It was not until the late eighties and early nineties when empirical studies attempted to
isolate the effects of labor migration on income inequality. Using labor migration survey data
from the 1980s, the relationship between international remittances and income inequality was
examined for rural Mexico, Egypt, Pakistan, and for some other Asian nations.
Stark, Taylor and Yitzhaki (1986) studied the effect of interna tional remittances on the
income distribution of two rural Mexican villages with different migration histories. The results
were contrasting. Remittances were found to reduce inequality in the village that had a longer
history of migration to the United States and hence a more ready access to United States labor
markets. On the other hand, remittances increased inequality in the village that had only a few
households that had experienced migration to the United States. Meanwhile, the more recent
study by Stark and Taylor (1991) further argued that the propensity of households to participate
in international migration from Mexico to the United States was directly related to the
households’ initial relative deprivation. This implied that relatively deprived households were
6
more likely to engage in international migration than were households more favorably situated in
their village’s income distribution.
A series of investigative studies organized by the International Labour Organization in
the mid-eighties examined the economic impact of overseas migration on labor-sending Asian
countries (Amjad 1989). As far as testing the relationship between income inequality and
international remittances was concerned, however, these studies were mainly descriptive, since
they did not use regression equations or decomposition analyses to test the said relationship. The
study on South Korea, for instance, merely reported of a highly positive correlation between
migration and the income disparity of urban households: The Gini ratio in the 1970s when
migration prevailed was higher than in the 1960s, implying a worsening of income distribution
(Hyun 1989). However, mere correlation does not necessarily mean that migration cannot
improve income distribution at all or will automatically aggravate it, since overseas employment
does allow some of those in the lower- income brackets to earn more than otherwise possible.
While also mainly descriptive, the studies on Bangladesh (Mahmud 1989), Thailand
(Tingsabadh 1989), and Sri Lanka (Rodrigo and Jayatissa 1989) suggested that the income-
distributional consequences of migration and remittances depended on the relative income
classes from which migrants’ families originally came. In Bangladesh and in Thailand, migrants
were found to have come from a more affluent background, since those from the poorer classes
could not afford the initial costs of migration. This implies that migration benefits the higher-
income classes and only serves to widen the income gap between them and the lower- income
classes. In contrast, most of Sri Lankan migrants came from the lower-income and low-skilled
groups, and this was seen to have moderated income disparities in Sri Lanka.
7
Meanwhile, Adams (1991) conducted a different study, this time focusing on the effects
of international remittances on poverty, inequality and development in rural Egypt. His study
used predicted income equations to estimate the changes that would occur between two
contrasting situations: excluding remittances, where the remittances of the correspondents with a
still-abroad migrant were excluded, and including remittances, where the remittances of these
still-abroad migrants were included. The study found that while international remittances helped
alleviate poverty, they nevertheless had a negative, worsening effect on income distribution.
Furthermore, the studies by Adams and Alderman (1992) and by Adams and He (1995)
on rural Pakistan provided a good foundation in the decomposition analysis of the sources of
income inequality. The decomposition analyses of these studies determined exactly what sources
of income contributed to the total income inequality, as well as how much these income sources
actually contributed to the observed inequality. It should be stressed that the main objective of
these two studies was only to analyze the determinants of rural poverty in Pakistan and not to
focus on and assess the impact of international remittances on income inequality, as evident in
both the data set and the methodology used. The study by Adams and Alderman, for instance,
used a 1986–1989 survey which, as the authors noted, “was not designed to be representative of
the rural population as a whole in Pakistan” (1992). Furthermore, in the decomposition of the
Gini coefficient, all transfers—international remittances, domestic or internal remittances,
pensions, and payments of the government to the poor—were lumped together as one source of
income. The effect of international remittances on income inequality was thus not isolated.
8
Philippine Literature
Just as empirical studies abroad remain in their early stages, Philippine studies about
international migration and its effects on income distribution have been few. In fact, some of the
more substantial studies did not come out until the early nineties. For instance, the work by
Untalan in 1980 looked at the composition and supply of workers who go to the Middle East and
analyzed the costs and benefits of labor exports. However, it barely dealt with the direct effects
of labor migration on domestic income inequality. Also, Macaraig and Sarino (1980) wrote an
undergraduate thesis on the remittances of contract workers, although the study focused more on
the direct effects of labor exports on welfare than it did on income distribution. By examining the
decile movements in income shares of overseas contract workers, the study showed that the
remittances of contract workers generally tended to decrease the welfare of families when the
30% mandatory remittance (under the 30% Remittance Law) served as a floor.
More than a decade later, Tan (1991) studied overseas employment, saving rate and
income distribution. Using the 1988 Family Income and Expenditures Survey, or FIES, she
observed that overseas employment had an equalizing effect on family income. The
improvement in income distribution was shown by the apparently much better distribution of
income by size for families with income from abroad than for families without such income.
More precisely, the statistical analysis showed that much fewer families with income from
abroad belonged to the category ‘poor’ (taking the poverty income to be approximately
PhP3000): 15.1% vs. 39.4% in urban areas, and 41.1% vs. 73.4% in rural areas.
However, the results of both Macaraig and Sarino’s and Tan’s studies could no t be
considered conclusive, because these studies did not specify whether the families sampled were
already poor even before the income from abroad was added in. In addition, these two studies
9
relied mainly on decile movements and size distributions rather than on some direct measure of
inequality, such as the Gini coefficient, which could be source decomposable; i.e., the total
inequality could be broken down into a weighted sum of inequality by various income sources.
In effect, these studies might not have considered that households also received incomes from
other sources, such as domestic wages or entrepreneurial activities. Thus, these two papers could
not conclusively determine or claim exactly how much of the overall inequality was actually due
to remittances, if at all.
Rodriguez and Horton’s 1995 working paper arrived at a conclusion very much at odds
with those of Macaraig and Sarino’s and Tan’s. Using the 1991 FIES, Rodriguez and Horton
compared households with migrants who remitted and households that did not have migrants or
remittances. The paper found out that the former group was predominantly urban and had a
higher total income and expenditure than the latter group. The paper also noted that the heads of
the former group were slightly older and more likely to be female, and that although these
household heads were less likely to be employed, they nonetheless had more education. All these
suggested that migration might exacerbate income inequality, since migrants came from urban,
wealthier, better-educated households.
Furthermore, by decomposing the Gini coefficient, Rodriguez, in his 1998 article, only
furthered the conclusion he made with Horton earlier that remittances actually worsened income
inequality in the Philippines. In the same paper, Rodriguez also pointed out some methodological
concerns with earlier Philippine studies. When income shares of households were ranked by their
pre-remittance income (defined as total income minus remittances from migrants), the procedure
implicitly assumed that if the migrants had stayed at home, their contribution to household
income would have been zero. This method of calculation would thus underestimate inequality
10
and make it appear that when the distribution was kept constant and remittances were added,
remittances would dramatically lift households from poverty.
The most recent study that was available was a paper by Balisacan and Piza (2003).
Unlike Rodriguez’s article, Balisacan and Piza’s paper decomposed inequality using the
Generalized Entropy class of inequality measures. The paper provided an excellent general
discussion about the nature and causes of inequality in the Philippines, although it did not focus
specifically on the impact of remittances upon income inequality. The results of the
decomposition, however, suggested that there was “weak evidence to the common claim that
overseas remittances in recent years (have) accentuated income inequality.” More importantly,
the paper could serve as a basis for comparison in checking whether the results would be
consistent even if different measures of inequality were used in the decomposition.
This paper takes all these into consideration. Before proceeding with a detailed analysis,
though, this paper first sets the context within which such analysis is to be made. The next
section therefore discusses the theoretical framework used.
11
III. Theoretical Framework
Migration is essentially a factor movement. Labor responds to employment opportunities
and wage differentials, which are determined by developments both in the sending economy and
in the receiving markets.
Migration and Motivations
Migrant labor responds to job opportunities abroad, but the opportunities tend to be
limited because of various economic, political and even cultural barriers. For example, workers
and labor unions in the host or receiving economies often resist the inflow of migrant labor due
to fears of unemployment among their ranks. Labor exporters like the Philippines, however,
respond to these openings by having a flexible or mobile labor force, since these openings tend
to be limited to a select set of occupations with strict qualification requirements. In this sense,
migration is seen as a human capital problem of simply maximizing the returns to the education
or training that the potential migrant has undertaken (Tan and Canlas 1989).
Relative wage rates also determine the direction of labor flows, and indeed, this monetary
benefit is often assumed to be the main motivation that encourages migration. The Philippine
case illustrates this concept. The recession after the balance-of-payments crisis in the 1980s
resulted in massive unemployment. Real wages then sank to levels that were lower than real
wages elsewhere, and as expected, many took jobs abroad (Tan 1991).
In addition, when people compare incomes within a reference group—such as their local
community, for example—the comparisons create psychic costs or benefits, or feelings that Stark
and Bloom (1985) call “relative deprivation or relative satisfaction.” A person may then decide
to migrate from one location to another either to change his relative position within his reference
12
group or to move to another reference group altogether. Generally, a person who feels more
relatively deprived is expected to have a stronger incentive to migrate than a person who feels
less so. As particular individuals reap the benefits of migration, the relative deprivation
perceived by non-migrants may change, thereby inducing them to migrate as well. Migration can
thus be viewed alternatively as “a diffusion process [wherein] the level of migration at any point
in time is likely to be positively related to past migration” (Stark, Taylor and Yitzhaki 1986).
Migration and Income Distribution
At the beginning of a country’s migration history, when few households have just started
establishing contacts at the receiving economy, the distribution of remittances—and possibly of
income as well—is “necessarily unequal” (Stark, Taylor and Yitzhaki 1986), which can be
attributed to a number of observable facts. One is that, during this time, information about the
working conditions abroad is scarce and costly; because of the uncertainty that this creates,
migration, in a sense, becomes a high-return yet high-risk investment. Moreover, the initial costs
of migration—the fees of recruiting agents, the costs of processing passports, visas and airline
fares—preclude the poorer sectors from migrating. Thus, the first households that will invest in a
migration decision are likely to come from the upper rungs of the income distribution, since they
are more financially capable to assume its costs and risks.
In addition, given that migration is demand-driven, any inequality in the distribution of
income puts those in the lower- income classes at an even greater disadvantage. Working abroad
may require specific skills, a certain level of education or experience, and even English
proficiency, depending on which foreign labor market or occupation opens its doors to migrant
labor. Thus, unless the low-income classes can afford to acquire the skills and levels of education
13
demanded by overseas work, then migration may remain to be a valid option only for those in the
high- income classes.
The impact of remittances in the overall household income distribution at this initial stage
therefore depends firstly upon the magnitude of remittances in relation to income from other
sources. If remittance-receiving households rely significantly on remittances from abroad to
augment their income, then the distribution of remittances is expected to greatly contribute to the
distribution of income. If, on the other hand, remittances represent only a small part of household
incomes, then remittances will have a minimal effect on income distribution.
In addition, the impact of remittances on income inequality also depends on which
income classes migrants originally come from. If migrants primarily come from the relatively
well-off, then their remittances may worsen income inequality. However, if migrants also come
from the lower classes, then remittances may contribute to greater income equality.
Over time, successful migrants provide valuable information to those left behind, and this
may raise the latter’s propensity to migrate. These would-be migrants, now feeling more
relatively deprived, will also want to migrate in order to shift the distribution of returns to
migration in their favor. Moreover, the early migrants make migration a much more worthwhile
undertaking because the information they bring reduces the risks and uncertainty associated with
it. In some instances, too, early migrants provide direct assistance to new migrants by financing
the costs of migration, especially for family and friends. In other words, migration flows tend to
create a “migration chain” that generates opportunities for later migration (Tan 2000).
Therefore, the effect of remittances on income inequality over time depends on how
information and factors that facilitate migration become diffused through the population. If
access to these factors easily spread across households, then migration—and hence the receipt of
14
remittances—by those at the lower ends of the income distribution is likely to occur, possibly
reducing the initially unfavorable effect of remittances on income inequality (Stark, Taylor and
Yitzhaki 1986).
Furthermore, the types of labor or skills demanded abroad may change over time,
depending on the developments in the labor markets of the receiving economies. These shifts in
demand for migrant labor may in turn influence the income distribution in the sending countries.
For instance, if the demand for labor abroad shifts towards those occupations that require lower
skill and education levels, then migration opportunities are opened up especially for those in the
lower income brackets, who may be constrained from acquiring higher skill levels. In this case,
migrants’ remittances may contribute to greater income equality.
More importantly, if migration is indeed a diffusion process as suggested above, then its
impact must now be seen as stage-specific (Stark and Bloom 1985). The effect of remittances on
income distribution may vary depending on which point in a country’s migration history is being
considered, because the level of diffusion of migration-facilitating factors as well as the skills
demanded by overseas employment vary at each point. This should partly explain the divergence
of views about the effect of remittances on income inequality, because the empirical studies did
not temper their conclusions with the fact that the observations were made at distinct and specific
stages of the migration process. Moreover, because the impact of international remittances and
income inequality is stage-specific, looking at longer time spans logically suggests that this
impact may also change over time. More explicitly, while remittances do seem to worsen income
inequality at the start of the migration chain, they may be expected to gradually reduce
inequality, as more people from the lower ranks of the distribution are enticed and are able to
migrate.
15
IV. Methodology
Taking this stage-specificity into consideration therefore requires, not only looking at the
context of a specified period of time, but also dissecting the income stream of households over
this period. Households receive income from various sources, such as wages and entrepreneurial
activities. It is in this regard that migration is a source of remittance income, for the decision to
migrate only “reflects a desire to diversify income sources” (World Bank 1995). For example, a
rural household may choose to send one of its members to work in the city or abroad while the
rest stay and work in the family farm. In any case, international remittances are just one source
of household incomes for any period in time, and an income decomposition ana lysis will isolate
their impact on total income inequality. To account for migration as a stage-specific diffusion
process, the decomposition will also be done repeatedly for several periods.
Decomposition Analysis
While any measure can easily allocate income inequality among the components of total
income, Adams and Alderman (1992) have shown that it is appropriate to use the Gini
coefficient, since this measure satisfies the five basic properties or axioms of decomposability:
(1) Pigou-Dalton Transfer Principle, (2) Income Scale Independence, (3) the Principle of
Population, (4) Anonymity, and (5) Decomposability. (A discussion of these five properties is
found in Appendix A, while the implications of using different inequality measures are
summarized in Appendix B.)
Following Stark, Taylor and Yitzhaki (1991) and, more recently, Adams and He (1995),
the source decomposition based on the Gini coefficient of total income, G, can be written as
G = 2[cov(y,r)]/? , (1)
16
where y is the series of total incomes, and r is the cumulative distribution of total incomes. The
Gini coefficient of the ith source of income, Gi, can therefore be expressed as
Gi = 2[cov(yi,ri)]/? i, (2)
where yi and ri refer to the series of incomes from the ith source and their corresponding
cumulative distributions, respectively. Since total income is the sum of all source incomes, the
covariance between total income and its cumulative distribution can be expressed as the sum of
covariances between each source income and cumulative distribution of total income. Equations
(1) and (2) can then be used to express the total income Gini as a function of the source Ginis:
G = ? SiRiGi, (3)
where Si is the share of each source of income to total income, or
Si = ? i/? ? ? ? ? ? ? ???
and R is the Gini correlation ratio, expressed as
Ri = cov(yi,r)/cov(yi,ri) (5) = covariance between source income amount and cumulative distribution of total income covariance between source income amount and cumulative distribution of source income
The Gini correlation ratio, R, ranges between +1 and –1. An income source is an
increasing (decreasing) function of total income as R approaches +1 (–1). When the income
source is a constant, R equals zero, implying that that income source’s share of the Gini is zero.
The decomposition corresponding to the Gini coefficient can therefore be expressed by
the following terms:
? wigi = 1; wi = ? i/? ; gi = Ri(Gi/G) (6)
where wigi is the factor inequality weight of the ith source in overall inequality, and gi is the
relative concentration coefficient of the ith source in overall inequality.
17
Assuming that additional increments of an income source are distributed in the same
manner as the original units, an income source can be defined as inequality- increasing or
inequality-decreasing on the basis of whether additional shares of income from that source lead
to an increase or decrease in overall income inequality. Based on Equation (6), it follows that the
ith income source is inequality- increasing or inequality-decreasing according to whether gi or the
relative concentration coefficient is greater than or less than unity.
The Data
Data for constructing the decomposed Gini indices are taken from the various Family
Income and Expenditure Survey (FIES) rounds from 1985 to 2000, with each round conducted
every three years. The FIES is an integrated survey of households that seeks to gather data on
family income and expenditures, determine the degree of inequality among families, and provide
benchmark information for the estimation of the consumer price index. While the FIES for 1961,
1965, and 1971 are also available, they are excluded from the analysis since their unit-record
data are not available for the decomposition.
Definition of Terms
The empirical derivation of the decomposed Gini indices is accomplished using data from
the Philippines, as categorized into ‘all Philippines’, urban, and rural. The FIES of 1997 defines
urban areas as follows:
1. In their entirety, all cities and municipalities having a population density of at
least 1,000 persons per square kilometer.
18
2. Poblaciones or central districts of municipalities and cities which have a
population density of at least 500 persons per square kilometer.
3. Barangays having at least 1,000 inhabitants where the occupation of inhabitants
is predominantly non-farming or non-fishing.
All areas not falling under any of the above classifications are considered rural.
Four components of household income, and therefore of income inequality, are
considered: wages, entrepreneurial income, other income, and remittances from migrants or
income from abroad. Wages include all forms of compensation received by family members who
are either regular or occasional workers in agricultural and nonagricultural industries.
Entrepreneurial income includes earnings from farming, trading, manufacturing and other
entrepreneurial activities engaged in by any member of the family as operator or as self-
employed. Other income includes the family’s net share of crops produced, rental value of
owner-occupied dwelling units and other family sustenance activities. Finally, Income from
abroad includes cash receipts from family members working abroad as contract workers, as well
as gifts and other forms of assistance from other migrants, particularly family members with
permanent residence abroad (or emigrants).
Data Limitations
Because of the tendency of some households to underreport their incomes, especially
richer families, it is possible that income data from the FIES are understated (Rodriguez 1998).
As a consequence, conclusions about the contribution of international remittances to total income
inequality may reflect this downward bias; the magnitude of income inequality may be
underestimated.
19
V. Presentation of Results and Analysis
As evident in Equation (3), the decomposition exercise outlined above can illustrate how
(a) the magnitude of remittances to total income, (b) the unequal distribution of remittances, and
(c) the correlation of remittances with total income all contribute to the observed inequality. The
first part of this section sets the context of the analysis against the theoretical framework by
describing the changes and trends in the decomposition variables. Having set the context, the
paper then proceeds to discussing the contribution of remittances to overall income inequality.
The Migration Chain
The Philippines has had a fairly long history of migration, which dates back to American
colonial times (Sto. Tomas 1984). With increasing demand for migrant labor and with some
encouragement from the Philippine government, migration has gradually become a worthwhile
option for Filipino workers over the years. (Appendix C discusses the country’s history of
migration and policies on remittances in more detail.) This supports the theory that migration is a
diffusion process wherein the benefits of migration induce non-migrants to migrate in the future.
Because of this diffusion process, one may expect to see a gradual rise in the number of
deployed workers over time—or a migration chain. This phenomenon can be observed from
Table 1, which shows that the number of deployed OFWs grows as time passes. Assuming that at
least some of these new migrants remit some fraction of their earnings back home like permanent
migrants and contract workers do, one may infer that this migration chain logically—although
not always necessarily—results in more households receiving income from abroad, as Table 2
indicates.
20
Table 1. Deployed Overseas Filipino Workers and Total Remittances from Overseas Filipinos, 1984 – 2000.
Year Deployed OFWs* Total Remittances (US$ Million)
1984 350982 602.9 1985 372784 687.2 1986 378214 680.4 1987 449271 791.9 1988 471030 856.8 1989 458626 973.0 1990 446095 1181.1 1991 615019 1500.3 1992 686457 1769.5 1993 696630 2229.6 1994 719602 3008.1 1995 654022 3868.3 1996 660122 4306.6 1997 747696 5741.8 1998 831643 4926.0 1999 837020 6794.6 2000 841628 6050.4 Source: Department of Labor and Employment. Yearbook of Labor Statistics, 2000.
* Land-based and Sea-based OFWs.
Table 2. Number of Households Receiving Income from Abroad.
1985 1988 1991 1994 1997 2000 Urban Philippines 343909 332890 493600 557121 523704 681266 Rural Philippines 224615 273805 281400 304404 357558 425239
All Philippines 568524 606695 775000 861525 881262 1106505
Urban/Rural 1.5 1.2 1.8 1.8 1.5 1.6 Source: National Statistics Office. FIES, various years.
Thus far, the statistics in Tables 1 and 2 have shown that past migration flows have
created opportunities for later migration. The growing numbers of deployed OFWs may imply
that the risks associated with migration have diminished, possibly because the information about
the working conditions and successes experienced by earlier migrants has accumulated
sufficiently to become public information (Tan 2000). However, the growing numbers of
deployed OFWs cannot yet be automatically attributed to an improvement in the labor mobility
of those in the lower ends of the income distribution in the sense that they have gained more
21
access to education or skill training. What can be implied from the statistics, especially from
Table 2, is that the opportunities for migration have not been equally distributed among urban
and rural households. In fact, the urban-rural ratios in Table 2 show that remittance recipients in
urban areas have continued to outnumber those in rural areas, and that they may even continue to
do so since the ratios show hardly any decline. All in all, urban households are more able to
pursue jobs abroad, as various other studies have already claimed and proven (see, for example,
Rodriguez and Horton 1995, and Tan 1991).
Among the reasons found regarding why more migrants typically come from urban
households is that urban households are wealthier (Rodriguez and Horton 1995), which Tables
3A and 3B confirm. With higher incomes, urban households can more easily assume the initial
costs of migration, such as those of acquiring the education required by overseas jobs.
Table 3A. Average Family Income (in Pesos), Urban – Rural.
1985 1988 1991 1994 1997 2000 All Philippines 31,052 40,408 65,186 83,161 123,168 144,039
Urban Philippines 46,127 60,330 89,571 113,121 178,121 204,977 Rural Philippines 21,875 28,284 41,199 53,483 73,319 85,373
Urban/Rural 2.11 2.13 2.17 2.12 2.43 2.40
Source: National Statistics Office. FIES, various years.
Table 3B. Average Family Expenditure (in Pesos), Urban – Rural.
1985 1988 1991 1994 1997 2000 All Philippines 26,865 32,521 51,991 67,661 99,537 118,002
Urban Philippines 39,134 47,299 70,551 91,115 140,955 164,794 Rural Philippines 19,397 23,529 33,733 44,427 61,966 72,953
Urban/ Rural 2.02 2.01 2.09 2.05 2.27 2.26
Source: National Statistics Office. FIES, various years.
22
Because of differences in wealth, urban and rural households are also presumed to differ
in education and skill levels (although the causation between wealth and education may have
worked the other way as well). In addition, urban households also have an advantage over rural
households because the former group faces lower transaction and placement costs when it comes
to migration. Placement offices are usually located in urban centers. Information about overseas
employment openings and conditions is also more difficult to access from rural areas because of
the limited availability of media, especially access to the Internet. Because of these, Tan (1991)
surmises that workers in rural areas probably resettle into urban centers first before migrating to
foreign destinations. This hypothesis is not entirely unexpected, given that wages are also higher
in urban areas, since the theory of human capital does predict that workers tend to move to where
wages are higher and where employment is available (Ehrenberg 2000).
Given also that urban households generally have better opportunities to migrate than rural
households, an analysis of the distributional impact of remittances as migration opportunities
spread across the population can therefore be carried out by considering separately the effect of
overseas remittances on urban and on rural income inequality. Such an analysis is actually a
variation of the one done by Stark, Taylor and Yitzhaki (1986) on the income distribution of two
Mexican villages, since the villages also differ in terms of their access to foreign labor markets.
In this sense, comparing the composition of urban and rural income inequalities will be
analogous to considering two contrasting conditions where migration opportunities in one
scenario are relatively more diffused than in the other.
23
The Composition of Income Inequality
The theoretical framework argues that the impact of migrants’ remittances upon income
inequality tends to become more favorable (or less unfavorable) as migration opportunities are
diffused through the population over time. When viewed in this context, the results of the
decomposition exercise seem to support the argument.
Shares of Income Sources in Total Income (Si)
Table 4 shows the shares of the four income sources in total income for 1985 – 2000. The
proportions are similar to those found by Rodriguez (1998) for 1991 Philippines. Paid wages
have been the most important income source in urban areas, representing, on the average, about
47% of total household incomes. In the rural areas, however, entrepreneurial income is a more
significant source of income, averaging about 38% as against only 35% for paid wages. This is
so because of a higher incidence of self-employment in the rural areas (Rodriguez 1998).
Nevertheless, paid wages exhibit an increasing trend for the entire country, while entrepreneurial
income shows the opposite. Other sources of income consistently represent just about a fifth of
total incomes for the entire Philippines.
Remittances, on the other hand, represent less than 10% of total incomes, but it appears to
be a more important income source in the urban areas than in rural areas. Moreover, the share of
remittance income in total income does not exhibit a clear trend for the entire Philippines, which
Figure 1 illustrates. It was as high as 8.3% in 1985 then it dropped to 6.8% in 1997. The share
then spiked back up to 8.6% in 2000. This erratic trend may most probably be due to movements
in the shares of the other domestic income sources. The share of remittances in total income has
dropped, for instance, because the shares of domestic incomes like wages have increased.
24
Table 4. Shares of Income Sources in Total Income.
Income Source 1985 1988 1991 1994 1997 2000
All Philippines Wages 0.3825 0.4336 0.4300 0.4518 0.4647 0.4586 Entrepreneurial income 0.2836 0.2822 0.2844 0.2530 0.2618 0.2444 Other income 0.2513 0.2139 0.2018 0.2197 0.2056 0.2109 Remittances from migrants 0.0826 0.0703 0.0838 0.0755 0.0680 0.0861
Total 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
Urban Philippines Wages 0.4213 0.4671 0.4585 0.4756 0.4939 0.4843 Entrepreneurial income 0.2247 0.2328 0.2476 0.2173 0.2287 0.2121 Other income 0.2625 0.2228 0.2036 0.2266 0.2058 0.2119 Remittances from migrants 0.0915 0.0773 0.0903 0.0804 0.0716 0.0916
Total 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
Rural Philippines Wages 0.3044 0.3627 0.3356 0.3707 0.3732 0.3776 Entrepreneurial income 0.4023 0.3867 0.4063 0.3746 0.3654 0.3459 Other income 0.2286 0.1953 0.1960 0.1963 0.2047 0.2077 Remittances from migrants 0.0646 0.0553 0.0621 0.0585 0.0567 0.0688
Total 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
*NOTE: Figures may not add up to totals correctly due to rounding.
Figure 1. Share of Remittances in Total Income.
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
1985 1988 1991 1994 1997 2000
Year
Sha
re
All Philippines
Urban
Rural
25
Furthermore, that the share of remittances is higher for urban households is expected.
Stark, Taylor and Yitzhaki (1986) saw that the shares of international remittances were higher in
the Mexican village that had had more migrants to the United States. Hence, it seems reasonable
to expect the share of remittances in urban households’ incomes to be higher than in rural
households’, since migration is a more practicable or possible employment option in urban areas.
This, in turn, may ultimately be due to the fact that migration-facilitating factors, such as access
to information, are more available and relatively more diffused in urban areas.
In addition, Stark and Taylor (1991) hypothesized that the more relatively deprived—in
this case, the rural households relative to the urban households—were more inclined to migrate
to improve their living conditions. We can also expect the same to happen in the Philippines,
since rural households do appear to be more deprived with their lower average incomes and
limited access to education and media. The decomposition results as shown in Table 4, however,
do not necessarily suggest that rural workers have not migrated as predicted. It is highly possible
that rural households migrate to the cities first, where information about and placement for
overseas jobs are more accessible. With rural-to-urban migration, the otherwise rural incomes
will be considered urban incomes, and this is necessarily reflected in higher incomes—and
higher income shares—for urban areas. It should be stressed, though, that this is just a
hypothesis, and that it may need to be tested further.
On the other hand, the shares of remittances in total household incomes do not exhibit a
clear-cut trend, both for urban and rural Philippines. This, however, does not mean that the
irregular trend in the share of remittances has little or no impact on the contribution of
remittances to total income inequality, as will be seen later.
26
Gini Coefficients (G i)
The contribution of an income source to overall income inequality does not rely solely on
its share in total incomes. It also depends on the distribution of income from that particular
source.
Table 5 summarizes the inequality in total incomes, as well as the inequality within the
four income sources. It shows that total urban income is more unevenly distributed than total
rural incomes. In addition, the total- income Ginis indicate that, over time, income distribution
has improved in urban areas but it has worsened in rural areas. For the entire Philippines,
however, it seems that income distribution has generally improved.
Table 5 also shows that remittance income is the least evenly distributed income source;
that is, of the four income sources, it has the highest average Gini coefficient of about 0.94 for
the entire Philippines. The inequality even appears to be more acute in the countryside: The rural
Gini coefficient s have an average of about 0.96, while the urban Gini coefficients have an
average of about 0.92.
Table 5 further reveals that the changes in the coefficients are very minimal—in fact, the
changes do not exceed 0.01 between the three-year periods. Thus, it should be prudent to remark
that the inequality in remittance incomes has remained relatively unchanged. The remittance
Gini coefficient for urban areas, for example, remains almost constant at 0.92, although the rural
coefficients show slightly more evident deviations. Indeed, looking at the trend lines plotted in
Figure 2 below will support these observations.
27
Table 5. Gini Coefficients.
Income Source 1985 1988 1991 1994 1997 2000
All Philippines Wages 0.6986 0.6941 0.6809 0.6832 0.6666 0.6725 Entrepreneurial income 0.7066 0.7405 0.7376 0.7258 0.7357 0.7121 Other income 0.6319 0.6591 0.6386 0.6510 0.6142 0.5961 Remittances from migrants 0.9463 0.9426 0.9359 0.9262 0.9348 0.9311
Total Income 0.4970 0.5080 0.4998 0.5012 0.4841 0.4769
Urban Philippines Wages 0.6372 0.6304 0.6252 0.6369 0.6012 0.6134 Entrepreneurial income 0.7946 0.8253 0.8002 0.7867 0.7948 0.7673 Other income 0.6750 0.7187 0.6628 0.6719 0.6253 0.6065 Remittances from migrants 0.9173 0.9187 0.9187 0.9064 0.9197 0.9151
Total Income 0.5052 0.5153 0.4927 0.4963 0.4621 0.4536
Rural Philippines Wages 0.7060 0.7055 0.7151 0.7028 0.7203 0.7181 Entrepreneurial income 0.5891 0.6101 0.5886 0.5840 0.5963 0.5849 Other income 0.4695 0.4642 0.4942 0.2847 0.5115 0.4849 Remittances from migrants 0.9715 0.9615 0.9588 0.9529 0.9542 0.9515
Total Income 0.3878 0.3910 0.3999 0.3921 0.4187 0.4115
Figure 2. Gini Coefficient of Remittances.
0.00
0.50
1.00
1985 1988 1991 1994 1997 2000
Year
Gin
i Co
effic
ien
t
All Philippines
Urban
Rural
28
Table 5 and Figure 2 show that the distribution of remittance income is very skewed,
which means that very few Philippine households actually receive remittance income. In fact,
less than 10% of Philippine households receive some income from abroad. Moreover, the Gini
coefficients for rural Philippines are higher than for urban Philippines because no more than
about 5.5% of rural households receive remittances, whereas as much as 9.2% of urban
households receive income from abroad. One may also imply from the figures in Table 6 below
that the change in the Gini coefficient of remittance income is negligible since the changes in the
proportion of remittance-receiving households within populations are small.
Table 6. Percent of Households Receiving Income from Abroad.
1985 1988 1991 1994 1997 2000 All Philippines 5.8 5.8 6.5 6.8 6.2 7.2
Urban Philippines 9.2 8.4 8.3 8.8 7.8 9.1 Rural Philippines 3.7 4.2 4.7 4.8 4.8 5.5
Urban/Rural 2.5 2.0 1.8 1.8 1.6 1.7
Source: Authors’ estimates based on NSO-FIES, various years.
One may recall that the urban-rural ratios in Table 2 indicate that remittance recipients in
urban areas continue to outnumber their rural counterparts. The urban-rural ratios in Table 6
show a more complete picture, though. The ratios in Table 6 show an evident downward trend,
which implies that of the total number of workers deployed for overseas employment, the
proportion that comes from rural areas seems to have increased over the years. Migration thus
seems to have gradually become a valid labor option for the relatively poorer rural areas as well.
Before discussing its impact on income distribution, though, the paper considers the last
determinant of an income source’s contribution to income inequality, or the Gini correlation
ratios.
29
Gini Correlation Ratios (Ri)
According to Equation (3), an income source’s share in total incomes and its distribution
among the population reflect only a part of its contribution to total income inequality. The
contribution also depends on where the recipients of that income source are situated in the
income spectrum.
Table 7 reports the Gini correlation ratios between income sources and total income. The
significance of these Gini correlation ratios depends on the distribution of the income source,
which its Gini coefficient indicates. Given a very unequal distribution of an income source,
higher Gini correlation ratios imply that more households at the upper ends of the distribution
receive income from that income source. Lower correlation ratios, on the other hand, imply that
even households at the lower ranks of the income distribution receive income from that source.
Table 7. Gini Correlation Ratios.
Income Source 1985 1988 1991 1994 1997 2000
All Philippines Wages 0.7383 0.7736 0.7688 0.7916 0.7896 0.7890 Entrepreneurial income 0.5497 0.5631 0.5643 0.5171 0.5300 0.4850 Other income 0.7980 0.7677 0.7553 0.7842 0.7446 0.7234 Remittances from migrants 0.8046 0.7440 0.7519 0.7118 0.6834 0.7266
Urban Philippines
Wages 0.7206 0.7474 0.7427 0.7705 0.7529 0.7539 Entrepreneurial income 0.6022 0.6297 0.6033 0.5563 0.5682 0.5059 Other income 0.8159 0.7961 0.7605 0.7925 0.7363 0.7108 Remittances from migrants 0.7112 0.6587 0.6954 0.6469 0.6146 0.6675
Rural Philippines
Wages 0.6055 0.6764 0.6648 0.6990 0.7313 0.7363 Entrepreneurial income 0.5727 0.5362 0.5684 0.5137 0.5296 0.4985 Other income 0.6340 0.5577 0.5969 0.7998 0.6365 0.5995 Remittances from migrants 0.8590 0.7683 0.7827 0.7490 0.7402 0.7740
30
According to Table 7, all income categories are positively correlated with total income.
In the case of remittances, the correlation ratios indicate that inequality in remittance income is
positively and highly correlated with overall income inequality, and even more so for rural areas,
where remittances are more unevenly distributed, than in urban areas. This implies that in rural
areas, primarily households from the higher classes receive remittances from migrants. For urban
areas, though, remittances also accrue mostly to the upper classes (since the ratios all exceed
0.5), but to a lesser degree. It appears that more households at the lower ends of the distribution
receive remittance income in urban areas than in rural areas.
Although remittances accrue mostly to the upper classes for the Philippines as a whole,
the ratios shown in Table 7 illustrate a downward trend for the entire country. Figure 3 illustrates
this trend, which suggests that the proportion of lower-class households receiving income from
abroad has increased over time. This lends further support to the observation that an increasing
proportion of migrants has come from rural households, especially since rural households have
been shown to be less wealthy than urban households and at a disadvantage when it comes to
migration opportunities.
Migration theory actually predicts these results. The reason why remittances primarily
accrue to those in the upper ranks of the income spectrum is because migration entails huge
costs, both monetary and psychic. Hence, although the less-privileged are more inclined to
migrate, they are more constrained from doing so.
In addition, the lower urban ratios, when compared with the higher rural values, further
imply that as migration opportunities become more diffused, more people from the lower ranks
are able to migrate and benefit from remittance income. This should explain the observed
31
downward trend as well, because as more people from the lower ranks seize job opportunities
abroad, the Gini correlation ratio drops.
Figure 3. Gini Correlation Ratio of Remittances.
The Contribution of Remittances to Total Income Inequality
The context is now set for describing the contribution of remittances to total income
inequality. Two other variables can be derived from the three decomposition variables to
describe the contribution of each income source to overall income inequality. One variable is an
exact measure of the contribution of an income source to overall income inequality, expressed as
a percentage of the total- income Gini. The other is the relative concentration coefficient, which
tells exactly if an income source increases or decreases total income inequality.
0.00
0.10
0.200.30
0.400.50
0.60
0.700.80
0.901.00
1985 1988 1991 1994 1997 2000
Year
Gin
i Cor
rela
tion
Rat
io
All Philippines
Urban
Rural
32
Percentage Contributions to the Gini
Table 8 reports the percentage contributions of each income source to the total Gini,
which depend on the magnitudes of all the three decomposition variables. The table reports that
paid wages make the biggest contribution to total income inequality for the entire country, and
that their contribution even increases over time. This result is to be expected because of the trend
in the share of paid wages in total income: Wages are the most significant income source,
especially in urban areas, and their share in total income for the entire country also demonstrates
an upward trend.
On the other hand, although entrepreneurial incomes make up the bulk of rural incomes,
their contribution to rural income inequality still falls below that of wages, and it exhibits a
downward trend. This is so because even in rural areas, the share of entrepreneurial income
declines over time, while that of wages rise. Moreover, for rural areas, entrepreneurial income is
less correlated with total income than wages are, as shown by the higher Gini correlation
coefficients for wages. Rural entrepreneurial incomes are also more evenly distributed than rural
wages, as they have lower Gini coefficients.
In the case of remittances, these have made the smallest contribution to total income
inequality, even if they are the most unevenly distributed income source. Figure 4 helps describe
this contribution of international remittances to total income inequality. The shapes of the lines
closely resemble those in Figure 1, which show that the share of remittances in total incomes has
fluctuated over the time frame considered. This emphasizes the fact that an income source’s
share in total income determines its percentage contribution to total income inequality, and that
the two generally move together. Remittances have thus made the smallest contribution to
overall inequality because they have always had the smallest share in total income.
33
Table 8. Contributions of Income Sources to the Gini (in Percent)*.
Income Source 1985 1988 1991 1994 1997 2000
All Philippines Wages 39.69 45.82 45.04 48.75 50.52 51.02 Entrepreneurial income 22.16 23.17 23.68 18.95 21.09 17.70 Other income 25.50 21.31 19.48 22.38 19.42 19.07 Remittances from migrants 12.65 9.70 11.79 9.92 8.97 12.21
Total 100.00 100.00 100.00 100.00 100.00 100.00
Urban Philippines Wages 38.29 42.71 43.20 47.03 48.38 49.37 Entrepreneurial income 21.28 23.48 24.27 19.16 22.35 18.15 Other income 28.62 24.73 20.83 24.31 20.51 20.14 Remittances from migrants 11.82 9.08 11.71 9.50 8.76 12.34
Total 100.00 100.00 100.00 100.00 100.00 100.00
Rural Philippines Wages 33.56 44.27 39.90 46.44 46.96 48.51 Entrepreneurial income 35.00 32.35 33.99 28.66 27.56 24.51 Other income 17.55 12.93 14.46 11.40 15.92 14.67 Remittances from migrants 13.90 10.46 11.66 10.65 9.57 12.31
Total 100.00 100.00 100.00 100.00 100.00 100.00
* Contributions are computed as [(Si × Ri × Gi)/G] × 100, where G is the total-income Gini. NOTE: Figures may not add up to totals correctly due to rounding.
Figure 4. Contribution of Remittances to Total Income Inequality (in Percent).
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
16.00
1985 1988 1991 1994 1997 2000
Year
Con
trib
utio
n (%
)
All PhilippinesUrbanRural
34
There is one important difference between Figures 1 and 4, though. While the plot line
for urban Philippines appears above that for rural Philippines in Figure 1, one can observe the
opposite in Figure 4. This implies that the contribution of an income source to total income
inequality does not rely solely on that income source’s share in total income, as the
decomposition equations above actually assert.
As Figure 4 shows, migrants’ remittances contribute more to rural inequality than they do
to urban inequality, even if remittances make up a smaller part of rural incomes, because
remittances are more unevenly distributed in rural areas than they are in urban areas. This is
reflected in a more evident, more-than-proportional increase in inequality in rural areas:
Remittances from abroad represent, on the average, only about 6% of rural incomes, but they
contribute an average of about 11.4% to rural income inequality. On the other hand, remittances
from abroad represent about 8% of urban incomes, but they contribute only about 10.5% to
urban income inequality.
Still, these percentage contributions do not seem to fully answer whether remittances—or
any income source for that matter—serve to increase or decrease income inequality. Since the
contributions of all income sources are positive, it may appear at first glance that all income
sources increase inequality. While these contributions may become negative because the Gini
correlation ratio can acquire a negative value, this possibility seems unlikely given that the Gini
correlation ratios of all four income sources have been consistently positive.
In addition, it is difficult to establish a discernible trend based on these percentage
contributions alone. Figures 1 and 4 seem to show tha t the contribution of remittances over time
depends solely on their share in total incomes; the contribution of remittances to the total- income
Gini fluctuates just because their share in total incomes also fluctuates. The plot lines in Figure 4
35
also cannot account for the declining trend in the Gini correlation ratios (and the theoretical
implications thereof), which, as Equation 3 states, should also explain the contribution of
remittances to the overall inequality.
Relative Concentration Coefficients (gi)
To describe more accurately the contribution of remittances to total income inequality
over time, one must therefore use another measure or variable. Derived from Equation (6) above,
the relative concentration coefficients of each income source are presented in Table 9. An
income source either increases or decreases income inequality, depending on whether the
coefficient is greater or less than unity.
Interestingly, paid wages appear to increase inequality, especially for rural areas. The
coefficients also show an increasing trend, with values again peaking in 2000 for the entire
country. This trend becomes even more apparent for urban Philippines, where paid wages
account for the bulk of household incomes: Wages become inequality- increasing only in 2000
after being inequality-decreasing since 1985.
In the case of entrepreneurial income, they have always been inequality-decreasing,
especially for rural Philippines where it accounts for the biggest share in total income. The
coefficients even exhibit a decreasing trend, which implies that entrepreneurial income actually
contributes to greater income equality over time.
36
Table 9. Relative Concentration Coefficients*.
Income Source 1985 1988 1991 1994 1997 2000
All Philippines Wages 1.0377 1.0569 1.0474 1.0790 1.0873 1.1125 Entrepreneurial income 0.7814 0.8208 0.8328 0.7488 0.8054 0.7242 Other income 1.0146 0.9960 0.9652 1.0186 0.9447 0.9041 Remittances from migrants 1.5319 1.3805 1.4079 1.3152 1.3198 1.4185
Urban Philippines
Wages 0.9089 0.9144 0.9423 0.9887 0.9796 1.0193 Entrepreneurial income 0.9471 1.0084 0.9798 0.8817 0.9774 0.8558 Other income 1.0900 1.1102 1.0230 1.0728 0.9964 0.9503 Remittances from migrants 1.2913 1.1744 1.2967 1.1813 1.2235 1.3466
Rural Philippines
Wages 1.1023 1.2205 1.1887 1.2528 1.2581 1.2847 Entrepreneurial income 0.8698 0.8366 0.8366 0.7652 0.7542 0.7084 Other income 0.7675 0.6620 0.7377 0.5807 0.7777 0.7065 Remittances from migrants 2.1520 1.8892 1.8767 1.8204 1.6868 1.7897
* Relative Concentration Coefficient = gi = Ri×(Gi/G), where G is the total-income Gini.
With values consistently greater than 1, remittances from migrants are undoubtedly
inequality- increasing. Furthermore, the fact that the rural coefficients are larger than the urban
values implies that remittances worsen rural more than they do urban income inequality, an
observation which is consistent with the calculated percentage contributions to the Gini.
Unlike the computed values for the percentage cont ribution of remittances, however, the
relative concentration coefficients of remittances reveal a more perceptible trend when graphed,
as seen in Figure 5. The coefficients for migrants’ remittances exhibit an obvious declining trend,
especially for rural areas and the Philippines as a whole; the urban trend is a bit more erratic,
although later years generally yield values lower than the 1985 level. This downward tendency
implies that, over time, remittances have gradually become less inequality- increasing, as
predicted by theory.
37
Figure 5. Relative Concentration Coefficient of Remittances.
Therefore, if the relative concentration coefficients reported in Table 9 are as predicted,
then it must be the case that the low-income groups have had greater access to migration-
facilitating factors such that an increasing proportion of migrants have now come from their
ranks. For instance, more placement centers may have been established in rural areas, although
statistics to support this claim are not available. In any case, the conclusion still seems to be
supported by the declining urban-rural ratio in Table 6 above, as well as by the downward trend
in the Gini correlation ratio of remittances in Figure 3—all of which show that low-income
groups have also benefited from remittances. This result would actually agree with those of other
studies on Asian countries, especially the one on Sri Lanka (Rodrigo and Jayatissa 1989), which
0.00
0.50
1.00
1.50
2.00
2.50
1985 1988 1991 1994 1997 2000
Year
Rel
ativ
e C
on
cen
trat
ion
Co
effi
cien
t
All PhilippinesUrbanRural
38
suggested that remittances would tend to improve the income distribution if migrants also came
from low-income households.
Moreover, the conclusion that remittances have gradually become less inequality-
increasing over time must now be seen in light of the fact that migration is a demand-driven
phenomenon and of the view that its impact on income distribution is stage-specific. The time
period covered in the paper was actually a stage in the migration history of the Philippines when
the prevailing demand for overseas workers had just shifted from the professional fields in the
sixties and seventies to the relatively low-skilled service sector in the eighties (see Appendix C).
In fact, the 1996 Survey of Overseas Filipino Workers by the NSO claimed that the service
sector “[had] been the sector with the most demand as early as 1990”—a trend that has continued
well into the nineties, as Table 10 indicates.
Table 10. Distribution of Overseas Filipino Workers, by Major Occupation, 1995 – 2000 (in Percent).
Major Occupation Group 1995 1996 1997 1998 1999 2000 Professional, Technical and Related Work 10.8 12.0 11.0 11.8 14.7 14.0 Administrative, Executive and Managerial 0.3 0.4 0.4 0.6 1.0 0.7 Clerical and Related Workers 2.9 3.4 3.0 4.3 3.3 4.3 Sales Workers 1.7 1.0 1.6 1.5 1.8 1.7 Service Workers 41.4 40.1 40.2 41.9 39.2 38.0 Agricultural, Animal Husbandry and Forestry Workers, Fishermen and Hunters 1.3 1.1 0.6 0.3 0.4 0.8 Production and Related Workers, Transport Equipment Operators and Laborers 41.0 41.1 43.2 39.3 39.6 40.1 Occupation Not Adequately Described 0.8 0.8 0.2 0.2 0.1 0.3
Total 100 100 100 100 100 100 Source: Authors’ estimates based on NSO-Survey of Overseas Filipino Workers, various years. NOTE: Figures may not add up to totals correctly due to rounding. Figures for years prior to1995 were unavailable.
Hence, it also seems plausible that remittances have become less inequality- increasing
over the period considered because the Philippines has faced a demand for migrant labor that
favors the relatively low-skilled occupations and the low-income classes. More importantly, if
39
the Philippines were to face the same demand in the future, then OFW remittances may be
expected to become even less inequality- increasing (or more inequality-decreasing).
The relative concentration coefficients seem to show further that of the three
decomposition variables, the positions of remittance-receiving households in the income
distribution—as manifested in the Gini correlation ratios—greatly determine whether
remittances will alleviate or worsen income inequality. This is so because over time, fluctuations
in the share of remittances in total income do not explicitly mean that remittances have actually
improved or exacerbated income inequality. The distribution of remittance incomes has also
hardly changed, even though migration-facilitating factors have presumably become diffused
through the population. Finally, the changes in the demand for certain types of overseas workers
are more likely to affect the income distribution by determining from which income classes
migrants will most likely come, which ultimately affects the Gini correlation ratios accordingly.
40
VI. Conclusions and Recommendations
In summary, decomposing total income inequality into its sources revealed that
remittances from international migrants contributed to income inequality. Remittances appeared
to increase inequality—and more so for rural than urban areas. However, remittances had
become less inequality- increasing over time, as predicted by the theory of stage-specific
migration and income distribution.
The paper likewise showed how the magnitude of remittances’ contribution to total
inequality in the Philippines was determined by the fluctuating share of remittances in total
income, the highly unequal distribution of remittance incomes, as well as the rank or position of
remittance-receiving households in the income distribution. Of these three determinants,
however, the income class from which migrants came appeared to be the most significant
element in explaining whether remittances increased or decreased total income inequality over
time.
Policy Recommendations
Therefore, if the goal of policy were to reduce income inequality, then policymakers
should best focus on how the income classes from which migrants came would contribute to
overall inequality. More explicitly, policies should aim to allow those in the low-income brackets
greater degrees of access to migration-facilitating factors, for remittances would only reduce
inequality if the less-privileged classes were also able to migrate. An example of such a policy
would include the establishment of more placement centers, especially in rural areas. This would
lower the transaction and placement costs faced by would-be rural migrants, since it would
reduce the need to travel to urban centers. However, doing so might seem inadequate in the long
41
run, especially since migration seemed to have become a viable option for the lower classes and
that the proportion of migrants coming from lower income brackets could even be expected to
continue increasing. In this regard, a complementary policy would be to upgrade existing
training centers and programs for overseas workers.
On the other hand, it would seem imprudent to continue encouraging overseas
employment without giving due consideration to the long-term effects of labor outflows. The
country might continue sending its workers abroad, since their remittances were found to be
becoming less inequality- increasing over time. However, if migrants were to continue coming
from urban areas and from higher- income households, migration might eventually exacerbate
income inequality. In addition, if only the more-skilled workers were able to migrate, an active
deployment of workers for overseas employment might result in brain drain, which had been
“often viewed as increasing inequality in the source [countries]” (Davies and Wooton 1992).
Most importantly, the migration of workers would seem very vulnerable to unanticipated
changes in the skills and occupations demanded abroad. In this regard, policymakers might want
to focus on fostering conditions that would either create domestic employment or make working
domestically more attractive relative to working abroad, with a long-term goal of making the
productivity of labor rise. This approach could take many forms, but two could be cited as
pertinent.
One way to increase the productivity of labor would be through the inflow of capital or
investments, whether foreign or domestic. The inflow of capital would, at least in theory,
increase the productivity of labor because it would increase its demand, and hence domestic
wages. This should address the issues of how to make skilled labor stay and how to absorb the
excess supply of labor if ever the demand for migrant labor slackened. Considering that wages
42
appeared to have a growing importance to households and an increasing contribution to
inequality over the years, though, discussions should be careful with how policies that aim to
stimulate investments would affect the wage structure. In the final analysis, too, domestic
investments should be encouraged primarily, for foreign direct investments would most probably
account for only a small part of total investments. In any case, policies should be geared towards
establishing and sustaining favorable macroeconomic conditions.
The other way to increase the productivity of labor would be by having an efficient
educational system—with an aim to reduce the spread of educational differences and skill levels
between workers and between income classes—so that those in the lower classes would be better
equipped if foreign demand shifted away from low-skilled labor. In this regard, the government
might want to look into its public education system, where many of the low-income classes
acquire the minimum skill requirements needed in securing a job, whether here in the Philippines
or abroad. All in all, by reducing the spread of educational differences, an efficient educational
system might also reduce the spread of migration opportunities between the upper and lower
income classes.
Recommendations for Further Studies
The paper has opened up many areas for future research. The paper has limited itself to a
discussion of income inequality in the context of international migration, but the analysis can be
taken a few steps further. For one, regression analysis can determine the effects of individual
households’ characteristics on income and its distribution. It may even be used to test the
significance of the results of the decomposition.
43
There also has been some mention of the possibility of rural-to-urban migration before
actual international migration takes place. The paper has not dwelled on the issue, although it
remains an important avenue towards understanding the impact of migration and remittances on
Philippine income inequality. In this regard, future studies may opt to decompose income
inequality on a smaller level—perhaps regional—instead of on a national level.
Finally, the paper has not given attention to the further implications of the changes in
inequality. Changes in the overall inequality may have ambiguous implications especially with
regard to welfare. For instance, a small increase in the income of household A, leaving all other
households’ incomes unchanged (ceteris paribus), may result in a worsening of income
inequality depending on household A’s relative position in the distribution. This may be
interpreted as a welfare gain, at least for household A, but not necessarily so for the entire
society. It is thus reasonable to investigate the effect of a small change in income from
remittances—or from any other source, as the decomposition equations allow it—on social
welfare. Doing so, however, will require some knowledge of social welfare functions, which is
beyond the scope of this paper.
•
44
Selected References
Adams, Richard H. Jr. “The Effects of International Remittances on Poverty, Inequality, and Development in Rural Egypt.” Research Report 86. International Food Policy Research Institute, 1991.
Adams, Richard H. Jr., and Harold Alderman. “Sources of Income Inequality in Rural Pakistan:
A Decomposition Analysis.” Policy Research Working Paper 836. World Bank. Agriculture and Rural Development Department, 1992.
Adams, Richard H. Jr., and Jane J. He. “Sources of Income Inequality and Poverty in Rural
Pakistan.” Research Report 102. International Food Policy Research Institute, 1995. Amjad, Rashid. Ed. To the Gulf and Back: Studies on the Economic Impact of Asian Labour
Migration. Geneva: ILO-ARTEP, 1989. Balisacan, Arsenio M., and Sharon Faye A. Piza. “Nature and Causes of Income Inequality in the
Philippines.” Paper presented at the “Conference on Comparative Analyses of East Asian Income Inequalities,” Bangkok, 27 – 28 January, 2003.
Davies, James B., and Ian Wooton. “Income Inequality and International Migration.” Economic
Journal. 102 (1992): 789 – 802. Department of Labor and Employment. Yearbook of Labor Statistics. 2000. Ehrenberg, Ronald G., and Robert S. Smith. Modern Labor Economics: Theory and
Public Policy. 7th ed. Boston: Addison-Wesley, 2000. Hyun, Oh-Seok. “The Impact of Overseas Migration on National Development: The Case of the
Republic of Korea.” To the Gulf and Back: Studies on the Economic Impact of Asian Labour Migration. Ed. Rashid Amjad. Geneva: ILO-ARTEP, 1989. 143 – 166.
Kuznets, Simon. “Economic Growth and Income Inequality.” American Economic Review. 45
(1955): 1 – 28. Layard, Richard, D. Piachaud, and M. Stewart. “The Causes of Poverty.” Tackling Inequality.
Comp. Richard Layard. London: Macmillan, 1999. 91 – 104. Macaraig, Mel A., and Michael B. Sarino. “An Analysis of the Remittances of Contract Workers
and Its Effects on Income Distribution and Welfare.” Thesis. UP School of Economics, 1980.
Mahmud, Wahiduddin. “The Impact of Overseas Labour Migration on the Bangladesh Economy:
A Macro-Economic Perspective.” To the Gulf and Back: Studies on the Economic Impact of Asian Labour Migration. Ed. Rashid Amjad. Geneva: ILO-ARTEP, 1989. 55 – 94.
45
Mangahas, Mahar. “Income Inequality in the Philippines: A Decomposition Analysis.”
Discussion Paper 74-15. Institute of Economic Development and Research. UP School of Economics, 1974.
National Statistical Coordination Board. Philippine Statistical Yearbook. 2002. National Statistics Office. Family Income and Expenditures Survey. Various years. ---. Survey on Overseas Filipino Workers. Various years. Available online. National Statistics
Office Website. http://www.census.gov.ph. Nygard, Fredrik, and Arne Sandstrom. Measuring Income Inequality. Hango, Finland: Hango
Trycken Ab, 1981. Ray, Debraj. Development Economics. Princeton University Press: New Jersey, 1998. Rodrigo, Chandra, and R.A. Jayatissa. “Maximising Benefits from Labour Migration: Sri
Lanka.” To the Gulf and Back: Studies on the Economic Impact of Asian Labour Migration. Ed. Rashid Amjad. Geneva: ILO-ARTEP, 1989. 255 – 302.
Rodriguez, Edgard R. “International Migration and Income Distribution in the Philippines.”
Economic Development and Cultural Change. 46 (1998): 329 – 350. Rodriguez, Edgard R., and Susan Horton. “International Return Migration and Remittances in
the Philippines.” Working Paper. University of Toronto, 1995. Sicat, Gerardo P. “Philippine Income Distribution: Facts for Economic and Social Policy.”
Discussion Paper 69-13. Institute of Economic Development and Research. UP School of Economics, 1969.
---. “Myths and Fallacies in Economic Policy Debates.” Discussion Paper 0001. UP School of
Economics, 2000. Stark, Oded, and David E. Bloom. “The New Economics of Labor Migration.” American
Economic Review. 175 (1985): 173 – 178. Stark, Oded, and J. Edward Taylor. “Relative Deprivation and Migration: Theory, Evidence, and
Policy Implications.” Policy, Research and External Affairs Working Paper 656. World Bank. Population and Human Resources Department, and Agriculture and Rural Development Department, 1991.
Stark, Oded, J. Edward Taylor, and Shlomo Yitzhaki. “Remittances and Inequality.” Economic
Journal. 96 (1986): 722 – 740.
46
Sto. Tomas, Patricia A. “Overseas Employment in the Philippines: Policies and Program.” Migration from the Philippines. Comp. Anthony Paganoni. Quezon City: Scalabrinians, 1984. 101 – 121.
Tan, Edita A., and Dante B. Canlas. “Migrants’ Savings, Remittances and Labour Supply
Behaviour: The Philippines Case.” To the Gulf and Back: Studies on the Economic Impact of Asian Labour Migration. Ed. Rashid Amjad. Geneva: ILO-ARTEP, 1989. 223 – 254.
Tan, Edita A. “Overseas Employment, Savings Rate and Income Distribution: The Philippine
Case.” Proceedings of Second Japan ASIAN Forum on International Labor Migration in East Asia, The United Nations University, Tokyo, September 26 – 27, 1991.
---. “Filipino Overseas Employment—An Update.” Discussion Paper 0003. UP School of
Economics, 2000. Tingsabadh, Charit. “Maximising Development Benefits from Labour Migration: Thailand.” To
the Gulf and Back: Studies on the Economic Impact of Asian Labour Migration. Ed. Rashid Amjad. Geneva: ILO-ARTEP, 1989. 303 – 342.
Untalan, Teodoro S. “The Middle East Demand for Philippine Labor: A Cost-Benefit
Evaluation.” Thesis. UP School of Economics, 1980. World Bank. World Development Report 1995: Workers in an Integrating World. New York:
Oxford University Press, 1995. ---. “Inequality Measurement.” World Bank Official Website. Accessed on 12 March 2003.
http://www.worldbank.org/poverty/inequal/methods/measure.htm. © 2000.
•
47
Appendix A Choosing a Measure of Inequality: The Five Properties of Decomposability
The decompositions of the measures of inequality are generally divided into two types:
(1) a functional decomposition by income-determining characteristics (subgroups), and (2) a
functional decomposition by income sources (Nygard and Sandstrom 1981).
Decomposition by income-determining characteristics determines how much of total
inequality is attributable to variability in each income-determinant or subgroup, as well as to in-
between-groups inequality. On the other hand, decomposition by income sources determines
how much of total inequality is attributable to income from specific sources such as wages,
entrepreneurial income, and remittances from abroad.
While there are many ways of measuring inequality, some measures are sensitive to
changes in the distribution and in the values of the data being computed. For example, the
variance, which is a basic measure of inequality, is not independent of the income scale:
Doubling all incomes would result in a quadrupling of the estimate of income inequality. Thus, it
would be prudent to choose a measure of inequality that would conform to a set of axioms,
which would in turn characterize a fair measure of inequality.
Axiomatic Approach in Choosing a Measure of Inequality
The Axiomatic Approach outlines five key axioms in choosing the appropriate measure
of inequality for decomposition: (1) The Pigou-Dalton Transfer Principle, (2) Income Scale
Independence, (3) the Principle of Population, (4) Anonymity, and (5) Decomposability (World
Bank 2000).
48
The Pigou-Dalton Transfer Principle. The inequality measure should rise (or at least not
fall) in response to a mean-preserving spread. An income transfer from a poorer person to a
richer person should register as a rise (or at least not as a fall) in inequality and an income
transfer from a richer person to a poorer person should register as a fall (or at least not as an
increase) in inequality.
Income Scale Independence. The inequality measure is scale independent if it is invariant
to uniform proportional changes. If each individual’s income changes by the same proportion
then inequality should not change.
Principle of Population. The inequality measure should also be invariant to replications
of the population. Merging two identical distributions should not alter inequality.
Anonymity. This axiom requires that the inequality measure be independent of any
characteristic of individuals other than their income (or the welfare indicator whose distribution
is being measured).
Source Decomposability. This requires overall inequality to be related consistently to
constituent parts of the distribution, such as specific income sources. For example, if inequality
is seen to rise amongst each income source, then overall income inequality is expected to rise as
well.
Inequality measures that satisfy all of these axioms provide a good foundation for
decomposition analysis. The Generalized Entropy (GE) class of measures, the Atkinson class of
inequality measures, and the Gini coefficient are the more popular inequality measures that
satisfy these axioms.
•
49
Appendix B Comparing the Decomposition Results of Using the Gini Coefficient and the Generalized Entropy (GE) Class of Inequality Measures
Varying measures of inequality will almost always rank distributions in different ways,
especially when using measures that are sensitive to changes that affect distributions. The Gini
coefficient, for example, appears to be a perfectly satisfactory index because it satisfies the five
axioms (or their equivalent, the Lorenz consistency) discussed in Appendix A (Ray 1998).
However the problem arises when two Lorenz curves that are being compared cross each other
(i.e., when sub-vectors of income overlap), in which case the Gini may fail to meet the
decomposability axiom and may therefore yield results that are not intuitively or mathematically
appealing (World Bank 2000).
Even the parameter used in the Generalized Entropy (GE), which represents the weight
given to distances between incomes at different parts of the income distribution, may give biased
results (World Bank 2000). For lower values of the parameter, the GE is more sensitive to
changes in the lower tail of the distribution, and for higher values the GE is more sensitive to
changes that affect the upper tail.
Clearly, there is no intuitive reason to prefer one measure to another. The logical move
then is not to rely entirely on one particular measure of inequality, but to compare the results of
using different inequality measures.
Figure A below plots the percentage contribution of migrants’ remittances to total income
inequality, computed using the Gini coefficient and the Generalized Entropy class of inequality
measures over the period of interest. The Gini coefficient yields values that are generally higher
than those of the GE class values. More interestingly, while the two measures indicate
contrasting trends, specifically from 1988 to 1994, a comparison of the results of using two
50
different inequality measures captures the fact that remittances’ percentage contribution to total
inequality has generally been erratic. Furthermore, even without reference to the theoretical
explanation as to where the inconsistency may have come from, this finding shows that
comparisons and conclusions may be sensitive to the choice of the index.
Figure A. Contribution of Remittances from Migrants to Total Income Inequality (in Percent).
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
Year
Per
cen
t
GE* 8.28 7.82 5.15 6.57 5.06 8.39
Gini** 12.65 9.70 11.79 9.92 8.97 12.21
1985 1988 1991 1994 1997 2000
*Source: Balisacan and Piza. 2003. **Source: Authors’ computations.
In this sense, it is important to realize that coming up with stricter criteria for choosing an
index of inequality measurement will be inevitably subjective and controversial. Relative
inequality may be perfectly discernible some of the time and difficult to judge at other times
(Ray 1998).
•
51
Appendix C Background Information on Philippine Migration
Historical Background of International Migration
First Wave
The first wave of migration occurred as early as the American colonization period. In the
1900s, the Philippines became a source of cheap plantation manpower when Hawaii experienced
a severe shortage of labor. The Hawaiian Sugar Planters Association increased its incentives for
recruiters to cope with the labor shortage, and the recruitment for overseas employment thus
became a lucrative business. However, there were cases of illegal recruitment for non-existent
jobs, or of low wages and inadequate living conditions.
Responding to the need to protect the interests of the Filipino worker, the Philippine
legislature passed Act 2486 in February 1915, the first manifestation of government regulation of
overseas employment and which remained the reference legislation for almost sixty years. The
Act provided, among others, that (a) no contracting of minors under fifteen and minors of
eighteen years without the written consent of parents or guardians was allowed; (b) return
transportation for workers who had finished their contracts or had become unfit for work was
guaranteed; and (c) recruitment of members of non-Christian tribes for exhibition purposes was
strictly prohibited (Sto. Tomas 1984).
Then, after World War II, Filipino contractors, engineers and technicians were recruited
to rehabilitate and reconstruct American strongholds in the Pacific region, particularly in Guam
and Vietnam.
52
Second Wave
The second wave of migration occurred in the 1960s as a response to the liberalization of
US immigration policy, which replaced racial quotas with occupation-based quotas. The
liberalization also aimed to weaken the monopoly of the American Medical Association on the
supply of medical workers. Filipino doctors and nurses responded favorably to this opportunity
to work abroad; in fact, according to Tan (2000), nearly half of the country's graduating doctors
went to the US in the said decade. Nevertheless, up until 1970, the volume of overseas
employment was assumed to be unsubstantial. Thus, Act 2486 was seen to suffice as the
government's main policy regarding overseas contract employment (Sto. Tomas 1984).
Third Wave
Then the Middle East phenomenon began in the 1970s, which started a new wave of
migration. The oil shock provided oil-producing countries with resources for construction and
development projects. However, these countries required more skilled manpower than their
natives could supply; hence, they started hiring overseas workers, such as seamen, construction
workers and engineers, from neighboring Asian countries like the Philippines.
At this time, however, the Philippines experienced worsening deficits in its current
account. This made the foreign exchange remittances of overseas Filipino workers valuable.
Thus, Presidential Decree 442-also known as the Labor Code of the Philippines-was signed into
law on 01 May 1974, ushering in a new period in government supervision of overseas
employment. Under this law, the government set up the Overseas Employment Development
Board, which aimed to lower the costs of emigrating through a reduction in the travel tax and the
establishment of one-stop shops that simplified the processing of travel and employment papers.
53
All these were aimed to encourage migration and hence the inflow of much-needed foreign
exchange.
Presently, the Philippine Overseas Employment Administration (POEA) and the
Overseas Worker's Welfare Administration (OWWA) have been taking care of the interests of
the OFWs. The POEA, for example, has worked to reduce the number of illegal recruitment
cases (Tan 2000). The OWWA has also established a fund to lend money to overseas workers
and their families, to provide pre-departure training grants for the migrant, and to extend
scholarships for the migrants' children. However, no information about the actual impacts of
these OWWA programs is available, whether in terms of the amount of funds spent or the
number of beneficiaries served (Rodriguez and Horton 1995).
Government Policy on Overseas Contract Workers’ Remittances
Government policy has underscored the importance of international remittances to the
Philippine economy.
This section draws heavily from Rodriguez and Horton (1995), because no other source
that described Philippine policy on remittances was available. In 1982, an interagency body
formed by the Central Bank of the Philippines, the Ministry of Foreign Affairs and Ministry of
Labor and Employment, passed Executive Order (EO) 857, which required the mandatory
remittance of 50% to 70% of overseas contract workers’ salaries. Sanctions imposed on
delinquent contract workers included the non-renewal of the worker’s passport or contract, and
the disapproval of new contracts for overseas employment.
After the government received numerous complaints from the contract workers, EO 1021
repealed the provisions of EO 857, although the target remittance figure remained in effect. The
54
government’s policy on compulsory remittances actually remains valid up to now, but it isn’t
widely observed because it is difficult to implement.
Perhaps to increase the overseas workers’ willingness to remit, the government started
giving incentives to migrants instead of imposing sanctions in the early nineties. There were
numerous programs, such as the Overseas Workers Investment program in 1991, but no data
about their performance or success were made available. (No data about the goals and objectives
of the said program were available.)
There were other programs that provided alternative channels that facilitated remittances.
For example, the OWWA provides remittance assistance to overseas contract workers: It takes
deposits abroad, then it delivers the money within 24 hours in Metro Manila (it takes two to three
more days for provinces). Many other Philippine banks work similarly.
Rodriguez and Horton conclude that the change from an imposing policy to a more
supportive one reveals the difficulty of increasing remittances through legislation, especially if
the determinants of overseas workers’ willingness or propensity to remit is not yet well
understood.
•
55
Appendix D Tables A1 – A6: Composition of Household Income Inequality, by Year
Table A1. Composition of Household Income Inequality, 1985.
Income Source
Share in Total Income (Si)
Gini Correlation (Ri)
Gini Coefficient (Gi)
Contribution Gini ̂ Contribution to Gini (%)
Relative Concentration Coefficients*
All Philippines
Wages 0.3825 0.7383 0.6986 0.1973 39.69 1.0377 Entrepreneurial income 0.2836 0.5497 0.7066 0.1102 22.16 0.7814 Other income 0.2513 0.7980 0.6319 0.1267 25.50 1.0146 Remittances from migrants 0.0826 0.8046 0.9463 0.0629 12.65 1.5319
Total 1.0000 1.0000 0.4970 0.4970 - - - - - - Urban Philippines
Wages 0.4213 0.7206 0.6372 0.1934 38.29 0.9089 Entrepreneurial income 0.2247 0.6022 0.7946 0.1075 21.28 0.9471 Other income 0.2625 0.8159 0.6750 0.1446 28.62 1.0900 Remittances from migrants 0.0915 0.7112 0.9173 0.0597 11.82 1.2913
Total 1.0000 1.0000 0.5052 0.5052 - - - - - - Rural Philippines
Wages 0.3044 0.6055 0.7060 0.1301 33.56 1.1023 Entrepreneurial income 0.4023 0.5727 0.5891 0.1357 35.00 0.8698 Other income 0.2286 0.6340 0.4695 0.0681 17.55 0.7675 Remittances from migrants 0.0646 0.8590 0.9715 0.0539 13.90 2.1520
Total 1.0000 1.0000 0.3878 0.3878 - - - - - - NOTE: Figures may not add up correctly due to rounding.
^ computed as (Si) * (Ri) * (Gi) * computed as Ri*(Gi/G), where G = total-income Gini
56
Table A2. Composition of Household Income Inequality, 1988.
Income Source
Share in Total Income (Si)
Gini Correlation (Ri)
Gini Coefficient (Gi)
Contribution Gini ̂ Contribution to Gini (%)
Relative Concentration Coefficients*
All Philippines
Wages 0.4336 0.7736 0.6941 0.2328 45.82 1.0569 Entrepreneurial income 0.2822 0.5631 0.7405 0.1177 23.17 0.8208 Other income 0.2139 0.7677 0.6591 0.1083 21.31 0.9960 Remittances from migrants 0.0703 0.7440 0.9426 0.0493 9.70 1.3805
Total 1.0000 1.0000 0.5080 0.5080 - - - - - - Urban Philippines
Wages 0.4671 0.7474 0.6304 0.2201 42.71 0.9144 Entrepreneurial income 0.2328 0.6297 0.8253 0.1210 23.48 1.0084 Other income 0.2228 0.7961 0.7187 0.1274 24.73 1.1102 Remittances from migrants 0.0773 0.6587 0.9187 0.0468 9.08 1.1744
Total 1.0000 1.0000 0.5153 0.5153 - - - - - - Rural Philippines
Wages 0.3627 0.6764 0.7055 0.1731 44.27 1.2205 Entrepreneurial income 0.3867 0.5362 0.6101 0.1265 32.35 0.8366 Other income 0.1953 0.5577 0.4642 0.0506 12.93 0.6620 Remittances from migrants 0.0553 0.7683 0.9615 0.0409 10.46 1.8892
Total 1.0000 1.0000 0.3910 0.3910 - - - - - - NOTE: Figures may not add up correctly due to rounding.
^ computed as (Si) * (Ri) * (Gi) * computed as Ri*(Gi/G), where G = total-income Gini
57
Table A3. Composition of Household Income Inequality, 1991.
Income Source
Share in Total Income (Si)
Gini Correlation (Ri)
Gini Coefficient (Gi)
Contribution Gini ̂ Contribution to Gini (%)
Relative Concentration Coefficients*
All Philippines
Wages 0.4300 0.7688 0.6809 0.2251 45.04 1.0474 Entrepreneurial income 0.2844 0.5643 0.7376 0.1184 23.68 0.8328 Other income 0.2018 0.7553 0.6386 0.0974 19.48 0.9652 Remittances from migrants 0.0838 0.7519 0.9359 0.0589 11.79 1.4079
Total 1.0000 1.0000 0.4998 0.4998 - - - - - - Urban Philippines
Wages 0.4585 0.7427 0.6252 0.2129 43.20 0.9423 Entrepreneurial income 0.2476 0.6033 0.8002 0.1196 24.27 0.9798 Other income 0.2036 0.7605 0.6628 0.1026 20.83 1.0230 Remittances from migrants 0.0903 0.6954 0.9187 0.0577 11.71 1.2967
Total 1.0000 1.0000 0.4927 0.4927 - - - - - - Rural Philippines
Wages 0.3356 0.6648 0.7151 0.1595 39.90 1.1887 Entrepreneurial income 0.4063 0.5684 0.5886 0.1359 33.99 0.8366 Other income 0.1960 0.5969 0.4942 0.0578 14.46 0.7377 Remittances from migrants 0.0621 0.7827 0.9588 0.0466 11.66 1.8767
Total 1.0000 1.0000 0.3999 0.3999 - - - - - - NOTE: Figures may not add up correctly due to rounding.
^ computed as (Si) * (Ri) * (Gi) * computed as Ri*(Gi/G), where G = total-income Gini
58
Table A4. Composition of Household Income Inequality, 1994.
Income Source
Share in Total Income (Si)
Gini Correlation (Ri)
Gini Coefficient (Gi)
Contribution Gini ̂ Contribution to Gini (%)
Relative Concentration Coefficients*
All Philippines
Wages 0.4518 0.7916 0.6832 0.2444 48.75 1.0790 Entrepreneurial income 0.2530 0.5171 0.7258 0.0950 18.95 0.7488 Other income 0.2197 0.7842 0.6510 0.1122 22.38 1.0186 Remittances from migrants 0.0755 0.7118 0.9262 0.0497 9.92 1.3152
Total 1.0000 1.0000 0.5012 0.5012 - - - - - - Urban Philippines
Wages 0.4756 0.7705 0.6369 0.2334 47.03 0.9887 Entrepreneurial income 0.2173 0.5563 0.7867 0.0951 19.16 0.8817 Other income 0.2266 0.7925 0.6719 0.1207 24.31 1.0728 Remittances from migrants 0.0804 0.6469 0.9064 0.0472 9.50 1.1813
Total 1.0000 1.0000 0.4963 0.4963 - - - - - - Rural Philippines
Wages 0.3707 0.6990 0.7028 0.1821 46.44 1.2528 Entrepreneurial income 0.3746 0.5137 0.5840 0.1124 28.66 0.7652 Other income 0.1963 0.7998 0.2847 0.0447 11.40 0.5807 Remittances from migrants 0.0585 0.7490 0.9529 0.0418 10.65 1.8204
Total 1.0000 1.0000 0.3921 0.3921 - - - - - - NOTE: Figures may not add up correctly due to rounding.
^ computed as (Si) * (Ri) * (Gi) * computed as Ri*(Gi/G), where G = total-income Gini
59
Table A5. Composition of Household Income Inequality, 1997.
Income Source
Share in Total Income (Si)
Gini Correlation (Ri)
Gini Coefficient (Gi)
Contribution Gini ̂ Contribution to Gini (%)
Relative Concentration Coefficients*
All Philippines
Wages 0.4647 0.7896 0.6666 0.2446 50.52 1.0873 Entrepreneurial income 0.2618 0.5300 0.7357 0.1021 21.09 0.8054 Other income 0.2056 0.7446 0.6142 0.0940 19.42 0.9447 Remittances from migrants 0.0680 0.6834 0.9348 0.0434 8.97 1.3198
Total 1.0000 1.0000 0.4841 0.4841 - - - - - - Urban Philippines
Wages 0.4939 0.7529 0.6012 0.2235 48.38 0.9796 Entrepreneurial income 0.2287 0.5682 0.7948 0.1033 22.35 0.9774 Other income 0.2058 0.7363 0.6253 0.0948 20.51 0.9964 Remittances from migrants 0.0716 0.6146 0.9197 0.0405 8.76 1.2235
Total 1.0000 1.0000 0.4621 0.4621 - - - - - - Rural Philippines
Wages 0.3732 0.7313 0.7203 0.1966 46.96 1.2581 Entrepreneurial income 0.3654 0.5296 0.5963 0.1154 27.56 0.7542 Other income 0.2047 0.6365 0.5115 0.0666 15.92 0.7777 Remittances from migrants 0.0567 0.7402 0.9542 0.0401 9.57 1.6868
Total 1.0000 1.0000 0.4187 0.4187 - - - - - - NOTE: Figures may not add up correctly due to rounding.
^ computed as (Si) * (Ri) * (Gi) * computed as Ri*(Gi/G), where G = total-income Gini
60
Table A6. Composition of Household Income Inequality, 2000.
Income Source
Share in Total Income (Si)
Gini Correlation (Ri)
Gini Coefficient (Gi)
Contribution Gini ̂ Contribution to Gini (%)
Relative Concentration Coefficients*
All Philippines
Wages 0.4586 0.7890 0.6725 0.2433 51.02 1.1125 Entrepreneurial income 0.2444 0.4850 0.7121 0.0844 17.70 0.7242 Other income 0.2109 0.7234 0.5961 0.0909 19.07 0.9041 Remittances from migrants 0.0861 0.7266 0.9311 0.0582 12.21 1.4185
Total 1.0000 1.0000 0.4769 0.4769 - - - - - - Urban Philippines
Wages 0.4843 0.7539 0.6134 0.2239 49.37 1.0193 Entrepreneurial income 0.2121 0.5059 0.7673 0.0824 18.15 0.8558 Other income 0.2119 0.7108 0.6065 0.0914 20.14 0.9503 Remittances from migrants 0.0916 0.6675 0.9151 0.0560 12.34 1.3466
Total 1.0000 1.0000 0.4536 0.4536 - - - - - - Rural Philippines
Wages 0.3776 0.7363 0.7181 0.1996 48.51 1.2847 Entrepreneurial income 0.3459 0.4985 0.5849 0.1008 24.51 0.7084 Other income 0.2077 0.5995 0.4849 0.0604 14.67 0.7065 Remittances from migrants 0.0688 0.7740 0.9515 0.0506 12.31 1.7897
Total 1.0000 1.0000 0.4115 0.4115 - - - - - - NOTE: Figures may not add up correctly due to rounding.
^ computed as (Si) * (Ri) * (Gi) * computed as Ri*(Gi/G), where G = total-income Gini