NATIONAL HIGHER EDUCATION RESEARCH INSTITUTE

Higher Education as a Source of Economic Growth:

Input-Output Analysis

Final Report

Research Team

Ooi Koon Peng

Ong Wooi Leng

Chan Huan Chiang

Penang Institute

(Formerly known as Socio-Economic & Environmental Research Institute)

10, Jalan Brown, 10350 Pulau Pinang

Tel: 04 228 3306 Fax: 04 226 7042

Email: enquiry@seri.com.my Website: www.seri.com.my

2

TABLE OF CONTENTS

TABLE OF CONTENTS ...................................................................................................................... 2

LIST OF TABLES ................................................................................................................................ 4

LIST OF FIGURES .............................................................................................................................. 5

LIST OF APPENDICES ....................................................................................................................... 6

GLOSSARY ........................................................................................................................................... 7

ABSTRACT ........................................................................................................................................... 8

PART 1 INTRODUCTION ............................................................................................................... 10

1.1 The business of higher education .......................................................................................... 10

1.2 Macro level analysis of higher education .............................................................................. 10

1.3 Malaysian input-output tables ............................................................................................... 11

1.4 Research questions ................................................................................................................ 12

1.5 Outline of this report ............................................................................................................. 12

PART 2 METHODOLOGY .............................................................................................................. 13

2.1 The case for the input-output analytical approach ................................................................ 13

2.2 The input-output framework ................................................................................................. 14

2.3 Derivation of input coefficients for higher education ........................................................... 15

2.4 Calculating output, income, and employment multipliers ..................................................... 17

2.4.1 Simple output multiplier ............................................................................................... 17

2.4.2 Total output multiplier.................................................................................................. 19

2.4.3 The import multiplier ................................................................................................... 20

2.4.4 Simple income multiplier ............................................................................................. 21

2.4.5 Total income multiplier ................................................................................................ 22

2.4.6 Type I and Type II income multipliers ......................................................................... 22

2.4.7 The employment multipliers ........................................................................................ 23

2.5 Summary of the various input-output multipliers ................................................................. 24

PART 3 CONTRIBUTION OF HIGHER INSTITUTIONS THE ECONOMIC GROWTH .... 28

3.1 Macroeconomic impacts of higher institutions ..................................................................... 28

3.2 Direct and indirect impacts of higher institutions on the economy of Malaysia ................... 31

3.2.1 Simple output multipliers ............................................................................................. 31

3.2.2 Total output multipliers ................................................................................................ 32

3.3 Leakages abroad from components purchased by higher institutions ................................... 33

3.4 Contribution of foreign students............................................................................................ 34

PART 4 INCOME AND EMPLOYMENT GENERATIONS ....................................................... 36

4.1 Introduction ........................................................................................................................... 36

3

4.2 Income and employment multipliers ..................................................................................... 36

4.3 Analysis of income and employment multiplier effects ........................................................ 36

4.3.1 Analysis of income multiplier effects .......................................................................... 36

4.3.2 Analysis of employment multiplier effects .................................................................. 37

4.4 Income and employment effects between higher education and other industries ................. 38

PART 5 STRUCTURAL CHANGES OF HIGHER INSTITUTIONS ......................................... 43

PART 6 CONCLUSIONS AND POLICY ISSUES ......................................................................... 45

6.1 The context of higher education investments ........................................................................ 45

6.2 Methodological approaches ................................................................................................... 45

6.3 Simple and total output multipliers ....................................................................................... 46

6.4 Income multipliers ................................................................................................................. 47

6.5 Employment multipliers ........................................................................................................ 47

6.6 Higher education as an economic sector ............................................................................... 48

6.7 Research issues ...................................................................................................................... 50

BIBLIOGRAPHY ............................................................................................................................... 51

APPENDICES ..................................................................................................................................... 52

APPENDIX A ...................................................................................................................................... 53

APPENDIX B ...................................................................................................................................... 54

APPENDIX C ...................................................................................................................................... 59

APPENDIX D ...................................................................................................................................... 61

APPENDIX E ...................................................................................................................................... 67

APPENDIX F ....................................................................................................................................... 75

APPENDIX G ...................................................................................................................................... 78

4

LIST OF TABLES

Table 1.1: Description of Malaysian Input-Output Tables .................................................................... 11

Table 2.1: Scaling of input coefficients of private higher institutions .................................................. 16

Table 3.1: Domestic and imported input of public higher education .................................................... 31

Table 3.2: Domestic and imported input of private higher education ................................................... 32

Table 3.3: Simple and total output multipliers for public and private higher education ....................... 33

Table 3.4: Output required as a result of a change in the expenditure by a typical foreign student ...... 35

Table 4.1: Income and employment multiplier of private and public higher education ....................... 37

Table 4.2: Comparison of income multiplier effects for selected industries with reference to private

higher education .................................................................................................................. 39

Table 4.3: Comparison of income multiplier effects for selected industries with reference to public

higher education .................................................................................................................. 40

Table 4.4: Employment multiplier effects of higher education ............................................................. 41

5

LIST OF FIGURES

Figure 2.1: Structure of a typical Input-output table ............................................................................. 14

Figure 3.2: Forward and backward linkages of private higher education ............................................. 29

Figure 3.3: Forward and backward linkages of public higher education .............................................. 30

Figure 3.4: Input composition of public and private higher education ................................................. 33

Figure 5.1: Input elements of education industry from 1987 to 2005 ................................................... 43

6

LIST OF APPENDICES

Table A.1: Input coefficients of private and public higher institutions ................................................. 53

Table B.1: Leontief Inverse of private and public higher education (Domestic) .................................. 54

Table B.2: Simple and total output multiplier (inclusion of household consumption) ......................... 55

Table B.3: Domestic, import and total output multiplier ...................................................................... 57

Table C.1: Import coefficient of private higher education .................................................................... 59

Table C.2: Import coefficients of public higher education.................................................................... 60

Table D.1: Forward and backward linkages of private higher education .............................................. 61

Table D.2: List of industries of backward and forward linkages for private higher education model .. 63

Table D.3: Values of forward and backward linkages of public higher education ............................... 64

Table D.4: List of industries of backward and forward linkages for public higher education model ... 66

Table E.1: Simple and total income multiplier of private higher education.......................................... 67

Table E.2: Direct, indirect and total income multiplier effects of private higher education ................. 69

Table E.2: Simple and total income multiplier of public higher education ........................................... 71

Table E.4: Direct, indirect and total income multiplier effects of public higher education .................. 73

Table F.1: Classifications of 94 categories to 16 categories ................................................................. 75

Table F.2: Simple and total employment multiplier .............................................................................. 77

Table G.1: Leontief inverse matrix of public higher education model, (I-A)-1 ..................................... 78

Table G.2: Output generated as a result of the expenditure by a typical foreign student .................... 102

7

GLOSSARY

Production Function: The process adopted in which inputs are combined and converted into outputs.

Business model: Strategies and procedures devised by businesses to maximise profits and achieve

long term sustainability.

Multiplier: The direct, indirect and induced economic impacts measured in how much increase in

Ringgit across all industry sectors following one Ringgit of initial investment in the reference

industry.

Input-output table: Rows and columns showing sales and purchases of intermediate inputs from

industry i to industry j to facilitate production by industry j.

Direct linkage: Industry i selling to industry j.

Indirect linkage: Industry i buying from another industry in order to facilitate its sale to industry j.

Induced linkage: Changes to household consumption of output from industries resulting from

increase sales of industry i that led to higher incomes or employment of workers.

Final demand: Sales to finish products to satisfy consumption, investments and exports.

Intermediate demand: Sales of components inputs from one industry to another industry.

Power industry: An industry that buys large amounts of inputs from other industries giving it power

as a lead industry upon which other industries depend on.

Sensitive industry: An industry that sells a lot of inputs to other industries and will thus be sensitive

to how much these other industries will buy as they are affected by economic circumstance.

Type I and Type II multipliers: Type I and Type II multipliers measure changes to income (or

employment) resulting from each Ringgit (or worker) increase in income (or employment) directly

and indirectly as a result from the initial RM 1 investment to the reference industry.

8

ABSTRACT

The impact of public and private higher education in Malaysia is examined by calculating

simple, total, Type I and Type II input-output multipliers. The table published by the Department of

Statistics distinguishes between private and public delivery of education but makes no distinction

between higher schools and college or universities. Adjustments had therefore to be made to the

input-output coefficients to reflect higher education using expenditure data of Universiti Sains

Malaysia as a baseline.

Input-output coefficients reflect fixed (i.e., non-substitutable) inputs to production output,

often referred to as the Leontief production function which is constant returns to scale but is able to

accommodate a large number of industry inputs. This enabled a comprehensive view of inter industry

linkages and therefore useful in the assessment of economic inputs. The table analysed contained 94

industry sectors showing sales and purchases between them.

The simple multipliers show direct and indirect impacts. The numbers show (backward links)

and forward sales for both public and private higher education and these were found to be slightly

below the median among the 94 industries, indicating that higher education is not a powerful (buys a

lot from other industries) nor is it a sensitive (sells a lot to other industries) industry. In this regard, it

is not a high impact industrial sector of the economy. Leakages due to imported inputs were also

analysed and it was found that public higher education show a noticeably high import content.

Ignoring such imported leakage, real estate is the main input component used by higher education in

Malaysia.

Type I and Type II income and employment multipliers were calculated after treating the

household sector as an additional, i.e., 95th sector that will also have inter-industry connections with

the other industries. Income and employment multipliers show the increase in employment income

and increase in numbers employed that are induced by the increase in industry sales that occurred

because of the initial RM1 investment made on the higher education sector.

Education in general and higher education in particular are public as well as private goods. It

is a public good because an educated society has a higher literacy rate and potentially more

productive. Collective social benefits will increase with increase in the supply of education. It is also

a private good, because an individual investing in education by paying for and attending courses in

order to attain higher qualifications can potentially earn a much higher lifetime income. In this

connection, the question arises in Malaysia over its coexistence of both public and private delivery of

higher education and adopting different production functions as revealed by their input-output

9

linkages with other industry sectors. This dual delivery evolved from legislations made in during the

90s that replaced the 1961 Education Act that began to allow private higher education delivery in the

country. By doing so, the social purpose of providing higher education as a collective social benefit in

the quest of nation building can be pursued alongside democratized (open to all) education provisions

as a market product.

Such a circumstance leads to several research questions that will need further investigations:

will the dual delivery system converge into single delivery system not in terms of whether it will be a

public or private sector business but whether the new delivery system will be modelled closely to

current public higher education model or the current private model? Will higher education in

Malaysia foster closer interconnections with the other industry sectors, in other words result in higher

multiplier values when similar analysis is conducted using new input-output tables. How do the

current inter industry connections between higher education and other industry sectors in Malaysia

compare against best practices found in other countries as revealed by their input-output coefficients?

This will allow a view as to the best possible expansion path for Malaysias higher education sector in

which its Leontief production function are adjusted over time and evolved into a best-practice model.

10

PART 1 INTRODUCTION

1.1 The business of higher education

This study on the impact of higher education on the economy was commissioned by the

National Higher Education Research Institute. Conceptually, making the link between a specific

sector and the rest of the economy is a tall order, because how weak or how strong this link is will

depend on the complex interconnections of sales and purchases across business and industry. The

interest of inter business linkages between education and the rest of the economy has both a micro as

well as macro view that unfortunately cannot be achieved in a single study. Higher education is a

business. The business of achieving its organizational aspiration will involve investments as well as

everyday spending. Expenditures on inputs and the output of private higher education becomes a

formula called the production function. The formula applied by any one institute of higher education

will depend on its business model that is developed according to the institutional strategy of the

individual organisation. A study pursued along these lines will be a micro level study in which the

aim would be to explore prospects of inter business linkages whereby tweaking strategies might lead

to changing inputs and alter the pattern of the industrial cluster involving higher education. Such a

study will be an industry outlook that envisions social and business development possibilities in the

light of current and recommended policies.

1.2 Macro level analysis of higher education

The macro level analysis, on the other hand, that was adopted by this study looked at data

compiled on inter-industry input-output tables compiled by Malaysias Department of Statistics.

Macro analysis is chosen to ensure economy-wide inclusiveness, not possible had the study been done

by polling from a selected sample of institutions. When higher education buys from another sector of

the economy, two things happen. First that sector will have to produce more to meet this supply

request from higher education. Second, higher production will also require additional inputs from yet

other sectors. Increased sales in these other sectors will require them also to increase production

which will then lead to more buying of inputs across the industry. Every new purchase from higher

education thus creates direct, indirect as well as induced impacts on the economy. Households also

make additional consumption purchases due to income effects and employment creations resulting

from increased investments into higher education. It all starts with higher education spending the first

Ringgit. The total accumulated buying and selling in the economy that spins off from this Ringgit

becomes the multiplier, which can then be used as a measure of the total impact resulting from that

one Ringgit of spending.

Upon making the decision to approach this study using input-output analysis the

methodological issues were immediately addressed. The theoretical basis of input-output methods has

11

gone through many decades of development since it was proposed by Wassily Leontief in 1941 for

which he was awarded the Nobel Memorial Prize in Economic Sciences in 1973. Instead of a

production function which would be somewhat unique to individual organization, the Leontief

production function is the sum of fixed coefficients (inputs per unit of output) representing inputs of

purchases from all included sectors of the economy. The Leontief production function is

characteristically non substitutable in terms of inputs and constant returns to scale. This limitation is

offset by the ability to include a wide variety of inputs not possible with the more popular Cobb-

Douglas of CES production functions.

Input-output analysis has its antecedence in classical economics. The French economists

Francois Quesnay in his Tableau Economique or economic table (published in 1758) constructed a

hypothetical table representing the relationship between output and expenditures of farmers,

landowners and manufacturers. Later, Leon Walrus developed a general equilibrium model

demonstrating the interdependence of markets comprising many different industries.

1.3 Malaysian input-output tables

In Peninsular Malaysia, although the Department of Statistics compiled its first input-output

tables for the year 1960, the tables were not officially published. Only input-output tables for the

years of 1965 and 1970 were compiled and released to the public. From 1970 onwards, the input-

output tables were modelled after the United Nations System of Nation Accounts 1968. With the joint

collaboration with the Economic Planning Unit (EPU), the Department then began to assemble the

Malaysian input-output tables for the year 1971. The first comprehensive set of Malaysian input-

output tables (inclusive of Sabah and Sarawak) was subsequently published for the year of 1978. It

was the collaborative effort of the Department of Statistics, the EPU and the United Nation

Development Program (UNDP). Table 1.1 shows the different number of industries/commodities and

availability of private and public education breakdown in various issues of Malaysian input-output

tables over the years.

Table 1.1: Description of Malaysian Input-Output Tables

No. I-O Year Year of

Publication

n x n

industries/commodities

Availability of private and public

education

1. 1978 1987 60 x 60 Yes

2. 1983 1988 60 x 60 Yes

3. 1987 1994 60 x 60 Yes

4. 1991 2002 92 x 92 No

5. 2000 2005 94 x 94 Yes

6. 2005 2010 120 x 120 No

12

1.4 Research questions

In the attempt to address the issue of the link between higher education and the rest of

Malaysias economy, the following research questions were addressed in this study:

a) What are the direct, indirect and induced impacts of one dollar rise of investment (or

consumption) in higher education in Malaysia?

b) How much of the above impacts will occur within the Malaysias economy and how much of

it will be leaked to industries abroad?

c) What are the other industries (in Malaysia and abroad) that have strong backward linkages to

higher education in Malaysia? These are industries that are dependent on the performance of

higher education in Malaysia, because they sell to higher education which form an important

intermediate market for those industries.

d) What are the other industries (in Malaysia and abroad) that have strong forward linkages to

higher education in Malaysia? These are industries whose performances are important to

higher education in Malaysia, because these industries form an important market for higher

education.

1.5 Outline of this report

This report is divided into five parts. Part 1 introduces the development of input-output

analysis as well as the research needs of this study. Part 2 reports on the methodology and outlines the

input-output framework adopted in the study. Input-output coefficients for higher education were

derived and using the Leontief inverse function, the output, import as well as income and employment

multipliers could then be calculated.

In Part 3, the prospective contributions made by higher education to Malaysias economic

growth are discussed. An assessment was made of the macro view of the overall economy and then

the inter-industry linkages between higher education and the other sectors in the economy were

established. Since Malaysia is a highly open economy in which exports plus imports together form

nearly two times the size of Malaysias gross domestic product (GDP) leakages of the multiplier

effects resulting from purchases made by higher education from industries located abroad were also

analysed.

Part 4 is dedicated to analysis of income and employment generation based on Total, Type I

and Type II multipliers, following which possible changes to the inter-industry links between private

higher education and the economy resulting from structural transformation were considered in Part 5.

Finally policy issues identified in this study are listed in Part 6.

13

PART 2 METHODOLOGY

2.1 The case for the input-output analytical approach

The Input-Output approach was dismissed by sceptics in most cases due to the restrictive

assumptions in its calculations. Among others, the model assumes constant returns to scale (that is,

input coefficients are independent of the level of output) when an exogenous demand is made upon an

industry.1 Furthermore the input coefficients are also fixed for each industry meaning that using less

of one type of input and more of another through substitution is not possible. This is not always true

in reality. If for example the pattern of incremental investment deviates from the norm (for example if

incremental investment focuses only on developing infrastructure for higher institutions or only on

hiring highly qualified lecturers) the use of a reference input-output table would not be able to

accurately reveal the actual impact produced. Instead the input-output model assumes that

investments are proportionate to the input coefficients. Many have also argued that the advancement

of technology and innovation may also have rendered past input-output tables obsolete since

innovation will likely alter the type of inputs used in production.

A comparison of the input-output table for the year 2000 and the input-output table for the

year 2005 verified the fact that input coefficients did vary over time. Within five years, the gross input

of higher education institutions had increased from RM13, 563,769.00 to RM 21, 794, 245.00, but the

total intermediate input coefficient, together with many of its components, had not remained the same.

In fact, it shifted from 0.1364 to 0.2869, suggesting a proportionally higher intermediate transaction

than primary input spending.2 The static nature of input-output analysis can thus to be overcome by

repeating the analysis using tables from different years to examine changes over time.

The input-output analytical approach, however, remains as the most pragmatic way forward.

First, it allowed for impact forecasting on a national scale by imbedding into the model a matrix that

represented the interdependency of all the industries in Malaysia. Second, this approach took into

account of the wealth of data by levying on official information diligently compiled in the form of

input-output tables. Third, the input-output model revealed the direct, indirect and induced linkages

amongst industries, making both intra-industry analysis between private and public higher

institutions and inter-industry analysis between higher institutions and the other industries possible.

Fourth, when closed in with information on income and employment, this method was able to

project values of income and employment multipliers, thus increasing the breadth of this researchs

scope and painting a more in-depth picture of the higher education industry in Malaysia today.

1 Hewings Geoffrey J.D., Regional Input-output Analysis, Sage Publication, 1985, pg 28. 2 Department of Statistics Malaysia, Input-output Tables 2000, 2000, Table 17; Department of Statistics Malaysia, Input-

output Tables 2005, 2005, Table 23. This point will be further explored in Part 4.

14

Investment on higher education industry was far from being an instantaneous project3 and

under circumstances such as demand and price fluctuation over time4, an impeccably accurate

forecasting mechanism will be impossible.5 But again, application of the model at different points in

time allowed us to circumvent the problem. The input-output analytical approach also offered

flexibility that made altering some parameters possible in order to satisfy the scope of this research.

2.2 The input-output framework

Figure 2.1: Structure of a typical Input-output table

Industries Final Demand Total Output

1 2 3 4 5 6

1 0.2 0.3

2 0.4 0.1

Industries 3

4

5

6

Primary Inputs 0.1 0.2

Total Input

The raw input-output table shows all the direct linkages amongst industries in the economy; a

row of the table shows the sales made by a selected industry to the other industries whereas a column

of the table shows the purchases made by a selected industry from the other industries. It is the latter

that we are particularly interested in here for further analysis. By dividing all these intermediate

transactions (section A in Figure 2.1) by the total input (section D in Figure 2.1), we would arrive at a

table containing the percentage of total inputs required from each industry for a Ringgits worth of

production. This is theoretically sound as industry js demand for inputs from other industries will

have been related to the amount of outputs produced by industry j. With reference to Figure 2.1, 0.3

and 0.1 Ringgits worth of output for example, are needed from industry 1 and industry 2 respectively

3 In the research paper The Economic Impact of Colleges and Universities, the authors, John J. Siegfried, Allen R. Sanderson and Peter McHenry, have argued that it is impossible to identify a period of time over which the difference

between the presence and the absence of a college can be discerned because most colleges start small and grow slowly over

time in the United States. There is no reason to suggest why the higher institutions in Malaysia would defy this trend. 4 Sajal Lahiris notion of scale-dependent coefficients and The Hudson and Jorgensons translogarithmic production function had respectively asserted that the level of demand and the relative level of price determined the value of input coefficients. 5 Gerking had long challenged, in multiple papers, the notion of the input-output model being developed without error.

E A B

C

D

15

for a Ringgits worth of final output produced by industry 2. Looking down the columns then, we

would know how a Ringgits worth of output required of a selected industry is being assembled from

different industries or through the use of labour information on labour can typically be gleaned by

analysing the components of primary inputs (section C in Figure 2.1).

Assuming that these coefficients do not vary with the scale of production (i.e., constant

returns to scale), we have in essence arrived at the recipe for production. It goes without saying, of

course, that the accuracy of this recipe depends on its deviation from the reference input-output

table as the scale of production changes. The question is: if 0.3 Ringgits worth of output is purchased

from industry 1 by industry 2 when a Ringgits worth of output is required of industry 2, how likely is

it that say, 300 Ringgits worth of output are purchased from industry 1 by industry 2 when 1000

Ringgits worth of output is required of industry 2? By having this nonetheless, we have the ability to

make estimation on the inter-industry transactions that will happen when a known volume of

exogenous demand is placed upon a selected industry. This is only a start because this piece of

valuable information can be further processed to portray the indirect linkages amongst all the

industries, local industries reliance on import and the creation of income and employment as a

corollary of production activities.

2.3 Derivation of input coefficients for higher education

This study was focused on the impact of private and public higher education on the Malaysian

economy. Unfortunately, the input-output table for the year 2000 published by Malaysias

Department of Statistics contained only details on public and private institutions but did not segregate

them further between schools and tertiary education. The first task of this study was therefore, to

produce two columns of private and public higher institutions by adjusting the input coefficients

provided in the input-output table for the year 2000.

16

Table 2.1: Scaling of input coefficients of private higher institutions

Activity

listed in

the I/O

2000 table

Private

Education Input

Coefficient

Total

Coefficient

(A)

Activity

listed in

USMs spending

data

USMs

spending

amount

(B)

USMs

Input

Coefficient

(C)

Scale

Factor

(D)

Scaled

Input

Coefficient

(E)

Hotel and

Restaurants 0.0118777 0.009796812

Transport 0.0104356 0.0223133

Travelling

Expenses

and

Delivery

17198062 0.01840417 0.82480715 0.008607357

Formulae and Clarifications:

(A): The addition of both input coefficients of Hotel and Restaurants and Transport as listed in column 1

(B): The actual spending made by USM based on the details obtained

(C): The ratio of USMs spending on Travelling Expenses and Delivery to its total spending of RM 934,465,523

(D): The ratio of USMs input coefficient to total coefficient (A)

(E): The product of each input coefficient and scale factor (D)

All 94 input coefficients of both private and public education listed in the Input-output 2000

tables. Table 2.1 were scaled using a two-pronged method: first, they were matched against the 9

categories listed in USMs spending data. Second, their input coefficients were scaled by the

corresponding USM input coefficients. Theoretically, it seemed almost as if the input coefficients of

USM had been used as weights for the activities found in the input-output table because it

represented higher educational institution. In the example given above, the scaled input coefficients of

the two items were found to be more than those of the actual Input-output 2000 tables by a factor of

approximately 0.82. We thus assumed that private higher education spends less in those items as

compared to private education as a whole. The same process was applied on public higher education

and the two columns were then substituted into the original input-output table for further adjustments.

The substitution occurred separately for both because our aim was to trace the relationships between

private higher education and the other industries (inclusive of public education in this case) and vice

versa.

This step necessarily assumed that USMs spending pattern was representative of the

contribution of higher institutions to the economy. Spending details of higher institutions are highly

confidential materials they were hardly accessible for the purpose of research. The input-output

table for the year 2000 was chosen in preference over the input-output table for the year 2005 because

the latter, albeit being the latest and most updated, merged both private and public institutions into a

single sector.

17

There was also an issue in accurately collapsing the 94 industries listed in the Input-output

table 2000 into the 16 categories (fewer) according to USMs spending data. Here, uncertainty

persists in less obvious cases. For example, USMs spending on delivery and travelling individually

would contribute to the transport industry in Malaysia. How can we decide then, the exact proportion

of contribution to the transport industry each made? Although this problem could be solved by

merging the delivery and travelling expenses into a single category (as shown above), this was done at

the cost of achieving higher precision because then the scale factor for both delivery and travelling

expenses would be identical and not according to their respective weights.

2.4 Calculating output, income, and employment multipliers

2.4.1 Simple output multiplier

Recall that a raw input-output table portrays only the direct linkages among industries. In

reality, the interactions among industries are more complicated than what the direct flows of output

seemed to suggest. For example, where an industry did not have a direct linkage with another industry

in the economy, this second industry may draw benefit from the expansion of the first industry if there

was a third industry that bound the two together. That would occur if the second industry sold to the

third industry, which in turn, sold to the first industry. We call this an indirect linkage. Direct and

indirect linkages can be numerically shown by making X, the final demand, as the subject of the

equation as follow:

AX + Y = X [1]

Y = X AX

Y = (I A)X

X = (I A)-1Y [2]

X = (I A)-1Y

X = (I + A + A2 + A3 + A4 + A)Y [3]

X = (I)Y + (A + A2 + A3 + A4 + A)Y [4]

A is the square matrix of the inter-industry flow expressed in terms of coefficient per dollar of output

(section A in Figure 2.1); X is the gross output by industries (section E in Figure 2.1); Y is the final

demand (section B in Figure 2.1); I is an identity matrix.

I is associated with the initial output effect on the economy because it reflects the initial

Ringgits worth of an industrys output needed to satisfy the final demand as shown in Equation [4].

Direct linkages (or what we interpret directly from the values of input coefficients, A) show only the

first round of spending. Indirect linkages or the subsequent rounds of spending are taken into account

by the expanded mathematical expression, (A2 + A3 + A4 + A), in Equation [3]. If we total up all

18

the entries of any column, we would arrive at what is known as the simple output multiplier.

Formally, it was the ratio of direct and indirect effects to the initial output effect alone, obtained from

a model in which households are exogenous. That figure stated the total value of production of all

industries that was necessary to produce a Ringgits worth of final demand for that industrys output.

To illustrate this concept further, we can draw information from Table 2.1 and assume that

the economy has only two industries, A and B, with the following input coefficients:

A =

And the Leontief Inverse, according to the formula, would be:

(I A)-1 =

If we use Y and X to represent changes in final demand and changes in gross output

respectively, from equation [2], we know that X = (I A)-1Y. Therefore mathematically, a Y of

(a Ringgits worth of final demand for the output of industry 1 and none for the output of

industry 2) would give rise to X of and Y of (a Ringgits worth of final demand for

the output of sector 2 and none for the industry of sector 1) would give rise to X of .

From the Leontief Inverse, it is shown that additional outputs of RM 1.5 from industry 1 and

RM 0.667 from industry 2 were required for a Ringgits worth of final demand for the output of

industry 1 and additional outputs of RM 0.5 from industry 1 and RM 1.333 from industry 2 were

required for a Ringgits worth of final demand for the output of industry 2. Note that the amount of

RM 1.5 represented RM 1 from industry 1 to satisfy the initial Ringgit of final demand plus an

additional RM 0.5 from intra- and inter-industry transaction.

More formally, if we represented the elements of (I A)-1 as ij, where i and j refer to the row

and column of an element in a matrix respectively, then the output multiplier of a selected industry

j, Oj, can be calculated from the formula:

Oj = [4]

As a proof, if j refers to 2 in this case, then simply, O2 = 12 + 22 = 0.5 + 1.333 = 1.833

19

2.4.2 Total output multiplier

The model that we had dealt previously depended on the existence of an exogenous industry

and the kinds of transactions that constituted the activity of this industry were consumption purchases

by households, sales to government, gross private domestic investment and net export. In fact, the

exclusion of households from the productive industries may be considered as a strain on economic

theory because an increase in labour inputs due to increased output would lead to an increase in the

amounts spent by households as a group of consumers. This in turn, leads to an increase of demand on

industrial output and the cyclical pattern continues.

To reflect the fact that additional outputs were necessary to satisfy the anticipated increase in

consumer spending, the household industry can be moved from the final-demand column to the

interrelated production table, thus making it one of the endogenous industry. This is generally known

as closing the model with respect to households. This would require a row and a column for the new

household industry the former showing the how labour services is used as an input by the other

industries and the latter showing the consumption pattern of private consumers. Finally, the element

in the (n+1) row and the (n+1) column would represent the household purchases of labour services.

Resuming the previous example, lets assume the augmented matrix to be as follow:

=

Hence, (I )-1 would be:

(I )-1 = (I A)-1 =

In the above matrix, we let the industry 1s payment for labour services to be 0.1; industry 2s

payment for labour services to be 0.2; households spending on industry 1 to be 0.15; households

spending on industry 2 to be 0.05 and households payment for labour services, to be 0.1.

In a model with household endogenous, the value of each element was higher because the

added impact of more household consumption due to increased income was explicitly taken into

consideration in the model. For example, instead of 0.667, industry 1 would now purchase 0.723 from

industry 2 with respect to a change in final demand only now the final demand is exclusive of that

from households. In the calculation of total output multiplier though, we would not include the last

20

element, the household industry, in the summation because we were only interested in the total output

effect. This was unlike that of simple output multiplier where the approach was to total up the whole

column. If we denoted total output multiplier as j, the total output multiplier for industry 1 would

thus be:

1 =

Therefore, a Ringgits worth of final demand for industry 1s output and none for industry 2

would result in a X of 2.307 (1.584 + 0.723). This figure was in direct comparison to the simple

output multiplier of 2.167. The general formula for the total output multiplier, for industry j, is given

by:

j = [5]

As a proof, if j refers to 2 in this case, then, 2 = 12 + 22 = 0.594+1.396 = 1.99

2.4.3 The import multiplier

The total multiplying effect would actually be much larger than what the above calculations

show. Imports had already been winnowed out from the intermediate input coefficients as a separate

component of primary inputs (section C of Figure 2.1). That is to say, the 0.3 worth of output

purchased by industry 2 from industry 1 did not include inputs purchased from foreign countries. No

country was absolutely self-sufficient because industries in a country usually did rely on foreign

import that fed into domestic production. The only question here is, to what extent was the

dependency? The answer can be found by analysing the import multiplier.

The total multiplying effect, one that necessarily transcended across the boundaries of a

country, had also to include information on imports purchased from each sector as well. Fortunately,

the amount of import purchased from each industry can be gleaned directly from the input-output

table. Adding this to the inputs purchased from domestic industries, we would then arrive at the total

input requirement needed for production. We may find for example that industry 2 purchases not only

0.3 worth of output from industry 1 for a unit of production but more due to the inclusion of import.

The calculation of total output multiplier was identical to the steps and formulae mentioned above.

Import multiplier was then calculated by subtracting domestic output multiplier from total output

multiplier.

21

2.4.4 Simple income multiplier

Industries do not restrict their purchases to only the other industries; they also purchase

labour from the economy. Rather than just translating changes in final demand into total value of

industrial output, as in the previous section, it is also of importance to translate changes in final

demand into the creation of income.

An approach to calculate this would be to convert each of the elements in any column of (I

A)-1, which measured the value of direct and indirect output, into Ringgits worth of household

income via household input coefficients. These coefficients that made up the (n+1) row, previously

used to close the model with respect to households, represented income paid to workers per Ringgits

worth of industrial output.

In the example used above, if we assumed that the two sectors have returns to labour of 0.1

and 0.2 respectively. Direct and indirect income created can be calculated as follow:

V = (I A)-1 =

V (I A)-1 =

V (I A)-1 = (0.2834 0.3166)

A closer look at the mathematics would reveal how direct and indirect income created by an

exogenous demand was accounted for. Again, the figure 1.5 in the Leontief Inverse can be interpreted

as the direct and indirect transaction from Industry A to itself to meet a Ringgits worth of exogenous

demand for the output from Industry A. If Industry A pays 0.1 to its workers for a Ringgits worth of

exogenous demand for the output produced, then logically, the mathematical expression of

would reveal the income paid to the workers of Industry A. Likewise, the figure 0.667 represented the

direct and indirect transactions from Industry B to Industry A that met a Ringgits worth of exogenous

demand for the output from Industry A. If Industry B paid 0.2 to its workers for a Ringgits worth of

exogenous demand for the output produced, then the mathematical expression showed

the income paid to workers of Industry B. Adding up both values would allow us to obtain simple

income multiplier of Industry A.

If we denote simple household income multiplier for sector j as Hj, then:

Hj =

As a proof, if j refers to 2 in this case, then simply, J2 = 31. a12+ 32. a22 = 0.3166

22

2.4.5 Total income multiplier

The above picture, however, was yet not complete because the matrix had not taken into

account that wages and salaries received by employees that would then be spent purchasing more

goods and service, thereby generating demand for additional output and by extension, creates

additional income. To calculate this induced income effects, we have to do the same for the elements

in (I )-1. As before, using an over bar to denote the multiplier derived from , in which household

is endogenous to the matrix, the parallel of Hj is:

j =

We have, from previous sections, the following two matrices:

= (I )-1 =

Following our numerical example,

1 = (0.1) (1.584) + (0.2) (0.723) + (0.1) (0.337) = 0.337

2 = (0.1) (0.549) + (0.2) (1.396) + (0.1) (0.376) = 0.376

Note that these total income multipliers are equal to the first two elements of the last row of (I

)-1. This is a mathematical property resulting from the inverse of matrix . But recall too that any

element in (I )-1 measures the direct, indirect and induced effects on sector is output of a Ringgits

worth of demand for industry js output. Thus, measures the direct, indirect and induced effects

on the output of household industry, labour services, when there is a Ringgits worth of demand for

industry js output. Therefore alternatively, the formula for total income multiplier is:

j =

2.4.6 Type I and Type II income multipliers

With output multipliers it was fairly clear that the initial effect of a Ringgits worth of final

demand for the output of industry j is that industry js production must increase by a Ringgit (and

subsequently of course, more than a Ringgit). But for income, there was another option of what

should be logically termed the initial effect of new demand because a Ringgits worth of new output

from sector j also means an additional income payment of to workers in industry j. Therefore

23

could be viewed as the initial income effect of demand for industry js output. If this was the

case, then we have two more multipliers to adopt for analysis: Type I and Type II.

Type I income multiplier had the direct and indirect income effects, or the simple income

multiplier, as a numerator and used as a denominator not the initial Ringgits worth of output but

rather its initial labour income effect, If we denoted Yj as Type I income multiplier for

industry j, then:

Yj =

Following our numerical example,

Y1 = (0.2834/0.1) = 2.834

Y2 = (0.3166/0.2) = 1.583

If instead, direct, indirect and induced income effects, or the total income multiplier, were used as a

numerator, then Type II income multiplier can be found. If we denote j as Type II income multiplier,

then:

j =

Following our numerical example,

1 = (0.337/0.1) = 3.77

2 = (0.376/0.2) = 1.88

2.4.7 The employment multipliers

To calculate employment multipliers, it was necessary to estimate the number of workers

employed in an industry relative to that industrys value of output. Note that unlike the (n+1)

(household) row, these labour coefficients were computed in physical, not monetary, terms.

Details on the number of workers hired in different industries were obtained from the

Department of Statistics in Malaysia. However, they were not segregated into the 94 industries of the

input-output table for the year 2009. Instead, they were only classified into 16 industries following the

24

Malaysian Standard Industrial Classification (MSIC) 2000. In light of this, the 94 industries of out

input-output table had to be reclassified to fit into the 16 MSIC codes.

With the reclassified 1616 input-output matrix, the input coefficients of employment were

calculated by dividing the number of employees of an industry by the value of the output produced.

This gave us the value of an employee for a Ringgits worth of output. If we assumed X1 to be RM

1,000,000 and e1, the number of workers in industry 1, to be 3,000. Then the physical labour input

coefficient, wn+1,i, is simply 0.003 (e1/X1).

The calculations of simple, total, Type I and Type II employment multipliers paralleled those

of income multipliers described above. The only thing was that physical labour input coefficients,

wn+1,j, was used here instead of monetary labour input coefficients, an+1,j. If we denoted Ej, j, Wj, j

as simple employment multiplier, total employment multiplier, Type I employment multiplier and

Type II employment multiplier respectively, then the formulae are as follow:

Ej = j =

Wj = j =

2.5 Summary of the various input-output multipliers

There are ten different types of multipliers in total that we will summarise here:

1. Simple output multiplier

n

i

ijjO1

The simple output multiplier measures how much additional output would be required for a

Ringgits worth of final demand for the output of private and public higher education institution. It

can be obtained by summing up the column entries of Leontief Inverse. The analysis can be found in

Section 3.2.

2. Total output multiplier

1

1

n

i

ijjO

The total output multiplier, on the other hand, describes the amount of additional output

needed resulting from a Ringgits worth of final demand for the output of private and public higher

education institution but also takes into account the induced effect of household income generation

through payment for labour services and associated with consumer expenditures or goods produced

25

by the various sectors. This is explained by the element (n+1) row and the element (n+1) column

which represent the household purchases of labour services. The findings of this multiplier are

presented in Section 3.2.

3. Simple income (household) multiplier

n

i

ijinj aH1

,1

The amount of additional direct and indirect income created for every Ringgits worth of

output produced is measured by the simple household income multiplier. The analysis of how much

income to be created across different industries when investments are made in higher education is

presented in Section 4.3.

4. Total income (household) multiplier

1

1

,1

n

i

ijinj aH

Alternatively, the total household income multiplier explains how much additional income to

be created as a result of the spending by household sector on goods and services produced. In other

words, it measures the direct, indirect and induced income effects on output purchased by household

sector when there is a Ringgits worth of new demand. The analysis of these effects can be found in

Section 4.3 and Section 4.4.

5. Type I income multiplier

in

j

ia

HY

,1

The Type I income multiplier measures the change of direct and indirect income generated by

the industry with respect to a change in the initial (direct) income payment to workers in the industry.

In short, it can be calculated by taking the ratio of simple household income multiplier to initial

labour income. The results of this analysis can be obtained in Section 4.3 and Section 4.4.

6. Type II income multiplier

in

j

ja

HY

,1

The Type II income multiplier, conversely, assesses the direct, indirect and induced income

generated by the industry with regard to a Ringgit change in initial labour income. It can be estimated

by taking the ratio of total household income multiplier to initial labour income. This analysis is

provided in Section 4.3 and Section 4.4.

26

7. Simple employment (household) multiplier

n

i

ijinj WE1

,1

From the employment perspective, the simple household employment multiplier is used to

gauge the number of additional jobs to be created for every Ringgits worth of output produced. The

analysis of the number of job vacancies made available across different industries is discussed in

Section 4.3 and Section 4.4.

8. Total employment (household) multiplier

1

1

,1

n

i

ijinj WE

As for the total household employment multiplier, it measures the number of additional jobs

to be created as a result of the expenditure by household sector on goods and services produced. In

other context, this multiplier is also taken into account of direct, indirect and induced employment

effects for every Ringgit of new demand. The result and analysis is shown in Section 4.3 and Section

4.4.

9. Type I employment multiplier

in

j

jw

EW

,1

The Type I employment multiplier describes the change of direct and indirect employment

generated by the industry with respect to a change in the initial (direct) vacancies available in the

industry. In brief, it can be calculated by taking the ratio of simple household employment multiplier

to initial vacancies available. The results of this analysis can be obtained in Section 4.3 and Section

4.4.

10. Type II employment multiplier

in

j

jW

EW

,1

The Type II employment multiplier, on the other hand, assesses the direct, indirect and

induced employment generated by the industry with regard to a Ringgit change in initial vacancies

available. It can be estimated by taking the ratio of total household employment multiplier to initial

vacancies available. This analysis is deliberated in Section 4.3 and Section 4.4.

27

Where i and j are indices for rows and columns of the input-output matrix contains n number

of sectors or industries.

ij are elements of the 1)( AI inverse matrix and ij are elements of the

augmented inverse matrix that includes the household sector as an additional

industry. The coefficient of the household sector is ina ,1 .

These few pieces of information are all that are needed to calculate all ten different

combinations according to their formulae in order to address different issue, i.e., the direct and

indirect impacts of output resulting from investments or the induced impacts caused by additional

household spending as well as the creation of additional employment.

28

PART 3 CONTRIBUTION OF HIGHER INSTITUTIONS THE ECONOMIC

GROWTH

3.1 Macroeconomic impacts of higher institutions

If the government has a fixed amount of money to spend, a comparison of output multipliers

across different industries would show where this spending would produce the greatest level of

output. Recall from Part 2.3 that output multiplier shows the total value of production that is

necessary to satisfy a unit of exogenous demand. An industry with a large output multiplier therefore

would indicate that it has strong backward linkages compared to industries with smaller multiplier

values because it purchases substantially more from the other industries.

The ability to absorb outputs produced by the other industries grants this high multiplier

industry a measure of indispensability. Should the industry choose to push the brakes and halt its

production totally, its suppliers would confront a sudden fall in demand. Of course, this extremism

seldom happens in reality. Even so, when an industry that has a high output multiplier decides to take

a breather and scales down its level of production, its interconnected chain of suppliers would surely

be affected. For this supremacy the industry wields in the market, it is also known as a power

industry. Of course, a detailed analysis of its input coefficients (origin of supply) would still be

necessary to judge whether it is associated to just a few key industries or indeed, many.

Yet, this only tells half of the story. If tracing along a typical column of an input-output table

shows the purchases made by an industry from the others (i.e., backward linkages), then tracing along

a typical row of an input-output table shows the sales made by an industry to the others (i.e. forward

linkages). The row total of these coefficients would be able to illustrate a selected industrys forward

linkages to the other industries. Although this figure does not qualify as a multiplier of any sort

(unless investments for all the other sectors investment remain constant) it does serve to show the

link of an industry with the others in a reverse order.

Its analysis too, proves to be the opposite of that of output multiplier. A high row total

indicates that an industry has strong forward linkages to the other industries; it sells substantially to

satisfy the demand of its products or services. This is not necessarily a negative attribute (its

performance may be bolstered during a boom, say) but it does mean that the industry is more

susceptible to fluctuations in the economy whilst a power industry seems to be sitting as a

determinant of fluctuation. Due to that, an industry with strong forward linkages is termed as a

sensitive industry. Again, while the figure of row total equips us with a yardstick for comparative

analysis, a final verdict would require a detailed assessment of the sales made by an industry to the

others.

29

Our attempt was to classify industries according to the value of their forward and backward

linkages depending on whether these values were above or below the medians (that is whether they

fall into the top 50% or bottom 50 of all industries). The median values were arbitrary lines drawn

from the sample of all output multipliers calculated from the input-output table. The median was

preferred over mean as a measurement of central tendency simply due to the presence of extreme

values. With the two median values, all the 94 industries were segregated into four quadrants and the

industries of each quadrant were presented in Appendix D, Table D.2 (private higher education) and

D.4 (public higher education).

As indicated by the red star, Figure 3.2 shows that private higher institutions with output

multiplier and row total of 1.8010 and 1.0366 respectively were both lower than the median values of

2.6176 and 1.5758 calculated from all the industries (see Appendix D: Table D.1). This suggests that

private higher institutions have relatively low forward and backward linkages; backward linkages

were in fact, too distant from the median line compared to forward linkages. This result is very similar

to that of public higher institutions whereby the output multiplier and row total were 1.8030 and

1.0199 and the median values were 2.6176 and 1.2401, accordingly (see Appendix D: Table D.3). By

this analysis, private and public higher institutions appear to be not powerful and sensitive.

Figure 3.2: Forward and backward linkages of private higher education

30

If the value of output is the sole consideration of the Malaysian Government, based on the

figures then, the two industries studied would not be a preferred choice for investment not even for

a partial investment. Many others would not hesitate to assert too that higher institutions are valuable

because they are characteristically stable and less prone to economic fluctuation than other industries.

This argument seems to resonate with the reasonable output multiplier (indication of backward

linkage) and low row total (indication of forward linkage) calculated in this study.

Figure 3.3: Forward and backward linkages of public higher education

The top positions of the output multiplier are still, by and large, dominated by manufacturing

industries, with a few exceptions of transcendental service industries such as Radio and TV

broadcasting, Recycling, Electricity and Gas and Hotels & Restaurants. This may be due to the fact

that manufacturing industries have proportionally higher intermediate output transaction than that of

primary inputs. The same goes for those service industries. Both industries studied in this paper defy

this trend because they invest heavily on human resources and not on tangible goods and services.

Therefore, their linkages with other industries appear to be weaker. Based on this study, the ratios of

intermediate output transaction and primary input are 0.31 to 0.69 and 0.35 to 0.65 for higher private

education and higher public education respectively (see Appendix A: Table A.1).

Given that the output multipliers of the both industries are low relative to others this should

not lead us into dismissing them as not being important to the economy. The truth is that the sizes of

output multiplier may not indicate the actual contribution made by these industries to national growth

31

because by examining only their coefficients (i.e. per unit output) we have not taken into account the

industries level of output.

3.2 Direct and indirect impacts of higher institutions on the economy of Malaysia

3.2.1 Simple output multipliers

Recall from Section 2.4.1 that the simple output multiplier measures the additional direct and

indirect output required from respective industries in order to satisfy a unit of final demand. The

multipliers can be analysed from how much the input were consumed locally and abroad. The

decompositions of the simple output multiplier (SOM) into its domestic and imported components for

selected industries on the basis of their strong backward linkages with public and private higher

education are shown on Tables 3.1 and 3.2 for public and private higher education, respectively. The

breakdown between the domestic and imported components of the multipliers results from the input

sourcing by industries in Malaysia from either domestic or imported locations. The choice between

the two is not necessary due to preference but oftentimes; imported inputs give an indicator of

missing industries, in other words, input components that can be obtained from local sources. The

numbers shown here, however, will indicate how much of the multiplier will benefit local industries

in terms of business sales and how much might instead impact on industries abroad.

Table 3.1: Domestic and imported input of public higher education

Industry Domestic Industry Import Industry SOM

Output Multiplier 1.4228 Output Multiplier 0.3801 Output Multiplier 1.8030

Education - Public 1.0002 Business services 0.0404 Education - Public 1.0002

Real estate 0.0911 Paper & board industries 0.0376 Real estate 0.0947

Printing 0.0595 Manufacture radio, TV etc. 0.0258 Printing 0.0759

Wholes.&retail trade 0.0563 Manufacture industries

chemic.

0.0238 Wholes.&retail trade 0.0732

Electricity & gas 0.0448 Building, construction 0.0172 Business services 0.0578

Business services 0.0173 Wholes.&retail trade 0.0169 Paper & board

industries

0.0522

Hotels & restaurants 0.0170 Petrol & coal industries 0.0166 Electricity & gas 0.0520

Paper & board

industries 0.0146 Printing 0.0164 Manufacture radio, TV

etc. 0.0276

Transport 0.0132 Iron & steel industries 0.0143

Transport 0.1024

Communication 0.0130 Transport 0.0118

Manufacture motor

vehicle 0.0879

SOM simple output mutliplier

Table 3.1 and Table 3.2 show the top ten output-generating industries from initial RM 1

investment made in the respective industries for public higher education and private higher education,

respectively. The total simple output multipliers of private and public higher education made up about

1.8010 and 1.8030 of additional output generated that resulted from one Ringgits worth of final

demand made by public and private higher education, respectively. The small differences of the total

32

simple output multipliers between public and private higher education would not imply much. While

domestic output multipliers show larger impact in private higher education (1.4595) compared to

public higher education (1.4228), public higher education (0.3801) recorded to have higher import

multiplier than that of private higher education.

Table 3.2: Domestic and imported input of private higher education

Industry Domestic Industry Import Industry SOM

Output Multiplier 1.4595 Output Multiplier 0.3415 Output Multiplier 1.8010

Education - Private 1.0001 Paper & board industries 0.0487 Education - Private 1.0218

Real estate 0.0908 Manufacture radio, TV

etc.

0.0326 Real estate 0.0950

Printing 0.0797 Manufacture industries

chemic.

0.0224 Printing 0.0855

Wholes.&retail trade 0.0512 Education - Private 0.0217 Paper & board industries 0.0730

Electricity & gas 0.0310 Business services 0.0208 Wholes.&retail trade 0.0663

Communication 0.0285 Petrol & coal industries 0.0157 Electricity & gas 0.0379

Paper & board

industries

0.0243 Wholes.&retail trade 0.0152 Business services 0.0366

Business services 0.0158 Iron & steel industries 0.0114 Manufacture radio, TV

etc.

0.0355

Hotels & restaurants 0.0156 Transport 0.0112 Communication 0.0340

Transport 0.0155 Crude petrol, natural gas

& coal

0.0112 Petrol & coal industries 0.0279

SOM simple output mutliplier

Also, it should not be too surprising to see the education industry forking out supply of

1.0218 and 1.0002 (that is, above the amount of 1) for private and public higher institutions but less

than 0.1 for the rest. The fact was that the 1.0000 of both 1.0218 and 1.0002 were allocated to satisfy

the new unit of final demand whilst only the remaining of 0.0218 and 0.0002 accounted for inter- and

intra-industry use. Therefore, it seemed to suggest, superficially, that inter- and intra-industry

transactions for private as well as public higher institutions were in reality, only miniscule.

3.2.2 Total output multipliers

In addition to the simple output multiplier, the total output multipliers were also calculated.

These differ from simple output multiplier in that the input-output table of inter-industry linkages has

an additional row and column that includes the household sector as one of the industries. The

household row (with coefficients jna ,1 ) shows sales (employment salaries) to each of the other

industry sectors and the household column (with coefficients 1, nia ) shows household consumption

from each of the other industries.

The total multipliers for both public and private higher education are shown in Table 3.3.

Notice that total multipliers 4.5871 for private and 4.5851 for public higher education are

33

substantially much larger than simple multipliers because on top of the direct and indirect impacts

resulting from the initial one Ringgits worth of investments into either public or private higher

education, the induced impacts are also measured by the total multipliers. Induced impacts are the

result of increased household consumption across the industry from additional incomes and

employment.

Table 3.3: Simple and total output multipliers for public and private higher education

Higher Education

Simple Output Multiplier Total Output Multiplier

Private

Higher

Education

Public

Higher

Education

Private Higher

Education

Public

Higher

Education

Private Higher Education 1.8010 4.5871

Public Higher Education 1.8030 4.5851

3.3 Leakages abroad from components purchased by higher institutions

Analysis of output multiplier alone does not take into account the use of inputs imported from

countries abroad to meet production requirements needed to satisfy a given unit of final demand.

Some of the impacts discussed therefore may not be occur within the boundaries of Malaysia. In this

study, the domestic output multiplier is found to be 1.4595 (Figure 3.4). This means that for a unit of

final demand made upon private higher institutions, out of the 1.8010 value of output produced, only

1.4595 would be localised within the Malaysian economy. The import multiplier amounted to 0.3415.

This at least paints a slightly more optimistic picture than public higher institution where a higher rate

of leakage was detected. Out of 1.8030 value of output produced by public higher institutions, only

1.4228 would be retained in the local economy. 0.3801 would flow out of the Malaysian economy.

Figure 3.4: Input composition of public and private higher education

34

3.4 Contribution of Foreign Students

In our multiplier analysis, X multiplier value is driven by specifying Y, in which Y, RM

1 change in public higher education and therefore the multiplier calculated X will be how many time

more than one. In analysing foreign student expenditure in the economy, Y will be the total

expenditure per student in the economy across the different sectors. The resulting impact or X is

calculated exactly the same way.

In this example, it is estimated that a typical foreign student spends RM 30,000 a year in a

public higher institution. Of these, RM 15,000 is spent on tuition fee, RM 6,000 on real estate activity

or rental, RM 2,000 on Manufacture radio, TV etc or electronic devices and the remaining RM 7,000

on food, printing, transport, communication, health and recreation. The input-output model can be

used to analyse the impact of this spending by foreign student in terms of final demand in the rest of

the economy. To recall, the new output generated (X) for the sum of expenditure made by a typical

foreign student can be calculated through the multiplication of Leontief inverse matrix, that is (I-A)-1,

with the final demand, which is the amount of expenses by a typical foreign student or (Y). In short,

X = (I-A)-1 (Y). The 94 x 94 Leontief inverse matrix can be obtained from Appendix G: Table G.1.

The amount of new output required as a result of the expenses by a typical foreign student is

summarised in Table 3.4. On the whole, a total of RM 30,000 spending made by a typical foreign

student would require approximately additional RM 15,250s worth of output in order to satisfy the

exogenous increase in the expenditure of foreign student. In other words, a total impact of about 1.5

times worth of new output would be needed from the whole economy as a result of the total

expenditure of RM 30, 000 made by a typical international student. Out of RM 45,247.03 new output

required, over RM 30,000s worth of new output would be required outside the higher education

industry. With the exclusion of higher education, the result shows that other industries plays a vital

role in generating sufficient amount of new output so as to fulfil the amount of new demand

consumed by an international student.

Note that while sectors such as higher education, real estate and manufacture of radio, TV etc.

were estimated to have the major expenses made by a typical foreign student, the effects in terms of

new necessary output required would not be as large as those sectors with smaller expenses. For

instance, more than two-fold of additional output were needed from the activities of electricity & gas

(RM 1,805.77), livestock breeding (RM 660.54), printing (RM 1,376.26), manufacture oils & fats

(RM 738.32), and paper & board industries (RM 663.59) if RM 600, RM 300, RM 400, RM 300 and

RM 200s worth of new output were to generate, accordingly. This may suggest that a typical foreign

student would regard these goods and services as their necessity merchandises. By looking at the

linkages analysis in Section 3.1, with the exception of electricity & gas, all other industries were

35

seemed to have very strong backward and forward linkages, thus suggesting very powerful and

sensitive industries.

Other industries, on the other hand, would require an additional total of about RM 6,800s

worth of output so as to facilitate the amount of new demand needed by the foreign student. It was

also found that wholesale & retail trade (RM 1,530.37), agriculture other (RM 661.93), petrol & coal

industries (RM 625.76) and business services (RM 534.06) were estimated to generate the greatest

impacts if the final demand of RM 30,000 were to be met (See Appendix G: Table G.2).

Table 3.4: Output required as a result of a change in the expenditure made by a typical foreign

student

Industry

Foreign Student

Expenditure (RM)

(Y)

Output Required

(RM)

(X)

Higher education 15,000.00 15,008.77

Real estate 6,000.00 8,770.00

Manufacture radio, TV etc. 2,000.00 2,207.54

Grain mills 1,000.00 1,071.93

Dairy production 600.00 792.46

Electricity & gas 600.00 1,805.77

Transport 600.00 1,140.91

Communication 600.00 1,048.26

Fishing 400.00 663.95

Meat & meat production 400.00 468.30

Printing 400.00 1,376.26

Livestock breeding etc. 300.00 660.54

Crude petrol, natural gas & coal 300.00 609.07

Pres. of seafood 300.00 313.63

Manufacture oils and fats 300.00 738.32

Health - Public 300.00 301.14

Preservation of fruits &veg. 200.00 222.44

Bakeries 200.00 227.24

Paper & board industries 200.00 663.59

Recreation 200.00 217.62

Entertainment 100.00 128.66

Others (sum of 73 industries) 0.00 6,810.63

Total (sum of 94 industries) 30,000.00 45,247.03

36

PART 4 INCOME AND EMPLOYMENT GENERATIONS

4.1 Introduction

To further examine how much private and public higher education contributes to the

countrys economy, it is also possible to estimate the amount of additional income and jobs across

different industries. Note that income and employment generations do not depend on industries which

have high or low investments. As indicated in the previous section, despite private and public higher

education are found to be fairly stable industries and less volatile due to economic fluctuation, it does

not follow that income and employment opportunities generations will have similar effects as output

is created. Hence, the decision on how much investment to make may also be made as the basic of the

magnitude of income created or the number of additional employment from private and public higher

education investments.

This chapter provides an in depth analysis on how much income to be generated and how

many job opportunities created from private and public higher education investments by also

examining the direct, indirect and induced income and employment multipliers.

4.2 Income and employment multipliers

The income and employment multipliers have been described in details in Sections 2.4.4

through 2.4.7. They differ from output multipliers discussed in Part 3 by the treating the household

sector as an additional industry denoted by the coefficients jna ,1 for a total of n industries. There are

several types of income and employment multipliers but their analysis all revolves around induced

impacts on the economy in terms of changes in terms of increased household incomes resulting from

initial investments made in the reference industry which is public and private higher education in this

case. See Section 2.5 (pages 23-26) for the computational procedures and Appendix E for the results.

4.3 Analysis of income and employment multiplier effects

4.3.1 Analysis of income multiplier effects

The income and employment multipliers of private and public higher education are shown in

Table 4.1. The simple income multipliers show additional incomes generated as an induced impact

above the direct and indirect impacts that were produced from the initial RM 1 investments made to

the reference industry (i.e., public and private higher education). The simple income multiplier is 85

cents per private higher education and 69 cents per public higher education. What these numbers

mean are that after RM1 has been invested into private and public higher education respectively, the

household sector receives an additional induced income rise by 85 cents and 69 cents respectively in

terms of new income generated. Private higher education was thus found to have a stronger impact on

37

income generation compared to public higher education. The total income multipliers of private and

public higher education constituted about 1.97 and 1.94 of new income generated that resulted from

the additional Ringgits worth of final demand made by private and public higher education,

respectively. The difference between the simple and total income multiplier is due to the amount of

direct and indirect impacts that input-output model generates. The simple income multiplier has n

number of industry sector. The input-output model used to calculate the total income multiplier has

n+ 1 sectors in which the additional sector is the household sector selling its services as labour inputs

and buying from industries as consumption. Since the household sector is a large industry as reflected

by the relatively high input coefficients, the total multiplier appears large when compared to simple

multipliers.

Table 4.1: Income and employment multiplier of private and public higher education

Effect Multiplier Higher Education

Private Public

Income

Simple 0.8513 0.6940

Total 1.9706 1.9435

Type I 1.3361 1.3208

Type II 3.0929 3.0504

Employment

Simple 0.0001

Total 0.0001

Type I 1.01285

Type II 1.01287

While the simple and total income multipliers shows the additional or new income generated

as a result of RM 1 investments made in the reference industry, Type I and Type II multipliers show

by how much has incomes risen (in terms of magnitude) from such RM 1 investments. As shown in

Table 4.1, Type I and Type II income multipliers did not differ as much between private and public

higher education. Likewise, private higher education show higher values of Type I and Type II

income multipliers, suggesting the importance of secondary effects in generating income to the

economy. For every Ringgit change in initial income payment to the workers, the direct and indirect

income change of private higher education is multiplied by 1.34, whereas public higher education was

estimated to produce 1.32 times (i.e., 1 plus 0.34 and 1 plus 0.32 times) more than the initial level of

income after factoring the induced impacts that resulted from the initial RM 1 of investments in the

reference industry.

4.3.2 Analysis of employment multiplier effects

Employment multipliers are calculated very similarly to income multipliers and differ only in

terms of the coefficients used in the input-output model that represents labour content (in terms of

persons working) instead of wages. The simple employment multipliers thus show additional jobs

created that resulted from the initial RM 1 investments into higher education.

38

The numbers on Table 4.1 show employment created in the economy based only on

employment multipliers for higher education without breaking down between the private and public

sectors. Employment data that separate between private and public higher education were not

available. The simple employment multiplier for higher education appear ver