A Test of Leser's Model of Household Consumption Expenditure in Malaysia and Singapore 9789814377515

Citation preview

1

I

I....J

The Institute of Southeast Asian Studies

The Institute of Southeast Asian Studies was established as an autonomous organization in May 1968. It is a regional research centre for scholars and other specialists concerned with modern Southeast Asia. The Institute's research interest is focused on the many-faceted problems of development and modernization, and political and social change in Southeast Asia. The Institute is governed by a twenty-four member Board of Trustees on which are represented the National University of Singapore, appointees from the government, as well as representatives from a broad range of professional and civic organizations and groups. A ten-man Executive Committee oversees day-to-day operations; it is chaired by the Director, the Institute's chief academic and administrative officer. The responsibility for facts and opuuons expressed in this publication rests exclusively with the author and his interpretations do not necessarily reflect the views or the policy of the Institute or its supporters.

"Copyright subsists in this publication under the United Kingdom Copyright Act, 1911, and the Singapore Copyright Act (Cap. 187). No person shall reproduce a copy of this publication, or extracts therefrom, without the written permission of the Institute of Southeast Asian Studies, Singapore."

A TEST OF LESER'S MODEL OF HOUSEHOLD CONSUMPTION EXPENDITURE IN MALAYSIA AND SINGAPORE

by

Sritua Arief

Research Notes and Discussions Paper No. 23 Institute of Southeast Asian Studies

1980

CONTENTil,

LIST OF TABLES

INTRODUCTION

1

I:

TilE DATA

2

II:

TilE MODEL AND ESTIMATION PROCEDURE

10

III: THE EMPIRICAL RESULTS

13

IV: CONCLUSIONS

33

REFERENCES

34

LIST OF TABLES Table No. \

1

Parameter Estimates for Rural Malaysia (Model I)

14

2

Parameter Estimates for Urban Malaysia (Model I)

15

3

Parameter Estimates for Malaysia (Model I)

16

4

Parameter Estimates for Singapore (Model I}

17

5

Parameter Estimates for Rural Malaysia (Model II)

18

6

Parameter Estimates for Urban Malaysia (Model II)

19

7

Parameter Estimates for Malaysia (Model II)

20

8

Parameter Estimates for Singapore (Model II)

21

9

Mean Budget Shares for Rural Malaysia, Urban Malaysia and Malaysia (Urban and Rural)

24

10

Mean Budget Shares for Singapore

25

11

Expenditure and Size Elasticities for Rural Malaysia, Urban Malaysia and Malaysia (Urban and Rural)

26

12

Expenditure and Size Elasticities for Singapore

27

13

Expenditure Elasticities for Rural Malaysia Based on Model I and Model II

29

Expenditure Elasticities for Urban Malaysia Based on Model I and Model II

30

Expenditure Elasticities for Malaysia (Urban and Rural) Based on Model I and Model II

31

Expenditure Elasticities for Singapore Based on Model I and Model II

32

14

15 16

INTRODUCTION

Econometric studies of household consumption expenditure date back to Engel's pioneering work based on Belgian household budget data. 1 Later, the subject came to occupy an important place not only in the applied econometric literature but also in the field of development planning and public policy formulations. Estimates of expenditure elasticities have proved useful in economic planning models2 because planners often set up consumption growth as a goal and want to know the growth patterns of the different components of consumption expenditure. For the successful implementation of development programmes, it is necessary to have some idea about the changes in consumption that are likely to occur with the rising income level. 3 For the purpose of projecting consumption, ideally the time senes data should be used. In most developing countries, however, reliable time series data on consumption are not usually available. Moreover, in most cases, available time series data do not have detailed commodity classifications. In order to avoid these problems, most of the studies in developing countries use cross-section data in analysing consumption patterns. The present study which also uses cross-section data presents an investigation of the household consumption patterns in Malaysia4 and

1 E. Engel, "Die Productions- und Consumptionsverhi\t nisse des Konigreichs Sachsen," Bulletin de l'Institut International de Statistique 9 (1895). 2 See, inter alia, Hollis B. Chenery and W.J. Raduchel, "Substitution in Planning Model~' in Hollis B. Chenery et al., eds., Studies in Development Planning (Cambridge: Harvard University Press, 1971); Hollis B. Chenery and H. Uzawa, "Non-Linear Programming in Economic Development," in Kenneth J. Arrow et al., cds., Studies in Linear and NonLinear Programming (Stanford, California: Stanford University Press, 1958); L. Johansen, "Savings and Growth in Long-Term Programming Models. Numerical Examples with a NonLinear Objective Function," mimeo. (Oslo: Oslo Institute of Economics, 1964). 3 H.S. Houthakker, "An International Comparison of Household Expenditure Patterns Commemorating the Centenary of Engel's Law," Econometrica 28 (1957). 4 Malaysia here refers to West Malaysia or Peninsular Malaysia. As far as we are aware, no cross-section study of household consumption expenditure has been done for Malaysia.

- 2 -

Singapore. 5 The Malaysian study is broken down into rural and urban patterns since in dualistic economies the need to identify "rural-urban dualism" of consumption has been fully appreciated. 6 The model used in this study is the one advocated recently by Leser. 7 The time series counterpart of Leser's model, known as the "Almost Ideal Demand System" model, was developed by Deaton and Muellbauer. 8 Other models of the Engel function where consumption expenditure on a particular item is put as a dependent variable and aggregate household consumption expenditure is put as an independent variable, that is, linear, semi-log, log-inverse, semi-log inverse, double-log, hyperbolic and double-log inverse, are tested in a larger work which is now under preparation. The organization of the study is as follows. First, we give a description of data sources and the concepts and definitions used in the data base. Next, we outline the model adopted and the procedure used for its estimation. Following this, we present the empirical results of the parameter estimates of the model, together with the interpretations of these results. Finally, we offer the conclusions of the study and the policy implications of the findings.

1:

THE DATA

Malaysia The Household Expenditure Survey of 1973, published by the Department of Statistics, Malaysia, is the basis for this study. This survey provides information 5 A cross-section and time series studies of household consumption expenditure for Singapore based on 1972/73 household budget survey were conducted by Hank Lim et al. See Hank Lim et al., A Study of Consumer Demand Patterns in Singapore 1960-1976 (Singapore: Institute of Economics and Statistics, Nanyang University, 1979). 6 See, for example, A.C. Kelley et al., Dualistic Economic Development (Chicago and London: University of Chicago Press, 1972). 7 See C.E.V. Leser, "Income, Household Size and Price Changes 1953-1973," Oxford Bulletin of Economics and Statistics 38, No.1 (February 1976); 1-10. 8 Sec A.S. Deaton and J. Muellbauer, "An Almost Ideal Demand System" (paper presented at SSRC Econometrics Study Group, London School of Economics, June 1978).

- 3 -

on average consumption expenditure on various consumption items by household expenditure class and average household size for various households corresponding to various expenditure classes. These sets of data are useful for the model of household consumption expenditure behaviour formulated in the next section of this paper. Following Is the classification of household consumption expenditure class, according to average monthly household consumption expenditure in Malaysian dollars.

Expenditure CJass 1

2 3 4 5 6 7 8 9 10 11

Average Monthly Consumption Expenditure (in M$) 49 50 100 150 200 300 400 500 600 700 800

and below -- 99 -- 149 -- 199 -- 299 -- 399 -- 499 -- 599 -- 699 -- 799 and above

This expenditure class applies to rural and urban areas, as well as to Malaysia as a whole. For the purpose of our analysis, the following groups of consumption expenditure items are used:

Group Group Group Group Group Group Group Group Group Group Group

1:

Rice 2: Bread and Other Cereals 3: Meat 4: Fish 5: Milk, Cheese and Eggs 6: Oils and Fats 7: Fruits and Vegetables 8: Sugar 9: Coffee, Tea and Cocoa 10: Other Foods 11: Beverages and Tobacco

Group 12: Clothing and Footwear Group 13: Rent, Fuels and Power Group 14: Furniture, Furnishings and Household Equipment Group 15: Medical Care and Health Expenses Group 16: Transport and Communication Group 1 7: Recreation, Education and Services Group 18: Miscellaneous Goods and Services

- 4 -

Concepts and Definitions The concepts and definitions adopted in Household Expenditure Survey, 1973, with regard to households, household members, consumption expenditure, location of household and consumption expenditure groups, are given below.

Household and Household Members. A household is defined as a unit where a group of persons normally live together, pool their financial resources and have common eating arrangements. Nonfamily members such as servants are included as members of the household. Persons are considered as members of a household if they have stayed in that household for more than 15 days in the month in which the household was surveyed.

Consumption Expenditure. All consumption expenditure data recorded m the survey are those on the acquisition basis. Goods on hire-purchase arc considered to have been bought at the time when the hire-purchase contracts are signed; or where there are no contracts, when the goods are delivered. Consumption expenditure includes items purchased as well as those produced on own account and used in final consumption. Imputed rentals of owner occupied dwellings, consumption from own production or stocks, goods and services furnished to employees free of charge or at reduced rates, and goods received as wages in kind and as gifts are included in the estimation of household consumption expenditure. Purchase of goods and services by households from government bodies is also considered as consumption expenditure if there is a clear link between the payment and the acquisition of the services or goods and if the decisions to make such payments are voluntary. Items which are regarded as financial and capital transactions are not part of household consumption expenditure. The main items in this category are income taxes, disbursements in the nature of investments, gambling losses, cash grants and donations.

- 5 -

Location of Household. The designation of the place of residence as rural or urban in Peninsular Malaysia is based on population size and the occupations of the population. Urban areas have been defined as: (a)

towns having a population of more than 75,000 in 1957;

(b)

towns having a population between 10,000 and 75,000 m 1957;

(c)

towns having a population of at least 7,670 but under 10,000 m 1957;

(d)

areas which do not fall into any of the above categories but have urban characteristics and where the population dependence on nonagricultural occupations is likely to be more than 60%.

All other areas are considered rural.

Consumption Expenditure Groups. The eighteen consumption expenditure groups mentioned earlier are defined as follows: (1)

Rice: rice of various qualities and types;

(2)

Bread and Other Cereals: flour, biscuits, bread and bakery products, other cereals, and other cereal products;

(3)

Meat: fresh meat, frozen meat and processed meat;

(4)

Fish: shellfish, fresh fish, iced, chilled or frozen fish and processed fish;

(5)

Milk, Cheese and Eggs: fresh milk and cream, condensed milk, milk powder, cheese and other dairy products and eggs;

(6)

Oils and Fats: butter, prepared animal oils and fats, vegetable oils and prepared vegetable oils;

(7)

Fruits and Vegetables: fresh fruits, frozen fruits, berries, nuts, preserved fruits, fresh vegetables, preserved vegetables, potatoes and other tubers;

(8)

Sugar: various types of sugar;

(9)

Coffee, Tea and Cocoa: coffee, tea, spices, chocolate, sugar confectionery, jam, marmalade, honey and other sweet ingredients;

(10)

Other Foods: foods not elsewhere classified;

(11)

Beverages and Tobacco: mineral water and other soft drinks, beer, wines, spirits and liquors, cigarettes, cigars, tobacco and other tobacco products;

- 6 -

(12) Clothing and Footwear: shirts, dresses, blouses, suits, jackets, trousers, underwear, stocking, socks, bathing suits, hats, rain coats, fabrics, yarns, sewing thread and other sewing articles, tailoring, knitting, leather footwear, nonleather footwear and repairs for footwear; (13) Rent, Fuels and Power: imputed rent, rent of leased dwellings (including subsidized and free housing), water charges, electricity, gas, liquid fuel, and other fuels; {14} Furniture, Furnishings and Household Equipment: furniture, floor covering, sculptures, paintings and other art objects, household textiles, other furnishing, cooking appliances, room airconditioning units, washing machines, refrigerators and freezers, sewing machines, other electrical appliances, repairs, glassware, china ceramic tableware, other kitchen utensils, electric bulbs, tools, washing powder and other cleaning materials, laundry, dyeing, insurance of household property and domestic services; (15) Medical Care ahd Health Expenses: medical and pharmaceutical products, therapeutic appliances and equipment, medical services, service charges on accidents and health insurance; (16) Transport and Communication: cars, motorcycles, bicycles, tyres, tubes, parts and accessories, repairs, petrol, motor oil, grease, insurance premiums, other expenditure in operation of personal transport equipment, railway fares, ship fares, airline fares, bus fares, taxi fares, other expenditure on purchased transport, postal services, telephone and telegraph services; (17) Recreation, Education and Cultural Services: boats and other major durables, television sets, radios, record players, tape recorders, cassettes, musical instruments, photographic equipment, sports equipment, play equipment and toys, gramaphone records, films, other recreational goods (semidurables), parts and repairs, other recreational goods (durables), cinemas, theatres and other public entertainment, television and radio licenses, lotteries and other gambling, books, newspapers, magazines and periodicals, school and study fees, boarding fees, and other recreational, education and cultural services; (18) Miscellaneous Goods and Services: services of barbers, beauty shops, goods for personal care, jewellery, rings and precious stones, watches, other personal goods, writing and drawing equipment and supplies, expenditure on food and beverages away from home, packaged tours, expenditure abroad, financial services and other miscellaneous services.

Singapore The 1977/78 Hmsehold Expenditure Survey published by the Department of Statistics, Singapore, is used as the data base for this study. It contains information on household expenditure for various consumption items by expenditure

- 7 -

class. The classification of the expenditure class expressed in average monthly household consumption expenditure in Singapore dollars is as follows:

Expenditure Class

1 2

Average Monthly Consumption Expenditure (in S$) Less than 300 300 -- 399 400 -- 499 500 -- 599 600 -- 799 800 -- 999 1,000 -- 1 ,499 1,500 -- 1,999 2,000 and over

3

4 5 6

7 8 9

For each household expenditure class, data on household consumption expenditure for a number of expenditure items are given. For the purpose of our analysis, we reclassified the various household consumption expenditure items into the following expenditure groups: Group 1: Group

2:

Rice and Other Cereals Meat and Poultry

Group 3:

Fish

Group 4:

Dairy Products and Eggs

Group 5:

Vegetables and Vegetable Products

Group 6:

Fruits

Group 7:

Other Foods

Group 8:

Clothing

Group 9:

Housing

Group 10:

Transport and Communication

Group 11:

Miscellaneous Expenditure

Group 12:

Nonassignable Expenditure

The Household Expenditure Survey provides data on percentage of households of different household sizes for each expenditure class from which the average household size for various expenditure classes can be derived.

- 8 -

Concepts and Definitions The following arc the concepts and definitions used m Singapore's Household Expenditure Survey.

Household. A private household is defined as a person living alone or a group of two or more persons living together in the same house and sharing common food arrangements or other essentials for living. A person who lives with others but has separate food arrangements constitute a separate household.

Household Members. Household members include boarders who pay food and lodging, visitors or guests who stay with the household for at least one month, live-in domestic servants who are provided with food and members of family temporarily absent from the household. Lodgers who pay for room or living space only and not for food, visitors or guests who stay with the household for less than one month, domestic servants who do not live in or have their own food arrangements and members of the family who are away for more than one year are excluded.

Consumption Expenditure. Consumption expenditure refers to actual payments made for goods and services during the survey month, irrespective of whether the goods and services paid for have been delivered or consumed. The exceptions are: ( 1)

goods acquired on credit such as those from provlSlon shops, clubs, etc., in which case consumption expenditure refers to the value of the purchases as and when they are acquired;

(2)

goods obtained by households from their own shops or farms which are valued at current market prices at time of withdrawal;

(3)

the costs of vacation -- travel, purchases of consumer durables and major repairs and renovation to houses carried out during the past twelve months including the survey month for which one-twelfth is taken as the expenditure.

- 9 -

Definitions of Consumption Expenditure Groups. expenditure groups are further subdivided as follows:

The consumption

( 1)

Rice and Other Cereals: other cereals;

rice, flour, bread, noodles, biscuits, cakes, pastries and

(2)

Meat and Poultry: fresh pork, fresh beef, fresh mutton, fresh poultry, frozen poultry, processed meat products, canned meat and prepared meat;

(3)

Fish: fresh fish, dried and salted fish, canned fish, fresh and frozen sea products, dried and salted sea products and other fish preparations;

(4)

Dairy Products and Eggs:

(5)

Vegetables and Vegetable Products: fresh leafy vegetables, fresh fruit vegetables, fresh root vegetables, dried, salted and preserved vegetables, canned vegetables, beans, peas, nuts, soya bean and other vegetable products;

(6)

Fruits: fruits;

( 7)

Other Foods: cooking oils and fats, sugar, sweets, chocolate, ice-cream, syrups, sugar preserves, coffee, tea, cocoa, soft drinks, sauces, salt and other condiments, spices and other related preparations, cooked food bought and consumed at home, cooked food bought from hawker stalls, meals taken at restaurants and other food not elsewhere classified;

(8)

Clothing: men's outer clothing, men's underclothing, men's clothing accessories and hosiery, women's outer clothing, women's underclothing, women's clothing accessories and hosiery, boys' outer clothing, girls' outer clothing, infants' clothing, clothing materials, tailoring, haberdasheries, leather and travel goods, watches, men's footwear, women's footwear and children's footwear;

(9)

Housing: rented accommodation, owner-occupied accommodation, minor repairs and maintenance, water, electricity and gas, hiring charges, other fuel, furniture, floor coverings, household furnishings, household appliances and equipment, radio and television, kitchen appliances and utensils, crockery and cutlery and other household durables;

(10)

Transport and Communication: purchases of vehicles, repairs and maintenance of vehicles and other necessary costs of vehicles, public transport, other travel and transport, postage, telephone and telegram;

(11)

Miscellaneous Expenditure: educational tuition and other fees, school textbooks, stationery, newspapers, magazines and periodicals, medical treatment, dental treatment, proprietary medicines and supplies, cinema and other entertainment fees, radio and television licences, sports goods and toys, optical and photographic goods, beer and stout, wines and spirits, cigarettes, other tobacco products, domestic services, laundry and cleaning, soap and detergents, other nondurable household goods, food for animals and pets, seeds, plants and flowers, stampcollecting and other hobbies, holiday expenses and others.

milk, butter, cheese and eggs;

tropical fruits, nontropical fruits, canned fruits and dried and preserved

- 10 -

(12)

Nonassignable Expenditure: pocket money to children, personal expenses of members and other nonassignable expenditure.

II:

THE MODEL AND ESTIMATION PROCEDURE

Lesser's model of household consumption expenditure 1s presented m the following forms: (1)

w·t

(2)

wi

a i + 8 i log C + ll i (Model l) =

a i + 8 i log C + A. i log Z + Jl i (Model II)

where w i is the budget share of the household for the i-th item, C is the aggregate household consumption expenditure, Z is the average household size and ll i is the residual error term. ai , 8 i and A. i are the parameters to be estimated. A multivariate regressiOn model representing the above forms of the Engel function can be written as:

(3)

Y

=

xB +

Jl

where y IS a vector of n individual observations on the of the particular Engel function, X is the n X n matrix on the n independent variables. n being equal to 2 for 3 for the second model. 8 is the coefficient vector and disturbances.

dependent variable of n observations the first model and ll is a vector of n

It is assumed in this model that the disturbances have zero mean and constant variance and are independently distributed, which implies

If the

n individual observations were available, we could apply the Ordinary

- 11 -

Least Squares (OLS) method to (3) in order to obtain the best linear unbiased estimates of ai , B i and II. i . In Malaysia and Singapore, the household expenditure surveys do not provide us with such data. Because of this, we have to estimate these parameters of the Engel function using the grouped data. In the context of grouped data of household consumption expenditure, the original n observations are assumed to be grouped into K expenditure classes giving arithmetic means of different expenditures in each expenditure class. The regression model depicted by equation (3) in terms of grouped means can then be written as (5)

y

xs +

=

jj

where y = Gy, X = Gx, and U = G J.l , G being the Prais-Aitchison's groupmg matrix.9 This grouping matrix is such that fl - - - - - - - - - - - - - - - - - - 0 I 'I I I II II

I

(6) {GG ) -

I I I

''-,

"

I I II II

'-, " '-,

f2

I I I 'I

'-

'-, '-

I I I

I I I

'-, ",

'-,

0 -----------------~

where fl, f2, ... , fK are the relative frequencies. 1 by {GG f% we obtain (7)

y*

=

where y* = ( GG I

9 See

S-J.

I

I

fKJ

Multiplying equation (5)

x*B+ J.l*

r¥2 y '

x*

Prais and Aitchison, "The Grouping of Observations in Regression Analysis,"

Review of International and Statistical Institute 22, No. 1 {1954):

1-22.

- 12 -

This transformation of equation ( 5) is equivalent to multiplying each class mean expenditure by the square root of the corresponding relative frequency of households in the class. With this transformation, it can be seen that

(8}

Ell

*=D

and E fl * fl*

=

2

0

I.

The application of the ordinary least squares procedure to the transformed variables as given in equation ( 7) will result in

(9)

s = £x

A

'

(cc l

r x1

]

1

[ _I

x

(GG l

) _1 - ]

y

The grouped data giVen for different expenditure classes indicate the arithmetic means of the observations relating to the households in the group. The use of arithmetic means is inappropriate when the original variable is transformed to logarithm where, conceptually, geometric means of the variable are relevant. Since there is no widely accepted way of estimating these means from grouped observations, we use the arithmetic means within each expenditure class as proxy for the geometric means. The expenditure elasticity ei, and household size elasticity oi are given by

(10) ei

=

1 + _fu_ wi A.·z

w·z If ei < 1, the item i is considered a "necessity" and if ei > 1, the item i is considered a "luxury". If ei is close to unity, we consider the item as "semiluxury".

Since the mam concern here is the estimate of the elasticities which could be used for prediction, the possible multicollinearity of the explanatory variables in Model II should not present a serious problem.1°

10

See the reasoning on this matter in A. Koutsoyiannis, Theory of Econometrics (London:

Macmillan Press, 1977), p. 252.

- 13 -

Ill:

THE EMPIRICAL RESULTS

This section presents the findings of the application of the above models of household consumption expenditure to the Malaysian and Singapore data under the following subsections: (a)

Performance of the Models;

(b)

Estimates of Expenditure Elasticities and their Implications

Performance of the Models Tables 1, 2, 3 and 4 show the estimates of Model I for Malaysia (rural, urban and both rural and urban) and Singapore respectively. The results show that Model I is quite valid for Malaysia and Singapore as shown by the values of R 2 except for Milk, Cheese and Eggs, Fruits and Vegetables, Other Foods and Beverages and Tobacco in rural Malaysia, for Meat, Milk, Cheese and Eggs, Oils and Fats, Coffee, Tea and Cocoa, Other Foods and Beverages and Tobacco in urban Malaysia, and for Milk, Cheese and Eggs, Fruits and Vegetables, Other Foods and Beverages and Tobacco in Malaysia as a whole, and for Fruits and Other Foods in Singapore. The budget shares are, m general, quite responsive to the aggregate expenditure variable. Judging from the responsiveness of the budget shares to aggregate expenditure, Model I proves to be equally applicable in rural and urban areas of Malaysia. The coefficients of the aggregate expenditure are not significant (using greater than unity of t-values as the criterion) for Fruits and Vegetables and Other Foods in rural Malaysia, for Oils and Fats, Fruits and Vegetables, and Beverages and Tobacco in urban Malaysia, and for Other Foods in Malaysia as a whole. In Singapore, Model I proves to be valid for all items except Fruits and Other Foods. The coefficient of the aggregate expenditure is not significant only for Fruits. Looking at the performance of Model II, as shown m Tables 5, 6, 7 and 8, we obtain the following facts:

- 14 -

Model I Table 1: Parameter Estimates for Rural Malaysia ( t-values m parentheses)

Item

a. i

Bi

R2

Rice

0.2668 (0.87)

-0.0444 (-3.26)

0.7080

Bread and Other Cereals

0.0704 (7.88)

-0.0060 (-3.80)

0.6167

Meat

-0.0357 (-3.33)

0.0145 (7.69)

0.8681

Fish

0.1587 (17.09)

-0.0179 (-10.99)

0.9306

Milk, Cheese and Eggs

0.0024 (0.25)

0.0039 (2.40)

0.3921

Oils and Fats

0.0343 (8.82)

-0.0028 (-4.17)

0.6595

Fruits and Vegetables

0.0754 (4.01)

-0.0027 (-0.82)

0.0707

Sugar

0.0979 (15.87)

-0.0129 (-11.98)

0.9409

Coffee, Tea and Cocoa

0.0272 (7 .63)

-0.0029 (-4.58)

0.6998

Other Foods

0.0127 (3.35)

-0.0001 (-0.01)

0.0601

Beverages and Tobacco

0.0675 (7 .28)

-0.0046 (-2.83)

0.4722

Clothing and Footwear

-0.0374 (-2.25)

0.0168 (5.80)

0.7889

Rent, Fuel and Power

0.3910 (10.44)

-0.0445 (-6. 79)

0.8366

Furniture, Furnishings and Household Equipment

-0.0281 (-4.96)

0.0116 (11.75)

0.9388

Medical Care and Health Expenses

-0.0132 (-4.27)

0.0049 (8.97)

0.8995

Transport and Communication

-0.2111 (-3.91)

0.0567 (6.00)

0.8004

Recreation, Education and Cultural Services

-0.0939 (-5.51)

0.0259 (8.69)

0.8935

Miscellaneous Goods and Services

-0.0220 (-1.6 7)

0.0262 (11.39)

0.9351

- 15 -

Model I Table 2: Parameter Estimates for Urban Malaysia (t-values m parentheses)

Item

a;

Rice

0.1985 (19.49)

-0.0231 (-13.02)

0.9496

Bread and Other Cereals

0.0839 (15.52)

-0.0084 (-8.95)

0.8991

Meat

0.0328 (1.91)

0.0049 (1.64)

0.2318

Fish

0.1147 (13.69)

..0.0111 (-7 .59)

0.8650

Milk, Cheese and Eggs

0.0261 (2.61)

0.0005 (0.30)

0.0105

Oils and Fats

0.0160 (3. 73)

0.0294

Fruits and Vegetables

0.0982 (8.99)

Sug&·

0.0288 (6.63}

-0.0004 (..0.52) ..().0069 (-3.63) -0.0030 (-4.02)

Coffee, Tea and Cocoa

0.0154 (7 .39)

-0.0011 (-2.9 7)

0.4964

Other Foods

0.0145 (10.34) 0.0370 (1.94)

0.3583

8£

R2

0.5943 0.6425

Furniture, Furnishings and Household Equipment

..0.0470 (-8.56)

Medical Care and Health Expenses

-0.0065 (-1.3 7)

Transport and Communication

-0.1775 (-3.96)

-0.0005 (-2.24) -0.0001 (..0.04) 0.0135 (7 .26) -0.0342 (-5.8 7) 0.0139 (14.49) 0.0043 (5.29) 0.0455 (5.83)

Recreation, Education and Cultural Services

..0.0505 (-2.40)

0.0213 (5.82)

0.7905

0.2767 (10.19)

-0.0169 (-3.56)

0.5859

Beverages and Tobacco Clothing and Footwear Rent, Fuel and Power

Miscellaneous Goods and Services

..0.0349 (-3.28) 0.3864 (11.56)

0.0050 0.8544 0.7931 0.9589 0.7573 0.7911

- -------- -

---~~

- 16 -

Model I Table 3: Parameter Estimates for Malaysia (t-values m parentheses}

Item

ex·

'

Si

R2

Rice

0.4713 {25.38)

-0.0633 {-19.49)

0.9768

Bread and Other Cereals

0.0732 {9.10)

0.7077

Meat

-0.0232 {-1.57)

Fish

0.1390 (9.02)

-0.0066 (-4.66) 0.0129 (5.01) -0.0140 (-5.22)

Milk, Cheese and Eggs

0.0057 (0.56)

Oils and Fats

0.0335 {10.49)

Fruits and Vegetables

0.0773 (4. 76)

0.0035 (2.03) -0.0029 (-5.23) -0.0032 (-1.13)

Sugar

0.0895 {16.47)

-0.0120 (-12.67)

0.9469

Coffee, Tea and Cocoa

0.0246 (10.66) 0.0127 (4.72)

-0.0024 (-6.00) -0.0001 (-0.23)

0.8000

Beverages and Tobacco

0.0647 (6.68)

0.4228

Clothing and Footwear

-0.0272 (-1.94)

Rent, Fuel and Power

0.3425 (8.08)

-0.0043 (-2.56) 0.0138 (5.65) -0.0319 (-4.31)

Furniture, Furnishings and Household Equipment

-0.0306 (-7.79)

0.0117 (17.11)

0.9701

Medical Care and Health Expenses

-0.0150 (-4.23) -0.1716 (-3.80)

0.0053 (8.60) 0.0472 (5.99)

0.8917

-0.0961 (-5.50) 0.0187 (2.14)

0.0274 (8.97) 0.0213 (13.95)

0.8994

Other Foods

Transport and Communication Recreation, Education and Cultural Services Miscellaneous Goods and Services

0.7360 0.7518 0.3143 0.7527 0.1243

0.0060

0.7806 0.6741

0.7995

0.9558

- 17 -

Model I Table 4:

Parameter Estimates for Singapore (t-values m parentheses)

Item

a.·z

Bi

R2

Rice and Other Cereals

0.2309 (32.28)

-0.0251 (-23.34)

0.9873

Fish

0.1590 (29.03)

-0.0162 (-19.65)

0.9821

Dairy Products and Eggs

0.0912 (24.43)

-0.0097 (-17.33)

0.9772

Vegetables and Vegetable Products

0.1255 (35.79)

-0.0132 (-25.04)

0.9889

Fruits

0.0193 (2.96)

0.0004 (0.36)

0.0187

Meat and Poultry

0.1556 (9.91)

-0.0132 (-5.59)

0.8170

Other Foods

0.3179 (7.07)

-0.0170 (-2.51)

0.4747

Clothing

0.1501 (6.22)

0.0293 (8.08)

0.9032

Housing

0.4415 (14.66)

·0.0406 (-8.98)

0.9202

Transport and Commrmication

-0.2339 (-4.96)

0.0524 (7.40)

0.8868

Miscellaneous Expenditure

-0.2009 (-8.19)

0.0571 (15.51)

0.9717

Nonassignable Expenditure

0.0439 (5. 74)

-0.0042 (-3.67)

0.6582

- 18 -

Model II Table 5:

Parameter Estimates for Rural Malaysia (t-values m parentheses)

A.·t

Item

-0.0490 (-0.08)

0.8089

0.0356 (8.55)

0.9622

-0.0011 (-0.07)

0.8682 0.9856

-0.0282 (-5.06)

0.0347 (5.54) 0.0360 (6.82) 0.0148 (6.16) 0.0679 (4.82)

0.1043 (19. 70) 0.0314 (11.99) 0.0179 (9.91)

-0.0196 (-7.50)

0.0179 (2.69)

0.9690

-0.0073 (-5.61)

O.ol17 (3.58)

0.8848

-0.0055 (-6.12)

0.0146 (6.44)

0.8385

Beverages and Tobacco

0.0785 (11.74)

-0.0162 (-4.88)

0.8043

Clothing and Footwear

-0.0161 (-1.!>8) 0.3408 (17.11)

-0.0055 ( -1.08)

0.0308 (3.68) 0.0595 (4.68)

0.0080 (0.81)

-0.1403 (-5.63)

0.9671

Furniture, Furnishings and Household Equipment

-0.0266 (-4.02)

0.0101 (3.10)

0.0040 (0.48)

0.9405

Medical Care and Health Expenses

-0.0122 (-3.40)

0.0030 (0.67)

0.9049

Transport and Communication

-0.2793 (-8.15)

-0.1907 (-4.45)

0.9426

Recreation, Education and Cultural Services

-0.0909 (-4.53) -0.0384 (-4.44)

0.0037 (2.12) 0.1282 (7 .57) 0.0228 (2.30) 0.0433 (10.15)

0.0083 (0.33)

0.8949

-0.0458 (-4.23)

0.9800

Rice Bread and Other Cereals Meat Fish Milk, Cheese and Eggs Oils and Fats Fruits and Vegetables Sugar Coffee, Tea and Cocoa Other Foods

Rent, Fuel and Power

Miscellaneous Goods and Services

0.2525 (0.69) 0.0831 (25.01) -0.0361 (-2.84) 0.1711 (34.20) 0.0153 (3.62) 0.0396 (20.59) 0.0997 (8.84)

-0.0406 (-2.32) -0.0193 (-11.74) 0.0149 (2.37) -0.0309 (-12.49) -0.0096 (-4.58) -0.0084 (-8.83)

0.9108 0.9407 0.7620

0.9436

- 19 -

Model II

Table 6:

Parameter Estimates for Urban Malaysia (t-values m parentheses)

Item

Cl£

S£

I. t.

Rice

0.2331 (21.27)

-0.0380 (-9.51)

0.0317 (3.87)

0.9825

Bread and Other Cereals

0.1023 (17.71)

-0.0164 (-7.77)

0.0169 (3.91)

0.9654

Meat

0.0945 (5.85)

-0.0216 (-3.66)

0.0564 (4.69)

0.7951

Fish

0.1476 (26.94) 0.0647 (9.39)

-0.0252 (-12.64) -0.0161 (-6.41)

0.0301 (7.3 7) 0.0354 (6.89)

0.9826

0.0331 (14.14) 0.1203 (6.87) 0.0427 (8.39) 0.0204 (6.50) 0.0142 (5.55) 0.0851 (3.05)

-0.0078 (-9.09) -0.0164 (-2.57) -0.0090 (-4.87)

0.0157 (8.99)

0.9126

0.0202 (1.54) 0.0128 (3.36) 0.0045 (1.94) -0.0003 (-0.15)

0.6880

Milk, Cheese and Eggs Oils and Fats Fruits and Vegetables Sugar Coffee, Tea and Cocoa Other Foods

-0.0032 (-2.81) -0.0004 (-0.44)

R2

0.8575

0.8522 0.6581 0.3602

0.0440 (2.11)

0.3593

0.0135 (1.98)

0.8702

-0.1152 (-5. 71)

0.9593

Clothing and Footwear

-0.0201 (-1.09)

Rent, Fuel and Power

0.2605 (9.62)

-0.0208 (-2.04) 0.0071 {1.06) 0.0199 (2.02)

Furniture, Furnishings and Household Equipment

-0.0665 (-12.41)

0.0223 (11.40)

-0.0179 (-4.4 7)

0.9882

Medical Care and Health Expenses

0.0068 (1.07)

-0.0014 (-0.60)

0.0122 (2.58)

0.8675

Transport and Communication

-0.3396 (-8.17)

0.1152 (7.61)

-0.1482 ( -4. 79)

0.9460

Recreation, Education and Cultural Services

-0.0365 (-0.96)

0.0153 (1.11)

0.0127 (0.45)

0.7957

Miscellaneous Goods and Services

0.2447 (5.14)

-0.0031 (-0.17)

-0.0293 (-0.82)

0.6184

Beverages and Tobacco

- 20 -

Model II Table 7:

Parameter Estimates for Malaysia (t-values m parentheses)

Item

eli

8·z

>.·z

Rice

0.4739 (21.3 7)

-0.0659 (-6.01)

0.0071 (0.24)

0.9770

Bread and Other Cereals

0.0856 (38.78)

-0.0190 (-17 .45)

0.0340 (12.03)

0.9846

Meat

-0,0126 (-0.80)

0.0022 (0.28)

0.0292 (1.45)

0.7911

Fish

0.1495 (8.98)

-0.0246 (-2.99)

0.0289 (1.35)

0.7981

Milk, Cheese and Eggs

0.0203 (4.58) 0.0382 (28. 73)

Fruits and Vegetables

0.1010 (13.86) 0.0927 (15.34) 0.0277 (20.15)

0.0403 (7.09) 0.0130 (7.62) 0.0653 (6.99) 0.0088 (1.13) 0.0087 (4.92)

0.9058

Oils and Fats

0.0166 (12.02) 0.0764 (10.37) -0.0088 (-0.95)

-0.0112 (-5.11) -0.0077 (-11.68) -0.0271 (-7.53) -0.0153 (-5.11) -0.0056 (-8.22) -0.0040 (-5.85) -0.0162 (-4.45) -0.0047 (-1.03)

0.0106 (5.99) 0.0324 (3.43) 0.0507 (4.28)

0.8190

Rent, Fuel and Power

0.2800 (15.40)

0.0310 (3.44)

-0.1717 (-7 .3 7)

0.9582

Furniture, Furnishings and Household Equipment

-0.0336 (-8.16)

0.0148 (7 .25)

0.9771

Medical Care and Health Expenses

-0.0124 (-3.29)

0.0028 (1.47)

-0.0082 (-1.56) 0.0071 (1.45)

Transport and Communication

-0.2316 (-7.93)

0.1077 (7.46)

-0.1650 (-4.41)

0.9415

Recreation, Education and Cultural Services

-0.0920 (-4.44)

0.0232 (2.27)

0.0114 (0.42)

0.9017

Miscellaneous Goods and Services

0.0173 (1.65)

0.0227 (4.41)

-0.0039 (-0.29)

0.9562

Sugar Coffee, Tea and Cocoa Other Foods Beverages and Tobacco Clothing and Footwear

R2

0.9701 0.8770 0.9543 0.9503

0.7669 0.9335

0.9143

- 21 -

Model II Table 8:

Parameter Estimates for Singapore (t-values m parentheses)

S.I

A

I

Rice and Other Cereals

0.2430 (42.82)

-0.0308 (-17.16}

0.0153 (3.45}

0.9957

Fish

0.1688 (44.58)

-0.0208 (-17.38)

0.0124 (4.19)

0.9954

Dairy Products and Eggs

0.0924 (18.24)

-0.0103 (-6.41)

0.0015 (0.37)

0.9777

Vegetables and Vegetable Products

0.1322 {69. 72)

-0.0164 (-27.28)

0.0085 (5. 71)

0.9982

Fruits

0.0310 (6.99)

-0.0052 (-3.71)

0.0148 (4.28)

0.7581

Meat and Poultry

0.1848 {19.38)

-0.0271 (-8.97)

0.0369 (4.96)

0.9641

Other Foods

0.3959 (11.71)

-0.0541 (-5.05)

0.0986 (3. 73)

0.8422

Clothing

-0.0132 (4.27)

0.0021 (2.11)

0.0023 (1.15)

0.9159

Housing

0.0381 (33.85)

-0.0012 (-3.3 7)

-0.0076 (-8.64)

0.9940

Transport and Communication

-0.0314 (-8.40)

0.0090 (7 .65)

-0.0101 (-3.46)

0.9623

Miscellaneous Expenditure

-0.0236 (-9.57)

0.0074 (9.45)

-0.0044 (-2.28)

0.9848

Nonassignable Expenditure

0.0052 (5.66)

-0.0008 (-2. 74)

0.0010 (1.38)

0.7409

Item

a.

R2

- 22 -

( 1) The budget shares, in general, are also quite responsive to the

aggregate expenditure variable. Again using the criterion of greater than unity of t-values, we observe that the coefficients of the aggregate expenditure are not statistically significant for Rent, Fuel and Power in rural Malaysia, for Other Foods, Medical Care and Health Expenses and Miscellaneous Goods and Services in urban Malaysia, and for Meat in Malaysia as a whole. In Singapore, the coefficients of the aggregate expenditure are significant for all items; (2) The coefficients of the size variable are in general statistically significant, indicating the responsiveness of the budget shares to household sizes. The coefficients of the size variable are not significant only for Rice, Meat, Furniture, Furnishings and Household Equipment, Medical Care and Health Expenses and Recreation, Education and Cultural Services, in rural Malaysia; for Rent, Fuel and Power, Recreation, Education and Cultural Services and Miscellaneous Goods and Services in urban Malaysia; and for Rice, Recreation, Education and Cultural Services and Miscellaneous Goods and Services in Malaysia as a whole. In Singapore, the budget shares are not responsive to household sizes for Vegetables and Vegetable Products and Clothing. For rural Malaysia, the coefficient of household size is negative and significant for Rent, Fuel and Power, Transport and Communication and Miscellaneous Goods and Services while for urban areas of Malaysia, the coefficient of household size is negative and significant for Rent, Fuel and Power, Furniture, Furnishings and Household Equipment and Transport and Communication. For Malaysia as a whole, the coefficient of household size is negative and significant for Rent, Fuel and Power, Furniture, Furnishings and Household Equipment and Transport and Communication whereas for Singapore, this coefficient is negative and significant for Housing, Transport and Communication and Miscellaneous Expenditure. The negative coefficient of household size for a particular item implies that the larger the household size, the smaller is the per capita consumption of this item. We notice the similarities in household behaviour for Rent, Fuel and Power and Transport and Communication in rural and urban Malaysia as the household size increases. The positive coefficient of household size for a particular item implies that the larger the household size, the larger the per capita consumption of this item. The coefficient of household size is posttlve and significant for Bread and Other Cereals, Fish, Milk, Cheese and Eggs, Oils and Fats, Fruits and Vegetables, Sugar, Coffee, Tea and Cocoa, Other Foods, Beverages and Tobacco and Clothing and Footwear in rural Malaysia, whereas in urban

- 23 -

Malaysia this coefficient is positive and significant for Rice, Bread and Other Cereals, Meat, Fish, Milk, Cheese and Eggs, Oils and Fats, Fruits and Vegetables, Sugar, Coffee, Tea and Cocoa, Beverages and Tobacco, Clothing and Footwear and Medical Care and Health Expenses. It is interesting to note here that m urban Malaysia, a significant positive effect on per capita consumption is found for Rice, Meat and Medical Care and Health Expenses compared to rural areas of Malaysia where no such effects are found on per capita consumption of these items as the household size mcreases. A possible explanation for this fact is that in large urban households, more household members have an opportunity for employment and, therefore, with the extra income, the higher income families with larger family sizes are purchasing larger quantities of higher quality rice, purchasing more meat and incurring more medical expenses. Although the negative coefficient of household size for Rice is not significant enough for rural Malaysia, this fact should not make us contented, bearing in mind the possible existence of malnutrition among the rural people of Malaysia. For Malaysia as a whole, the significant positive effect on per capita consumption as household size increases is found for Bread and Cereals, Meat, Fish, Milk, Cheese and Eggs, Oils and Fats, Fruits and Vegetables, Sugar, Coffee, Tea and Cocoa, Other Foods, Beverages and Tobacco, Clothing and Footwear and Medical Care and Health Expenses. In Singapore, significant positive effects on per capita consumption as the household size increases are found for Rice and Other Cereals, Fish, Vegetables and Vegetable Products, Fruits, Meat and Poultry, Other Foods, Clothing_ and Nonassignable Expenditure. (3) Judging from the values of R2, Model II proves to be valid for all items except Rice in rural Malaysia. This model performs poorly for Other Foods and Beverages and Tobacco in urban Malaysia. In Malaysia as a whole, as well as in Singapore, Model II is found to be valid for all items. The findings show that Model II in general is superior to Model I in describing the household expenditure behaviour in Malaysia and Singapore.

The Elasticity Estimates and Their Implications The expenditure and size elasticities calculated at the mean budget shares (see Tables 9 and 10 for the mean budget shares) using the formulae

- 24 -

Table 9: Mean Budget Shares for Rural Malaysia, Urban Malaysia and Malaysia (Urban and Rural)

Rural Malaysia

Urban Malaysia

Rice

0.1314

0.0680

0.1152

Item

Malaysia (Urban and Rwal)

Bread and Other Cereals

0.0370

0.0363

0.0363

Meat

0.0455

0.0606

0.0494

Fish

0.0583

0.0521

0.0572

Milk, Cheese and Eggs

0.0244

0.0291

0.0256

Oils and Fats

0.0183

0.0138

0.0170

Fruits and Vegetables

0.0551

0.0592

0.0592

Sugar

0.0252

0.0116

0.0217

Coffee, Tea and Cocoa

0.0111

0.0093

0.0110

Other Foods

0.0127

0.0114

0.0121

Beverages and Tobacco

0.0417

0.0362

0.0402

Clothing and Footwear

0.0572

0.0411

0.0506

Rent, Fuel and Power

0.1410

0.193.3

0.1628

Furniture, Furnishings and Household Equipment

0.0373

0.0313

0.0.355

Medical Care and Health Expenses

0.0141

0.0180

0.0150

Transport and Communication

0.1074

0.0796

0.0943

Recreation, Education and Cultural Services

0.0518

0.0697

0.0580

Miscellaneous Goods and Services

0.1250

0.1815

0.1387

represented by equations (9) and (10) are presented in Tables 11 and 12. In rural Malaysia, Meat, Rent, Fuel and Power, Furniture, Furnishings and Household Equipment, Medical Care and Health Expenses, Transport and Communication, Recreation, Education and Cultural Services and Miscellaneous Goods and Services are considered as luxuries whereas in urban Malaysia, Clothing and Footwear, Rent, Fuel and Power, Furniture, Furnishings and Household

- 25 -

Table 10:

Mean Budget Shares for Singapore

Item Rice and Other Cereals

Mean Budget Share 0.0650

Fish

0.0520

Dairy Products and Eggs

0.0269

Vegetables and Vegetable Products

0.0382

Fruits

0.0216

Meat and Poultry

0.0683

Other Foods

0.2056

Clothing

0.0438

Housing

0.1726

Transport and Communication

0.1130

Miscellaneous Expenditure

0.1770

Nonassignable Expenditure

0.0160

Equipment, Transport and Communication and Recreation, Education and Cultural Services are luxuries. The luxury items for Malaysia as a whole are Meat, Rent, Fuel and Power, Furniture, Furnishings and Household Equipment, Medical Care and Health Expenses, Transport and Communication, Recreation, Education and Cultural Services and Miscellaneous Goods and Services. In Singapore, Clothing, Transport and Communication and Miscellaneous Expenditure are considered to be luxury items in household budgets. Size elasticities are found to be negative for Rice, Meat, Rent, Fuel and Power, Transport and Communication and Miscellaneous Goods and Services in rural Malaysia, for Other Foods, Rent, Fuel and Power, Furniture, Furnishings and Household Equipment, Transport and Communication and Miscellaneous Goods and Services in urban Malaysia and for Rent, Fuel and Power, Fumitu~ez Furnishings and Household Equipment, Transport and Communication and Miscellaneous Goods and Services in Malaysia as a whole. The n~gative values of size elasticities for these items imply that they exhibit diseconomies of scale and the positive values of size elasticities for the rest of the items indicate the presence of economies of scale in the consumption of these items as income and household size increase. In Singapore, Housing, Transport and

Table 11:

Ex~enditure

and Size Elasticities for Rural Malaysia, Urban Malaysia and Malaysia {Urban and Rural)

Urban Malaysia

Rural Malaysia

Malaysia (Urban and Rural)

Item

Expenditure Elasticity

Size Elasticity

Expenditure Elasticity

Size Elasticity

Expenditure Elasticity

Size Elasticity

Rice

0.6910

-0.3044

0.4412

0.4662

0.4280

0.0616

Bread and Other Cereals

0.4784

0.9622

0.5482

0.4656

0.4766

0.9366

1.0445

0.5911

Meat

1.3275

-0.0242

0.6436

0.9307

Fish

0.4700

0.5952

0.5163

0.5777

0.5699

0.5052

Milk, Cheese and Eggs

0.6066

1.4754

0.4467

1.2165

0.5625

1.5742

Oils and Fats

0.5410

0.8087

0.4348

1.1377

0.5471

0.7647

Fruits and Vegetables

0.4882

1.2323

0.7230

0.3412

0.5422

1.1030

Sugar

0.2222

0.710-3

0.2241

1.1034

0.2949

0.4055

Coffee, Tea and Cocoa

0.3423

1.0577

0.6559

0.4839

0.4909

0.7909

Other Foods

0.5669

1.1496

0.9649

-0.0263

0.6694

0.8760

Beverages and Tobacco

0.6115

0.7386

0.4254

1.2155

0.5970

0.8060

Clothing and Footwear

0.9038

1.0402

1.1727

0.3285

0.9071

1.0020

Rent, Fuel and Power

1.0567

-0.9950

1.1029

-0.5960

1.1904

-1.0547

Furniture, Furnishings and Household Equipment

1.2708

0.1072

1.7125

-0.5719

1.4169

-0.2310

Medical Care and Health Services

1.2624

0.2128

0.9222

0.6778

1.1867

0.4733

Transport and Communication 2.1937

-1.7756

2.4472

-1.8618

2.1421

-1.7497

Recreation, Education and Cultural Services

1.4402

0.1602

1.2195

0.1822

1.4000

0.1966

Miscellaneous Goods and Services

1.3464

-0.3664

0.9829

-0.1614

1.1637

-0.0281

!\:)

0\

- 27 -

Table 12:

Expenditure and Size Elasticities for Singapore

Expenditure

Size

Elasticity

Elasticity

Rice and Other Cereals

0.5262

0.2354

Item

Fish

0.6000

0.2385

Dairy Products and Eggs

0.6171

0.0558

Vegetables and Vegetable Products

0.5767

0.2225

Fruits

0.7593

0.6852

Meat and Poultry

0.6032

0.5403

Other Foods

0.7369

0.4796

Clothing

1.0479

0.0525

Housing

0.9930

-0.0440

Transport and Communication

1.0796

-0.0894

Miscellaneous Expenditure

1.0418

-0.0249

Nonassignable Expenditure

0.9500

0.0625

Communication, and Miscellaneous Expenditure exhibit some degrees of diseconomies of scale. The findings show some urban-rural differentials in household behaviour in Malaysia. The differences are quite obvious (as shown by substantial differences in expenditure elasticities) in the household behaviour on Meat, Fruits and Vegetables, Coffee, Tea and Cocoa, Other Foods, Beverages and Tobacco, Clothing and Footwear, Furniture, Furnishings and Household Equipment, Medical Care and Health Expenses and Miscellaneous Goods and Services. The relatively higher expenditure elasticities for Meat, Beverages and Tobacco, Medical Care and Health Expenses and Miscellaneous Goods and Services in rural, as compared to urban, Malaysia imply that with increase in income, the rural household will have a higher demand for these items than their urban counterparts. The higher expenditure elasticities for Fruits and Vegetables, Coffee, Tea and Cocoa, Other Foods, Clothing and Footwear, Furniture, Furnishings and Household Equipment in urban areas of Malaysia compared to those in rural areas imply that the

- 28 -

households in urban Malaysia will demand more of these items when their incomes rise, compared to the rural households. It might be useful to present here the sensitivity . of the expenditure elasticity estimates to the model specification. The comparisons of expenditure elasticity estimates based on Model I and Model II are given in Tables 13, 14, 15 and 16 for rural Malaysia, urban Malaysia, Malaysia as a whole and Singapore respectively.

In all areas, the elasticity estimates are quite sensitive to model specification. The sensitivity is particularly high for Bread and Other Cereals, Milk, Cheese and ~ Oils and Fats, Fruits and Vegetables, Sugar, Coffee, Tea and Cocoa, Other Foods, Rent, Fuels and Power and Transport and Communication in rural Malaysia. In urban areas of Malaysia, the sensitivity is particularly high for Meat, Milk, Cheese and Eggs, Oils and Fats, Sugar, Beverages and Tobacco, Rent, Fuels and Power, Furniture, Furnishings and Household Equipment and Transport and Communication. In Malaysia as a whole, the sensitivity is relatively high for Bread and Other Cereals, Milk, Cheese and Eggs, Fruits and Vegetables, Other Foods, Beverages and Tobacco, Clothing and Footwear, Rent, Fuels and Power and Transport and Communication. A relatively high sensitivity of expenditure elasticity to model specification is found for Fruits, Clothing, Housing, Transport and Communication and Miscellaneous Expenditure in Singapore. It is important to appreciate the degree of specification sensitivity registered by the expenditure elasticity estimates. Given the importance of expenditure elasticity in demand projections and planning models, it is imperative to use expenditure elasticity estimates as accurately as possible.

The adoption of Model II leads to the treatment of Milk, Cheese and Eggs and Clothing and Footwear as "necessities" when they should really be regarded as "luxuries" and the treatment of Rent, Fuels and Power as "luxuries" when they should be regarded as "necessities" in rural Malaysia if Model I is used. Model II also leads to the treatment of Meat, Milk, Cheese and Eggs and Medical Care and Health Expenses as "necessities" when they should really be considered as "luxuries" and the treatment of Rent, Fuels and Power as "luxuries" when they should really be regarded as "necessities" in urban areas of Malaysia if Model I is adopted. In Malaysia as a whole, Milk, Cheese and Eggs, Clothing and Footwear should have been treated as "luxuries" rather than as "necessities" and Rent, Fuel

- 29 -

Table 13:

Expenditure Elasticities for Rural Malaysia Based on Model I and Model II

Item

Expenditure Elasticity Model I

Expenditure Elasticity Model II

Rice

0.6621

0.6910

Bread and Other Cereals

0.8378

0.4784

Meat

1.3187

1.3275

Fish

0.6930

0.4700

Milk, Cheese and Eggs

1.1598

0.6066

Oils and Fats

0.8470

0.5410

Fruits and Vegetables

0.9510

0.4882

Sugar

0.4881

0.2222

Coffee, Tea and Cocoa

0.7387

0.3423

Other Foods

0.9921

0.5669

Beverages and Tobacco

0.8897

0.6115

Clothing and Footwear

1.2937

0.9038

Rent, Fuel and Power

0.6844

1.0567

Furniture, Furnishings and Household Equipment

1.3110

1.2708

Medical Care and Health Equipment

1.3475

1.2624

Transport and Communication

1.5279

2.1937

Recreation, Education and Cultural Services

1.5000

1.4402

Miscellaneous Goods and Services

1.2096

1.3464

- 30 -

Table 14:

Expenditure Elasticities for Urban Malaysia Based on Model I and Model II

Item

Expenditure Elasticity Model I

Expenditure Elasticity Model II

Rice

0.6603

0.4412

Bread and Other Cereals

0.7686

0.5482

Meat

1.0809

0.6436

Fish

0.7869

0.5163

Milk, Cheese and Eggs

1.0172

0.4467

Oils and Fats

0.9710

0.4348

Fruits and Vegetables

0.8834

0.7230

Sugar

0.7414

0.2241

Coffee, Tea and Cocoa

0.8817

0.6559

Other Foods

0.9561

0.9649

Beverages and Tobacco

0.9972

0.4254

Clothing and Footwear

1.3285

1.1727

Rent, Fuel and Power

0.8231

1.1029

Furniture, Furnishings and Household Equipment

1.4441

1.7125

Medical Care and Health Expenses

1.2389

0.9222

Transport and Communication

1.5716

2.4472

Recreation, Education and Cultural Services

1.3056

1.2195

Miscellaneous Goods and Services

0.9069

0.9829

- 31 -

Table 15:

Expenditure Elasticities for Malaysia (Urban and Rural) Based on Model I and Model II

Item

Expenditure Elasticity Model I

Expenditure Elasticity Model II

Rice

0.4505

0.4280

Bread and Other Cereals

0.8182

0.4766

Meat

1.2611

1.0445

Fish

0.7552

0.5699

Milk, Cheese and Eggs

1.1367

0.5625

Oils and Fats

0.8294

0.5471

Fruits and Vegetables

0.9459

0.5422

Sugar

0.4470

0.2949

Coffee, Tea and Cocoa

0.7818

0.4909

Other Foods

0.9917

0.6694

Beverages and Tobacco

0.8930

0.5970

Clothing and Footwear

1.2727

0.9071

Rent, Fuel and Power

0.8041

1.1904

Furniture, Furnishings and Household Equipment

1.3296

1.4169

Medical Care and Health Expenses

1.3533

1.1867

Transport and Communication

1.5005

2.1421

Recreation, Education and Cultural Services

1.4724

1.4000

Miscellaneous Goods and Services

1.1536

1.1637

- 32 -

Table 16:

Expenditure Elasticities for Singapore Based on Model I and Model II

Expenditure Elasticity Model I

Expenditure Elasticity Model II

Rice and Other Cereals

0.6138

0.5262

Fish

0.6885

0.6000

Dairy Products and Eggs

0.6394

0.6171

Vegetables and Vegetable Products

0.6545

0.5707

Fruits

1.0185

0.7593

Meat and Poultry

0.8067

0.6032

Other Foods

0.9173

0.7369

Clothing

1.6689

1.0479

Housing

0.7648

0.9930

Transport and Communication

1.4637

1.0796

Miscellaneous Expenditure

1.3226

1.0418

Nonassignable Expenditure

0.7375

0.9500

Item

and Power as "necessities" rather than as "luxuries" if Model I is used. In Singapore, Fruits should have been treated as "semiluxuries" rather than as "necessities", Housing as "semiluxuries" rather than "necessities", and Clothing, Transport and Communication and Miscellaneous Expenditure as "luxuries" rather than as "semiluxuries" if Model I is adopted to describe the household expenditure behaviour.

- 33 -

IV:

CONCLUSIONS

Our study finds that the model incorporating household size proves to be better than the one without it in describing household consumption expenditure behaviour in Malaysia and Singapore. The recorded sensitivity to model specification of estimated expenditure elasticities for various consumption items indicates the importance of the household size variable m determining the more accurate magnitude of expenditure elasticities. The relatively higher expenditure ela~ticity for Rent, Fuel and Power and Medical Care and Health Expenses in Malaysia, which classifies them as "luxuries", should be taken as a warning to public policy makers in this country, bearing in mind that these two components should have fallen under basic necessities. The higher expenditure for Rice in rural, compared to urban, Malaysia, which implies that rural households will demand more Rice as their incomes increase, may be taken as an indicator of possible inadequate levels of rice consumption among the rural poor. The findings of the study which indicate elements of rural-urban dualism with regard to household consumption expenditure in Malaysia may be useful for development policy making in Malaysia. This is especially important in view of the growing need to apply a "basic needs strategy" directed mainly at the poor segments of the population. The expenditure elasticities for Clothing and Housil}g_ in Singapore which are around unity may indicate that these items are to a certain extent difficult for the low income group to consume adequately. These findings may have implications for policies with regard to subsidy and direct as well as indirect taxes. Some of the obvious limitations of this study include the ignoring of household composition, the assumption that all households face the same set of prices at a point in time, and the carrying out of the entire exercise within the framework of Leser's model. However, such issues are best left to future research workers.

. 34 .

REFERENCES

Chenery, H.B. and H. Uzawa. "Non-Linear Programming in Economic Development." In Kenneth J. Arrow et al., eds., Studies in Linear and Non-Linear Programming. Stanford: Stanford University Press, 1958. Chenery, H.B. and W.J. Raduchel. "Substitution in Planning Models." In H.B. Chenery et al., eds., Studies in Development Planning. Cambridge: Harvard University Press, 19 71. Cramer, J.S. "Efficient Grouping Regression and Correlation in Engel Curve Analysis." journal of the American Statistical Association 59 (1964): 233-50. Deaton, A.S. and J. Muellbauer. "An Almost Ideal Demand System." Paper presented at SSRC Econometrics Study Group, London School of Economics, 1978. Engel, E. "Die Productions- und Consumptionsverh'al.t mss des K'onigreichs Sachsen." Bulletin de L'Institut International de Statistique 9 (1895). Houthakker, H.S. "An International Comparison of Household Expenditure Patterns Commemorating the Centenary of Engel's Law." Econometrica 28 (1957): 532-51. Johansen, L. "Savings and Growth in Long-Term Programming Models. Numerical Examples with a Non-Linear Objective Function." Mimeo. Oslo: Oslo Institute of Economics, 1964. Kelley, A.C. et al. Dualistic Economic Development. University of Chicago Press, 1972. Koutsoyiannis, A.

Theory of Econometrics.

Leser, C.E.V. "Forms of Engel Functions." 694-703.

London:

Chicago and London:

Macmillan Press, 1977.

Econometrica 31 (1963):

- 35 -

Leser, C.E.V. "Income, Household Size and Price Changes, 1953-1973.'' Oxford Bulletin of Economics and Statistics 38 (1976): 1-10. Lim, Hank, et al. A Study of Consumer Demand Patterns in Singapore, 19601976. Singapore: Institute of Economics and Statistics, Nanyang University, 19 79. Malaysia.

Department of Statistics.

Household Budget Survey, 1973.

1978.

Prais, S.J. and J. Aitchison. "The Grouping of Observations in Regression Analysis.'' Review of the International Statistical Institute 22, No. 1 (1954): 1-22. Singapore. Department of Statistics. 1979.

Household Expenditure Survey 1977/78.

INSTITUTE OF SOU111EAST ASIAN STUDIES LIST OF PUBLICATIONS IN THE RESEARCH NOTES AND DISCUSSIONS PAPERS SERIES 1

M. Mainguy, Economic Problems Related to Oil and Gas Exploration, 1976. 39pp. S$6.00 (Out of print)

2

R. William Liddle, Cultural and Class Politics in New Order Indonesia, 1977. 21pp. S$4.00 (Out of print)

3

Raja Segaran Arumugam, State and Oil in Burma, 1977. (Out of print)

4

Hilman Adil, Australia's Policy Towards Indonesia During Confrontation, 1962-66, 1977. 90pp. S$8.00 (Out of print)

5

Albert D. Moscotti, Burma's Constitution and Elections of 1974: Book, 1977. 184pp. S$13.00 (Out of print)

6

Thamsook Numnonda, Thailand and the japanese Presence, 1941-45, 1977. 142pp. S$13.00 (Out of print)

7

Nguyen The Anh, The Withering Days of the Nguyen Dynasty, 1978. 33pp. S$4.00

8

M. Rajaretnam, Thailand's Kra Canal: (Out of print)

9

R.O. Whyte and Pauline Whyte, Rural Asian Women: Environment, 1978. 34pp. S$4.00

10

Ismail Kassim, The Politics of Accommodation: Malaysian General Election, 1978. 110pp.

11

Leo Suryadinata, The "Overseas Chinese" in Southeast Asia and China's Foreign Policy: An Interpretatt've Essay, 1978. 45pp. S$5.00 (Out of print)

12

Y. Mansoor Marican, Publt'c Personnel Administration in Malaysia, 1979. 21 pp. S$4. 00

13

Norbert Hofmann, A Survey of Tourism in West Malaysia and Some SocioEconomic Implz'cations, 1979. 48pp. S$5.00

Some Issues, 1978.

36pp.

82pp.

S$5.00

A Source

S$8.00

Status and

An Analysis of the 1978 S$11.00 (Out of print)

14

B.A. Hamzah, Oil and Economic Issues in Brunei, 1980.

15

Lee Yong Leng, The Razor's Edge: Boundaries and Boundary Disputes in Southeast Asia, 1980. 29pp. S$7.50

16

Anton van Naerssen, Location Factors and Linkages at the Industrial Estates of Maiacca Town: Implications for a Regional Development PoHcy in Peninsular Malaysia, 1980. 3~pp. S$7.50

17

Pradumna B. Rana, Exchange Rate Risk Under GeneraHzed Floating: Asian Countries, 1980. 20pp. S$5.00

18

Parsudi Suparlan and Hananto Sigit, Culture and Fertility: Indonesia, 1980. 41pp. S$7.50

19

Nor Laily Aziz et al., Culture and Fertilz'ty: 92pp. S$12.50

20

Amelia B. Alfonso, Leda L. Layo and Rodolfo A. Bulatao, Culture and Fertility: The Case of the Philippines, 1980. 67pp. S$9.50

21

Chang Chen-Tung, Ong Jin Hui and Peter S.J. Chen, Culture and Fertility: The Case of Singapore, 1980. 95pp. S$12.50

22

Suchart Prasithrathsint, Likhit Dhiravegin and Chavalit Si~ipirom, Culture and Fertility: The Case of Thailand, 1980. 68pp~ S$9.50

23

Sritua Arief, A Test of Leser's Model of Household. Consumption Expenditure in Malaysia and Singapore, 1980. 35pp. S$ 7.50

34pp.

S$7.50

Eight

The Case of

The Case of Malaysia, 1980.

THE AUTHOR Sritua Arief, Ph.D., is Chief Editor of the South East Asian Economic Review (Australia), and a Consultant to the United Nations, New York, on public policies and income distribution in ASEAN countries. A Visiting Fellow at the Institute of Southeast Asian Studies from December 1979 to July 1980, he has written extensively on various aspects of the Indonesian economy for several international journals including Cultures et developpemmt (Louvain, Belgium) and Asian Economies (Seoul, Korea). Among his publications are Consumption Patterns in Indonesia~· An Econometric Study (Jakarta, 1978) and Indonesia: Growth, Income Disparity and Mass Poverty (Jakarta, 1978).