The Determinants of Ethnic Inequality in Peninsular Malaysia

14
THE DETERMINANTS OF ETHNIC INEQUALITY IN PENINSULAR MALAYSIA by CHARLES HIRSCHMAN Assistant Professor of Sociology, Duke University, Durham, North Carolina, USA. 7’HY do different ethnic or racial communities share unequally in the distribution of rewards in society? This question has perplexed social scientists as well as aroused citizens for many years and will most probably continue to do so for the foreseeable future. Perhaps no other issue poisons the social relations of as many societies as does the problem of inter-ethnic inequality. The search for a general theory of the evolution of plural societies and the sources of inequality between racial and ethnic groups has been one of the dominant themes of sociological writing since the days of Robert Park (1950; also see Frazier, 1957; Cox, 1948; van den Berghe, 1967; w1 Lieberson, 1961; Gordon, 1964; Blalock, 1967). In recent years with the development of more specified models of the process of stratification (or status attainment), it has become possible to test. the relative explanatory power of several more modest theories of the middle-range. Among the competing explanations for differential socio- economic achievement by ethnic groups are: inheritance of poverty, differential cultural orien- tations towards achievement, and the selective practice of discrimination against certain ethnic groups. Most of this literature has followed from the application of multivariate analysis to the question of social mobility as developed by Blau and Duncan (1967, Chapter 6; also see Duncan, 1969; Duncan and Duncan, 1968; Featherman, 1971; Duncan and Featherman, 1972). In this research effort, I attempt to extend this form of analysis in an investigation of the process of stratification among the ethnic communities of Peninsular Malaysia. The first section of this paper gives a brief .overview of the plural structure of Peninsular Malaysian society. The second section develops the concept of the socioeconomic life cycle model and its application to the in- vestigation of ethnic stratification, followed by an empirical analysis of the processes of educa- tional and occupational attainment by ethnicity in Peninsular Malaysia. The final section will elaborate on the theoretical utility of this model for the comparative study of ethnic stratification. The Plural Society of Peninsular Malaysia The character of Malaysian society has been largely shaped by her ethnic diversity. Although Peninsular Malaysia may become a more homo- genous society in a few generations, for the present, ethnic divisions are strongly reinforced by cultural, religious, linguistic, and social differences. According to the 1970 Population Census, slightly more than half of the population (53.2%) identi- fied themselves as Malays, while more than a third (35.4%)were classified as Chinese and one out of ten (10.6%) were Indians. The remainder of less than one percent consisted of Thais, Eura- sians, and Europeans (Department of Statistics, 1972, p. 27). Although the Malaysian peninsular had been an historical crossroad of Southeast Asia, and her early ports contained communities of various Asian nationalities, most of the major immigration 25

Transcript of The Determinants of Ethnic Inequality in Peninsular Malaysia

Page 1: The Determinants of Ethnic Inequality in Peninsular Malaysia

THE DETERMINANTS OF ETHNIC INEQUALITY

IN PENINSULAR MALAYSIAby

CHARLES HIRSCHMANAssistant Professor of Sociology,

Duke University, Durham,North Carolina, USA.

7’HY do different ethnic or racial communitiesshare unequally in the distribution of rewards

in society? This question has perplexed social

scientists as well as aroused citizens for manyyears and will most probably continue to do so

for the foreseeable future. Perhaps no other issue

poisons the social relations of as many societies

as does the problem of inter-ethnic inequality.The search for a general theory of the evolution

of plural societies and the sources of inequalitybetween racial and ethnic groups has been one

of the dominant themes of sociological writingsince the days of Robert Park (1950; also see

Frazier, 1957; Cox, 1948; van den Berghe, 1967;

w1

Lieberson, 1961; Gordon, 1964; Blalock, 1967).In recent years with the development of more

specified models of the process of stratification(or status attainment), it has become possible to

test. the relative explanatory power of severalmore modest theories of the middle-range. Amongthe competing explanations for differential socio-economic achievement by ethnic groups are:

inheritance of poverty, differential cultural orien-

tations towards achievement, and the selectivepractice of discrimination against certain ethnic

groups. Most of this literature has followed fromthe application of multivariate analysis to the

question of social mobility as developed by Blauand Duncan (1967, Chapter 6; also see Duncan,1969; Duncan and Duncan, 1968; Featherman,1971; Duncan and Featherman, 1972). In thisresearch effort, I attempt to extend this form ofanalysis in an investigation of the process of

stratification among the ethnic communities ofPeninsular Malaysia. The first section of this

paper gives a brief .overview of the plural structureof Peninsular Malaysian society. The secondsection develops the concept of the socioeconomiclife cycle model and its application to the in-

vestigation of ethnic stratification, followed byan empirical analysis of the processes of educa-tional and occupational attainment by ethnicityin Peninsular Malaysia. The final section will

elaborate on the theoretical utility of this modelfor the comparative study of ethnic stratification.

The Plural Society of Peninsular Malaysia

The character of Malaysian society has beenlargely shaped by her ethnic diversity. AlthoughPeninsular Malaysia may become a more homo-genous society in a few generations, for thepresent, ethnic divisions are strongly reinforced bycultural, religious, linguistic, and social differences.

According to the 1970 Population Census, slightlymore than half of the population (53.2%) identi-

fied themselves as Malays, while more than a

third (35.4%)were classified as Chinese and one

out of ten (10.6%) were Indians. The remainderof less than one percent consisted of Thais, Eura-sians, and Europeans (Department of Statistics,1972, p. 27).Although the Malaysian peninsular had been an

historical crossroad of Southeast Asia, and her

early ports contained communities of variousAsian nationalities, most of the major immigration

25

jamill
1976. UMBC Economic Review (United Malayan Banking Corporation) (Kuala Lumpur) 12(1):25-38
Page 2: The Determinants of Ethnic Inequality in Peninsular Malaysia

Charles Hirschman

streams occurred during the period from 1850 to1930 (Kernial Singh Sandhu, 1969; Saw SweeHock, 1963). This was a period of increasinginvolvement by the British political and eco-nomic colonialists. The main industries duringthis period of increasing foreign economic penetra-tion were tin mining and commercial agriculturalplantations. After 1900, natural rubber produc-tion on large plantations became the main pillar ofthe national economy. These industries requiredlarge supplies of cheap labour, and migration fromthe poorer areas of Asia helped to fulfill thatneed (Jackson, 1961; Farmer, 1960). Althoughmost immigrants stayed only a few years inMalaysia before returning to their homelands, asignificant number became permanent residents ofthe country. By 1957, over j: of the Chinese, and

^ of the Indians in Peninsular Malaysia had beenbom locally (Fell, 1960, vol. 14, p. 15). Therehas been relatively little cultural assimilation orinter-marriage among the three major ethnic com-munities in Peninsular Malaysia. Language andreligious differences have been strong barriersagainst inter-ethnic unions of culture or identity.(Arasaratnam, 1970; Purcell, 1967; Wilson, 1967).As with most other plural societies, the ethnic

communities in Peninsular Malaysia are not evenlydistributed throughout the social structure.Chinese are more likely to live in urban areas,have higher education, engage in non-agriculturaloccupations, and have higher incomes than areIndians and Malays. Indians are generally closerto Chinese on most socioeconomic characteristicsthan are Malays who are predominantly a rural-agricultural population.

Table 1 gives a statistical overview of the socio-economic inequality among married adult malesfor the three major ethnic communities and itserves to introduce the variables which willbe utilized in the subsequent analysis.1 Thedifferences in both socioeconomic attainment andsocioeconomic origins are quite substantial be-tween all three communities, particularly betweenMalays and Chinese. The average monthly cashincome of a Malay man is less than half that

of a Chinese. Current residence is divided intothree categories, metropolitan cities (populationover 75,000 in 1957), non-metropolitan towns(population between 7,760 and 74,999) and ruralareas (the remainder of the country). Over 85percent of the Malay men live in rural areas ascompared to only 32 percent of the Chinese men.The Malaysian education system has changed

considerably over the years, but it was basicallypatterned after the British model of 6 years ofprimary schooling, 5 years of secondary schooling,and 2 years of university preparatory classes withnational examinations being a key determinantof progress through the system. There werevernacular schools which served primarily eachethnic community as well as an English medium-school system which was by and large an urbanphenomenon. On the average, Indians receiveda year more of schooling, and Chinese a yearand a half more of schooling than did Malays.Additionally, Chinese and Indians were muchmore likely to have attended English medium-schools than Malays. An English medium edu-cation was often essential for participation in thegovernmental sector until quite recently.Almost 90 percent of Malay men were born

in rural areas as compared with only 61 percentof Indians, and 35 percent of Chinese men. Thelower two panels of Table 1 show the distributionof occupations of the sample of married men ofeach ethnic group as well as that of their fathers.Malays are much more likely to be in agriculturaloccupations and to have come from agriculturalorigins than are Chinese and Indians. A sizeableminority of Chinese men (24.5 percent) are em-ployed in sales or commerce occupations.

1 These data are based on a sample of men marriedto women in the childbearing years who were interviewedin a fertility/family planning survey taken in 1966-1967.The survey was based on a national probability samplewith a sample size of 5,457, which was stratified by sizeof place and was weighted to conform to the actualdistribution of the population. For an excellent over-view of the survey design and questionnaire, see NationalFamily Planning Board of Malaysia, 1968. For a dis-cussion of the representativeness of the sample ofhusbands for the adult male population, see Hirschman,1972b, chapter 4.

26

Page 3: The Determinants of Ethnic Inequality in Peninsular Malaysia

The Determinants of Ethnic Inequality in Peninsular Malaysia

Table 1

ETHNIC DIFFERENCES IN THE SOCIOECONOMIC CHARACTERISTICS OFMARRIED MEN FROM 1966- 1967 WEST MALAYSIAN FAMILY SURVEY.

Mean Monthly Income

Current ResidenceRuralNon-Metropolitan UrbanMetropolitan

Years of Schooling01- 34- 567- 89-10

1112 or moreNot Reported

Mean Years of Education

Medium of EducationEnglishOther or Not Reported

BirthplaceRuralSmall TownTown or CityNot Reported

Father’s OccupationProfessional and AdministrativeClericalSalesServiceTransport and CommunicationCraft and Production ProcessLabourerAgriculturalNot Reported

Occupation of the Married MenProfessional and AdministrativeClericalSalesServiceTransport and CommunicationCraft and Production ProcessLabourerAgriculturalNot Reported

Unweighted Sample Size

Malay

$98.37

85.9%9.34.8

25.6%14.225.722.14.42.81.51.72.0

3.95 years

6.1%93.9

89.2%6.44.20.2

3.1%1.23.91.81.73.41.1

72.011.7

6.3%2.83.96.45.24.54.3

66.00.6

2,219

Chinese

$226.61

32.1%31.436.5

5.8%24.116.619.57.4

10.53.46.46.4

5.47 years

12.5%87.5

34.9%31.429.14.6

6.5%4.1

22.02.03.79.08.3

26.518.0

9.1%7.2

24.54.77.519.110.916.30.6

2,332

Indian

$159.44

62.9%16.720.4

14.6%19.823.111.613.17.74.92.92.2

4.95 years

22.7%77.3

60.8%16.419.43.4

4.2%4.97.75.33.2

12.48.3

41.013.0

4.4%6.75.35.09.0

12.511.845.20.1

795

27

Page 4: The Determinants of Ethnic Inequality in Peninsular Malaysia

Charles Hirschman

This brief overview of some of the majordimensions of ethnic inequality will serve to pro-vide a background for the subsequent investiga-tion into the determinants of these patterns.

The Socioeconomic Life Cycle Model

The primary goal of this research is to explainwhy Chinese and Indians are more successfulthan Malays in achieving higher education, higherincomes, and higher status occupations. Theapproach adopted in this research can most aptlybe called the Socioeconomic Life Cycle Model(Duncan, 1967), although it has been referred toas "the process of stratification" (Duncan, 1968)and the "status attainment model" (Sewell andHauser, 1972). The basic model portrays the

pattern of influence from Socioeconomic originsto Socioeconomic attainment later in life. Thus,the most simple model might be represented inFigure 1 which shows three stages: (1) originstatus, (2) educational attainment, and (3) socio-economic attainment. One’s origin status is theposition in the Socioeconomic hierarchy into whichone is born usually measured by the occupa-tional and educational status of one’s father.Origin status is usually considered to be an im-portant influence upon education because one’sfamily position can give economic as well ascultural support for continued schooling. In turn,one’s education is seen as the most importantinfluence on the occupational attainment of mostadult men.

Figure 1THE BASIC MODEL OF THE SOCIOECONOMIC LIFE CYCLE

Origin Status Socioeconomic

EducationalAttainment

Although this model is deceptively simple, itprovides an avenue for investigating a numberof important substantive questions in the studyof stratification. Using this approach with a

multiple regression form of analysis, one canaddress such fundamental questions as: the degreeof rigidity in the stratification system, and whetherwealthy parents provide any direct influence onson’s Socioeconomic career, independently of edu-cation (for the classic exposition, see Blau and

Duncan, 1967; for an elaboration of the modelincluding a variety of different variables, seeDuncan, Featherman, Duncan, 1972).

Although this model of the soeioeconomic lifecycle has usually been applied to studies ofstratification of the total population, it can alsobe applied to studies of differential achievementfor population subgroups such as ethnic com-munities. From this analytical perspective, it ispossible to measure the relative importance of

28

Page 5: The Determinants of Ethnic Inequality in Peninsular Malaysia

The Determinants of Ethnic Inequality in Peninsular Malaysia

ethnicity as a determinant of one’s achievementas compared to other prior determinants, educa-tion and social origins; or in other words, the net

influence of ethnic membership after "statistically

controlling" for other background factors. Thisform of analysis speaks to a profound theoreticalissue regarding the role of ethnicity in social

inequality. The theoretical question is whethera disadvantaged ethnic group, in this analysis,Malays, has an unfavorable position in the socialhierarchy simply because they started out at a

lower level. In other words, is the Malay dis-advantage only one which affects poorer Malays,just as it affects all persons of poor origins,irrespective of ethnicity or is there a special dis-

advantage which is common to all Malays? Ifthis is true, then our regression analysis will showthat there is a net effect due to ethnicity aftercontrolling for other social background factors.

If the ethnic differences can be explained en-

tirely by the unfavourable social origins of the

disadvantaged group then we have a situationwhich might be called "inheritance of poverty."In this situation ethnic inequality cannot be main-tained indefinitely into the future because similar

processes of change (social mobility) affectingdifferent population structures (each ethnic groupdistributed by occupation or social class) willultimately result in a convergence of the popula-tion structures (for a discussion and applicationof this demographic principle to study of social

mobility, see Matras, 1960). In an analysis ofBlack-White occupational differentials in the

United States, Lieberson and Fugitt (1966) useda Markov chain analysis to demonstrate the effectsof the same social mobility process applied to

both blacks and whites (thus allowing only thedisadvantages of origins to be the source of in-

equality) and found that occupational differentialsbetween races would largely disappear in twogenerations.

If there is a situation with a "net effect"of ethnicity on socioeconomic achievement, howmight this be interpreted? Perhaps the two major

theories which might be proposed to accountfor this ethnic disadvantage are what I call the"cultural explanation" and the "discriminationexplanation". The cultural thesis suggests thatethnic communities have different cultural tradi-tions which offer a greater or lesser motivation

for achievement. In the simplest form, this theorywould posit that certain ethnic groups emphasizevalues of thrift, hard work, and deferred gratifica-tion which gives them an advantage over othercultures which are more inclined to take thingseasy. These ideas have often been expressed inthe American sociological literature under thegeneral theory regarding "the need for achieve-

ment" (Rosen, 1959). From this perspective, one

would argue, that even with equivalent socio-

economic backgrounds (same origin status andeducation) a disadvantaged ethnic group is less

likely to achieve because its members have less

motivation for achievement.An alternative explanation for the same em-

pirical finding (ethnicity has a net effect on

achievement) would be that a disadvantaged ethniccommunity encounters discrimination in the pro-cess of stratification. Discrimination can occur

in a variety of institutional environments, suchas schools, the job market, or within economicinstitutions. This would mean that teachers, em-

ployers and job supervisors give preferential treat-

ment to certain ethnic groups and not to others.It is perhaps more easy to discriminate in certain

employment markets where formal qualificationsare not required. It is probably more difficultin educational institutions and certain occupa-tional streams, such as professions where "paperqualifications" and other objective criteria are

probably used.These two diverging explanations are based on

the same empirical finding, yet have radicallydifferent interpretations for analyzing society.According to the cultural explanation, the in-

dividuals of the disadvantaged groups must

change their attitudes and motivations in order to

achieve full participation while the discrimination

29

Page 6: The Determinants of Ethnic Inequality in Peninsular Malaysia

Charles Hirschman

hypothesis suggests that the structure of societyand rules of the game, so to speak, must bechanged in order to achieve ethnic equality.Only rarely do we have data which contain

information on both the structural factors andsocial psychological orientations which wouldmake it possible to test between these alternativeexplanations. (For example, see Featherman,1971.) In this study the data are not full enoughto permit a definitive investigation of this issue,but we will be able to make some tentative con-clusions on the basis of inferential evidence inthe few occupations where a "net effect" ofethnicity is found.

Data and Model of Ethnic Stratification inPeninsular Malaysia

The data used originated from a populationsurvey taken during 1966 and 1967 in PeninsularMalaysia. The national probability survey wasfocussed upon gathering data on the fertility andcontraceptive behavior and attitudes of Malaysianmarried women, (for previous analyses of thesedata, see Palmore and Ariffin, 1969; Palmore,Klein, and Ariffin, 1970; and Palmore, Hirsch,and Ariffin, 1971). The survey interviewed 5,457

married women in the childbearing years (age15-44). In addition to the data on fertility, thewomen were asked several questions about theirhusband’s characteristics including his ethnicgroup, birthplace, education, occupation, income,and his father’s occupation when her husbandwas about twelve years old. Thus my analysisis based upon the sample of Malaysian husbandswhose characteristics were reported by their wives.White not a random sample of adult men, it isprobably fairly representative of such a group.Using the socioeconomic life cycle model and

the variables contained in this survey, I haveformulated a model of ethnic stratification whichis represented in Figure 2. A man is assumed tobegin his life with three important characteristicswhich will influence his subsequent socioeco-nomic career: his ethnic membership, his father’soccupation, and his birthplace. Then next followseducation, measured in years of schooling, andeducational medium, whether the education wasin an English-medium school. Then in successivestages come current residence, occupation, andincome. While such a model might not reallydescribe the temporal ordering of every man’scareer, it is probably a reliable enough approxi-

Figure 2

SOCIOECONOMIC LIFE CYCLE OF THE PROCESS OF STRATIFICATIONIN PENINSULAR MALAYSIA

Ethnicity

Education

Father’sOccupation

CurrentResidence Occupation Income

Educational Medium

Birthplace

30

Page 7: The Determinants of Ethnic Inequality in Peninsular Malaysia

The Determinants of Ethnic Inequality in Peninsular Malaysia

mation to guide our analysis. The categories anddistribution of each of these variables have beenpresented in Table 1. Now let us turn to theactual analysis of ethnic stratification.

The Case of Education

First, let us consider the impact of ethnicmembership and social background upon edu-cational attainment, then turn to occupationalattainment.The technique of analysis used here will be a

variant of regression analysis known as multipleclassification analysis which simply means thateach independent variable will be coded into aset of categories rather than a single variable.Thus, the variable Birthplace becomes threevariables: "rural", "small town", and "town orcity".2 Each category is coded "one" if thatcharacteristic is present, otherwise "zero". Theraw regression coefficients are transformed so asto present the effect of being in that categoryexpressed as a deviation from the grand meanof the dependent variable (Melichar, 1965).

Table 2 shows the effects of social back-ground upon the dependent variables "Years ofSchooling" and "English-Medium Schooling".The independent variables are Ethnicity, Father’sOccupation, and Birthplace. For each dependentvariable, three models or equations were run, thefirst one contains only the regression of thedependent variable on Ethnicity, the second modeladded in Father’s Occupation, and the thirdmodel adds Birthplace for a total of three in-dependent variables. This sequence of modelswith ethnicity entered first allows us to see whatthe gross effects of Ethnicity are (as observed inthe population) and the net effects of ethnicity(after controlling for differences in social back-ground)^As mentioned earlier, the coefficients are ex-

pressed as deviations from the grand mean ofthe dependent variables. The grand mean ofYears of Schooling is 4.4 years for the entiresample. The first column shows the effects ofa model with only Ethnicity as an independent

variable which shows that the Malay mean is.4 a year less than the overall mean of 4.4 years,while the Chinese mean is 1.1 years more than thesample average of all married men. This gap of1.5 years of schooling between the average Malayman and average Chinese man is narrowed to0.8 of a year in the second equation with Father’sOccupation introduced as an additional indepen-dent variable. This change in ethnic coefficientscan be interpreted as the result of the fact thatChinese and Malays have different distributionsof father’s occupations. The third model showsthat the gap is further reduced to 0.3 of a yearof schooling when Place of Birth is introducedas an additional independent variable. It seemsclear that most of the ethnic gap (80 percentbetween Malays and Chinese) is a result ofdifferential social origins, or what might be calledthe handicap of a rural-agrarian background forMalays.The proportion attending English-medium

schools is analyzed in a comparative fashion onthe right hand side of Table 2. Here, the grandmean of the dependent variable is the proportionof the entire sample who have attended English-medium schools (9.5 percent). The coefficientsin the table represent deviations from this figurein percentage terms. In the first column, we seethat Indians are 13 percent more likely to attendEnglish-medium schools than the average, whileChinese are 3 percent more likely, and Malays 3percent less likely than the average for the entirepopulation. However, after introducing Father’sOccupation and Birthplace in the third model, theeffects for Malays and Chinese are reversed with

2 Actually a fourth category "birthplace unknown"was entered into the regression analysis to utilize caseswhich had missing data for some variables. This wasalso done for other independent variables, but thecoefficients obtained are not presented here. If a casehad missing data for the dependent variable. Education,the person was assigned the value of the grand mean.

3 More complex models incorporating ethnic-socialbackground interaction terms have been run (Hirschman,1972b, Appendix G). While it is possible to add oneor two points of explained variance with more complexmodels, the basic pattern of coefficients and interpreta-tion remain the same.

31

Page 8: The Determinants of Ethnic Inequality in Peninsular Malaysia

Charles Hirschman

Table 2

EFFECTS OP SOCIAL BACKGROUND ON THE NUMBER OF YEARS OFSCHOOLING AND ON THE PROPORTION HAVING AN ENGLISH-MEDIUM

EDUCATION, PENINSULAR MALAYSIA, 1966- 1967.

Independent Variables Dependent VariablesYears of Schooling, English-Medium EducationGrand Mean 4.4 Grand Mean 9.5%

(1) (2) (3) (1) (2) (3)

Ethnicity

Malay 0.4 0.2 0.1 3.4 1.5 0.2

Chinese 1.1 0.6 0.2 3.0 1.4 5.9

Indian 0.5 0.3 0.2 13.1 10.1 9.0

Father’s Occupation

Professional and Administrative 3.1 2.7 30.4 26.5

Clerical 3.2 2.8 45.2 39.8

Sales 1.2 1.0 6.1 33.5

Service 1.5 1.2 13.4 9.4

Transport and Communication 1.3 1.0 14.4 11.2

Craft and Production Process 0.8 0.6 5.5 3.2

Labourer 0.1 0.4 1.9 4.8

Agricultural 0.5 0.4 5.7 4.2

Birthplace

Rural 0.3 3.8

Small Town 0.9 9.6

Town or City 1.2 14.5

R2 (Variance Explained) 3.9 13.4 15.7 3.5 16.1 19.9

Notes:

For each dependent variable:

Model includes only Ethnicity as an independent variable.

Model 2 includes Ethnicity and Father’s Occupation as independent variables.

Model 3 includes Ethnicity, Father’s Occupation, and Birthplace as independent variables.

32

Page 9: The Determinants of Ethnic Inequality in Peninsular Malaysia

The Determinants of Ethnic Inequality in Peninsular Malaysia

Malay being slightly positive and Chinese almost6 percent negative. The net effect of being anIndian remains 9 percent above the average ofthe entire population.Another way to compare the relative explana-

tory power of ethnicity with social origins is toexamine the proportion of variance explained(R2) for each of the models in Table 2. UsingEthnicity alone as an independent variable ex-plains only 3.9 percent of the variance in educa-tional attainment and only 3.5 percent of thevariance of the proportion attending English-medium schooling. The inclusion of the othertwo social background variables in the regressionanalysis raises the amount of variance explainedto 15.7 percent and 19.9 percent, respectively.

It seems that most of the Malay disadvantagein the educational process is a result of theirorigin status and birthplace which retards accessto schooling. Being more likely the sons offarmers with probably fewer resources (monetaryand cultural) and being born in rural areas whereschools were not common until recent years,Malays have not had the same level of attainment(in terms of years of schooling or attendance atEnglish-medium schools) as non-Malays. Thereseems to be no strong ethnic effect in the finalmodels which might be interpreted as a uniquecultural orientation or discrimination with the ex-ception of the high Indian coefficient for English-medium schooling. Perhaps Indians have a strongmotivation to attend English-medium school orpossibly English-medium schools were simplymore accessible to them (for a similar study basedon 1957 Census data with similar conclusions,see Hirschman, 1972a).

The Case of Occupational Attainment

Table 3 shows the analysis of the relative effectsof ethnicity and social background upon occupa-tional attainment separately for eight different

occupations. That is, each different occupationwas run separately as a dependent variable forsix sequential and cumulative models of socialbackground variables. The first model includes

only Ethnicity as an independent variable, thesecond added in Father’s Occupation, the thirdadded in Birth-place, the fourth EducationalAttainment, the fifth Educational Medium, andthe sixth included all previous variables plusCurrent Residence. Rather than presenting a fulltable for each occupational attainment categorywith the effects from all social variables. Table 3represents a condensed version of eight tableswith only the coefficients or effects of Ethnicityupon each occupational attainment presented foreach of the six models. All coefficients are ex-pressed as deviations from the grand mean, whichis the proportion of the total sample in anyparticular occupation.For example, the proportion of the total sample

in professional and administrative occupations is6.8 percent. The first model shows the grosseffects of ethnicity (as observed in the population)with Chinese having a positive effect of 2.1 per-cent, Malays a negative 0.3 percent effect, andIndians a negative 2.3 percent effect. The neteffects of ethnicity are shown in Model Six afterall other variables are entered along with ethnicity.These results show that the signs of the coefficientsfor Malays and Chinese have been reversed withMalay being positive and Chinese being negative.Clearly the Malay disadvantage shown in Model1 (smaller proportions in professional occupations)result from disadvantaged social origins as com-pared to Chinese. If Malays and Chinese hadequivalent social backgrounds as measured in thevariables through Model 6, Malays would prob-ably be overrepresented in professional occupa-tions. If one scans down the table looking onlyat the difference between Malay and Chinesecoefficients in Model 1 (gross) and Model 6 (net)one can divide the urban occupations (excludingagriculture) into two groups according to thepattern of change in the ethnic coefficients. Inone group we can include Professional, Clerical,Service, and Transport and Communicationoccupations, where the handicap for Malays canbe entirely explained by their disadvantaged socialorigins inheritance of "rural agrarian back-

33

Page 10: The Determinants of Ethnic Inequality in Peninsular Malaysia

Charles Hirschman

Table 3

EFFECTS OF ETHNICITY IN THE OCCUPATIONAL ATTAINMENT PROCESSFOR SIX SEQUENTIAL MODELS OF SOCIAL BACKGROUND FOR EACH OFEIGHT OCCUPATIONS. EFFECTS ARE EXPRESSED AS DEVIATIONS FROMTHE PERCENTAGES OF THE TOTAL SAMPLE IN THAT OCCUPATION.

GrandMean (1) (2) (3) (4) (5) (6)

Professional andAdministrative

MalayChineseIndian

ClericalMalayChineseIndian

SalesMalayChineseIndian

ServiceMalayChineseIndian

Transport andCommunication

MalayChineseIndian

6.8%

4.2%

8.4%

5.8%

6.1%

0.32.12.3

1.43.12.5

4.516.03.1

0.51.10.8

0.91.42.9

0.40.63.4

0.71.41.2

3.412.42.8

1.01.92.0

0.30.02.3

0.90.83.7

0.21.00.6

3.412.32.8

1.22.42.4

0.21.42.0

1.21.44.3

0.41.10.1

3.412.42.6

1.32.82.0

0.21.52.2

1.21.24.4

0.30.31.5

3.412.02.2

1.32.82.0

0.21.42.1

1.21.44.3

0.71.01.5

3.110.92.1

2.25.82;0

0.73.32.2

Craft andProduction Process 8.5%

MalayChineseIndian

3.9 3.2 2.9 2.8 2.9 2.010.7 9.2 8.4 8.3 8.4 5.84.1 2.4 2.1 1.9 1.8 1.8

LabourerMalayChineseIndian

AgriculturalMalayChineseIndian

6.5%2.24.55.4

53.3%12.7

-37.08.1

2.14.34.9

7.326.12.4

2.14.35.0

5.0-19.1

0.4

2.04.15.0

4.4-17.6

0.4

2.03.85.2

4.5-18.7

1.5

1.73.05.3

1.57.61.3

Notes:The independent variables in these models are:

Model Ethnicity.Model 2 Ethnicity, Father’s Occupation.

Model 3 Ethnicity, Father’s Occupation, Birthplace.Model 4 Ethnicity, Father’s Occupation, Birthplace, Years of Schooling.

Model 5 Ethnicity, Father’s Occupation, Birthplace, Years of Schooling, Educational-Medium.

Model 6 Ethnicity, Father’s Occupation, Birthplace, Years of Schooling, Educational-Medium, Current Residence.

34

Page 11: The Determinants of Ethnic Inequality in Peninsular Malaysia

The Determinants of Ethnic Inequality in Peninsular Malaysia

ground". For all these occupations the net effectfor being a Malay is positive after all othervariables are included in the model.

In a second group, one can place Sales, Craftand Production Process, and Labourer occupationstogether. For those occupations, the Malay dis-advantage is decreased after controlling for socialbackground, but it still remains sizeable in thesixth model. Why do Malays not seem to beable to enter these occupations as frequently asothers even if they had the equivalent patterns ofurban residence, education, and other social back-ground characteristics. I do not think that thecultural hypothesis provides a very satisfactory ex-planation. One would need to posit that Malayshave sufficient goals for achievement in somewhite and blue collar occupations, but not others.The alternative explanation, earlier noted as "dis-crimination" might provide some insight herethough it needs to be considered in the light ofseveral factors. First, all three of these occupa-tions are largely concentrated in what might becalled the "small shop" sector. Such small shops,engaged in commerce, manufacturing, or someother activity, usually only employ very fewpaid workers, sometimes only a single family. Itis not surprising that many such shops and indus-tries employ members of one ethnic group. Theemployer may wish to hire members of his ownfamily or because of familiarity or languageconsideration choose to hire other members ofhis own ethnic group in preference to those fromother ethnic groups. Recent evidence has in-dicated that the proportion of employed Malayswho are employers (0.7%) is much less than thecomparable figures (5.0% and 2.1%) for em-ployed Chinese and Indians (Choudry, 1970, p.92). While the practice of hiring members ofone’s own ethnic group for familial reasons maynot have the invidious connotations of discrimina-tion, it is a selective social process which operatesto inhibit the dispersion of all ethnic groupsthroughout the occupational structure. An alter-native explanation to the one I have presentedhere might suggest that Malays are less likely to

apply for jobs in the "small shop" sector becausethey do not like to work in surroundings wherelanguage and cultural practices are unfamiliar.Most likely, this is a part of the complex over-all process of selection into various occupationaland industrial sectors.

Conclusion and Discussion

This paper has reviewed the basic model ofthe socioeconomic life cycle and how it can beapplied to the study of ethnic inequality. I havesuggested that there are two theoretical questionswhich arise from such an investigation. The firstis to measure if there is an ethnic "effect" ordifferential that remains after statistically con-trolling for other social background characteristics.If the ethnic differential approaches zero after"equalizing" social backgrounds, this means thatan inheritance of poverty situation exists, whereone ethnic group is disadvantaged because a largerproportion of its members grow up in poorerfamilies. A social strategy to deal with thissituation would be to extend educational andsocioeconomic achievement opportunities to allthose who come from poorer or disadvantagedhomes. If the effects of childhood poverty onadult socioeconomic achievement could be elimi-nated, the ethnic differentials in status could beeliminated with no explicit public policies focussedon ethnicity per se.

However, if an ethnic "effect" or differentialdoes remain after controlling for social back-ground characteristics then a second theoreticalissue must be confronted. What is it aboutethnicity that causes differentials in achievement?In this paper, I have suggested a rather crudedistinction that either the ethnic groups in questionhave differential goals and thus behave differently,or else they are treated differently by gate-keepersin the system. I have labeled these two con-flicting interpretations as the cultural hypothesisand the discrimination hypothesis, respectively.Whereas, in reality these factors or explanationsmay be intertwined, I have tried to make theconceptual distinction between them because both

35

Page 12: The Determinants of Ethnic Inequality in Peninsular Malaysia

Charles Hirschmmt

are theories which have a sociological traditionas well as popular images. If there is an ethniceffect, how is it possible to test whether thecultural or discrimination hypothesis is morecorrect? One way might be to measure social

psychological variables of ambition as well as

perception of discrimination by the disadvantagedethnic group. If such variables "intervene" or

transmit the effect of ethnicity on socioeconomicattainment, then we might have an adequate basis

for interpretation. The work of David Deather-man (1971) and Duncan and Featherman (1972)have been illustrative of combining psychologicaland structural variables in models of socioeco-nomic achievement.The results of this empirical analysis of ethnic

stratification in Malaysia suggest that the primaryexplanation for the lower socioeconomic attain-ment of the Malay men as compared to non-

Malays in Malaysia is because they are more

likely to come from disadvantaged social originsas measured by father’s occupation and birth-

place. Over 89 percent of Malays were bom inrural areas as compared with 61 and 35 percent,respectively, for Indians and Chinese. Equallysignificant was the fact that 72 percent of our

sample of adult Malay men were sons of farmerswhile the comparable figures for Indians andChinese were only 41 and 26 percent. I havesuggested that these handicaps of social originprovided less access to education and the urbanopportunity structure for Malays. If Malaysand Chinese had equivalent social backgroundsaccording to the variables measured here, over

80 percent of the educational attainment gapwould be eliminated. However, there did seem

to be a significant net effect of ethnicity in severalsectors of the occupational attainment process.Malays were still less likely to enter Sales, Craft,and Labourer occupations even after all othersocial variables were controlled. From the patternof occupational differentials, I have inferred thatthis "net effect" as being more likely the resultof discrimination than of differential cultural

ambitions. The selective process of discrimina-tion may be no more than the preference of an

employer to hire his own son in a family business.But if a significant proportion of the jobs incertain occupational sectors are created by smallfamily run businesses, then this might result indifferential access by ethnicity into these occupa-tion.The study of comparative race and ethnic

relations has been hindered by a lack of a com-mon reference point. In some theories or em-

pirical studies, the focus has been on assimilation,in others on inter-racial conflicts, and in otherson the degree of power held by one group over

another. While I would not want to discouragethe diversity of sociological inquiry, I would liketo suggest that the mode of analysis utilizedhere the study of ethnic stratification is a

particularly appropriate method of comparativeanalysis. Its focus on inequality in terms ofvarious status dimensions such as income, educa-tion, and occupation, which is often the under-lying cause behind much of inter-ethnic hostilityand conflict. With the appropriate data and theconceptual tools of the regression analysis andthe socioeconomic life cycle model, research can

distinguish how much inequality results fromfactors of differential social background and howmuch is due to factors associated with race orethnic membership itself. I regard this sort ofdistinction as a major theoretical advance com-pared to the present conventional wisdom whichis filled with notions of differential racial/ethnicpropensities to do this or do that. Additionally,such research might provide both practical direc-tion as well as legitimation for public policy whichis designed to reduce the association betweenethnicity and inequality.

36

Page 13: The Determinants of Ethnic Inequality in Peninsular Malaysia

The Determinants of Ethnic Inequality in Peninsular Malaysia

ACKNOWLEDGEMENTS

I am grateful to Dr. Ariffin bin Marzuki, formerDirector of the National Family Planning Boardof Malaysia and Professor James Palmore of theUniversity of Hawaii for making a copy of the datatape from the 1966-1967 West Malaysia FamilySurvey available to me. I am also indebted to

David Brown, Avery Guest, Patrick Haran, JamesHouse, and Alan Kerckhoff for constructive com-ments on an earlier draft of this paper. This re-search was partially supported by a training grant(GM 01190) from the Center for Demographyand Ecology of the University of Wisconsin.

REFERENCES

Arasarathnam, Sinnappah. Indians in Malaysia andSingapore. Kuala Lumpur: Oxford UniversityPress, 1970.

Blalock, Hubert. Towards a Theory of Minority-GroupRelations. New York: John Wiley, 1967.

Blau, Peter and Otis Dudley Duncan. The AmericanOccupational Structure. New York: John Wiley,1967.

Choudry, N. S. Socio-Economic Sample Survey ofHouseholds-Malaysia, 1967-1968: Employment andUnemployment. Kuala Lumpur, Malaysia: De-partment of Statistics, 1970.

Cox, Oliver C. Caste, Class and Race. New York:Monthly Review Press, 1948.

Department of Statistics, Malaysia. The 1970 Censusof Population, Malaysia A Preliminary ReportBased on Field Count Summary. Kuala Lumpur:Department of Statistics, Malaysia, 1971.

Department of Statistics, Malaysia. Community Groups.Kuala Lumpur: Department of Statistics, Malay-sia, 1972.

Duncan, Otis Dudley. Discrimination against Negroes.The Annals 371 :85-103, 1967.

Duncan, Otis Dudley. Social stratification and mobility:problems in the measurement of trend. In EleanorSheldon and Wilbert Moore (eds.) Indicators ofSocial Change. New York: Russell Sage Founda-tion, 1968.

Duncan, Otis Dudley. Inheritance of poverty or in-heritance of race? In Daniel P. Moynihan (ed.)On Understanding Poverty: Perspectives from theSocial Sciences. New York: Basic Books, 1969.

Duncan Beverly and Otis Dudley Duncan. Minoritiesand the process of stratification. American Socio-logical Review 33:356-364, 1968.

Duncan, Otis Dudley, David Featherman and BeverlyDuncan. Socioeconomic Background and Achieve-ment. New York: Seminar Press, 1972.

Duncan, Otis Dudley and David Peatherman. Psycho-logical and cultural factors in the process ofoccupational achievement. Social Science Research1:121-145, 1972.

Featherman, David. The socio-economic achievement ofwhite religio-ethnic sub-groups: social and psycho-logical explanations. American Sociological Re-view 36:207-222, 1971.

Fell, H. 1957 Population Census of the Federation ofMalaya, Report No. 14. Kuala Lumpur: Depart-ment of Statistics, Federation of Malaya, 1960.

Frazier, E. Franklin. Race and cultural contacts in themodern world. Boston: Beacon Press, 1957.

Gordon, Milton. Assimilation in American Life. NewYork: Oxford Press, 1964.

Gullick, J. M. Malaysia, London: Emest Benn Limited,1963.

Hirschman, Charles. Educational patterns in colonialMalaya. Comparative Education Review 16:486-502, 1972.

Hirschman, Charles. Ethnic Stratification in WestMalaysia. Unpublished Ph.D. dissertation. Uni-versity of Wisconsin, Madison, 1972.

Hirschman, Charles. Ethnic and Social Stratification inPeninsular Malaysia. Arnold and Catherine RoseMonograph. Washington, D.C.: American Socio-logical Association, 1975.

Jackson, R. W. Immigration Labor and the Develop-ment of Malaya 1786-1920. Kuala Lumpur:Government Printer, 1961.

Kernial Singh Sandhu. Indians in Malaya: SomeAspects of their Immigration and Settlement.London: Cambridge University Press, 1969.

Lieberson, Stanley. A societal theory of race and ethnicrelations. American Sociological Review 26:902-910, 1961.

Lieberson, Stanley and Glen Fugitt. Negro-whiteoccupational differences in the absence of dis-crimination. American Journal of Sociology 75:188-200, 1966.

Matras, Judah. Comparison of intergenerational occupa-tional mobility patterns: an application of theformal theory of social mobility. PopulationStudies 14:163-169, 1960.

Melichar, Emanuel. Least squares analysis of economicsurvey data. Proceedings of the Annual Meetingof the American Statistical Association. Businessand Economics Section, 1965.

National Family Planning Board of Malaysia. Reporton the West Malaysia Family Survey: 1966 1967.Kuala Lumpur: National Family Planning Board,1968.

37

Page 14: The Determinants of Ethnic Inequality in Peninsular Malaysia

Charles Hirschman

Palmore, James A. and Ariffin bin Marzuki. Marriagepatterns and cumulative fertility in West Malaysia,1966-1967. American Journal of Sociology 76:375-398, 1969.

Palmore, James A., Robert E. Klein and Ariffin binMarzuki. Differential participation in extendedfamilies: West Malaysia 1966-1967. AmericanJournal of Sociology 76:375-398, 1970.

Palmore, James A., Paul M. Hirsch, and Ariffin binMarzuki. Interpersonal communication and thediffusion of family planning in West Malaysia-Demography 8:411 -425, 1971.

Park, Robert. Race and Culture. Glencoe, Illinois:The Free Press, 1950.

Farmer, J. Norman. Colonial Labor Policy and Admini-stration. New York: Monograph of the Associa-tion for Asian Studies, 1960.

Purcell, Victor. The Chinese in Malaya. Kuala Lumpur:Oxford University Press (originally published in1948), 1967.

Roff, William R. The Origins of Malay Nationalism.New Haven: Yale University Press, 1967.

Rosen, Bernard C. Race, ethnicity and the achievementsyndrome. American Sociological Review 24:4760, 1959.

Saw Swee Hock. Trends and differentials in inter-national migration in Malaya. Ekonomi 4:87113, 1963.

Sewell, William H. and Robert M. Hauser. Causes andconsequences of higher education: models of thestatus attainment process. American Journal ofAgricultural Economics 23:851-861, 1972.

Silcock. T. H. and E. K. Fisk (eds.). The PoliticalEconomy of Independent Malaya. Berkeley: Uni-versity of California Press, 1963.

van den Berghe, Pierre L. Race and Racism: AComparative Perspective. New York; John Wiley,1967.

Wang, Gungwu (ed.). Malaysia: A Survey. NewYork: Frederick A. Praeger, 1964.

Wilson, Peter J. A Malay Village and Malaysia:Social Values and Rural Development. New Haven:HRAF Press, 1967.

38