Ethnicized Social Mobility in Taiwan:

Mobility Patterns Among Small and Medium-Scale Businesspeople

This paper discusses the mobility paths of two ethnic groups in Taiwan, “Taiwanese” and “Mainlanders,” and seeks to explain why they opt for particular paths. The nature of Taiwan’s political and economic structure and the relationship of that structure to differences in educational attainment and thus differences in occupation between these two ethnic groups will be outlined. Differential educational opportunities resulting from government policy have set limits for upward mobility according to academic credentials, especially among ethnic Taiwanese, and from the 1960s-70s, running one’s own small business became the most significant mobility path for less educated Taiwanese.

The data for my study are based on two sources: interview fieldwork conducted in Taiwan from March 1993 to January 1994 and from September 1997 to March 1998;[1] and raw data from the General Survey of Social Change in Taiwan, 1992.[2]

In the first part of the paper I will describe the general pattern of educational and occupational development in Taiwan and will focus on the difference between Taiwanese and Mainlanders. The term ‘Taiwanese’ refers to Han Chinese people present in Taiwan before 1945 and their descendants, namely the Minnan (Hokkien-speaking) and Hakka groupings, while ‘Mainlander’ refers to those who came to Taiwan from China after 1945 and their descendants.[3] In the second part of the paper I will discuss the influence of political and economic structures on educational development in the postwar period. There I will identify important characteristics of Taiwan’s education system, the role of the state and the ways in which a family's socioeconomic background strongly influence education and use path analysis to chart these effects. I will finish with a discussion of why those who were less educated, mostly Taiwanese, did not adopt collective strategies to climb the social ladder but instead chose to become independent proprietors.

Literature Review

The relationship between a family’s socioeconomic status, the parents’ education, and the child’s education and occupational attainment has been a topic of frequent interest in sociological research. A clear pattern has emerged of socioeconomic status and educational level of parents having a significant influence on their children’s education, which in turn affects occupational attainment. Following this line of research, we can easily trace the mobility patterns of different classes (see, e.g., Treiman, 1970; Blossfeld and Shavit, 1993; Blau and Duncan, 1967; Bourdieu and Passeron, 1977; Bowles and Gintis, 1976).

Oddly, research on businesspeople in Taiwan has been divided into two camps. Social stratification scholars describe general mobility patterns, but they do not explain why Taiwanese and Mainlanders often adopt different strategies to climb the social ladder. In particular, no-one has previously discussed the mobility patterns of small and medium-sized businesspeople in Taiwan (see, e.g., Chen and Yeh, 1998; Tsai and Chiu, 1993a, 1993b; Huang Yizhi, 1995; Yang Ying, 1988). On the other hand, those who conduct research on small and medium-sized businesspeople based on interviews or fieldwork almost never touch on the relationship between socioeconomic background, educational level and occupations held by parents. They do not analyze in detail who is more likely to become self-employed or an employer, though some of them have had nascent ideas about mobility paths (see, e.g., Shieh, 1989; DeGlopper, 1979; Amsden, 1991; Gates, 1981). Much research shows that Chinese seek to be proprietors, and the most common, and uninformative, explanation has been the ‘Chinese mentality’. Some say that Chinese have been, and generally still are, willing to exert enormous amounts of effort in search of improvement and security for their families. Though the cultural explanation sometimes talks about history, that history is static and unchanged. The authors seem to believe that social institutions never influence culture, or at most just play a minor role (Wong, 1995; Harrell, 1985). One obvious defect of these explanations is that they fail to explain why certain groups are more likely to run their own businesses than others, e.g., Taiwanese viz. Mainlanders in Taiwan.

Some research shows that there is a different probability of Taiwanese and Mainlanders running a business, even though they hold the same ‘Chinese cultural background’. Susan Greenhalgh’s research on Taiwanese urban society found that ‘being Taiwanese’ means drawing on family, community and religious ties to climb through small-scale business and then larger-scale entrepreneurship into the commercial and industrial elite. “Being Mainlander” means eschewing ties of kinship and community and using contacts with people in the bureaucracy and a higher education to obtain a white-collar job in government, and hence a secure position in the bureaucratic elite’ (Greenhalgh, 1984). Marshall Johnson argued that Taiwanese held a comparative advantage in the small business world where networks, familiarity and trust are required. The factors that these authors suggest are not altogether convincing. If Taiwanese were able to use family or community ties to find their way in the business world, one cannot see any reason why Mainlanders could not have done so, especially since relatively good opportunities were available to the latter throughout the Kuomintang (KMT) era. Such opportunities should have been in abundant supply from this Mainlander-dominated administration, which held a firm grip on the economy in the 1950s and on the public sector until as late as the 1980s.[4] For example, it should have been easier for Mainlanders to do business with Mainlander-dominated public enterprises, if – as Marshall Johnson argued – networks, familiarity and trust were required. He and other authors, however, made these claims without analyzing the initial conditions of the two groups in Taiwan in the 1950s; nor did they analyze structural factors – such as family status or state educational policies – which would largely determine subsequent mobility paths.

Other authors look for structural factors to explain why Chinese people strive to be proprietors. Donald DeGlopper found in a study conducted in Lukang that almost all proprietors were working as craftsmen before they set up their own business. He also found that those who desired to be ‘independent’ or ‘mobile’ were also those who had been hired in small factories where employment was unstable (DeGlopper, 1979: 304). Another more comprehensive study by Shieh Gwo-shyong found three factors determined the likelihood of a blue-collar worker becoming a boss: the logic of dependent capitalism, the specific labor regime in Taiwan, and a worker’s age and marital status (Shieh, 1989). Though DeGlopper and Shieh have noted the mobility paths of laborers, both of their studies are subject to criticism for not differentiating Taiwanese from Mainlanders.

Based on the above review and on the following research, I argue that different ethnic groups, facing different structural restrictions – particularly family socioeconomic background and educational policies shaped by political and economic structures – proceed along different mobility paths.

Education, Occupation and Ethnicity

Let us first examine the distribution of these two groups across different industries after the 1960s. To exclude the effect of older generations on occupational distribution, I will examine the occupations of those who entered the formal labor market after 1960. Table 1 shows people who were first formally employed in the 1960s, 1970s and 1980s. Of these people, 50%, 51% and 54% respectively joined the secondary industry workforce, and 38%, 27% and 45% respectively joined tertiary industries in the 1960s, 1970s and 1980s.[5] This trend reflects rapid industrialization in the period between 1952 and 1992: the share of agricultural output in the Gross Domestic Product fell from 32.2% to 3.5%, while industrial output increased from 19.7% to 41.4%, with tertiary output increasing from 48.1% to 55.1% (CEPD, 1994: 42). The fast-growing industrial sector, especially the manufacturing sector, absorbed most of the new labor.

Table 1 about here

From this table it can be seen that more Mainlanders were first employed in tertiary rather than secondary industries, while first employment for Taiwanese tended to be in secondary industries. The data reveal that the Taiwanese were, on average, 3.4 years younger than Mainlanders when they commenced formal employment. Two explanations might be offered for why Mainlanders entered the formal labor market later than Taiwanese. One is that many could not find formal employment after graduating from school, and so either remained unemployed or at best found informal employment, which resulted in late entry into the formal labor market; a second possibility is that they stayed in school longer than Taiwanese. We find that the Mainlanders obtained an average of 2.4 more years of education than Taiwanese, though this difference narrowed over time.[6] This supports the second possibility raised above, i.e., Mainlanders stayed in school longer and so entered the formal job market later than Taiwanese. Their advanced education also explains why most gained employment in the tertiary sector. Although a supervisor in secondary industries requires more education than a clerk in tertiary industries, and although hawkers and pedicab drivers in tertiary industries are not highly educated, tertiary industries usually require a higher education. After controlling for ethnic background, I found that the number of years of education largely accounts for the different rate at which Taiwanese and Mainlanders entered different economic sectors. When this educational difference is taken into account, the difference in the chance of the two groups entering tertiary industry narrows considerably.

Table 2 shows the ethnic distribution for the self-employed (including employers) across time. The Mainlander percentage is consistently lower than their sample distribution over the past forty years. In other words, Mainlanders tended to work as employees instead of running a shop or factory. If the more highly educated are more likely to be successful in business because of their knowledge, why did the more highly educated Mainlanders not have a higher probability of becoming self-employed or an employer? This problem will be addressed below.

Table 2 about here

From the survey data, we note three features. Firstly, the Mainlanders’ education is of a consistently higher level than that of the Taiwanese. Secondly, the proportion of Taiwanese in the self-employed or employer categories is much higher than that of Mainlanders. Thirdly, Mainlander occupations tend to be “white-collar.” In the following sections, I will discuss how occupational distribution has been affected by political, economic and social forces and the differential “ethnic” effect of education.

Education After the War

According to human capital theory, education is a kind of capital that one can accumulate to facilitate finding employment (Becker, 1975). What one spends on education will be rewarded through a higher income or prestige in a future job. Related to this approach is the issue of who pays for the cost of this education. It is usually not the educated individual who pays, but instead his/her family and the government. Educational inequality can be reproduced through the effects of socioeconomic inequality, as many findings have shown.

In other words, we have to distinguish between three different kinds of educational costs: one is the opportunity cost of getting an education, another is the tuition fee paid by the family, and the third is the cost covered by the government. Meritocracy is not the only principle underlying the education market. Social and political forces also intervene, which may reduce or reinforce the effects of personal merit. In this section I will discuss three features that characterize postwar education in Taiwan: firstly, the hierarchical ranking of schools; secondly, the uneven distribution of educational opportunity between Taiwanese and Mainlanders due to quantifiably unequal socioeconomic status; and thirdly, tight government control of the education system. The ranked school system intensified competition among students to enter a better school, and restrictive government education policies before the 1990s reinforced this trend. Therefore, attending a buxiban (“cram school”) became a preferred method of beating the competition. The extra cost of attending a cram school, though affordable, is heavy for a family with an average income. Mainlanders, generally positioned with a higher socioeconomic status, thus tended to see their children enjoying an advantage in competing for educational opportunity.

The Hierarchically Ranked School System

National Taiwan University has traditionally been the best university in Taiwan in the eyes of the general public with Tsing Hua University or Chiao Tung University also highly regarded, followed by other public and then private universities. Reputations of high schools are based on their ‘examination passing rate’ to university. The Taiwan system resembles the Japanese education system, but is more centralized because of streamlined secondary and tertiary matriculation requirements and state education policy[7] (Dore, 1976: 46-50).

In the entrance exam, students from the few ‘superstar’ schools have a higher probability of matriculating. For example, the overall entry rate into tertiary education in 1981 was about one-third of applicants, but the entry rate for the elite Chien-kuo Senior High School was over 95%. In 1971, only two thirds of graduating junior high school students could enter senior high school or junior college. The other third had to enter the labor market. In 1981, 34 out of 187 academic senior high schools (18.2%) saw none of their students pass the university examination. Fewer than 10% of graduating students from another 76 schools passed the exam. In other words, 58.8% of the total number of academic high schools do not fulfill their function as preparatory schools for university (Zhang Chunxing, 1985:407-8).

Because of the stratified school system, a vocational education has become the ‘least’ prestigious. For those who enter high school, the first choice is normally an academic high school. Junior vocational schools, except for a small number of prestigious institutions, are usually the last resort for those who are bright and can afford further study. As David Schak has found, many students from low- and even middle-income families choose vocational schools, unless a student is very bright and can enroll in a public school (Schak, 1989: 365). A survey conducted by the Ministry of Education in 1971, the first year to produce graduates after the change to a nine-year compulsory education program, showed that 85.21% of graduating junior high school students in Taiwan wished to study further. Most of them would have liked to study in an academic senior high school, and few preferred a vocational junior college or high school (Sun Bangzheng, 1971:300). In 1984, the two ‘best’ vocational senior high schools in the Taipei area could recruit only 40% of their student quota. The situation in private vocational senior high schools was worse, with over seven thousand vacancies (Zhang Chunxing, 1985:410).

This hierarchy of schools reflects occupational prestige. Students graduating from more prestigious schools obtain more prestigious employment. From this perspective, education becomes a selection mechanism for the business community to determine who enters which job and what rank. Thus, young people spend their energy trying to enter a ‘better’ school.

To enhance their competitiveness, students have to start their preparation as soon as possible. Family socioeconomic background is crucial at the very beginning. In colonial times, education was a privileged mobility path for upper-class Taiwanese. After the war, increasing numbers of children were able to attend elementary and junior high school. Elementary education was no longer a 'luxury', but it was still difficult for students to enter junior high school before 1968, the year when nine years of education became compulsory. This difficulty was due in part to school capacity and in part to economic difficulties faced by the individual family.

In 1955 and 1956, Hsinchu County and Kaohsiung City experimented with a system in which prospective junior high students were accepted without taking an entrance exam. The survey reported that only 55% of graduating students in Hsinchu and 62% in Kaohsiung continued their studies (Wang Zhiting, 1958:350). The report did not say why 40% of pupils did not want to study further, but from my interview data, every interviewee from that generation who today owns a small and medium-sized enterprise (SME) emphasized that his/her family was too poor to allow him/her to attend junior high school. Some of these informants had to work as apprentices and at the same time went to junior high school at night. Table 3, which is derived from raw survey data, shows the average number of years of education by date of birth and occupation of father. The relationship between the father’s occupation and length of education is clearly the same as in the Japanese era (1895-1945). It shows that in the 1950s and 1960s it was still not easy for a poor family to provide their children with better education. The entry rate into junior high school in 1955 was only 39% (Yu Shulin, 1977: 350). As can be seen, children of farmers and manual workers lost out at the start. After the government extended compulsory education from six to nine years, the difference narrowed, but these groups remained less educated.

Table 3 about here

The combined effect of this hierarchical school system and differential socioeconomic status was the emergence of private ‘after school classes’. A wealthy or above-average income family can send their children to private elementary and junior high schools famous for their ‘examism’. These schools only teach subjects that are to be examined and tend to ignore all other activities (Sun Bangzheng, 1971: 203). Another phenomenon is the hiring of private tutors who teach children after school. In Taiwan, in 1995, the private tutoring wage per month at four hours per week was about US$230.[8] In 1995 the average wage was US$1,200 a month, so a family with an average income could not afford private tuition. A cheaper way was to send children to cram schools, a more common option for middle income families, but as a tool for learning they are far less effective.

Socioeconomic and Educational Differences Between Taiwanese and Mainlanders

Can the persistent difference in educational achievement between Taiwanese and Mainlanders over the past forty years be called institutionalized discrimination or just the result of difference in socioeconomic family background? I will review the evidence and then discuss the causes of this difference.

Table 4 shows “occupational prestige” scores by ethnicity. The occupational prestige of Taiwanese fathers increased steadily from 29.96 in the 1950s to 39.76 in the 1980s. Over the same period, the Mainlander figure remained at around 50. In short, Mainlander fathers enjoyed higher social prestige than Taiwanese, but this difference has decreased progressively over time.

Table 4 about here

From the winter of 1965 through the spring of 1966, Richard W. Wilson conducted research at selected elementary schools in Taiwan. He evaluated the socioeconomic status of the children at three schools: a city public school, an urban private school and a rural public school. Fathers of the children at the urban public school generally held medium-to-low occupational positions with corresponding incomes, while at the private school they enjoyed medium-to-high incomes and the rural public school low-to-medium incomes (Wilson, 1970:149-152). The status classification, though based on children's vague responses and the author's own subjective estimates, did reflect the status difference of different schools. Normally the educational facilities in urban areas are better than in the countryside, and those at private elementary schools are better than those at public schools. Many private elementary schools are the so-called 'superstar schools'.

The ethnic ratio (Mainlander to Taiwanese) in the three schools was 47:53, 92:8, and 40:60 respectively. The private schools mainly enrolled Mainlanders, and their fathers' occupational status was 'medium-to-high'. Wilson's description is in agreement with the class structure Wang observed in another paper, where he argued that Mainlanders (except for serving or retired soldiers) held better positions in the state bureaucracy and state enterprises or ran businesses in the 1950s (Wang, 1999).

The most 'competent' students, as was noted earlier, could enter academic high schools. A survey by Stephen M. Olson in 1968 showed that the numbers of Mainlander students in one municipal academic senior high school was much higher than at a private academic and commercial-vocational school. Education levels for fathers of students were highest at the municipal academic school, followed by the private academic and the private commercial-vocational schools. A correlation between ethnicity and education of fathers was clear. The effect of ethnic socioeconomic difference could therefore be seen in junior high school (Olson, 1972: 261-295).

Entering a ‘better’ academic senior high school provided a better chance of entering university. A survey conducted by the Stanford Institute in 1961 showed that 37% of students graduating from universities between 1957 and 1959 were born in Mainland China (N=850), while only 6% of those graduating from senior vocational high school were Mainlanders (N=1,994) (Stanford Research Center 1962: 128).[9]

The statistical survey of education of Taiwan Province revealed the same situation. In 1967, 'when Mainlanders constituted about 14% of the total population and probably a somewhat lower proportion of school-aged youth, almost a quarter of the island's non-vocational high school students, 30% of its college students, and almost 40% of those enrolled in the top-ranking national universities were Mainlanders (Appleton, 1976: 706).

Tables 5 and 6, again derived from raw national survey data, show the effects of ethnic, demographic and stratification factors on education and occupational attainment. Table 5, at least at first glance, suggests that ethnicity seems to be an important factor affecting one’s educational achievement, even after considering gender and date of birth. But after allowing for stratification variables, i.e., education of the father and mother, its effect is reduced sharply from 2.24 years to 0.79. This means that the total effect of ethnicity on education was that the Mainlander, on average, was 2.61 years more educated than Taiwanese, but also that the indirect effect, composed of the demographic effect (0.37) and parents’ educational background (1.45), contributes most to the total effect.[10] If we include the variable “father’s occupation,” which contributed 0.36 to educational attainment, then the ethnicity variable becomes insignificant. In other words, the ‘obvious’ ethnic cleavage phenomenon in educational attainment is in fact caused by stratification variables. We have found that the educational difference between Taiwanese and Mainlander was narrowing, and Table 4 also shows the occupational prestige difference narrowing. We can expect that accompanying the improvement of the Taiwanese populace’s socioeconomic status, the ethnic educational difference would be gradually eliminated.

Table 5 about here

People without higher educational credentials have to enter lower prestige occupations. Table 6 shows the effect of different variables on occupational attainment. As in Table 6, ethnicity seems at first sight important to occupational prestige. The prestige score difference is 8.76, which represents the total effect of ethnicity on occupational prestige. After factoring in demographic variables, its effect is reduced by only a small amount. But after adding the variables of parents’ education and the occupation of the father, the ethnic variable and demographic variables both become insignificant. Yet all these variables together can explain only 24% of the total variations. If we include one’s own education in the regression equation, R square increases from .24 to .42, that is, it almost doubles, and education of parents becomes insignificant. This means that one’s occupation status is very highly correlated with education. With each additional year of education, her/his occupational prestige score increases 2.11 points. The direct effect of parents’ education diminishes to insignificance because its effect on an informant’s occupational attainment is transmitted through one’s own education. So we can surmise that the occupational difference between Taiwanese and Mainlanders will be eliminated if the gap in the educational difference between the two groups diminishes in the future.

Table 6 about here

Why then does the father’s occupation have a significant influence on his children’s occupations? Analysis of the 1992 survey shows that 61.8% of respondents got their first jobs through personal networks, e.g., were introduced by family, friends or relatives, while only 38.2% found their jobs through an official employment office, newspaper advertisements or by passing a state examination.[11] If so, a father’s social network could be an important resource in finding one’s first job, and there is a tendency for that social network to be rooted amongst those of similar social background. If one’s child is incorporated into his/her network, s/he might find a job similar in prestige to that of his/her father.

The State’s Education Policy

Ethnic, demographic and stratification factors explain 50% of the variation in a person’s education, as Table 5 shows. The other 50% can be attributed to a student’s own efforts or talent, state education policies, and so on. In modern society, the state plays an important role in educational planning, which sets constraints on the educational opportunities of its people. Education is incorporated into the modern political system to achieve certain ‘national goals’ set by the state, such as economic development and the “recovery of the mainland.” Inevitably, then, education is a function of economic and political programs.

The postwar, KMT government-controlled education system has had two notable features: (i) the government used education as an ideological tool; and (ii) educational development was based on economic planning.

Education is the easiest and cheapest means to steer ideology. The KMT regime attempted to legitimize its rule as the sole ‘representative’ of China, and it felt it had to tightly control any ‘heretical’ discourse challenging its claims (Wakabayashi, 1987: 368).[12] Private education was seen as a potential ‘challenger’, so the establishment of new schools was put under strict control. This policy meant that although there was a high social demand for higher education throughout the postwar period, only a small number of private schools was allowed to be established. The government argued that there were not sufficient funds to expand higher education, but at the same time it did not allow the private sector to do so. It was not a problem of any lack of private capital. For example, after an ‘open door’ policy was proclaimed to set up private junior colleges in 1964, forty-one new private colleges were set up in five years, increasing the numbers in Taiwan from 8 to 49 (Jiaoyubu, 1976: 677; Chen Shunfen, 1991: 231). This earlier policy of limiting higher education was aimed at political and ideological control, but a side effect was to intensify competition between students.

Though the enrolment rate of graduating elementary school students in junior high school increased from 33% in 1953 to 62% in 1967 (the year before the introduction of a compulsory nine-year education), the absolute number of ‘failing’ students in junior high school exams increased steadily from 89,367 to 134,076. ‘After-school lessons’ became pervasive, which the education ministry in 1976 admitted was ‘one of the most serious problems in recent educational development’ (Jiaoyubu, 1976: 227-8). A rich family in this competition was more ‘competitive’ because it could better afford the after-school lessons. Mainlanders generally held better occupations, so they could better afford the extra education cost for their children. But the Mainlander children were not only helped by their higher socioeconomic family status, but were also heavily subsidized by the government. All children of state employees received exemptions from tuition fees and were in line for scholarships at all educational levels, while other students had to pay tuition fees above the junior high school level, which was not part of the compulsory education domain (Tsai and Chiu, 1993b: 218). Civil servants, mainly Mainlanders, benefited the most from this subsidy. From 1950 to 1974, only 59% of newly recruited civil servants who passed the state examination were Taiwanese, while 41% were Mainlanders, a figure considerably higher than the Mainlander proportion (15%) of the population.[13] Thus, Mainlanders received a significantly disproportionate subsidy from the government. At the same time, a special tax based on agricultural income was levied to finance educational expenditure (Wang Zhiting, 1959: 176). In other words, the government’s financing policy saw educational resources flowing from a disadvantaged class, the (mainly Taiwanese) peasants, to a privileged class, the (mainly Mainlander) public servants. Little wonder, then, that Shu-ling Tsai and Hei-yuan Chiu reported a ‘farming background exerts a negative impact on educational attainment at all [school-level] transitions’ (Tsai and Chiu, 1993b: 218).

Another important influence was related to economic planning. In the 1950s and 1960s, vocational education dominated mainstream educational thinking all over the world, which reflected its function in promoting economic development. Taiwan was no exception (Yang Yirong, 1991: 136). The determining factor for the government to develop vocational education was a report by the Stanford Research Institute in 1962, which proposed that Taiwan should increase its skilled labor supply to promote rapid economic development (Jiaoyubu, 1976: 628). Since then, entry into academic schools has been strictly controlled. To increase the supply of skilled labor, the government encouraged the private sector to establish vocational schools to ease demand for higher education. In addition, the government set a target that adjusted the academic and vocational senior high student ratio from 80:20 to 30:70 (Gai Zhesheng, 1969: 135). This former ratio was successfully reversed to 32:68 in 1988 (Yang Yirong, 1991: 149).

This was a dramatic change. The effect of such an economic policy is not hard to imagine: fewer students were able to study at senior academic high schools, which in turn intensified the competition among students and among junior high schools to succeed in the student race to enter academic senior high schools. Analysis of the raw data contained in ‘The General Survey Data of Social Change in Taiwan, 1992-93’, yields an ethnic ratio (Taiwanese: Mainlanders) at different educational levels for the highest degree earned: for academic senior high school, the ration was 76:24; for junior college, 79:21; and for vocational senior high school, 87:13. The ratio deflates the number of Mainlanders in the case of academic senior high school because those who entered university or senior college were not included. The ethnic ratio for university is 74:26. The overall demographic ratio is 86: 14.

Education itself was sufficient to function as the main mechanism for reproducing socioeconomic inequality in a competitive education market (Bowles and Gintis, 1976). State policy could have reduced the negative effects of socioeconomic differences, but in Taiwan, educational policy has reinforced inequality.

Mobility Path Patterns Among SME Businesspeople

Workers of low prestige occupations have to ‘help themselves’ to become upwardly mobile. Laborers in various countries adopt different strategies to improve their conditions. As we have seen, over the past forty years Taiwanese have received less education than Mainlanders. In the following I will examine the mobility patterns of owners of small and medium-sized enterprises, concentrating on the self-employed in non-primary industries. I will sketch a general employment history, and then discuss the occupational category of the self-employed, which includes employers who hire workers, as well as the petty bourgeoisie such as small shopkeepers or taxi-drivers, who do not hire workers. Then I will discuss the structural influences on mobility patterns.

Education frequently determines one’s lifetime occupation. In Taiwan, occupational prestige scores in an individual’s work life have remained relatively stable, though most experienced at least slight upward mobility (see Table 7).[14] This suggests that social stratification in Taiwan was largely determined by schooling, with students classified through the examination system. Their future occupations were largely determined by the level of education they attained.

Table 7 about here

Table 8 provides a general sketch of employment history. The most distinguishing feature is that the percentage of self-employed increase over time. The typical employment history of the self-employed is: find a job in the labor market, then marry, and become a housewife (if female) or set up a small shop (if male). This tendency occurs across generations.

Table 8 about here

The number of self-employed in non-agricultural sectors increases as time passes, but does not change much for those in primary industries after marriage. With regard to the relationship between education and self-employment in the non-agricultural sectors, it is clear that the higher the education, the lower the percentage of those at that level of educational attainment who are self-employed. With regard to ethnic distribution, 23.2% of Taiwanese, excluding the self-employed in primary industry, are self-employed, while only 11.1% of Mainlanders are. In other words, Taiwanese with less education are more likely to be self-employed over time.

What sorts of jobs are held by the self-employed in Taiwan? Analysis of the 1992 survey data reveals 370 self-employed in non-primary industries. Of these, 142 (38.4%) hired workers to work for them and so could be called small employers.[15] However most of the non-agricultural self-employed simply ran a small shop of their own. Of the remainder, 116 out of 370 (31%) are wholesale or retail trade operators, 35 (9.5%) are street vendors, 19 (5.1%) are taxi-drivers and 18 (4.9%) are restaurant owners. In other words, more than 50% of the non-agricultural self-employed were earning their living based on their ex ante labor which depends on market supply and demand (Offe, 1985: 28).

These employers have the same sort of mobility path as the person who runs a small shop. The first jobs most employers held were as employees (71.8%), but by the time they married the percentage who remained so had dropped to 37.3%, while the percentage of self-employed increased from 11.3% to 40.8%.

Table 9 shows that the hiring of workers has nothing to do with age; the clear pattern is that the more educated are more likely to hire people. This is consistent with a survey conducted in 1996 by the Directorate-General of Budget, Accounting and Statistics. This survey found that employers are more educated than self-employed workers. The average number of workers that employers hire is 12.3, whereas for the self-employed that number is only 2.8 (Directorate-General of Budget, Accounting and Statistics, 1997:171).

Table 9 about here

Even though less educated Taiwanese are more likely to be self-employed, their occupations center on low prestige jobs such as street vendors, taxi drivers or small shop owners. One possibility is that the less educated do not have sufficient human capital to learn new skills to compete, so their work is mostly dependent on subcontracting. Those who received a mid-level education unable to find good jobs or be promoted to a high position in an organization, but they were able to learn new skills with the help of what they learned in high school. Those interviewees who emphasized competitive production skills were usually those with vocational senior high school credentials or higher, while those who talked about their guanxi (connections) with others were usually less educated. Here it seems that education played a role in promoting economic success.

In summary, we find that:

1. Occupational distribution has been largely determined by education, and the education system in Taiwan functions as a mechanism that reproduces inequality in social status.

2. People with different educational levels have different mobility paths. The general mobility pattern in Taiwan is for the less educated to work for others, then later look for opportunities for self-employment, usually in a marginal, low prestige business. For the higher educated, the pattern tends toward obtaining a job with prestige as an employee in an organization, and to stay in that position, hoping for promotion.

3. Employers are distinguishable from the self-employed, and the data show that the former are generally better educated.

4. Mainlanders are better educated than Taiwanese, and they are less likely to be self-employed. The ethnic cleavage in occupational distribution is chiefly the result of educational factors and guanxi.

Obviously the prevalence of small and medium-sized enterprises in Taiwan is not induced by any ‘cultural mentality’ or familial and community ties. Different educational levels set general limitations on upward mobility, and for the less educated, one of the means to overcome this limitation is to withdraw one’s labor power from the wage market and to start a small business.

Sociopolitical Control of Collective Action

An interesting question is why the less-educated, i.e. manual laborers, adopted the strategy of abandoning their class status individually instead of trying to improve their conditions collectively. The answer seems to lie in structural factors that influence the strategies manual laborers adopt.

Education as a kind of ideological control

Education is usually considered to be a means of combating socioeconomic inequality. Yet the accepted values regarding upward mobility have direct and indirect social control effects. The direct effect is that the examination selection system is widely and unquestioningly accepted as legitimate. People are apt to believe that it is the most fair and efficient way to select elites for the management of society, and that it is not prejudiced against poor families, as every student receives the same form of assessment (Yang and Ye, 1984: 375).

Such a widely accepted belief, that educational opportunity is equally provided through such exams (as well as the belief that a higher education is important for a successful career),[16] produces an indirect effect: those who are not able to pass entrance exams and obtain higher education are deemed failures in Taiwan’s very competitive society. This is not only the social image of the less-educated (Yang and Ye, 1984: 372-373), but also a self-image in many if not most cases. These social and individual images are congruent because the less educated can only find low-prestige jobs, which reinforces the negative image of the less-educated. In the 1992 survey, 78% of interviewees were not satisfied with their own educational performance.[17] In Taiwan, teachers assess everything, including “morality,” on a one hundred percent scale, and those who receive higher marks are always regarded as more intelligent, hard-working and as models for other students. In junior high school, almost every school classifies students in two ways: a “front” section which is established to drive students along to pass the senior high school examination, and a “back” section, which is said to accommodate those students who do not want to continue their studies.[18] Analysis of the 1992 survey data shows that among those whose highest educational attainment was an academic high school degree, 21% had been placed in the ‘back section’; the figure for vocational senior high school was 32%, while for those who went on to receive a university degree it was only 2.0%.[19]

In sum, education is recognized as a mechanism to distinguish the ‘most capable’ people from the ‘incapable’ on the one hand, and to legitimize the existing social hierarchy on the other.[20] One case illustrates this popular attitude towards the function of education. A colleague of the author had been explaining to students at a junior college why the percentage of Mainlanders attending tertiary education is much higher than Taiwanese. She asked students to write an answer to the question in a final examination. Even after they had been told about the subject in class, a number of students replied: because Mainlanders are more clever than Taiwanese.

People without higher educational credentials must try to find alternatives if they seek to be upwardly mobile. Since the education system in Taiwan is inflexible, it is almost impossible for someone who is out of the school system to return to school, so virtually nobody can achieve upward mobility via this route. For example, one interviewee’s daughter observed that her father, born in 1938, told her that his first aspiration had been to study at university. Experiencing financial difficulties at the time, however, he was forced to abandon that dream. He said that since it was impossible to attain a higher education, the alternative way open to him to raise his status was to strive to make money. Opening a shop seemed to be a strategy to climb the social ladder. Ka (1993: 31) also found that ‘it is impossible to promote the less-educated in a bureaucracy, so 82% of the informants believed that to be one’s own boss was the best way to be successful.’

But even if one is motivated to improve one’s inferior social and material status, why does s/he adopt individualistic measures instead of collective action? We must therefore examine the mechanisms that have excluded this possibility of collective action.[21]

The State-Controlled Labor Regime

Taiwan has been ranked as a soft authoritarian regime prior to its democratization in the mid-1980s (Winckler, 1988: 151-171). This means that any collective action, including labor movements, was viewed as subversive, a threat to social stability, and was kept under surveillance and suppressed when necessary (Wang and Fang, 1992: 9-14).

Methods to control the labor movement in Taiwan were established in the period of conflict between the Chinese Nationalists and the Chinese Communists, after the former retreated to Taiwan. The implementation of martial law consolidated the control of labor by the party-state (Xu Zhengguang, 1988: 4). A series of laws were implemented to control labor, including a Trade Union Law, the Labor-Capital Disputes Law, the National Mobilization Law, martial law, Provisions for Controlling Agriculture, Industry and Commerce in the Emergency Period, etc. The most important measure was to preempt workers’ collective-dispute rights. According to clause six of the Trade Union Law, industrial laborers over twenty years old in a district or within a factory could organize an industrial or occupational trade union. Until 1988, however, 85% of Taiwan factories employed fewer than thirty workers, and the laborers in these firms were excluded from the right to organize a union (Wang, 1988).

Even when laborers were eligible to organize a union, government bureaucracy would combine with employers to prevent the establishment of an ‘unpopular’ union. In one case, laborers working in a shoe-manufacturing company wanted to establish a union. As soon as they had submitted the application form, the government office in charge of the matter notified the company, and the company mobilized pro-company workers to establish a rival union. The office forged the date of receipt on the first group’s application, thereby invalidating it, since the law provided that only one union could be established per factory or area. Apart from this arbitrary administrative power, the law forbade any coalition of local level unions. Local level unions needed to be subordinate to a higher level union, which was always controlled by the party-state (Hsiao, 1992: 155).

Because Taiwan was effectively a one-party state, the state administration was in fact controlled by the party, and orders from the party superceded administrative rules (Bosco 1992: 174). For example, Kaohsiung fishermen submitted an application to organize a union, which was ratified by a government office, but a KMT cadre disagreed and overthrew the ratification, which caused disputes between fishermen and the government for several years. Further, a KMT cadre was placed inside every large enterprise, located in the personnel office, to supervise labor/management relations (Xu 1988:5)

Through controls such as these, trade unions in Taiwan had no power to independently strive for workers’ welfare. Instead, they tended to become peripheral organizations within the party-state. If working conditions became unbearable or one’s income remained low, the only option was to resign and find a new means of employment.

This kind of control is a precautionary means for bringing social discontent under a net of strict surveillance. However, personal social networks can also play an effective and less costly means to drain off that discontent.

In Taiwan, most people seek employment through personal networks. Under these circumstances, all labor-capital conflict is neatly transformed into ‘personal’ relationships that can be mediated by friends, family or relatives. One example of the effects of such personal mediation can be seen in the strike against the Far Eastern Chemical Enterprise in May 1989. Many workers had found their jobs in this company through the help of relatives or friends, and, during the strike period, some workers had to resist pressure from their parents and relatives, who persuaded them not to join the strike (Hsia, 1993: 257). Another case involved workers who joined an independent union and came under a lot of pressure from parents and friends who had introduced them to work in the company. ‘In the process of labor-capital conflict, the employer mobilized all social relations surrounding the union leaders to press them to stop fighting. In addition to their parents, the main pressure came from those friends who had introduced them for them to work here’ (Hsia, 1993: 217). Such social networks play a key role in preventing the formation of collective class-consciousness; in other words, personal social networks mitigate against worker collectivity and atomize the workers. For there to be any successful collective action by laborers, it must go beyond the reaches of the social network.

When conflict escalated into collective action, local politicians intervened. As a rule, local businesspeople had good relationships with local politicians. Local politicians and factions form on the basis of social networks, especially in rural areas, and bosses would ask politicians to mobilize their social networks to ‘de-mobilize’ the involved parties (Jacobs 1980:78-82).[22]

In his research on social protests in Taiwan in the 1980s, Wu Chieh-min found that local faction politicians, who were mainly representatives from the city, county, and urban and rural townships, played key early roles vis-a-vis collective action. They intervened in collective action through mediation or on-the-spot consultations. Social conflicts were then placed in the context of patron-client networks, which offered a negotiation ‘pipeline’ between protesters and the company, and between protesters and local government. Such a patron-client network could function because protesters and companies were mindful of harassment from intelligence agencies, and they therefore preferred local politicians, with whom they were more familiar, to ‘serve the people’ (Wu Chieh-min, 1990: 80-135). As James C. Scott has observed, ‘the effect of machine rule under universal suffrage is to submerge growing collective policy demands with immediate payoffs, thereby retarding the development of class-based political interests among the lower strata’ (Scott, 1972:151).

From the above analysis, we can understand why workers in Taiwan preferred to adopt the strategy of abandoning their class status individually rather than the strategy of class mobilization to improve their lot. Paid employees comprised 67.5% of the total labor force in 1996,[23] but almost every collective problem became transformed into an individual problem or became smothered through the mechanisms described above, i.e., the state administrative and policing apparatus preempted the emergence of any possible collective action, and, with the help of local factions, could ‘demobilize’ any collective action.

Conclusion

In general, educational opportunities correlate highly with family socioeconomic background. In Taiwan, as noted, this socioeconomic background is also related to ethnicity. Mainlanders tend to be more highly educated than Taiwanese, though the gap has been narrowed. To explain the phenomenon, I traced back to the 1950s the fact that the postwar political structure determined occupations and the shape of the educational system. Mainlanders were mostly placed in public offices or public enterprises, which gave them better economic benefits. Most Taiwanese were farmers at that time. Such an occupational difference entailed economic inequality between the two groups, which in turn affected educational opportunities for their children. Such an unequal opportunity structure was reinforced by the government’s education policy. Educational institutions were tightly controlled according to economic planning, which meant students had to face strict examination selection. The upper classes’ economic resources helped their children to pass the examinations, while the government heavily subsidized the education of the children of public servants. These government policies benefited Mainlanders more than Taiwanese.

Since children from better socioeconomic backgrounds obtained a better education, they could get better jobs in the labor market. The strategy adopted by the less educated was to try to escape their class milieu, where possible, and try to become their boss, which was the best path of upward mobility open to them. Such a strategy was nurtured by Taiwan’s specific political and economic context. Two main reasons for a lack of alternative paths were: the widely accepted ideology of education as a mechanism to distinguish the ‘capable’ from the ‘incapable’; and the state-controlled labor regime’s forestalling of the emergence of a labor movement. With these mechanisms in place, and if employment was unstable, wages low, and working hours long, their only chance was to leave wage employment. As Hill Gates has noted, capitalism ‘offered a social model of upward mobility based directly on wealth rather than on connections with the state through the highly limited channels of degree- and office-holding” (1987: 261). Thus, the general mobility pattern in Taiwan for the less educated has been to get a wage job first and then to become self-employed where possible. The booming world economy in the 1960s gave entrepreneurial blue-collar workers, mainly ethnic Taiwanese, a chance to move out of this ‘inferior’ class.

NOTES