Social Mobility in Modern China

Transcription

1 Social Mobility in Modern China Jing YANG University of Oxford Most socialist countries in East Europe and Asia implemented economic reforms in the early 1980s. Among them, China is a unique case in that it has achieved not only fast and continuous economic growth, but has also managed to retain the socialist regime. Together with impressive economic success, the Chinese class structure has transformed from a simple, bounded, and rigid occupational hierarchy under Mao, to an open, modernized, and evolving system in the early industrialization period of the nation. The situation in China is, however, complicated due to several institutional barriers, such as the household registration system, cadre system, work-unit system, and one-child policy, all of which still persist and are exacerbating the gap between urban and rural areas. These social conditions are influential in shaping social mobility in China. This paper investigates the patterns of intergenerational mobility in contemporary China using data from Life Histories and Social Change in Contemporary China, The trends of class mobility in terms of absolute and relative rates are analyzed based on two different class schemata: the 9-class schema developed by the Chinese Academy of Social Science (CASS); and the CASMIN developed by Goldthorpe and Erikson. (Erikson and Goldthorpe 1992; Goldthorpe, Llewellyn, and Payne 1987) By the late 1990s in China, the total mobility has increased, but the trends of vertical mobility fluctuated, corresponding to structural transitions. Social fluidity is changing, but from the results we cannot see if the class mobility is becoming more open or more restricted. Finally, a modified multi-matrix model of social fluidity is applied to intergenerational mobility tables of men and women. The results show how the patterns of class mobility have changed during the market transition period. 1. Introduction Studies on class structure and social mobility in industrial societies are relatively abundant. As liberal theorists have suggested, the processes of industrialisation and economic expansion have brought enormous changes to occupational structure, particularly the growth of the middle class and the decline of the agricultural sector. Structural transformation led by the development of science and technology requires a more complex process of the division of labour. Thus, the more developed the society, the more mature and complicated the occupational structure becomes, and the more vital and open the social mobility of the society becomes. (Lipset and Bendix 1959) When the society becomes more open, liberal scholars expect to find an increasing total mobility 1

2 (upward mobility in particular), and more equal opportunities for people of all social origins to move between classes. In contemporary studies on social mobility, there are two basic measurements. The first measure, absolute mobility, is where mobility is understood simply as movement between class origins (the social class in which someone was brought up, or empirically, the father s occupation when respondent was 14) and class destinations (the class the respondents occupied at the time of the survey). The second measure, relative mobility, also known as social fluidity, concerns the relationship between class origins and current class position. More specifically, it is based on the comparison between people of different class origins and of their chances of being found in one destination class rather than another. Therefore, from absolute mobility we can observe the general trends of people moving between classes, but from the study of the trends in social fluidity, we can see if the society has become more open in its patterns of class allocation or not, regardless of any structural transformation caused by economic, political or ideological factors. (Breen 2004) Regarding social mobility in industrial society, FJH hypotheses (Featherman, Jones, and Hauser 1975) revise the argument of Lipset and Bendix by focusing on relative rates of mobility across countries rather than the observed mobility rates. They discovered that the opportunities for movement between classes are more equal for people of different social origins, and suggested that because most societies in the West have undergone the process of industrialization (since the 18 th century), the cross-national mobility rates are also expected to converge and reach a similar level. All the results led to their major statements that the inequality in mobility chances ascribed from class origins (i.e., relative mobility chances) will be roughly constant across the nations. Several studies have proven the broad similarities in social mobility patterns between Western countries, though variations do exist under different social contexts. Recently, Breen et al (2004) reviewed the trends of social mobility in 11 European countries 1 by using the class schema developed from CASMIN 2. They argued that the 1 The countries they studied include Britain, France, Ireland, West Germany, the Netherlands, Italy, Sweden, Norway, Poland, Hungary, and Israel. Their data covered the period from the early- or mid-1970s to the mid- or late-1990s. 2

3 convergence in class structures has been driven by some internationally consistent trends, such as the growth in the service class (I+II), and the decline in manual work, especially unskilled labour in particular. Among women, increased rates of labour force participation have been associated with a reduction in international variation, as more and more of women enter occupations in the white-collar classes (I+II and III). The biggest cause of this convergence has been the declining significance of the agricultural classes, especially in Poland, Ireland, and Hungary, where a large farm sector persisted until the last quarter of the 20 th century. (Breen 2004) The authors believe that societies follow a developmental path that incorporates two major transitions: first, from an agricultural to an industrial society, and second, from an industrial toward a post-industrial society. The consequences for the class structure of the former transition are a decline in the proportions in classes IVc (self-employed 2 CASMIN is the class schema developed by Goldthorpe and Erikson from The CASMIN project (Comparative Analysis of Social Mobility in Industrial Nations). The Sevenfold version of CASMIN is commonly used in studies on social mobility (from Social Mobility and Class structure in Modern Britain): Class distinctions are classified as follows: I: All higher-grade professionals, self-employed or salaried; higher grade administrators and officials in central and local government and in public and private enterprises; managers in large industrial establishments; large proprietors. People of Class I have secure jobs and high income, which is likely to rise over their lifetime. Their positions involve exercising authority and expertise within a range of discretion, and thus offer considerable autonomy and freedom from control by others. Therefore Class I is taken as middle or higher level of service class of modern capitalist societies. (raised by Dahrendorf, following Karl Renner) II: Lower-grade professionals and higher-grade technicians; lower-grade administrators and officials; managers in small business and industrial establishments and in services; and supervisors of non-manual employees. Class II can be regarded as complementing Class I of CASMIN in representing service class as a whole. III: Routine non-manual workers: clerical employees in administration and commerce; sales personnel; other rank-and-file service workers. Class III is to be regarded as separate from the service class, and might rather be seen as representing a white-collar labour force that is associated with, but marginal to the service class. IV: Petty bourgeoisie: small proprietors (including farmers and smallholders), with and without employees; self-employed artisans; and all other own account workers apart from professionals. Economic security and prospects of this class are generally less predictable than in the case with salaried employees. V: Lower-grade technicians; supervisors of manual workers. VI: Skilled manual wage-workers in all branches of industry (including apprentice) VII: All manual wage-workers in industry in semi- and unskilled grades; and agricultural workers. Class VI and VII can be taken as the working class because both their market situation and work situation are subordinate to the other classes. 3

4 farmers) and VIIb (unskilled manual-working class) and growth in the remaining classes, especially (among men) the skilled manual working classes (V+VI and VIIa). The transition to a post-industrial society witnesses a further decline of working class and agricultural labour (V+VI and VIIa) and the growth of white-collar workers (I+II, III). Still, the move away from agriculture is crucial for this transition from the study of absolute mobility. Moreover, the authors found that the class mobility in countries such as Hungary and Poland, where the class distribution in the 1970s differed from others, deviated from the common pattern in many respects. 3 For instance, as far as absolute mobility is concerned, data from Hungary shows a higher rate of recruitment to all classes from farm origins; a lower rate of recruitment from skilled worker-class origins; a lower rate of outflows from class origins to service-class positions; and a higher rate of outflows to skilled or unskilled working class positions. 4 Meanwhile, data from Poland shows that for both sexes neither late state-socialist developments nor early capitalist ones decreased downward mobility. Downward mobility increased, in fact, under late state-socialism as well as under early capitalism. This increase was more profound among men. In terms of upward mobility, the position of women seemed to fair better. Upward mobility was increasing both in and among women, but remained stable among men. Detailed outflows tell, in general, the same story. 5 For both countries, women are found concentrated in the white-collar class even more efficiently. When turning to relative mobility rates, Robert and Bukodi found no steady increase in the openness of Hungarian society: it fluctuated corresponding to different periods, and only the eldest cohort showed an increase in social fluidity, while it stayed constant for the younger cohorts. When a Hungarian version of the core model was applied to a mobility table in different time, they found an increase in hierarchical effect for men, and a significant increase in inheritance effects for both men and women. Therefore, the mobility chances of women did not deteriorate as much as did those of 3 Chapter 11, by Bogdan.W.Mach; and chapter 12, by Peter Robert and Erzsebet Bukodi. Breen, Richard "Social Mobility in Europe." Oxford; New York: Oxford University Press. Pp Ibid. 5 Ibid. Pp

5 men. 6 In Poland, the trends of social fluidity stayed constant for men, but for women, a clear tendency of increasing fluidity was found in both state socialism and capitalism periods. Moreover, when the core model of social fluidity was applied, the hierarchy effects were found generally more effective for men, while the inheritance effects were more effective among women. 7 Some scholars have also argued that the social stratification system in the advanced communist regime is not the same as it is in Western capitalist societies, where the hypotheses on trends of class mobility would apply. Goldthorpe (Goldthorpe 1964) suggested that the stratification system cannot be viewed only as the reflections of advancing technology and economy of a society. Instead, stratification is also affected by a range of other social factors, especially political actions and state interventions. Goldthorpe and Parkin (Parkin 1969) have summarized a number of arguments with regard to the class stratification developed in totalitarian systems of the communist regime. They argued that in the societies where the monopolistic political power is highly concentrated, the mobility patterns generated from the stratification system are fundamentally different from what is typically found in modern Western capitalist societies, nor could the mobility patterns can be captured by the concepts developed for capitalist societies. In this respect, China has entered the second phase of transition from a socialist to an industrial society, and the model of classlessness is no longer applicable to the emerging market economy. China has become a unique case as an industrialized, yet socialist state where the tensions between political power and a market economy have persisted throughout the transitional period. Inspired by Polanyi and Szelenyi s study on market reform in post-socialist countries in East Europe, Victor Nee raised the Market Transition theory (Cao and Nee 2000; Nee 1989; Nee 1991; Nee 1992; Nee 1996; Nee and Cao 1999; Nee and Lian 1994; Nee and Matthews 1996), thus initiating the debate on how a market economy altered income differences in China. He argued that market opportunities enhance returns to both education and entrepreneurial initiatives, and market transitions diminish the advantages 6 Ibid. Pp Ibid. Pp

6 that political elites derive from their redistributive power. Therefore, the inequality caused by rising returns to human capital and initiative are offset by decreased inequality on the basis of political capital. Nee mainly focuses on the role market forces play in shaping the paths of institutional change and stratification order. Andrew Walder provided more detailed analyses on interactions between a central government and local leading cadres on the development of local economies. Walder summarizes four models of elite mobility in post-socialist countries during a market transition period. Among his arguments, he stressed that redistributive power in China is still pervasive and that cadres are privileged to be entrepreneurs themselves, allocating economic opportunities to their relatives, or obtaining benefits from competition among entrepreneurs with limited resources. He also argued that the impact of market transition varies on different cohorts in rural and urban China during different periods. On the other side, many empirical studies have been completed by Chinese scholars. In 2002, the Chinese Academy of Social Science (CASS) conducted a nationwide survey on the changing life histories of Chinese people during the market transition period. Soon after gathering and analyzing the data, they raised the idea of a new class structure, redrew the class boundaries and released the results to the public. They outlined the class scheme based on the estimation of three social resources attached to occupations; Chinese people were sorted into a 9-class schema 8 (CASS02): Leading cadre, manager, private entrepreneur, professional, clerk, service worker, manual worker, and peasant. The release of this new class schema from CASS is no doubt a promising start for advanced studies on social stratification and class mobility in state socialism, which experienced fundamental economic reform and institutional changes for local researchers. The new class structure tends to be less politically sensitive considering China s special national circumstances. Referring to previous statistical records from five national censuses, CASS showed the changes of class structure across different stages of economic reform. (See Tables 3 and 4) However, we can see that its latent rationale is too general and mechanical to be efficient at this stage. Most crucial effects of social factors 8 Lu, Xue Yi "Report of Social Class Study in Contemporary China." Beijing: Social Sciences Documentation Publishing House. Originally there should be 10 classes, and the last one is semi- /unemployed, which is usually omitted from the empirical analysis. 6

7 are possibly buried under the truth. For this reason my current intent is to compare CASS09 class schema with Goldthorpe s class schema, which may also reveal a slightly different picture of social stratification in China. We hope that this comparison will reveal a rational and fresh perspective into the Chinese class structure once again. The main questions of my research, therefore, are: first, what is the class structure like if we apply a new class schema to a nationally representative dataset? Then, what are the patterns of intergenerational mobility in modern China? And finally, whether, and to what extent, the preceding mobility features derived from industrialism that can shed light on the social mobility in modern China. 2. Data Data for my empirical analysis was taken from the survey Life Histories and Social Change in Contemporary China, This survey was conducted on a national probability sample in China from June to October, Interviews were completed for 6,090 Chinese adults, aged 20 to 69 (3,087 urban residents and 3,003 rural residents). The questionnaire covered many aspects of people s life, and the information collected was sorted into 14 parts: 1) respondent s life history, 2) father, 3)mother, 4) paternal grandfather, 5)maternal grandfather, 6) spouse, 7) other family members, 8) education and activity history, 9) family economic activities, 10) parental property, 11) residential history and housing conditions, 12) material consumption, 13) cultural and social activities, 14) additional information of their life. This project is the first survey on the life history of Chinese people, and also the first one that has sampled from the population of the whole nation. In this paper, the absolute mobility rates are based on Chinese class schema CASS02. For the relative mobility rates, it was necessary to obtain the appropriate coding for the occupations, based on which we can get CASS02 class schema and CASMIN. 9 The project was initiated in 1994 by Donald J. Treiman and Ivan Szelenyi, at UCLA, and Andrew Walder, then at Harvard and now at Stanford University, in cooperation with staff from the Department of Sociology, People s University, Beijing. 7

8 1) Recoding Occupational Categories into CASS02 (Classification of Occupation released by Chinese Academy of Social Science in 2002) The 1996 survey data contains three occupational codings: CSCO95 (Chinese Standard Classification of Occupations 1995), ISCO-68, and a list of general classifications of 14-occupational categories. 10 Among them, only CSCO95 and ISCO-68 are provided with detailed descriptions of subordinate categories. 11 Thus, I could either recode from CSCO95 or from the 14-occupational categories to the CASS02 9-class schema. The 14-occupational category was mostly used in previous studies of elite dualism on occupational attainment. (Walder, Li, and Treiman 2000; Walder 2003) Walder and Li combined the category 6 (middle level management personnel) and 7 (high level management or leader of unit) into administrative elites, while category 9 (middle level professional/technical personnel) and 10 (high level professional/technical personnel) were defined as professional elites. Because Walder and Li s classification was rather general, there were some problems I needed to address: First, we cannot distinguish leading cadres and managers from the 14- occupational categories since they are combined into the same group. Also, if we distinguished this group of respondents and crosstabulate those with CSCO95 coding and ISCO-68 coding separately, most of them fall in more specific occupations in both coding systems. Second, soldiers and policemen 12 constitute one category in the 14-occupation coding. However, it is not necessary to distinguish them in the CASS02 acquired 10-class 10 This list is a 14-category general classification of occupations in China at the time of inquiry, which is on the show-card-11 for the interview. The interviewers were asked to select the most concrete one from them for the respondent from this list when they went through the job details of any time period during the course of interview. The list contains: (1) Agricultural worker (farming, fishing, forestry, animal husbandry); (2) Unskilled manual worker; (3) Skilled Manual worker; (4) Sales and Service worker; (5) Ordinary office worker; (6) Middle level management personnel; (7) High level management or leader of work-unit; (8) Ordinary or lower level professional/technical personnel; (9) Middle level professional/technical personnel; (10) High level professional/technical personnel; (11) Individual entrepreneur (getihu, also called self-employed ); (12) Head of private enterprise; (13) Soldier or policemen; (14) Others. 11 Known as International Standard Coding of Occupation in 1968, the detailed description of ISCO-68 code is not included in the codebook. 12 Though there are only 15 cases in total. 8

9 schema. Basically, soldiers should be included with the sales and service working class, and the policemen should be included in the ordinary office workers class. Third, most cases coded as indefinable Others in the 14-occupational category can be identified with a specific occupational code in either CSCO95 or ISCO-68. For the above reasons we constructed the 9-class version of CASS02 class schema based on CSCO95 coding. The details of recoding are listed in Table 10 in the Appendix, which shows the distribution of CSCO95 (with occupational titles) within the CASS02 9- class schema. Unfortunately, there was no particular coding for self-employed in CSCO95. To avoid having an empty class when we were sure there were a certain number of self-employed people at the time of our inquiry, we partially solved the problem as follows: in the 14-occupational categories from the show-card in the questionnaire, there was a category called getihu, which could be taken as equivalent to the self-employed. There were 175 cases that fell in this category. Therefore, this part of the sample could be transferred into CASS02 as self-employed directly before the recoding from CSCO95 to CASS02 actually took place. Thereafter, the recoding process listed in Table 8 was applied. Inevitably, there might be some sample still misplaced in other classes, but given the downturn of small, self-employed businesses around the 1990s, this will not affect our analysis significantly. Table 1.1 presents the distribution of class of origins and destinations of male and female respondents within the full 9-class version of CASS02. From the table we see that the proportion of private entrepreneurs experiences the greatest increase for both men and women. Especially for men, it increases more than five times from 0.18 to 1 percent of the total sample size; and for the female respondents, the size of private entrepreneurs as class destination also increases about three times from what it was as class origin. Similarly, the self-employed class, as one of the components of a private economy, also increases from 1.3 to 3.8 percent for men and from 1.2 to 2.4 percent for women. The distinctive differences between men and women on allocation of class destinations are found in leading cadres, administrative personnel, and the professional class. For men, the three classes are quite stable. For women, however, the proportion of both leading cadres class (from 2.2 to 0.8 percent) and administrative personnel s class (from 2.7 to 0.6 percent) experience a great decrease. Meanwhile, the proportions of the 9

10 other two components of the new middle class, private entrepreneurs and professionals show a substantial increase: for the former, it increases from 0.1 to 0.4 percent, and for the latter, it increases from about 8.3 percent to 11.1 percent. The greatest changes in distribution of class destinations for men and women are found in the manual worker class, the sales and service working class, and the peasants class: for men, the proportion of the manual working class increases from 15.9 to 26.2 percent, while the proportion of the sales and service working class only shows a mild increase from 6.7 to 9.4 percent. On the whole, the sharp decrease of the proportion of the peasants class from 62 to 41 percent contributes to the increased proportion of almost all other classes ranging from 0 to 10 percent. For women, the reduction of the peasants class from 62 to 52 percent leads to a jump in proportion of sales and service workers from 5.6 to 13.6 percent. However, in comparison to men, the proportion of females in the manual working class remains around 16 percent for class of origins and destinations. Table 1.1 presents a general image on the changing social structure and a preliminary idea on trends of class mobility. On the whole, men appear to be more upwardly mobile than women, especially in the upper class and the manual working class. Women seem to be more active in the professional class and the sale and service class. Table 1.2 displays the distribution of class of origins and destinations within a collapsed 4-class version of CASS02, from which we can see the emergence of the new middle class as a whole and from a more straightforward aspect. 2) Collapsed 4-class Middle Class Version of CASS02 Class Schema We further collapse the 9-class CASS02 class schema into a 4-class schema, in which the new middle class can be distinguished. According to the definition of Chinese scholars, the Chinese new middle class includes leading cadres, managers, private entrepreneurs, and professionals; then we combined the clerical workers, self-employed, and sales and service working classes as non-manual workers; lastly, the manual workers and peasants remain as they are defined in full version of CASS02. Among the new middle class, the leading cadres still occupy the privileged positions attached with various social resources as they were before the reform. Some of 10

11 them, however, transferred to become managers during the process of the restructuring of the S/COE, and some of them quit their job to either work in private firms or start their own business, which still valued the networks they established in the past. Private entrepreneurs are evidently classified as new middle class in China, which is different from its Western counterpart. However, considering the rather small sample in the 1996 survey, we sort them into the new middle class, which also consists of leading cadres, managers, and professionals. In my point of view, the ambiguous definition of the new middle class here is inevitable, due to the immature occupational structure and growing regional diversities in China. In addition, self-employed, clerical workers, and sales and service workers are combined as non-manual workers. Self-employed are those people who run their own business, which is usually quite small and within the service sector (i.e. street vendor, small barbershop, retail store, and restaurant). Their income fluctuates due to frequently changing circumstances. Therefore, it is more appropriate to combine the self-employed with the other two classes as non-manual workers. Table 1.2 reports the distribution of class of origins and destinations of male and female respondents within the 4-class CASS02. From the table we can see that, for men, the 20 percent decrease of agricultural labour has led to a slight increase of the new middle class as a whole from 12 percent to 15 percent. This is much less when compared to the 8-percent-increase of non-manual workers class and the 10-percentincrease of the manual workers. For women, there is only 10 percent reduction in agricultural labour; the proportions of the new middle class and manual workers are almost the same as the distribution of the class of origins. Again, the non-manual workers class shows the greatest increase from about nine percent to 19 percent. The changing distribution of class for the male and female samples clearly shows the gender differentiation in the division of labour. 3) Recoding Occupational Categories into Goldthorpe Class Schema (CASMIN) The ISCO-68 coding was also conducted in the original data for possible cross-national comparison, based on which we can obtain CASMIN class schema. We first recode ISCO-68 to EGP10 (Erikson, Goldthorpe and Portocarrero class code), then we collapse 11

12 the EGP10 to 7-class CASMIN with few error. In addition, Walder and Treiman created a list of additional coding in ISCO-68 especially for leading cadres (coding list from 2051 to 2082). There are essential distinctions among some specific rankings of cadres in China, shown as high, middle, and lower level of administration in Table 10. The cadres of the high administrative level are sorted to Class I of CASMIN, while cadres of middle and lower administrative levels are sorted into Class II. 13 Not surprisingly, the classes of petty bourgeoisie (IVa+b) and self-employed farmer (IVc) are both empty in this survey data. For the former, the size and type of the business, as well as the income of self-employed (getihu) are significantly inferior compared to the actual petty bourgeoisie. For the latter, operating under the state s control of land ownership, Chinese peasants could arrange production independently and keep surplus products after fulfilling the quota of public grain and agricultural tax assigned by the government. Therefore, the category of self-employed farmers is absent due to the private property issue, which varied the meaning of self-employed farming from how it was defined in the West. Consequently, we manage to obtain a 5-class version of CASMIN. The distribution of the male and female sample within CASMIN-5 is shown in Table 1.3. From this table, we can see that the service class, which to some extent is equivalent to the new middle class of CASS02, constitutes about 20 percent of the total sample, whereas the routine non-manual workers class seems to be the smallest class of all. Again, the declining agricultural labour from 62 to 41 percent contributes to the increases of other classes in various ranges. There are two points regarding the following empirical analysis that need to be noted here: First, the analysis will be conducted mainly on the basis of the 4-class version of CASS02. Second, in most of the previous research, the national sample of China is usually split into urban and rural parts. However, once being split, neither of the mobility tables can be taken as representing a complete occupational structure of the country. Therefore, in this part we will use the national sample that combines both parts for analysis. 13 In fact, class I and II are combined and classified as service class in sevenfold CASMIN. 12

13 4) Cross-classification of CASS02 and CASMIN The cross-distribution of the sample within 9-class CASS02 and 5-class CASMIN are shown in Table 2.1 and 2.2. From Table 2.1 we find that almost the whole new middle class of CASS02 falls in service class of CASMIN, and all the peasants fall in Class VIIb of CASMIN. The skilled and unskilled manual working class (V+VI, and VIIa) in CASMIN are combined as manual workers class in CASS02. Despite that, certain discrepancies are expected in describing a population according to the two class models. First, 40 percent of the clerical are found in the service class in CASMIN; 32 percent of them are actually classified as routine non-manual class; and about 27 percent of them fall into the manual working classes in CASMIN. In the sales and service class, 23 percent of them also fall into the service class of CASMIN; 54 percent of them are sorted into routine non-manual class; and about 23 percent of them are also classified as manual workers. From the class composition of 5-class CASMIN shown in Table 2.2, we find that the new middle class of CASS02 make up 72 percent of the service class, and the rest is comprised of clerical workers (nine percent), sales and service workers (14 percent), and manual workers (two percent). Among the new middle class of CASS02, the professionals make up 52 percent of the total service class of CASMIN. In addition, sales and service workers make up 71 percent of routine non-manual workers class, and manual workers make up about 80 percent for both skilled and non-skilled manual workers class of CASMIN. In conclusion, because the two class schemata share the same origins and logics, we find more overlap than divergence between them based on the relatively small sample size of the 1996 survey. 3. Absolute Mobility Rates Tables 3.1 and 3.2 present the inflow rates of men and women within the 9-class version of CASS02 class schema, which illustrates the distribution of respondents of various origins within each class of destination. The substantial decline of agricultural labour from about 60 percent to 40 percent contributes to the enormous outflow of peasants to other sectors. From the marginal percentages of the tables we find that both men and 13

14 women of peasants origin occupied the largest proportion of each class of destination, despite that for women, 30 percent of administrative personnel and 32 percent of clerical workers class were of sales and service working origin. The Peasants class itself showed the highest rate of self-recruitment for both men (90 percent) and women (89 percent). However, there remained a certain proportion of children of peasants or manual workers that moved up and found jobs in other sectors, especially for men, which we can see from the outflow rates shown in Table 3.3 and 3.4. From this we can map out the mobility patterns of men and women of each class of origins into the classes of destination. First, the results prove that the rates of self-recruitment of agricultural labour and manual workers for both men and women are particularly higher than the other occupations. For men, 58 percent of the sons and 71 percent of the daughters of peasants still worked on a farm, which indicates that females are more likely to be involved in farm-working than male peasants. As we mentioned in the introductory chapter, most surplus labour from the agricultural sector usually tended to work in township or village owned enterprises or state-owned enterprises in the cities. This is also reflected in our data: that 20 percent of sons and 10 percent of daughters of peasants became manual workers. Compared to the findings in late 1980s, the outflow from agricultural origins into the manual working sector for men has almost doubled; however, it stays more or less constant for women. In addition to the upward mobility of labour from the agricultural sector to manual occupations, industrialization also entails the relative expansion of the service sector and contraction of manufacturing industries. In China s case, we find that about 53 percent of sons and 36 percent of daughters of the working class still held manual work positions. However, the outflow patterns of men and women appear to be controversial. For men, 13 percent moved down to farm work and 11 percent moved up to the sales and service sector. For women, 14 percent moved down to farm work and 21 percent found work in the sales and service arena. Most importantly, 16 percent of women of manual origins even found professional jobs, which is 10 percent higher than is the case for men. Thus, women from a manual working background are more likely to move up to either the sales and service sector or even to professionals posts, while for men, staying within the 14

15 manual sector appears to be more preferred than either working on a farm or working as sales and service personnel. Similar outflow patterns of men and women are also found in the sales and service sector. First, 30 percent of men and 31 percent of women of sales and service origins stayed within that sector. Moreover, 33 percent of sons of sales and service personnel became manual workers, which is 10 percent higher than women. However, of the people who moved down to the agricultural sector, the rate for women (14 percent) was about three times that of men (five percent). In terms of outflows to higher occupations, especially into the middle class, 11 percent of men and 16 percent of women within the sales and service background became professionals. In addition, eight percent of men became administrative personnel, which was about seven times that of women. And, five percent of women of sales and service working origins became leading cadres, which is about three times that of the men. Of those who moved out of the manual working class, the upturn is most distinctive in the sales and service sector: 11 percent of men and 21 percent of women with a manual working-class background ended up doing sales and service work. On the contrary, the outflow of respondents of sales and service working origins, the downturn is found most distinctive in manual destination: 33 percent of sons and 23 percent of daughters of sales and service workers ended up with manual work. This further indicates the common inter-mobility between the workers of these two sectors. The outflows of middle class appear to be different for men and women: people with the background of leading cadres, administrative personnel, and professionals tend to obtain either professional jobs or started as manual workers. However, for men, the outflow rates of these three origins to manual workers appear to be higher than the outflow rates to professional class, while for women it was totally opposite. About 35 percent of sons or daughters of new middle class or non-manual working origins found their job within the sector. In addition, compared with females, men of non-manual working origins were more likely to move to manual sector. On the contrary, women of manual working origins were more likely to find job as lower whitecollar. 15

16 Overall, for both sexes, enormous outflow of the agricultural labour contributes to the increase in middle, non-manual, and manual working class. More specifically, women of new middle class background are 5.5 percent more likely than men to find their jobs within the sector. Compared with women, men with farm origin are 15 percent more likely than women to end up in middle class. Secondly, the composition of routine non-manual workers class for men and women are quite similar, except that working women are five percent more likely than men to move upwardly into the non-manual sector. For the manual worker s class, women from new middle class families are five percent more likely than men to find jobs in manual sector, whilst male peasants are seven percent more likely than female peasants to end up in manual sector. Lastly, a large proportion of peasants are of peasant origins, which are consistent with the results based on the data taken from RSSCC and CASS surveys. Thirdly, the rate of self-recruitment of new middle class and non-manual workers class for men and women are quite similar. However, within the manual sector, men are about 17 percent more likely than women to stay as manual workers; while within agricultural sector, women are about 13 percent more likely to maintain their peasant s work. However, if we look at the long-range downward mobility rates, women of new 14 The first attempt to capture the intergenerational mobility patterns in contemporary China is Cheung (1997), Cheng and Dai s (1995) investigation on the data from survey Research on Social Structure in Contemporary China (RSSCC) in RSSCC took the form of a multi-staged sampling survey of social life in four provinces (Liaoning, Hebei, Shandong, and Guizhou) and two municipalities (Beijing and Shanghai) of China in 1988, including both urban (which contains 4,658 cases) and rural areas (which contains 4,723 cases), where the questionnaires with substantial differences in occupational classification were applied. In practice, Cheng and Dai used the eight main occupational categories created in the original data to look at the absolute mobility rates and trends of social fluidity. Meanwhile, referring to the idea of CASMIN, Cheung sorted the sample into a 5-class version class schema: white-collar workers, manual workers, self-employed farmers, peasants and other workers in primary production, and the petty bourgeoisie with or without employees. Considering the private sector had not been well developed at the time the survey was conducted, especially the private entrepreneurs, Cheung omitted the large proprietors class from the CASMIN and the sample affiliated to any kind of private business was converted to class of petty bourgeoisie. Lastly, the self-employed farmers and agricultural workers are distinguished as two classes. 15 The survey data on The changing social structure in contemporary China that was conducted by CASS in 2001 based on national probability sampling (12 provinces, 72 cities, with a sample size around 11,000) is not available to the public. Most of the information is taken from the reports released by CASS based on this survey. Therefore my current research is only a preliminary observation on the changing patterns of occupational mobility in China in late 1980s, middle 1990s, and early 21st century. 16

17 middle class origins are 10 percent more likely than men to end in agricultural sector. Moreover, compared with men, women of routine non-manual origins are eight percent less likely to be found in manual sector in terms of short-range downward mobility, but six percent more likely to find job in agricultural sector. In terms of long-range upward mobility, women of manual origins are 6.5 percent more likely to be found in the new middle class destination than men and nine percent more likely to move to routine nonmanual sector than men. The graphs in the Figure 1.1 show the components of absolute mobility rates for men obtained from the data of 1988, 1996, and 2001 surveys. Total mobility rates (TMR) are the percentages of all respondents who are found off the main diagonal in the mobility table. The line on the top indicates a sharp decrease of TMR from the year 1988 to 1996, and stayed constant in From the trends of TU and TD we can see that it is largely caused by the declining TD from 26 to 9. However, the effect of the increasing TU from 34 to 39 is comparatively modest. The trends are the same for women. 16 From the Figure 2.1 we can see that for men, total upward mobility rates show a slight increase firstly from 34 in the late 1980s to 36 in the mid-90s, and then to 39 in the most recent survey in Meanwhile, total downward mobility shows a sharp decrease firstly from 26 to 15, and a further decreases to nine, which contributes to the overall about 10 percent decrease of total mobility rates from late 1980s (60) to 2001 (52). The Figure 1 demonstrates a more direct vision of the trends of main features of absolute mobility rates a cross time. For women, Figure 2.2 shows that across time, despite the similar trends shown in all features of absolute mobility rates, the rates of total mobility and total upward mobility of women stay lower than men; while the rates of total downward mobility appear to be slightly higher than men. From 1988 to 1996, we find that total upward mobility rates increase from 25 to 28 percent, and the total downward mobility rates show a 10 percent decrease from 27 to 17. Figure 3 demonstrates total mobility rates of men and women over three age cohorts aged from 19 to 69: the oldest cohort consists of respondents who were born 16 The female figure from the CASS 2001 survey is not available on the report. 17

18 before 1949, aged from 48 to 69; the second cohort includes respondents who were born after 1949 and before the Cultural Revolution, aged from 31 to 47; and the youngest cohort is composed of those who were born during the decade of Cultural Revolution, when the occupational allocation were strictly controlled by the central redistributive power and particularly favoured people of peasants or working class origins rather than the merit of personal achievements. The respondents of this cohort aged from 19 to 30 at the time of the inquiry. First, there is a big increase of TMR for both men (by 16 percent) and women (by 10 percent) of the second cohort, followed by a relatively moderate increase for the youngest cohort. (Seven percent for men and four percent for women) The gap between TMR for men and women is stable for eldest and youngest cohorts, except that it became slightly narrower for the Cultural Revolution cohort. 4) Absolute Mobility between Four Major Class Categories i) Outflow from Farm Origins In Figures we show the graphs, for men and women respectively, which summarise the trends of outflow from farm origins to manual, non-manual, and middle class destinations by cohorts. In terms of short-range upward mobility from the agricultural sector to the manual sector, outflow rates of men show a sharp increase from 10 percent for the eldest cohort to 29 percent for the youngest cohorts. For women, outflow from farm origins to manual working class also increases from 4 to 15 percent cross three cohorts, although the overall increase is not as profound as men. The trends of outflow from farm origin into non-manual destinations are not so different between men and women across three age cohorts. For men it has been fluctuating around 12 percent, while for women it gradually increases from 12 to 16. In Figure 4.3 we observe a slight decline of the outflow from farm origin into middle class destination from 12.5 to 12 percent in second cohort, followed by a sharp decrease from 12 to 6 percent for men. Similarly, for women, it declines from 7 to 6.5 percent, then to 3.5 percent. Then it is followed by a further decrease of 5.5 percent for men and a roughly 3 percent decrease for women for the youngest cohort. From the 18

19 Figure we can see that the long-range upward mobility of men from the agricultural sector to the middle class sector has almost doubled compared to women across time. Overall, in terms of the outflow rate from farm origins to the other three destinations, (except a similarly smooth pattern into the non-manual working class), there are twice as many men than women outflow from the farm origins to the manual and middle class destinations. ii) Outflow from Manual Origins Interestingly, we found slightly opposite trends for men and women in Figures , which present the outflow of men and women from manual origins to farm, nonmanual, and middle class destinations by cohorts. On the whole, women with manual working origins appear to be more likely to move out of that sector than men, regardless of whether they move upwardly or downwardly. In Figure 5.1, men show a slow but steady increase from 19 to 23.5 percent in outflow from the manual to the non-manual sector across three cohorts. For women, there is a sharp drop, from 38 to 30 percent for the second cohort, but they stayed about the same for the youngest cohort. The outflow rates for men and women became closer for the Cultural Revolution cohort, though the outflow of women from manual origins to a non-manual destination has been higher than men cross the cohorts. On the other hand, Figure 5.2 shows the outflow from the manual sector to the middle class, men first experience a sharp decrease from 24 to 10 percent, and then stabilize around 11 percent. Conversely, the outflow rate for women stays steady around 21 percent for the first two cohorts, and then falls to about 14 percent. Figure 5.3 presents the graphs for outflow from manual working origins to farm destinations across three cohorts for both men and women. Again, we observe similar trends for men and women: a sharp decrease from the first to the second cohort, followed by a small rise in the outflow rates. For men, the figure shows a sharp decrease from 22 to nine percent for the second cohort, followed by a roughly five percent increase for the youngest cohort. Similarly, the outflow for women first shows a sharp decrease of from 19

20 37.5 to 11 percent for the second cohort, and slightly drops to 14 percent for the youngest cohort. Moreover, if we compare Figure 5.3 with Figure 4.1, from which we have observed the constantly increasing upward mobility of people with farm origins into the manual working sector, we can see two significant trends. First, there has been a steady mobility of the labour force between the manual and farm sectors across three cohorts, particularly for men. Second, for the youngest cohort, people with farm origins are more likely to move out of that sector rather than the other way around, here again, particularly for men. The results partially confirm that in contemporary China, like other industrial societies, the decline of the labour occupation within the agricultural sector has begun, with the surplus labour force from agricultural sector gradually being absorbed into the growing industrial sector. Although there has been 20 percent reduction in the agricultural population, agriculture is still playing an important role in the state economy. iii) Outflow from Non-manual Origins The outflow from non-manual origins for men and women are shown in the Figures , from which we can see there are constant changes cross three cohorts for both men and women. Figure 6.1 graphs the upward mobility from the non-manual to the middle class. The trend for men steadily decreases from 32 to 10 percent cross three cohorts. However, for women, it first slightly increases from 25 to 30 percent from the eldest to the second cohort, which is slightly higher than men, then falls to 17 percent for the youngest cohort, which is almost double that of men. Figure 6.2 shows the downward mobility from non-manual origins to the manual sector. The trend for men shows a steady increase from 24 to 38 percent cross three cohorts; for women, it firstly experiences a sharp increase from 4 to 28 percent, which again, surpasses men s figure; then is followed by a minor decrease to 24 percent for the youngest cohort. On the whole, men of non-manual origins are more likely to find a job in the manual sector than women, although the outflow has increased for both sexes. 20