Conservatism or liberalism? A study on sociopolitical attitudes of Chinese middle class Li Chunling Institute of Sociology Chinese Academy of Social Sciences Since the beginning of this century, a social group with higher income, higher education and higher occupational prestige has been emerging in Chinese cities. It has been called by the public media as “middle class”. Even though people dispute about the definition of middle class, there is no doubt that this group exists in Mainland China and it is expanding quickly. The group has attracted more and more attentions from the public, businessmen and policy-makers as well as researchers of sociology, economics and politics. Sociologists especially have had a great interest in the group and made many discussions about it. The recent focus of this topic is on the sociopolitical function of middle class. Chinese sociologists are debating on such questions as these: What are the sociopolitical consequences of the emergence of middle class? Is middle class a stable force or unstable force for the existing authority? Will middle class promote a democratic transition or preserve the existing political order? Analysts gave two opposite answers. Some people say middle class is a social force to promote a democratic transition so that it is a destabilized factor for the authority. That means middle class is an unstable force for the government. These analysts suggest that the government should take carefully measures to control this group. But others consider middle class as a social force to support the existing political and social order. They say that middle class has a function of sociopolitical stabilizer and is a stable force for the authority. They advise the government to take a policy to enlarge middle class. Based on a measurement of sociopolitical attitudes of Chinese middle class, this article tries to answer above questions. Controversy on sociopolitical function of middle class There has been a heated and long controversy on sociopolitical function of middle class since 1980s when Chinese researchers began to discuss a possibility of emergence of Chinese middle class. Radical intellectuals in 1980s were expecting a democratic movement promoted by a rising middle class because sociologists and political scientists described middle class as “the most active supporters of democratization”. In 1990s most sociologists had changed their view and described middle class as a stable force for the political order, a positive supporter for government’s policies of the economic reform and a driving force for the economic development. However, in the recent years, a few sociologists questioned the view of “stable force” and claimed that middle class was probably a potential time-bomb for the authority. Middle class as an unstable force Sociopolitical theorists from the West ( such as Lipset, Huntington, Glassman etc.), based on the experience of the Western societies, proposed a logic of correlation between the emergence of middle class and development of political democracy. S.M. Lipset (1963) developed a wealth theory of democracy, namely, “the more well-to-do a nation, the greater the chances that it will sustain democracy” (Lipset 1963, p.31). Democracy is related to economic development because there is a diamond-shape social stratification with an expanded middle class in wealthy countries. Samuel P. Huntington (1991, p.67) support Lipset’s argument by reporting that “the most active supporters of [the third wave of] democratization came from the urban middle class.” Other theorists mentioned cases of the East Asian and Southeast Asian countries and regions (South Korea, Taiwan, Philippines and Thailand, etc.). In these societies, rising middle class has brought about democratic movement and a series of sociopolitical turbulence. A few of Chinese sociologists (Li Lulu 2008; Zhang Yi 2009) follow this idea and consider middle class will become a potent agent of the sociopolitical transition toward democracy and civil society. Middle class as a stable force However, the most influential Chinese sociologists (Li Qian 2001, Zhou Xiaohong 2005) think that Chinese middle class is a sociopolitical stabilizer because they hold a political conservatism. They provide some reasons: 1) Chinese middle class is a group of the most beneficiary during the economic reform and rapid economic growth. 2) Chinese middle class strongly depend on the State and have a propensity for state authoritarianism. 3) Chinese middle class at present shows apolitical attitudes and prefer materialism. In addition, some researchers who study middle class of East Asian society propose an explanation of East Asian specificity. Because of traditional forms of political culture, middle class of Asia prefers political conservatism and relies on the state for economic security (Jones and Brown 1994). Middle class of Asia shows a dual political personality or contradictory attitude. On the one hand, they is inclined to support liberalism and democracy, on the other hand, they desire sociopolitical stability and are subservient to authoritarian state for economic security. (Hsiao 2001； Hsiao and So 1999） . An ideal dichotomy of sociopolitical attitude Based on these literature mentioned above, we may propose a relationship between sociopolitical function and sociopolitical attitudes of middle class. Middle class is a stable force if its members take conservative attitudes. On the contrary, middle class become an unstable force if middle class take liberalist attitudes. In table 1, I frame an ideal dichotomy of sociopolitical attitude of conservatism and liberalism by putting forward three dimension criteria to classify different attitudes of conservatism and liberalism toward certain issues. For each criterion, there are two indexes to test the certain attitudes. Table 1 Ideal dichotomy of sociopolitical attitude of conservatism and liberalism Criteria Basic judgment toward the current situation Basic attitudes toward the state and government Basic value on social justice Conservatism Feeling Satisfying and disliking change Liberalism Feeling dissatisfying and hoping some change Low confidence in government; preference for political democracy Low acceptance of inequality; sympathy for disadvantaged groups and low class Index of measurement ① satisfaction degree of individual living (SDIL index) ② satisfaction degree of social situation (SDSS index) ③ confidence degree in government (CDG index) ④degree of state authoritarianism (DSA index) ⑤acceptance degree of social inequality (ADSI index) ⑥ degree of conflict consciousness (DCC index) High confidence in government; preference for authoritarian state High acceptance of inequality; concealing interest conflict between groups Measurements of indexes SDIL index SDIL index means the satisfaction degree of individual living. There are two questions for the measurement of this index. Question 1: Comparison with five years ago, your current living condition is: 1) much better 2) better 3) no change 4) worse 5) much worse Question 2: You think your living condition after five years will be: 1) much better 2) better 3) no change 4) worse 5) much worse Each question has five selective answers. Each answer is represented by a certain score. For example, question 1, “much better” is 2 score, “better” is 1 score; “no change” is 0, “worse” is minus 1; “much worse” is minus 2. Total score of 2 questions for a person is the score of this index. SDSS index SDSS index means the satisfaction degree of social situation. There are two questions for the measurement of this index. Question 3: Do you think social situation in our country is stable? 1) very stable 2) stable 3) I don’t know 4) unstable 5) very unstable Question 4: How do you feel about social circumstance in our country? 1) very harmonious 2) harmonious 3) I don’t know 4) inharmonious 5) very inharmonious CDG index CDG index means the confidence degree in government. There is one question including 6 items for the measurement of this index. Question 5: Do you believe below governmental organizations or affaires? A. The central government B. Local government C. News press of the government D. Statistics published by the government E. Appeal organization F. Judge and policeman 1)strongly believe 2)believe 3)I am not sure 4)don’t believe 5) strongly don’t believe DSA index DSA index means the degree of state authoritarianism. There is one question including 6 items for the measurement of this index. Question 6: Do you agree with statements below: A. Democracy means the government makes decisions for people. B. Government is responsible for managing important affairs of our country, so people should not care about these affairs. C. People should follow the government’s instructions, just like employees should obey the commands of boss. D. The government and party have capabilities to manage our country. E. People should move out of their houses if the government want to build public constructions in the location of their houses. F. People pay tax and the government may decide how to spend it without taking into account of people’s viewpoint. 1)Strongly agree 2) agree 3) I am not sure 4)don’t agree 5) strongly don’t agree ADSI index ADSI index means the acceptance degree of social inequality. There is one question including 14 items for the measurement of this index. Question 7: Do you think it is fair or unfair in the below aspects of our society? A. Wealth and income distribution B. Public finance and tax policies C. Opportunities of job and employment D. Opportunities of individual development E. College entrance examination system F. Promotion of the government’s officials G. Public health care H. Compulsory education I. Political right J. Judicatory system K. Welfare in different regions and industries L. Welfare in urban and rural areas M. Social security N. Total situation of social equality 1) very fair 2) fair 3) I am not sure 4) unfair 5) very unfair DCC index DCC index means the degree of conflict consciousness. There are two questions for the measurement of this index. Question 8: Do you think there is interest conflicts among social groups in our society? 1) no conflict 2) I don’t know 3) a little conflicts 4) many conflicts 5) very much conflicts Question 9: Do you think it is possible the interest conflicts among social groups will become serious? 1)definitely impossible 2)impossible 3)I don’t know 4)maybe possible 5) definitely possible Class classification and definition of middle class Before starting the measurement of attitudes, a classification of middle class has to be defined for the analysis. There are so many classifications of classes and so many concepts of middle class. It is not the issue this paper wants to make clear. The class classification used here is developed by the EAMC project. EAMC project means the East Asian Middle Class Project which is a comparative research of Asian middle class running by researchers from some Asian countries including South Korea, Japan, Hong Kong, Taiwan and so on. EAMC class scheme classifies 6 classes based on John Goldthope’s class scheme. The researchers of EAMC project think this class classification is appropriate for study of middle class of Asian societies. Table 3 Class classifications of Goldthope and EAMC project Goldthope’s class scheme Ⅰ Higher-grade professionals Ⅱ Lower-grade professionals Ⅳa Small employers with employees Ⅳb Small employers without employees Ⅲa Routine non-manual employees Ⅲb Personal service workers Ⅴ Technician and supervisors Ⅵa Skilled workers Ⅶa Semi-/non-skilled workers Ⅳc Farmer Ⅶb Agricultural workers 6) Farmers/farm labor 5) Working class 4) Marginal middle class EAMC project’s scheme 1) Capitalist (employers hire 20 or more employees) 2) New middle class 3) Old middle class Among 6 classes in the class scheme of the EAMC project, three classes (new middle class, old middle class and marginal middle class) are considered as middle classes. That implies the definition of middle class is plural, namely middle classes, not middle class. The further presupposition proposed by the EAMC project is that there is internally intra-class diversity among middle classes in attitudes. Different groups of middle class have different sociopolitical attitudes. A little revise is made when this class classification is used to define Chinese middle class. Capitalist class, which is not one part of middle class in the EAMC project, is supposed to be a key part of Chinese middle class. That is because Chinese capitalist class is a new class and its appearance has been changing the original class structure and symbolizing a rising of middle class. It is unlike the emergence of middle class in other societies, in which expanding social group of professional and managerial employees represented a rising of middle class and changed the original class structure that was composed of capitalist class and working class, In addition, capitalist class is not the most dominant or most advantageous social group in China. Most of capitilists are owners of small or medium size enterprises. Even though they possess a large number of economic capital, there is a limit to their social and political influences. In China, the most dominant and advantageous group is high-ranking government officials and CEO of state-owned enterprises. They have great powers and control most of social-economic resources but capitalists have not. In this context, capitalists are deemed as a part of middle class in China. Data, variables and methods Data The data for this research is a national sample survey data on Social stability collected in 2006 by the Institute of Sociology, Chinese Academy of Social Sciences. The sample size is 7061. Taking into account most members of Chinese middle class residing in cities, only cases of urban area is included for analysis. Case number of urban area is 2894. Methods and variables 6 OLS linear regression models are used for examining sociopolitical attitudes of classes. Each model is for each attitude index. Dependent variables of regression models are 6 indexes. Independent variables are five classes, including capitalist class, new middle class, old middle class, marginal middle class and working class. Controlling variables are sex, age and education. Table 4 shows the descriptive statistics of variables for regression analysis. Table 4 Descriptive statistics of variables （N=2894） Variable Age Schooling years Scores of SDIL index Scores of SDIL index Scores of SDIL index Scores of SDIL index Scores of SDIL index Scores of SDIL index Minimum 18 0 -4 -4 -12 -12 -28 2 Maximum 69 20 4 4 12 12 28 10 Average 39.8 9.7 .7786 1.1399 3.7487 .1703 -.2880 6.2458 Ratio（％） Sex（male） Capitalist class New middle class Old middle class Marginal middle class Working class 45.1 0.2 10.8 11.3 14.4 20.3 Standard error 13.0 4.1 1.6146 1.6596 4.0239 4.0923 8.8481 1.8043 Results and interpretation Table 5 shows the results of regression analysis. Table 5 Unstandardized OLS Coefficients for the Linear Regression of Attitude indexes on Classes Dependent variable Independent variable Class (reference group: working class) Capitalist class New middle class Old middle class Marginal middle class Sex（male） Age Schooling years Constant .542 (.768) .787** (.105) .424** (.095) .326** (.089) -.002 (.059) -.017** (.002) .008 (.009) 1.179** (.153) .706 (.812) .231* (.111) .188 (.100) .002 (.094) .007 (.063) .000 (.003) -.015 (.009) 1.231** (.161) 1.031 (1.966) .072 (.268) -.020 (.243) .039 (.228) -.106 (.152) .017** (.006) -.039 (.022) 3.475** (.390) 1.543 (1.903) -.806** (.260) .222 (.235) -.577** (.221) .004 (.147) .031** (.006) -.222** (.021) 1.226** (.378) -6.339(4.329) -.415 (.591) -1.378** (.534) -.033 (.502) -.401 (.335) .006 (.014) .062 (.030) -.730 (.860) -.844 (.857) .104 (.117) -.054 (.106) .211* (.099) .109 (.066) .001 (.003) .098** (.010) 5.170** (.170) SDIL index SDSS index CDG index DSA index ADSI index DCC index Adjusted R2 N .056 2894 .001 2894 .004 2894 .098 2894 .002 2894 .058 2894 Note: Standard error shown in parentheses. **:P≤0.01; *:P≤0.05 The first model is for SDIL index (satisfaction degree of individual living). That is to examine if there is any difference among classes in this attitude. Since the number of capitalists’ cases in survey data is too little, all coefficients for capitalist are not significant. Consequently, this data results can not show clear attitudes of capitalists. Three coefficients of middle classes all are significant. That means there is difference in the satisfaction degree of individual living between working class and three middle classes (new middle class, old middle class and marginal middle class). In addition, these coefficients are positive figures and the coefficient of the new middle class is the largest one. That means three middle classes have higher satisfaction degree of individual living than working class. New middle class has highest satisfaction degree of individual living among middle classes. Sex coefficient and education coefficient indicate that there are no gender difference and no difference among different educated groups in satisfaction degree of individual living. However, age coefficient is significant and negative. That means there is difference among different age group. Older people have low satisfaction degree of individual living than younger persons. The second model is to examine the difference among classes in SDSS index (satisfaction degree of social situation). All coefficients are not significant except the coefficient of new middle class. It is positive figure. That means new middle class has higher satisfaction degree of social situation than other middle classes and working class. Old middle class and marginal middle class have same satisfaction degree as working class. There are no gender difference, no age difference and no educational difference. The third model is to examine the difference among classes in CDG index (confidence degree in government). All coefficients are not significant except age. That shows no class difference, no gender difference and no educational difference in the confidence degree in government. But young people have higher confidence than older people. The fourth model is to examine the difference among classes in DSA index (degree of state authoritarianism). For this index, there are significant and large differences among classes, age groups and educational groups. New middle class and marginal middle class have negative coefficients which indicate that new middle class and marginal middle class have much less degree of stat authoritarianism than working class and old middle class and capitalist class. In other words, new middle class and marginal middle class have more democratic consciousness. In addition, higher educated group has more democratic consciousness and less state authoritarianism than lower educated group. Older people have more state authoritarianism than young persons. The fifth model is to examine the difference among classes in ADSI index (acceptance degree of social inequality). All coefficients are not significant except the coefficient of old middle class. The coefficient of old middle class is negative figure and quite large figure. That means old middle class has much higher acceptance of social inequality than other classes. In addition, capitalist’s coefficient, even though it is not significant because of few cases, is very large negative figure. That seems to imply that middle classes with economic capital (such as capitalist class and old middle class) have higher acceptance of social inequality or lower expectation of social justness. The final model is to examine the difference among classes in DCC index (degree of conflict consciousness). Only two coefficients for marginal middle and education are significant. That shows marginal middle class has more conflict consciousness than other classes. And higher educated group has more conflict consciousness than lower educated group. Summary and conclusion Summary of results Middle classes show significantly higher satisfaction degree of individual living. Among middle classes, new middle class has highest satisfaction degree. 71.8% of new middle class say their living condition has improved in the recent five years (the highest percentage in five classes). 68.6% of new middle class expects their lives will be better in the future five years (the highest percentage in five classes). That implies middle classes, especially new middle class, are likely to remain the situation unchanged and do not like any sociopolitical change that might influence their living condition. So that middle classes probably take conservative attitude toward a potential sociopolitical change. New middle class has higher satisfaction degree of social situation than working class and other middle classes (old middle class and marginal middle class). 80.1% of new middle class say social situation in our country is stable. 76.6% of new middle class say social circumstance in our country is harmonious. That implies new middle class is most likely to maintain unchanged social order and object to changes that might bring about any social turbulence. The further implication is that new middle class may be inclined to take conservative attitude in this aspect. There is no significant difference between classes in confidence degree in government. All classes express a relatively positive evaluation of the government even though there are differences of confidence degree in different items. All classes show quite high degree of confidence in the Central government while expressing relative lower degree of confidence in official statistics. 94.1% of new middle class, 94.8% of old middle class, 95.4% of marginal middle class, 94.2% of working class and 89.9% of capitalist class say they relatively believe or very believe in the Central government. That implies middle classes has high confidence degree in government and further implies they are likely to preserve the existing political situation. New middle class and marginal middle class show much less degree of state authoritarianism than working class and old middle class. New middle class is least likely to support for an authoritarian state. That indicates new middle class and marginal middle class have more democratic consciousness than other classes. They are more expecting a democratic government even though they do not like a sociopolitical change that might bring about any turbulence. Capitalist class and old middle class have much higher acceptance of social inequality and much less expectation of social justice than new middle class, marginal middle class and working class. That seems to show middle classes diverge in attitude toward social inequality. Middle classes with economic capital have less consciousness of social justice while middle class with cultural capital have more consciousness of social justice. That implies middle classes with economic capital are more likely to take conservative attitude while middle classes with cultural capital are more likely to take liberalist attitude in this aspect. Marginal middle class shows significantly higher degree of conflict consciousness than other classes. Capitalist class and old middle class seem to have lower degree of conflict consciousness. That implies middle classes with economic capital are more likely to take conservatism in this aspect while marginal middle class (lower strata of middle class) is more likely to take liberalism and have more sympathy for the disadvantaged groups and lower class. Final conclusion There are indeed differences between middle class and working class in sociopolitical attitudes, especially between new middle class and working class. However, we can not conclude that middle class is more conservatism or liberalism than working class. In some aspects, such as in basic judgment toward the current situation, middle class is more conservatism than working class. In other aspects, middle class is more liberalism than working class such as attitudes toward the authoritarian state. More importantly, there is existing differences among middle classes in sociopolitical attitudes. New middle class, a middle class with more cultural capital, holds a contradictory sociopolitical attitude. On one hand, they have most democratic consciousness, on the other hand, they display highest satisfaction of current situation and desire to remain such situation. Capitalist class and old middle class, middle class with economic capital, hold relatively political conservatism. They are more likely to support state authoritarianism and have least consciousness of social justice. Marginal middle class, lower strata of middle class, is probably most liberalism. They have most consciousness of social justice, more democratic consciousness and most sympathy for lower class. Age and education show partly effect on sociopolitical attitudes. Education has a significantly positive correlation with liberalism, especially with democratic consciousness. With the increase in the schooling years of middle class, liberalism and democratic consciousness might advance among middle class in the future. It is worthy to note that younger people have more democratic consciousness and lower confidence in the government. That probably implies liberalism and democratic consciousness might develop among middle class when more young people enter middle class in the future. Chinese middle class, as a whole group, has a divergent attitude. It is a mixture of conservatism and liberalism. They have high satisfaction of the current situation and high confidence in the government. A part of them has high expectation for political democracy and social justice. Such contradictory attitudes imply that Chinese middle class may be incline to choice the Third Roadlass, slowly stepwise sociopolitical transition. In that case, Chinese middle class is a stable force at present. However, in the future, there isass, possibi that it become an lity unstable force. References Pacific Review 7(1):79-87. Li, C. (2003) “The Composition of Middle Class of the Contemporary China”, China Population Science, No.6. Li, C. (2005) Cleavage or Fragment : A Empirical Analysis on the Social Stratification of the Contemporary China. Beijing: Social Science Academic Press. Li, C. (2009) Formation of Middle Class in Comparative Perspective: Process, Influence and Socioeconomic Consequences. Beijing: Social Science Academic Press. Li, L. (2008) “Social function of middle class: a new perspective and multi-dimension framework”, Transaction of Renming University, No.4. Li, Q. (2001) “Middle class and Middle stratum”, Transaction of Renming University, No.4. Li, P. and Zhang Y. (2007) Size, identity and attitudes of Chinese middle class, Society, No.6. Lipset, S.M. 1963. “Economic Development and Democracy.” Pp. 27-63 in Political Man. Garden City, NY: Anchor. Zhang, Y. (2009) “Is middle class a social stabilizer?” In C.Li (ed.) Formation of Middle Class in Comparative Perspective: Process, Influence and Socioeconomic Consequences. Beijing: Social Science Academic Press. Zhou X. (2005) Survey on Chinese middle class, Social Science Academic Press.