The rapid growth of the middle class in contemporary China has become one of the most striking phenomena in its social structural transformation. However, compared with the middle classes in Western societies, the Chinese middle class is different in terms of its mode of formation and its structural composition. Its class character and function also have their own distinctive attributes. In the process of a social transformation with a high degree of space–time compression, the Chinese middle class has brought together diverse groups, such as the new old middle class, the main marginal middle class, the middle class that emerged from the free market and the middle class from the redistribution sector.Footnote 1 To a large extent, the diverse composition of the middle class determines the complexity of its social and political functions.
According to the highly regarded “Global wealth report” released by the Credit Suisse Research Institute in 2017, China's share of the world's 1.1 billion middle class individuals (wealth ranging from $10,000 to $100,000) has risen rapidly, from 12.6 per cent in 2000 to 35 per cent in 2017, making China's share the world's largest. While enjoying the results of a rapid improvement in material living standards, China's middle class also experiences pressure, confusion and contradictions. In recent years, China's economic growth has slowed down, market competition has sharply intensified, the cost of living has continued to rise, and the stock market and housing market have fluctuated drastically. All these changes have increased the pressure and anxiety felt by the Chinese middle class.Footnote 2 In addition, a series of social incidents involving private property rights protection, judicial justice, healthcare and child safety has further exacerbated the insecurity of the middle class. At the same time, the middle class's class consciousness, rights awareness, public participation and even the ways it resists are also quietly changing.Footnote 3 Members of China's middle class show more and more significant class power in the struggle for property owner's rights, resistance to environmental pollution, NIMBY movements, feminist actions and other struggles to expand the space of action within civil society.Footnote 4
What kind of socio-political influences will emerge from the large-scale rise of the Chinese middle class? Will these influences be conducive to maintaining social stability or will they promote social change? Theorists have provided two opposing answers to these questions. The radical position advocates that in an affluent society in which the middle class dominates, it will, inevitably, actively promote liberalization and political democracy.Footnote 5 In contrast, the conservative position asserts that the middle class is the avant-garde in economic consumption but the defender in political actions.Footnote 6 This is because the middle class benefits from the existing social order, its thoughts and ideologies are moderate, and it is unlikely to promote radical political changes. In short, each position has different perspectives and the debate continues.
I argue that the discussion about the political function of the middle class in China should not be generalized but instead should focus on multidimensional analysis. This is because the overall position of the middle class is such that its members share many similarities but the complexity of the middle class's formation mode and its internal composition also make the group heterogeneous. In view of this, I used system segmentation and market situation to categorize the middle class into several basic types. I then examined the overall consistency and internal heterogeneity of the political values of the middle class from multiple indicators of freedoms, government satisfaction and political voting.
Theoretical Background and Research Hypothesis
The political function of the middle class
Analysis of social and political functions has always been a core issue in research on the middle class. Additionally, there is also debate on the radical and conservative aspects of the middle class's political functions. One school of thought argues that the middle class has a radical political orientation and is a “booster” for social democratic change. According to Seymour Lipset, “the richer a country is, the more likely it is to promote democracy.”Footnote 7 This view was later termed the “iron law” by theorists.Footnote 8 Ronald Glassman demonstrates the connection between the middle class and democratic politics from a historical perspective and believes that the middle class provides the basis for the structure and institutions of democratic politics.Footnote 9 Samuel Huntington suggests that “the most active supporters of the third wave of democratization came from the urban middle class.”Footnote 10 There can be but two explanations for this assertion. First, the middle class has a higher critical social consciousness and tendency for action because its members tend to be highly educated. Second, the class experience in market economy activities encourages members of the middle class to make more demands for freedom and justice, actions that actively promote liberalization.
The opposing point of view argues that the middle class has a conservative political orientation and is a stabilizer of social politics. The logic of this assertion is that the particular interests of the middle class (for example, power, income, prestige) are derived from the existing order (middle position). Therefore, members of the middle class will tend to maintain the existing order. C. Wright Mills terms this group the “political rearguarders.”Footnote 11 The many scholars who share this view contend that the middle class acts as a buffer between the upper and lower classes, eases social conflicts and promotes social stability.
In studies on the political function of the middle class in East Asian society, most scholars also favour the conservative argument. David Jones and David Brown believe that the East Asian middle class has a tendency towards political conservatism and is attached to strong government owing to the influence of traditional Eastern political culture.Footnote 12 Research by Michael Hsiao and Alvin So suggests that the rise of the middle class in East Asia has occurred during a period of rapid economic growth led by a strong state which implements policies to promote that growth. As the middle class benefits from this economic growth and strong government, it is invested in maintaining social and political stability and protecting its own economic interests.Footnote 13
There are also two opposing views on the political function of the middle class in China. The mainstream view holds that the middle class has conservative political tendencies. That is, its members widely agree with the existing system and are unlikely to become promoters of democratization in the near future.Footnote 14 The opposing school of thought is that the Chinese middle class has a tendency towards social change. Zhang Yi asserts that the middle class is more critical and has the potential to subvert society, stating that it “will never quietly embark on the historical stage.”Footnote 15 A recent study by Andrew Nathan claims that at present, the Chinese middle class supports the existing order and system because its members desire social stability. However, he maintains that as the values of the middle class become more liberalized, the political alienation of its members will increase and radical and extreme tendencies will come into prominence when economic growth slows or the security of their socio-economic status diminishes.Footnote 16
Although there has already been much discussion on the political function of the middle class, some issues require further examination. First, most of the research on the Chinese middle class is based on a Western context and tends to discuss the middle class as a homogenous entity, ignoring its internal heterogeneity. Second, studies that do highlight the particularities of Chinese society focus more on market differentiation;Footnote 17 however, the emphasis on the institutional division of redistribution and the market environment is not enough. Third, the empirical indicators used for predicting political function, such as satisfaction, feelings of inequality and sense of fairness, are indirect and do not address the issues adequately.Footnote 18 Some indicators focus only on perceived attitudes and lack the elements of political action. This study attempts to remedy these issues by providing a comprehensive index system for understanding the multidimensional political values of the Chinese middle class.
The diversified composition of the middle class in China
Although generally situated between the upper and lower levels of society, the middle class in China is not a class entity with uniform characteristics. Instead, it is a multilevel collection of groups. The diversity of the groups depends on the complexity of the class formation mode. The growth of the middle class is influenced not only by the market but also by the redistribution system.
The first factor is market situation (new middle class vs marginal middle class). The Weberian tradition of class analysis has always argued that class conditions depend to a large extent on market situation and that the economic benefits from the market determine an individual's life chances and class position. The differentiation caused by market situation is typically reflected in the distinction between the new middle class and the marginal middle class. The new middle class forms the greater part of the middle class and comprises middle and senior administrative and management personnel as well as professional and technical personnel. According to Karl Renner, these groups share the same employment status because they form a service relationship with their employer.Footnote 19 Their incomes are significantly higher than those of lower-level employees because (a) members of the new middle class play an important role in management and the professional services, and (b) their employment relationship contains more trust and moral implications.Footnote 20 In contrast, the marginal middle class plays more of a supporting role for the new middle class and members are often regarded as routine, non-manual employees. They are also engaged in indirect productive work but use fairly standardized techniques to carry out repetitive work. Therefore, there are significant gaps between these two groups in terms of income, promotion opportunities, authority and job security.Footnote 21
The category termed “small owners” may also be referred to as the petty bourgeoisie according to Marx's definition. Although members of this group have special characteristics in terms of employment status, possess property and have a high degree of work autonomy, their economic status is still determined by their position in the market. Mills pointed out that the transition from old middle class (small owners) to new middle class is a transition from property to a new stratified axis – occupation.Footnote 22 In terms of market conditions, the economic benefits for the small owners are lower than those enjoyed by the new middle class and also some of the marginal middle class.
The second factor is the institutional division (the middle class that came via the redistribution sector vs the middle class that emerged from the free market). On the macro level, China's market-oriented transformation was initiated and promoted under the leadership of the government. The path-dependent inertia of the original system, the logic of resource allocation led by political power and the controlled transformation process all deeply affected social stratification and the relationships between class groups.Footnote 23 At the micro level, the unit organization within the system is still an effective tool for the state to rule. Resource allocation, control of ideology and social integration remain prominent functions of the unit organization. This is obviously different from the market economy outside the system, which is full of individualized free competition and interest-oriented action.
Owing to the division of China's redistributive system and market economy, two middle classes with different class experiences have formed. One comprises managers, professional technicians and routine non-manual employees in government departments and state-owned/collective economic sectors; the other segment of the middle class is composed of the groups engaged in management, technology and service work in the market economy sector and in non-profit organizations. There are significant differences between the above two groups in terms of employment relationships, resources, power and working conditions.
Thus, combining the two factors of institutional division and market situation, five typical middle classes can be distinguished. The system division comprises two different labour market environments in which the middle class is further differentiated by the market situation: the new middle class within the redistributive system; the new middle class outside the redistributive system; the marginal middle class within the redistributive system; the marginal middle class outside the redistributive system; and finally, the small owners.
Research hypothesis: external consistency and internal heterogeneity of the political values of the middle class
In response to the corresponding relationship between the middle class and political values, I propose the following three research hypotheses.
First, the middle class as a whole is generally situated between the upper and lower classes of society such that its political orientation may be different from other classes. The upper class is not included in the study because of the unavailability of survey data for this segment of the population. I believe that the middle class may have more liberal political values than the manual class class for two reasons. The first reason has a cultural and ideological interpretation. That is, the middle class pursues civil rights and has a higher level of political participation because it has a higher degree of education and awareness of freedom. The higher level of education and better quality of life help to cultivate individuals’ critical consciousness, leading to the emergence of citizens with a tendency to be critical about political authority. A second reason concerns the market because a considerable proportion of the middle class has a class experience involving market economic activities. Market transactions are based on the equal and free rights of individuals and groups. Therefore, a middle class that developed during the process of marketization will form a clear sense of the rights and benefits of liberalization, will express a strong sense of fairness and justice and will actively promote democratic change.Footnote 24 Based on this argument, the following hypothesis is proposed:
Hypothesis 1: Compared with the manual class of workers and peasants, the political values of the middle class, as a whole, will be more liberal.
Second, under the two different economic systems – the redistribution system and the market system – two types of middle class with distinct class experiences have formed. The units within an economic system are important organizational forms for the control of ideology and resource allocation. The close relationship between members of a class and their organizations is reflected in two aspects: political loyalty and interest attachment. As a result, members continue the inertia of the redistributive system in many aspects such as working methods, lifestyle and even thinking habits.Footnote 25 In contrast, a middle class outside the redistribution system emerged in the market economy. As mentioned above, the class experience of market economic participation makes participants more inclined to pursue freedom, equality and fairness. In addition, most of the members of this class are young with a high level of education and cultural capital. Their distinctive features are liberal thinking, high liquidity, wide dispersion, poor organization and active non-institutionalized participation (such as in social movements or internet expression).Footnote 26 They may also exhibit more enthusiasm for changes that result in more freedom. Therefore, I propose the following hypothesis:
Hypothesis 2: Compared with the middle class within the system, the political values of the middle class outside the system are more liberal.
Third, the three middle class types in the market economy (new middle class, marginal middle class and small owners) are somewhat heterogeneous in terms of their political values. Among them, the new middle class is a special sub-group whose members have achieved high economic status owing to their professional talents and managerial roles. In addition, members of the new middle class are very knowledgeable and possess a keen sense of fairness and justice. Because of these traits, they make appeals for greater political participation. However, the ruling party regards them cautiously and guards against the actions of and demands by the new middle class. The political attributes of the new middle class in the official system are not clear. The institutionalized channels of expression of interests are narrow, the policy space is limited and only a small number of outstanding members of the new class (for example, the media/legal elite, NPC deputies/CCPCC members) possess high-quality political resources. Most of the ordinary members of the new middle class are still on the edge of the political system and are excluded from the power structure.Footnote 27
Thus, the relationship between high economic status and low political status constitutes an asymmetric status discrepancy, which may lead to more liberal tendencies and resistance by the new middle class. In contrast, the market economy and political statuses of the marginal middle and small owners are not high and the status discrepancy pressure is low, so their political liberalism may not be as high as that of the new middle class. Therefore, we propose the following hypothesis:
Hypothesis 3: Among the middle classes outside the system, the new middle class may have a higher liberal tendency than the other two classes.
Research Design
Data
The data used in this study come from the 2015 Chinese General Social Survey (CGSS). This project, launched in 2003, is the earliest nationally representative continuous survey run by Renmin University of China. The project systematically and comprehensively collects data biannually from various levels of society – community, family and individuals – to trace trends in social change. The survey is seen as one of the most important sources of data for studying contemporary Chinese social issues. The CGSS project was accepted as part of the International Social Survey Programme (ISSP) in 2006.
The 2015 survey used the probability proportional to size (PPS) and stratified sampling methods (district/county, street/township, village/residential committee, residents) to complete 10,968 valid questionnaires in urban and rural areas in 28 provinces across the country. In the current study, the manual class was used as a reference group and, when added to the middle class, the final sample size was 8,782.
It should be noted that this study focuses on the comparison between the manual class and middle class, and does not include the upper class (elite social group). This is because the upper class is composed mainly of senior officials from the Party, government officials, senior managers of large state-owned enterprises, large business owners and senior intellectuals with social influence. This study is based on the analysis of sample survey data and the likelihood of being able to access upper class members is small, so the upper class is not included in the comparative analysis.
Measurement
(1) Middle-class categories
The positioning of the middle class depends on its structural position in the social system and not the amount of resources its members possess. Therefore, the definition of the middle class in this study is not based on the classification of income, education and prestige but on occupation, because typically, occupation embodies power relations based on resource possession.
Based on institutional and market position, the operational definitions of the five middle classes are as follows. First, the new middle class (management personnel, professionals and technicians) and the marginal middle class (general office staff, professional and technical assistants, and business service personnel) were distinguished by the occupational categories of the respondents in the data. Second, according to the ownership and unit type (state-owned sector and non-state-owned sector), they were then were subdivided into four categories: (a) the new middle class within the system, (b) the marginal middle class within the system, (c) the new middle class outside the system, and (d) the marginal middle class outside the system. Together with (e) the small owners (including the self-employed), who are themselves outside the system, a total of five basic group types were distinguished.
(2) Political values
Although theorists from opposing factions have different judgements with regard to the political functions of the middle class, most agree with the inference that if middle class members have a certain degree of political conservatism, they may become a force for maintaining social and political stability; contrarily, if middle class members hold a certain degree of political liberalism or political radicalism, they may become a force for promoting political change.Footnote 28 Therefore, the key to grasping the political function of the middle class is to judge its political values – that is, its citizens’ cognition and evaluation of the political system that are formed through specific class experiences and political socialization.Footnote 29
To measure political values, it is best to directly ask respondents whether they support or oppose specific social and political reform programmes and policy measures. However, the data in this research do not cover these kinds of questions. In fact, because of their sensitivity, it is difficult to ask such questions in current social surveys and obtain real answers from respondents. Therefore, this study uses measurements of public attitudes and behaviour commonly used by political scientists, and extracts resources from existing questionnaires. Based on theoretical and practical considerations and the assistance of a statistical test, I measure political values using three indicators: support for freedom, government satisfaction, and participation in political elections.
These indicators are an important reference point from which to observe an individual's political values. Support for freedom represents the individual's inner consciousness with regard to pursuing freedom from government intervention. As a political value, it advocates for the protection of individual freedom of thought and speech, promotes a transparent political system, restricts the powers of government by law, and guarantees the right of individuals to work and live freely. The second indicator, government satisfaction, represents the individual's external evaluation of the government's performance and trust in government agencies.Footnote 30 Participation in political voting is a practical reflection of political values. Grassroots elections provide an institutional channel for political participation. However, many members of the middle class believe that the voting process is just a formality so they do not participate in voting. Instead, they seek out ways of expression outside the system.
To summarize, if individuals are highly supportive of freedom, are dissatisfied with the government and do not participate in regular political voting, they are classified as political liberals. On the other hand, if they have little support for freedom, high government satisfaction and participate in regular political elections, they are classified as conservative.
To measure the indicators of support for freedom and government satisfaction, I use a set of questions for each indicator and employ Cronbach's alpha to measure the internal consistency of these questions. Cronbach's alpha measures how closely related a set of question items are as a group for measuring a specific variable or variables. The more similar or closely related the items, the higher the reliability of the questionnaire for measuring what it says it will measure. In general, a value that is higher than 0.7 implies that the scale or questionnaire has a high reliability.
The process for measuring the three indicators is as follows.
Support for freedom. In contemporary China, the number of births a couple can have is dictated by the family planning policy, work mobility and migration are regulated by the hukou system, and freedom of speech is subject to political censorship. Therefore, support for freedom of family planning, migration and speech can reflect the leanings of an individual towards political freedom, as well as alienation from the existing political system. In the 2015 questionnaire, three related items were set up to examine the degree of acceptance of statements on freedom of speech, freedom to have children and freedom of choice in work and life. Each statement was evaluated on a five-point scale, from completely disagree to completely agree. The Cronbach's Alpha value for these three items was 0.711.
Government satisfaction. This was measured by respondents’ evaluations of government performance in nine areas such as people's livelihoods, fairness and justice, environmental protection, public and safety. The respondents evaluated these items on a five-point scale, from very dissatisfied to very satisfied. The Cronbach's Alpha value for these nine items was 0.882.
Participation in political voting. Responses to the question “Did you vote in the last village committee election/community residents committee election?” were used as the operational indicator of whether or not an individual participated in political voting.
Since these original variables are categorical variables, especially “political voting,” which is a dichotomous variable, in order to better represent the structural relationship between these variables, I unify the three indicators of political values into a two-category encoding form: high/low support for freedom; high/low government satisfaction; and participated in political voting/did not participate in political voting. Table 1 presents the detailed measurements of the three indicators of political values.
Table 1: Measurement of Political Values
Source:
Questionnaire for the Chinese General Social Survey 2015.
Analysis model
The analysis strategy presented below uses the latent class model (LCM) to find typical political values from the three explicit categorical variables (see online appendix). It then uses the logistic regression model to examine the different political values of class groups, including both the difference between the middle class as a whole and the manual class as well as differences within the middle class.
(1) Latent class model (LCM)
The LCM relates a set of observed (usually discrete) multivariate variables to a set of latent variables. It is a type of latent variable model and is often used to find groups or subtypes (latent classes) in multivariate categorical data. A latent class is characterized by a pattern of conditional probabilities that indicates the chance that variables will take on certain values. Specifically, it finds a set of groups with common features from the eigenvalues of the joint distribution of categorical explicit variables (observed variables), the probability of each set, and the conditional probability of the explicit variables of each set that can be derived from it. Therefore, in the general form, the LCM is often referred to as the probability model.
The logic of LCM is to classify individuals based on their response patterns on explicit variables. For example, a mathematics test has ten judgement questions. Individuals with high mathematical ability might answer all questions correctly. Students with low ability presumably will only answer easy questions correctly. Students with medium ability might correctly answer all easy questions and solve moderately difficult questions. Students with different abilities show some similarity when answering questions of varying levels of difficulty. Therefore, students can be divided into different groups using their responses to the questions. In this study, there are three political values explicit variables: support for freedom, government satisfaction and political voting. In the analysis process, I first discover the latent classes behind the characteristics and distribution of the three explicit variables, determine the optimal number of latent classes, the meaning of each class according to the model-fitting condition, and then expand the relationship between the variables.
First, the number of latent categories based on the model-fitting indicators is determined. Table 2 lists the model-fitting indicators for different numbers of latent categories. It can be seen that the BIC and AIC values of the two-category model are the lowest. In addition, from the results of the significance test of the likelihood ratio statistic and the chi-square statistic, the two-category model has the best fit compared to the one-category and the three-category models. Therefore, I focus on the response probabilities of the different dimensions of the three explicit variables on the two latent categories to determine the characteristics of different latent categories.
Table 2: Fitting Indicators of Different Latent Models
Table 3 shows the results of the two-category latent class model. The basic characteristics of the two latent classes can be generalized from the conditional probability of the explicit variables on the two latent classes. Latent class 1 reflects the characteristics of low support for freedom (72.2 per cent), high government satisfaction (72.7 per cent) and more active voting participation (57.7 per cent), and so can be summarized as a conservative orientation. Latent class 2 reflects the characteristics of high support for freedom (57.0 per cent), low government satisfaction (68.4 per cent), and non-participation in political voting (75.9 per cent), which are typical of a liberal orientation. Of these two orientations, the percentage of latent class 1 (conservative) was much higher than that of latent class 2 (liberal).
Table 3: Results of Two-category Latent Class Analysis
(2) Logistic regression model
After identifying the typical patterns of political values through the LCM, I further investigate the relationship between the middle class and political values through the logistic regression model. In this model, the core variable of class position (middle class, manual class) is introduced and a series of control variables incorporated. These include social demographic variables such as gender, age, marital status and ethnic categories as well as socio-economic variables such as education, income and Party membership. In addition, internet use is employed as a variable to represent how information is obtained. Studies have shown that the media channels through which people receive information influence their political values. New media such as the internet provide diverse and heterogeneous information, including international cultural information, which may increase an individual's awareness of political freedom and social critical consciousness.Footnote 31
Empirical Results
Descriptive analysis
The descriptions of the variable are reported in Table 4. In the analysis sample, the middle class as a whole accounted for 41.6 per cent of the sample, and the manual class, including workers and rural workers, accounted for 58.4 per cent of the sample. Within the middle class, the marginal middle class outside the system was the largest group, at 10.6 per cent of the sample. Other middle-class groups were roughly the same size, around 7–8 per cent of the sample. In other control variables, the gender ratio was essentially balanced. The average age of respondents was 46.4 years. Most of the respondents were married (82.8 per cent). The Han nationality accounted for the vast majority (92.1 per cent) of respondents, and the percentage of ethnic minorities was less than 10 per cent. The percentage of ethnic respondents was consistent with the national distribution of ethnic populations in China. In addition, 9.7 per cent of respondents were Party members. In terms of socio-economic status, 37.9 per cent of respondents had a secondary-level education and 16.2 per cent of the respondents were college graduates or above. The average value of the annual income logarithm was 8.6. In terms of internet use, 34.8 per cent of respondents frequently used the internet to obtain information.
Table 4: Summary Statistics of Variables (N = 7683)
Source:
Chinese General Social Survey 2015.
Logistic model analysis: the differentiated political values of the middle class
After identifying the two typical political values (conservative and liberal), I construct a binary logistic model to analyse the differences in political values between class groups. Table 5 shows model results of the differences between the middle class as a whole and the manual class, while Table 6 shows model results of the differences among the various groups within the middle class. In both models, the control variables described above are used.
Table 5: Logistic Regression Model for Political Values between the Middle Class and Manual Class
Source:
Chinese General Social Survey 2015.
Notes:
Standard error in parentheses; significance level:+ p < 0.1, * p < 0.05, ** p < 0.01, *** p < 0.001.
Table 6: Logistic Regression Model for Political Values within the Middle Class
Source:
Chinese General Social Survey 2015.
Notes:
Standard error in parentheses; significance level: + p < 0.1, * p < 0.05, ** p < 0.01, *** p < 0.001.
It can be seen from Table 5 that the middle class as a whole has a significantly more liberal political orientation (coefficient 0.407) than the manual class, a difference that is highly significant. Thus, hypothesis 1 is supported by the empirical data. I believe that this liberal value can be attributed to middle class subjects’ high level of literacy and career experiences, resulting in a greater level of critical social awareness. These traits make it more likely for middle class subjects to demand freedom and justice than for manual class subjects to do the same.
With regard to control variables, the difference between men and women in terms of political values is not significant. Older respondents tend to have a more conservative political orientation than younger respondents. Married respondents are more conservative than unmarried respondents, and the difference in political values between Han and ethnic minorities is not significant. Consistent with general expectations, the political values of CCP members are conservative. With regard to socio-economic status, the higher the level of education and income, the more liberal the respondents’ political values. This is consistent with the findings of previous studies. One explanation for this finding could be that people with higher levels of education tend to have a greater level of social critical consciousness. Respondents who use the internet more frequently show a significantly higher liberal orientation than those who do not. This finding could be attributed to the diverse information on the internet and its liberalizing effect on the middle class.
Table 6 presents the heterogeneity of the political values within the middle class. The model results show that with reference to the new middle class within the system, the coefficients of other middle-class groups outside the system are all positive and statistically significant. This is presented as follows: (a) the new middle class (0.547); (b) the marginal middle class (0.203); and (c) the small owners (0.363). However, the coefficient of the marginal middle class within the system is not statistically significant. These results support hypothesis 2 that the political values of the three types of middle-class groups outside the system are more liberal than the two types of middle-class groups within the system. Institutional segmentation does form a basic dividing line for the political values of the middle class. The middle-class groups within the system all show a low sense of political freedom, higher political support and regular political participation. This could be because of their organizational power dependence, practical consideration of interest dependence or value preference for ideological political correctness.
Further, among the three types of middle-class groups outside the system, the new middle class has the highest degree of political liberalization, the small owners the second highest, and the marginal middle class comes third. The results support hypothesis 3 that among the middle classes outside the system, the new middle class may have a higher liberal tendency than the other two classes. It can be said that the results are in line with the logic of market differentiation. Because new middle class members have higher incomes from their jobs as managers and their professional talents, the economic status of the new middle class is significantly higher than that of the other two groups. However, the channels for new middle-class members to express their interests and political participation are limited. The contradiction between high socio-economic status and low political status creates tension, which then stimulates political liberalism. In contrast, the marginal middle class and small owners are excluded from the system so their political liberalization is higher than that of the groups inside the system. On the other hand, their socio-economic status is not high, being roughly the same as their political status, and they do not have many political and economic status deviation pressures. Therefore, their political liberalization is not as high as that of the new middle class.
Another point worth mentioning is that the middle class as a whole, including the divided groups, not only has a higher liberal value than the manual class but also the political liberalization of each type of middle class is higher than that of the manual class. This finding is derived by comparing regression coefficients. Specifically, the political liberalization of each class group can be sorted as follows: the middle class outside the system > the middle class within the system > the manual class. To a large extent, the results undermine the argument made by many scholars that the Chinese middle class is a social stabilizer. The results reported here do not support this assumption. Whether as a whole or as an internally subdivided group, the political attitudes of the middle class are more, rather than less, liberal.
So far, the relationship between the middle class and political values has been presented in a relatively clear pattern. The middle class as a whole shows a more liberal political orientation than the manual class, and the orientation of groups within the middle class is highly heterogeneous. The middle class in the redistribution sector, whether it is the new middle class or marginal middle class, shows a more conservative tendency, while the political values of several types of middle-class groups outside the system are more extreme, especially the political values of the new middle class. In short, the political function of the middle class in China cannot be generalized. Instead, discussion should focus on multidimensional analysis. The results do not support the argument that the middle class is a stabilizer; however, they do not fully support the argument that the Chinese middle class is a force for promoting political liberalization either, because the middle class within the system shows a higher conservative value than the middle class outside the system. It appears that of all the groups, the middle class outside the system, especially the new middle class, may be the group with the most potential for driving political liberalization in China.
Conclusion and Discussion
Unlike the middle classes in Western societies, the mode of formation and structural composition of the Chinese middle class are unique. In the process of modernization and market transformation, a variety of middle-class groups have been brought together in contemporary China. Therefore, studies of the social and political functions of China's middle class need to explore multidimensional issues in order to understand its integrity and internal heterogeneity. Generally, the internal heterogeneity of this class is no less than the difference between it and other classes. When discussing the heterogeneity of China's middle class, it is important to understand two factors – market situation and institutional division – and the role that state power and the political process play in shaping class structure. These socio-political issues are key to research on China's middle class because they distinguish it from the growth of the middle classes in Western societies and in other societies in East Asia.
The empirical results show that two typical socio-political values (conservative and liberal) can be identified through degrees of support for freedom, government satisfaction and political voting. A conservative value is typically embodied as low support for freedom, high government satisfaction and regular political participation. A liberal value represents high awareness of political freedom, low government satisfaction and low political participation. In the correspondence between political values and various group categories, the following characteristics are observed.
First, the middle class as a whole is not necessarily a social stabilizer and its liberal orientation is significantly higher than that of the manual class. The dominant perspective of previous research is that China's middle class is characterized by conservativism. However, the research findings presented in this study do not support this view. One possible explanation for the divergence in findings is that the prevailing views asserting middle-class conservatism are based more on theoretical inference than on reliable empirical data. Another explanation is that the political values of the middle class are constantly changing. With the growth of the middle class and the awakening of class consciousness, it may well be the time for collective action, with the middle class as the main actor. The widespread use of online media has meant that more and more members of the middle class have been affected by a fervour for collective action and their radical consciousness has also increased significantly. In short, whether based on cultural interpretation or market interpretation, the Chinese middle class is showing a more radical tendency than in previous times.
Second, the still forming and developing Chinese middle class is a highly diversified group and its differentiated political values are centred around the two factors of institutional division and market differentiation.
The political values of the three middle class groups outside the system are significantly more liberal than those of groups within. The redistribution system acts as an effective tool of state rule. It integrates the state's powers of command, resource distribution and ideological control, resulting in a dominance and obedience type of relationship with its subjects. Even in the post-danwei 单位 period, the interest dependence and political loyalty of middle class members inside the system with regard to state organization have not changed.Footnote 32 Thus, unlike the middle classes in the West and other East Asian societies, the function of the Chinese middle class is not only driven by the logic of modernity and the market but also by the direct and powerful interventions of government. State power continues to play a huge role in shaping and reshaping China's class structure.
The middle class outside the system, especially the new middle class, constitutes the driving force for potential change. This may be owing to higher education levels and class experience in market competition. It may also be owing to the contrast between its members’ higher economic status and lower political status. At present, the new middle class outside the system is officially defined as the “new social class” by the government. It has experienced changes in its attributes in the past 40 years, from the political exclusion of the 1980s and 1990s, the political absorption of the early 2000s, and the political integration of the new era.Footnote 33 In general, the new middle class's institutional channels for political expression and participation are still narrow, the policy space is still limited and most members are still excluded from the power structure. If and when the new middle class's economic security is no longer guaranteed and the political participation aspirations of its members are not realized, it may have a greater ability to mobilize its resources than the marginal middle class. The new middle class may also use the tenets of ideological discourse and modern values to influence and mobilize the public and join other dissatisfied groups to advocate for reform or even revolution.
Acknowledgements
The research on which this paper is based was supported by the National Social Science Foundation of China (Grant No. 17CSH031).
Supplementary material
Supplementary material is available online at https://doi.org/10.1017/S0305741020000296.
Conflict of interest
None.
Biographical note
Guangqiang QIN is associate professor of sociology at the Minzu University of China. He studies the formation of the middle class and its social and political functions in contemporary China. He is also interested in citizens’ subjective response to the past forty years of changes in China.
