Liberalism and Postmaterialism in China: The Role of Social Class and Inequality

Abstract

The postmaterialist thesis holds that postmaterialist and liberal values tend to be strongest in affluent locations and among people in higher socioeconomic positions. We demonstrate the degree to which China fails to conform to these expectations and seek to account for Chinese exceptionalism. We use multilevel models fitted to 2006 Chinese General Social Survey data to test the postmaterialist thesis. In general, we find expected associations for postmaterialism but not for liberalism. Indicators of individual-level status, including household income, middle/upper class status, urban residence, and majority ethnic group status are not associated with liberalism. Provincial-level affluence is not positively associated with either postmaterialism or liberalism, while income inequality is positively associated with liberalism. We conclude that in highly collectivist cultures like China's, economic development can have unexpected effects on value change. Growing inequality, which people in lower-status positions perceive as a threat, can promote liberalism, while people who benefit most from rising affluence and growing inequality may be more inclined to support traditional than liberal values.

About the Authors

Tony Huiquan Zhang is a PhD candidate in the Department of Sociology, University of Toronto. He is presently completing his dissertation on postmaterialism and liberalism in China and cross-nationally using data from the Chinese General Social Survey and the World Values Survey.

Robert Brym is SD Clark Professor of Sociology in the Department of Sociology, University of Toronto, and a Fellow of the Royal Society of Canada. He has published widely on politics in Canada, Russia, and Israel/Palestine. The winner of numerous awards for his teaching and scholarly work, his writings have been translated into nine languages. His recent research projects focus on collective and state violence in Israel and Palestine; democracy and intolerance in the Middle East and North Africa; and student protest in Toronto and Montreal.

Robert Andersen is Dean of Social Science, Western University, London, Ontario, Canada. The author of numerous awards for his publications, he studies the effects of economic inequality on attitudes and political actions. His most recent research examines the interaction between national-level income inequality and individual-level economic position (as measured by social class or income) in their effects on various outcomes such as attitudes toward income inequality, social tolerance, support for capitalist ideology, and civic participation.

Notes

“Traditional vs. liberal” is sometimes called “traditional vs. secular-rational” (Inglehart and Baker 2000; Bomhoff and Gu 2012) or “authoritarian vs. liberal” (Flanagan and Lee 2000; 2003). “Materialist vs. postmaterialist” (Flanagan 1980) is sometimes called “survival vs. self-expression” (Inglehart and Baker 2000).

For all but one item, we converted responses to a 1-to-10 scale according to the different levels of agreement (from “strongly disagree” to “strongly agree”). The question on “preference regarding government investment” asks, “Assuming the local government is going to invest/provide funding for some projects/programs, please tell us the top three ones you would support.” Response options include: (1) environmental protection, (2) medical services, (3) policing, (4) education, (5) gyms and stadiums, (6) retirement insurance, (7) fighting poverty, (8) reducing unemployment, (9) unemployment insurance, and (10) museums and the arts. Following earlier definitions of postmaterialism, we categorized responses (1), (4), (5) and (10) as postmaterialist. Respondents received 3 points if a postmaterialist item was ranked first, 2 points if a postmaterialist item was ranked second, and 1 point if a postmaterialist item was ranked third. Materialist items (options 2, 3, 6, 7, 8, and 9) received negative scores (−3 if ranked first, −2 if ranked second, and −1 if ranked third). Summing the scores produced a measure of support for postmaterialism. After standardizing all six variables, we calculated mean values for the materialist-postmaterialist items and for the traditionalism-liberalism items.

The Han form the overwhelming majority at nearly 92 percent of the population. The Manchu constitute less than 1 percent of the population. Their socioeconomic position is higher than that of the Han.

For all covariates, we used multiple imputations (n = 5) to handle missing data (Allison 2002). For reporting hierarchical linear model results, the MI package in R was used to pool the results and produce estimates from imputed data sets.

We also used robust regression diagnostics (Andersen 2008) within and between provinces to explore for influential cases, multicollinearity, and heteroscedasticity. We did not detect any significant problems.