Middle Class Identity in the Modern World: How Politics and Economics Matter
Curtis, Josh, Canadian Review of Sociology
ALTHOUGH THE STUDY of class identity has a strong tradition in research on politics and inequality (Centers 1949; Davis and Robinson 1988; Hodge and Treiman 1968; Jackman and Jackman 1983; Kelley and Evans 1995; Vanneman and Pampel 1977; Wright 1997), it has received little attention in recent years. The recent work focuses on either individual countries (Davis and Robinson 1998; Goldman, Cornman, and Chang 2006; Hayes and Jones 1992; Oddsson 2010; Shirahase 2010; Surridge 2007) or small groups of countries (e.g., Baxter 1994; Evans, Kelley, and Kolosi 1992; Johnston and Baer 1993; Wright 1997), making it difficult to uncover systematic cross-national differences in people's perceptions of their class position. In short, a rigorous analysis of data from many countries is necessary in order to systematically test the influence of national context on class identity. This article attempts to fill this gap.
Using World Values Survey data, the present paper explores the correlates of class identity in 15 modern societies. My purpose is to uncover the ways individual-level economic conditions, political ideology, and economic context interact in their effects on class identity. The ultimate goal is to demonstrate the conditions under which middle class identity bias--that is, identifying with the middle class regardless of objective class position--is most common.
MIDDLE CLASS IDENTITY BLAS
Research on class identification evolved from the work of Cantril (1943; 1944), Centers (1949), and others (Hodge and Trieman 1968; Ossowski 1963). Research typically finds that most people understand class labels and believe that social classes still exist (Argule 1994; Gilbert 2003; Jackman and Jackman 1983; Popitz et al.  1969; Reid 1998; Vanneman and Cannon 1987). Yet, it is also commonly found that people's self-reports of social class can be poor measures of actual class positions (Evans and Kelley 2004; Kelley and Evans 1995; Oddsson 2010; Rothman 2002). In fact, people often identify with the middle classes, regardless of their position in the social structure (Evans and Kelley 2004; Kelley and Evans 1995). Dominant middle class identities have been reported to occur in both affluent Western countries (Kelley and Evans 1995; Oddsson 2010) and former Communist countries (Evans and Kelley 2004; Evans et al. 1992).
This bias is often explained by "reference group theory," which suggests that many people cannot distinguish between classes, and for various reasons develop a dominant "middle class" consensus (Evans and Kelley 2004; Kelley and Evans 1995). This is alleged to occur because people consider themselves in relation to those around them. The homogeneity of references groups distorts people's perception of the class structure. If has also been suggested that when people refer to the "middle class" they tend to consider similarities in consumption patterns rather than production relations, income, or wealth (Clement and Myles 1994). This process of embourgeoisement is a common explanation for declining class consciousness, which is influenced by postindustrial economic change (Goldthorpe and Lockwood, 1963).
Reference group theory gets us part of the way to understanding dominant middle class identification, but it fails to explain cross-national differences. Factors at the national-level may influence the ways that people think about class and where they perceive their fit in the class structure to be. For example, some argue that when people feel that they share common interests, typically associated with economic experiences, their class identities are strengthened (Weakliem 1993). This can occur as a result of the political climate within countries or the ways in which income and wealth are distributed (Wright 1985).
Several studies argue that there is very little difference in class identity and awareness across the modern world (Evans and Kelley 2004; Kelley and Evans 1995). …