LABOR
Unions Cause People to Identify as Working Class
Data show union members more likely to identify as working class, but not the reverse
Photo via Creative Commons
By: Profs. William W. Franko and Christopher Witko
With Labor Day just behind us it’s a good time to reflect on the role of unions in the U.S. The research is clear: unions raise worker wages, improve benefits, reduce income inequality, and shrink both the racial and gender wage gaps. They also push policy in a more egalitarian direction.
In addition to producing these economic improvements and policy effects, unions also provide information and education to members, which might influence a variety of attitudes and beliefs. We examined the role that unions play in shaping class identity, and the relationship between class identity and policy attitudes in the U.S., in a recent article, “Unions, Class Identity and Policy Attitudes”, in the Journal of Politics.
What we find may surprise you: unions cause people to identify as working class.
Thinkers since at least Marx have argued that class identity — the subjective identification with a particular class group, such as the working class — is an important precursor of political attitudes and behavior. Marx, of course, thought that the working class identifying as such was essential to the overthrow of capitalism.
In modern democracies, class conflict has been channeled into mostly peaceful political activities and thus the interest in how class shapes policy attitudes and political behavior remains. Indeed, among scholars and journalists in the U.S., in recent years there has been keen interest in the attitudes and voting of different class groups, like the white working class, and educated, affluent professionals. Furthermore, scholars, pundits and average Americans have become more interested in identity politics. Yet, class identity remains understudied in the U.S., perhaps reflecting, as the eminent historian Sean Wilentz wrote of the closely related concept of class consciousness, “[it] is not so much studied and written about as it is written off from the start.”
Despite this, when asked in surveys, most Americans readily identify with a class group, most often the middle or working class. Further, while it’s often believed that Americans overwhelmingly identify as middle class, in fact about as many people identify as working class. Research also shows that Americans are about as aware of class distinctions as people in countries believed to have stronger class identities. Additionally, based on research conducted by ourselves and others it appears that class identity still shapes policy attitudes very much the way it did decades ago, with working class identifiers preferring more liberal economic policies, though this does not always translate into voting for the more liberal candidate based on other research we have conducted.
This raises the question of what shapes class identity?
Class identity develops starting early in childhood but continues to form into adulthood. Building off of research into interest groups, we argue that unions have a strong incentive to try and get their members to identify as workers or to think of themselves as belonging to the working class. Union leadership wants more people to identify with the working class because it makes them more likely to contribute to union political and organizing activities. If this is correct, we should see that union members are more likely to identify as working class than middle or other class groups, controlling for their objective class position and other factors.
To test this, we used data from the General Social Survey (GSS). The GSS asks people whether they identify with the lower, working, middle or upper class. Using an original measure of objective class that includes education, income and occupation, we find that union members are more likely to identify as working class regardless of their objective class group and other demographic controls. The effects are shown below.
One question that the GSS data doesn’t help us answer is whether people who identify as working class are also more likely to enter unionized jobs and seek union membership. If this is the case, then union membership is correlated with class identity but does not cause it.
To investigate this possibility, we also examined panel data gathered from the 1970s-1990s in multiple waves (the Youth-Parent Socialization Survey), which allows us to test whether union membership precedes class identity, or vice versa. We find that joining a union predicts future working class identity, but working class identity does not predict future union membership. This suggests that unions cause a greater propensity to hold a working class identity.
These class identities are politically important because they may influence policy attitudes and political behavior. Theoretically, we would expect that the working class would prefer more egalitarian economic policies than the middle or upper class and this is generally confirmed in existing studies. Thus, we examine the association between identification with different class groups and various broad policy attitudes using GSS data.
The questions ask whether the government should do more to reduce inequality, help the poor, help the sick, and do more to solve the country’s problems, which measures general support for government intervention into the economy and society. We can see below that in each case the working class has more left or liberal policy attitudes than the middle or upper class, though less liberal policy attitudes than those who identify as lower class. Note that these effect sizes are after controlling for objective class and demographics. For reducing inequality, working class identity rather than middle class identity has an effect roughly the size of being Black (around 5 percentage points). The other effect sizes are much smaller but are in the expected direction.
Our findings have important implications.
As most are aware, a far lower percentage of the workforce is unionized now compared to the 1970s. Our research suggests that this decline of unions likely has important implications for shifting class identities over time, as well as policy attitudes and voting, which we are currently exploring in a book project. For now, this research adds to the growing volume of literature that shows that identities are central to politics, and that the decline of unions in the U.S. has had wide-reaching consequences for society and our democracy.
William W. Franko is Associate Professor of Political Science at West Virginia University.
Christopher Witko is Professor of Public Policy and Political Science at the Pennsylvania State University.
