No alt causes – minimum wage raises increase union membership
Clemens 20 [Jeffrey & Michael R. Strain; October 2020; associate professor of economics at UC San Diego; director of Economic Policy Studies and the Arthur F. Burns Scholar in Political Economy at the American Enterprise Institute; National Bureau of Economic Research; “Public Policy and Participation in Political Interest Groups: An Analysis of Minimum Wages, Labor Unions, and Effective Advocacy”; 10.3386/w27902 KF]
We investigate the relevance of these threats to our estimation framework through a standard set of best-practice robustness checks. To investigate the relevance of preexisting trends in union membership, we use event-study estimators to check for such trends directly. We find no evidence that trends in union membership had diverged before the enactment of minimum wage increases. We also investigate whether we obtain similar results if we implement synthetic control methods. These methods are designed to ensure that “treatment” and “control” states followed similar trajectories before the implementation of minimum wage increases. We find that synthetic control methods yield results similar to our more basic, baseline approach. We further investigate the relevance of economic shocks that might shape both union membership trends and preferences over minimum wage increases. When we control for proxies 4 for macroeconomic conditions, we find that the relationship between minimum wages and union membership rates strengthens. Further, we find that variation in minimum wages is uncorrelated with changes in other labor market regulations. Our results are also robust to controlling for the partisan composition of state governments. Finally, we implement a simple matching approach to check for the relevance of differences in states’ baseline union membership rates. Our findings are robust to using these alternative methods. Taken together, our initial analysis finds robust evidence that recent minimum wage increases have led to increases in union membership rates. We next take our analysis further in an attempt to differentiate among alternative theories of the causes of union membership. To do so, we begin by investigating which groups of workers become more likely to be members of unions following minimum wage changes. To analyze the mechanisms through which minimum wage increases shape union membership, we consider three distinct groups of workers. The first group consists of individuals who are directly affected by minimum wage increases. The second group consists of individuals who may be indirectly affected due to their employment as high-skilled workers in minimumwage-intensive industries. The third group consists of individuals in industries that are unlikely to be affected by the minimum wage. Our proxies for these groups involve differences across industries, between the public and private sector, and between high- and low-education workers. As we show through analyses of wage data, high-education workers and public-sector workers have wages that are not affected by the minimum wage. Wage data reveal that the minimum wage’s direct impacts are concentrated among young and low-education individuals in low-wage food service and retail industries. 5 Our analysis reveals that minimum wage increases affect union membership rates among the groups we analyze in distinctive ways. First, we find strong evidence for gains in union membership among workers with high levels of education, in particular when they are employed in the public sector. Second, we find declines in union membership among low-skilled individuals in low-wage industries. 3 This second group contains the minimum wage’s direct beneficiaries. Importantly, this second group thus consists of workers for whom a minimum wage increase may substitute for the bargaining power associated with union membership. Finally, we assess several mechanisms through which alternative theories of union membership might work. First, we investigate whether minimum wage increases predict increases in union wage premiums. We find that they do not. Second, we find that news coverage tracks key moments in the legislative histories of minimum wage increases.4 Minimum wage changes predict spikes in the number of articles that jointly discuss the benefits of minimum wages and key players in the labor movement. Further, we show that events in the legislative histories of minimum wage changes do not increase total coverage of organized labor, but shift the character of that coverage towards articles about minimum wages. Finally, using polling data, we find that minimum wage increases predict increases in the favorability of public perceptions of labor unions. 3 Notably, this is true of a sample of individuals who remain employed in those industries, implying that they experienced direct positive effects of minimum wage changes. Indeed, we provide direct evidence that this group’s wages are quite strongly and positively impacted by the minimum wage changes we analyze. 4 Using counts of local news articles assembled using LexisNexis, we find that minimum wage increases are associated with increases in articles about the minimum wage, including those that quote labor advocates or otherwise discuss key players in the labor movement. We also observe increases in counts of articles about the minimum wage at crucial stages in the legislative process, including the months in which state legislators introduced minimum wage legislation and the months in which such legislation was passed. 6 Our findings relate to several literatures of potential interest. First, we add to a long history of work on the determinants of union membership. A relevant slice of this literature considers Right to Work laws. These laws enable workers to benefit from a union’s presence without paying dues, which raises free-rider considerations (Lumsden and Petersen, 1975; Ellwood and Fine, 1987; Moore, 1998). Our emphasis on the role of unions as participants in social or political movements connects to Freeman (1998), who analyzes the “spurt” of union membership growth connected to society-wide economic change during the Great Depression. In our analysis, we find evidence that union membership rates can be shaped by the enactment of popular labor market policies with which unions have actively associated themselves. That said, we also find that the free-rider considerations raised by studies of Right to Work laws are a relevant concern, as the minimum wage’s direct beneficiaries become less likely to join unions following minimum wage increases. Second, we contribute to the literature on the question of what unions actually do. This question is perhaps most famously associated with the aptly named “What Do Unions Do?” by Freeman and Medoff (1984). Much of this literature focuses on estimating union wage premiums (Freeman and Medoff, 1981; Hirsch and Schumacher, 2001) and assessing how changes in union density affect wage inequality (Card, 1996; Card, Lemieux, and Riddell, 2004; Farber, Herbst, Kuziemko, and Naidu, 2018). Some work in the economics literature has focused on the role of unions as “a voice both at the work place and in the political arena” (Freeman and Medoff, 1979). Our analysis of news coverage provides evidence that the labor movement participates in the political process, in part, by shaping news coverage of minimum wages. In this news environment, minimum wage increases predict increases in union membership and improvement in public sentiment toward unions. 7 Third, we relate our evidence to potential channels through which an interest group might accumulate members and influence. One channel, for example, is for a group to work towards the material benefit of its members (Buchanan, 1965; Olsen, 1965). Interestingly, we find that the minimum wage’s most direct beneficiaries become less, rather than more, likely to join unions. As noted above, this finding is consistent with models that emphasize a classic “freeriding” concern. In addition, we find that increases in union membership are prevalent among high education workers and public sector employees, neither of which benefit directly from the minimum wage. This finding is consistent with models in which membership is responsive to an interest group’s reputation for effective, public-spirited advocacy (Clark and Wilson, 1961; Wilson, 1973). Our analyses of news coverage and public opinion polls provide suggestive evidence on the mechanisms through which these effects might operate. Finally, research in political science and political economy has long analyzed the interplay between interest groups and policy. Within this broad area of inquiry, a relatively small set of papers has focused on the impacts of policy on subsequent politics.5 Research in this vein has focused primarily on cases in which policies created constituencies out of their direct beneficiaries.6 Anzia and Moe (2016) analyze a case in which policy directly shaped the potential influence of an already organized group — namely, public-sector unions. They show 5 Research on the activities of interest groups and other political factions has tended to focus on characterizing their effects on the political process rather than on the drivers of their membership prospects. Relevant theoretical studies in the political economy literature include work by Baron (1994), Grossman and Helpman (1996), Persson (1998), Besley and Case (2003), and Dewan and Shepsle (2011). Empirical evidence consistent with important roles for interest groups is somewhat sparse. Anzia (2019) argues that the literature’s lack of evidence on the importance of interest groups may reflect its lack of focus on subnational politics, which can yield more opportunities to execute credible empirical strategies than can research focused exclusively on federal politics. 6 Perhaps the most famous example involves the enduring constituency created by the Social Security system (Campbell, 2003). Schattschneider (1935) similarly emphasizes the interest groups created by the Smoot-Hawley Tariff of 1929. More recently, Clinton and Sances (2018) and Baicker and Finkelstein (2019) have analyzed the effects of access to Medicaid on political participation. The latter analyses find nontrivial but transitory effects. 8 that changes in public-sector labor law nontrivially shaped future politics and that legislators’ votes on these laws suggest a sophisticated understanding of their long-run effects. In our analysis, we find that minimum wage policy influences the subsequent fortunes of the labor movement. We provide evidence that labor groups can attract members by contributing to policy victories and shaping media narratives, which can elevate their status in public opinion. Our analysis thus connects to a broad line of research on how the efforts of organized groups can shape policy, public opinion, and their own future prospects. 7 Our paper proceeds as follows.