Ilona Babenko is Associate Professor at W.P. Carey School of Business at Arizona State University. This post is based on a recent paper by Professor Babenko, Viktar Fedaseyeu, Assistant Professor in the Department of Finance at Bocconi University, and Song Zhang, University of Lugano and Swiss Finance Institute.
Do CEOs affect political choices of their employees? Using a large sample of U.S. firms, we find evidence that they do. First, we document that employees donate significantly more money to CEO-supported political candidates than to otherwise similar candidates not supported by the CEO. In 2012, for example, Barack Obama raised three times more money from employees of firms whose CEOs donated to him than from employees of firms whose CEOs donated to Mitt Romney (see Figure). We find similar effects for all federal elections (House, Senate, and President). Second, we find that employees located in congressional districts where CEOs support political candidates are more likely to vote in elections, suggesting that CEOs can affect not only their employees’ campaign contributions but also voter turnout.
Interactions between firms and politicians have received substantial attention in the literature. Firms can invest in political capital by establishing direct connections with legislators, for example through employment of current or former politicians (Faccio (2006), Faccio, Masulis and McConnell (2006)), by spending money on lobbying (Borisov, Goldman, and Gupta (2015)), and by financing candidates’ political campaigns through corporate political action committees, known as PACs (e.g., Cooper, Gulen, and Ovtchinnikov (2010), Akey (2015)). These activities may create value for firms’ shareholders because of subsidies, preferential allocation of government contracts and external financing, less strict regulation, and lighter taxation. Yet another, and largely unexplored, mechanism through which firms can establish political connections is campaign contributions made by their employees. This mechanism can be important since political contributions made by individuals in the United States far exceed those made by corporate PACs (e.g., Ansolabehere, de Figueiredo, and Snyder (2003)).
Our first result is that CEOs and employees tend to donate to the same political candidates. This relation per se is not necessarily causal, but neither is it to be expected. On one hand, the livelihoods of both the CEO and the firm’s employees are at least partially tied to the success of their firm, which makes them susceptible to common economic forces and may result in a contemporaneous relation between their political donations. On the other hand, participation in the political process is a high civic duty of an individual and is not determined by economic factors alone. Moreover, employees typically have different socioeconomic characteristics from those of CEOs and may favor different political outcomes. Thus, we further analyze the relation between CEO and employee contributions and investigate the mechanisms that may give rise to such a relation.
The relation between CEO and employee political contributions may exist either because they share a common set of political and economic goals or because CEOs explicitly advocate for their preferred candidates. We find evidence consistent with the second explanation. In particular, we show that CEO influence cannot be explained by common geographic factors or candidate strength. We also show that the relation we document holds around CEO turnover (including plausibly exogenous turnovers caused by natural retirement or death), which indicates that changes in employee contributions result from changes in how their CEOs donate and not the other way around. Also consistent with the idea that CEOs have direct influence over their employees’ political choices, we show that the link between CEO and employee campaign contributions is strongest in firms that explicitly advocate for political candidates. While we cannot observe all communication by CEOs on election matters, the U.S. federal law requires corporations that spend more than $2,000 per election on express advocacy of the election or defeat of a political candidate to report such communication costs to the Federal Election Commission. We find that the estimated effect of CEOs on employee contributions goes up by more than five times in firms that report communication costs. Thus, our results cannot be fully explained by common economic shocks or reverse causality (i.e., by CEOs observing political preferences of their employees and contributing to the candidates favored by them).
We also analyze cases when CEO impact is likely to be most effective and find that politically connected CEOs and CEOs of firms from heavily regulated industries are more successful in influencing how their employees donate. Further, employees are more likely to contribute money to the candidates supported by the CEO if those candidates are members of congressional committees with direct jurisdiction over the firm.
Campaign contributions are not the only way in which employees participate in the political process. Since most employees are voters, they can also directly affect electoral outcomes by going to the polls and voting for a particular candidate. While anecdotal evidence suggests that CEOs attempt to influence how their employees vote, we are not aware of any study analyzing the success of such attempts. Prior research does show, however, that voter turnout can be substantially increased through direct contact and communication with the voters (e.g., Gerber and Green (2000)). To study whether CEOs are effective in mobilizing their employees to vote, we use individual employee survey data from the NBER’s Shared Capitalism Research Project, which contains self-reported information on voting behavior (Kruse, Freeman, and Blasi (2010)). We find that employees located in areas in which CEOs make campaign contributions are approximately 11.5% more likely to vote than employees in other areas.
Overall, our evidence indicates that CEOs are a political force, with potentially important welfare implications for firms they manage and for the nature of democracy. These welfare implications depend both on whether CEOs promote their own political agenda or act in the interests of the firm, and on whether the interests of the firm coincide with the interests of its employees. If CEOs only promote their own political agenda, then their impact on employee contributions is likely to be welfare decreasing. While we cannot rule out that CEOs pursue personal political goals, our evidence does indicate that at least some of their political impact is driven by the interests of the firms they run: CEOs appear to exert greater influence on employee contributions in regulated industries and to candidates with jurisdiction over the firm, both of which are likely to be correlated with firm-specific benefits from political participation but not necessarily with CEOs’ political ambition.
The full paper is available for download here.