I recently came across an interesting paper in the Review of Finance by and
While previous research has focused on the advantages firms can derive from maintaining connections to politicians, we consider an orthogonal channel, that is, that CEOs with better political access can use corporate resources to help an incumbent politician stay in power. We use France as our research setting since a large fraction of publicly traded assets are managed by CEOs whose past professional experience involved serving in government. Our results suggest that political connections between CEOs and politicians may indeed factor into important corporate policies, such as job (plant) creation and destruction. Publicly traded firms managed by politically connected CEOs adjust their employment and plant creation (and destruction) practices in ways that are consistent with helping incumbent politicians in their bid for re-election. Specifically, both employment growth and the rate of plant creation increases at connected firms in election years, while the rate of plant destruction decreases. These practices are particularly strong in election years and in cities that are traditionally more contested. Consistent with the idea that these employment practices might be detrimental to firm performance, we find that accounting performance at firms managed by connected CEOs is lower than non-connected firms and decreases as the fraction of plants that are located in contested areas increases. We show that the lower performance is mostly driven by higher labor costs.
While politicians may, in part, return favors to connected firms through the granting of subsidies, we do not find a net positive effect of political connections on firm performance in the French context. Moreover, we do not think that our employment results are driven by increased access to government contracts for connected firms since there is no accompanying increase in sales or value added around election times for these firms. In fact, analysis of both the cross-section of firms and CEO turnover reveals a negative correlation between firm performance and CEOs’ connections to the political leadership. While our research design does not allow us to verify the causal effect of political connections on firm performance (since performance effects cannot be tied closely to the timing of elections), it does seem to suggest that in the French context—different from analysis in other countries—political connections might have large costs or lower benefits.
One can conjecture that the difference between our findings for France and some of the earlier papers is driven by the quality of the institutions across countries or the fact that France is a stable democracy. Maybe it is easier for corrupt politicians to bestow large favors on connected businesses in countries with a powerful dictator who has a large amount of discretion or where the rule of law is less established, while this political patronage is more limited in established democracies such as France. However, more research will be needed to understand the nature of this political gift exchange and the institutional and other constraints on the level of interaction between politics and business.