Lobbying and the collective action problem: Comparative evidence from enterprise surveys

Industry lobbying is traditionally thought of as a non-excludable good subject to collective action problems that are most easily solved by concentrated industries. However, there is very little empirical support for this hypothesis. In this paper, we address a major shortcoming of existing work on the topic: Its nearexclusive reliance on data from the US. Using comparative firm-level survey data from up to 74 countries, we construct an industry-level indicator of concentration and test its effect on firms' lobbying activity. Using multilevel Extreme Bounds Analysis and Bayesian Variable Selection techniques to account for model uncertainty, we find no evidence that industry concentration is a predictor of lobbying activity. We discuss the implications of these non-findings for the literature and outline possible avenues for further research.