Political Connections and Market Structure (with Camilla Roncoroni)

This paper empirically investigates how political connections affect entry into a retail industry. Political connections can result in a privileged relationship with the regulator, thus allowing the connected firm to profit from political advantages. However, political connections are likely to hurt non-connected competitors, and imperfect competition can ultimately decrease consumer welfare. We focus on the largest grocery retailer in Italy, Coop Italia, a network of consumer cooperatives which has historical links to political parties. We measure the strength of its political connections across geographical markets by counting the number of cooperatives’ directors who have also been local government officials.

Our model allows political connections to distort firms' ability to enter a market in three ways: by imposing higher entry costs for non-connected firms, by lowering entry costs for the connected firm, and by helping the connected firm produce more informed forecasts of competitors’ profitability and entry decisions.

We estimate a game-theoretic model that accounts for both the interdependence among firms’ entry decisions and the effect of market variables, including a measure of political connections. The informational environment is affected by the presence of political connections, and the connected player may be better informed than its competitors about market conditions. To take this into account, we adopt a new method to estimate the entry game under weak assumptions on the informational environment. We allow the connected firm to have privileged information, without assuming that it necessarily has it. Our approach provides confidence sets for parameters, including parameters that measure the effect of political connections on firms’ entry decisions.

We find a positive effect of political connections on cooperatives’ profits, and a negative effect on some competitors. In a counterfactual, we examine the effect on market structure of a policy that makes entry regulation more transparent, thus removing the effect of political connections. We also link market structure outcomes to consumer welfare and estimate a model of demand and pricing for supermarkets. We obtain bounds on the expected welfare change and quantify the welfare cost of connections.