Draft 7.0: Comments Welcome

Special interests attempt to influence lawmakers through campaign contributions and through informational lobbying. Both avenues have been explored extensively in theoretical models. Only the former, however, has received much empirical scrutiny. We provide the first empirical tests of the major class of models of costly legislative lobbying, the Potters-van Winden-Grossman-Helpman (PWGH) model. To do so, we extend a simple PWGH model to encompass situations in which a legislature adjusts a pre-existing policy only periodically. We then test predictions of the model using data derived from over 50,000 observations of annual lobbying expenditures by special interest groups in the American states. We find that, as predicted, special interest groups 1) increase lobbying expenditures when the legislature is controlled by “enemies ” rather than by “friends”; 2) increase lobbying expenditures in budget years in states with biennial budgeting, relative to budget years in states with annual budgeting; and, 3) increasingly exit the lobbying process as lobbying costs rise. Overall, the results provide substantial empirical support for the PWGH class of signaling models of interest group lobbying in legislative settings.