(When) are lobbying expenditures a good proxy for lobbying activity?

Résumé

Lobbying expenditures have become a commonly used proxy for measuring lobbying activity in research examining the role of money in politics. However, analyses of the assumptions underlying these approaches have been scant. I use data from unique lobbying disclosure requirements in the U.S. state of Wisconsin — clients report the hours their lobbyists worked for them and by how much the lobbyists were compensated for these hours — to examine the extent to which expenditures can be a valid proxy for activity, using the correlation between activity and expenditures. To address differences in levels across dyads, I also examine the relationship between changes in hours worked and changes in compensation amounts within lobbyist employer-client dyads. I find that lobbying expenditures can serve as a reasonable proxy, as indicated by strong overall correlations. However, this finding masks substantial heterogeneity by type of lobbyist. In particular, subset analyses show strong correlations for in-house lobbyists but moderate correlations for contract lobbyists, indicating the potential for systematic measurement error. I provide additional evidence pointing to the type of compensation by the client as a plausible mechanism and offer suggestions to improve the measurement of lobbying activity.