Assessing what lobbyists do from an empirical perspective is the focus of a recent and interesting paper by Marianne Bertrand (Chicago-Business), Matilde Bombardini (U. British Columbia), and Francesco Trebbi (U. British Columbia). Is it Whom You Know or What You Know? An Empirical Assessment of the Lobbying Process exploits an array of publicly-available data and considers competing models of lobbying. The abstract follows.
"Some believe that lobbyists’ main role is to provide issue-specific information and expertise to congressmen to help guide the law-making process. Others believe that lobbyists mainly provide the firms and other special interests they represent with access to politicians in their “circle of influence” and that this access is the be-all and end-all of how lobbyists affect the lawmaking process. This paper combines a descriptive analysis with more targeted testing to get inside the black box of the lobbying process and inform our understanding of the relative importance of these two views of lobbying.
We exploit multiple sources of data covering the period 1999 to 2008, including: federal lobbying registration from the Senate Office of Public Records, Federal Election Commission reports, committee and subcommittee assignments for the 106th to 110th Congresses, and background information on individual lobbyists.
A pure issue expertise view of lobbying does not fit the data well. Instead, maintaining connections to politicians appears central to what lobbyists do. In particular, we find that whom lobbyists are connected to (through political campaign donations) directly affects what they work on. More importantly, lobbyists appear to systematically switch issues as the politicians they were previously connected to switch committee assignments, hence following people they know rather than sticking to issues. We also find evidence that lobbyists that have issue expertise earn a premium, but we uncover that such a premium for lobbyists that have connections to many politicians and Members of Congress is considerably larger."