The following is the fifth, and last, in a series of guest posts from job market candidates working in Institutional and Organizational Economics (check out the first, second, third, and fourth). Watch for the rest of the series over the next couple weeks, and think about interviewing one of these fine students if you have an opening. (-PLW)
guest post by Miguel Espinosa
Rosen introduces Smith to Coase: Firms use in-house lobbyists when they need expertise on the firm, and they use external lobbyists when they need expertise on topics.
A great range of firms across diverse industries use knowledge workers such as IT personnel, managers, and lawyers. Firms sometimes hire them internally and sometimes outsource their services. How do firms choose one versus the other? In my job market paper, I answer this question using US federal lobbying data. The data are particularly suitable for answering this question because, contrary to previous studies that contain information only on outsourced activities, the lobbying data have detailed information on which activities are conducted inside the firm and which activities are outsourced.
I first propose a theoretical framework to understand this decision and then conduct extensive empirical work to test the validity of its predictions. In the model, two key differences between in-house and external lobbyists emerge in equilibrium. Internal staff specialize in the firm’s industry—that is, they are familiar with its culture, its internal practices, etc. External staff have scarce knowledge of the firm’s industry, but they are experts on the topics on which they work.
The use of an in-house lobbyist provides knowledge of the industry at the expense of paying compensation proportional to the skills that the lobbyist has acquired. I call this payment the cost of acquiring skills. This cost is proportional to how easy it is for lobbyists to access the information to conduct their work. When it is easy to access this information, then it is less costly to have someone in-house. For lobbyists, the information they need to perform their job includes knowing which firms are interested in which bills, which lobbyists are working for which interest groups, what other firms are interested in and so on.
If the firm hires an external lobbyist, it has access to a higher skill level than the in-house lobbyist provides. External lobbyists achieve these skill levels because they can leverage the acquisition of skills across several clients, while in-house lobbyists acquire fewer skills because they work for only one client. Despite their superior skills, the drawback of hiring external lobbyists is that they do not know what happens in the firm’s industry as well as in-house lobbyists do. Thus, when the firm hires an external lobbyist, it has to pay a cost to communicate the culture or the internal practices of its industry. I call this the communication cost.
Therefore, the main trade-off that the firm faces in choosing in-house versus external staff is between these two costs: the cost of acquiring the skills for the in-house lobbyist or the cost of communicating the culture of its industry to the external lobbyist. I show theoretically that firms use in-house lobbyists when they need to solve tasks that occur very frequently or that have low skill requirements. If the tasks are not that frequent or have high skill requirements, the firms outsource. Finally, if the culture of the industry is too expensive to communicate, firms are more likely to use in-house rather than external lobbyists.
To confirm the empirical validity of the theoretical framework, I conduct several empirical exercises. I first show some correlations using firms and time fixed effects estimations. Here I provide only two examples. I measure skill requirements of lobbyists’ activities using information on the witnesses interviewed in Congressional hearings. In the US Congress, bills are introduced and sent to committees, where they are studied in hearings. In these hearings, Congressmen invite witnesses with experience or expertise in the subject to testify. I obtain each witness’ occupation and classify it as either high- or low-skill. For example, a witness with a Ph.D., a professor, or someone high up in the hierarchy of a large firm (CEO, Senior advisor) is classified as a high-skill witness. Then, I show that firms use external rather than internal lobbyists when they lobby for bills that are studied in committees with a larger fraction of high-skill witnesses. Although the classification of high and low skills is supported by external measures such as O*NET measures, I show that the results are robust to different ways of defining high-skill occupations.
A second correlation exercise seeks to understand the relationship between the use of in-house lobbyists and communication costs. As mentioned, when firms outsource, they spend time explaining the culture and the internal workings and restrictions of the industry to external lobbyists. One measure of how lengthy or costly that process can be is the extent of regulation in that industry: industries with more regulation will take more time to explain. I use RegData, which provides estimates of the extent of regulation at the industry level. Then, confirming the main prediction about the relationship between communication costs and the integration decision, I show that firms belonging to industries with more regulation use in-house rather than external lobbyists.
Although these exercises provide a first step toward validating the theory, they do not use exogenous variation to determine the extent to which the predictions of the model are valid. Therefore, my next step is to use quasi-experiments to test two main theoretical predictions. First, firms use more external lobbyists when facing more skill-intensive activities. Second, firms are more prone to use in-house lobbyists when the cost of acquiring skills decreases.
I test these predictions using two different events. First, the BP oil spill of 2010 changed the type of problems that companies in the oil and gas extracting industry were facing. Before the event, these firms lobbied mainly for bills to improve their funding opportunities. They applied for grants, tax cuts, subsidies and so on. After the spill, although they continued to lobby for these benefits, they started facing more-technical bills. These new types of bills studied changes in safeguard requirements and technological investments to avoid another oil spill. Thus, the type of problems that the affected firms faced became more skill-intensive. Confirming the prediction of my model, I show that firms in the oil and gas extracting industry used external lobbyists more heavily after the oil spill.
The second event I use is the Honest Leadership and Open Government Act (henceforth OGA), which amended some parts of the Lobbying Disclosure Act of 1995. It brought about several reforms to increase transparency in the lobbying industry, and one of those changes involved the cost of acquiring relevant information to conduct lobbying activities. Related to this cost, there were two important amendments. First, lobbyists must file reports about their activities electronically. Before the Act, they could choose between filing the reports electronically or by hand. Second, and more importantly, the lobbying reports were put online in a public, searchable database soon after the reports were filed. Thus, as it became cheaper to access information about other lobbyists’ activities, the cost of acquiring skills to conduct advocacy activities decreased. Next, I empirically show that firms started using in-house lobbyists more frequently after the Act took effect.
All of the results in this paper are based on a simple and insightful economic force: increasing returns to the use of skills. This force has been present in economic debates at least since Adam Smith, but research has largely ignored its relevance to study the problem of the integration decision in a skill-intensive context. Aside from the current lobbying industry application, my paper underscores the relevance of studying alternative forces to understand the integration decision. Given how little we know about the workings of this force on the integration decision and how this force can interact with incentive-alignment issues, the mechanisms and patterns presented in my Job Market paper are the first step toward better understanding the seminal Coase question from a different angle.
Miguel Espinosa is a Ph.D candidate in Economics at the London School of Economics.