By Christopher Cotton, Queen’s University

U.S. Rep. Mick Mulvaney, Donald Trump’s pick for budget director, has attracted attention for asking, “Do we really need government funded research at all?” Although it’s not clear whether Mr. Mulvaney is against all government research funding, the question raises concerns of academics and proponents of evidence based policymaking.

Coauthor Arnaud Dellis (UQAM) and I speak to the impact of government funded research (or the lack thereof) in a research article (ungated) published in the most recent issue of the Journal of Law, Economics and Organization.

In that article, we develop a game theoretic model of policymaking in which both privately- and publicly-funded research helps policy makers learn about the costs and benefits of alternative reforms. We show how the issues and reforms on which policy makers focus their attention depends on the extent to which they rely on private sources to learn about policy. As public research funding increases, their reliance on private interests to fund research decreases, and policy improves.

This is because politicians who must rely on special interests to produce research are more likely to prioritize the issues and reforms supported by well-funded organizations, where research will be available. Even if these are not the same issues where refrom is expected to provide the most benefit to constituents.

Our model takes a very optimistic view of the role that interest groups play in politics. In our model, interest groups only produce research to help a politician better understand the costs and benefits of alternative policies. That’s it. They can’t distort or hide unfavorable evidence. There are no political contributions or elections or future career concerns. The traditional channels through which interest groups are seen to distort policy have been eliminated.

This may be seen as a best case scenario for the role interest groups play in policymaking. But even with the deck stacked in favor of interest groups being beneficial, we show how government reliance on them for research leads to worse policy.

The problem is not that the private provision of research is itself bad. Special interests produce helpful evidence and lobby politicians on their issues. This helps politicians make better decisions on issues with active lobbies.

Rather, the problem comes from a shift in attention away from more important issues and to issues with active research lobbies. It shifts the policy agenda to focus on more-wealthy or better-organized special interests, which can on average hurt voters.

This shift in priority is most likely to arise when politicians find it difficult or costly to learn about issues without the help of special interests. The distortion does not arise when the availability of public research drives down the costs to politicians of learning about policy without relying on special interests.

Imagine a simple example where a politician must decide whether to make educational reform or the reform of bankruptcy laws the focus of her term in office. The politician knows that education is the most important issue for her constituents, and that in the absence of private research lobbying, she will devote the few resources she has to learn about and implement education reform.

The interest groups recognize this. Those in support of bankruptcy reform anticipate that their issue will be sidelined in favor of education reform if they don’t work to influence the politician. And, so they organize. They commission research and produce evidence about the benefits of bankruptcy reform and the best way to implement it. They lobbying, offering information and support if the politician is to focus on their issue.

An active lobby in support of bankruptcy reform could have one of two effects on education lobbying. First, it could spur an education lobby to organize, and to produce its own research, in order to offset bankruptcy lobbying and maintain priority of education on the politician’s agenda. This is known as counteractive lobbying in political science. If it happens, then the politician will learn about both issues before choosing on which to act. In this case, the private sector produces research on both issues. Here, relying on the private sector to produce research does not lead to worse outcomes.

A more-troubling possibility is that the special interests in support of the education reform choose not to mobilize. This could be because a collective action problem makes organization difficult, or because the education lobby finds it strategically rational to refrain from costly political participation, sit back and wait, and hope that bankruptcy reform fails and that priority shifts back to their issue.

In this case, the reliance on the private sector to produce research about the impact of policy can lead to worse outcomes. When the bankruptcy lobby succeeds in shifting the agenda, the politician ends up acting on the less important issue without even learning about the more important issue. She no longer prioritizes the more promising education reform, and as a result, her constituents can be worse off.

As public research funding increases on the most-promising issues, the potential for such distortions disappears. In our example, the distortion depends on it being relatively difficult for the politician to learn about the education reform without help from a special interest.

If public research on the education reform is readily available through publicly funded academic or government sources, then the politician is unlikely to shift priority away from the more-promising education reform. The bankruptcy lobby may still produce research in the hope that, counter to expectations, the research shows that bankruptcy reform should be given priority.  But, priority will not shift unless that shift is good for constituents.

Even when one takes the most optimistic view of interest groups and lobbying, and assumes that their sole means of influence is through the production of research, the rich and powerful special interests have a disproportionate impact on policy. Improving politician ability to learn about issues without relying on special interests and lobbyists will lead to better policy.

This is an argument against recent trends to reduce congressional staff, or limit the capacity of internal research offices such as the Congressional Research Services or the Government Accountability Office. It is a reason to support federal funding of independent research through the National Institute of Health or the National Science Foundation. It is reason to be cautious of calls to eliminate government funded research.