When do policymakers listen to policy analysis, and when do they ignore it?

Greg Nash

All politicians like to claim that their preferred policies are backed by "sound science" or "good analysis." The fact that they do this so often, and that politicians on opposite sides of an issue make these claims, increases public cynicism. The public justifiably doubts both the claim that a policy is backed by good analysis and, eventually, the worthiness of policy analysis itself. Recent debates on the Keystone XL pipeline, for example, included competing claims about the impacts of approval of the pipeline on both the environment and on the number of jobs created.

But sometimes good analysis does affect policy. The George W. Bush administration required the labelling of trans fatty acid (TFA) content in foods, and the Obama administration has moved toward banning TFAs in processed foods because of convincing evidence relating the incidence of cardiac disease to TFA consumption. Going further back, the Reagan administration was convinced by a cost-benefit analysis to require the removal of lead from gasoline.

In my new book, "Analysis and Public Policy: Successes, Failures and Directions for Reform," I examine the question of when policymakers listen to good analysis and when they ignore it. I studied four types of policy analysis: cost-benefit analysis, risk assessment, environmental impact assessment and impact analysis. I found numerous examples for each of these when analysis made significant differences in policy decisions. I also found examples where analysis was used to postpone decisions for decades and where it was ignored completely.

There are four broad categories of factors that determine the influence of analysis on policymaking:

Politics. The image of the politician demanding that analysts come up with a particular result persists in the popular imagination. I found no examples of this happening. That does not mean that politics is irrelevant, however. It is arguably the biggest element in the policy analyst's universe. The more attention a policy decision gets, the less room there is for analysis. For example, after a tragic accident, there is intense pressure to come up with a solution for the accident, and little attention paid to the merit of the solution. In addition, politically powerful interests often use analysis to argue that decisions that would burden them should be delayed, sometimes indefinitely.

Law. Policy analysts work in an environment shaped by congressional requirements. Sometimes Congress mandates a particular solution to a policy problem. All the analysis in the world won't allow federal agencies to override a bad idea from Congress. Congress can also require analysis of a policy questions, or sharply limit the use of analysis in a policy question.

Bureaucracy. Analysts must learn to advocate their views within the vast government bureaucracy. Their location in the bureaucracy can assist or impede their ability to do so. Access to decision-makers is important, but so is a degree of independence from program heads who have a particular agenda. The existence of an external office (such as the Office of Information and Regulatory Affairs, or OIRA) that supports analysis can be very helpful to agency analysts fighting for their ideas within their organizations.

The limits of analysis. The public (and some political leaders) expect analysis to give them answers. But policy analysis, whether it is of scientific questions or economic ones, is usually charged with making predictions about the future. And predictions are, by their very nature, uncertain. This means that policy analysts must become adept at explaining uncertainty to political leaders. These political leaders will then have to make decisions on questions that don't have clear answers, and then themselves be answerable to a public that wants certainty.

The heartening message of my book is that despite these challenges, sometimes policy analysis makes a difference. I suggest a number of ways that we can increase the influence of good analysis in policymaking. The most important of these suggestions is promoting an understanding that in a democracy, analysis is a tool to assist in decision-making. Good analysis can't give us the only answer to a policy question, but it can almost always shed light on the implications of different answers.

Shapiro is an associate professor and director of the Public Policy Program at Rutgers University and a member of the Scholars Strategy Network.