Reflections on Economic Research and Public Policy John R. Commons Award Lecture, 2002

Article excerpt

An early draft of the bylaws of the American Economic Association states that economists who have attained a certain age are entitled to present papers that begin "Reflections on --." The exact age is unclear because the manuscript is faded, but I have been assured by competent authorities that whatever that age is, I have surpassed it. The purpose of entitling a paper "Reflections on --" is to warn the audience not to expect either a comprehensive or a systematic treatment of the subject. "Reflections" seems preferable to its synonyms, "meditations" or "ruminations." The former is too "new age," while the latter suggests the chewing of cud.

I begin by noting that economists undertake research for a wide variety of reasons. Many wish to pursue a teaching career, but they know some publication is essential if employment is not to be temporary. At one time the median number of publications of economists after the dissertation was zero. By contrast, some economists make their living by doing research or supervising the research of others. Approximately one third of the members of the AEA (and a somewhat larger proportion of all professional economists) are not academics; they are employed in government, nonprofit organizations, or private industry. For many of them, the programmatic needs of their employers shape their research agenda.

Another significant motive for research is intellectual curiosity. Most economists like to solve puzzles. This may involve finding solutions to abstract problems or discovering significant patterns in data. Still another motive is the approval of one's peers. Paul Samuelson concluded his Presidential Address to the American Economic Association by saying, "In the long run, the economic scholar works for the only coin worth having--our own applause.'" Finally, some economists undertake research because they want to influence public policy.

This taxonomy must be immediately qualified in three ways. First, the same economist may have different motives for different pieces of research. Secondly, even a single piece of research may be undertaken for more than one reason. Finally, even research that was not undertaken to influence public policy may have that as a result. With this as background, I come to the principle question of this paper, "Does economic research have a significant influence on public policy?" My answer is more than you might think--but it's complicated."

To be candid, "more than you might think" is a bit of a fudge. What I really should say is "more than I thought when I started to write this paper." As an economist who has published many outstanding books and papers (so my wife tells me), I have frequently been frustrated by the slow pace at which policymakers embrace my ideas and put them into practice. I suspect that I am not the only economist with this feeling. A simple division of economists into two groups according to whether they think their own research has or has not had a significant influence on public policy would probably find the "has nots" in the majority. Frankly, in some cases this shows the good judgement of the policymakers, but that's another story.

One source of complication is the varying time lag between research and its effect on public policy. Some research affects policy almost immediately, while a half century or more may elapse before the influence of other research becomes evident. Also, some research has a direct, almost palpable effect on a particular policy, while other research has an indirect, albeit broader, effect. Additional complications arise in determining whether there is a firm causal chain between the research and the policy.

Some Examples From Public Finance

As a prime example of rapid direct effect, consider the paper by John Shoven and David Wise, "The Taxation of Pensions: A Shelter Can Become A Trap." Among other results, Shoven and Wise (1998) showed that the marginal tax rate on income in some sheltered retirement accounts was close to 100 percent. …