Reagan and the Wonks

Article excerpt

I recently testified before the U.S. Senate on the issue of capital punishment, a topic on which I have done some empirical research. In addition to myself, there was another social scientist who had also done research on the topic, and two women whose daughters had been murdered. That is, there were witnesses who had empirically and scientifically studied the issue, and two individuals with personal experience with homicide. The stories told by both women were heartbreaking, but nonetheless one wonders exactly what sort of useful information those two people could provide on the issue of optimal policy with respect to murder.

This pattern is not unusual. Congressional hearings generally include witnesses who have some identifiable connection with the issue under study, even though they may have no particular expertise in analyzing the issue. Humans seem to find anecdotes and tales of identifiable individuals useful in discussing policy issues. Indeed, I began this column with an anecdote about myself in an attempt to catch the interest of readers. Overall, however, economists such as me are generally bad at using such evidence, which may be why we are generally not successful politicians (with some exceptions, such as Phil Gramm and Dick Armey). We are known as "wonks" or "bean counters," and individuals who deal in anecdotes are more successful as politicians. For example, lawyers, who may not understand empirical analysis, deal with cases involving individuals in their day-to-day professional lives. For this reason, they may be more successful in democratic politics.

Why do we observe this pattern of political decisionmaking? What are its implications?

NEIGHBORS AND NUMBERS To understand why humans pay so much attention to identifiable individuals and so little attention to data about masses of individuals, it is useful to consider the environment in which we and our decisionmaking methods evolved. For most of our history as humans and all of our history as prehumans, our ancestors lived in small groups probably no more than 100 individuals. In such groups, everyone knew everyone else, at least by sight and reputation. More importantly, if something happened to a particular individual, this was by definition a high-probability event. That is, if a neighbor in a population of 100 was eaten by a saber-tooth tiger in a certain locale, then the risk of death from going to that locale was at least 1 percent, and it would have been wise to avoid that place.

On the other hand, we had no need to understand probabilistic events involving millions or even thousands of people, since no decision would have had any observable impact on such a conglomeration. We have absolutely no evolved intuition for understanding an event involving 1 million people. We can, of course, use empirical and statistical methods to analyze such events, but the understanding of that analysis is wholly intellectual, not intuitive.

This dependence on observation of individuals is what we bring to the political process. It explains why successful politicians (who must appeal to voters) rely more on stories about individuals than would seem rational to social scientists. This strategy is available to liberals and conservatives; Ronald Reagan was adept at using politically adroit anecdotes.

SEEN AND UN SEEN There are several implications for political decisionmaking from this understanding of our evolved decisionmaking tools. I discuss a few.

First, elections are themselves about identifiable individuals. We pay excessive attention to the personality of a politician. Most of us will never meet the president, and will have no more than a passing acquaintance with our senators or congressmen. Nonetheless, we are concerned with aspects of their character (are they haughty and aloof or friendly, do we enjoy listening to their speeches) that are irrelevant for making important political decisions. …