This article is adapted from Clifford S. Russell's introductory essay in a new RFF book, Collective Decision Making: Applications from Public Choice Theory.
Public choice is concerned with the mechanisms by which human societies make decisions about their collective lives. It is distinguished from the older disciplines of "government" and political science by matters of substance and style. And, though many of its practitioners have come to it from economics, its interests are not the same as those of traditional economics, even that part dealing with such government concerns as the theory of public finance. Public choice is a new field. Its birth is usually dated from 1951, the year Arrow's pathbreaking Social Choice and Individual Values was published, but many earlier scholars contributed to its development, especially in the area of voting systems and their properties.
It seems to me useful to think of public choice theory as having two major distinct, yet occasionally intertwined threads. The first of these is the study of rules (institutions) for arriving at a collective choice or ranking of alternatives (which may be individual issues or comprehensive combinations of such issues on the basis of the choices or preferences of the individuals making up the collective unit (committee, neighborhood, state, or nation). Such choices are made all the time by actual governments as, for example, when by one means or another we decide how much pollution to tolerate, how much national defense to provide, and how much to tax ourselves to pay for the desired "outputs." Arrow's contribution was a very powerful and mildly depressing impossibility theorem in this area. He showed that there is no mechanism for aggregating individual preferences that satisfies certain quite plausible and even innocuous-sounding conditions. It would not be unfair to say that a large fraction of the energy of public choice theorists since Arrow has been devoted to qualifying, expanding, and interpreting this fundamental theorem, an enterprise which has involved the use of set theory, symbolic logic, and other mathematical tools that have given the field a special flavor of abstraction.
This part of public choice theory begins with the assumption that the individuals making up the collective have well-defined (and, almost always, rational) preference among the alternatives to be ranked or chosen from. Participation in the preference aggregation exercise is assumed and usually, but not always, it is further assumed that each individual's true preferences are what get aggregated. The central question may be thought of as, what properties will the aggregated or social preference have under different sets of assumptions about the individual preferences and the aggregation technique?
Private decisions and public choices. The other major threat to the theory begins at a different point and asks quite different questions. Here the interest has been to insert into the actual world of collective choices the homo economicus who is so useful and familiar in the world of private, market choices, and to see how the creature behaves.
Because in real life the market and political worlds are linked at innumerable points and many people are active In both simultaneously, it may seen odd to think that a special theory should grow up around the application of economic concepts to the study of politics and public decisions. But it is really not so strange, for there is a strong tradition in our political theory that people ought to act differently when engaged on public business and that they in fact do act differently. Public choice theory says: Let us assume that people act as rational maximizers of their own wellbeing whether they are deciding which shirt to buy or which candidate to vote for or, indeed, whether to vote at all. Put these self interested people into the real contexts of public life, as voters, candidates, bureaucrats, and so forth, and see how they behave. Are the predictions of the theory borne out by experience? Do they help us to understand what happens better than the more traditional view?
These questions have prompted public choice research on the motivation and behavior of legislators, bureaucrats, and on interaction of the two. The message of this research generally is that the self interested public actor will behave in ways quite different from those anticipated in the conventional wisdom of our political culture. Legislators will not be motivated to master issues, but rather to help constituents and contributors in dealing with the federal government. Bureaucrats will aim at maximizing their power (budgets, staff, and so forth), not at making socially optimal decisions.
As these predictions have the ring of accuracy, at least when the empirical test is the casual variety based on current events, it is tempting to think that the theory has got hold of a key that has eluded the traditional approaches. The task now should be to use this powerful insight to redesign our institutions to take explicit account of the self interested behavior of those who will run and staff them. There will be more assumptions that those in public service will act to serve the public.
A general impression of accurate prediction is not, however, the same thing as careful empirical examination of the theory, and it is the latter that should available before we go very far in redesigning our institutions. When such a careful examination is done, we find that there are problems. Consider, for example, the area of collective life in which the two threads identified above are most obviously tangled: the activity of voting, whether for candidates for legislative or executive office or in referenda on specific issues. Voting is far and away the most commonly used method of aggregating individual preferences to arrive at social decisions, and all conceivable voting systems suffer from the basic infirmity captured by the Arrow theorem. At the same time, the process of voting for candidates and in referenda is as close as most people get to politics and public decisions and so provides the major opportunities for us to behave either as selfish maximizers or selfless public citizens. A problem for public choice theory is that when the rational pursuit of self-interest is taken to be equivalent to the maximization of the expected value of an action, it can be shown to be irrational for individuals to vote at all whenever more than a handful of other voters are involved. This is because the probability that a single vote will affect the outcome is always so small that even large prospective differences in wellbeing that depend on the election outcome will not outweigh the small costs of actually casting a ballot. However, even public choice theorists know that people vote, so we must obviously look closely at the theory that leads us to this false theorem. On the other hand, the part of the theory (axiomatic social choice theory, as it is called) that looks at the process of arriving at decisions through voting, given participation, has had considerable success in predicting outcomes under particular structures.
Application of public choice. Therefore it seems clear that "application" of public choice theory must involve two distinct kinds of activities. First, there must be efforts to probe the reliability of the theory through empirical testing of its implications. Second, if we gain confidence in the theory, its lessons may be used to critique existing institutions and to design new ones. One important way to do this would be to begin with a set of properties that we would like a collective choice institution to satisfy. Then, on the basis of our theory, we would construct an institution satisfying those requirements. However, while it may be possible to obtain agreement— even unanimous agreement—on a set of principles or characteristics we want our new institution to satisfy, these may often turn out to be inconsistent. When they do, we face the much more difficult task of trading off one principle for another or one set for another set. The implicit values attached to each principle by each voter or citizen then become important, and the possibility of unanimous choice must decline sharply.
It is also interesting to observe the key role that is played by strategic behavior, which is nothing more than a consequence or manifestation of self interested behavior. Arrow's principle of "independence of infeasible alternatives" is simply a statement that people will behave strategically. That is, people will reveal their preferences incorrectly or insincerely when it is in their self-interest to do so. There seems to be a sense in which self interested maximizing behavior is a fundamental problem for human societies collectively—a conclusion which, if justified, is very different from the conclusions of microeconomic welfare theory for market systems.
There is no lack of major challenges for those who would make a contribution through the application of public choice theory. These challenges include empirical testing of the theoretical structure, analysis of existing institutions, and design of alternatives to established ways of making collective decisions. The prospective payoff is there, for we find wide agreement that specific public policy problems, ranging from energy through pollution and medical care to inflation and unemployment, are not being addressed adequately. Indeed, they may be caused or at least exacerbated by our current public choice arrangements.
Whether or not public choice theory can really help us with these difficult problems remains, as I stressed above, an open question. More seriously, it will almost certainly remain an open question for a long time because the necessary research is difficult and time consuming. There will be few payoffs along the way, though it has been demonstrated that some very real payoffs are possible where problems have manageable boundaries, and able researchers can devote significant time and effort. Typically, however, the would-be applier of public choice theory cannot promise short-term help with the problems so widely perceived; and the proposals coming from this field may not seem particularly responsive to the pressures felt by suppliers of research funds, both public and private. This is not, of course, a dilemma unique to public choice. It is really only a slight variation on the "basic" versus "applied" research debate which receives periodic airing in Science and the daily newspapers. The pendulum of consensus on the relative weights of near and certain as opposed to distant and speculative payoffs has moved steadily toward the former over the past decade. Its appropriate position is a public decision, though not one subject to a well-defined choice process, and claims from interested parties must be viewed with skepticism in this as in other debates.
A version of this article appeared in print in the January 1980 issue of Resources magazine.
