Source: Journal of Economic Literature, Vol. XXXVII (December 1999)
Sugden, Robert (University of East Anglia)
Review of "On Voting" by G. Tulock.
"According to its blurb, this is a comprehensive book, a formal analysis of the foundations of [collective] decisionmaking, and an examination of the many techniques of voting and th different outcomes in different situations. This completely misdescribes Gordon Tullocks informal, idiosyncratic, and thought-provoking book. As I read it, Tullock is presenting a critique of the directions that the sub-discipline of public choice has taken since his and James Buchanan path breaking book, The Calculus of Consent (University of Michigan Press, 1962)
Tullock trades heavily on his status as a founding father of public choice. He writes in an unbuttoned style, mixing theoretical analysis and snippets of fact with anecdotes, personal opinions, reminiscences, minor grievances against fellow scholars, and gratuitously abrasive remarks (of which my favorite is his way of pointing the reader to the literature on voting methods: ⌠If the reader is still curious, I should warn him that research in the field is both tedious and likely to be erroneous [p. 183]. He is unapologetic about the absence of references to other peoples work. And he is candid about the fact that he is not presenting the products of new theoretical or empirical research, but merely suggesting topics on which other people might work. This is the kind of book that is read because of who the author is. But, given the importance of this particular authors lifetime contribution to public choice, and the sheer brilliance of some of his early work (most notably on rent seeking), that is not such a bad reason for being interested in what he has to say about the state of his subdiscipline.
Tullock sees public choice as a subject in which there was a burst of interest from the 1950s to the 1970s, but which has now died out (p39). The cause of death was the set of unremitting negative conclusions that issued from the analysis of Condorcet and Arrow paradoxes. This conception of democracy as inherently inconsistent is unpalatable to most scholars, and does not square with the observation that democratic decision making generally does not result in choice cycles. Tullock suggests that the gap between theory and observation can be closed if we recognize that there is much more to democractic decision making than majority voting.
To begin with, there are rules of order, and various frictions that restrict the number of votes that are taken in assemblies and committees. Tullock hints at a model of collective decision making in which there is a constraint on the rate at which proposals are voted on. Proposals can be made at any time, but are stacked in a queue. When a proposal reaches the front of the queue, there is a simple vote: for or against. In this kind of model, Condorcet cycles do not fatally disable collective decision making. Interestingly, too, this model is consistent with less-than-majority voting rules (for example: a proposal passes if it receives at least 40 percent of the votes). This is significant for Tullock, who is keen to defend the claim famously made in The Calculus of Consent, that simple majority voting is just one point on a spectrum of possible voting rules.
For Tullock, the central mechanism of public choice is not voting according to preferences, but logrolling Ѓг the trading of votes on one issue for votes on another. Logrolling, he claims, is dominant in all democratic societies (p. 139), although it takes different forms in different political systems. Where party discipline is weak, as in the U.S., there are straightforward exchanges of votes between elected representatives and (a useful term I had not heard before) the mechanism of the soup kitchen through which various apparently irrelevant small items are added to a bill so as to attract the votes of wavering representatives. Where party discipline is strong and the electoral system produces one-party governments, as in the U.K., logrolling takes place within the governing party. Where party discipline is combined with proportional representation, as in Israel, there is logrolling when coalitions are formed. Constraints on the rate at which votes can be taken induce a further kind of logrolling, when the proposers of a bill or referendum proposition negotiate about just what that bill or proposition should contain.
This leads to a picture of politics that is much more economic than the picture presented in more public choice theory. What public choice theory took from economics was the formal apparatus of rational choice, and what it looked for Ѓг and failed to find in politics was the collective analogue of the standard assumptions of individual rationality. Tullock urges us to see politics as a particular kind of market. In evaluating political mechanisms, we should look not for logical coherence but for efficiency. Like other markets, the political market leads to efficiency within the constraints of the system (p. 139). This gives us a standpoint from which we can consider proposals for improving the efficiency of the system.
Unfortunately, the style in which Tullock presents his argument is unlikely to persuade public choice scholars to adopt his agenda. The informality, disjointedness, and occasional repetitiveness of the text not to mention unusually poor copy editing make it easy to overlook the significance of what is being said. But public choice theorists might to well to reflect on what has been lost by not following up the central insight of The Calculus of Consent: that politics is a market."