THREE MODELS OF GOVERNMENT Notes for an article by Frank Forman Buchanan has been tirelessly propounding the distinction between the "truth model" and the "exchange model" of government. The first is held by someone who believes that he is in possession of the truth about what the government ought to do (government "for the people"), while the second is held by those, like Buchanan, who take the preferences and values of citizens as the starting point and try to mold a government as a result of a constitutional omni-lateral exchange, corresponding to bilateral exchange in the market (government "by the people"). I shall not try to improve upon Buchanan's many elegant statements of this dichotomy and, indeed, shall presume they are well- known to the readers here. However, Public Choice theory has produced two different varieties of non-truth models to explain the governments we see. The first is the constitutional framework, ideally where there is unanimity at the constitutional stage. The ideal constitution would then provide legislative rules to balance the costs imposed on those who object to a given piece of legislation and the costs of reaching an agreement. Thus, Thecalculusofconsent, a single word by now. The U.S. Constitution of 1789 may very well have approximated unanimity, and the Confederate Constitution of 1861 even more so, but by today the cost of forming rent- seeking coalitions has so far fallen that the constitution needs revision in the direction of requiring substantial supra-majorities. And this is familiar also. But there is another model beloved of Public Choice scholars, and this is that the government serves only to serve itself, usually to maximize its revenues, and which cannot be construed to be either government for the people or government by the people. According to this model, public goods are not the aim of government but rather its incidental byproducts. The great John C. Calhoun beat the scholars in saying that the basic division is between the taxpayers and the taxeaters (not his term). Both models indeed explain much of the world around us. These two views have been termed those of consensus vs. conflict by Joseph A. Tainter in his remarkable book, The Collapse of Complex Societies (Cambridge UP, 1988), which all should read for its panoramic presentation of theories of government alone. Buchanan takes a consensus view, as in his Calculus of, while Marx takes a conflict view. But where is the truth model, the one pre-Public Choice political scientists believed in and which is the most common view held generally and by socialists and libertarians alike (Buchanan said they were equally bankrupt in The Limits of Liberty)? There is a trichotomy here, one that I finally found elucidated in James Dale Davidson and Lord William Rees-Mogg, The Sovereign Individual: How to Survive and Thrive during the Collapse of the Welfare State (NY: Simon & Schuster, 1997). It's an exciting and upbeat book, until you realize that the authors conflate their analysis of 500-year cycles of change driven by the technology of governments with their prophecies of the immediate future at most twenty years away. (Recall that John Perry Barlow said that cybercash would doom central governments within six months. We tend to overestimate short-run change and underestimate long- run change.) This conflation dampens the considerable excitement generated by the book. Davidson and Rees-Mogg speak of governments controlled by a king, by its employees, or by its consumers. Clearly, they say, government is nowadays controlled by its employees. Under capitalism (that economic system controlled by consumers, not as Marx would have it by capitalists), a telephone call costs the same whether done by a rich man or a poor man. But governments operate to squeeze the maximum out of its so-called customers and charge and get away with charging far more for a stock market transaction by a rich man. An excellent point, an exciting point, since the information revolution is going to do away with the capacity of governments run by their employees, though 500 years is a bit of a wait. But where is the government by a proprietor? Only mentioned in passing, I'm afraid, but it corresponds perfectly to Buchanan's truth model, according to which the economist's job is to find this truth and convey it to the king, a hopefully benevolent despot. The world we see is, as always, a mix of things, and all three models describe political realities. Everyone thinks that there is some truth about politics, for everyone want to keep at least some things beyond the reach of politics. Everyone thinks also that some things are always within the reach of politics, surely to provide domestic tranquillity. All governments, too, serve their own bureaucrats, and all of them do manage to rest on some modicum of consensus. It would be hard to find an example of any government that perfectly represents a single pure type. If so, then a little caution is in order before we rush in an proclaim a single model to be the way things ought to be. Government by king, or the truth model, is objectionable both ontologically (there is no truth out there, or at least none that Buchanan can see) and epistemologically (no committee of experts can find it). Government ruled by its employees is objectionable, not just to the exploited but to the exploiters, who would ultimately gain more satisfaction in life if they were to earn their incomes honestly. Government ruled by its customers is objectionable, on Mencken's famous characterization of democracy as "the theory that the common people know what they wants, and deserve to get it good and hard." There are good things about governments under each model also. A benevolent despot can do good, at least if his activities are contained, and there are certain truths about politics that stem from human nature. Elites, to take up the conflict model, have been beneficial, again within limits. And the public provision of public goods, as seen by the public, is a good thing to have. Alas, while writing this, I am reminded of a much earlier formulation, by Aristotle. He spoke of three kinds of polities, control by the one, by the few, and by the many. He spoke of good forms and degenerate forms of each. I would say our current governments are a mix of degenerate types. But consensus remains pretty high.