Bureaucracy, The Supply Side, and Empirical Public Choice

Demand analysis—that is, the interconnections between voting and the de­mand for public goods—has taken center stage in the contemporary public-choice literature. The primary focus on this issue implies that goods and ser­vices demanded in the public sector are automatically supplied. Public-goods supply, however, takes place through government bureaucracies, and the in­centive mechanisms of "bureaus" have, with very few exceptions, not been the subject of much inquiry in public choice. Two exceptions have been the work of the Austrian economist Ludwig von Mises (Bureaucracy, 1944) and the more recent study by Gordon Tullock, The Politics of Bureaucracy (1965). These books, especially the latter, originated the attempt to model the process of bureaucratic output and most particularly the motivations through which "public-sector supply" takes place.

How do bureaucrats behave? What are their motivations? Is there a dis­cernible quantity that they optimize in their public-good-supplying operations? The works mentioned above, especially through Tullock's intellectual influ­ence, resulted in a singularly interesting recent contribution in this area. In 1971, William A. Niskanen, Jr., published his Bureaucracy and Representative Government, which, in the author's words, "focuses on the relations between a bureau and its environment, particularly the environment of representative government, and develops the consequences of these relations for the bureau's budget and output" (Bureaucracy and Representative Government, p. 9).

Niskanen views the bureaucrat as an "endogenous" maximizer in the sys­tem, not unlike the entrepreneurial suppliers of private goods in the economic system. But one crucial difference stands out. While private entrepreneurs can maximize profits, government bureaucrats cannot, at least not legally. Though illegal side payments are not unknown in the political arena, it is far more rea­sonable to point to such variables as income, prestige, the size of the bureau, the bureau's budget, job promises after retirement, and so on as candidates for the bureaucratic maximand. Niskanen assumes that bureaucrats are budget maximizers, and he models government bureaus as individual budget-maximizing units. Budget maximization enables the individual bureaucrat to increase his or her salary, have an easier (or more "pleasant") working envi­ronment, or both.

Bureaus, in this scenario, are "nonprofit organizations which are financed... by a periodic appropriation or grant" (Bureaucracy and Represen­tative Government, p. 15). In essence, a total budget is transformed into a level of total output, since marginal adjustments are not feasible within the bu­reaucratic context. One of the (many) implications of the model is that in their attempt to maximize budget size (and thus the size of the bureau), suppliers will "eat up" the consumers' surplus that results from public-goods supply. The sheer growth of bureaucracy is also an obvious implication of this theory. There have been difficulties, moreover, in integrating the theories of public-goods demand and Niskanen's notion of supply into a "general-equilibrium model." Niskanen's model has stimulated a good deal of research into the "supply problem," however, and it has become an ongoing research concern in the economics of public choice.

Positive public choice has yielded a large number of testable implications and extensions. Economists, especially since 1970, have been hard at work ex­panding and empirically estimating some of these propositions. A very large literature, some of which might be called "empirical public choice" has grown up in the field. The list of contributions delving into these matters is long and pertains to such issues as (1) what the economics of campaign contributions are and how they affect political competition, (2) how self-interest leads to length of political terms in office and to the rules of succession, (3) how the independent judiciary affects cartel behavior, (4) how entry barriers into poli­tics are determined by economic variables, (5) how and why coalitions are formed within legislatures, and (6) why state and federal legislatures contain more lawyers as representatives than any other occupation. A whole branch of literature has developed on the "political business cycle," that is, how self-interested politicians acting under reelection constraints can cause swings in inflation, income, and employment.