Incentive Compatible Planning and Budgeting of "Distributive" Federal Programs
Edward H. Clarke
Commerce and Lands Branch,
U. S. Office of Management and Budget,
Office of Information and Regulatory Affairs*
* These views should not be regarded as representing those of the Office of Management and Budget or any other Federal entity. An earlier version of this paper was presented at the November, 1996 Annual Meetings of the Southern Economics Association.
A planning/budgeting mechanism has recently been described by Bailey (1996, 1997) that attempts to mimic the competitive market through appropriately designed voting rules by competitive citizen committees. This paper explores variants of Bailey's approach in the context of the design of grants-in-aid that motivate subnational entities (state and local governments) to more efficiently allocate grant-in-aid resources, accounting also for interjurisdictional spillovers where actions in one jurisdiction positively or adversely affect other jurisdictions. Specific applications are illustrated in the context of grants and revolving loans for environmental management and transportation in ways that illustrate the utility of incentive compatible design for planning (agenda setting) and subsequent budgeting (resource allocation) decisions.
- On Intergovernmental Coordination
- Application of the General Progress Program to Particular "Distributive" Programs: Transportation and the Environment.
The paper draws on examples of potential applications of the demand revealing process. I began this project, aimed at applying the idea in public decisionmaking, about 20 years ago when the idea (demand revealing) initially addressed to the "free rider" problem of public goods provisioning (Clarke, 1971) was billed as "a new and superior process for making collective choices" (Tideman and Tullock, 1976). One would have thought that Tideman and Tullock had found a philosopher's stone in that the limitations of demand revealing were, in their words, that it will not "cure cancer, stop the tides, or solve many other problems." The procedure, of course, subsequently elicited a host of criticisms, particularly concerning its practicality.
In responding to criticisms of the idea, particularly those concerned with practicality, I have, in subsequent years, worked on methods of applying it in current institutions -- for example, internal budgeting for information technology within the Federal government or even the "governance" of the Nation's airports and air traffic control system. In several essays that form the basis of this paper, I extend the idea to intergovernmental arrangements, both within a Nation-state and among Nation-states, so as to achieve "subsidiarity" through "incentive-compatible" means. Based on recent work by Bailey (1997), I also introduce some new ways of combining this mechanism, often called the "pivotal mechanism," with other tools for encouraging "strong democracy" along the lines of my original attempts at application (Clarke, 1977, 1980).
The 1977 paper acknowledged the need to, in Buchanan's words, "place the method in an appropriate constitution that will limit and define the range of applicability" (Clarke, 1977, p. 39), although there was a tendency to defend rather wide-ranging applicability on the pure utilitarian grounds of "cost-avoidance" relative to current institutions.
A current debate centers on whether the method is appropriate for the "constitution of a future country" (Bailey, 1997) or whether it could fit comfortably within the institutions of a present country, a debate that is implicit at least in this paper. In any case, the debate promises to be a lively one -- centering on the place of mechanism design in the design of political institutions generally.
The issue of practicality reaches to the conflict between the modern public choice variant of classical political economy and those who approach this topic and public policy generally from more traditional perspectives. The conflict, though clearly favoring the more traditional perspectives, is clearly presented in Anderson (1990). Even if one adopts, as I do, much of the pragmatic liberal perspective concerning matters of institutional design, there are others who would cast a particularly critical eye on the entire enterprise as being fundamentally at odds with legal institutions and the entire political science driving these institutions.
To give an example, I introduce Edward Rubin's recent assessment (Rubin 1993) of the practical impact of public choice on modern legal theory and practice. The review casts a particularly critical eye on the behavioral assumptions underlying the public choice vision, which rejects "the romantic notion often proposed by civic republicans that both voters and legislators are, or can be motivated, by public spirit rather than self interest, and that they can effectuate their desires through rational discourse rather than strategic, self-maximizing behavior."
Professor Rubin then notes both an optimistic and pessimistic strand in public choice, including a principal theme (in the optimistic strand) that is the "development of mechanisms for resolving collective choice problems," (referring among other examples to my original 1971 article on demand revealing, "Multipart Pricing of Public Goods" (Clarke, 1971).
Later, in assessing the strengths and weaknesses of public choice, he notes (Farber and Frickley's assessment, in Law and Public Choice, of) the strengths "which lie in the implications for pragmatic, incremental solutions to contemporary problems of governance; its weaknesses in its empirical inaccuracies and in the impracticality of its other implications for these same contemporary problems."
Specifically, at footnote 44: "The impracticality of public choice recommendations should not be underestimated. Consider, for example, Edward Clarke's idea that a taxation scheme could be employed to reveal voter demand for public goods," (Clarke, 1977). "To induce honest declaration of expected gains, he suggests that the person who casts the deciding vote in favor of a given policy should pay a tax equal to the net gain of other voters that would have resulted if the deciding voter had not voted." (In the same footnote, Rubin adds an example of "an equally creative approach" discussed in Chapter 19 of the Calculus of Consent, which the authors, Buchanan and Tullock themselves described as a farfetched but novel way of dealing with special interest legislation).
This pessimistic assessment of Rubin on the optimistic strand of solutions for contemporary problems is echoed, even very recently and in part, by one of the fathers of public choice, Gordon Tullock (a strong advocate of demand revealing). In the New Palgrave (p. 1044), for instance, Tullock notes that "at the concrete level, those who study Public Choice have been able to provide more in the way of suggestions for reform within the bureaucratic structure than in the higher level parts of democracy where the voters control the legislature, and the legislature and the executive control the bureaucracy. Nevertheless, there are suggestions for improving the whole structure of government and, with time, it is hoped, there will be both more ways of making improvements and better scientific evidence that the 'improvements' are indeed improvements."
The shaping of such improvements among the executive, the legislature and the bureaucracy (including the intergovernmental arena) is the principal subject of the several essays, accompanied by critical thoughts of my own concerning the problems motivating the recent criticisms of public choice (and rational choice, more generally). I found myself trying to assimilate much of this criticism (see, for example, Rubin, 1991) while bringing my own critical theory perspective to the practice of public choice, particularly in the realm of institutional design and the management of public regulation.
I should note that, in private communications, Professor Rubin has modified his observations on the perceived lack of practicality as follows: one might distinguish between an arrangement which is "(1.) capable of being put into practice or (2.) capable of being put into practice by real political actors in the foreseeable future. ... I see demand revelation as impractical in the second sense... Administrative agencies may use them to resolve particular problems, but with respect to the general political science, it seems like a visionary rather than a practical approach."
In another communication, Tullock has largely concurred with Rubin: "I think Rubin's remarks about the impossibility of putting it in with present day personnel is correct. However, if you can get people in the habit of using something similar on other programs, it may be possible to gradually expand it over a larger and larger area. Perhaps in 200 years, we will have it as our basic constitutional method."
Following the work of Mueller (1996) and Bailey (1997), my attempts at practical applications have focused increasingly on constitutional design within existing institutional arrangements. Admittedly, there is an optimistic sense underlying this work. As Mueller (1996, page 47) states: "the notion that individuals have the capacity to design a set of political institutions that advances their general welfare, embodying them in a constitution, agree to it and abide by it expresses a good degree of optimism about human intelligence and capacity for self governance." The skeptics also, at many junctures, express strong reservations about voting rules that communicate the intensities of preferences (i. e. point voting, demand revealing), noting that while they may sometimes immune from individual strategizing, they are subject to coalitions (Riker, 1979). The real concerns of the skeptics, however, are, I believe more philosophical (perhaps even political) than technical.
This paper builds on Bailey's recent attempts to mimic the competitive market through appropriately designed voting rules in making decisions about the supply of public goods and the control of externalities. See also Clarke, 1980 for an earlier attempt to achieve this objective. Bailey combines the use of the "pivotal mechanism" with strong incentives for putting positive rather than zero sum games on the agenda. The process is driven by competitive citizen committees which might strike one as some kind of Utopian fiction of sorts, resembling Bentham's own Utopian-fictional "public opinion tribunal" (POT) which formed the heart of his "constitutional codes" (Rosen, 1983). In this work, the agenda setting of the POTs, or citizen advisory committees (weak legislatures), are driven by the pivotal or Vickrey-Clarke-Groves (VCG) mechanisms, and are regulated by a Commission responsible for final allocations of Lindahl tax/transfers.
Both the committees and the Commission are rewarded by the success with which their allocations approximate the rule of unanimity in the public economy. The rewards (to the committees) are based on their contributions to social welfare as measured by departures from the status quo and the next best preferred alternatives (the proposals of a second competitive legislature). The final selection among competitive budgetary and regulatory agendas is made by use of the principles of a Thompson referendum (Bailey 1996, Thompson 1966) which also motivates accuracy in the provision of willingness to pay information by citizens or their direct representatives. The use of the pivotal mechanism with the Thompson referendum, described here by way of utilizing Congressional districts as the unit of account, exploits the advantages of each of these basic truth telling mechanisms while minimizing their comparative disadvantages (Bailey, 1996).
In this paper, drawing on several examples from the essays, I explore the chances for practical application of this conceptual approach to incentive compatible design. The examples illustrate an approach to the design of grants-in-aid that motivates subnational entities (state and local governments) to more efficiently allocate grant-in-aid resources. The proposal effectively separates the allocation from the distribution in that an initial entitlement level of spending may be greater or less than the amount that the jurisdiction will actually spend. The paper also illustrates an incentive-compatible approach towards accounting for interjurisdictional spillovers where actions in one jurisdiction benefit or adversely affect another. Specific applications of the approach are illustrated in the context of grants-in-aid for environmental management (construction grants and Superfund) and transportation.
In this and the related contexts elaborated above, I attempt to show how further experimentation with these "truth telling" concepts in limited administrative (budgetary and regulatory settings) could truly promote democracy on a human scale, promoting more meaningful involvement by citizens in the social art -- the science and practice of government.