In defense of excessive government.

Lee, Dwight R.

1. Introduction

I don't have much practice defending government, and I certainly would not have written this article if it were going to be subjected to a rigorous review process. But because it will not be subjected to any review process, rigorous or otherwise, I have the opportunity to try something different and irritate almost everyone, and that is not an opportunity to be passed up.

In mounting a defense of excessive government, I risk annoying two groups of people with very different political perspectives. The first group consists of those who think the term "excessive government" is an oxymoron, or if they do recognize the possibility of excessive government, they certainly do not think it describes any democratic government. The second group consists of those with whom I have long identified, classic liberals working in the field of public choice who advocate if not a minimal government, certainly one that is limited far more than in any existing democracy. In writing this article, I kept anticipating the objections of the latter group and trying to guard myself against them. In this effort I have attempted to use arguments drawn from, and consistent with, the public choice perspective to make my case.

Before diving into the details of my defense of excessive government, I will outline briefly the main points. I recognize that we rely on government to do some pretty heavy lifting in the form of collective action that, although socially desirable, cannot be motivated by market incentives. Given the serious obstacles to efficient collective action, it should not be surprising that this action is typically performed poorly by government. Of course, the difficulty in carrying out collective action and the deficiencies in the political process are commonly used to make a strong case for restricting government to a very few functions that it can perform cost-effectively. Unfortunately, the very problems that make it desirable to impose effective constraints on government make it extremely difficult to do so.

The flaws inherent in political decision making force a crucial trade-off that philosophical discussions about the ideal role for government typically overlook, the trade-off between the desirability of preventing government from doing things it does poorly and the desirability of securing advantages that only government action can provide. Precisely because government is not very responsive to the general interest, we lack the control over it to get rid of all the inefficient government programs while keeping the efficient ones. "Washington Monument syndrome" is a political reality that public choice economists delight in pointing to,(1) but the implication of that reality is that to get government to provide things we want, like the Washington Monument, we will have to put up with excessive spending on other government activities that cost more than they are worth. Recognizing the full implications of having to accept the bad with the good in government leads to at least a cautious appreciation for the political influence of organized interest groups. Public choice economists correctly see their influence as a major source of excessive government spending, but following the public choice explanation for this influence to its logical conclusion suggests that it is also the source of important benefits. Also, public choice economists may be mistaken in seeing some government spending as excessive because they overlook the fundamental distinction between marginal and inframarginal effects. Finally, I argue that a minimal state may be either desirable or possible, but not both. If it is desirable, it is not possible, and if it is possible, it is not desirable.

2. Compared to What?

"Compared to what?" is a completely natural question to those who instinctively think in terms of scarcity and opportunity costs. This explains why economists tend to miss the humor in the following joke: When asked, "How's your wife?," the fellow responded, "Compared to whom?" To most economists, that seems like a perfectly reasonable response, so where's the humor? Nothing can be usefully evaluated without reference to relevant alternatives. Although the joke may suggest a certain callousness, it actually reflects a forgiving perspective, one based on an understanding that the world is an imperfect place, and pointing out that something is flawed is not a meaningful criticism of it unless better alternatives are possible.

Much criticism of the market, for example, is completely misplaced because it ignores this elementary point. The first thing that many people fail to recognize when criticizing the market is the difficulty of the task assigned to it. Coordinating the decisions and harmonizing the interests of millions of individuals, who know little and care less about the concerns and circumstances of one another, is extraordinarily difficult. I should note that a primary reason this difficult task goes unappreciated is that the market performs it so remarkably well that the resulting benefits seem part of the natural order of things. Karen Vaughn (1996, p. 839) mentioned the problem in her 1995 presidential address: "Given the daunting nature of the economic problem that humans face, that markets should ever succeed is more of a puzzle than that they might from time to time fall short of our expectations." Not realizing the difficulty of what the market accomplishes makes it easy for those who criticize inevitable market failures to ignore the importance of realistically examining the alternatives. Fortunately, this failure to ask "compared to what?" has not gone unanswered.

Austrian economists have earned their keep by persistently pointing out that the perfectly competitive model is not a realistic alternative to the warts and blemishes of the real-world marketplace. Indeed, it is not even a particularly desirable alternative. As Robert Bork (1978, p. 92) - admittedly no Austrian - has stated, with less exaggeration than it may seem, "A determined attempt to remake the American economy into a replica of the textbook model of competition would have roughly the same effect on national wealth as several dozen strategically placed nuclear explosions."

The response to the failure of market critics to ask "compared to what?" that I am most familiar with is the development of public choice. Buchanan and Tullock (1962) revitalized and redirected the study of political economy by subjecting political alternatives to the same analytical scrutiny that economists had long applied to the market. One measure of the success of public choice is that most professional economists would now be embarrassed to call for corrective government policies solely because the market is not working with textbook perfection.

But, if criticism of the market process should be tempered by comparisons with realistic alternatives, so should criticism of the political process. Just as with the market, we rely on government to perform an extremely difficult task. Indeed, its task is at least as difficult, and probably more so, than that assigned to the market. Unlike the market, in which most exchanges involving agreements between two parties with only moderate, if any, spillover affects others, politics involves exchanges that directly affect everyone in the relevant polity and requires reaching at least de facto agreement from them all.(2) According to Buchanan (1967, p. 4), "Political decision-making is a complex and intricate process, much more complicated than is the nonpolitical decision-making in market institutions" (emphasis added).

In other words, I can defend political failure on the same grounds that we defend market failure. Let me do so by paraphrasing Vaughn's earlier comment: Given the daunting nature of the political problem that humans face, that government performs as well as it does is more of a puzzle than that it is excessive when measured against textbook standards of efficiency. Or, citing Buchanan (1988, p. 12) once more, "[my] argument might be taken as a criticism of the naivete of both the market-failure welfare economists and the market-works-politics-fails stance of many modern public choice new neo-classical economists. By comparison with idealized standards, both markets and politics fall" (emphasis added).

3. The "Invisible Hand" Behind Rent-Seeking Waste

Private Interest Dominates the Political Process

The theory of rent seeking, and the resource dissipation that results, is one of Gordon Tullock's most important contributions to our understanding of political reality. The coercive power of government attracts organized interests intent on capturing private profit at public expense by influencing political decisions. At worst, political rent seeking destroys value; at best, it reduces the value that government programs could otherwise create, as with unnecessary military bases or excessively costly procurement of military hardware. Whereas private-sector profit seeking creates value, political rent seeking creates a larger, less efficient government, and the term "rent seekers" is commonly used as a pejorative. Clearly, any attempt to defend excessive government is incomplete if it does not come to the defense of rent seeking and rent seekers.

Let me first take on the easier of these two by defending rent seekers. One of the most powerful insights of economics is that we should not depend on self-sacrifice for the realization of desirable outcomes. Conversely, we should not blame self-interested people for undesirable outcomes. Most people, most of the time, will put their interests ahead of the interests of others, and little can be done to alter this fundamental fact of life. Thus, the only hope for realizing a persistent pattern of desirable outcomes is through institutional arrangements that, as if by an invisible hand, align individual interests with those of the general community. When outcomes are perverse because of a misalignment of interests, economists advocate improving social institutions, not complaining about deficiencies in human nature. Indeed, the great insight of public choice is that reliance on an invisible hand to guide self-interest is just as important in achieving political outcomes as in achieving market outcomes, and if private interest is just as dominant a motivator in politics as in markets, we should be neither surprised nor particularly offended by political rent seekers.(3)

This suggests that any defense of rent seeking must also be a defense of the political arrangements within which it occurs. I argue that the invisible hand created by the prevailing democratic institutions in the United States, although not as impressive as the invisible hand of the marketplace, leads rent seekers to behave in generally beneficial ways. This defense should not be taken as an endorsement of the political status quo, and I comment on the possibilities for improvements at the end of this article. I make no claim that political outcomes are efficient according to the standard textbook criteria, and I see little agreement between my argument and Wittman's (1995) claim (from the subtitle of his book) that "democratic institutions are efficient." I acknowledge the excesses and waste of the political process, which seems to be necessary if I am going to defend them.

The Problem of Providing Public Goods

I begin my defense of excessive government by considering the problem of providing public goods, a problem we leave to government because market incentives leave these goods greatly underprovided, if provided at all. Unfortunately, shifting this task to government does not diminish the fundamental problem: motivating people to voluntarily contribute to the provision of public goods. Everyone sees an opportunity to "free ride" on the contribution of others, but when people yield to this temptation in significant numbers, the good cannot be provided, and there is no ride for anyone. Of course, a primary justification for the coercive power of government is that it allows us to collectively overcome the free-rider problem in providing public goods.

Unfortunately, the demands for public goods are difficult to transmit through the political process for much the same free-rider reasons as in the market. When considering the political provision of public goods, people often overlook this because they see the political process as possessing an important advantage over markets in communicating preferences: the ability of voting to overcome the free-riding distortions in market communication. Consider an idealized version of this advantage. Citizens express their preferences for public goods through the political process by voting, either directly in referenda or indirectly by choosing political representatives. Unlike a decision to contribute to a public good in the marketplace, a decision to contribute to a public good in the voting booth imposes no cost on the decision maker unless the election obligates everyone to contribute. So, preferences are communicated candidly, because the voter who favors a public good has no fear of being exploited by free riders by being honest and no hope to free ride on the sacrifices of others by being dishonest. Politicians, anxious to please voters, respond to the expressed preferences of the electorate by providing the desired public goods as efficiently as possible.

Although simplistic, this version of the political process contains some truth. After all, without government action, we would find ourselves with less national defense, less crime prevention, and less environmental quality than we have and certainly less than people desire at the cost of providing them. But this discussion of voting is also misleading. Although voting does overcome some free-rider obstacles to accurate expression of preferences, it creates others. Because one vote is unlikely to be decisive, voters have little motivation to become well informed on the issues. This "rational ignorance" is a free-rider problem, with each voter hoping to benefit from other voters who make more effort to become informed.(4) Even if fully informed, voters will often yield to the free-rider temptation of "expressive voting," voting to satisfy expressive preferences rather than voting for outcomes they actually prefer.(5) Also, informed and honest voting is not enough for government to respond efficiently to the public's desire for a public good. If a cost-effective proposal for a public good is put on the ballot with costs distributed over voters in proportion to benefits and is presented clearly to voters receptive to such information, it will surely receive an overwhelmingly, and likely unanimously, favorable vote. But people must go to the trouble of formulating and comparing alternative public good proposals, getting the best of the proposals on the ballot, and then, after the vote, performing the monitoring and oversight necessary to ensure that the public good is supplied at the desired level and at least cost. Who will go to all this trouble? Not likely consumers, who can free ride on the contributions of others. Public provision of public goods is plagued with the same type of free-rider problems as private provision.

A counterargument might be that elected representatives are anxious to put forth public good proposals that mobilize constituent support. I agree, but this is not the same as saying that they will put forth proposals that best serve the interests of the general citizen. The biggest payoff to politicians comes from providing programs that concentrate benefits on politically active groups organized around narrow interests. The public's general desire for public goods justifies granting government the power to tax and regulate to provide them, but the private interests of organized groups determine the details of how those public goods are provided. These organized groups, and their political agents, typically benefit the most by reducing the efficiency of public good provision.(6) Unnecessary expenses in providing public goods are an extra cost to society but are the source of rents to public-sector bureaucrats and private-sector contractors who supply the goods. Because the social cost of the reduced efficiency is spread over the general public, any effort to mobilize opposition to excessive cost is frustrated by the free-rider problem.

This sounds like a pretty standard public choice criticism of excessive government and special-interest waste. Where's the defense of excessive government that I promised? My defense is based on the fact that inefficient public good provision results from the lack of consumer involvement in the details of public goods supply, a lack that exists in the private market and that political provision does little to overcome. If consumer demands for public goods could be transmitted to government suppliers with the same clarity and incentive to respond as the demands for private goods are transmitted to market suppliers, criticism of government waste in providing those goods would be justified. But, because transmitting demands for public goods to government suppliers is so difficult, not only is much of this waste understandable, it can also serve a useful purpose.

The reality is that suppliers are the only groups that can be motivated to exert themselves on behalf of making public goods available because of their concentrated interest, and they will do so only when it is possible for them to take advantage of the coercive power of government to ensure payment for their efforts. This obviously is not the way things would be in a perfect world, but it is the way things are in the real world, and attempts to reduce the inefficiency that comes inevitably with supplier control over providing public goods can easily result in greater inefficiency by reducing the amount of public goods provided.

The Case of the Military-Industrial Complex

Consider the case of national defense and the rent-seeking influence of the military-industrial complex (MIC). There can be no doubt that the rent seeking of the contractors, military brass, politicians representing districts containing military bases, and others making up the MIC has been the source of wasteful government spending. A common reaction is to see the MIC as a malignancy in our democracy that exploits the public for private gain and weakens our national security in the process. But the public choice insight into the power of special-interest groups suggests to me a more sympathetic view of the MIC. Consider the consequences for national defense in a democracy that eliminates the rent-seeking influence of the MIC.

Politicians would continue to provide their constituents that mix of public goods and services that motivates the greatest political support. In deciding on this mix, each congressional representative would recognize that his or her constituents value national defense. But each would also know that his or her constituents could benefit from the national defense that results from the contribution of other constituencies. Residents of South Carolina receive as much protection from a submarine patrolling the oceans as do residents of California no matter whose taxes pay for it. Therefore, representatives from all districts would be strongly motivated to divert military spending into programs that concentrate benefits in their district. If the MIC were unable to convert some of the general benefits from national defense into concentrated special-interest benefits, the political process would leave the military budget defenseless against the demands for government spending that do concentrate benefits. The case for the MIC, then, is based on the real possibility that it is better to have an adequate supply of national defense at excessive cost than an inadequate supply at least cost.

The Case of the Environmental-Industrial Complex

The case for rent-seeking waste in the military can be extended to the inefficiencies found in environmental policies. Economists have made compelling arguments that current environmental policies make it excessively costly to reduce pollution. The dominant command-and-control approach to pollution reduction has been found in numerous studies to be several times more costly than approaches relying on market incentives,(7) and the blame for the political popularity of the command-and-control approach has been placed on the rent-seeking of groups that capture private benefits from inefficient pollution control.(8)

Again, the natural tendency is to criticize what, in this case, may be called the environmental-industrial complex (EIC), which is composed of private industries and government environmental agencies, for being more concerned with protecting its private interests than protecting the environment.(9) In an ideal world, environmental quality would be provided in response to consumers who communicated their preferences clearly and compellingly to obedient suppliers. But, as with the MIC, we are dealing with a public good, and one can defend the waste in environmental policy as providing the rents necessary to mobilize interest groups to take more political action on behalf of environmental protection than would be taken otherwise. Attempts to increase the efficiency of pollution-control policies by reducing the influence of the EIC can increase inefficiency by reducing the amount of pollution control. In other words, there is a trade-off between reducing pollution at less cost and reducing pollution.

I want to make sure that my point here is not misinterpreted. I am not arguing that without the MIC and the EIC there would be no effective political demand for national defense and environmental quality. Politicians cannot ignore completely the general public's desire for public goods as expressed through voting, which explains why special interests must use the cover of public concerns and actually provide some public goods to collect their rents. Even in this regard, however, the MIC and the EIC may serve a useful function by exaggerating the threats to be protected against.

But without the MIC and the EIC, the public demand for national defense and environmental quality would motivate about the same type of political action as does the public demand for campaign finance reform that does more than just protect incumbent politicians. In general, the public would like to see more competitive congressional elections. But the benefits of better legislation would be diffused over the entire population, while the costs would be concentrated on members of Congress and their constituencies, each of which sees advantage in a representative experienced in pork-barrel politics. So without a way of concentrating some benefits from genuine campaign finance reform on an organized group with rent-seeking ability, we can expect the rhetoric of campaign reform out of Washington but not the substance. Who would be satisfied with the level of national defense or environmental quality if our politicians were no more motivated to fund military and environmental programs than they are to promote competitive elections?

The Case of Crime Control

National defense and environmental protection, although maybe the most obvious, are certainly not the only examples of how important rent seeking is to the political provision of public goods. Consider crime control. It is true that there are more privately hired security guards than publicly hired police officers, but the former create primarily private goods. The public good aspects of crime control depend crucially on government provision (or financing) of courts and prisons. The public wants these components of a criminal justice system, but much of the effective political demand for them comes from organized interest groups. For example, according to a recent issue of The Economist (1996; p. 25),

American crime policy seems to have become an area where the arguments. . . take second place to the lobbying power of special-interest groups. . . One is the prison-building lobby. . . The prison guards' union has also become a powerful voice. According to a study of campaign contributions in California in 199192, the local version, the California Correctional Peace Officers' Association, was the second-largest donor in the state. It spends around $1 million on political contributions for the governorship and the legislature in each election cycle.

No serious person would argue that this rent seeking pushes in the direction of least-cost prison space. But this is not necessarily an indictment of such rent seeking once one recognizes that the private-interest reality of politics forces us to confront trade-offs between the desirability of getting public goods provided and the desirability of reducing the waste and excess in their provision.

Some Criticisms

There are at least three criticisms of my argument at this point. The first is that people do contribute voluntarily to the provision of public goods. I do not deny this. Many people, for example, voluntarily contribute to environmental organizations that do things to protect the environment directly or that take political action to improve environmental quality. Much of this voluntarism does some good. But I think most economists would agree that such voluntary action alone would leave us with woefully inadequate quantities of public goods such as national defense, environmental quality, and crime control. Certainly public choice economists, who emphasize the dominant role of self-interest, are skeptical that much good will come in either the private or public sector from people sacrificing private concerns to promote social objectives. So I don't see this as a serious criticism of the argument that we have to rely on government and the private motivation of organized interest groups to secure adequate (not optimal) quantifies of public goods.

A second criticism, which I take more seriously, is that without government action to supply public goods, private arrangement would emerge to provide those goods. In some cases, this is clearly true. Where the benefits from a public good are easily excludable, such as television programming, private arrangements surely provide them more efficiently than government. Even when excluding people is more difficult, "private" arrangements may still be more efficient than the typical government approach. But I placed the term private in quotation marks, because any arrangement that succeeds in providing a public good to many people, where excludability is a concern, requires some means of making collective decisions and then enforcing those decisions on those who either disagree with them or attempt to be a free rider. In other words, such a private arrangement for providing public goods takes on characteristics of governments. Boudreaux and Holcombe (1989) and Holcombe (1994, chap. 5) develop this point by considering the similarities and differences between private associations and governments. Most relevant to my argument is their emphasis on the importance of the ease of exit, with there being a continuum between private organizations, from which exit is easy, and governments, from which exit is difficult. But ease of exit depends to some degree on the size of the relevant jurisdiction, with it relatively easy to exit a homeowners' association (to use one of Boudreaux and Holcombe's examples) because it is small (the collective consumed benefits do not extend very far). A move a few blocks away is all that is required. If the public good provides nonexcludable benefits over an entire nation, as with national defense, then exit is far more difficult because the size of the relevant jurisdiction is much larger. So, as I read Boudreaux and Holcombe, any arrangement capable of providing public goods generating widespread benefits resembles a government in important ways, with some degree of coercive power being exercised by those with a comparative advantage in taking advantage of that power for private purposes.

Let me now consider a third, and I believe most potent, criticism of my case for relying on special-interest rent seeking to provide public goods. Special-interest groups have undeniably lobbied successfully for government to do a whole host of things that have nothing to do with providing public goods and that should either be left to the private sector or not be done at all. A dominant activity of the federal government is currently the transfer of wealth from the unorganized to the organized. These transfers are invariably justified in the name of achieving some noble public objective, such as saving the family farm, making sure that everyone has access to the best medical care at close to zero cost, saving American jobs against foreign competition, or providing our children with a good education. Quite apart from the fact that family farming, medical care, jobs, or even education are not public goods, government attempts to promote these objectives have been counterproductive. Here I strongly disagree with Wittman (1989, 1995), who tries to defend government by arguing that it takes on only those activities that it handles most efficiently and leaves the rest to the private sector.(10) This opens my argument to the criticism that it applauds the productive consequences of special-interest rent seeking while minimizing the counterproductive consequences. However, this criticism assumes we can get rid of the harmful rent seeking and keep that which is beneficial. That sounds good to me. I want to go on record right now as being in favor of getting rid of the bad and keeping the good. But there is one small problem with this noble sentiment: How do we carry it out?

Even assuming some way to cut government fat while leaving the lean (and this could never be done with surgical precision), we are talking about a tremendous improvement in public policy. As Tullock (1971) has observed, good public policy is a public good. But the only way to get this policy public good provided would be for private-interest groups to ignore their private interests. Good luck. In the public choice world, where private interests motivate political decisions, good programs are a by-product of special-interest pursuits that invariably lead to plenty of bad programs. For example, Tullock (1983) has argued that government transfer programs do little to help the poor partly because the controlling political support for these programs comes from interest groups that trade their support for a host of other transfer programs that benefit the nonpoor.

It is naive to hope that we can somehow compartmentalize political self-interest so that it leads only to programs that promote broad social goals, such as poverty reduction. Haveman (1988), for example, has argued that we could do more to help the poor while reducing the costs of government antipoverty programs by eliminating the transfers to the nonpoor and keeping, and in some cases expanding, those to the poor. This is correct, of course, as far as it goes. The problem is that by ignoring the reality of private-interest politics, it doesn't go very far.

I'm not suggesting here that we can never get rid of bad government programs. Obviously we can. When I first started teaching economics at the University of Colorado in the early 1970s, I loved to point out the perversities generated by the policies of the Civil Aeronautics Board, the Interstate Commerce Commission, and the Selective Service System. I thought I would be able to use these examples for my entire teaching career. But these agencies have been either abolished or their most harmful activities eliminated. We got rid of some bad programs, and I had to revise my lecture notes. Unfortunately, shifts in the influence of competing political coalitions that undermine the support for some lousy government programs increase the support for others. This is not an argument that improving government in general is impossible. Indeed, Richard McKenzie and I (McKenzie and Lee, 1991) have argued that global competition is putting pressure on governments around the world to become more efficient. But the progress that is possible will not be achieved by simply deciding to get rid of the bad programs and keep the good ones. Successful rent seeking for government programs not worth what they cost is a price we will always pay for government programs that are worth more than they cost.

Assuming, as public choice economists do, that self-interest is the dominant motivator in politics as in markets, our only hope for consistently beneficial political outcomes is the invisible hand of political rent seeking. Admittedly, the invisible hand of the political process does not do as well channeling the pursuit of self-interest into socially desirable outcomes as does that of the market, and no doubt improvements can, and I hope will, be made in political institutions that bring self-interest and the public interest into tighter alignment. But, given the difficulty of efficiently making and implementing collective decisions that have to be made in a viable community, especially nation-sized communities, the incentives created by existing democratic institutions, as perverse as they often are, are on balance doing an impressive job of promoting the public interest.

4. The Marginal Versus the Total

Some may think I am being a bit naive by saying that government is doing an impressive job of promoting the public interest. I even find myself cringing a bit when I say this for the same reason other economists are probably doing the same thing. The reason is that economists are trained to think in marginal terms, and I do not believe there can be any doubt that government is destructive at the margin. But I believe most people, public choice economists included, will accept that governments, at least those constrained by constitutional democracy, as imperfect as those constraints are, create positive net value in total. Citizens in the industrial democracies are better off with their existing levels of government than they would be without government. Accepting this proposition is not the same thing as defending government, as any good economist knows. As just indicated, economists are primarily concerned with what takes place at the margin, so even though government may be providing total benefits far in excess of total costs, it can still be criticized on the basis of marginal considerations. Public choice criticisms of the political process are not that it leads to a government that adds less value than it costs in total but that it results in government expansion beyond the point where its marginal value is equal to its marginal cost. The recommendation is not for the elimination of government but for the downsizing of government. But there can be a positive relationship between the total net value that government creates and the degree to which government expands beyond the marginal limits of efficiency, at least as those marginal limits are conventionally considered. By considering a plausible version of this relationship, it is possible to make a case for a government with a marginal value significantly less than its marginal cost, again, as those margins are most likely to be considered.

Consider police protection. The total benefit we receive from police protection is very large, but we all are offended by police brutality. Police brutality is easily seen as a particularly offensive example of excessive government, government activity beyond the point where marginal value is equal to marginal cost. I don't want to be accused of favoring police brutality, but I do think that, as someone who has taken on the task of defending excessive government, I am obligated to say something more useful about it than a predictable, although largely vacuous, expression of outrage. The only way to ensure no police brutality would be to restrict the discretionary authority of the police so much that their ability to protect the public would be extremely limited, if not eliminated. This situation is illustrated in Figure 1 by the downward-sloping curve [MV.sub.P1], which represents the marginal value of police activity when the police have a level of power given by [P.sub.1] (a level sufficiently small that prevents police from engaging in excessive activity). So police activity is expanded only to level [A.sub.1] (where [MV.sub.P1] = [MC.sub.p], which represents the marginal cost of police activity), which appears to be the efficient amount, as we depict efficiency with our standard diagrams.

But if we gave the police more power, we would expect the marginal value curve to shift to the right, to perhaps [MV.sup.P*], as the inframarginal value of police protection increased. At every level of activity up to [A.sub.1], the total value of police protection is greater than before. But because of the additional power, the police can now expand their activity to [A.sup.*], well into the region where marginal value is less than marginal cost. Indeed, as I have constructed it, at [A.sup.*] the marginal value of police activity is negative and includes some police brutality. Police activity now appears excessive. But I have also constructed the diagram so that it is easily seen that the total net value of police protection is now greater than before, when policy activity was at its efficient level ([A.sub.1]).

Obviously, the curves in Figure 1 do not capture all the relevant marginal values. The marginal value of police power is greater than indicated by the marginal value of the additional police activity that power makes possible because of the inframarginal benefits more police power creates. So [P.sup.*] is a more efficient level of police power than [P.sub.1] (and I assume it is the efficient level). Thus, as long as there are the assumed connections (or any positive connections for that matter) between police power, the ability of the police to provide greater protection, and the ability of the police to expand their level of activity, then probably the best we can do is give the police the power to expand their activity to the point of occasional police brutality. The best we can do is have a police force that appears to almost everyone to be excessive.

This defense of police brutality generalizes to other government excesses. Give any agency of the government the power required to provide the maximum value over its ideal range of activity, and you have almost surely given it, and its special-interest clients, the power to expand its activity beyond that desirable range. An MIC that can adequately defend the nation has the power to expand its spending beyond the level required to do so. An Environmental Protection Agency with the power to adequately protect the environment is one that can stymie pollution-control approaches that would reduce the size of its budget. A public university with the power to adequately conduct basic research and disseminate knowledge is one that can conduct a lot of bogus research and disseminate a lot of nonsense, and do so at a high cost. As always, there is a trade-off to consider. To obtain a reasonable level of inframarginal benefits from government activity, government must have enough power to expand its activity beyond the point where, at the margin, that activity is destructive.

Determining the best point for this trade-off between inframarginal benefits and marginal excess is obviously not easy. But it is worth noting that it creates a perceptual bias against excessive government, Public choice economists consistently point out, quite correctly, that there is a bias toward excessive government spending because the benefits from many spending programs concentrate benefits that are easily noticed, much appreciated, and clearly connected to the programs that generate them. On the other hand, spending programs spread costs so thinly over the general public that they are easily ignored and difficult to trace to their source even if they are noticed. But an opposing perceptual bias deserves mention in any discussion of excessive government. This bias is derived from the emphasis that people place on marginal, as opposed to inframarginal, considerations, an extremely useful emphasis, as explained by economic theory. But we commonly ignore the tremendous total value we receive from people and things while grousing about the aggravations they cause at the margin. Spouses, cars, and colleagues come to mind. Similarly, it is easy to see the marginal excesses of government, the police brutality, the costly overruns for military hardware, the overreaching of the Environmental Protection Agency, and the tenured professors who abuse their light teaching schedule. It is much more difficult to see the inframarginal benefits that would not be possible without the operational discretion and political power that generate the marginal excesses. I am not arguing, and do not believe, that this perceptual bias against excessive government is as strong as the one favoring it. But it is a bias that works, at least to some degree, against government growth and that would create the perception that government is excessive even if it were not.

5. Is a Desirable Minimal State Possible?

Let me next make a somewhat different case for what is typically seen as excessive government by asking whether a desirable minimal state is possible.(11) In addressing this question, I argue against the minimal state as either a practical objective or an ideal. My argument depends on two propositions widely accepted by economists. The first, integral to my discussion throughout this article, is that the case for a minimal state comes from the public's inability to prevent organized interest groups from capturing political power and using that power to achieve private benefits at public expense. The second is that the better an agent, the more that agent can desirably do.

Although uncontroversial standing alone, together these two propositions cast serious doubt on the possibility of a desirable minimal state, a conclusion that I suspect is quite controversial. In a nutshell, the conclusion is that if a minimal state is desirable, it is impossible to achieve, and if it were possible to achieve, it would not be desirable.

The minimal state is considered desirable because the public cannot control political decisions to prevent government from being used by special-interest groups to trump the public interest when it expands beyond minimal state limits. But the special-interest influence that makes a large government such a threat to the general public is also the dominant force behind the growth in government. In other words, for the very reason a minimal state would be so desirable, achieving it is politically difficult if not impossible. On the other hand, if somehow the public acquired sufficient control over government to achieve a minimal state, the government would be subject to sufficient control to usefully address problems not within the purview of a minimal state. A government that is a good agent can obviously expand its reach farther with socially beneficial results than a government that is a poor agent. So, the public control over government necessary to make a minimal state possible would render such a state no longer desirable. The minimal state may well be a fantasy that, like most fantasies, owes its appeal to the fact that it is beyond our reach. If we ever achieved it, we would no longer want it.

There are two objections to my argument. It might be argued that it is possible to exert the type of broad control over government necessary to achieve a minimal state but not exert the detailed control required to make government a better agent at solving particular problems. The trouble with this objection is that it presumes that government solutions require detailed involvement, or micromanaging. But this is not the case. Indeed, government can typically accomplish more by establishing incentives that motivate private action than by attempts to solve problems and provide services directly. For example, given the inefficiency associated with the command-and-control approach to pollution control, there are certainly pollution problems that it makes no economic sense for government to address but that could be beneficially addressed by market-incentive approaches. Surely the type and level of control over government necessary to impose and maintain a minimal state would be sufficient to shift environmental policy to far more reliance on market incentives.

A second objection to my minimal state argument involves a rejection of any utilitarian or contractarian considerations. As long as there exists some basis for evaluating the social costs and benefits of government action, whether behind the "veil of ignorance" or otherwise, it makes sense to infer a positive relationship between the efficiency of government as a public agent and the number of desirable government activities. But some who embrace a natural-rights justification of government reject such efficiency-based arguments when it comes to determining the acceptable role of government. From this perspective, the only justification for government is as a means of protecting the natural rights of individuals, and the only acceptable way of providing this protection is one that does not violate anyone's natural rights.(12) Natural-rights theorists will no doubt agree that the lack of public control over government makes achieving a minimal state unlikely but will reject the second half of my argument that if such a state were possible, it would cease to be desirable. To those who refuse to accept a positive connection between public control over government and the amount it is desirable for government to do, I have no response except to say I think they are wrong.

6. Conclusion

At this point, some of you may be wondering what happened to the Dwight Lee whose work has been largely devoted to pointing out and critiquing the excesses of government, not defending them. Let me make clear that the old Dwight Lee is still here. My defense of excessive government is not that the existing level of government excess is efficient or even acceptable. I hope no one takes my remarks as a defense of the status quo or as an argument against trying to rein in the size and scope of government activities. Certainly, I will continue to make the case for reining in government, and I will do so for two reasons. First, I believe that government is larger than is warranted by my defense of excessive government, and second, I think public choice arguments can help restrain, and possibly reverse, government growth, at least marginally,

Public choice is easily interpreted as a message of pessimism. Rational ignorance and apathy characterize the general public; narrowly focused interests always trump the general interest; and attempts at reform are either feeble failures or soon subverted by politics as usual. I think this is a misreading of public choice. I remember sitting around a table at a Liberty Fund conference a number of years ago with Jim Buchanan during a decidedly pessimistic discussion about the possibility of political reform. Suddenly, Jim cleared his throat, and everyone went silent in anticipation. Jim said something to the effect of "If its all hopeless, then just why in the hell are we here?" Good question.

I believe the reason we were there, and the reason for much of the work economists do, is that improvement is possible, and increasing and communicating our understanding of how the political economy works can lead to that improvement. Just communicating some basics of public choice to a wider audience can reduce the influence of politically organized interest groups. For example, being well organized in pursuit of a highly concentrated benefit whose cost will be thinly spread over the unorganized public is seldom sufficient for political success. To be effective, special-interest political activity must masquerade behind some plausible claim of public concern. Consider that Gordon Tullock has long favored a policy of imposing a tax of $1 on every American with the resulting $260 million going directly to him. He actually favors a higher tax, but not being greedy, is willing to settle for $1. His proposal has all the characteristics of a successful special-interest program. Gordon is a small group and has always been well organized. The benefits would be extremely concentrated and the costs so diffused and insignificantly burdensome to any one person that no one would be motivated to oppose the proposal. Despite these advantages, the Tullock proposal has never achieved political liftoff because it lacks any plausible pretense that it would advance some noble public purpose. Gordon is still working on a public-interest rationale, but so far without success.

The point is that if you strip away the public-interest facade cloaking a special-interest proposal, political support for the proposal quickly collapses. Public choice, by creating a coherent framework for understanding the political process in terms of private interests, is helping to penetrate the public-interest rhetoric and to expose the private-interest reality. Obviously the public-interest rhetoric is not easy to penetrate. Clever advocates can almost always make plausible arguments tying their programs to noble-sounding social objectives (the failure of Gordon Tullock to accomplish this notwithstanding). Rationally ignorant voters will find it convenient to accept these arguments, although in case after case, the programs help a few at the expense of many while working against the very objectives they are supposedly promoting. But public choice, and the realistic view of the political process it has given intellectual legitimacy, can have the long-run effect of restraining government by undermining, at least marginally, special-interest political influence.

You are excused at this point for asking yourself whether I am for special-interest political influence. In the first part of this article, I argued that such influence is essential in motivating the provision of important government services, and now I am criticizing that influence. So, which is it, good or bad? The answer is both, and there is less inconsistency in praising and condemning special-interest influence in politics than may first appear.

For all the reasons discussed earlier, private interest and the waste and excess that go with it are not only unavoidable but also necessary to stimulate desirable government actions. The question is, how can we reduce the waste and excess of special-interest politics without also greatly reducing the government's power to do useful things? The answer is to reduce the power of political interest groups across the board rather than to attack the power of such groups individually. Drain the swamp, don't wrestle the alligators one at a time.

For example, if political reform concentrates on reducing the influence of the MIC then, as discussed earlier, the government's ability to provide national defense will be undermined with little, if any, reduction in government waste and excess as military spending gets diverted into other special-interest programs. On the other hand, if political reform reduces the general influence of narrowly focused interest groups, then the reduction in rents from waste and excess in one activity will be accompanied by corresponding reductions in other political activities. In this case, the rents available in important government activities will not be reduced relative to those in less important, or harmful, activities, so meaningful reform will do less to threaten the good activities relative to the bad. In fact, if reform is based on, or accompanied by, a heightened skepticism of special-interest political activity, then interest groups lobbying for genuine public goods will likely have an advantage over those lobbying for what are primarily transfers. If so, we could end up with not only less rent-seeking waste overall but also a more efficient mix of government services.

Of course, under any realistic political reform, there will always be excessive government that someone with nothing better to do can defend. But with luck and public choice scholars attacking government excess rather than defending it, there will soon be less excess to defend.

1 Washington Monument syndrome refers to government agencies responding to the prospect of budget cuts by threatening to cut the services that the public demands the most. For example, the U.S. Park Service claims it will have to shut down the Washington Monument if its budget is cut.

2 It is true that decisions often become politicized and thus require the general agreement of a uniform approach when they could be handled better through private-party exchanges in the marketplace. Mandating that employers provide particular fringe benefits is an example. Political overreaching of this type is easily seen as excessive government and will be discussed later as I move into my defense of such excess.

3 I qualify this statement by saying "particularly offended," because it can be difficult to tolerate the pursuit of self-interest when social institutions do a poor job aligning individual interests with the general interest. Because political incentives do not, and probably cannot, bring individual and collective interests into as tight an alignment as do market incentives, it can be argued that we should be offended by political rent seeking.

4 See Downs (1957, chap. 11) for the first systematic development of the concept of rational ignorance.

5 For example, if I realize expressive satisfaction by voting for a program to help the poor, I will likely vote for it even though I know that if it passes, the cost to me will be far greater than the benefit I will receive. Because the outcome of the election is, with a probability of almost 1, independent of how 1 vote, it costs me almost nothing to vote for the proposal. So, if I receive a modest benefit from expressing myself in favor of the poor, I will vote for the proposal even though I would prefer the proposal be defeated (if my vote were decisive, I would vote against it). I can free ride off the negative vote of others while I get the benefit of feeling good about my generosity. For a detailed discussion of expressive voting and its implications, see Brennan and Lomasky (1993).

6 For a discussion of how organized interest groups benefit from inefficiencies in providing national defense, see Cowen and Lee (1992).

7 For a summary of some of these studies, see Table 15.2 and accompanying discussion in Tietenberg (1992, pp. 402-405).

8 Relevant articles include Buchanan and Tullock (1975), Ackerman and Hassler (1981). Maloney and McCormick (1982), and Parshigian (1984).

9 At least some private environmental groups should also be included in the environmental-industrial complex because they often subordinate their concern for environmental quality to the private interests of their leadership. See Boerner and Kallery (1995).

10 For example, Wittman (1989, p. 1421) argues that "democratic political markets tend toward efficiency. [which] implies that democratic governments will allocate to the economic markets those tasks in which the economic market is most efficient."

11 My discussion here is based on a previous article of mine (Lee 1989). Although there may be some disagreement about exactly what the allowable functions of a minimal state should be, such a state is clearly best described as a "protective state," in the terminology of Buchanan (1967, pp. 68-70) as opposed to a productive state, again in Buchanan's terminology (1975, pp. 97-98).

12 Nozick (1974, chaps. 1-6) argues that a minimal state can arise and operate without violating the natural rights it is responsible for protecting.

References

Ackerman, Bruce A., and William T. Hassler. 1981. Clean air/dirty air, or how the Clean Air Act became a multibillion dollar bailout to high-sulfur coal producers and what should be done about it. New Haven, CT: Yale University Press.

Boerner, Christopher, and Jennifer C. Kallery. 1995. Restructuring environmental big business. Policy Study No. 124. St. Louis, MO: Center for the Study of American Business, Washington University.

Bork, Robert H. 1978. The antitrust paradox: A policy at war with itself. New York: Basic Books.

Boudreaux, Donald J., and Randall G. Holcombe. 1989. Government by contract. Public Finance Quarterly 17:264-80.

Brennan, Geoffrey, and Loren Lomasky. 1993. Democracy and decision: The pure theory of electoral choice. Cambridge, England: Cambridge University Press.

Buchanan, James M. 1967. Public finance in democratic process: Fiscal institutions and individual choice. Chapel Hill, NC: University of North Carolina Press.

Buchanan, James M. 1975. The limits of liberty: Between anarchy and Leviathan. Chicago: University of Chicago Press.

Buchanan, James M. 1988. Market failure and political failure. The Cato Journal 8(1):1-13.

Buchanan, James M., and Gordon Tullock. 1962. The calculus of consent. Ann Arbor, MI: University of Michigan Press.

Buchanan, James M., and Gordon Tullock. 1975. Polluters' profits and political response: Direct controls versus taxes. The American Economic Review 65:139-47.

Cowen, Tyler, and Dwight R. Lee. 1992. The usefulness of inefficient procurement. Defense Economics 3:219-27.

Crime in America. 1996. The Economist, 8 June, pp. 23-5.

Downs, Anthony, 1957. An economic theory of democracy. New York: Harper and Row.

Haveman, Robert. 1988. Starting even: An equal opportunity program to combat the nation's new poverty. New York: Simon and Schuster.

Holcombe, Randall G. 1994. The economic foundations of government. New York: New York University Press.

Lee, Dwight R. 1989. The impossibility of a desirable minimal state. Public Choice 61:277-84.

Maloney, Michael T., and Robert E. McCormick. 1982. A positive theory of environmental quality regulation. The Journal of Law and Economics 25:99-123.

McKenzie, Richard B., and Dwight R. Lee. 1991. Quicksilver capital: How the rapid movement of wealth has changed the world. New York: Free Press.

Nozick, Robert. 1974. Anarchy, State, and Utopia. New York: Basic Books.

Parshigian, Peter B. 1984. The effects of environmental regulation on optimal plant size and factor shares. The Journal of Law and Economics 27:1-28.

Tietenberg, Tom. 1992. Environmental and natural resource economics. New York: HarperCollins Publishers.

Tullock, Gordon. 1971. Public policy as a public good. Journal of Political Economy 79:913-8.

Tullock, Gordon. 1983. Economics of income distribution. Boston: Kluwer Nijhoff Publishing.

Vaughn, Karen I. 1996. Economic policy for an imperfect World. Southern Economic Journal 62:833-44.

Wittman, Donald. 1989. Why democracies produce efficient results. Journal of Political Economy 97:1396-424.

Wittman, Donald. 1995. The myth of democratic failure: Why democratic institutions are efficient. Chicago: University of Chicago Press.

Dwight R. Lee is Professor of Economics and the Bernard B. and Eugenia A. Ramsey Chair of Private Enterprise Economics, University of Georgia.