William Niskanen on Cutting Taxes and Shrinking Big Government

Reason's 1978 interview with the late libertarian economist


Editor's Note: This interview originally appeared in the November 1978 issue of Reason.

With a background in economics and political studies and experience in the university, the think tank, the government, and the corporation, William Niskanen brings varied skills and insights to the tax-cutting movement. He has been involved in the effort to reduce taxes since 1972, when he helped to draft the unsuccessful Proposition 1 in California. More recently, he has been instrumental in designing the Michigan tax-limitation measure, which will be on the ballot in that state in November, probably along with competing tax-reduction proposals. And now he is working with the National Tax Limitation Committee to draw up a constitutional amendment limiting federal taxation.

Mr. Niskanen moved from an undergraduate education at Harvard to the University of Chicago, where he earned master's and doctorate degrees in economics. He has worked for the Rand Corporation and the Institute for Defense Analysis as well as for the government, in the Defense Department and the Office of Management and Budget. He taught for several years at the Graduate School of Public Policy at the University of California at Berkeley and is the author of Representative Government and Bureaucracy. Since July 1975 he has been chief economist for a major U.S. corporation.

At a recent conference in Virginia at which Mr. Niskanen was a featured speaker, Senior Editor Tibor Machan heard of Niskanen's recent efforts in the tax-cutting department. Along with conference participant Lester Hunt, he took Niskanen aside for some of his thoughts on taxes and governments.

REASON: Would you tell us a little bit about the specifics of this bill or document you are working on?

NISKANEN: We have just formed a committee to draft one or more versions of a national tax-limitation amendment to the US Constitution. Several of us have our own recommended versions that we bring to this committee, but the committee has yet to resolve the specific language of the versions that it will endorse. Bob Bork, who was formerly solicitor general of the United States and is now a professor of law at Yale, is the chairman of tis drafting committee, and we expect to have one or more versions of such an amendment ready by the end of September.

REASON: There are some other people who are probably known to our readers who are part of this effort.

NISKANEN: Yes, the group includes Milton Friedman, formerly at the University of Chicago and now connected with the Hoover Institution; Walter Williams of Temple University; James Buchanan of the Virginia Polytechnic Institute; and a number of members of the original group associated directly with the National Tax Limitation Committee–Lewis Uhler of California, Professor Craig Stubblebine of the Claremont Colleges.

REASON: Is there any connection with the Jarvis-Gann movement that brought about Proposition 13 in California?

NISKANEN: There's no organizational connection. Several members of our group, most importantly Mr. Friedman, strongly supported the Jarvis-Gann initiative, but that came from a different group and is addressed to a quite different problem than the issue to which a general tax-limitation amendment is addressed. Jarvis-Gann is not a general tax limitation. It limits specific tax rates but does not limit overall taxation, even within the state of California.

REASON: Is it pretty clear that, whatever document emerges from your effort, it will be a tax-limitation provision and not like the one in California?

NISKANEN: We would not attempt to limit any specific tax rate. It would be a general limitation on the total taxing plus borrowing by the federal government. Now the issue as to whether that would be limited in terms of some constant share of the total income in the United States or some formula that would lead to a declining share over time is not yet resolved.

REASON: What do you think are the major and popular objections to this effort–that is, what do you anticipate would be some of the most important political obstacles for you to overcome?

NISKANEN: Well, there are two obvious sources of opposition to the idea of the tax-limitation amendment. One is by the fairly large group of people who want government spending to increase. The other group of people, who may have no strong preferences for outcomes in the sense of either larger government or smaller government, believe that it is inappropriate to constrain elected representatives in this particular manner. They feel that the selection of representatives is the proper basis for electoral or voter control over government.

The primary reason why we believe that the tax-limitation amendment ought to be added to the Constitution is much the same reason as for any other amendment–certain features of the government should not be subject to decisionmaking by the body of elected representatives. The whole spirit of the tax-limitation movement is not necessarily to reduce government or to increase it or whatever but to shift one important decision from elected representatives to the people–and that is the share of their income that they are prepared to have government allocate–and to leave to government the entire responsibility for determining the composition of spending and the setting of individual tax rates. The amendments that we have drafted in California and Michigan and that we will be drafting for consideration as a national amendment in no case have constrained the authority of governments to determine how much money goes to particular purposes or groups within the total budget, nor have they constrained any specific tax rate. It leaves those questions of detail to legislatures, where the process of vote trading, or logrolling, is an important feature in assuring effective representation of minority interests. Buut on the question of the budget within which the political officials are operating, we feel that that decision should be made by the electorate themselves and not by the elected representatives.

REASON: Are there any constitutional obstacles that you foresee on grounds that this usurps some of the legislators' powers or responsibilities as defined in the Constitution?

NISKANEN: No. The populace, through constitutionally approved procedures for constitutional change, may constrain the legislature and the executive in any way they choose, and that's the case even if one may feel that the amendment has been poorly designed or whatever. Any amendment, of course, must be passed by constitutionally approved procedures, and at the federal level–I think quite appropriately–those procedures are very restrictive.

REASON: Why do you think that the amount of government spending should not be decided by elected representatives?

NISKANEN: Well, there's a fundamental asymmetry in the political process that leads to an overspending bias. It is primarily a consequence of the fact that political activity itself is costly. Any proposal, for example, which has concentrated benefits but very diffuse costs will be such that the people who bear the cost do not have sufficient incentives to register their concerns through the political authorities or change their votes among candidates on hat basis. Any number of subsidy programs or concentrated special-interest programs would be of that character. That's one important, and probably the most important, general bias in this process–the proliferation of special-interest programs with concentrated benefits and diffuse costs. Ten cents per American family per year finances a $7 million subsidy, and if that subsidy only goes to a few people they have an enormous incentive to use the political process to achieve that subsidy, and no one has much concern about opposing it. That's true of all levels of government almost regardless of how the service is supplied. It's true to the extent that the Constitution, either in fact or implicitly, authorizes those types of special-interest programs.

For a long period of time in American history, we took seriously the limitations on federal functions that are spelled out explicitly in Article I, Section 8, and the 10th Amendment, and as long as we took those limitations very seriously, the federal government was constrained in terms of the taxing activities that it could undertake. But beginning primarily in the '60s, there is no longer any effective constraint on the range of functions the federal government may undertake, and that increases the importance of subjecting the federal government to a more aggressive type of constraint. The 10th Amendment, as you remember, says that any program or any function that is not authorized by the Constitution is reserved for the states or the people, so that the presumption was that unless it was specifically authorized they could not do it. The effective constitutional constraint right now is that unless it is specifically prohibited, they may do it. That's an enormous change; it's happened in our lifetime.

There are a number of other biases, of course, in governmental decision processes which lead to overspending, one of the more important of them being the incentives of the bureaucrats, or the incentives of the people in the civil service, because their incentives are primarily a function of how much the government spends.

The combination of these several processes, I believe, and many people believe, leads to a greater aggregate level of spending than would be desired or would be approved by the electorate if they had a chance to vote on it directly. Now the test of that, of course, is whether such an amendment will pass when an informed electorate has a chance to vote on it. We have had several experiences to date in which they have not passed. Proposition 1 in California in 1973 did not pass by a narrow margin. A similar proposal in Michigan in 1976 did not pass, nor did another in 1974 in Arizona.

REASON: Is there some campaign effort being made now which, learning from the past, would perhaps incorporate better information dissemination?

NISKANEN: Well, we may have learned something about information dissemination, but I think that's less important than the fact that the electorate has changed in the meantime. The same proposal going before California or Michigan now is likely to be approved in a situation where it wasn't approved some years ago just because the mood of the electorate has changed. Just this year, for example, Tennessee has passed a spending limitation as a constitutional amendment. The state of New Jersey has passed a tax-limitation amendment as a statute. I expect that the proposed constitutional amendment in Michigan will pass this fall. Several of the Democratic legislators in California who opposed Proposition 1 in 1973 have expressed the present belief that they wished it had passed, in part to head off the kind of momentum that developed for Jarvis-Gann.

REASON: Given your present work on tax limitation, perhaps you could tell us a little about your prvious scholarly work on the issue of bureaucracy, politicians, and general political processes in this country.

NISKANEN: Well, the primary theme of my scholarly work in the area that is now called public choice is that the decisions that are made by public bodies, bureaus, and representative assemblies are likely to be biased relative to the interests of the population unless the underlying guidance of the rules under which they operate—the Constitution—is correctly formulated. The bureaucratic provision of public services is, in a theoretical sense, an anomaly in the contemporary world because we have no theory which justifies the bureaucratic supply of these government services as distinct, in many cases, from just the governmental financing of them. For reasons that are rather complex, in most cases we have chosen to supply government services from government firms or bureaus instead of, for example, using voucher systems to provide for public financing of education or the purchase of other government services from private firms, and so forth. But bureaus operate rather differently from private firms and particularly if they have a monopoly in the provision of a particular good or service. A profit-seeking private monopoly will sell too little of a product at too high a price, but a public monopoly, because people cannot collect the bureaucratic profits as direct income, is likely to oversupply the service relative to the interests of even the legistlature, let along the interests of the population.

Now one of the problems of the legislature is that it has to divide up its work among committees, but these committees are, to a substantial extent, self-selective. People who have the greatest interest in a particular governmental activity are those who are most likely to dominate that committee. In my book I call this "the case of the high-demand committee," and there is considerable evidence that those legislators who are most concerned about a particular function, in terms of preferring the highest level of spending for those functions, are the ones who dominate the committees who review those functions. And while the committees have some incentive to correct for, let's say, simple inefficiency of the bureaus in the sense that they would like to get more output for the same amount of money, they do not have much incentive to correct for the oversupply of those services because thy have the same kinds of preferences and outcomes—they want more of those services themselves. And one of my reccomendations with respect to legislatures is consistent with the sort of recommendations that I make relative to bureaus and that is to reduce the monopoly power of these review committees—for example, by assigning issues across communities to a number of different committees on occasion, or possibly by shifting people around among the committees.

REASON: Do you have some ideas about reforming the bureaus themselves?

NISKANEN: It has always been rather strange to me, and has led me to thinking through these issues in the first place, why there has been a general presumption that competition in the private sector is desireable but that competition in the public sector is undesirable. On thinking that through, I was led to conclude that competition in the public sector is also desirable and for very much the same reasons as it is in the private sector. So I would like to introduce more competition in the public sector by two means—both by increasing the competition among bureaus, providing similar if not identical output, and by increasing the competition between bureaus and private firms. I see no reason, for example, to make an either/or choice between having public schools and having a voucher plan, in that with a voucher plan parents may well choose to send their children right back to public schools, in which case I would expect the public schools to be better than they are now relative to the purposes of the parents.

Other actions that can be taken within a bureaucracy that I would favor would be to substantially increase the proportion of people within the bureaucracy who are chosen by patronage rather than by the civil service. I think that the substitution of civil service for patronage in essentially all government jobs has made the civil service an independent force in our community in a way in which their interests are served, not by how well they serve whatever party is in power, but by how big the government is. That substantially reduces the leverage of whoever is elected, whatever party is elected, to run the government. Patronage would increase the responsiveness of the civil servants to the elective authorities.

REASON: It seems, to put it roughly, that you would like to make government more efficient, and that may mean that what some of us might call unjust functions of government would be strengthened. And the question arises, isn't that somehow sacrificing the individuals who are subject to these unjust laws for the sake of some goal, and what is that goal and why ought this trade-off be made?

NISKANEN: Well, a twofold response: one is that my concern in general is to make government more responsive. Efficiency is part of responsiveness but is not the whole of it. Ultimately, of course, we would all like the government to be doing the right things in terms of the interests of the population and the community and not just doing the wrong things at a lower cost. I have a personal view, however—which comes less from analytic reasoning than from observations of political behavior—that the issue of whether government should not be doing something is often never forced unless you force th issue of its efficiency. There is a wide range of things that I would prefer the government not to do, but I think realistically an essential part of the political process to change those laws, regulations, programs, is to enforce them, and enforce them generally, and to do them as efficiently as possible so that the issue of whether the government out to do it is forced and is not confused with the question of arbitrary justice or inefficiency or something.

REASON: Your example of the prohibition issue should come in here, I think.

NISKANEN: That's right. The history of the elimination of state prohibition after the federal prohibition ammendment was finally repealed is that the repeal of the state prohibition happened only after the governor or the state government chose to enforce the prohibition law on a rather uniform basis. As long as legislators and public men and women could have both their law and their liquor, they maintained that law which they themselves violated. I see no prospect, for example, of eliminating or reducing what I would regard as the oppressive marijuana laws until the sons and daughters of mayors and of governors and of legislators are charged with violations of these laws. I think that enforcing that law is a necessary process for people to sort out whether the law is good or bad.

REASON: Thank you very much, Dr. Niskanen.