January/February 2012

The Growth of Government

By William A. Niskanen

In late November 1972, Governor Reagan convened a small group at the Century Plaza in Los Angeles. In his opening remarks, he observed that conventional politics, including the recent reelection of a Republican president by a large majority, would probably not be sufficient to slow the growth of government. Several other speakers summarized what little we knew about the reasons for the growth of government. My own remarks focused on the erosion of the constitutional restraints on the powers of the federal government, powers that were specifically enumerated in Article I, Section 8, of the Constitution and specifically circumscribed by the Tenth Amendment. Governor Reagan then asked us to consider the feasibility and desirability of a constitutional limit on the total tax and spending power of the state.

That meeting was the genesis of the contemporary tax-limitation movement in the United States, a movement whose history is not yet complete. The group formed in Los Angeles later drafted and promoted the first major proposal for a constitutional limit on the total tax and spending power of a state, a proposal that was appropriately termed "Proposition 1." The early record of this movement was not encouraging. Proposition 1 was defeated in California in 1973, and a similar proposal in Michigan was defeated in 1976. The year 1978, however, proved to be a watershed for the tax-limitation movement. General limits on state taxing authority were approved in Tennessee and Michigan, and property taxes were substantially reduced in California and Massachusetts.

Since that time, general tax limitations have been approved in an additional half-dozen states, and the legislatures of 32 states have called for a similar amendment to the federal Constitution. For state and local governments, a limitation on or reduction in taxes can be an effective limit on government spending because most such governments are constrained to have a balanced budget on their operating accounts. For the federal government, however, tax limitation is not sufficient to constrain spending because the federal government has, effectively, unlimited authority to borrow. An effective restraint on federal spending, thus, must constrain both taxing authority and borrowing authority.

A proposed amendment to the federal Constitution that would change the voting rules on tax and borrowing authority was approved by more than two-thirds of the Senate and by more than a majority of the House in 1982 and is being considered in the Senate again this year. For those of us who are concerned about the long-term growth of the federal government, approval of this proposed amendment should be the highest priority.

The election and reelection of President Reagan, however, have somewhat defused the pressure for more general restraints on the powers of government, based on a plausible belief that the election of the most conservative president in our lifetime would be sufficient to slow the growth of government. Governor Reagan knew better; real (inflation-adjusted) state spending in California grew more rapidly during his terms as governor than during any other equivalent period. President Reagan should know better; real federal spending increased about as rapidly during his first term as president as during the Carter administration.

Moreover, there is no general evidence that the growth of government spending depends on the party in office. An examination of federal spending during the years since World War II, based on some research that I completed early this winter, indicates the following.

First, real defense spending increased more rapidly under Democratic presidents, real domestic spending increased more rapidly under Republican presidents, but there was no significant difference in the growth of real total federal spending from one party to another during this period.

Second, the composition of the growth of federal spending under President Reagan was significantly different from that under other Republican presidents but was not significantly different from that under all the postwar presidents taken as a group.

Third, the amount and composition of the federal spending share of the gross national product can be explained almost entirely by three conditions: the level of real per capita GNP, the number of U.S. armed forces overseas, and the unemployment rate. Controlling for these conditions, federal spending was independent of the party of the president, the party that controlled the House, and the party that controlled the Senate.

A fourth observation bears on two competing hypotheses of the relation between federal spending and federal taxes. One view, shared by Milton Friedman and Ronald Reagan, is that reducing taxes will reduce spending. The contrary view, shared by James Buchanan and Herbert Stein, is that reduced taxes will increase spending by reducing the perceived price of government services. Over the postwar years, controlling for the three conditions mentioned earlier, federal spending appears to have been independent of federal tax receipts. In other words, a reduction in federal tax receipts, for a given level of GNP, increases the deficit by a roughly equal amount.

One interpretation of this evidence is that our political system is working quite well, responding to a change of preferences represented by both parties but independent of the governing party. Another interpretation is that our political system is not working very well, due to common biases in the behavior of elected officials of both parties. For the moment, the available empirical tests are not sufficient to choose between these contrary perspectives. My own perspective is that our political system does not serve us very well, a view that has been strongly shaped by my own government experience and by the developing contributions of the public-choice scholars.

The major biases in our political system are now more broadly perceived. Most people have little incentive to invest in information about government or, for that matter, to vote. Concentrated interests have more incentive to invest in political activity than the broader community of consumers and taxpayers. The regional basis for representation in Congress elevates regional concerns relative to more broadly shared national concerns. Elected officials have an unusually short time horizon, given their limited terms of office and the erosion of party control. The incentives of bureaucrats are not consistent with either the efficient supply of government services or the optimal level of these services. The elected officials and bureaucrats who set the agenda for subsequent decisions have unusual power over the outcomes of these decisions.

These major biases, however, have been inherent in our political system since the dawn of our republic, and they do not explain why the federal share of GNP has increased from 2.5 percent in 1929 to 25 percent in 1985. What does explain it is the erosion of the constraints on the powers of the federal government, a change in our effective constitution without any change in our formal constitution. This change is the most important political development in our lifetime.

James Madison, the primary architect of the Constitution, described the powers of the federal government in the Federalist No. 45 as "few and defined... to be exercised principally on external objects, as war, peace, negotiation, and foreign commerce." Thomas Jefferson, in his first inaugural address, described the powers of the federal government as limited to "the external and mutual relations only of these states." In 1936, however, the Supreme Court ruled in United States v. Butler that "the power of Congress to authorize appropriations of public money for public purposes is not limited by direct grants of legislative power found in the Constitution." The Court, apparently, paid no attention to Madison's defense of the proposed Constitution in the Federalist No. 41, where he argued that "the idea of an enumeration of particulars which neither explain nor qualify the general meaning, and can have no other effect than to confuse or to mislead, is an absurdity." After his retirement as president, Madison wrote that the General Welfare clause slipped into the Constitution as a consequence of "inattention to phraseology occasioned doubtless by its identity with the harmless character attributed to it in [the Articles of Confederation] from which it was borrowed."

My own reflections on this issue, summarized in an article in 1975, concluded that "at the present time, the enumerated functions do not even command lip service. The U.S. Constitution, in terms of its effectiveness in constraining the functions of the federal government, is a dead letter." Moreover, no coherent theory of the state has been substituted for the formal constitutional restraints. During the 1950s, economists developed a theory of the state, described as "welfare economics," based on the provision of public goods and the correction of externalities in market transactions. This theory — formulated by Arrow, Bator, Baumol, Samuelson, and others — provides a satisfactory framework for the role of the state, but only if government officials are all-knowing saints. Welfare economics overlooked the fundamental insight of our constitutional founders: that government should be so structured that the behavior of government officials in pursuit of their private interests also serves the public interest.

In any case, welfare economics did not provide any effective restraints on the developing functions of government. Many government activities are not public goods. Many government activities create additional externalities. Much of the activity of the modern state, like that of imperial Rome, consists of providing various forms of bread and circuses, goods and services that are adequately supplied by the market.

For some of you, this may be your first exposure to the Cato Institute. What can a small nonpartisan policy institute, operating in a city fascinated with power, do to address these issues? The role of Cato, like that of similar institutes, is based on the premise that changes in government policy should be based on informed consent. In pursuit of that role, we document the effects of a wide range of government activities and develop alternatives that will better serve the interests of a free community.

Most important, we try to shape the perspective by which people address policy issues. Does a specific policy increase personal choice? Does it reflect the consent of those affected? Does it respect the considerable diversity of the American community? As with other institutes, we bring a number of analytic perspectives to bear on these issues. Some of our resident and adjunct scholars have been influenced especially by the Austrian economic tradition, best represented in the writings of Mises and Hayek.

Others have been more influenced by the neo-classical tradition or the developing field of public choice. We differ from many other institutes primarily in our political values. These values have been characterized as "radical conservative" by the liberal press, as "anarchist" by the conservative press, and sometimes as "libertarian." I am personally uncomfortable with any of these labels. We share the values of Locke, Smith, and Mill — people who were known as Whigs and, later, as Liberals in British political thought. We share the values of Jefferson and Madison, the authors of the leading charters of American liberty, people who called themselves both Republicans and Democrats. We will differ from the dominant political traditions primarily when they try to use the powers of the state to impose their particular values on the larger community. We will oppose contemporary liberals when they fail to distinguish between a virtue and a requirement.

We will oppose contemporary conservatives when they fail to distinguish between a sin and a crime. This perspective — libertarian, if you wish — reflects a shared set of political values among people who may have widely different personal values or religious beliefs, a perspective, I contend, that was and is the essence of the original and continuing American revolution. The terms that are used to describe this perspective are not important. We ask you to read our publications. We will value any contribution you can make to our funding. Spell our name correctly, and call us what you will.

Governor Reagan was prescient in recognizing that the reelection of a Republican president by a large majority would not be sufficient to slow the growth of government. That is still the case. President Reagan will earn our support when the actions of his administration are consistent with his long-held convictions. He will deserve our opposition when they are not. The cause of liberty is not especially dependent on which party or person is elected to political office. It is more dependent on the views shared by both parties on the importance of limiting the role of the state to those activities that serve the common defense and the general welfare.

William A. Niskanen, distinguished senior economist and chairman emeritus of the Cato Institute, passed away on October 26, 2011. This article, the inaugural lecture he gave after assuming his position at Cato in 1985, first appeared in the July-August 1985 issue of Cato Policy Report. It is presented here unchanged, and is still timely.