Democracy in Deficit: The Political Legacy of Lord Keynes

Keynesian policy is centered on the use of the government’s budget as the primary instrument for ensuring the maintenance of high employment and output. The implementation of Keynesian policy, therefore, required both the destruction of former principles of balanced public budgets and the replacement of those by principles that permitted the imbalance necessary for Keynesian budgetary manipulation. But politicians, and the public generally, were not urged, by Keynes or by the Keynesians, to introduce deficit spending without a supporting logical argument. There was more to the Keynesian revolution than mere destruction of the balanced-budget principle as a permanent feature of the fiscal constitution. This destruction itself was a reasoned result of a modified paradigm of the working of an economy. And, in the larger sense, this is really what “Keynesian” is all about, as we have already noted in Chapter 3. The allocative bias toward a larger public sector and the monetary bias toward inflation are both aspects of, and to an extent are contained within, a more comprehensive political bias of Keynesian economics, namely, an “interventionist bias,” which stems directly from the shift in paradigm.
*13

In an inherently unstable economy, government intervention becomes practically a moral imperative. And there is no argument for allowing for a time period between some initially observed departures and the onset of policy action. “Fine tuning” becomes the policy ideal.
*14 The notion of an unstable economy whose performance could be improved through the manipulation of public budgets produced a general principle that budgets
need not be in balance. There would be years of deficit and there would be years of surplus, with these deficits and surpluses being necessary for macroeconomic management. A stable relation between revenues and expenditures, say a relatively constant rate of surplus, would actually indicate a failure on the part of government to carry out its managerial duties.

As we noted in our earlier discussion of the Keynesian political framework, the budgetary policies were to be applied symmetrically. In the Harvey Road political setting, it might even be said that Keynesian economics did not destroy the principle of a balanced budget, but merely lengthened the time period over which it applied. The Keynesian paradigm, in other words, would not be viewed as essentially changing the fiscal constitution within which economic policy is conducted. But what happens when we make non-Keynesian assumptions about politics?

Taxing, Spending, and Political Competition

The political process within which the Keynesian norms are to be applied bears little or no resemblance to that which was implicit in Keynes’ basic analysis. The economy is not controlled by the sages of Harvey Road, but by politicians engaged in a continuing competition for office. The political decision structure is entirely different from that which was envisaged by Keynes himself, and it is out of this starkly different political setting that the Keynesian norms have been applied with destructive results. Political decisions in the United States are made by elected politicians, who respond to the desires of voters and the ensconced bureaucracy. There is no center of power where an enlightened few can effectively isolate themselves from constituency pressures. Furthermore, since World War II, the national economy has never been appropriately described as being in depression of the sort idealized in the elementary Keynesian models. Throughout the three decades of postwar experience, increases in aggregate demand have always been accompanied by increases in price levels, by inflation.

In a democracy, the pressures placed upon politicians to survive competition from aspirants to their office bear certain resemblances to the pressures placed upon private entrepreneurs. Private firms compete among themselves in numerous, complex ways to secure the patronage of customers. Politicians similarly compete among themselves for the support of the electorate, and they do this by offering and promising policies and programs which they hope will get them elected or reelected. A politician in a democratic society, in other words, can be viewed as proposing and attempting to enact a combination of expenditure programs and financing schemes that will secure him the support of a majority of the electorate. These similarities suggest that political competition is not wholly unlike market competition.

There are also obvious and important differences between market and political competition. Market competition is continuous; at each instance of purchase, a buyer is able to select among alternative, competing sellers. Political competition is intermittent; a decision is binding for a fixed period, usually two, four, or six years. Market competition allows several competitors to survive simultaneously; the capture by one seller of a majority of the market does not deny the ability of the minority to choose its preferred supplier. By contrast, political competition has an all-or-none feature; the capture of a majority of the market gives the entire market to a single supplier. In market competition, the buyer can be reasonably certain as to just what it is he will receive from his act of purchase. This is not true with political competition, for there the buyer is, in a sense, purchasing the services of an agent, but it is an agent whom he cannot bind in matters of specific compliance, and to whom he is forced to grant wide latitude in the use of discretionary judgment. Politicians are simply not held liable for their promises and pledges in the same manner that private sellers are. Moreover, the necessity for a politician to attain cooperation from a majority of politicians produces a situation in which the meaning of a vote for a politician is less clear than that of a vote for a private firm. For these reasons, as well as for several others, political competition differs in important respects from market competition, even where there is also a fair degree of similarity.
*15

The properties of political competition and the characteristics of the budgetary policy that emerge may be examined in several ways. Indeed, the public-choice literature now possesses a variety of analytical models designed to explore and explain the properties of alternative institutional frameworks for political competition.
*16 A government can provide a single service, or it can provide a combination of services. It can finance its budget by a variety of tax forms, either singly or in combination, and, additionally, it can subject any particular tax to a variety of rate schedules and exemption rules. Furthermore, preferences for public services can differ as among individual citizens; particular features of the political system can vary; and budget imbalance can be permitted.

Changes in any of these particular features will normally change the budgetary outcomes that emerge. Changes in tax institutions, for instance, will normally change the tax shares and tax prices assigned to different persons. This, in turn, will alter individual responses to particular budgetary patterns. The number of services provided may also matter. With a single service, it is fruitful to conceptualize budgetary outcomes in a plurality electoral system as conforming to the preferences of the median voter. With multiple services, however, the conceptualization is not necessarily so simple, for a trading of votes may take place among persons over issues.

The essential features of democratic budgetary choice may be illustrated quite simply. This may be done by considering the gains and losses to politicians of supporting alternative-sized budgets. Suppose for now that a balanced-budget constraint exists. We can start with a budget of zero, and then take account of the gains and losses in terms of constituent support from expansions in the size of the budget. Under the assumption that public output enters positively into the utility functions of citizens, the expenditure by itself will secure support for the politician. The taxes, however, will reduce the disposable income of citizens, thereby affecting them negatively and reducing support for the politician. In a plurality electoral system, for given preferences and fixed tax institutions, the budget will be expanded so long as a majority would prefer the public service to the private goods they would have to sacrifice via taxation.
*17

A detailed description of the various analytical possibilities concerning the character of political competition in a democratic society would require a survey of a quite complex literature.
*18 For our purposes in this book, however, the central notions we have just described are sufficient. What this line of analysis suggests is that the consideration by politicians of the gains and losses in terms of constituent support of alternative taxing and spending programs shapes the budgetary outcomes that emerge within a democratic system of political competition. The size and composition of public budgets in such a system of competitive democracy, in other words, can be viewed as a product of the translated preferences of a subset of politicians’ constituents and the constitutional-institutional rules that constrain the political system.
*19

Unbalanced Budgets, Democratic Politics, and Keynesian Biases

With a balanced-budget rule, any proposal for expenditure must be coupled with a proposal for taxation. The elimination of this rule altered the institutional constraints within which democratic politics operated. Two subtly interrelated biases were introduced: a bias toward larger government and a bias toward inflation. Before examining the foundations of the inflationary bias, it may be useful to examine the relationship between these two biases.

The allocative bias stems from the proposition that, if individuals are allowed to finance publicly provided goods and services through borrowing rather than through taxation, they will tend to “purchase” more publicly provided goods and services than standard efficiency criteria would dictate.
*20 The inflationary bias stems from the proposition that, for any given level of public goods and services, for any size of the budget, individuals will tend to borrow rather than to undergo current taxation, at least to an extent beyond the financing mix that would be ideally dictated by either classical or Keynesian criteria. The first bias entails the hypothesis that, because of government borrowing, government spending will be excessive; the second bias entails the hypothesis that, regardless of spending levels, government
borrowing will be excessive.
*21 The public-choice analytical framework makes it possible to see how taxation and debt finance exert differing effects on observed political outcomes. In considering the nature of the pressures of political competition in a revised, Keynesian constitutional setting, we shall consider, in turn, budget surpluses and budget deficits.

Budget surpluses and democratic politics

The creation of (or an increase in) a budget surplus requires an increase in real rates of tax, a decrease in real rates of public spending, or some combination of the two. In any event, creating or increasing budget surpluses will impose direct and immediate costs on some or all of the citizens in the community. If taxes are increased, some persons in the community will have their disposable incomes reduced. If public spending is reduced, some current beneficiaries of public services will be harmed. In terms of direct consequences, a policy of budget surpluses will create losers among the citizenry, but no gainers.

Indirectly, there may be some general acceptance of the notion that the prevention of inflation is a desirable objective for national economic policy.
*22 It could be argued that citizens should be able to see beyond the direct consequences of budget surpluses. They should understand that a budget surplus might be required to prevent inflation, and that this would be beneficial. The dissipation of what would otherwise be a surplus through public spending or tax cuts, therefore, would not be costless, for it would destroy those benefits that would result from the control of inflation.

These direct and indirect consequences impact quite differently, however, on the choice calculus of typical citizens. The direct consequences of the surplus take the form of reductions in
presently enjoyed consumption. If taxes are raised, the consumption of private services is reduced; if expenditures are reduced, the consumption of public services is reduced. In either case, the policy of budget surplus requires citizens to sacrifice services that they are presently consuming.

The indirect consequences, on the other hand, are of an altogether different nature psychologically. The benefit side of the surplus policy is never experienced, but rather must be
creatively imagined, taking the form of the hypothetical or imagined gains from avoiding what would otherwise be an inflationary history. This is a gain that, by its very nature, can never be experienced, but can be only imagined as it is created in the mind of the individual citizen.
*23

A variety of evidence suggests that these choice alternatives are dimensionally quite distinct. Moreover, the sensed benefits from the surplus would be diminished by the severity of the information requirements that confound the citizen’s efforts constructively to imagine the benefits that would result from a surplus. The choice is not at all a simple matter of choosing whether or not to bear $100 more in taxes this year in exchange for $120 of benefits in two years, and then somehow to compare the two, historically distinct moments in being. The imagining process requires an additional step. The person must form some judgment of just how he, personally, will fare from the surplus; he must reduce his presumption of the aggregative impact of the surplus to a personal level. As such future gains become more remote and less subject to personal control, however, there is strong evidence suggesting that such future circumstances tend to be neglected, with “out of sight, out of mind” being the common-sense statement of this principle.
*24

In sum, budget surpluses would seem to have weaker survival prospects in a political democracy than in a social order controlled by “wise men.” Budget surpluses may emerge in a democratic political system, but there are institutional biases against them. A person may oppose the creation of a budget surplus for any one or any combination of the following reasons:

1.  He may be among those whose taxes will be directly increased, or among those whose benefits from publicly supplied goods and services will be reduced.

2.  He may be among those who consider their own economic position (as workers, as investors, and as owners of assets) to be vulnerable to downward shifts in aggregate demand.

3.  And he may be among those who anticipate the prospect of making economic gains from inflation.

For any one or any combination of these three reasons, a person may object to demand-decreasing budgetary adjustments. In offset to this, there remains only some generally hazy notion that price-level stability is preferred to inflation, along with some sort of understanding that this objective is supposed to be related to his own direct tax and benefit streams. Viewed in this light, there really should be no difficulty in understanding why we have never observed the explicit creation of budget surpluses during the post-Keynesian years.

Budget deficits and democratic politics

In a democratic society, there would be no political obstacles to budget deficits in an economy with genuine Keynesian unemployment. Budget deficits make it possible to spend without taxing. Whether the deficit is created (or increased) through reduced taxes or increased expenditures, and the particular forms of each, will, of course, determine the distribution of gains among citizens. The central point of importance, however, is that, directly, there are only gainers from such deficits, no losers. In the true Keynesian economic setting, of course, this is as it should be.

The problem with deficit finance is not that it would be supported in a Keynesian economic setting, but that it would also be supported even if the economy were distinctly non-Keynesian. Once the constitutional requirement of budget balance is removed, there are pressures for budget deficits, even in wholly inappropriate, non-Keynesian economic circumstances. If we assume that the money supply is at all elastic in response to budget deficits, a proposition that we support in Chapter 8, deficits must be inflationary in a non-Keynesian economy. The direct effects of budget deficits are sensed only in terms of personal gains. The creation or increase of a deficit involves a reduction in real tax rates, an increase in real rates of public spending, or some combination of the two. In any event, there are direct and immediate gainers, and no losers, regardless of whether the economy suffers from Keynesian unemployment or is blessed with full employment.

In a non-Keynesian economy, deficits will indirectly create losers through the consequences of inflation. These indirect consequences, however, are dimensionally different from the direct effects, just as they were with respect to budget surpluses. The direct consequences of deficit creation take the form of increased consumption of currently enjoyed services. These would be privately provided services if the deficit involves a tax reduction, and would be publicly provided services if the deficit involves an increase in public output.

The indirect consequences, by contrast, relate not to present experience, but to future, conjectured experience. The benefit of deficit finance resides in the increase in currently enjoyed services, whereas the cost resides in the inflationary impact upon the future, in a creatively imagined reduction in well-being at some future date. The analysis of these indirect consequences is essentially the same as that of the indirect consequences of budget surpluses, so there is little to be gained from repeating the analysis in detail. It suffices to say that the act of creatively imagining the future is compounded by the twin factors of remoteness and the absence of personal control.

Democratic societies will tend to resort to an excessive use of debt finance when they have permitted Keynesianism to revise their fiscal constitutions. Deficit finance will generate political support, even in a non-Keynesian economy, among those persons for whom either of the following two conditions holds:

1.  persons who are among those whose real taxes will be reduced or among those who expect to experience an increased flow of benefits from government, including receipt of direct monetary transfers; and

2.  persons who are among those who consider their own economic positions (as employees, as investors, as owners of real assets, or bureaucrats) likely to be improved as a result of the increase in aggregate demand.

These direct supporting consequences may even be supplemented by a general, but unnecessary, notion that increased real output and employment are worthy objectives for national economic policy coupled with some idea that these may be achieved by the creation of deficits. In opposition of deficits, there is only some future sensed anticipation of the benefits resulting from living in the absence of the inflation, with its various consequences, that would otherwise have taken place.

The post-Keynesian record in fiscal policy is not difficult to understand. The removal of the balanced-budget principle or constitutional rule generated an asymmetry in the conduct of budgetary policy in competitive democracy. Deficits will be created, but to a greater extent than justified by the Keynesian principles; surpluses will sometimes result, but they will result less frequently than required by the strict Keynesian prescriptions. When plausible assumptions are made about the institutions of decision making in political democracy, the effect is to create biases against the use of budgetary adjustments that are aimed at the prevention and control of inflation, and biases toward budgetary adjustments that are aimed at stimulating employment.

Deficit Finance and Public-Sector Bias

This bias toward deficits produces, in turn, a bias toward growth in the provision of services and transfers through government. Deficit financing creates signals for taxpayers that public services have become relatively cheaper. Because of these signals, voters will demand a shift in the composition of real output toward publicly provided services (including transfers). The “true” opportunity costs of public goods relative to private goods will not, of course, be modified by the use of the budget for purposes of stabilization. To the extent that voters, and their elected legislators, can recognize these “true” cost ratios, no public-spending bias need be introduced. It does not, however, seem at all plausible to suggest that voters can dispel the illusion of a relative price change between public and private goods.
*25

Consider the following highly simplified example. In the full-employment equilibrium assumed to have been in existence before an unanticipated shortfall in aggregate demand, the government provided
one unit of a public good, and financed this with a tax of $1. The restoration of full employment requires a monetary-fiscal response of 10¢. Suppose now that the response takes the form of reducing tax rates. Taxes fall so that only 90¢ is collected while $1 continues to be spent. The tax price per unit of public output is only 90 percent of its former level. At any tax-price elasticity greater than zero, equilibrium in the “market” for the public good can be restored only by some increase in quantity
beyond one unit, with the precise magnitude of the increase being dependent on the value of the elasticity coefficient. So long as individuals concentrate attention on the value of public goods, defined in the numeraire, there will be a clear bias toward expanding the size of the public sector in real terms, despite the presumed absence of any underlying shift in tastes.
*26

For this effect to be operative, individuals must confront tax institutions in which marginal tax price moves in the same way as average tax price. This requirement is met with most of the familiar tax instruments; proportional and progressive income taxation, sales taxation, and property taxation all possess this attribute. Tax reductions are normally discussed, and implemented, through reductions in
rates of tax applied to the defined base. So long as a deficit-facilitated tax reduction takes this form, the terms of trade between public goods and private goods will seem to shift in favor of the former.

The institutionally generated illusion, and the public-spending bias that results from it, can be dispelled if marginal tax prices are somehow held constant while tax collections are reduced inframarginally. If a deficit-facilitated tax cut could take this latter form, there would be no substitution effect brought into play; individuals would continue to confront the same public-goods-private-goods trade-off,
at the margin, before and after the shift in fiscal policy. It is difficult, however, to construct permanent institutional arrangements that will meet this marginal tax-price criterion. For temporary tax cuts, a pure rebate scheme accomplishes the purpose. Such action does not modify tax rates
ex ante, and, hence, marginal tax prices. A pure rebate scheme that is not anticipated offers an allocatively neutral scheme of injecting new currency into an economy during a temporary lapse into a pure Keynesian setting. If, however, the spending shortfall is expected to be permanent, and to require continuing injections, rebates will come to violate allocational neutrality for familiar reasons. As soon as persons come to anticipate the
ex post rebates in making their budgetary decisions
ex ante, they will act as if marginal tax prices are reduced. To forestall the bias toward public-sector spending in this permanent setting, some other institutional means of maintaining constancy in marginal tax prices would have to be developed.

The one-sided application of Keynesian policy remedies, which emerges from a democratic political setting, may itself create instability in the process. It has increasingly come to be realized that inflation may not generate employment. In fact, inflation may attract resources into employments that cannot be maintained without further inflation.
*27 The combined inflationary and interventionist bias of the Keynesian paradigm, therefore, may inject instability into a non-Keynesian economy. In this fashion, the application of Keynesian prescriptions may create a self-fulfilling prophecy, a possibility that we consider in Chapter 11. Keynesianism, in other words, may have changed the fiscal constitution in political democracy, and with destructive consequences.
*28