Can We Trust Federal Spending as an Economic Stabilizer?

It is basically inconsistent for people to vie for government contracts or clamor for local support from the federal government and then to dismiss fiscal policy as ineffectual “pump-priming.” The fiscal issue that concerns us here is whether government expenditures should be jockeyed about in response to changing economic circumstances or whether they should serve simply as a reasonably firm part of the economic framework within which other policies operate. This latter issue is the real issue, and much of what follows deals directly or indirectly with it.

Second, far too much of the debate over fiscal policy revolves around deficits and the federal debt. Deficits per se are neither good nor bad; they may be inflationary or deflationary. They may be automatic (determined by general economic conditions) or discretionary (determined by deliberate action or inaction on the part of the government). Theoretically and practically, a budget deficit or surplus by itself is no guide to action, no measure of fiscal impact, and no indicator of fiscal responsibility.

No one except a small minority really believes that it is possible or desirable for the federal government to balance its budget each and every year. The premises are palpably absurd: this minority asserts that the government should cut its spending simply because private demand is declining, or should raise tax rates in a recession to maintain a constant flow of tax dollars from falling levels of national income. Even the Chicago Tribune hauls up at the brink of this abyss. But this does not mean that the balanced-budget notion is nonsense; nor does it mean, as we shall see, that discretionary, ad hoc fiscal policies have very much technical or operational merit as economic stabilizers.

A rule of cost

The annually-balanced-budget dictum is essentially a simple and important rule of cost: that, when resources are transferred from private to public use, they should be paid for honestly, openly, and fully. Considering mankind’s long experience with fiscal chicanery, it is not a bad idea—only an unworkable one.

In a recession, deficits are inevitable because of our flexible tax structure and the automatic increases that occur in various welfare payments, and they are beneficial because disposable income, consumer spending, and employment are maintained at higher levels than would otherwise be the case. Furthermore, it is an incontrovertible fact that debt-financed government expenditures are no more inflationary, in a purely monetary sense, than are tax-financed expenditures—as long as the funds are borrowed in the open market at unpegged rates of interest. Both methods of finance take money from the public and take bank reserves out of the financial system. It is also a historical fact, however, that past deficits (at home and abroad) often have been financed by inflationary methods, or have led to inflationary policies at a subsequent date, because of political reluctance to pay market rates of interest on accumulated deficits, i.e., on the national debt. Hence, the balanced-budget idea contains a second grain of empirical truth embedded in its unworkable formula.

With this clearing of the decks, let us now review the principal policy issues. The difficulties of trying to use discretionary ad hoc shifts in government expenditures as an economic balance wheel are well known yet often ignored.

Three time lags

The biggest problem is one of time and time lags. According to a popular schema adopted from a now classic article by Lloyd Metzler on inventories, there are three time lags that will always plague the policy maker.

First is the information lag. We never know where the economy is at the moment—only where it was last month or last quarter. Because our statistical indicators are unavoidably late at best and because of the difficulties of short-term forecasting, considerable time elapses before the need for policy action is clearly perceived.

Second is the decision lag. After the need for action is recognized, decisions take time. What kinds of expenditures should be increased, by whom, and by how much? The Bureau of the Budget simply cannot “push the funds out.” Other policy questions also must be considered. For example, should certain kinds of military procurement proceed apace when the Department of Defense is yet uncertain as to whether a particular weapons system is really what is wanted in the long run?

According to one of several “Stigler’s Laws,” the government cannot do anything quickly. From observation I can tell you how this law operates in connection with fiscal policy. What usually happens is that internal pressures to accelerate the flow of obligations and expenditures build up gradually within policy councils. Then the various agencies must be consulted, and they generate a veritable snowstorm of paper. This process takes considerable time and, no doubt, yields considerable job satisfaction to government employees. There is much activity but little action. By the time a significant fiscal shift takes place, the need has often passed, and the effects are often perverse. Where congressional action is needed, as in the case of a tax cut, the decision lag can be interminable. More than two years of debate and harangue preceded the 1964 tax cut, and all during that time the administration stoutly maintained that immediate action was desperately needed for short-term economic stimulus. By the time the tax bill passed, fears were growing that it might “overheat” the economy in the short run.

The third lag is the time required for the economic adjustments to take place. The effects of any policy—a significant shift in federal expenditures, for example—spread gradually through the economy. Public works and major procurement items have a long lead time. Even allowing for some immediate stimulus to private activity when contracts are let and before expenditures actually are made, the process of adjustment is slow. In military procurement, expenditures tend to rise about nine months after new obligations are made. The secondary multiplier effects of a rise or fall in government spending build up 12 to 18 months after the shift in policy is made.

Other practical problems

In practical terms, these three lags severely limit the scope for stabilizing expenditure policies. Indeed, there is an ever-present danger that fiscal manipulations will operate in a procyclical manner to accentuate the swings of the business cycle, rather than moderate them.

To speed things up and reduce these inevitable time lags, a variety of proposals has been made over the years, ranging from a reserve shelf of public works to more-flexible administrative procedures in the Bureau of the Budget. The late President Kennedy, for example, requested from Congress executive authority to cut tax rates within a range and to undertake discretionary public works projects as economic stimulants when he deemed these actions to be desirable. As might have been expected, Congress did not rush to divest itself of its traditional prerogatives. For the present, at least, the forecasting, timing, and time-lag problems still loom as very large, if not insurmountable, obstacles to any full-blown stabilizing expenditure policy.

There are further difficulties. Stabilizing expenditure policy is clumsy. It calls for whipsawing governmental expenditures back and forth, expanding them in recession and cutting them back in a boom. In the process it seems highly likely that spending decisions will be even less rationally made than they are now. Governmental services are either wanted and worth what they cost or they are not, and spending decisions should be made on the basis of strict cost-benefit principles, as difficult as they are to apply in governmental budget making. As a practical matter, it is impossible to manipulate government spending in order to stimulate or depress aggregate demand without creating economic inefficiencies. For this reason, tax policy and monetary policy, which leave expenditures more or less fixed or to be decided upon the basis of independent criteria, have important advantages as stabilizing instruments.

Making fiscal policy work in context

In strict theory, of course, optimum combinations of several policies can be devised. For maximum output, resources employed in any particular manner should be as productive as in all possible alternative uses. This means that resources invested publicly should have the same marginal productivity as resources invested privately. Presumably, prospective government investments can be ranked in descending order of productivity, with fewer projects having higher rates of return (interest), and many more projects having lower anticipated yields.

Assume for the moment that the economy is in proper balance at high employment with both the private and the public sectors having equal marginal productivities. Now, private-investment demand shifts sharply downward. To stimulate private investment sufficiently to restore total demand to previous levels, monetary policy would have to be sharply expansionary and interest rates would have to fall far enough to induce the necessary private spending. However, at these lower interest rates, there would be underinvestment in the public sector, and our efficiency rule would be violated. Ideally, there should be some combination of money creation and increased government investment that would simultaneously restore full-employment demand and maximize overall production. Similarly, the efficiency rule requires a cutback of government spending and monetary restraint when private demands are shifting upward beyond the full employment level at current interest rates.

All this is well and good—and technical. In reality, no precise schedules are available to policy makers, nor are policy decisions made on the basis of a neat economic calculus. Budgets are, at best, crude economic instruments. During a recession, they become the vehicle upon which politicians load all their pet spending schemes. For this reason, many well-intentioned people distrust the use of expenditures as a stabilization device. They fear that government expenditures will be expanded willy-nilly in a recession (ostensibly to promote recovery) and expanded again in boom times (when tax receipts are permissibly high). In short, they fear that “stabilizing” expenditure policy will be used as a one-way street to permanent enlargement of the size and scope of government and that no sensible economic calculations will be used to control the traffic. Who would say that such fears are completely irrational or unfounded?

The last set of operational problems I want to mention deals with prediction, control, and substitutability. Much discussion of both expenditure and tax policy takes place within a vacuum. The effects of a change in expenditures or in tax-rate schedules will depend critically upon the direction of the accompanying monetary policy. Only to a limited extent are fiscal policies a substitute for monetary policies. In the main, they are complements, and the hoped-for stimulative or depressive effects of fiscal policy will not come to pass without cooperating monetary measures.

Consider the multiplier effects on total spending induced by an increase in government spending or by a cut in taxes. How is one to predict their magnitude? Will the Board of Governors of the Federal Reserve System be willing to increase the stock of money in sufficient amounts to finance additional transactions? If not, interest rates will rise and private spending will decline by an unpredictable amount, to offset, at least partially, the hoped-for build-up of secondary or induced spending.

This is a simple case. The difficulties are more fundamental and complicated. Besides capital-market offsets, there may be adverse price effects (bottlenecks), foreign-trade effects, inventory effects, and a host of other induced ripples, positive and negative, that will affect the outcome in any given situation.

Moreover, the size of the expected multiplier depends critically on the stability of the relation between private consumption expenditure and national income. If stability conditions do not hold, the value of the multiplier cannot be predicted with confidence. When one allows for other dynamic economic changes, the difficulties are compounded.

The President’s Council of Economic Advisers speaks blithely of “plugging in the best multiplier that economic science can produce.” Unfortunately, the best is not very good. Economic science simply cannot produce a very good number that will accurately predict multiplier effects in advance. This honest admission of uncertainty is, in itself, no bar to the use of fiscal policy, but it should cause economists to adopt a proper attitude of humility in making their recommendations.

Predictability is paramount

Prediction and control are crucial for economic policy. The best policy instruments are those that are properly subject to governmental control and that produce predictable results. In the final analysis, stabilizing expenditure policy must be judged in terms of these criteria. In a highly significant study for the Commission on Money and Credit, Milton Friedman and David Meiselman tested the predictability of the effects of changes in autonomous expenditures (net private investment and government) upon consumption and income as compared with the predictability of changes in the stock of money upon consumption and income. As Friedman and Meiselman write, “the results are strikingly one-sided.” It turns out that money matters very much. Monetary changes are much more highly correlated with subsequent income and consumption than are autonomous changes in expenditures.

Friedman and Meiselman do not claim their findings to be decisive; there are statistical difficulties in fitting data into appropriate theoretical categories. But the study is highly suggestive and has created some stir. If money, which is subject to control, is a better predictor than autonomous expenditures, which are under only partial control and hard to manipulate, is there really a stabilizing task left for expenditures policy? Friedman and Meiselman have to be taken seriously because they ask the right questions. By trying to resolve what is basically a question of empirical fact, they have helped to lift stabilization policy out of the realm of theology.

Where do we go from here?

Does all this mean we should abandon economic stability as a goal of budget policy? I think not. To do so would constitute another form of “fiscal irresponsibility.” Despite the existence of serious technical and operational constraints on the effective use of discretionary expenditure policy, the economic effects of the budget simply cannot be ignored. At minimum, the budget pressures should work in the right direction. There is no excuse for tolerating the perverse fiscal acts that have sometimes occurred in the past. Moreover, there does exist some modest scope for improving the administrative procedures in order to adapt spending policy somewhat more effectively to changing economic conditions.

There are many unsettled questions in the field of stabilization policy. Most of them are not theoretical issues but questions of empirical fact. In other words, they are questions that eventually may be answered by painstaking research. One of the serious deterrents to progress on this front lies in the fact that too many people, including many economists, have a vested interest in or strong commitment to a particular intellectual apparatus or policy position. Such policy positions are often unrelated to economic realities. They change slowly.

Walter D. Fackler was professor of business economics at Chicago Booth, where he also served as director of the Executive MBA Program. He passed away in 1993.