EMPLOYMENT TAX CREDITS AS A FISCAL POLICY TOOL By GARY C. FETHKE and SAMUEL H. WILLIAMSON 1 1. AN OVERVIEW 1.1. INTRODUCTION. Recent experience with concurrent inflation and unacceptably high levels of unemployment in the United States has provoked skepticism and concern regarding the short-term countercyclical effectiveness of traditional fiscal and monetary policy tools. This dissatisfaction has generated many policy suggestions by prominent economists, politicians, and businessmen on how the Nation can meet the twin goals of full employment and price stability expressed in the Full Employment Act of 1946. Four newer forms of policy which have been explored are: (1) Direct controls or guidelines on wages and prices (some form of permanent incomes' policy); (2) simultaneous deployment of expansionary fiscal policy and tight monetary policy; (3) public service employment; and (4) wider use of tax credits, subsidies, and selective taxes as counter-cyclical fiscal measures. This paper examines and evaluates one form of countercyclical credit policy, known as an employment tax credit. The basic rationale for a universal employment tax credit is straightforward. This program will provide employers tax credits on either wage rates, wage bills, or employment levels. These credits will directly reduce business labor costs without reducing labor incomes. By stimulating the demand for labor and the aggregate supply of goods and services, such a program would initiate increases in real GNP, and thereby increase employment. An employment tax credit program differs from the emphasis of traditional fiscal and monetary policies which act directly to increase aggregate demand, and only indirectly to increase aggregate output. Furthermore, since the direct effect of the credit decreases business labor costs, its use to stimulate employment does not place significant pressure on prices or contribute, via Phillips Curve relationships, to inflationary pressure. In the short-run, the stimulus to employment encourages firms to utilize otherwise idle capacity and therefore moves the economy closer to full employment. The program can be designed such that the level of the tax credit allowed can be varied in response to economic conditions. The universal nature of the program will not disrupt the operation of regional labor markets, and if there is any impact on the distribution of income, it will favor the low income and unemployed. 1 Associate professors at the University of Iowa in business administration and economies, respectively. The authors are indebted to their colleagues Carol Oliven, who read and helped revise much of the paper, and Andrew Policano and Calvin Siebert for their comments. Opponents of employment credit policy insist that the plan will simply grant windfall profits to business, distort the allocation of resources, be difficult to discontinue, and lead to widespread cheating and misrepresentation. Examination of each of these potential shortcomings of employment credits reveals that appropriate administration can overcome them, or that they are exaggerated and of minor practical importance. 1.2. BACKGROUND OF EMPLOYMENT TAX CREDIT POLICY To date, most employment tax credit proposals have been aimed at influencing the employment levels for particular groups or categories of labor, and thus are selective rather than universal in coverage. Selective wage credit programs have been analyzed in a number of contexts: (1) Depressed regions of a developed country (Borts, Lind, SerckHanssen, Archibald); (2) urban areas of developing countries (Hagen); (3) income maintenance (Kesselman, Barth, and Eisner); (4) job training of low-wage workers (Hammermesh); and (5) alternatives to tariff protection (Bhagwau and Ramaswani). Limited employment tax credit programs have been introduced in developed countries, and some experience is available for review. In Britain, the Regional Employment Premium and Selective Employment Tax provide labor incentives by industry and location. In the United States, employment tax credits have been enacted under the job opportunities program, the AFDC work incentive program, and the training incentive payments program in New York City. These categorical programs, however, induce firms to substitute eligible workers for ineligible ones; and their net impact on overall employment, output, and prices is slight. A universal employment credit program was first suggested by Nicholas Kaldor who argued that a wage subsidy would reduce labor cost, increase profit, and generally encourage private enterprise to expand employment. Kaldor also attempted to provide some empirical evidence regarding the net cost to government of a wage bill subsidy. Ragnar Frisch constructed simulations of tax and subsidy programs for the Norwegian economy, and was enthusiastic about possible counter-cyclical implications of wage-bill taxes and subsidies. Most recently, Berndt, Kesselman, and Williamson simulated the impact on U.S. manufacturing of replacing the investment tax credit with an equivalant cost-to-government employment tax credit for the period from 1962 to 1971. Their principal conclusions were that total employment would have been one-half to more than 1 percent higher in many of the years, and that use of capital would have been 1 to 6 percent lower. They also argued that an employment tax credit would induce firms to substitute production workers for capital and non-production workers. Subsequent empirical work by Berndt, using the same data, revealed that the removal of energy price ceilings and investment incentives and the adoption of a 4 percent employment tax credit would have decreased energy use in manufacturing by 5 percent and increased employment by over 2 percent. Elsewhere, we have examined the aggregate impact on employment, output, wages, prices, and net Government revenue of alternative employment tax credit programs (Fethke and Williamson). In the United States, a universal employment tax credit bill was introduced into Congress by Senator Jacob Javits in 1971. Legislation has recently been submitted by Senator Lloyd Bentsen in 1975 and Senator John Tunney in 1976. The concept has been reviewed by the U.S. Department of Labor, the Congressional Budget Office, the Council of Economic Advisers, and the Congressional Research Service of the Library of Congress. 1.3. TYPES OF EMPLOYMENT TAX CREDITS Employment tax credits have been proposed in many forms. Each type has a different impact upon the price of labor services to the firm as well as a different impact upon aggregate supply and demand in the economy. In structuring the credit there are two policy parameters to consider: First, the type and amount of the credit, and second, the level of base employment beyond which the credit is offered. One way of instituting the program is to offer a tax credit as a specified amount per worker employed. A single, universal credit of this form not only reduces labor cost but also lowers the cost of unskilled relative to skilled labor. This follows because a per worker credit represents a higher proportion of wages for low-paid, unskilled workers than for high-wage, skilled workers. This credit presumably stimulates employment of young people, women, and minority workers who comprise the major portion of marginally skilled workers and who represent a disproportionate share of the unemployed during periods of declining business activity. An alternative type of credit is one given on a specified percent of the worker's wage. This form, which is actually a type of negative payroll tax, is neutral as to its effect on skill categories of labor. The second policy parameter is the base level of employment beyond which employers qualify for the credit, and again there are several alternatives. First, a subsidy can be offered on the entire workforce. The advantages of this approach include ease of administration, low monitoring costs, and minimal difficulty in establishing criteria for inclusion. On the other hand, an across-the-board credit may prove expensive relative to its incremental impact on employment. Second, a credit can provide a subsidy for newly hired workers who are drawn from the ranks of the unemployed; that is, a marginal employment tax credit. In this case, the base of the program is the actual (current) level of employment. Finally, a variable base employment credit (VBEC) can be considered which permits a tax credit for workers hired in excess of a base level, with the base to be adjusted to reflect changing business conditions. During periods of rising unemployment, the base can be decreased; then during periods of recovery, it can be adjusted upward; and during boom periods, the credit can be abolished. It is also possible to adjust the base to reflect different employment conditions in regional or sectorial labor markets. Marginal employment tax credits have received the most legislative attention. These programs appear attractive since they reward only net additions to the workforce rather than some percentage of old and new workers. The marginal credit, therefore, does seem less vulnerable to the claim that employment credits simply provide a windfall profit to businesses. A marginal credit is not as flexible a counter-cyclical fiscal policy tool as the VBEC on at least three grounds. First, a marginal credit is less effective in stabilizing existing employment since it does not penalize firms for laying off workers. A VBEC, with the base set below the current employment level, would protect existing jobs, since each worker laid off would reduce a firm's employment tax credit earned. Second, establishment of the base at last period's employment, or at the previous peak of employment, may either nullify any impact the credit may have, or worse, turn the credit into a procyclical measure. This follows because firms will typically ignore a marginal employment credit when employment is declining and rush to accept it when employment is expanding. Third, the more the credit is restricted to particular workers or employers, the less impact the program will have. For these reasons, we do not feel that the program should be limited to presently unemployed workers. A reduction in unemployment will come through an increase in demand for all workers, and our results show that a VBEC is the most effective method of stimulating this demand. For the same reason, we do not feel the program should be restricted as to size or type of employer. A VBEC can be offered to any employer that makes social security contributions, including local governments, schools, and nonprofit organizations, as well as the business sector. In this way, a VBEC would have an even impact throughout the economy and would not distort employment patterns. A difficulty with the VBEC is that administration will require the ability and power to change the level of the base as the economy moves through different phases of business activity, as well as the ability to evaluate employment conditions in heterogeneous labor markets. Given the track record of U.S. fiscal policy, this is no small administrative task. 1.4. APPROACH OF THIS REPORT We prefer a variable base employment credit which provides a percentage reduction in the wage rate. A flexible base program has three desirable features: First, variation of the base can directly stabilize employment over the business cycle; second, the ability of administrators to set alternative bases for different sectors of the economy will assist in eliminating windfall profits; and third, the program's impact on aggregate demand can readily be altered through base and credit rate adjustment. In our analysis, therefore, we select a wage rate subsidy because it appears the least distortive, the most direct, and the easiest to administer. In section 2, the macroeconomic effects of a VBEC are examined; in section 3, the impact of employment credits on the individual firm is considered; finally, section 4 contains a summary and concluding remarks. 2. MACROECONOMIC EFFECTS OF EMPLOYMENT TAX CREDITS 2.1. INTRODUCTION This section describes the impacts of a VBEC program on key aggregate variables in the economy. In particular, we focus on the response of employment, real output, wages, prices, aggregate demand, and net cost to the government. The analysis is based on an evaluation of a multiequation model which is defined and briefly described in the appendix. Section 2.2 presents the analytical results that derive from mathematical analysis of the model. Of primary interest in this section is the specification of the set of conditions under which a VBEC can increase employment and output without increasing aggregate prices. Using the aggregate model and recent U.S. data, section 2.3 presents a set of calculations of the numerical changes in the aggregate variables in response to an employment tax credit. These calculations are reported for different bases and for alternative forms of credit financing. Numerical results highlight the analytical findings, and indicate the likely changes of those variables whose behavior cannot be predicted solely on the basis of an a priori reasoning. The final subsection, section 2.4, compares VBEC's with other fiscal policies. 2.2. ANALYSIS OF THE MACROECONOMIC MODEL1 The model we use differs from simple Keynesian representations in two ways: First, labor supply, even during recessions, is not assumed to be infinitely elastic, but will increase as wages go up. Second, employment, output, prices, wages, et cetera, are described as jointly determined variables. This means there is feedback between labor, commodity, and money markets.2 Mathematical analysis of the model leads us to the following proposition: An increase in a VBEC will expand employment, output, and real wages. These results hold for almost every configuration of personal and business tax rates, levels of unemployment compensation, labor supply responses, technological conditions, and forms of Government financing. The VBEC increases real demand for labor. Assuming some flexibility in the short-run supply of labor, employment expands in response to a rise in the real wage. The greatest rise in employment 1 The derivations on which this section is based are not presented in detail. They will be provided on request by the authors. 2 Simple Keynesian models presume that real output is determined by real expenditure on consumption, investment, and government purchases. Given an infinitely elastic shortrun supply of labor and constant product prices, unemployment is determined as the difference between full employment work force and the actual work force. The actual work force is the level of employment required to produce demand-determined real output. This sequence can be modified to accommodate changing wages and prices by adding a noninstantaneously clearing labor market (Peacock and Williamson). |