rate will increase consumption expenditure, and therefore stimulate aggregate demand. The increase in aggregate demand, and more specifically aggregate prices, leads to higher wages, employment, and real output. Through the multiplier process, larger income leads to higher levels of aggregate demand and further growth in employment. The entire scenario depends upon the initial increase in prices required to stimulate labor demand. Labor supply and productive capacity are presumed to be sufficiently flexible to accommodate these increases in demand.

These are familiar results of standard aggregate analysis, and they describe a major disadvantage of a reduction in personal taxes as compared to an increase in the employment tax credit rate under similar economic conditions. With a VBEC, it is possible to directly expand employment and real output without first stimulating an increase in aggregate demand and prices. Thus, employment tax credit programs possess desirable features especially for those periods, now commonly experienced by advanced industrial nations, when prices and unemployment are concurrently rising.

A supposed disadvantage of the VBEČ, as compared to personal income taxes, is that employment credits will cause distortions in interfactor allocation of resources, favoring the hiring of workers over capital. Thus, employment credits will presumably be less neutral than personal income taxes with respect to their relative impact on the allocation of resources. In our opinion, the temporary nature of a VBEC will stimulate intertemporal reallocation of employment, rather than permanent substitution of labor for capital. The subsidy will encourage firms to maintain employment and possibly add to their work forces during slack periods. When the economy expands, the base can be increased and eventually the credit eliminated.

B. Payroll Tares

If there is any merit to an argument that selective taxes and credits effect the long-term allocation of resources, it surely applies more to payroll taxes which add over $100 billion annually to the cost of labor. In the short run, however, an increase in payroll taxes is analytically similar to a reduction in the VBEC. Both actions will increase the cost of labor and contribute to downward pressure on employment.

Our model predicts that the currently proposed policy of increasing payroll taxes, while simultaneously reducing personal income tax rates, will precipitate a decline in employment and an increase in the price level. Basically, this proposal is the exact opposite of the numerical calculation presented as Case 3. Given a reduction in the income tax rate with a negative VBEC (that is, an increase in payroll taxes), the signs in Case 3 will all be reversed.

In our model, personal income taxes directly affect disposable income and government revenue; they do not directly affect either labor demand or labor supply. See the appendix. In a purely classical labor market, changes in the level of aggregate demand would have no effect on employment and real output.

This argument has been offered in a recent Council of Economic Advisor's report which is critical of job credit programs.

A short-run increase in labor costs resulting from higher payroll taxes will encourage a reduction in the demand for labor. The stimulus to aggregate demand provided by lower personal income taxes may not be sufficient to offset the payroll tax promoted decline in employment, and will contribute to rising prices. Thus, a Government program intended to stimulate employment while simultaneously increasing payroll taxes may instead contribute to unemployment and rising prices. At the very least, the Government should consider the timing of payroll tax changes and attempt to avoid increasing social security and unemployment contribution taxes during periods of declining economic activity. We will say more about payroll taxes in section 3.3.

10

C. Investment Tax Credit

The investment tax credit reduces the price of capital goods and thereby stimulates gross investment. Thus, an investment tax credit seems to offer the same countercyclical advantages as the VBEC.11 Compared to a VBEC, however, there are major drawbacks to using an investment tax credit as a short-run stabilization tool.

The critical parameters for assessing the effects of credit policy on capital and employment are price elasticities of demand and supply for labor and capital, and the rates at which the use of these inputs can be expected to change in response to changes in input costs. (Picou and Waud.) Empirical evidence reveals that while price elasticities of demand for labor and capital are of roughly similar magnitude, rates of input adjustments are not. In particular, capital adjusts to changes in relative prices, or to changes in output, at a much slower rate that does employment 12 (Rosen and Nadiri, Hickman and Coen, Picou and Waud.)

Therefore, the response by firms to a VBEC will be more rapid and more complete than their response to an equivalent investment tax credit. For example, it has been estimated that only 50 percent of a once-for-all change in the relative price of capital will be absorbed after five years; in contrast, nearly 80 percent of the adjustment between actual and desired labor will be accounted for in the first year.13 Thus, the immediate impact of the VBEC on real output is probably high, while the immediate impact of the investment tax credit is slight.

An employment tax credit will stimulate many areas of the economy where investment tax credits have little or no effect. For example, such labor intensive industries as services, wholesale and retail trade have relatively small capital stock but employ twice as many as the goods-producing sectors.

10 The pro-cyclical aspects of unemployment taxes are well-known. This tax, however, is a small and declining portion of the wage bill while the relative importance of social security taxes is accelerating.

11 Theoretical support for selective investment tax credits and accelerated depreciation was provided by Robert Hall and Dale Jorgenson; and their work, which relies on neoclassical theory of capital accumulation, remains with only minor modification the basic justification. See Hall and Jorgenson.

In our model, we assume that capital stock is fixed and that investment has no immediate impact on aggregate supply. Thus, the only current effect that an increase in the investnient tax credit has in the model is to increase aggregate demand. Even this impact will be mitigated if the government reduces purchases as a way to finance the credit.

13 Coen and Hickman (table 5, p. 297).

Many proponents of investment tax credits defend them not as a short-run fiscal tool, but for their growth implications; however, in an era when economists and conservationists are questioning the longterm emphasis on capital and resource-using technologies, public policy aimed at increasing capital stock rather than directly stimulating employment may be misguided. We recognize the need for new capital for economic growth. Permanent investment tax credits, accelerated depreciation schemes, depletion allowances, as well as continued increases in payroll taxes (the fastest growing source of Federal revenue), act in the long run to encourage increases in the capital/labor ratio. Recent empirical evidence reveals the elasticity of substitution between capital and production labor, and between energy resources and labor to be significantly positive. As an economy develops it will continue to try and reduce the cost of its most expensive inputs; therefore, a policy to stimulate only investment may contribute to lower long-run employment as firms substitute toward capital-intensive (and energy-using) techniques of production.

13

3. THE MICROECONOMIC EFFECTS OF AN EMPLOYMENT TAX CREDIT

3.1. INTRODUCTION

The primary purpose of an employment tax credit is to stabilize and expand employment by encouraging firms to retain present employees and to hire additional workers. Therefore, determination of labor demand and supply conditions is important in predicting the impact of a VBEC. In developing the aggregate model in section 2, some specific assumptions are made about the behavior of aggregate demand and supply of labor as well as the prescribed employment base. In section 3.2, the implications of these assumptions are examined at the level of the individual firm. In section 3.3, we discuss two commonly cited aspects of labor markets which are not directly incorporated in our aggregate model. The first of these concerns the impact of the VBEC on the length of the workweek. The second involves the effects of a VBEC when there is labor hoarding; that is, when labor is treated as a quasi-fixed input possessing attributes similar to those of capital. Finally, section 3.4 addresses the question of administration and scope of the credit.

3.2. THE RESPONSE OF THE INDIVIDUAL FIRM TO A VBEC

Under competitive conditions, the quantity of labor demanded by a business firm is determined by wages, other labor costs, and the value of the product labor produces. A reduction in the price of labor, ceteris paribus, increases the quantity of labor demanded as long as the net contribution to profit of the additional employment is positive.

This is illustrated in figure 3.1. The short-run equilibrium level of employment, L1, is determined where the market wage rate, W, equals the demand for labor, D. At this point, the cost of an additional worker, W, just equals that marginal worker's contribution to the value of the firm's product.

Introduction of a VBEC can encourage the individual firm to maintain or even increase employment when there is a decline in the demand for labor. For example, if the price of output falls, then the value of each worker's output is lower at all wage rates (the demand for labor declines to D2). Under these conditions, employment is ordinarily reduced to L2. If the firm receives a credit on the wage rate, however, employment need not decline. If the credit reduces

1 The demand curve for labor, known as the value of the marginal product of labor, is found by multiplying the price of the firm's product by the incremental output contribution of each worker. Short-run demand for labor is presumed to be downward sloping because of diminishing incremental returns to labor.

the wage rate paid by the firm by s percent of the wage (to W(1-8)), then employment remains at L1, rather than declining to L2. Larger credits will induce employment increases beyond L1.

It is not necessary to credit the wages of all workers in order to maintain employment at L. In particular, whether the firm can be induced to continue employment at L depends on the prescribed base of the program, L. For all bases from zero to L2, (0>L>L2), the credit will be accepted and employment will remain at L1. This follows because for employment rates from zero to LB, the firm pays W, which is below the value of labor's marginal product. Once Lg is reached, the effective cost of labor drops to W(1-8), and employment will continue to be profitable until L is reached.

If the base is greater than L2, however, the credit may or may not be accepted. As seen in figure 3.2, for LB>L2, the wage of LB-L2 workers will exceed the value of these workers' contribution to output, by Area I. The question then becomes whether or not this loss in profit is offset by the gain attributable to the employment of creditsubsidized workers.

The maximum base under which the firm will accept the credit, and still be willing to hire L1 workers, is where the profit lost by paying LB-L2 workers a wage which exceeds the value of their added product is just equal to the gain in profit attributable to employment L1-LB. If area I equals area II, then L is that base.