Ilw Liiboi- Mmlirt

safe employees firom the government-designated groups - individuals with a college educauon or subsiandal work experience - and to reduce the demand for those lacking such education and experienced.

Comparing the average incomes of minorides as a percentage of white incomes before and after quotas, we fmd that Puerto Ricans and Mexicans actually lost ground, while blacks stayed about the same despite a seeming abatement of racial prejudice everywhere. Many factors could be influencing these results, but at the very least, the gross evidence shows no posidve benefit from affirmative acdon policies and often seems to show harm,

Moving from the labor to the product markets, non-merit hiring policies injure consumers. Under such pohcies, both firms average total and mai-ginal costs are higher than would occur imder merit hiring. After all, non-merit hiring means discriminadng against the producdvely best person for the job in favor of the racially best person. The immediate and. obvious effect is the misuse of the exisdng productive skills of both the favored and the disfavored group. More subdy, racially based hiring diminishes the incentive of both groups to develop and maintain human capital. After all, hcunan capital is produced through work, effort, and investment It pays to make such investments when they are rewarded. Race, on the other hand, is inherited rather than produced. To the extent that race is subsututed for merit as a hiring criterion, the incentive to develop and maintain human capital is undermined. As such, less can be produced. Thus, in the product market under quotas, supply is decreased, prices are higher, and the quandty sold is lower.

INTERNAnONAL INCOME CQMPARISONS

Surprisingly enough, the degree of Income

inequality is not much affected by the t)pe of economic system. Table II shows the share of income going the various households ranging from the poorest to the wealthiest in the United States, Sweden, and the former Soviet Union. For example, the poorest 20% of American households received about 7% of all income after taxes had been deducted and government transfer payments such as Social Securit) and welfare had been added. On the other hand, the wealthiest 20% of American households received about 40% of all such personal income. One might expect that Sweden, a small country with ambitious welfare programs and a drastically less diverse populadon than the United States, would show far greater income equality. Similarly, after over a half-century of communism, one might expect considerable income equality in the former Soviet Union. Yet, when we look at the facts, all three countries have roughly equal distributions of income despite radical differences in government efforts to achieve it.

IMPERFECT LABOR MARKETS

As weve mentioned, the perfecdy competitive labor market requires that all parties be price-takers. This condition is violated on the supply side under unionization and on the demand side if only a single firm buys labor

MONOPSONY MARGINAL FACTOR COST

The latter situation is termed monopsony and is perhaps best characterized by a single large emplo)"-er in an extremely isolated community. Because the monopsonist is the only buyer of the communit)s labor, he faces the entire upvvard sloping supply curve of that community. Hence, in Figure VW, to hire N workers the monopsonist must offer the wage to increase employment by one worker to N - 1 , thej

Table II Income Shares in Three Countries

Data are rounded to the nearest percent

Source: L. Galloway. -Folklore of income Distribution. in Government Controls and the Free Marfcel

1

monopsonist must offer the higher wage, W. If this higher wage is to be paid to all employees, notjust the added worker, the increase in the total cost caused by employing another worker, i.e., the marginal factor cost of labor (MFC]), will exceed the new wage rate, W. Thus, in Figure VII the monopsony MFC] curve rises above the labor supply curve. Summarizing, if the monopsonist wants to increa.se employment from N to N + 1, his costs will rise not byW, the new wage, but by the higher amount of the marginal factor cost, MFC. We can estabhsh this result both numerically andthrough the marginal/average rules.

To see this point numerically consider Table III Here we have displayed an upward sloping labor sup-L ply schedule, so that hiring 3 workers requires a wage of $5.00, but hiring four requires a wage of $5.50. The total factor cost of labor (TFCj) is the wage rate dmes the amount of employment, i.e., W x N. For example, the total factor cost of hiring four workers is $22.00 (4 X $5.50). Suppose we want to find out how much this employers costs rise as a result of increasing employment from three to four workers, i.e., we want to calculate the marginal factor cost of labor. This cost ulcrease is composed of two parts. First, were hiring one more worker and paying him $5.50. In addidon, were also paying each of the first three workers 50 more, so the cost of employing them rises by $1.50 (50C X 3). The total increase in cost comes to $7.00 ($5.50 -F $1.50). Do you see that the MFC of the 5th worker represents the sum of $6.00 and $2.00? Note that in both cases the marginal factor cost (MFCj) exceeds the wage rate (W) as in Figure VlI. While this. method of calculating the MFC] highlights the relationship between the marginal factor cost of labor and

the wage rate, it is not computationally efficient. We could also derive the marginal factor cost of labor simply as the change in the total factor cost. As a result, the marginal factor cost of hiring the fourth worker is just $7 ($22 - $15); that of die fifth is $8 ($30 -$22).

The supply curve for labor can be thought of as the average factor cost curve of labor or, in other words, the total factor cost of labor divided by the number of workers hired (AFCj = TFCj/N). But die total factor cost of labor is the wage rate times the number of workers (W x N). Therefore, substituting for TFC] in the expression above and cancelling the N, it follows diat AFC = (TFC/N) - (W x N)/N = W. But firom the marginal/average rules, when this average function, the wage W, is rising, its associated marginal function, the marginal factor cost of labor (MFC]), lies above it as in Figure VII.

Returning momentarily to the competitive firms labor supply cunC back in Figure III, its easy to see that the marginal factor cost of labor for the competitive firm is equal to the wage rate. Since the average factor cost of labor (AFCj) is constant, the marginal factor cost of labor (MFCj) must equal it according to the marginal/average rules. Numerically, if the wage is constant at $5, the change in the total factor cost of labor from hiring an added worker must likewise equal $5. For example, suppose a price-taking firm is hiring six workers at a wage rate of $5 per hour, then its wage bill will be $30. Ifit hires a seventh worker, its wage bill will rise to $35, so that the change in the total factor cost of labor will be $5 ($35 - $30). Thus, the marginal factor cost of labor isjust the $5 wage in this case. All of this goes to show that a firms marginal factor cost of labor curve rises above its labor supply curve

Table I Derivation of the Marginal Revenue Product (Demand Curve) of Labor

(1) Number of Workers

0 1 2 3 4 5 6

(2) Wage Rate (W)

$0 3.00 5.00 13.33 20.00 30.00 40.00

Total Factor Cost ot Labor (TFCi)

Marginal Factor Cost of Labor (MFC,)

Marginal Revenue Product (MRPi)

only when its supply slopes upward . When a firms labor supply curve is perfecdy elastic, as it is for a compedtive buyer of labor, AFCi, = MFCi, = W.

monopsony EQUILIBRIUM

A profit maximizing monopsonist continues to hire labor until the MRPj just equals the MFCj. Consider Table FV. Here we have placed an upward sloping labor supply schedule in columns (1) and (2) alongside the MRP] schedule taken from Table 1. Should the firm hire the first worker? Of course, it should; from column (4) the first worker adds only $3 to the firms costs, (MFCj = $3) but from column 5, he adds $50 to its revenue (MRP] = $50). Similarly, the second worker adds only $7 to the firms costs, but $40 to its revenue, so he too should be hired. The third worker adds $30 to the firms costs and also $30 to its revenue (MFCj = MRPj = $30), so the firm would be indifferent to hiring him. But hiring a fourth worker at a wage of $20 adds $40 to costs but conuibutes only $20 to revenue. Hence the firm would cease hiring at three workers. In other words, the profit maximizing monopsonist will hire until his marginal revenue product of labor equals his marginal factor cost, i.e., until MRPi = MFC 1.

Now what will the equilibrium wage be? Clearly, it will not be the marginal revenue product of $30.00. Otherwise excess supply would occur: the firm only wants three workers but if it paid $30.00 per worker, five workers from column (1) would apply, instead, the firm will pay the lowest possible wage consistent with hiring three workers. From column (3) that wage is $13.33. The difference between equilibrium marginal revenue product and the wage rate is

termed exploitation by economists. Here exploitation is $16.67 ($30.00-$13.33). Exploitation is entirely consistent with very high wages. For example, the National Football League functions as a monopsonist even though the salaries and bonuses paid to football players are immense.

In Figure VIII the monopsonist finds that his MRP exceeds his MFC] until N„1 workers are hired. Therefore N is the equihbrium employment aiid requires the monopsonist to pay wages of W. But as we saw in Secdon III, the competitive level of employment occurs where the labor demand curve (the marginal revenue product curve) strikes the labor supply curve. Hence, in the competitive case, wages and employment are the larger amounts, Wc and Nc .

All of this makes intuitive sense. Unlike com-