A recent proliferation of theoretical papers on the minimum wage
has made it increasingly difficult to keep things straight. Various theories can be distinguished according to whether or not they assume the existence of an uncovered sector and whether or not unemployment is allowed to
exist, assumptions that drastically affect the outcome. The broad
question is whether the benefits of higher wage rates compensate for the costs of a reduced probability of working; but the comparison can become complicated, as will be seen below. In making statements about utility levels of groups of workers, I have tried to deal with the issue of uncertainty that is, whether the utility of a certain expected income must be downgraded if actual income varies; but I have omitted some
other, nonquantifiable, benefits and costs of minimum-wage legislation.
Most economists allege that these noneconomic considerations set up a prima facie case that increases in the minimum wage are harmful-because they reduce employment, eliminate opportunities for on-the-job training, saddle workers with "involuntary" intermittent job histories, impair the self-respect that comes from having a job, and the like. Yet equally impressive, and equally vague, considerations can be marshaled on the other side. A boost in the minimum wage.
The theory that includes a sector not covered by the minimum wage
but does not take note of the existence of any unemployment has been developed most fully by Finis Welch.2, the equilibrium wage in the absence of a minimum is assumed to be WO in both
sectors. The initiation of a minimum at W, in the covered sector creates
Sc-d, of excess labor there, and this labor is willing to transfer to the uncovered sector. But, given positively sloped supply curves, that much labor will be added to the uncovered sector only if the uncovered wage. As that wage is bid down, the addition to the uncovered supply, SU,
declines by an amount that depends on who gets the covered jobs. (If the covered jobs went to those with the lowest reservation wages, the workers on the covered supply curve to the right of d, would move to the uncovered sector and shift out the supply by an appropriate amount above their reservation wages. If the covered jobs were allocated randomly,
the uncovered supply would shift out in the manner drawn.) more realistic model can be constructed along lines applied to minimum wages by Jacob Mincer. This theory postulates that wages in the uncovered sector would not fall all the way to the level that equates demand and supply, but would remain above that level as those workers who do not have covered jobs prefer to remain unemployed until covered jobs open up.3 In equilibrium the free flow of labor between the two sectors should
operate to equate the utility of a relatively certain but lower-wage job in the uncovered sector with that of a less certain but higher-wage job in the covered sector. Since in equilibrium the utility of jobs in the two sectors is equated, one can determine whether lowage workers are better off as a result of a legislative change simply by measuring changes in the uncovered wage. (To borrow a phrase, that wage forms a "certainty equivalent" to the covered-sector package.
