The direct effects of OSHA regulations are obtained simply by suppressing the
interaction terms. For the price-cost margin (wages) these direct effects are sizeable and
negative (positive). Allowing for indirect effects, the interaction terms between the
regulation variable and the large firm proxy and a Frost-Belt location are both positive
and well determined, reflecting compliance and enforcement asymmetries. Also as
expected, the advantage of larger firms and a Frost-Belt location are dissipated in the
presence of unionism: estimated at mean coverage, unions gobble up almost one-half of
these regulation-induced rents. Turning to the wage equation, the interaction terms are
again as expected: the coefficients for the firm size and regional interaction terms are
positive and statistically significant. The direct effects of regulation on wages are now
negative and well determined, but there is no indication of successful union dissipation of
advertising rents (as was indicated in the price-cost margin equation).
To determine whether predators gain on net, the authors estimate the relative importance
of the direct effects and indirect effect of regulations at their mean values (and mean
values of the dependent variables). Estimates are provided for minimum, mean, and
maximum values of the large firm and Frost-Belt arguments. For an industry with the
maximum share of workers in establishments with at least 250 workers, the reported net
gain in profits is 2.9 percent; for an industry with the largest percentage of its workers in
the Frost Belt the profits gain is 9.1 the corresponding gains for unionized workers are
3.2 and 3.8 percent, respectively.
The bottom line from this study is that two distinct groups gain from regulation. The
logical inference is that they may be expected to actively support OSHA and that the
indirect effects of regulation are indicative of predation, not innocuous by-products of the
public pursuit of workplace safety (Bartel and Thomas, 1987, p. 241).
(ii) Legislation as a Corrective: The Political Compromise Hypothesis
In an interesting discussion of unjust dismissal legislation in the United States at state
level, Krueger (1991) argues that legislation is an antidote to if not the casuistic rulings of
the American courts then certainly to the (manner of the) attenuation of the common law