We re-examine the evidence presented by Neumark and Wascher (1992) on
the employment effect of the minimum wage. We find three critical flaws in
their analysis. First, the school enrollment variable that plays a pivotal
role in their specifications is derived on the false assumption that
teenagers either work or attend school. Measurement error biases
contaminate all the empirical estimates that use this enrollment variable.
Second, Neumark and Wascher measure the effect of the minimum wage by a
coverage-weighted relative minimum wage index. This variable is negatively
correlated with average teenage wages. Taken literally, their results show
that a rise in the coverage-weighted relative minimum wage lowers teenage
wages. Examining the direct effects of state-specific minimum wages, we
find that increases in state minimum wages raise average teenage wages but
have essentially no employment effects.
Finally, a careful analysis of Neumark and Wascher's data shows that
subminimum wage provisions are rarely used. This casts doubt on their
claim that subminimum provisions blunt any disemployment effect of the
minimum wage.
Neumark and Wascher contend that other minimum wage studies are biased
by failing to control for school enrollment, and by failing to consider the
lagged effects of minimum wages. We re-analyze the experiences of
individual states following the April 1990 increase in the Federal minimum
wage, allowing for a full year lag in the effect of the law and controlling
for changes in (properly measured) enrollment rates. Contrary to their
claims, allowing for lagged effects and controlling for enrollment status
actually strengthens the conclusion that the 1990 increase in the Federal
minimum had no adverse employment effect.
minimum wages
On April 1, 1992 New Jersey's minimum wage rose from $4.25 to $5.05 per hour. To evaluate the impact of the new law we surveyed over 400 fast food restaurants in New Jersey and Pennsylvania before and after the rise in the minimum. Comparisons of the changes in wages, employment, and prices at stores in New Jersey relative to stores in Pennsylvania (where the minimum remained constant at $4.25 per hour) provide simple robust estimates of the effect of the increased minimum wage. Our empirical findings challenge the conventional notion that a rise in the minimum causes employment to decline. Relative to stores in Pennsylvania, fast food restaurants in New Jersey increased employment by 2.5 employees per store. We also compare employment changes at stores in New Jersey that were initially paying $5.00 per hour or more (and were therefore largely unaffected by the new law) to the employment changes at lower-wage stores, where the new law raised wages by 10-15 percent. Stores that were unaffected by the minimum wage had the same employment growth as stores in Pennsylvania, while stores that had to increase their wages increased their employment. Finally, we evaluate theoretical models that might explain these results.
A Case Study of the Fast Food Industry in New Jersey and Pennsylvania
A Case Study of the Fast Food Industry in New Jersey and Pennsylvania
A Case Study of the Fast Food Industry in New Jersey and Pennsylvania
Throughout the post—war period, U.S. and Canadian unemployment rates
moved in tandem, but this historical link apparently ended in 1982.
During the past three years, Canadian unemployment rates have been some
three percentage points higher than their U.S. analogues, and this gap
shows no sign of diminishing. This paper is an empirical evaluation of
a variety of explanations for this new unemployment gap.
We first show that the demographic and industrial composition of
the two countries is remarkably similar, so that no simple mechanical
hypothesis explain the basic puzzle. It is also evident that the
increase in Canadian unemployment relative to U.S. unemployment cannot
be fully attributed to output movements. We find that the gap between
actual and predicted Canadian output, based on U.S. output, has fallen
dramatically since 1982 while the unemployment gap has widened. We also
find that unemployment in Canada was 2 to 3 percentage points higher in
1983 and 1984 than predicted by Canadian output.
We have investigated a variety of hypotheses to explain the slow
growth of employment in Canada after 1982. These hypotheses attribute
the slow growth of employment to rigidities in the labor market that
raise employers’ costs and restrict the flow of workers between
sectors. The evidence does not support the notion that the growth in
relative unemployment in Canada is due to differences in the regulation
of the labor market in the two countries. Minimum wage laws and
unemployment benefits are fairly similar in Canada and the U.S., and
neither has changed relative to the other in the last decade.
Unionization rates have increased in Canada relative to U.S. since
1970. Most of this divergence occurred before 1980, however, and does
not seem to have created an unemployment gap prior to 1980. Finally,
the hypothesis that differential real wage rates are a major determinant
of relative employment in the U.S. and Canada is soundly rejected by the
data. Real wage rates have been essentially uncorrelated with employ-
ment movements within each country and between the two countries.
The imposition of a national wage standard sets up a useful natural
experiment in which the "treatment effect" varies across states depending
on the fraction of workers earning less than the new minimum. I use this
idea to evaluate the effect of the April 1990 increase in the Federal
minimum wage on teenage wages, employment, and school enrollment.
Interstate variation in teenage wages was high at the end of the 1980s, in
part because 16 states had enacted state-specific minimums above the
prevailing Federal rate. Comparisons of grouped and individual state data
confirm that the rise in the minimum wage significantly increased teenage
wages. There is no evidence of corresponding losses in teenage employment
or changes in teenage school enrollment.