William Davidson Institute Working Paper 479
employment, respectively). At that point further increases in the minimum wage reduce
employment. At the highest level of toughness observed in Costa Rica over this period, a one
percentage point increase lowers total employment by 0.6 percent and covered sector employment
by 0.4 percent.
These findings may be interpreted as supporting the monopsonistic model, which predicts that
increases in wages can increase employment up to the point where the marginal cost is equal to the
marginal revenue. Any increases above that point force the employer to be on his/her demand
curve. However, it is difficult to imagine a labour market for unskilled workers as functioning as in
the monopsonistic model. Hence, another plausible interpretation of these findings is that they
support the traditional model of the labour market for less skilled workers and the monopsonistic or
efficiency wage models for the skilled workers. In particular, industries that use unskilled labour
and pay low wages have a high ratio of minimum to actual wage (i.e., high toughness). Firms in
these industries face an elastic supply of labour and when a rise in minimum wage forces these
firms to move up their demand for labour curves and lay off workers. This is consistent with the
negative employment effect documented in Tables 3 and 4. On the other hand, industries that
employ relatively more skilled labour (and where capital intensity may be higher), have a low ratio
of the minimum to the average wage (i.e., low toughness). These are also the industries that are
likely to face upward sloping supply curve of labour or where employers may pay efficiency wages
so as to attract and keep skilled workers. In these industries an increase in the minimum wage may
be expected to raise employment under the monopsonistic or efficiency wage scenario. This is
consistent with the positive effect of increasing minimum wage on employment found for these
industries in Tables 3 and 4.9