The results reported above impose that the effects of all the explanatory variables be the same for all industries (with the exception of the constant term and lagged employment). It is reasonable to think that some of the underlying parameters in the model outlined in section two, such as the industry demand and cost elasticities, to be industry specific. We have attempted to relax this restriction somewhat by splitting the sample between high and low-markup industries and, as seen in Table 1 and 2, and the results do differ. In this section we provide further insights in the differences in the exchange rate effects by industry there.
There are two ways that we attempt to shed light on the industry-specific behavior of adjustment of employment and wages to exchange rate changes. First, we use the estimated exchange rate coefficients in Tables 1 and 2 for the low and high-markup samples and the average export share and imported input share of each industry during the sample period to calculate the industry-specific elasticities of response to an exchange rate shock.
These elasticities are reported in Table 3. Second, we estimated our original wage and employment specifications separately for each industry using its individual data. The estimated coefficients from these regressions are reported in Appendix Tables 2 and 3.
Table 3 shows that the computed labor market response elasticities to an exchange rate shock vary considerably across industries. An exchange rate depreciation results in an increase in both wages (for 16 out of 20 industries) and employment (15 out of 20). in most industries. On average, a 10 percent permanent dollar depreciation will result in an increase in manufacturing wages of 0.4 percent and will increase total employment (number of jobs) by about 0.1 percent.
The effects of an exchange rate depreciation on overtime activity tend to be negative for most industries and its variability appears larger among different industries. However, its overall effect in the overall manufacturing sector seems also small. The average effect on overtime employment in manufacturing from a 10 percent dollar depreciation is around 0.46 percent, while it appears to decrease overtime wages by about 0.2 percent.
The average elasticity of response have changed considerably over time. Average employment elasticities have consistently increased during the sample period, and have changed from initial negative values during the 1970s to increasingly larger positive values. Average wage elasticities have also been in general increasing, except for a short period in the middle of the 1980s.