WAGE ADJUSTMENT: The Data and Regression Specification 2

Posted by Kathryn Schwartz on September 03, 2014

The total hours series is the product of the employment series and the reported average weekly hours in each industry. Our measure of wages per employee is the average hourly wage in each industry, constructed by dividing the total of wage and salary accruals to all employees in each industry by the number of non-farm employees. Overtime wage is defined as the difference between total average hourly wages and average hourly wages excluding overtime for production and non-supervisory workers as defined by the Bureau of Labor Statistics. Overtime hours are the average weekly overtime hours of production workers.

As detailed in Appendix A, Table 1, there is a fair amount of dispersion of wages and employment across industries and within industries over time. Wages are less variable within industries over time than across industries at points in time. The variability of wages across industries is almost four times higher than the average variability overtime of wages within each industry. Wage variability across industries is also higher than employment variability. Finally, industry overtime wages and employment are considerably more volatile than overall wages and employment.

Wage and employment variability across industries are also related to observable industry and worker characteristics. Wage variability across industries is positively correlated with the skill intensity of its workers, industry unionization rates, and industry capital intensity. High unionization rates are associated with higher wage and employment variability (both total and overtime) while industries with higher price-over-cost markups tend to have lower employment and wage variability.

Finally, in our analysis we decompose exchange rate movements into their permanent versus transitory components to capture our expectation that regular labor demand should be most responsive to permanent movements in exchange rates. We use the decomposition first suggested by Beveridge and Nelson (1981) and later employed to exchange rate data by Huizinga (1987), Campbell and Clarida (1987), Cumby and Huizinga (1990), and Clarida and Gali (1994), among others.

The procedure, described in the appendix, decomposes the real exchange rate series into a stationary component (the temporary component) and a nonstationary component (its permanent component). An analysis of the resulting variance decomposition shows that the temporary component of exchange rate changes accounts for only a small proportion of the variance of the real exchange rate series. The variance of the transitory component of the real exchange rate accounts for less than 40 percent of the total variability of the real exchange rates. Electronic Payday Loans Online

In this section we present the results of several types of regressions.

Tags: , ,