Reply to Skillman

Gil Skillman’s comments fall into two groups: specific comments on the model and broader questions concerning the fruitfulness and power of our approach in comparison to other approaches. We briefly address the specific points before turning to the some of the interesting and more general issues raised by Skillman. Skillman suggests that our account of the movements in relative wages implies that both highand low-skill workers should have experienced decreasing wages (p.3). We fail to understand this comment. Our argument in section 5 of the paper relates the rise in unemployment to slow adjustments in wage aspirations, following the observed decline in the growth rates of labor productivity and average real wages. This argument is compatible with rising or falling real wage rates. What matters in our account is the ratio Am/B, and this ratio may clearly decline even if A has increased. Our model predicts widening wage inequality in the US but the average growth rate of wages depends on the pace of technical progress, that is, on the growth rate of A.1 It is correct, secondly, that not all countries experienced the pronounced increase in wage inequality that characterized the United States or the United Kingdom. Germany, as pointed out by Skillman, had "declining income inequality along with increased unemployment" (p.3). Our numerical examples, however, included Germany. Moreover, using plausible parameter values the simple model in section 4 was consistent with the patterns of employment and wages in both the US and Germany. Thus, it is unclear how the German example can indicate a failure of our model to predict "the empirical connection between unemployment and wage trends in OECD countries" (p. 3). The rise in unemployment for high-skill workers in the US also casts doubt on our account, Skillman suggests (p. 3). But again, our numerical example in section 4 is based on US data that incorporates this trend increase.