ABSTRACT
Many of the comments focused on technical aspects whose solution was felt to be highly important for the interpretation of the empirical findings in both papers. One issue with respect to Robert Gordon’s paper was how to identify shifts in the trend of output and 179productivity, because these estimates were essential to identify the output gaps. Some discussants raised doubts as to whether the trend of output, estimated for the 1973–1979 period, should be extrapolated until 1984. They pointed to other studies which identified a downward shift in the trend; the same applied to estimates of labor productivity. If there was indeed such a downward shift, the size of the output gaps estimated in the paper would have an upward bias; in turn, the judgement on the wage gap would be biased downwards because, given the high growth of productivity, the actual behavior of real wages would seem more moderate. The same would be true if the years chosen as benchmarks were exceptional in the sense of being peak years of the business cycle or of the inflationary process. A further issue was whether observed productivity could really be taken as a factor independent of wage behavior; high observed productivity growth might be misleading since it might only reflect a substitution away from labor induced by, for example, excessive wage increases.
