Job flows and productivity dynamics: evidence from U.S. manufacturing

Through their influence on the cross-sectional distribution of productivity across firms and workers, job creation and destruction likely have an impact on the rate at which aggregate productivity changes over time. However, the nature of this effect is not, a priori, clear. While a broad consensus has emerged suggesting that job destruction enhances productivity by eliminating inefficient production units, theories disagree with regard to the effect of job creation. In particular, 'vintage-capital' theories of creative destruction suggest a positive influence since job flows are conjectured to represent the reallocation of labor from low- to high-productivity positions. Others suggest that job creation may, instead, represent the expansion of employment primarily along a lowskill (or low 'match-quality') dimension. In such a case, job creation would serve to lower aggregate productivity. This paper estimates the influence of job creation and destruction on total factor productivity (TFP) growth using annual data on 389 4- digit U.S. manufacturing industries over the period 1974-1993. As expected, the results reveal a positive association between job destruction and changes in TFP. Yet, they also indicate that, contrary to the creative-destruction view, job creation tends to have a negative effect on productivity growth.