Evidence of Induced Innovation in US Sectoral Capital’s Shares

We use annual data on capital's share and relative factor prices from 35 US industries from 1960 to 2005 to test the induced innovation hypothesis. We derive, from a productionfunction framework, testable implications for the effect of contemporaneous and lagged factor price ratios on capital's share of production. The predicted effect is positive or negativedepending on the elasticity of substitution between labor and capital. From panel regressions, the estimated effect of the contemporaneous factor price ratio implies an elasticity of substitution that is less than unity, consistent with the consensus from the literature. Based on this, our negative estimated effects for lagged price ratios are both statistically significant and consistent with the induced innovation hypothesis