Structural transformation, the mismeasurement of productivity growth and the cost disease of services

If workers self-select into sectors based upon their relative productivity in different tasks, and comparative advantage is aligned with absolute advantage, then as a sector's employment share increases (decreases) the average efficacy of its workforce will fall (rise). This provides a potential explanation for the differential in the measured productivity growth of contracting goods and expanding services. Using changes in defense expenditures as an exogenous shifter of employment shares, I estimate that the elasticity of worker efficacy with respect to employment shares is substantially negative. While conventional estimates indicate that productivity growth in goods is .8% and 1.4% faster than in services in the US and the OECD, respectively, regression point estimates suggest that the true difference might lie between a .5 percent advantage for goods and a .4 percent advantage for services. Taking the middle of this range, the view that goods and services have similar productivity growth rates provides a plausible alternative characterization of growth in developed economies.