Labor's Share of the Business Cycle: Do Biased Technology Shocks Cause Business Cycles?

Abstract

A recent paper by Young (2004) demonstrated that biased technical changes, in the form of shocks to labor's share/elasticity, can drive economically large fluctuations in a real business cycle (RBC) model. We examine the cyclical properties of 4 quarterly measures of US labor's share from 1959 through 2000. VARs employing M1 growth, inflation, changes in T-bill rates and government spending growth as non-technical/demand-driven instruments attribute between 60 and 88 percent of labor's share change volatility to exogenous shocks. Variance decompositions attribute only between 10 and 34 percent of US GDP volatility to these labor's share shocks. Alternatively, based on growth accounting theory, changes in the real wage rate, the real return to capital and the capital to labor ratio, can account for between 71 and 85 percent of labor's share changes. Variance decompositions controlling for these variables attribute an economically negligible portion of GDP volatility to biased technical changes. Both methodologies estimate a statistically significant negative response of GDP growth to biased technical changes.