Bargaining power, demand growth and the decline of the labor share

I develop a Keynesian growth model where conflict over income distribution determines the labor share. Changes in workers’ bargaining power can either raise or depress output, depending on their effects on aggregate demand. An increase in demand reduces slack in the labor market, leading to a higher labor share. I incorporate the empirical implications of the model into a structural vector autoregression (SVAR) with sign restrictions to assess whether shocks to workers’ bargaining power raise or depress output, and to decompose the long-run decline in the labor share into components driven by demand, technology, and bargaining power. I find that an increase in labor’s bargaining power is contractionary in the short run, accounting for a quarter of output fluctuations. Demand shocks explain over 60% of the long-run decline in the labor share, with changes in labor’s bargaining power accounting for the remainder.