We provide new evidence that industry financial conditions help determine wages in the US manufacturing sector. Ordinary least squares estimates of the effect of rents per worker on wages are significantly positive, but quite small. We show that this may stem from econometric difficulties that plague the OLS estimates. Using the US input–output tables to isolate demand shocks, we overcome these issues and identify the effects of the industry financial situation on wages. Our IV estimates reveal substantial rent sharing—much more than is consistent with a purely competitive labour market.