UC Davis

The ‘China shock’ operated in part through the housing market, which is one reason why its impact was so large on the U.S. labor market. We add housing to a multi-region multi-sector model, with individuals choosing whether and where to work. Controlling for housing reduces the negative direct impact of import exposure on manufacturing employment by 26%–31%, with a significant magnification through the housing market. A variance decomposition analysis shows that the indirect effect of the China shock through the housing market contributes an added 11% to the variance in manufacturing employment as compared to the direct effect of import exposure. For manufacturing and construction employment combined, the indirect effect even outweighs the direct effect.