IMF Working Papers

This paper theoretically and empirically investigates the role of spousal labor in buffering transitory shocks to husbands' earnings. To measure the amount of the shock that spousal labor absorbs, an instrumented cross-sectional variance decomposition is developed. Using data from the Panel Study of Income Dynamics, the paper finds that the smoothing resulting from the wives' labor response (both labor force participation and hours of work) is larger for households with limited access to credit. This finding, which is consistent with the model's prediction, indicates that because of the presence of liquidity constraints, the temporal change in family income (exclusive of wives' earnings) reinforces the substitution effect in explaining the effect of shocks to the husbands' earnings on spousal labor.