How well do people share risk? Standard risk-sharing regressions assume that any variation in households’ risk preferences is uncorrelated with variation in the cyclicality of income. I combine administrative and survey data to show that this assumption is questionable: Risk-tolerant workers hold jobs where earnings carry more aggregate risk. The correlation makes risk-sharing regressions in the previous literature too pessimistic. I derive techniques that eliminate the bias, apply them to U.S. data, and find that the effect of idiosyncratic income shocks on consumption is practically small and statistically difficult to distinguish from zero.

This appendix contains seven sections. Section A reports results from running regressions of labor earnings on GDP using data from the PSID, for comparison with the results using HRS data in the body of the paper. Section B examines the relationship between family income, aggregate shocks, and risk preferences in the PSID. Section C gives technical details on the Markov Chain Monte Carlo estimation employed in table 1 of the paper and reports the complete parameter estimates for the regressions summarized in that table. Section D reports results when the relationship between earnings and aggregate shocks is estimated with individual-specific coecients rather than common coecients for each risk-tolerance group. Section E reports results comparable to table 1 of the paper and table D.1 of this appendix using only Social Security covered earnings instead of the combination of Social Security and W-2 earnings. Section F reports robustness checks for tables 2 and 3 of the paper under alternative definitions of the household and the consumption and income variables. Section G reports robustness checks for tables 2 and 3 under an alternative definition of the leisure variable.