The Effect of Gender Differences in Risk Tolerance on the Gender Gap in Wealth Accumulation

This paper uses data of the Health and Retirement Study (HRS) on the asset allocation in IRAs to estimate a distribution of the risk tolerance of men and women separately. It uses an individual portfolio choice model where the asset allocation in IRAs was used as a risky asset. To estimate the distribution of the risk tolerance, the distribution is assumed to be log-normal and a simulation method is used to calculate the corresponding parameter values. Slight but no significant differences in risk tolerance were found between men and women on average. As a consequence, no significant evidence is found to conclude that gender differences in risk tolerance have an impact on the gender gap in accumulated wealth. The paper compares the methods in a longitudinal setting between 2006 and 2014; a period before and after the global financial crisis. No significant changes are found between the risk tolerance distribution of the HRS respondents in 2006 compared to the the risk tolerance distribution of the HRS respondents in 2014. Lastly, the paper adds to literature on imputation methods for the estimation of wealth and income data. It proposes an estimation method for the imputation of missing values using survey questions based on a survey method which uses follow-up brackets. The use of the imputation methods did not effect the conclusions about the distribution of the risk tolerance significantly.