What are the gender and race/ethnicity differences in pre-retirement withdrawals?Background: In part one of this “Research Minute” series, I looked at the ratio of account balances to salary by age, gender and race/ethnicity for a synthetic universe of defined contribution participants. The CFERS initiative attempts to analyze these differentials by looking at gender and race/ethnicity differences in the various components that drive the account balances (contributions, asset allocation, loan activity and pre-retirement withdrawals).In part two of the series, I focus on the pre-retirement withdrawals. Part three will use this information to investigate how treatment of pre-retirement withdrawals would help to mitigate the disparities seen in part one.Findings: The data for this analysis comes from the Collaborative for Equitable Retirement Savings. I compute the annual probability of a withdrawal prior to age 59for active participants with at least 15 years of tenure with the current employer.While the relative rankings of gender and race/ethnicity combinations vary by age, as seen in Figure 1, overall it appears that Hispanic females and Asian males have the lowest pre-retirement withdrawal activity with an overall differential relative to the mean of -34% and -33%, respectively. White females (-16%) and Asian females (-11%) also exhibit pre-retirement withdrawal activity lower than the overall mean whereas white males (+3%) and Hispanic males (-4%) appear to have close to average propensity in this regard.The findings in Figure 1 suggest a very different situation for Black participants, however. Black males and Black females have the highest pre-retirement withdrawal activity with an overall differential relative to the mean of +71% and +85%, respectively.It is also important to focus on what percentage of the existing account balance is removed in a pre-retirement withdrawal. Additional analysis (not shown here) indicates that Black males take out a percentage that is 50% greater than the mean while Black females have an average percentage 10% above the overall mean.* Synthetic universe. The sample filters out participants with a salary of less than $10,000.Source: Morningstar Center for Retirement and Policy StudiesBottom Line: It is clear that the ability to take pre-retirement withdrawals from a defined contribution plan may be extremely important in some situations. However, preliminary analysis of the CFERS data suggests that gender and race/ethnicity differentials in the annual probability of taking such a withdrawal is highly correlated with the observed disparities in the overall account balance to salary ratios.An obvious follow-up question to this analysis is: how much difference would "treatment" of pre-retirement withdrawals make with respect to mitigating gender and race/ethnicity disparities in account balances? Tune in next week as we provide some preliminary evidence in part three of this series!If you are interested in participating or would like additional information, please contact me at: jack.vanderhei@morningstar.com.