COVID-19 Related Job Loss and Early Withdrawals from Retirement Accounts: Mediation By Financial Hardships and Subjective Financial Wellbeing

Background and Purpose: The COVID-19 pandemic has exerted tremendous influence on households’ financial stability. As of July 2020, 31.3 million people had experienced pandemic-related job loss in the prior 4 weeks, representing 12 percent of the noninstitutionalized population. Meanwhile, more than half a million retirement account withdrawals were made at Fidelity, and about 2% of retirement plan participants at Vanguard withdrew a median of $10,413 from their accounts. Indeed, there are legitimate and compelling motives to make early withdrawals, such as to ease the impact of job and income losses. However, when these withdrawals are projected over one’s lifetime, they can then turn into a significant reduction in retirement savings. The purpose of this study is to understand mechanisms that lure households into tapping into their retirement assets during a major event or when making investment decisions and the indirect roles of financial hardship, financial well-being, and financial literacy in this relationship.

Methods: Data for this study came from the Socio-Economic Impacts of COVID-19 Survey fielded from April 27, 2020, to May 12, 2020, using a nationally representative sample of 4,765 respondents. Job loss was measured as a dichotomous variable (0 = No, 1 = Yes). Financial hardship was measured with number of hardships in the past three months and financial well-being was self-rated on a 20-point scale, with higher score indicating worse financial well-being. Retirement account withdrawals were measured as a dichotomous variable (0 = No, 1 = Yes). To correct for endogeneity, we used generalized boosted regression modeling (GBM) to estimate propensity scores to balance the groups who did and did not experience a COVID-19 related job/income loss. Stabilized weights, constructed based on propensity scores to adjust for potential mediation effects, were applied in path models.

Results: About 19% survey participants reported losing jobs due to COVID-19 and 10% reported making withdrawals. Households experiencing COVID-19 job losses ( = .048*) were more likely to make early withdrawals from retirement accounts. The effect of job loss on retirement account was partially transmitted by financial hardships ( = .035***) and financial wellbeing ( = .008**). The moderation effect of financial literacy on both indirect effects were examined. While the interaction term of financial literacy and job loss was not significant, the marginal effects displayed meaningful patterns. Both the mediation model and the moderated mediation model presented satisfactory global model fit.

Conclusion and Implications: Our results show that COVID-19 job losses increased the probability of making early withdrawals from retirement accounts through increased financial hardships and lower perceptions of financial wellbeing. Our findings suggest that policies to minimize the impact of job loss on long-term financial stability (e.g., retirement financial security) and asset accumulation may lower households’ risk for financial hardships and improve their perceptions about financial wellbeing, especially in the context of a major event or crisis.