Abstract
In this paper, we use a novel natural experiment and Norwegian tax data to quantify the causal impact of unemployment risk on individual savings. By comparing individuals who live in the same area but face different increases in risk, we show that a one-percentage-point increase in unemployment rates increases safe assets by 1.3%. Reassuringly, this effect is driven by low-tenured workers, who face the highest increase in risk. Savings in risky financial assets remain unaffected, implying a decrease in the overall risky share of individual portfolios. We use two independent approaches, relying either on cross-sectional variation or time series variation, to quantify the importance of job loss risk in accounting for higher savings during recessions. Our results suggest that unemployment risk can explain 60%–80% of the recession-induced increase in safe assets.