Measuring the Significance of Precautionary Savings: Empirical Evidence from the PSID

Many economic models assume frictionless worlds with perfect insurance markets. A notable exception is the standard incomplete markets (SIM) model which allows for incomplete insurance markets that result in precautionary savings among agents. In this paper, I test for evidence in support of precautionary savings using both macro and micro-level data. Using aggregate data, I find that the unemployment rate is negatively correlated with savings and consumption, and positively correlated with household debt. If the unemployment rate rises, savings and consumption fall the next period while household debt increases. At the micro level, panel data analysis shows strong evidence of precautionary savings following labor shocks, with household consumption and savings both falling after an adverse employment shock