Do Near Misses With a Disaster Affect Individual Demand for Credit? A Large-Scale Study of Credit Profiles

Abstract

We investigate the credit demand of individuals in areas narrowly missed by severe tornadoes. Leveraging detailed credit bureau data and a difference in differences matching model, we find that, compared with similar individuals far away from tornadoes, individuals living in areas neighboring tornadoes decrease credit demand following a tornado even without suffering from tornado-induced financial distress. Within one year after a tornado, the aggregate spillover effect relative to direct impact is roughly 49% for mortgages, 115% for personal loans, and 336% for student loans. These decreases are partly driven by individuals closing down accounts on the extensive margin and more prominent among individuals with more financial flexibility pre-tornado, despite them being better guarded against disaster fallout. Taken together, such behavior suggests elevated risks perception and salience-driven decision making. Finally, we confirm the results by developing a type of interpretable semiparametric deep neural networks. The findings inform us about the aggregate impact of adverse shocks by highlighting an often neglected psychological spillover effect. They also imply that studies using neighboring areas as control groups may misestimate a shock’s impact. Finally, they may promote more access to credit for individuals on the boundaries of disastrous events and better policy response to such events.

Keywords: Behavioral Bias, Natural Disaster, Mortgage, Household Finance, Climate Finance, Deep Neural Network

JEL Classification: D14, D81, G21, G51, H31, Q54