Get access

Mortgage Choice as a Natural Field Experiment on Choice under Risk


  • The authors wish to express grateful thanks to the editor and two anonymous referees for suggestions that have led to significant enhancements to the paper. They also wish to thank Dan Houser, Glenn Harrison, and other participants at the CEAR Workshop (Econometrics of Choice Under Risk and Over Time) in Denver in January 2011, for useful comments. The first author gratefully acknowledges sponsorship from the ESRC under studentship PTA-031-2004-00221. Data available from the UK Data Archive at Essex University: The standard disclaimer applies.


Data on approximately 280,000 borrowers from the UK Survey of Mortgage Lenders are used to model choices between variable and fixed rate mortgages. The choice is assumed to depend on three factors: risk attitude, interest-rate expectations, and individual discount rate. The ordered probit model is used for estimation, while taking account of a number of econometric issues including missing counterfactuals, selectivity, and endogeneity. A large number of strong effects are found, including: higher income borrowers are less risk averse and have a lower discount rate, and risk aversion rises with the amount borrowed, providing evidence of increasing relative risk aversion.