Mortgage Choice as a Natural Field Experiment on Choice Under Risk∗

Microdata from the UK Survey of Mortgage Lenders is used to model borrowers' choices between variable and fixed rate mortgages. The data is treated as a large-scale " natural experiment " on risky choice, with the choice of a fixed rate corresponding to the " safe choice " in a more conventional experimental setting. The choice is assumed to depend partly on risk attitude, and partly on expectations of future movements in interest rates. Approximately 280,000 choices, made by borrowers between 1992 and 2001, appear in the sample. The ordered probit model is used for estimation, while taking account of a number of econometric issues including missing counterfactu-als, selectivity, and endogeneity. Explanatory variables are divided into three groups: mortgage price variables; interest rate expectations; and borrower characteristics. A number of strong effects are found, including: fixing is more likely when agents expect interest rates to rise; a larger amount borrowed (i.e. higher pay-offs in the choice problem) increases the propensity to fix; the presence of female borrowers increases the propensity to fix; older borrowers are less likely to fix; high-income borrowers are less likely to fix, particularly so if income is " self-certified ". The results offer fresh insights into the analysis of risky choice, particularly with regard to the roles of incentives and subjects' income.