Abstract
We examine peer-effects in mortgage borrowing decisions. We analyze how characteristics of borrower’s colleagues are related to the chosen mortgage loan. Our main interest is in whether peers’ basic skills are related to individual loan taking decisions. Comparing borrowers’ choices within firms including industry - time fixed effects we find that those borrowers whose colleagues have high skills in mathematics tend to take mortgage loans with low interest rates. Relationships of the interest rate with peers’ other basic skills such as writing, reading, speaking or listening are less obvious. The measured impact is stronger when the borrower has lower skill. There is also heterogeneity in the results with respect to gender and the number of available banks in the district. Although with our empirical strategy we cannot rule out all endogeneity issues, placebo tests and robustness checks suggest that the most serious problems of sorting and time varying common shocks are not distorting the results. The magnitude of the effect is economically significant, roughly one fifth of the standard deviation of mortgage loan interest rates.