Heterogeneity in demand for credit payment protection insurance

Abstract: Household indebtedness has increased remarkably around the world in the last two decades (IMF, 2023) e.g., in Finland, where households borrow more, but also end up with economic difficulties where they cannot afford to amortize their loans (Tilastokeskus, 2023). A way to shield one’s loan payments in case of a harder economic situation, is through a credit payment protection insurance. The purpose of this current paper is to study which factors are important in the insurance choice, and to give insight into differences in genders, age groups and income, when applied to the insurance choice. I will go through literature relating to insurance choices, risk preferences and heterogeneity related to the topic. I will investigate whether there are differences in the factors affecting the insurance choice. These results could give insight into which factors are important to include in macroeconomic models, where heterogeneity is a recurring research area. The analysis is done using a binary probit regression. The dependent variable is whether the individual has the insurance or not. The independent variables relate both to the loans, the individuals’ characteristics, and their earlier existing loans. The data used in the current paper was collected from a bank in the unsecured loans market and an insurance company operating in Finland. The results show that the insurance choice is on one hand affected positively by the loan size, nominal interest rate, increased age and if the individual is self-employed. On the other hand, the choice is negatively affected by the monthly payment, no. of dependents, net income, old loans, and if the individual is temporarily employed or has a co-applicant. All these factors are statistically significant. Further studies should look closer at the characteristic variables, for example, they could divide age into distinct categories to see more closely which age groups are important in the question. A gender divide with different age groups could also prove useful. Similarly for income, different income classes should be analysed more closely.