Födas, åldras, låna

Sammanfattning: Household debts is interesting for a number of institutions in society. One reason for this is that it gives a picture of how severely a potential economic crisis might affect the country. If a model can be found which precisely explains the household debt levels, it can be used to predict dangerous situations at an earlier stage, so that resources to counteract the situation can be employed before it turns into a large-scale meltdown. This thesis tries to find such a model, by examining whether there is a relation between debt level (measured as the quota between total debt and disposable income) and the natural variation in the number of births each year over a longer period of time. The thesis also examines whether the number of births, when used as explanatory variable, can outperform a model where interest rate is used as explanatory variable instead, with regard to explanatory power. The models used are estimated via application of time series analysis (AR(p)-models) and LASSO regression. The author finds signs indicating that there might be a connection between the number of births and debt levels, but nothing indicates that number of births is a better explanatory variable than for instance interest rate levels. It is also shown that the time series analysis gives the largest contribution in explaining debt levels, and a more rigorous analysis of time series is proposed for potential future studies on the topic.