Determinants of household savings rates: Logistic quantile regression approach

Abstract

Saving rates have a fundamental economic importance that affects the economic performance of countries and the welfare level of individuals. Savings have been addressed in various ways with alternative economic approaches. The determinants of household savings rates were examined at the level of quantiles in this study. For this purpose, the logistic quantile regression approach proposed for bounded dependent variables was used. Since savings rates have a bounded and continuous structure, it is appropriate to analyze them with this method. Income level has been considered the principal determinant of savings rates and the change in the effect of income on savings rate at the level of quantiles was examined in details. As a result of the analysis performed separately for homeowners and tenants, it was determined that there were differences between the two groups. The change in the income effect was non-linear at the quantile level in both groups. While income was more effective at high savings rates for homeowners, it was more effective at low savings rates for tenants. On the other hand, the effects of other characteristics of the households also differed between the homeowners and the tenants at the level of the quantiles.