Tax concessions for owner-occupied housing and tenure choice in Belgium

This paper analyses the influence of the tax treatment of mortgage loans on the tenure choice of households in Belgium. It is very common e.g. in OECD countries that owner-occupation is subject to a favorable tax treatment: homeowners are allowed to deduct mortgage interest and imputed rents are exempt from income taxation or significantly underestimated. This bias in favor of owner-occupation is in most cases an heritage of past tax rules, but it persists deliberately in many countries. Defenders see it as a mean of promoting homeownership, which brings positive externalities, is a part of an ideal, and offsets partially imperfect social security systems. However, this tax expenditure is costly and its effectiveness is controversial. In the U.S., several authors argue that the mortgage interest deduction has only a marginal effect on the rate of homeowners, or even no effect at all. More specifically, the effect would even be negative in urban areas or for young households. Few analysis concern European countries.This paper focuses on Belgium. In this country, homeowners can deduct at marginal taxation rate an annually-limited amount of mortgage repayments for their only own home, including interests and capital under an aggregated form. In order to assess the effect of this tax relief on households' decision to buy or rent, we estimate a model of tenure choice using a cross-sectional sample of 800 young households who are "recent movers". Microdata come from the 2012 EU-SILC survey. We compute the net present value of the total tax benefits for owners or losses for renters and use it in the model, while controlling for the effects of income, demographic characteristics, housing prices and location. Regarding income, we compute an estimate of permanent income and we consider both the effect of current and permanent income. Our results indicate that the tax relief has a significant positive effect on the choice of owner-occupation, but that it is secondary when compared to some other variables, notably demographic characteristics. The results are interpreted with regard to the objectives and the cost of that policy.