The effect of housing loan subsidies on affordability: evidence from Croatia

In this paper, we explore whether a program of housing loan subsidies introduced in Croatia contributed to housing price increases. The subsidy was designed to cover a portion of annuities in the initial period of the housing loan repayment, with a distinct feature that a household could apply for the subsidy only during a month-long period. Using a dataset on housing transactions we document that the subsidy disrupted the usual intra-annual dynamics of residential transactions as they became very concentrated in the month when housing loan subsidy applications ended. However, we do not find any descriptive evidence that the aggregate homeownership rate increased. Using an event study approach we find that the housing prices increased just around the introduction of the subsidy, especially in the most developed municipalities which absorbed more than 60% of the subsidies. Exploiting the subsidy built-in implementation rules we identify the effect of the subsidy via a dynamic difference-in-difference approach which corroborates our baseline conclusions. Using the regional variation in the subsidy intensity, we also find that housing price capitalization was driven by prices in areas and regions with already active housing markets. Finally, to discard the possibility that the observed housing price increase is confounded by tourism, we also explore the nexus between tourism and the housing market in Croatia and conclude that our baseline results hold.