PurposeThe purpose of this paper is to analyze the effects of Swedish rent controls on observed vacancy rates for rental housing.Design/methodology/approachHousing vacancy rates are unevenly distributed among Swedish municipalities. In large expansive municipalities, such as Malmö, Göteborg and Stockholm, vacancy rates are very low, while in declining or smaller municipalities such as those in the northern and interior parts of Sweden, vacancy rates are considerably higher. This implies welfare losses not only in growing municipalities with queues for rental apartments but also in municipalities that are shrinking since the controlled rents there are higher than market rents and cause higher vacancy rates than with market rents. The authors estimate the influences of various determining factors, such as population growth, population size, rent levels, construction, demolition and market orientation of rents, on the observed vacancy rates.FindingsThe authors find that that these factors affect the vacancy rates differently depending on whether a municipality is large or small, growing or shrinking. Population growth, in percent per year, plays an important role in explaining the observed vacancy rates in declining regions.Research limitations/implicationsA research task that remains to be done is to calculate the welfare losses due to rent higher than the market rent for municipalities in contraction.Practical implicationsTo reduce the welfare losses of rent control, both in expanding and contracting municipalities, economists' straightforward recommendation to deregulate the rent control should, in principle, be carried out.Originality/valueIn many countries, rent control regulations are limited to cities, such as New York City. The paper shows that the Swedish rent control system however, applies nationwide, except for annual rent increases, which are set locally through negotiation.