Second Homes and Demographic Change in Swedish Municipalities

Sammanfattning

Sweden is one of the countries in the world with the highest rate of secondhome ownership. These dwellings are disproportionately located in in ruralmunicipalities that are also experiencing a trend of population decline. Thisoverlap has been part of a long running debate over the effects of secondhomes and whether they help contribute to rural depopulation throughdisplacement. Or does the decline in population and high number of secondhomes simply come as a symptom of structural rural decline or is it highlycontext dependent and vary from municipality to municipality. This thesis aimsto test these three positions by doing a quantitative cross-sectional spatialanalysis of the relationship between second homes and population change inall of Sweden’s 290 municipalities. Second home share is measured in 2010 andused to predict population change over the period 2010-2024 with Agestructure, domestic migration, income, education, tax rate and populationdensity as controlling variables. The study combines ordinary least squaresregression, exploratory spatial data analysis, a spatial lag model, andgeographically weighted regression (GWR and MGWR), using open municipal-level data from Statistics Sweden.At the bivariate level, when only comparing second home share and populationchange there is a negative association, meaning that municipalities with highshares of second homes have worse demographic outcomes. When addingother control variables however this association is reversed, revealing thatwhen controlling for those above mentioned variables there is a positive andsignificant association (b = +0.143, p < 0.001). One of the core findings fromthe results of the spatial lag model and MGWR is that the results show nomeaningful spatial variation. These findings reject the displacement hypothesisat the municipal aggregate level and support a symptom-and-contributioninterpretation. An interpretation that can be made is that second homes clusterin already-declining areas and that their presence is associated with asomewhat better demographic outcomes than those areas' structuralcharacteristics would predict, plausibly by sustaining the seasonal demand thatkeeps local services viable. The effects of second homes are however rathersmall, and can not be seen as a way to reverse depopulation in dyingcommunities. The results indicate that municipal planners should see secondhomes as a small piece of the puzzle not as the solution but also not as a threat,and that should be dealt with in accordance to local housing conditions ratherthan an overarching displacement narrative that is not supported by the data.