Yves here. Note that this article originally had the anodyne headline, “Housing affordability is reshaping Europe’s social fabric.” Surprisingly, the authors tell us that few studies have been done on how unreasonably expensive housing harms society. So this analysis is very important to be seen.
By Jean-Jacques Hallaert, International Monetary Fund Macroeconomics, and Iglika Vassileva, International Monetary Fund Senior Economist. Originally published at VoxEU
Housing affordability has been at the forefront of the European policy debate. While the economics literature focuses more on the drivers of housing unaffordability, this column examines its effects in the EU. The authors find that increased housing cost burdens have a significant negative impact on housing adequacy, which has a significant impact on poverty and health. Effects on fertility and labor force participation are small.
Housing affordability has been at the forefront of the European policy debate. Surveys, media reports, and protests across the EU suggest that housing is increasingly seen not only as an economic concern, but also as a threat to social cohesion. For example, in the European Affordable Housing Plan, the European Commission says: “What Europe is facing is more than a housing problem. It is a social problem” (European Commission 2025a).
However, most economic research focuses on why housing is so unaffordable and its distributional effects across groups and regions, and very little attention has been paid to its consequences (European Commission 2025b).1 In recent work (Hallaert and Vassileva 2026), we show that rising housing cost burdens are not just a symptom of broader pressures; and they have important consequences for individual well-being as well as for economic, demographic, and social conditions.2 Importantly, our analysis considers the impact of housing availability on multiple outcomes for the same time period (2010-2023) and for the same geographical area (EU27 countries).
The Three Keys to Real Estate Costs
We measure how high housing costs affect individuals and communities in three main ways.
First, high housing costs reduce housing affordability, forcing households into small, substandard, or poorly located residences. This is a particularly difficult problem for young people (European Commission 2025b, Eurofound 2023). We capture this process by measuring the impact of rising housing costs on housing affordability and severe shortage rates.3
Second, high housing costs lead to lost opportunities. They make it difficult for individuals to approach work and educational opportunities. Reduced opportunities may increase poverty and contribute to low labor force participation and inefficient labor distribution, all of which affect well-being and productivity and economic growth (Glaeser and Gyourko 2018, Nguyen et al. 2026). We test the importance of this mechanism by measuring the impact of increases in housing cost burdens on the number at risk of poverty and labor force participation.4
Third, higher housing costs suppress non-housing consumption. When a large portion of income is spent on housing, some households may have to cut back on other expenses. This can reduce access to health care, nutrition, and education, and affect birth decisions. We illustrate this mechanism by estimating the effect of housing availability on fertility.
The impact on health shows that the channels can be connected: poor housing conditions can affect physical and mental health directly, but health status may deteriorate if one reduces spending on health care.
Using Machine Learning to Measure Housing Affordability Outcomes
Measuring the effects of the housing cost burden is challenging because the forces that shape housing availability also drive the outcomes we care about. In a relatively short panel, the list of potentially important factors is long compared to the number of years available. A two-part least-squares textbook approaches the struggle in this setting. So we use a dual machine learning method with dynamic instruments (DML-IV),5 which allows us to include a richer set of controls, while allowing the relationship to be nonlinear and allowing the effect of housing costs to vary across countries and over time.
Figure 1 summarizes the results.6 Each coefficient is best described as an elasticity. For example, a coefficient of 2.3 in the density ratio implies that a country where housing costs as a share of disposable income is 1% above the EU average tends to have a density level about 2.3% above the EU average, all else being equal.
Figure 1 Estimated impact and significance of a 1% deviation in housing costs (as a share of disposable income) from the EU average
Be careful: Whiskers report statistical significance at 10%.
The source: Author’s calculations.
The Major Impact of Housing Affordability Is on Housing Conditions
High housing costs quickly translate into a noticeable deterioration in living standards. Figure 2 shows that even a small increase of 1% in the share of housing costs in disposable income compared to the EU average causes a 2.3% increase in the level of overcrowding and a 1.4% increase in the level of severe housing deprivation compared to the EU average. We find no evidence that the inability to find affordable private housing holds young people back from leaving the parental home. Therefore, it is not a logical route to a deterioration in housing conditions or other dimensions.
Broad Economic and Social Outcomes: Jobs, Poverty, Population, and Health
We estimate that when the burden of housing costs is 1% higher than the EU average, labor force participation is lower than the EU average by about 0.55% (0.4 percent) for all women and the population (Figure 2).
Low labor force participation and lost job or education opportunities due to housing costs reduce individual income and thus can increase the poverty rate. We estimate that a housing cost burden 1% higher than the EU average increases the poverty rate by 0.9% (equivalent to 0.15 percent) compared to the EU average of 16.8% (Figure 2). Thus, housing affordability problems are not only indicative of poverty; they also contribute to it (Desmond, 2016)
The fertility effect is present but small, operating in part due to delayed parenting (Kearney and Levine 2025, van Doornik et al. 2025). We estimate that when housing cost burdens are 1% higher than the EU average, women will have their first child on average less than a month later than the EU average and that the fertility rate will be 0.1 lower than the EU average of 1.51 (Figure 2).
Consistent with the medical literature, we find that housing costs also affect health outcomes. Housing costs per disposable income 1% above the EU average are associated with a 1% increase in the share of people reporting “bad” or “very bad” health compared to the EU average of 9.17%. This is another channel through which the availability of housing affects the well-being of individuals (including children’s physical and emotional development; see Hallaert et al. 2023) and, by affecting the main part of the human economy, also affects productivity and potential growth.
Figure 2 Estimated impact of a 1% deviation in housing costs on the net impact from the EU average
<
Be careful: The ratio of EU housing costs to disposable income was 20.8% in the period considered. The chart reports an average effect of 1 percent more (21.0%) on all variables on the horizontal axis. The EU27 reports the average value of each variable during that period.
The source: Author’s calculations.
The conclusion
By influencing where and how people live, their access to work and education, their consumption choices, and even their decisions to have children, housing is reshaping Europe’s social fabric and affecting economic prospects. Although we have presented an EU-wide measure of results, the DML-IV framework also produces country- and year-specific estimates, revealing significant cross-country variation and how the negative effects of housing availability have eased after the pandemic. This may partly reflect the increase in remote work, which may alleviate housing-related issues, but this requires further investigation.
Authors’ note: The views expressed here are those of the authors and should not be attributed to the IMF, its Executive Board, or its management.
See the original post for clues
