The impact of housing prices on residents’ health: a systematic review

Abstract

Background

Rising housing prices are becoming a top public health priority and are an emerging concern for policy makers and community leaders. This report reviews and synthesizes evidence examining the association between changes in housing price and health outcomes.

Methods

We conducted a systematic literature review by searching the SCOPUS and PubMed databases for keywords related to housing price and health. Articles were screened by two reviewers for eligibility, which restricted inclusion to original research articles measuring changes in housing prices and health outcomes, published prior to June 31st, 2022.

Results

Among 23 eligible studies, we found that changes in housing prices were heterogeneously associated with physical and mental health outcomes, with multiple mechanisms contributing to both positive and negative health outcomes. Income-level and home-ownership status were identified as key moderators, with lower-income individuals and renters experience negative health consequences from rising housing prices. This may have resulted from increased stress and financial strain among these groups. Meanwhile, the economic benefits of rising housing prices were seen to support health for higher-income individuals and homeowners – potentially due to increased wealth or perception of wealth.

Conclusions

Based on the associations identified in this review, it appears that potential gains to health associated with rising housing prices are inequitably distributed. Housing policies should consider the health inequities born by renters and low-income individuals. Further research should explore mechanisms and interventions to reduce uneven economic impacts on health.

Introduction

In contemporary society, the structures we live in, as well as our legal relationships to these structures, are intertwined with our fundamental senses of self and belonging [1,2,3]. For decades, homeownership has been recognized as a core measure of success [4, 5]. Recognizing the importance of housing, studies have variously examined the effects of wide-ranging housing-related factors on health, including housing quality, overcrowding, neighbourhood deprivation, social cohesion, housing density, housing suitability or sufficiency, and neighbourhood socioeconomic status [6, 7]. While these effects continue to be explored, it is generally agreed that housing is a fundamental determinant of health [7], which broadly exerts impacts on health through a variety of mechanisms.

Indeed, housing-related health effects arise from specific housing conditions, as well as the legal conditions that define our relationships to these spaces, and our emotional attachments to these various factors. For example, living and owning a home can create access to opportunities that can further bolster health [8]. Similarly, housing related factors—such as indebtedness, mortgage stress, and credit problems—can cause severe mental health problems, depression, and suicide ideation [9, 10]. With these factors in mind, people in most countries face numerous barriers to securing their right to a home [5, 11], and a wide array of policies have been proposed and implemented to address these barriers [12,13,14]. In addition to these factors, the location of a home, the quality of a building, or the neighbourhood context in which a home exists are also hugely influential to health [7, 15, 16].

In conceptualizing these varied mechanisms, it is important to consider both direct and indirect mechanisms through which the relationship between housing and health manifests. Direct effects predominantly emerge from psycho-physiological stress responses. Elevated housing costs can induce chronic stress, leading to mental health conditions, like anxiety and depression, and other health problems [17]. Indirectly, escalating housing prices exert economic pressures that limit individuals' capacity to allocate resources towards health-promoting activities and necessities. This economic strain can result in compromised nutrition, reduced access to healthcare services, and diminished ability to manage chronic conditions, therefore, exacerbating health disparities. Moreover, the financial burden can lead to other lifestyle changes that further impair physical and mental well-being, such as increased substance use or reduced physical activity.

Despite these effects being documented in previous studies, there are no systematic reviews on the impact of rising housing prices on health. The present review aims to examine the effect of housing price on health by considering whether changes in housing market price impact the health of residents living in an area. To accomplish this aim, we conduct a systematic review. This review is especially timely since housing prices have risen in the past five years at an alarming rate.

Methods

Article search

The first step in our multi-stage systematic literature review was to manually identify relevant articles through a rudimentary search on SCOPUS and PubMed (Appendix B). We then created a list of keywords to use for our search. Keywords aimed to identify articles that measured changes in housing prices and health impacts, Appendix A outlines how we identified keywords and provides a complete list of selected keywords. After conducting the keyword search in PubMed and SCOPUS, duplicates were removed and the remaining articles were then uploaded to Rayyan, an online software that aids in systematic reviews [18]. To assess whether our search is comprehensive, AG confirmed that the articles identified in the rudimentary initial search, mentioned earlier, were also included in this search. For the purposes of this literature review, we define health using the language provided by the World Health Organization (1948): “health is a complete state of mental, physical, and social well-being, and not merely the absence of disease.” As such, no additional inclusion or exclusion criteria were used to exclude or include specific health conditions. We felt this was appropriate given that this is the first literature review on this topic and because after a review of included articles, it was apparent that a wide variety of health outcomes have been considered. Furthermore, the biopsychosocial models of health that we engage to inform our view that housing prices have direct and indirect effects on health underscore that diverse and nuanced pathways across various mental and physical domains of health are likely important to consider. Using Rayyan, AG and LW reviewed the titles of each manuscript to remove articles that were clearly not relevant to this review [18]. The application of inclusion and exclusion criteria resulted in 21 articles that were directly relevant to this review. AG and LW also searched the reference lists for these 21 articles to identify any additional articles. These missed articles were added to our final inclusion list, creating a total of 23 included articles.

Data extraction

Data were extracted by AG and LW from each of the identified and included articles and AG re-reviewed the data extraction to verify accuracy. Extracted variables included: first author name, year of publication, years of data collection, sample size, location(s) of study, study design (e.g., case control, cohort, cross-sectional, serial cross-sectional study), analysis type (e.g., regression), outcome, explanatory factor, confounders/mediators/moderators, and a summary of primary findings (including effect size measures). This data extraction is provided as Table 1.

Table 1 Data Extraction for Included Studies

Risk of bias assessment

During the data extraction process, we conducted an assessment based on the Joanna Briggs Institute Critical Appraisal Tools [42]. Each study was classified according to its study design and rated using the appropriate tool designed for each study. However, despite varying methodological quality, no studies were excluded based on risk of bias assessment, as there were no clear sources of systematic bias with sufficient likelihood of challenging the conclusions of the source studies.

Narrative synthesis

During the data extraction and risk of bias assessment phases, AG and LW recorded general notes on each of the studies. These notes, along with the extracted information, were used to construct a narrative synthesis of the evidence. This process was guided by Popay et al.’s [43] Guidance for Narrative Synthesis in Systematic Reviews. A narrative approach was selected to allow for an examination of the potential complexity inherent in the synthesis of findings across contexts, time periods, and populations to provide a nuanced discussion of what roles housing and rental markets might play in shaping health, with attention to both outcomes and potential mechanisms. Findings within study classes were reviewed to determine potential mediation and moderation. These explorations informed the development of a list of key points used to organize the presentation of our results. We then integrated and contextualized these findings with those from other relevant (though excluded) studies identified through our review process and from the texts of the included articles.

Results

Included studies

Our keyword search returned 6,180 articles. Of these, 5,590 were removed based on review of the abstract and title as they were not directly related to our review topic (i.e., they did not measure changes in housing price and/or health outcomes). The remaining articles were reviewed based on their full-texts and a final list of 26 articles were considered for inclusion. However, five articles were not able to be retrieved (even after emailing the original authors), leaving us with 21 articles. The reference lists and bibliographies for these 21 included articles were then screened and two additional articles were thus included in our review resulting in a final sample of 23 articles. Figure 1 shows the flow diagram for included studies and these studies are listed in Appendix B.

Fig. 1

PRISMA systematic review flow diagram

Dates and locations of studies

A full description of studies is included in Table 1. Studies were published from 2013–2022. Ten studies were from East Asia, eight from the United States, three in Europe, one from Australia, and one included nine countries (France, Japan, Netherlands, Spain, Switzerland, Sweden, United Kingdom, USA).

Study design

Of the included studies, ten had a longitudinal study design, and thirteen studies were serial cross-sectional studies. Studies examined the effect of housing prices on health over time by repeatedly surveying a specific geographical area or population. One study included both qualitative and quantitative data collection.

Outcome variable measurement

Most studies compared multiple outcomes. Seven studies focused on mental health as the outcome variable—utilizing various measures, including self-rated mental health, standardized scales for depression or anxiety, and receipt of pharmaceutical prescriptions [22, 32, 33, 35, 40, 41]. Nine studies analyzed the impact of housing prices on physical health—utilizing various measures of physical health, including objective assessments of physical health (e.g., body mass), self-rated physical health assessments, reports of specific health conditions (e.g., COVID-19), reported health behaviours (e.g., alcohol use, smoking), and mortality ( [24, 28, 29, 37, 44, 30, 36, 38, 39]). Seven studies included both physical and mental health measures as their outcome variable [19, 20, 23, 26, 27, 34, 45].

Explanatory variable measurement

Housing prices were measured using many different types of data, including house price index, self-reported housing price (extracted from surveys), and average market price. Many studies used house price index as a measure of housing prices [19, 21, 22, 30, 38, 39, 41]. Zhang & Zhang [33] included self-reported housing price. Alternatively, many studies examined housing market prices using existing survey data [20, 23,24,25,26,27,28,29, 31, 32, 34,35,36,37, 40].

Key findings

Studies included in our review highlighted a plurality of results when testing the relationships between housing prices and health. As shown in Tables 2 (physical health) and Table 3 (mental health), the included articles reported mixed findings across the outcomes explored. Given the heterogeneity of findings regarding the associations between housing prices and health outcomes, several authors examined potential moderators and mediators in attempt to understand the mechanisms at play. These included studies examine the role of wealth effects (by comparing effects on homeowners and renters), socioeconomic status (e.g., income level), and broader economic forces (e.g., area-level improvements). While keeping these pathways in mind, there are likely other alternative explanations beyond those explored. However, these appear to be the most dominant frameworks used to understand the effects in our included studies.

Table 2 Narrative overview of primary themes and findings, with citations for relevant articles (Physical Health)

Table 3 Narrative overview of primary themes and findings, with citations for relevant articles (Mental Health)

Wealth effects

The first major pathway has been described as a “wealth effect” – which produces different effects for homeowners and renters [19, 20, 23, 25, 27, 33, 35, 37, 45]. For example, Hamoudi & Dowd [37] report that homeowners living in areas with steep price increases, perceive this as an increase in their overall wealth, resulting in positive health outcomes (not observed for renters). Similarly, Zhang & Zhang [33] show that increases in house prices has a positive effect on homeowner’s subjective well-being. De & Segura [30] specifically notes that price depreciation causes homeowners to experience feelings of a loss of wealth, leading to increases in alcohol consumption. Among studies that fail to show a wealth effect, Daysal et al. [29] shows that rising prices in Denmark do not impact households due to the buffering effects of government supports. Conversely, when examining the effects among renters, Wang & Liang [31] argue that rising housing prices have detrimental "strain" effect, which is also observed in several studies included in our review [25, 27, 38, 39, 45].

Income level

In addition to the wealth and strain effects illustrated through studies among homeowners and renters, many studies also examined the mediating effects of income [19, 20, 22, 28,29,30, 32, 33, 35, 38,39,40]. Several of these studies show that housing unaffordability constrains spending and that low-income individuals are particularly impacted [22, 24, 33, 38]. For example, Wong et al. [39] show that housing prices lead to reduced fruit consumption. However, results also show positive impacts for low-income homeowners – as exemplified by work showing that low-income homeowners are more sensitive to housing price gains [38, 40].

Broader economic forces

In considering both mechanisms described above, authors of included studies have also considered whether housing prices are merely an indicator of broader economic trends merits consideration. The most common strategy for accounting for this has been to include other indicators that might capture area level improvements. Indeed, most studies controlled for both individual characteristics or variables, such as age, gender, marital status, years of education, race/ethnicity, and employment status [20, 23,24,25,26,27, 29, 30, 32, 34, 35, 37, 39,40,41, 45], and a variety of economic factors, including individual income, country-level median income, and local area characteristics [19, 22,23,24,25,26,27,28, 30, 32, 36,37,38,39,40, 44, 45]. These factors are important to control for because rising housing prices can indicate a growing economy in which there are substantial improvements to neighbourhoods and communities [27, 29, 33, 38, 40]. As such, the observed improvements in health could simply arise from broader economic benefits (rather than being specifically attributable to housing prices) [31, 33]. However, generally speaking studies showed that there were independent effects of housing price or value, even after controlling for local area level improvements, and wider economic conditions [19, 27, 38].

Strength of effects

Given heterogeneity in the direction of effects, the lack of standardization in the reporting of effect sizes from study to study, differences in the measurement of exposure and outcome variables, and variation in the inclusion of mediators, moderators, and confounders, we did not conduct a meta-analysis to describe the effect size of housing price on health. However, housing prices appear to exert influence on health and wellbeing with statistically significant effects across various health-related outcomes (See Table 1 for range of effect measures). The effects generally appear to be smaller when considering specific health conditions and greater when considering more subjective and more broad definitions of health (e.g., self-rated health). Of course, at a population-level, even relatively small effect sizes may pose a considerable challenge. For example, Xu & Wang [36] report that a 10% increase in housing prices is associated with a 6.5% increase in probability of reporting a chronic disease – a relatively small increase on a person-level, but when scaled could easily pose a considerable burden to the health system. In summary, further careful measurement and methodological refinement is needed to quantify the effects of housing prices on various health conditions. For any given health condition, this will require multiple well-designed studies across place and time. Such replication is particularly important given the observed sensitivity of findings to the inclusion of confounders, moderators, and mediators.

Discussion

Primary findings

While examining whether changes in housing prices are associated with changes in health, we recognized it is difficult to establish a directional and causal relationship between housing price and health. This is particularly true given that health may increase opportunities for home ownership and economic success [3, 46,47,48,49]. Nevertheless, given the wealth of literature highlighting housing affordability as a key determinant of health [7, 9, 10], it is reasonably anticipated that rising housing prices would be associated with worse health outcomes among individuals who do not own housing. However, based on analyses of the studies included in this review, the relationship between housing price and health is complex and nuanced, with a significant degree of heterogeneity across outcomes and populations.

First, this review illuminates that changing housing prices impacted different people differently, depending on for example, income level, gender and/or homeownership status [19, 22, 25]. The negative impact of housing price on health for renters and low-income individuals could be due to the existential angst from being excluded from home ownership, which is often considered an important indicator of social class and success [8]. However, this could also be due to the cost effect that is created from the rising house prices, subsequently raising low-income owners and renters’ cost of living [31]. Additionally, renting may be associated with lower neighbourhood tenure, especially when individuals are priced out of a neighbourhood [8]. As a result, they may experience deleterious health effects associated with loneliness, social isolation, lack of neighbourhood cohesion, and community disconnectedness [8]. Likewise, the positive effect observed among homeowners and high-income individuals may be explained by increases in psychological safety leading to changes in health behaviors, for example, knowing they have invested in a home that will support them or their heirs financially, people may be better able to focus on their well-being. As well, homeowners may be able to directly leverage the value of their home to gain access to additional capital and investment opportunities – which could support increased financial wellbeing [8].

The effects of housing on health can be conceptualized as arising from two sources. It appears that rising “cost effects” (the increased costs of houses and the costs passed on to tenants) are inversely correlated with health while “wealth effects” (the contributions of housing price to person wealth) contribute positively to health (for example, for homeowners and investors whose wealth increases due to the rising cost of housing). The balance of these effects differs depending on their unique impact on individuals – with lower income people and renters more strongly impacted by cost effects, and higher income people and homeowners more strongly impacted by wealth effects.

In considering these effects, we note that there is likely considerable geographic, temporal, and contextual variation in the health effects of rising housing prices. For example, rising housing prices may occur alongside neighbourhood improvements (or degradation) and economic booms (or recessions) [45], which themselves are associated with improvements (or deterioration) to health [8]. As such, the presence of these factors may obscure or interact with the gains to health. Similarly, variations across cultures and countries may change how individuals internalize the rising housing prices [50], causing them to experience greater or lesser distress in reaction to rising prices.

Limitations of included studies and directions for future research

Given these two primary factors, research highlights several opportunities for improving this literature. First, future studies should give more careful attention to how moderators and mediators are conceptualized. For example, “home-owners” are hardly a homogenous class of individuals: some own their homes outright and others are paying mortgages that offer varying levels of security (e.g., fixed vs. variable mortgages, 5-year vs. 30-year mortgages) [8]. Second, a broader range of effect moderators should be explored. For example, few studies specifically examined the health effects of rising home prices on vulnerable populations, including young adults and first time home buyers who may be especially disadvantaged by rapidly growing housing prices [50]. Similarly, isolating effects as arising from economic, legal, environmental, and social pathways can help identify strategies for mitigating health harms. For example, it may be important to understand whether changes to neighbourhood environments drive health harms as opposed to changes in personal financial status. Third, more within-person studies are needed to understand the potential mechanisms and situational factors that promote or mitigate the health effects of rising housing prices. Along with use of appropriate, theoretically informed moderators, isolating the within-person effects can help us better quantify the effects of interest to inform policies and prevention strategies. Fourth, longer follow-up times may allow for better understanding regarding the time-horizons of the effects explored. Indeed, it is possible that rising housing prices could have differential effects on the health of a population in the short versus long term. This is particularly important given the interaction between housing prices (which may act as a price signal for investments) and other economic factors with strong potential to increase health [8]. Fifth, the studies used a variety of measures for housing price and health outcomes – which varied in quality. For example, health outcomes were primarily measured using self-reported measures [19, 20, 23, 25,26,27, 30, 31, 33,34,35, 37,38,39,40,41] – which may be highly sensitive to bias due to the likelihood that individuals might report worse health when they are unhappy with economic factors. Improving measurement of outcomes can be done by leveraging administrative and other data sources. Sixth, it can be difficult to link area-level and individual-level factors, particularly in the context of limited cross-sectional studies or in longitudinal studies with only a few follow-up points. Likewise, many cohort-based studies have limited geographic coverage or insufficient temporal scope. As such, longer, larger, and wider studies are needed to fully ascertain the relationships under consideration.

Implications of findings

Although further research is required to overcome the limitations mentioned, existing evidence indicates that increases in housing prices may significantly influence health outcomes. Future studies should aim to exclude alternative explanations, examine the effects over longer periods, and establish consistent measurement methods to predict the impact of housing prices more accurately on health. The findings of those students will aid policymakers in creating strategies that address the health implications of rising housing prices. Policy makers should develop frameworks that respond to the impacts of rising housing prices on health. Such approaches could be facilitated through frameworks such as the WHO’s Health in All Policies (HiAP) policy, which advocates for the inclusion of health and social impacts among other criteria used throughout decision making processes [51]. Many studies in this review support this view and describe their work as having important implications for housing and health policy [19,20,21, 24, 26,27,28,29, 31, 34, 35, 37,38,39]. For example, Yuan et al. [26] notes the importance of directing government support and housing subsidies towards vulnerable groups – though these should be packaged with other policies [26]. Such supports can apparently buffer against the negative effects of rising housing prices by creating a saftey net that reduces the psychosocial and cognitive effects associated with economic changes in one's personal circumnstance. Arcaya et al. [21] also recommends governments investigate establishing more mental health facilities in areas where housing price fluctuations impact people's mental health but warns economic development that allows for greater investment in health infrastructure can also lead to increases in housing prices.

Of course, other types of interventions may also be warranted, including broader financial interventions (e.g., direct loans; [52], those which promote community, neighborhood, and social cohesion among residents [53], or those that aim to change how people value home ownership [26]. With respect to this final option, communities should consider whether renting may in fact be a desirable outcome for some individuals and therefore promote a culture in which individuals realize the variety of investment opportunities available to them rather than being overly-focused on a traditional model of investment [26, 32]. For example, Zhang & Zhang [33] writes that homeowners should be provided financial and economic knowledge to better manage wealth gains, however, this could be taken one step further and include the importance of educating people on the dangers on the commodification of housing, to prevent an over reliance on the importance of housing wealth gains.

Limitations of our review

In addition to the limitations specific to studies included in our review, our review itself also has several limitations. First, while we trained two reviewers to conduct article screening, assessed inter-rater reliability as greater than 80%, and adjudicated conflicts with the help of a third reviewer, it is possible some articles that could have been included were excluded due to the many different forms of outcome and explanatory variable measurement. Second, while we have searched multiple databases, used comprehensive key words for our search, and conducted manual searches of the reference lists, it is possible there were studies that we missed and failed to include in this study. That said, it is unlikely the exclusion of these articles would change our conclusion that the literature is currently mixed and that there is a need to flesh out the mechanisms and moderators that link housing prices to health. Third, we were not able to conduct a meta-analysis, and our numeric reports of the number of studies with each characteristic should not be treated as a meta-analysis. Rather, these findings and analyses of these studies should be interpreted as a descriptive analysis that highlights significant heterogeneity of findings and critical inconsistencies in the mechanisms, mediators, and moderators that give rise to these associations. Fourth, we did not exclude any studies based on article quality because we did not find there to be sufficient heterogeneity in the quality of the observational studies to merit exclusion according to variable inclusion, study design, or sampling method. In other words, we sought to avoid introducing bias by arbitrarily excluding articles – particularly given that the number of articles captured here was already relatively low (at least given the diversity of methods, measures, and populations captured). However, future reviews might consider more narrowed inclusion and exclusion criteria when a sufficient body of literature is available for a given outcome. For example, limiting analyses to only well-designed cohort studies might support a more careful selection of articles. Finally, we note that we included studies across a wide variety of health outcomes. While this was done to maximize inclusion (given the wide heterogeneity of measures used), we acknowledge that future research might be strengthened by studying specific pathways linking housing prices to specific health and social outcomes. Such detailed research is greatly needed so as to not only highlight the relevance of housing prices to health but identify strategies for mitigating potential harms of rising housing prices.

Conclusion

Our review shows that there are complex relationships between housing prices and health – with studies arriving to mixed conclusions across a wide-variety of health outcomes and populations. Yet, there is insufficient evidence for a causal relationship, but it appears that if such a relationship exists it likely differs according to homeownership status, income-level, and as a factor of the broader economic and structural forces in play, including the level of economic supports provided by governments for low income individuals. Future research should explore these pathways, moderators, and confounders using long-term, geographically diverse, cohort studies that account for a broad diversity of causal or alternative mechanisms. Such future research will allow for a more nuanced understanding of health and health inequities related to rising housing prices.