8. The impact of housing policy on housing inequality

Gerard Domènech-Arumí
ECARES, Université Libre de Bruxelles

The opinions expressed and arguments employed herein are those of the author and do not necessarily reflect the official views of the OECD, its Member countries, or the KIPF.

Housing inequality is relevant, yet we know very little about it. Housing is crucial to understanding income and wealth inequality and, as the “door of entry to neighbourhoods,” critical to studying social mobility and other inequality dynamics. However, housing inequality has been largely overlooked in the literature. Perhaps because income and wealth inequality have received most of the attention, or possibly because, until recently, computing capabilities required to process large administrative datasets (such as a cadastre) were very limited. This chapter reviews existing housing inequality estimates and discusses how different housing policy instruments may shape it, with a strong focus on recent evidence for Belgium.

The chapter starts with a review of the reasons to study housing inequality and a series of definitions of several dimensions of housing inequality (e.g., housing consumption versus housing wealth). Housing is both a consumption good and an asset. Households spend a significant fraction of their income on housing and invest most of their wealth in housing assets. Thus, housing is crucial to understand income, consumption and wealth inequality. It is also crucial to the study of social mobility, as housing's physical locations (i.e., neighbourhoods) are extremely relevant to several short and long-term outcomes (Chetty, Hendren and Katz, 2016[1]; Chetty and Hendren, 2018[2]; 2018[3]; Chyn and Katz, 2021[4]; Durlauf, 2004[5]). These reasons alone justify investigating housing inequality per se, but they also constitute strong foundations for looking at housing when the objective is to understand the roots of income or wealth inequality better.

Cadastral data is an ideal tool to study housing inequality. Their main advantages are cross-country availability and homogeneity, geolocation capabilities, coverage and the potential to perform historical analysis with cross-sectional data. The availability of this excellent resource provides yet another reason to study housing, especially when the focus of research or policy is at the sub-national level.

The chapter then reviews some of the few existing estimates of housing value inequality, focusing on Belgium. Housing inequality in Belgium is relatively low. Domènech-Arumí, Gobbi and Magerman (2022[6]) (DGM henceforth) estimate an overall level of housing (value) inequality of 0.25 (Gini index) and document substantial geographic heterogeneity, with inequality in some regions, districts and cities significantly below or above the national level. New estimates for Massachusetts (MA) also reveal significant geographic heterogeneity in that state but suggest that the overall level of housing inequality in the United States (US) may be substantially higher (Gini of 0.4 in MA). The results at the sub-national level reveal the great extent of geographic heterogeneity in (at least) the two countries and make salient that granularity in the data matters in the study of inequality.

Within cities, housing inequality appears to be higher in downtowns and lower in residential areas. The chapter then investigates within-city patterns by reviewing and reproducing existing local housing inequality estimates for Brussels and Barcelona (Domènech-Arumí, 2021[7]), as well as new estimates for Boston. The three cities show similar spatial patterns in their distribution of housing inequality across neighbourhoods, with higher inequality levels in the central parts of a city.

The correlation between housing value and income inequality is high. DGM find a correlation above 0.6 between their housing inequality estimates and available income inequality estimates for Belgium, a result that is in line with Aladangady et al. (2017[8]) and Albouy and Zabek (2016[9]) for the United States. These correlations reinforce the idea that exploring housing inequality may be helpful to understand income inequality when income data are unavailable, e.g., at low levels of aggregation.

The final section of the chapter starts reviewing a set of common housing policies and discusses their impact on housing inequality based on existing evidence. These policies are housing allowances, housing vouchers, mortgage interest deductions, rental investment subsidies, rent control and housing transaction fees (or stamp duties). Among these, housing vouchers and allowances appear to be the most promising for reducing housing consumption inequality. Rental investment subsidies may also achieve that goal, but they may exacerbate housing wealth inequality.

Finally, the chapter reviews the effects of a recent reduction in Flanders' housing registration fees on housing prices and inequality. DGM estimate that the 3 percentage point reduction in transaction fees caused house prices to increase by almost 3%, a result in line with Han et al. (2022[10]). They also find that house price increases were more prevalent at the bottom of the dwelling value distribution, thus causing the overall housing value inequality in Flanders to decrease by 0.8%. They document significant spatial heterogeneities, with inequality reductions being more prevalent in the east of the country and less in the areas close to Brussels. That geographic heterogeneity makes the local nature of housing salient, as even policies designed at the regional or federal level can have very distinct impacts at the local level.

Income or wealth inequality cannot be fully understood without considering the role of housing. Housing expenditure account for 10 to 30% of household consumption in OECD countries (OECD, 2022[11]) – thus, it is a crucial component of consumption and income inequality – and housing is the most evenly distributed asset of the wealth distribution (Garbinti, Goupille-Lebret and Piketty, 2020[12]), and therefore a major determinant of wealth inequality.

As the “door of entry” to neighbourhoods, housing is crucial for the study of social mobility. There is overwhelming evidence pointing to the many short and long-term effects of exposure to different quality neighbourhoods, such as social mobility or lifetime earnings.1 Therefore, given that owning or renting a home in a neighbourhood is the only way to benefit from (or be harmed by) neighbourhood effects, housing is a key to understanding current and future inequality dynamics. Housing is the link connecting the research on inequality and neighbourhoods.

Housing is a complex good. It is a consumption good, that provides shelter (housing structure) and access to local amenities (e.g., schools) and labour markets. It is also an asset and capital good, the major source of wealth throughout the income distribution (Kaas, Kocharkov and Preugschat, 2019[13]; Martínez-Toledano, 2020[14]). Finally, it is a source of income, specifically a direct source for landlords leasing housing units in the rental market and an indirect source for owner-occupiers. Consequently, different measures of housing inequality are informative of different things.

With knowledge of ownership and owner-occupancy, measures of the total housing value per owner are informative about wealth inequality. It is also reasonable to consider renters as households with zero wealth (Albouy and Zabek, 2016[9]), and homeowners (of potentially multiple properties) as households with positive wealth, with their total wealth being equal to the total value of all owned properties. Since the wealthiest households hold a higher share of their wealth in financial assets (Kuhn, Schularick and Steins, 2020[15]; Zucman, 2019[16]), that approach will understate true wealth inequality. A caveat is that, ideally, the most accurate measure of housing wealth should rely on market-value house prices, net of outstanding mortgage payments. While the former can be estimated with supplementary data (e.g., combining a transactions database with machine learning techniques), the latter is more difficult to supplement, as mortgage information is typically unavailable in cadastres. Despite its shortcomings, it is the view of the author that pursuing this measurement agenda can be fruitful for researchers and policymakers alike.

Housing value inequality is also informative about income inequality. In a world with homothetic preferences,2 housing consumption (rents or imputed rents) would be perfectly correlated with income. Because lower-income households spend relatively more on housing than richer households, the relationship between income and housing inequality is not one-for-one. Still, recent research suggests that there is a strong link between the two – see Albouy and Zabek (2016[9]) and DGM for more elaboration. When comparing dwellings of different values, it is useful to think of each home as representing a household. Most of the estimates reviewed in this chapter refer to this type of inequality and can therefore be interpreted as informative of household income inequality.

Finally, the inequality of housing space is informative about real income inequality. Recent work highlights that inequality may be overestimated because higher-income workers tend to live in cities with high housing costs (Diamond and Moretti, 2021[17]; Moretti, 2013[18]). Thus, net-of-housing consumption (and income) inequality may not be as stark as they seem after factoring in housing and other costs of living. In this sense, looking at housing space is informative of the extent of real housing consumption inequality at the national level. At a more local level (where amenities and land value are approximately held constant), it is also informative about income inequality.

Studying housing inequality with cadastral data offers several advantages. First, cadastral data are typically very homogeneous and available in many countries. This implies a high degree of replicability across contexts. Second, the data are typically geolocated, thus implying that time-varying or arbitrary administrative boundaries are not a problem. Third, the data typically contain information on the universe of real estate, thus implying that censoring or the need for imputations is not an issue. Finally, because real estate is durable and the year of construction is typically observed, it is possible and meaningful to construct a panel dataset from a simple cross-section.3

A disadvantage of cadastral data is that official value assessments of properties may be outdated, as it is the case in Belgium (see below). That limitation can be overcome with supplementary data. For example, real estate transactions or listing data may be combined to impute the contemporary market value of properties. A second, related, disadvantage is that some characteristics of the properties may not be appropriately updated (e.g., energy efficiency).

DGM (2022[6]) studies housing value (and space) inequality in Belgium using data from the Belgian cadastre. It should be noted upfront that the official cadastral values in Belgium date back to the 1970s and are outdated. However, DGM use the universe of Belgian real estate transactions from 2006-2022 to estimate current property values, rather than using the outdated official cadastral values. By leveraging this extensive transaction data, the housing inequality estimates presented here reflect contemporary market values and avoid the limitations of the archaic official cadastral valuations. They estimate an overall level of housing value inequality (as of 2022), as measured by the Gini index, of 0.25 – a number very close to the OECD estimate (0.26) of income inequality (OECD, 2022[19]). They take advantage of the granularity of cadastral data to further study inequality at different sub-national levels, from the region and down to the statistical sector level (the smallest administrative unit in Belgium). Figure ‎8.1, borrowed from DGM, offers a visualisation of their results in maps.

Inequality estimates vary substantially with the level of aggregation. At the regional level, Wallonia (in the south) is the most unequal (Gini of 0.265) and Flanders (in the north) the least unequal (Gini of 0.208). Inequality within regions is also substantial. For example, within the relatively equal region of Flanders, the communes (municipalities) and statistical sectors in the vicinity of Antwerp exhibit high levels of inequality. Aggregation matters and significant heterogeneities exist within country and regional borders.

Inequality within a city can also be substantial. Applying the methodology developed in Domènech-Arumí (2021[7]), the author also estimates the Local Neighbourhood Gini (LNG) – a Gini index capturing inequality in the immediate vicinity of a given building – for several cities in Belgium. Two significant advantages of the LNG are that it is independent of (arbitrarily drawn and changing) administrative boundaries and offers excellent visualisation of local inequality when plotted on a map. Figure ‎8.2 shows the estimates for Brussels.

Local inequality varies substantially within Brussels. Inequality is especially high in the southern neighbourhoods of the city, particularly in the areas close to Bois de la Cambre, the park in the south, some areas in the East of Ixelles, as well as central parts of the city close to Avenue Louise. Inequality is lower in the outskirts of the city. Particularly in Watermael (in the southeast) and Anderlecht (northwest). These are predominantly residential areas.

The previous figures make clear that inequality estimates vary substantially as we zoom in or out in terms of aggregation, at least in Belgium. Subsection 2.5 will discuss the possible causes, but before that, the chapter briefly reviews housing inequality estimates in other countries.

The chapter first keeps the focus at the very local level and the original LNG estimates from Domènech-Arumí (2021[7]) for Barcelona, Spain are reproduced in Figure ‎8.3.

As in Brussels, local inequality varies substantially across Barcelona’s neighbourhoods. Local inequality is especially high in the central parts of the city, particularly in the streets close to La Rambla and Diagonal Avenue, as well as in the neighbourhoods of Sarrià and Sant Gervasi. Local inequality is generally lower in the neighbourhoods of Sants, Sant Martí and Sant Andreu – in the southwest and east of the city. Heterogeneity in local inequality is substantial in Barcelona.

We next switch continents to look at housing inequality within Boston and Massachusetts (MA). Figure ‎8.4 and Figure ‎8.5 show housing value inequality estimates (Gini index) across Boston’s census tracts and MA’s municipalities, respectively.4 Estimates are computed using the 2021 assessed value of the universe of dwellings in the state, obtained from the Massachusetts property registry (the cadastre equivalent in the United States).

As in Brussels and Barcelona, inequality heterogeneity is substantial in Boston. Similar to its European counterparts, inequality is higher in the central parts of the city (Back Bay, the South End and East Boston) and lower in the neighbourhoods farther from downtown (Dorchester, Roxbury and Jamaica Plain). In contrast with Barcelona and Brussels, overall housing inequality is substantially higher in Boston. The citywide housing value Gini index was 0.226 in Brussels, 0.295 in Barcelona, but 0.427 in Boston.

Housing inequality in Massachusetts is significantly higher than in any Belgian region or Barcelona. The overall level of inequality in the state is 0.405, as measured by the Gini index. The same pattern and ordering appears when looking at income inequality. For example, income inequality (Gini) is 0.354 in Catalonia (Encuesta de Condiciones de Vida, 2019) and 0.48 in Massachusetts (American Community Survey, 2019).5

Finally, heterogeneity across MA’s municipalities is also very significant. Inequality is high in the towns close to Boston, Cape Cod, or the border with New York state. Inequality is lower in the central areas of the state. In ongoing and future work, the author of this chapter will extend and analyse housing inequality in the rest of the United States.

Housing inequality is largely unexplored. Some work has taken into account the role of housing in income inequality in terms of imputed rents (Frick et al., 2010[20]; Frick and Grabka, 2003[21]; Piketty, Saez and Zucman, 2017[22]), or provided nuances to the idea that differences in the standard of living in the United States are as stark as they seem by accounting for geographic disparities in housing costs (Diamond and Moretti, 2021[17]; Moretti, 2013[18]). However, there is very little work directly quantifying cross and within-country differences in housing inequality. The closest example of such work is probably Albouy and Zabek (2016[9]), who show that housing consumption inequality in the United States closely mirrored trends in income inequality in the second half of the twentieth century. They also show that housing inequality is primarily driven by differences in the value of land. For France, it is worth mentioning the work by André and Meslin (2021[23]), who use rich data from the French cadastre to explicitly look at the role of housing in explaining wealth inequality. For China, some work has shown a positive relationship between housing consumption and socio-economic status and has documented increasing inequality since the 1990s (Huang and Jiang, 2009[24]; Logan, Bian and Bian, 1999[25]). To the best of the author’s knowledge, there is no similar work that quantifies within-country housing inequality, as was done in the above sub-sections.

The housing inequality estimates for Belgium, Barcelona and Massachusetts revealed two patterns. First, within-country heterogeneity is large. At least in Belgium and Massachusetts, some regions and municipalities are more unequal than the country (or state) as a whole, whereas some others are significantly less unequal. Second, cities themselves are highly unequal and heterogeneous, but inequality tends to be higher in areas closer to downtown areas and lower in residential areas further from them. That same pattern emerged in the cities of Brussels, Barcelona and Boston.

Housing value inequality is (probably) highly correlated with income inequality. DGM (2022[6]) find a correlation between the two above 0.5 for Belgium and 0.6 for Brussels.6 The income and housing inequality estimates from Boston and Massachusetts, as well as the work by Albouy and Zabek (2016[9]), suggest that the correlation is also high in the United States.

Fully characterising the relationship between income and housing inequality will be crucial. Thanks to the OECD, the World Inequality Lab and others, our knowledge of cross-national inequality has significantly improved in recent years (Alvaredo et al., 2020[26]; 2022[27]; Chancel et al., 2021[28]; Kuhn, Schularick and Steins, 2020[15]; OECD, 2022[19]; Piketty and Saez, 2003[29]; Solt, 2016[30]). We know that (particularly northern) Europe is the least unequal region in the world, whereas Sub-Saharan Africa and Latin America are the most unequal. We also know that inequality in North America is significantly higher than in Europe and that disparities have grown since the 1980s. Unfortunately, we know very little about the current state and dynamics of inequality within those regions and countries, largely due to data availability (e.g., survey data may not have exact location identifiers, or sample sizes may be too small to obtain accurate measurements at low aggregations). Since cadastral data are typically geolocated and contain information on the universe of real estate, it is relatively straightforward to compute inequality estimates at any desired level of aggregation. Thus, if housing and income inequality are highly correlated and addressing a research or policy question requires granularity of the data, focusing on housing may be the best way around it.

It will also be critical to study whether the spatial patterns revealed in Belgium, Barcelona and Massachusetts also appear in other contexts. As discussed in the last sub-section, housing inequality is largely unexplored, and therefore more work is needed to see whether the previous findings are truly general or specific to the Western cities analysed.

Most importantly, from a policy perspective, it is paramount to fully understand the causes and consequences of housing inequality and how policy can affect it. We already have a good idea about the causes. At the core, the underlying mechanism driving housing inequality is sorting. Households and firms choose where they live or operate, which mechanically creates inequality across regions and within cities – for example, households sort across cities and neighbourhoods based on income or skill. Moreover, the interaction between these factors in the physical space creates externalities (e.g., agglomeration economies) that may reinforce or reverse inequality dynamics (Baum-Snow and Pavan, 2013[31]; Behrens, Duranton and Robert-Nicoud, 2014[32]; Behrens and Robert-Nicoud, 2015[33]; Fogli and Guerrieri, 2019[34]; Glaeser, Resseger and Tobio, 2009[35]; Puga, 2010[36]). We do not know as much about the consequences of housing inequality, but it is well-known that segregation and consuming poor-quality or overcrowded housing are linked with several negative outcomes (Akbar et al., 2022[37]; Ananat, 2011[38]; Billings, Deming and Rockoff, 2013[39]; Currie and Yelowitz, 2000[40]; Cutler and Glaeser, 1997[41]; Goux and Maurin, 2005[42]; Shertzer and Walsh, 2019[43]). Policy interventions may affect housing inequality and its dynamics in different ways. The last section of this chapter provides a short overview of several widely used housing policies and studies the effects of a recent reform in Belgium.7

Housing allowances are transfers to low-income households (typically tenants) to assist them in affording housing expenses. Commonly, they are progressive, means-tested, and not conditional on residing in some pre-determined area (in contrast to housing vouchers).

Some studies estimate that housing allowances reduce post-tax as well as transfer income inequality (Bozio et al., 2015[44]; 2018[45]). However, their impact is not as large as it could be as landlords capture a fraction of the subsidy in the form of higher rents (Fack, 2006[46]; Gibbons and Manning, 2006[47]; Susin, 2002[48]). The exact impact of this policy (and all policies discussed in this section) critically depends on the elasticities of housing supply and demand (Eerola et al., 2022[49]; Eriksen and Ross, 2015[50]).

Housing vouchers are similar to housing allowances, but they are conditional on leasing or purchasing affordable housing in specific (good) neighbourhoods. The Moving to Opportunity program implemented in 1994 in the United States is probably the most prominent and well-studied example of this policy.

Research suggests that vouchers reduce rent burdens, overcrowding and homelessness, while also being effective in keeping or moving disadvantaged households to good neighbourhoods, thereby allowing these households to benefit from all the advantages associated with living in better neighbourhoods. Vouchers also reduce housing consumption inequality (Chetty, Hendren and Katz, 2016[1]; Chyn, 2018[51]; Katz, Kling and Liebman, 2001[52]; Ludwig et al., 2013[53]). Scalability and programme take-up are areas with room for improvement (Ellen, 2020[54]).

Mortgage interest deductions (MID) allow households to deduct mortgage payments from their taxes. MID policies are common in many OECD countries, and one of their stated goals is to increase homeownership.

MID policies are regressive as they target homeowners – individuals in the middle and upper parts of the income and wealth distribution (Poterba and Sinai, 2008[55]). Moreover, research suggests MID policies tend to increase home prices without significantly affecting homeownership rates (Damen and Goeyvaerts, 2021[56]; Gruber, Jensen and Kleven, 2021[57]; Hilber and Turner, 2014[58]). With little effect on homeownership, MID policies increase post-transfer income and wealth inequalities.

Rental investment subsidies are tax deductions for landlords purchasing or investing in new housing to be rented on the market as affordable units for low and middle-income tenants. The most prominent example of such a policy is the Low-Income Housing Tax Credit (LIHTC) in the United States.

Evidence from the United States suggests that the LIHTC effectively keeps lower-income households in better neighbourhoods (Diamond and McQuade, 2019[59]), thereby reducing housing consumption inequality. However, because these policies are transfers to homeowners, housing wealth inequality may increase.

Rent control is a policy that aims to make rental units affordable to keep low and middle-income renters in good neighbourhoods. This policy has many versions: from absolute price ceilings affecting all dwellings in a city or a neighbourhood to restrictions on the growth rate of rents that vary with local or dwelling characteristics (e.g., their age).

The effects of rent control depend on the characteristics of the policy, but most evidence finds predominantly negative impacts on several dimensions. For example, tenants become less mobile, and owners become less likely to provide units in the regulated rental market – and when they do, they tend to invest less in dwelling maintenance (Diamond, McQuade and Qian, 2019[60]; Diamond and McQuade, 2019[59]; Glaeser and Luttmer, 2003[61]; Sims, 2007[62]). Rent control policies allow a few households to significantly increase housing consumption (as they remain in good or gentrifying neighbourhoods). However, rents in the unregulated market increase, thus harming most renters, benefiting landlords, and increasing inequality (Diamond, McQuade and Qian, 2019[60]).

Housing transaction fees are levies on the parties transferring real estate. Unlike the previous policies, which regulated or subsidised housing consumption, housing transaction fees’ primary purpose is typically tax revenue collection. They exist in many OECD countries.

Research suggests that transaction taxes are highly distortionary, affecting house prices, sales volumes, and therefore, homeownership rates (Besley, Meads and Surico, 2014[63]; Best and Kleven, 2017[64]; Han, Ngai and Sheedy, 2022[10]). Since homeowners have relatively higher incomes and wealth, these policies can reduce post-tax and transfer income inequality. However, as they may discourage some households from becoming homeowners, they can exacerbate housing wealth inequality. This paper’s following and last sub-section discusses the effects of a recent housing transaction fee reform in Flanders.

DGM (2022[6]) study the effects of a 3 percentage point reduction in home registration fees in the Belgian region of Flanders. On January 1, 2022, the reform was introduced in Flanders but not in the regions of Brussels and Wallonia. Using a differences-in-differences framework, they find that the 3 percentage point reduction in registration fees caused an increase in house prices of almost 3%. This result is in line with findings from similar policies in other contexts (Besley, Meads and Surico, 2014[63]; Best and Kleven, 2017[64]; Han, Ngai and Sheedy, 2022[10]).

The reform compressed the dwelling value distribution. The paper finds that the increase in home prices was more pronounced at the bottom of the dwelling value distribution, with homes at the bottom decile appreciating as much as 7%. Homes above the median value only experienced a negligible change in prices. The authors estimate that the compression in the dwelling value distribution decreased housing inequality in Flanders by 0.8%.

The geographic heterogeneity effects of the policy (a dimension not often studied in the literature) were substantial. Given the uneven distribution of dwellings of different values in Flanders and the rest of Belgium, the reform affected housing value inequality differently across geographies. Figure ‎8.6 illustrates the estimated changes in inequality across varying sub-national levels. Inequality decreased more in the east and west of Flanders. At the province level, the authors estimate the reform reduced housing inequality by more than 2% in Limbourg and between 1 and 2% in West Flanders. The image is more nuanced when going below the province level. At the district level, the same spatial patterns persist at the country's borders, but the maps reveal a slight increase in inequality (less than 1%) in the district (arrondissement) of Halle-Vilvoorde – the district immediately adjacent to Brussels. At the municipality or statistical sector level, the number of administrative units with small increases in housing inequality is much larger. It is worth noting, however, that most of these "increases" in inequality are negligible and not statistically different from zero (particularly at the statistical sector level).

The authors discuss two implications of their results. First, the likely winners of the policy were low-value homeowners – who saw the value of their real estate (and therefore wealth) increase thanks to the reform. Second, and this is just speculation, the reform might have reduced wealth inequality if a fraction of liquidity-constrained wishing-to-be homeowners acquired a property thanks to the reduction in transaction fees (part of a home down payment).

A shortcoming of the study is that the authors cannot make factual statements about housing wealth inequality. In that regard, they can only speculate. They were unable to access ownership information in the cadastral data. Without such information, their analysis was limited to discussions on housing value and space inequality, thus providing only a partial picture of two dimensions of housing inequality.

Nevertheless, their analysis illustrates well the important extent of local heterogeneities induced by a regional housing policy and makes the local nature of housing salient. For policymakers, it provides valuable insights about the spatial scope of policies they might not have previously considered, and it opens the door for the consideration of compensatory mechanisms following the identification of (local) winners and losers from the policy.

The chapter reviewed some of the existing housing inequality estimates, with a strong emphasis on recent evidence for Belgium. Housing inequality in Belgium is relatively low (e.g., compared to estimates from the United States), but heterogeneity across subnational entities is substantial. Within cities, housing inequality tends to be larger in downtowns and lower in residential areas, a pattern that emerged in Brussels, Barcelona and Boston. More cross and within-country evidence are required, but housing and income inequality appear to be highly correlated.

The chapter also reviewed some of the most common housing policies and discussed their impact on housing inequality. Housing allowances, housing vouchers and rental investment subsidies are among the most promising avenues to reduce housing (consumption) inequality, although the latter may also exacerbate housing wealth inequality. Finally, a recent reduction in housing registration fees in Flanders increased housing prices in that region and reduced housing value inequality. The impact of the policy was highly heterogeneous across sub-national units, thus highlighting how all housing policies involve a strong local component. Therefore, enhancing the granularity of housing and income inequality estimates may assist governments in improving their policies' design, targeting and implementation.

Finally, as the OECD’s (2021[65]) Brick-by-Brick report argues, and the Belgian case study makes clear, regularly updating cadastral values is important for fairness, avoiding distortions and keeping track of housing inequality.


[37] Akbar, P. et al. (2022), “Racial Segregation in Housing Markets and the Erosion of Black Wealth”, The Review of Economics and Statistics, pp. 1-45, https://doi.org/10.1162/rest_a_01276.

[8] Aladangady, A., D. Albouy and M. Zabek (2017), “Housing inequality”, NBER Working Paper Series, No. 21916 (Revised), https://dornsife.usc.edu/assets/sites/1003/docs/D._Albouy_Paper.pdf.

[9] Albouy, D. and M. Zabek (2016), Housing Inequality, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/w21916.

[26] Alvaredo, F. et al. (2020), Distributional National Accounts Guidelines, Methods and Concepts Used in the World Inequality Database, https://hal.archives-ouvertes.fr/hal-03307274.

[27] Alvaredo, F. et al. (2022), World Inequality Database, https://wid.world/.

[38] Ananat, E. (2011), “The Wrong Side(s) of the Tracks: The Causal Effects of Racial Segregation on Urban Poverty and Inequality”, American Economic Journal: Applied Economics, Vol. 3/2, pp. 34-66, https://doi.org/10.1257/app.3.2.34.

[23] André, M. and Meslin (2021), “Et pour quelques appartements de plus: Étude de la propriété immobilière des ménages et du profil redistributif de la taxe foncière”, Document de Travail, Insee, https://www.bnsp.insee.fr/ark:/12148/bc6p07c0xkt.pdf.

[31] Baum-Snow, N. and R. Pavan (2013), “Inequality and City Size”, The Review of Economics and Statistics, Vol. 95/5, pp. 1535-1548, https://doi.org/10.1162/rest_a_00328.

[32] Behrens, K., G. Duranton and F. Robert-Nicoud (2014), “Productive Cities: Sorting, Selection, and Agglomeration”, Journal of Political Economy, Vol. 122/3, pp. 507-553, https://doi.org/10.1086/675534.

[33] Behrens, K. and F. Robert-Nicoud (2015), “Agglomeration Theory with Heterogeneous Agents”, in Handbook of Regional and Urban Economics, Elsevier, https://doi.org/10.1016/b978-0-444-59517-1.00004-0.

[63] Besley, T., N. Meads and P. Surico (2014), “The incidence of transaction taxes: Evidence from a stamp duty holiday”, Journal of Public Economics, Vol. 119, pp. 61-70, https://doi.org/10.1016/j.jpubeco.2014.07.005.

[64] Best, M. and H. Kleven (2017), “Housing Market Responses to Transaction Taxes: Evidence From Notches and Stimulus in the U.K.”, The Review of Economic Studies, Vol. 85/1, pp. 157-193, https://doi.org/10.1093/restud/rdx032.

[39] Billings, S., D. Deming and J. Rockoff (2013), “School Segregation, Educational Attainment, and Crime: Evidence from the End of Busing in Charlotte-Mecklenburg”, The Quarterly Journal of Economics, Vol. 129/1, pp. 435-476, https://doi.org/10.1093/qje/qjt026.

[45] Bozio, A. et al. (2018), “Inequality and redistribution in France”, 1990-2018: evidence from post-tax distributional national accounts (dina). WID. World Working Paper, No. 10, http://www.ecineq.org/ecineq_paris19/papers_EcineqPSE/paper_58.pdf.

[44] Bozio, A. et al. (2015), “Reforming French housing benefits: why not merging benefits?”, Notes IPP, No. 18, https://shs.hal.science/halshs-02539844/.

[28] Chancel, L. et al. (2021), World inequality report 2022.

[67] Chapelle, G., G. Domènech-Arumí and P. Gobbi (2023), “Housing, Neighborhoods and Inequality”, CEPR Discussion Papers, No. DP17969, https://cepr.org/publications/dp17969.

[2] Chetty, R. and N. Hendren (2018), “The Impacts of Neighborhoods on Intergenerational Mobility I: Childhood Exposure Effects*”, The Quarterly Journal of Economics, Vol. 133/3, pp. 1107-1162, https://doi.org/10.1093/qje/qjy007.

[3] Chetty, R. and N. Hendren (2018), “The Impacts of Neighborhoods on Intergenerational Mobility II: County-Level Estimates*”, The Quarterly Journal of Economics, Vol. 133/3, pp. 1163-1228, https://doi.org/10.1093/qje/qjy006.

[1] Chetty, R., N. Hendren and L. Katz (2016), “The Effects of Exposure to Better Neighborhoods on Children: New Evidence from the Moving to Opportunity Experiment”, American Economic Review, Vol. 106/4, pp. 855-902, https://doi.org/10.1257/aer.20150572.

[51] Chyn, E. (2018), “Moved to Opportunity: The Long-Run Effects of Public Housing Demolition on Children”, American Economic Review, Vol. 108/10, pp. 3028-3056, https://doi.org/10.1257/aer.20161352.

[4] Chyn, E. and L. Katz (2021), “Neighborhoods Matter: Assessing the Evidence for Place Effects”, Journal of Economic Perspectives, Vol. 35/4, pp. 197-222, https://doi.org/10.1257/jep.35.4.197.

[40] Currie, J. and A. Yelowitz (2000), “Are public housing projects good for kids?”, Journal of Public Economics, Vol. 75/1, pp. 99-124, https://doi.org/10.1016/s0047-2727(99)00065-1.

[41] Cutler, D. and E. Glaeser (1997), “Are Ghettos Good or Bad?”, The Quarterly Journal of Economics, Vol. 112/3, pp. 827-872, https://doi.org/10.1162/003355397555361.

[56] Damen, S. and G. Goeyvaerts (2021), “Housing market responses to the mortgage interest deduction”, SSRN Electronic Journal, https://doi.org/10.2139/ssrn.3854883.

[59] Diamond, R. and T. McQuade (2019), “Who Wants Affordable Housing in Their Backyard? An Equilibrium Analysis of Low-Income Property Development”, Journal of Political Economy, Vol. 127/3, pp. 1063-1117, https://doi.org/10.1086/701354.

[60] Diamond, R., T. McQuade and F. Qian (2019), “The Effects of Rent Control Expansion on Tenants, Landlords, and Inequality: Evidence from San Francisco”, American Economic Review, Vol. 109/9, pp. 3365-3394, https://doi.org/10.1257/aer.20181289.

[17] Diamond, R. and E. Moretti (2021), Where is Standard of Living the Highest? Local Prices and the Geography of Consumption, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/w29533.

[7] Domènech-Arumí, G. (2021), “Neighborhoods, Perceived Inequality, and Preferences for Redistribution: Evidence from Barcelona”, SSRN Electronic Journal, https://doi.org/10.2139/ssrn.3863594.

[6] Domènech-Arumí, G., P. Gobbi and G. Magerman (2022), “Housing inequality and how fiscal policy shapes it: Evidence from Belgian real estate”, National Bank of Belgium Working Paper, No. 423, https://ideas.repec.org/p/nbb/reswpp/202210-423.html.

[5] Durlauf, S. (2004), “Neighborhood effects”, in Handbook of Regional and Urban Economics, Cities and Geography, Elsevier, https://doi.org/10.1016/s1574-0080(04)80007-5.

[49] Eerola, E. et al. (2022), The Incidence of Housing Allowances: Quasi-Experimental Evidence, https://www.doria.fi/handle/10024/185961.

[54] Ellen, I. (2020), “What do we know about housing choice vouchers?”, Regional Science and Urban Economics, Vol. 80, p. 103380, https://doi.org/10.1016/j.regsciurbeco.2018.07.003.

[50] Eriksen, M. and A. Ross (2015), “Housing Vouchers and the Price of Rental Housing”, American Economic Journal: Economic Policy, Vol. 7/3, pp. 154-176, https://doi.org/10.1257/pol.20130064.

[46] Fack, G. (2006), “Are housing benefit an effective way to redistribute income? Evidence from a natural experiment in France”, Labour Economics, Vol. 13/6, pp. 747-771, https://doi.org/10.1016/j.labeco.2006.01.001.

[34] Fogli, A. and V. Guerrieri (2019), “The end of the American dream? Inequality and segregation in US cities”, Becker Friedman Institute for Economics at the University of Chicago, https://bfi.uchicago.edu/insight/research-summary/the-end-of-the-american-dream-inequality-and-segregation-in-us-cities/.

[21] Frick, J. and M. Grabka (2003), “Imputed Rent and Income Inequality: A Decomposition Analysis for Great Britain, West Germany and the U.S.”, Review of Income and Wealth, Vol. 49/4, pp. 513-537, https://doi.org/10.1111/j.0034-6586.2003.00102.x.

[20] Frick, J. et al. (2010), “Distributional effects of imputed rents in five European countries”, Journal of Housing Economics, Vol. 19/3, pp. 167-179, https://doi.org/10.1016/j.jhe.2010.06.002.

[12] Garbinti, B., J. Goupille-Lebret and T. Piketty (2020), “Accounting for Wealth-Inequality Dynamics: Methods, Estimates, and Simulations for France”, Journal of the European Economic Association, Vol. 19/1, pp. 620-663, https://doi.org/10.1093/jeea/jvaa025.

[47] Gibbons, S. and A. Manning (2006), “The incidence of UK housing benefit: Evidence from the 1990s reforms”, Journal of Public Economics, Vol. 90/4-5, pp. 799-822, https://doi.org/10.1016/j.jpubeco.2005.01.002.

[61] Glaeser, E. and E. Luttmer (2003), “The Misallocation of Housing Under Rent Control”, American Economic Review, Vol. 93/4, pp. 1027-1046, https://doi.org/10.1257/000282803769206188.

[35] Glaeser, E., M. Resseger and K. Tobio (2009), “INEQUALITY IN CITIES”, Journal of Regional Science, Vol. 49/4, pp. 617-646, https://doi.org/10.1111/j.1467-9787.2009.00627.x.

[42] Goux, D. and E. Maurin (2005), “The effect of overcrowded housing on children’s performance at school”, Journal of Public Economics, Vol. 89/5-6, pp. 797-819, https://doi.org/10.1016/j.jpubeco.2004.06.005.

[57] Gruber, J., A. Jensen and H. Kleven (2021), “Do People Respond to the Mortgage Interest Deduction? Quasi-Experimental Evidence from Denmark”, American Economic Journal: Economic Policy, Vol. 13/2, pp. 273-303, https://doi.org/10.1257/pol.20170366.

[10] Han, L., L. Ngai and K. Sheedy (2022), “To Own or to Rent? The Effects of Transaction Taxes on Housing Markets”, Entre for Economic Policy Research, https://personal.lse.ac.uk/ngai/HanNgaiSheedy.pdf.

[58] Hilber, C. and T. Turner (2014), “The Mortgage Interest Deduction and its Impact on Homeownership Decisions”, Review of Economics and Statistics, Vol. 96/4, pp. 618-637, https://doi.org/10.1162/rest_a_00427.

[24] Huang, Y. and L. Jiang (2009), “Housing Inequality in Transitional Beijing”, International Journal of Urban and Regional Research, Vol. 33/4, pp. 936-956, https://doi.org/10.1111/j.1468-2427.2009.00890.x.

[13] Kaas, Kocharkov and Preugschat (2019), “Wealth Inequality and Homeownership in Europe”, Annals of Economics and Statistics 136, p. 27, https://doi.org/10.15609/annaeconstat2009.136.0027.

[52] Katz, L., J. Kling and J. Liebman (2001), “Moving to Opportunity in Boston: Early Results of a Randomized Mobility Experiment”, The Quarterly Journal of Economics, Vol. 116/2, pp. 607-654, https://doi.org/10.1162/00335530151144113.

[15] Kuhn, M., M. Schularick and U. Steins (2020), “Income and wealth inequality in America, 1949-2016”, Journal of Political Economy, Vol. 128/9, pp. 3469-3519.

[25] Logan, J., Y. Bian and F. Bian (1999), “Housing inequality in urban China in the 1990s”, International Journal of Urban and Regional Research, Vol. 23/1, pp. 7-25, https://doi.org/10.1111/1468-2427.00176.

[53] Ludwig, J. et al. (2013), “Long-Term Neighborhood Effects on Low-Income Families: Evidence from Moving to Opportunity”, American Economic Review, Vol. 103/3, pp. 226-231, https://doi.org/10.1257/aer.103.3.226.

[14] Martínez-Toledano, C. (2020), “House price cycles, wealth inequality and portfolio reshuffling”, WID. World Working Paper, No. 2,8.

[18] Moretti, E. (2013), “Real Wage Inequality”, American Economic Journal: Applied Economics, Vol. 5/1, pp. 65-103, https://doi.org/10.1257/app.5.1.65.

[11] OECD (2022), OECD Affordable Housing Database, https://www.oecd.org/housing/data/affordable-housing-database.

[19] OECD (2022), OECD Income Inequality Database, https://data.oecd.org/inequality/income-inequality.htm.

[65] OECD (2021), Brick by Brick: Building Better Housing Policies, OECD Publishing, Paris, https://doi.org/10.1787/b453b043-en.

[29] Piketty, T. and E. Saez (2003), “Income Inequality in the United States, 1913-1998”, The Quarterly Journal of Economics, Vol. 118/1, pp. 1-41, https://doi.org/10.1162/00335530360535135.

[22] Piketty, T., E. Saez and G. Zucman (2017), “Distributional National Accounts: Methods and Estimates for the United States”, The Quarterly Journal of Economics, Vol. 133/2, pp. 553-609, https://doi.org/10.1093/qje/qjx043.

[55] Poterba, J. and T. Sinai (2008), “Tax Expenditures for Owner-Occupied Housing: Deductions for Property Taxes and Mortgage Interest and the Exclusion of Imputed Rental Income”, American Economic Review, Vol. 98/2, pp. 84-89, https://doi.org/10.1257/aer.98.2.84.

[36] Puga, D. (2010), “The Magnitude and Causes of Agglomeration Economies”, Journal of Regional Science, Vol. 50/1, pp. 203-219, https://doi.org/10.1111/j.1467-9787.2009.00657.x.

[66] Sharkey, P. and J. Faber (2014), “Where, When, Why, and For Whom Do Residential Contexts Matter? Moving Away from the Dichotomous Understanding of Neighborhood Effects”, Annual Review of Sociology, Vol. 40/1, pp. 559-579, https://doi.org/10.1146/annurev-soc-071913-043350.

[43] Shertzer, A. and R. Walsh (2019), “Racial Sorting and the Emergence of Segregation in American Cities”, The Review of Economics and Statistics, Vol. 101/3, pp. 415-427, https://doi.org/10.1162/rest_a_00786.

[62] Sims, D. (2007), “Out of control: What can we learn from the end of Massachusetts rent control?”, Journal of Urban Economics, Vol. 61/1, pp. 129-151, https://doi.org/10.1016/j.jue.2006.06.004.

[30] Solt, F. (2016), “The Standardized World Income Inequality Database”, Social Science Quarterly, Vol. 97/5, pp. 1267-1281, https://doi.org/10.1111/ssqu.12295.

[48] Susin, S. (2002), “Rent vouchers and the price of low-income housing”, Journal of Public Economics, Vol. 83/1, pp. 109-152, https://doi.org/10.1016/s0047-2727(01)00081-0.

[16] Zucman, G. (2019), “Global Wealth Inequality”, Annual Review of Economics, Vol. 11/1, pp. 109-138, https://doi.org/10.1146/annurev-economics-080218-025852.


← 1. See Sharkey and Faber (2014[66]) or Chyn and Katz (2021[4]) for recent reviews.

← 2. Households have homothetic preferences if they consume the same basket of goods in the same proportion, independently of their income.

← 3. With some caveats. Namely, demolitions and property modifications (e.g., parcel partitions) might be an issue, especially when intending to extrapolate to long time horizons.

← 4. A census tract in Boston contains roughly 4 000 people and is one of the smallest administrative units in the United States Census.

← 5. According to EU-SILC data, the income Gini index for the Brussels Capital Region was 0.345 in 2022, substantially higher than Flanders (0.226) and Wallonia (0.242).

← 6. They compare the Income Inter-quantile Range (IQR) at the statistical sector level (the only measure of income inequality at that level of aggregation) with the housing value IQR.

← 7. See Chapelle et al. (2023[67]) for a more thorough description of these policies and their incidence on homeowners and renters.

Legal and rights

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.

© OECD/KIPF 2023

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at https://www.oecd.org/termsandconditions.