How large is Spain's public housing stock? Tracing the official figures and the EU averages they are compared with (2019–2026)

Abstract

Nearly every policy text describes Spain's public housing stock as small, but the texts do not agree on how small. We collected every figure we could find in official Spanish and EU documents for the size of Spain's public, social or protected housing stock, and every EU average set beside it: 32 figures for Spain and 14 EU averages, from 25 documents published between 2019 and 2026. For each we recorded:

• the source and the date;
• the reference year;
• the definition, the unit and the denominator.

We checked all 46 quotations literally against the downloaded texts. We then traced each figure as far as the documents allow and tried to reproduce it.

The counts. All counts of Spain's public stock come from two surveys by the Ministry's Observatorio de Vivienda y Suelo (OVS) of the dwellings that regions and municipalities own and let: 290,000 dwellings (2019 survey) and 318,000 (2023 survey).

The shares.

• 1% to 1.7% follow from dividing these counts by households or by all dwellings.
• 2.5%, 3.3%, 3.4% and 3.5% each equal a year of INE's living-conditions survey: the share of households that pay a rent below the market price, whoever the landlord.

The EU averages.

• 9%: the Ministry's 2020 average for the EU-28, including the United Kingdom.
• 8.0%: its 2024 bulletin's average for the EU-27.
• 7%: the OECD's unweighted mean over the countries it plots.

The reconciliation. We list 17 differences between figures and count one as explained only when every number in it reproduces. 7 are explained, 6 partly explained, 2 are errors in a source and 2 are not explained. The two not explained are the 9% in the 2026–2030 Plan, which gives no source, and a possible double count of Ceuta and Melilla.

Art. 32. Art. 32 of the 2023 housing law requires the State to inventory the housing it owns and to publish an annual report on it. We found none:

• in the Ministry's electronic office, which the law names;
• at CASA 47, the State housing entity;
• in 1,054 BOE issues;
• in the national open-data catalogue or the Ministry's publications.

No national total includes the State's own stock; only partial figures have been published.

1. Introduction

That Spain has very little public housing is a premise shared by:

• the 2023 housing law and the 2026–2030 State Housing Plan;
• the European Commission's country reports and the Council's recommendations to Spain;
• the OECD and the Banco de España.

The numbers attached to it vary. The housing law's preamble says 290,000 dwellings, 1.6% of households. The Plan says 318,000, 1.7% of households, and adds that «todo el parque de alquiler social en España (inferior a precio de mercado) suma el 3,5 %, situando a España lejos de la media de la Unión Europea del 9 %». Less than three months later, the recitals of the Council's recommendation to Spain stated that its social housing represents «less than 2 % of the total housing supply compared to an EU average of roughly 7 %».

Read together, the figures look inconsistent: two counts, about ten shares and four EU averages. We ask:

1. What official figures exist, and what does each measure (definition, unit, denominator, reference year)?
2. Where does each come from?
3. Which differences between them are explained by definitions, years or denominators, and which are not?
4. Does the annual report on the public housing stock required by art. 32 of Ley 12/2023 exist?

We describe the figures and their origins. We do not assess housing policy. A figure that disagrees with another is a fact to describe, not evidence of intent.

2. Prior work

OECD. The Affordable Housing Database (indicator PH4.2) is the reference for cross-country comparison. It warns that «the definition of social housing, as well as the size of the sector, varies considerably across countries». For Spain it notes that «the figures may also contain other types of reduced rent housing, e.g. employer-provided dwellings». The press reports that the OECD's 2025 Economic Survey of Spain puts social rent at «tan solo el 3% del parque inmobiliario». We could not open the survey.

European Commission. SWD(2025) 1053, which accompanies the Affordable Housing Plan, says the OECD database «relies on national definitions that differ substantially». It adds that «Specific social housing data is not collected by Eurostat».

Ministry of Housing. The Observatory's bulletins (2020, 2024) are the primary sources of the Spanish counts. They state their survey method but do not discuss other official figures.

Scholarly work. We ran 10 queries on Crossref and OpenAlex (about 450 titles screened). The closest works:

• Trilla and Bosch, El parque público y protegido de viviendas en España: un análisis desde el contexto europeo (Fundación Alternativas, Documento de trabajo, 2018). It examines the history of protected housing and estimates needs. Its site answered us with a CAPTCHA, so we read only the abstract.
• Pareja-Eastaway and Sánchez-Martínez, More Social Housing? A Critical Analysis on Social Housing Provision in Spain (Critical Housing Analysis, 2017).
• García-Montalvo, La vivienda en España: diagnóstico y propuestas desde una perspectiva internacional (Ekonomiaz, 2026).
• IKEI and Corral, on the Basque public rental stock (Ekonomiaz, 2010).
• A 2026 master's thesis on the loss of protected status (Badia del Vallès).

Judging by their abstracts, none of these compares the official figures with each other. Site searches of Funcas and Housing Europe found no reconciliation either; Funcas's article «La vivienda social» gives no figures in its web page.

Fact-checkers.

• Newtral, 22 November 2018, rated «Verdadero» a statement by the then minister that social housing «representa apenas el 2,5% del parque de nuestro país». It cited the European housing observatory (Housing Europe).
• Newtral, 17 January 2023, repeated the Ministry's 2020 EU table (Spain 2.5%).

Press. We read 11 articles found through Bing News RSS and one headline. Datelines are the articles' own.

• El País, 29 September 2026, gives 318,000 as «apenas un 1,7% de las viviendas principales y un 1,3% del parque total». It adds that the comparison is «complicada, porque la definición de vivienda social varía en cada país y las cifras cambian ligeramente según la fuente». Over the Ministry's 27.0 million dwellings, 318,000 is 1.2%.
• La Gaceta de Canarias, 4 October 2026, notes that national sources differ «con metodología y referencia distintas». It puts the OECD's figure for Spain at «alrededor del 1,2%»; the OECD's current table gives 1.13%.
• Others repeat a figure. EFE Verifica (30 September 2026: 1.7% against the OECD's «7,1 %» for the EU), Público (3 October 2026: 1.7% against 7%), El Periódico (28 July 2026, headline only: 1.72% against 8%), La Marea (8 January 2025: 2.5% against 9.3%).
• Cadena SER, 3 October 2026, attributes to the Banco de España a share equal «al 1,5 % de la oferta total», the same mismatch as the 2025 EU texts (§4.5).
• Two articles describe the survey share as public housing.
  • El País, 7 May 2026: «cerca de 600.000 viviendas públicas de alquiler, lo cual supone el 3,3% de todas las casas».
  • La Vanguardia, 28 September 2025: «un 3,3% del stock actual está protegido».

As §4.3 shows, 3.3% is the share of households paying a below-market rent.

What we add is narrow:

1. the quoted and dated trail of each figure;
2. a reconciliation in which each "explained" is checked by recomputation;
3. the errors and mismatches we found in the sources;
4. the record of the search for the art. 32 report.

That definitions differ was already known.

3. Data and method

Full detail is in METHOD.md.

Sources.

• Spanish law: the consolidated texts of Ley 12/2023 and the state housing plans (BOE open-data API), and two Constitutional Court rulings.
• The Ministry: the OVS Boletín Especial Vivienda Social 2020 and 2024.
• CASA 47: its planning documents and two web pages.
• Statistics: INE's living-conditions survey (ECV) and Eurostat's EU-SILC tenure table; the OECD database (PDF and workbook, updated 24 November 2025).
• Others: Housing Europe's reports (2019, 2021, 2025); two Banco de España publications.
• EU texts (Cellar): the Council's recommendations to Spain (2019–2026), the Commission's country reports on Spain (2019–2026), the European Affordable Housing Plan and SWD(2025) 1053.

Every request used our User-Agent, read robots.txt first and kept to one request per second per host.

Figures table. One row per figure:

• the value as printed and its unit (dwellings, % of households, % of principal dwellings, % of all dwellings, % of persons);
• one of nine concept classes:
  • A: publicly owned rental stock;
  • B: below-market rent, households;
  • C: reduced-price or free rent, persons;
  • D: Housing Europe national estimates;
  • E: OECD definition;
  • F: protected housing including sale;
  • G: attributed figure that does not match its source;
  • H: no number;
  • I: State-side partial figure;
• the source the document itself cites;
• a literal quotation.

A script checks every quotation against the raw text.

Recomputation. From the Ministry's bulletins we extracted the EU, regional and municipal tables and recomputed:

• their totals;
• their EU averages, with and without the United Kingdom;
• the scaling of the municipal stock.

We also recomputed the OECD's averages from its workbook, and every Spanish share over every available denominator: the Ministry's stock estimate (all and principal dwellings, 2019–2025), Census households, and the households the law states.

Reconciliation. Each difference between figures gets:

• a category: denominator, reference year, definition, scope, method, attribution, arithmetic, or unexplained;
• a status: explained, partly explained, not explained, or error in a source.

A difference counts as explained only if every number in it reproduces. If the concept is clear but a number does not reproduce, it is partly explained.

Art. 32. We read the article in its consolidated version, the law's amendment record and two Constitutional Court rulings. Then we searched for the report:

• in the Ministry's electronic office, the place art. 32.2 names;
• at CASA 47;
• in every BOE issue from the law's entry into force to 3 October 2026 (1,054 issues over 1,227 days, 242,047 item titles);
• in datos.gob.es, the publications centre's housing catalogue (234 titles) and the Transparency Portal;
• in the press, and in the text of the Ministry's 2024 social-housing bulletin.

4. Results

4.1 Two counts, both from Ministry surveys

The housing law's preamble repeats the 2020 bulletin's paragraph almost word for word. It cites «las últimas estimaciones del Observatorio de Vivienda y Suelo» and defines the stock as «exclusivamente, la vivienda en alquiler de titularidad pública». The Plan of April 2026 uses the 2023 survey without naming it. Its preamble describes the stock as «de ámbito autonómico y municipal».

Between the two surveys the regions' rental stock went from 178,493 to 196,842. The 2023 regional table has a column absent in 2019: 34,920 dwellings «sujetas a modelos de colaboración público-privada», built on public land under surface rights or concessions. Art. 3 of Ley 12/2023 counts these as social housing. The bulletin does not say whether the 2019 survey counted such dwellings as rental, so part of the change may be scope rather than new dwellings.

4.2 Same counts, different denominators

Spain had 1.4 dwellings per household in 2024: the Ministry's stock estimate gives 27,004,765 dwellings, 19,278,057 of them principal. A share of households is therefore about 40% larger than the same count as a share of all dwellings.

Two figures do not reproduce exactly:

• The OECD's 1.13% implies 25.6 million dwellings, a stock figure we did not find in any source. Housing Europe 2021 prints 25,793,323, in a cell labelled 2020 with a footnote reading «2019 data».
• The Commission's «1.5-1.7% of households» (2026). Its lower bound is not reproduced by 318,000 over any of the Ministry's principal-dwelling counts for 2019–2025, which give 1.62–1.72%. It equals the Banco de España's share for about 300,000.

4.3 2.5%, 3.3%, 3.4% and 3.5% are a different measure

The 2024 bulletin adds a figure «en sentido amplio»: social rental «representa un 3,3% del parque de viviendas principales u hogares». Its EU table explains: «En el caso de España el porcentaje de alquiler social se ha estimado a partir de la Encuesta de condiciones de vida del INE». The Plan updates the figure to 3.5% for 2025.

The ECV asks households about their tenure. Its category is a rent below the market price, whoever the landlord: a public agency, a private landlord, an employer or an old controlled lease. INE's series:

• 2.5% in every year from 2012 to 2016;
• 2.7% (2019), 3.3% (2023), 3.4% (2024) and 3.5% (2025).

It excludes «Cesión», free accommodation, at 6.5% in 2025.

Four other figures equal values of this series, although their documents do not name it:

• 2.5%: Spain's cell in the Ministry's 2020 EU table, and Housing Europe 2019's Spain profile («of total occupied housing» stock). The 2024 edition of the same table fills the cell from the ECV and says so.
• 2.5–3.4%: the Commission's 2026 range (2012–2016 and 2024 values). The Commission describes it as «all protected or regulated housing (rental and sale)».
• 3.4%: Housing Europe 2025's «around 3.4%», attributed to Spanish government estimates and described as protected housing.

These figures are not counts of the public or protected stock.

Eurostat publishes the EU-wide version of the survey. Its category «Tenant, rent at reduced price or free» counts persons and includes free accommodation. On it, Spain was at 9.0% in 2025 and the EU-27 at 10.7%. In Spain most of that category is free accommodation, typically provided by family. It is not a measure of social housing and is not comparable with the stock-based EU averages.

4.4 Four EU averages

The two OVS averages differ by scope: removing the United Kingdom (4,627,402 social dwellings) from the 2020 table gives 8.0%, the 2024 result. The 2024 bulletin presents its 8.0% as the EU-27 average of that edition, not as a revision of the 2020 figure.

The OECD figures reproduce exactly from its workbook:

• EU: 7.99, the unweighted mean of the 19 EU members plotted in Figure PH4.2.1.
• OECD: 7.07, the unweighted mean of all 31 countries plotted, Colombia (0.006%) included.
• Weighted by dwellings, the same 19 EU countries give 6.8%.

The Commission's «6-7% of the EU housing stock» is close to that weighted figure, but the Commission does not say how it averaged. Its staff document pairs the range with «around 28 million homes». That is the total for all countries with recent OECD data (27.8 million recomputed), not for the EU.

4.5 What does not reconcile

The Plan's 9%. The 2026–2030 Plan writes «la media de la Unión Europea del 9 %» and gives no source. The 9% figures we could trace:

• the OVS 2020 EU-28 average (9.3%, which includes the United Kingdom);
• Eurostat's EU-27 share of persons renting at a reduced price or free in 2018–2019 (9.3% and 9.1%).

We cannot tell which, if either, the Plan used. The Observatory's 2024 bulletin, dated January 2025, gives 8.0% for the EU-27. In the same sentence, the Plan sets the 9% beside Spain's 3.5% from the ECV, a household tenure share. The 2024 bulletin's own EU table makes the same kind of comparison: Spain's 3.3% (ECV) beside other countries' Housing Europe estimates.

An attribution that does not match its source.

• The Commission's 2025 country report on Spain (SWD(2025) 209) says: «According to the Bank of Spain( 4 ), Spain’s social rental housing stock represents just 1.5% the total stock, compared to the EU average of 9%». Footnote 4 cites the Banco de España's El mercado del alquiler de vivienda residencial en España (2024).
• That paper gives the stock as «en torno al 1,5 % de las viviendas principales y el 1 % del stock total de viviendas». It puts the European and OECD average «en ambos casos, en torno al 7 % del parque de viviendas».
• The Council's recommendation of 8 July 2025, recital (34), repeats the attributed claim without a reference: «According to the Bank of Spain, the stock of social rental housing is only 1,5 % of the total housing stock, well below the EU average of 9 %».

We found no Banco de España text that gives 9%, and neither EU text states another origin for it.

The same Commission report has a second pair in its annex, «3.3% vs 8% EU average». These match the Ministry's 2024 figures. The report cites Provivienda for them, whose page is behind a CAPTCHA.

4.6 The Ministry's 2024 bulletin: arithmetic and method

Tabla 2.3 heads its second column «2024» for figures that the rest of the bulletin dates to 2023.

Growth of the regional rental stock. The text's «aumento del 5%» matches its printed total (178,493 to 187,252, +4.9%). Over all regions the rows give 196,842, +10.3%, but 34,920 of those are in the public-private column that is new in 2023. The public-ownership column alone, 161,922, is 9.3% below 2019. The like-for-like change therefore lies between −9.3% and +10.3%, depending on whether the 2019 survey counted public-private dwellings as rental. The bulletin does not say. The text's for-sale and «other» changes (−18% and +15%) use totals that omit Ceuta.

Ceuta and Melilla. The two autonomous cities appear with the same figures (1,941 rental dwellings together) in two places:

• the regional table, which feeds the 197,000;
• the municipal table, which feeds the scaled 121,000.

If both entered the total, the second count adds about 3,831 dwellings after scaling, 1.2% of 318,000. The bulletin does not publish the scaling, so we cannot tell whether it did.

The municipal component. The bulletin says it scaled «los datos aportados por un total de 168 municipios que representan 24,5 millones de habitantes». Our parse of its Tabla 2.8 finds:

• 409 municipalities of over 20,000 inhabitants;
• 181 municipalities with figures (some of them zeros): 62 from the 2023 survey and 119 carried over from 2019;
• together, 24.0 million people and 61,207 rental dwellings.

Scaled to Spain's population (47.4 million in the bulletin's table), that gives 120,805; with the bulletin's own 24.5 million base, 118,419. Our sample is therefore not exactly the one the bulletin describes. Either way about half of the municipal component, and about a fifth of the 318,000, is extrapolated. The extrapolation covers municipalities under 20,000 inhabitants, which were not surveyed.

4.7 The State's own stock and the art. 32 report

Article 32 of Ley 12/2023, «Parque público de vivienda», reads (in force since 26 May 2023, one version, no amendment):

1. El Estado, en cumplimiento del principio de colaboración y cooperación, así como en garantía del de trasparencia en la forma que se instrumente de acuerdo con las administraciones competentes, deberá elaborar y mantener actualizado un inventario del parque público de vivienda de su titularidad y de sus entes adscritos o dependientes, en el que se incluirá, al menos, la siguiente información: a) Identificación de las viviendas que lo componen. b) Características fundamentales de las viviendas. c) Situación de uso de las mismas. d) Características de las personas u hogares usuarios. 2. Anualmente se elaborará y publicará en la sede electrónica prevista en el artículo 38 de la Ley 40/2015, de 1 de octubre, de Régimen Jurídico del Sector Público, una memoria sobre las características del parque público de vivienda y su utilización, debidamente actualizada, que permita la conformación de un mapa de la vivienda social para facilitar su acceso por parte de la ciudadanía. Asimismo, se recogerá la información sobre las actuaciones previstas para reforzar el parque de vivienda social en los próximos cuatro años.

Scope. Paragraph 1 concerns the State's own stock: housing owned by the State and its attached or dependent entities. Paragraph 2 names no owner, so read alone «parque público de vivienda» and «mapa de la vivienda social» could suggest a wider scope. The Constitutional Court has read the article as concerning State-owned stock:

• STC 190/2025 describes the chapter as obliging the State «a proporcionar información periódica sobre cuestiones tales como el parque público de vivienda de titularidad estatal (art. 32)».
• STC 79/2024 records the State's pleading that «el art. 32 de la Ley 12/2023 se refiere a patrimonio público estatal». It records Andalusia's challenge to arts. 32–36, which it did not examine: «Por ello, los arts. 32 a 36 de la Ley 12/2023 quedan fuera del objeto de análisis de esta sentencia».

We follow that reading: the report concerns the stock owned by the State and its entities, such as CASA 47, the State housing entity.

Deadline. The article sets no date for the first report. It requires one «Anualmente» from the law's entry into force on 26 May 2023 (final provision 9).

What we found.

We found no art. 32 report or inventory by any route open to us, including the electronic office the law names. A report published only on the Ministry's main website, which blocks our client, cannot be excluded.

State-owned figures that do exist. No national total includes the State's own stock: the national counts (§4.1) cover regions and municipalities only. Partial figures have been published:

• Banco de España (Informe Anual 2023): 47,000 dwellings, citing figures for July 2023: «unas 47.000 viviendas para alquiler social disponibles o en desarrollo por parte del Ministerio de Vivienda y Agenda Urbana, además de 14.000 de la Sareb y unas 10.000 del Fondo Social de Viviendas en alquiler».
• CASA 47 (2025): 40,000 dwellings for transfer from Sareb: «se han identificado de forma preliminar más de 40.000 viviendas y cerca de 2.400 suelos».
• CASA 47 (7 September 2026): 816 dwellings; its two calls «suman 816 viviendas en alquiler asequible», and its portal «cuenta ya con más de 800 viviendas».

These are a programme pipeline, a planned transfer and the units offered on the State entity's portal. None is a count of the dwellings the State owns.

4.8 The reconciliation table

data/reconciliation.csv lists 17 differences between figures. 7 are explained, 6 partly explained, 2 are errors in a source and 2 are not explained.

5. Discussion

The headline numbers in the public debate rest on two surveys of regional and municipal owners by a single Observatory. Most of the apparent disagreement disappears once denominators, survey years and concepts are explicit. Three findings remain.

1. The EU comparison is the least consistent part.
   • The averages differ by country set, averaging method and base.
   • Two official texts set a Spanish household-survey share beside a stock-based EU average.
   • The 2026–2030 Plan uses a 9% average with no stated source.
2. An attribution mismatch was repeated. A figure attributed to the Banco de España in a Commission staff document reappears in the Council's recommendation of the same Semester cycle. Neither its denominator nor its EU average is in the cited paper. A radio report a year later attributes the same «1,5 %» of the total to the Banco de España.
3. The national count has no owner-level published record.
   • The national count is about half extrapolated at the municipal level.
   • The bulletin's tables contain arithmetic slips.
   • The State's own stock is outside the count. It is the subject of art. 32, which asks for an inventory and an annual report in the Ministry's electronic office; we found neither.

We do not claim that any of these choices was meant to mislead. A common definition, a dated source and a stated denominator in each official text would remove most of the spread.

6. Limitations

• The Ministry's main website blocks our client (HTTP 403). We checked the electronic office that the law names, but a report published only on the main website cannot be excluded.
• We reproduce the surveys; we cannot audit them. We cannot check the regions' and municipalities' answers. Our reproduction of the municipal scaling uses a sample that differs slightly from the one the bulletin describes.
• Matching values is not proof of origin. We describe the 2.5%, 3.4% and 2.5–3.4% figures as equal to ECV values, and the Plan's 9% as having two possible origins. Only the 2024 Ministry table states its use of the ECV.
• Some sources could not be read:
  • the OECD's 2025 Economic Survey of Spain (403);
  • Provivienda and Fundación Alternativas (CAPTCHA);
  • El Periódico (406);
  • the Madrid housing agency's stock dataset, whose portal's robots.txt disallows all agents.

Figures from these are cited only as others report them.

• The census. INE's API was unreachable. We rely on the Census 2021 project document, which lists three tenure categories and no social rent, and on Housing Europe's statement that the census does not separate social from private rent.
• Regional figures cannot be summed independently.
  • Catalonia's asset inventory (13,552 items) does not itemise the housing agency's stock.
  • We found no regional stock dataset for Andalusia.
  • The Basque datasets are contract registers.
• Coverage. We may have missed figures in documents we did not search: regional laws and plans, parliamentary answers, speeches.
• Press search. Only Bing News RSS was used.
• Tables read from PDF. They are parsed from text. Seven unit tests cover the parsers, every printed total is checked against its rows, and the differences we report were checked by hand against the PDF.

7. Ethics

How we fetched.

• Every scripted request carried our User-Agent, read robots.txt first and kept to at most one request per second per host.
• Where robots.txt disallowed our client, we did not fetch: datos.comunidad.madrid, data.europa.eu's search API, part of the Banco de España's site, and the BOE's XML path. We read the Constitutional Court rulings from the BOE's /buscar/doc.php, which is allowed.
• Where a server blocked us (the Ministry's main website, the OECD, two CAPTCHA pages), we did not retry or try to get round it.
• Python could not verify one server's TLS chain. For that server we used curl, which verified it; TLS was never disabled.

Personal data. We used official texts, statistics and published news. The municipal table names municipalities, not people. The rulings and the press name public officials only in their public role. We processed no personal data.

Contacts. We contacted no one.

8. Data, code and licences

Code is Apache-2.0; our data and text are CC BY 4.0. Third-party figures and quotations are attributed row by row:

• Spanish legislation and court rulings are not protected by copyright;
• EU texts and Eurostat data are reused under the Commission's reuse policy;
• material from the Ministry, CASA 47, INE, the OECD, Housing Europe and the Banco de España is quoted with attribution.

Raw downloads are not redistributed.

9. Automation and review

AI agents (Claude models) did the work. They:

• found and downloaded the sources, under the rules in §7 and METHOD.md;
• wrote the parsers and the reconciliation;
• searched for the art. 32 report and for prior work;
• drafted the text.

Checks.

• scripts/figures.py checks every quotation in the figures tables against the raw source.
• scripts/check_quotes.py checks every «quotation» in the README and this paper.
• scripts/check_headlines.py checks the headline numbers against data/.

Review. An independent AI agent then reviewed the study adversarially. It re-fetched the 22 documents then cited and found every quotation literally. It reproduced data/ byte for byte.

Its findings:

• one blocker: we had checked a wrong hostname for the Ministry's electronic office;
• nine major:
  • the OECD's country set;
  • the ECV origin of 2.5% and 3.4%;
  • the Constitutional Court rulings on art. 32;
  • CASA 47 and the State-side figures;
  • the reconciliation tally;
  • an unsupported attribution of the EU texts' 9%;
  • non-literal quotations in the README;
  • the reading of the regional growth;
  • the reading of the Eurostat measure;
• several minor ones.

All were applied in this version, except one: a swapped regional/municipal split in Housing Europe 2021, which the page layout does not show. This version has not yet been re-reviewed.

10. Competing interests

EasyByte Hub S. Coop. Mad., which runs EasyxLab, has no contract with any of the following:

• the Ministry of Housing, CASA 47, or the regional or municipal housing agencies;
• the OECD, Housing Europe or the Banco de España;
• the European institutions.

It does not build or manage housing. No EasyByte product uses these data.

How to cite

EasyxLab (2026). How large is Spain's public housing stock? Tracing the official figures and the EU averages they are compared with (2019–2026). Study S23. EasyByte Hub S. Coop. Mad. https://github.com/easybytehub/easyxlab

References

Spanish law and courts

• Ley 12/2023, de 24 de mayo, por el derecho a la vivienda. BOE 124, 25 May 2023; consolidated text, BOE-A-2023-12203.
• Real Decreto 326/2026, de 22 de abril, Plan Estatal de Vivienda 2026-2030. BOE-A-2026-8872.
• Tribunal Constitucional. Sentencia 79/2024, de 21 de mayo (BOE-A-2024-12808). Sentencia 190/2025, de 16 de diciembre (BOE-A-2026-1756).

Ministry, CASA 47 and statistics

• Ministerio de Transportes, Movilidad y Agenda Urbana, Observatorio de Vivienda y Suelo. Boletín Especial Vivienda Social 2020. NIPO 796-20-148-8.
• Ministerio de Vivienda y Agenda Urbana, Observatorio de Vivienda y Suelo. Boletín Especial Vivienda Social 2024. NIPO 179-24-022-6.
• Ministerio de Vivienda y Agenda Urbana. Estimación del parque de viviendas (VDP002_01).
• CASA 47, Entidad Estatal de Vivienda. Plan Anual de Actuación 2026; Informe de gestión 2024; web pages on the Sareb transfer («Traspaso Vivienda y Suelo») and of 7 September 2026.
• INE. Encuesta de Condiciones de Vida, table 76845. INE. Censos de Población y Viviendas 2021. Proyecto.
• Eurostat. EU-SILC, ilc_lvho02 (updated 17 September 2026).
• OECD. Affordable Housing Database, indicator PH4.2, Social rental housing stock (updated 24 November 2025).
• Housing Europe. The State of Housing in the EU 2019; The State of Housing in Europe 2021; The State of Housing in Europe 2025, Spain profile.
• Banco de España. Informe Anual 2023, chapter 4 (2024). Khametshin, López Rodríguez and Pérez García, El mercado del alquiler de vivienda residencial en España: evolución reciente, determinantes e indicadores de esfuerzo, Documentos Ocasionales 2432 (2024).

EU texts

• European Commission. 2022, 2024, 2025 and 2026 Country Reports – Spain: SWD(2022) 610, SWD(2024) 609, SWD(2025) 209, SWD(2026) 209.
• European Commission. The European Affordable Housing Plan, COM(2025) 1025. SWD(2025) 1053.
• Council Recommendations to Spain of 14 July 2023, 8 July 2025 and 10 July 2026.

Scholarly work

• Trilla, C. and Bosch, J. (2018). El parque público y protegido de viviendas en España: un análisis desde el contexto europeo. Fundación Alternativas, Documento de trabajo.
• Pareja-Eastaway, M. and Sánchez-Martínez, M. T. (2017). More Social Housing? A Critical Analysis on Social Housing Provision in Spain. Critical Housing Analysis 4(1). doi:10.13060/23362839.2017.4.1.331.
• García-Montalvo, J. (2026). La vivienda en España: diagnóstico y propuestas desde una perspectiva internacional. Ekonomiaz. doi:10.69810/ekz.1529.

Fact checks and press (datelines from the articles; URLs in data/prior_work_search.csv)

• Newtral, 22 November 2018 and 17 January 2023.
• La Marea, 8 January 2025.
• La Vanguardia, 28 September 2025.
• Europa Press, 26 November 2025 (on the OECD Economic Survey).
• El País, 7 May 2026 and 29 September 2026.
• El Periódico, 28 July 2026 (headline only).
• EFE Verifica, 30 September 2026.
• Público, 3 October 2026.
• Cadena SER, 3 October 2026.
• La Gaceta de Canarias, 4 October 2026.

Cite this study

EasyxLab (2026). How large is Spain's public housing stock? Tracing the official figures and the EU averages they are compared with (2019–2026). Study S23. EasyByte Hub S. Coop. Mad. https://github.com/easybytehub/easyxlab/tree/main/studies/s23-public-housing-stock

@techreport{easyxlab_s23,
  title       = {How large is Spain's public housing stock? Tracing the official figures and the EU averages they are compared with (2019–2026)},
  author      = {{EasyxLab}},
  institution = {EasyByte Hub S. Coop. Mad.},
  number      = {S23},
  year        = {2026},
  url         = {https://github.com/easybytehub/easyxlab/tree/main/studies/s23-public-housing-stock}
}