Spain's 2024 record in centres for homeless people: what the +57.5% measures

INE, Spain's statistics office, reported a mean of 34,145 occupied places a day held by adults in centres for homeless people in 2024, +57.5% on the 21,684 of 2022. That is the highest figure, and the fastest rise, in the series INE charts from 2006. The figure is a mean of the occupied places that responding centres report on two days. It is not a count of homeless people: it leaves out people sleeping rough and people who come and go between the two days. This study asks what the rise is made of.

Most of the increase is in centres that INE classes as specialised in immigrants.

• The category, as INE defines it:
• specialisation means «Distintas situaciones que el centro atiende específica y exclusivamente»;
• in 2024 each centre ticks «Centro especializado en la atención a migrantes» on the questionnaire itself.
• INE does not break the category down. How many of its places are asylum or humanitarian reception and how many are housing for seasonal workers is unknown.
• Occupied places in these centres went from 8,045 to 18,173. That is 10,128 of the 12,461 increase, or 81.3%.
• INE's own explanation: «las llegadas masivas de inmigrantes a Canarias y el reparto de inmigrantes provocaron el aumento del número de plazas de alojamiento disponibles».

The definition changed in the record year.

• In 2020 and 2022, INE's methodology added immigrant centres to its definition of homelessness «para asegurar la continuidad de la serie histórica».
• In 2024 it moved to the ETHOS typology. Immigrant accommodation became one of its homelessness categories: «Personas inmigrantes que viven en alojamientos temporales por su estatus de extranjeros o trabajadores temporeros».
• The specialisation item appears on the questionnaire for the first time in 2024.
• INE's release does not mention the change.

How the increase splits. With a symmetric (Shapley) split, the 12,461 divides into four parts. The range column gives the result of each one-at-a-time ordering.

The published tables cannot separate centres that opened from centres that entered the directory or answered for the first time. Such frame changes can sit in either term. The 2025 revision is an example: it added one centre with 387 occupied places, and the split books 287 of them as "more people per centre".

In the other centres, occupied places grew faster than their number.

• Occupied places grew +17.1% and the number of these centres offering accommodation +11.9%, so people per centre rose +4.7%.
• That average mixes gender-violence shelters (−24.6% per centre) with the rest (+7.9%).
• Across the variants we computed, the per-centre rise runs from +3.3% to +11.9%.
• The total number of centres also rose +17.1% (1,175 to 1,376). That is an equality between unlike quantities: most of the new centres were immigrant-specialised.

The 81% share moves little under the three variants that can change it: places instead of occupied places, 81.0%; the figures INE first published, 80.7%; a 2020 base, 78.2%. The 2020 base crosses INE's 2022 switch to counting adults only. The December snapshot rose more (+65.8%) than the June one (+48.6%). In the Canary Islands, occupied places went from 2,692 in June 2024 to 5,001 in December, 40.5% of the national June–December gap.

The published figure changed after release.

• On 26 September 2025 INE published 33,758 (+55.7%).
• Its results page now reads «Datos revisados y modificados con fecha 17/10/2025». Nationally the change adds one centre, 387 places and 387 occupied places, all in immigrant-specialised centres.
• In the release's regional table only Navarre changed (403 → 790 mean places), and the changed cells are shown in red.
• INE published no note on what changed or why. Its methodological report lists «revisiones extraordinarias (por ejemplo, las debidas a un error en estadísticas ya publicadas)» among the revisions its policy covers, and also states «No existe revisión de datos».

Against the reception system, the arithmetic only.

• INE's immigrant-specialised places equal 17.1% of the year-end places of the state asylum reception system (SAPI) plus the humanitarian programme (PAH) in 2020, 26.6% in 2022 and 37.4% in 2024.
• This is not a coverage rate. We cannot show that INE's centres are a subset of SAPI and PAH, and the category also includes seasonal-worker housing.
• From 2020 to 2022 INE's figure moved with SAPI, at the time of the displacement from Ukraine. From 2022 to 2024 it moved with PAH and with arrivals (+104.9%). Each matches INE's own explanation for that edition.

The national strategy's coverage indicator mixes two populations.

• The Strategy 2023–2030 divides the survey's places by the 28,552 homeless people estimated by INE's 2022 survey of people. Of those people, «el 0,4% en centros de ayuda al refugiado».
• So the numerator includes the reception places and the denominator almost none of their residents.
• On the Strategy's own basis (December places) the indicator went from 70.7% to 98.2% to 148.1%, above its 2030 goal of 90%. This is our calculation; the Ministry has not published it.
• Immigrant-specialised places can be removed only from the mean. On that basis the indicator is 65.1% without them in 2024.
• The Ministry's 2024 progress report reads the survey as having «estimaba en 34.145 el número de personas sin hogar en España».

People in reception centres need shelter, and under the ETHOS typology adopted by the Strategy and by INE in 2024 they are houseless. The European Observatory on Homelessness notes that no EU state recognises this classification in law or administration, and that ETHOS Light excludes them. Either way, the record is real: more shelter places were occupied than in any year INE charts since 2006. What it does not show is a comparable rise outside immigrant-specialised centres. We measure what the statistic counts. We make no claim about any group of people.

Much of this was said before.

• HOGAR SÍ said on the day of release that the inclusion of migrant and asylum places «condiciona drásticamente el aumento», and questioned the selection of centres.
• INE named the Canary Islands.
• In the United States, Meyer, Wyse and Williams (2026) attribute 59–62% of a 43% rise in sheltered homelessness to new migrants.

What we add:

• the decomposition and its range;
• the definition change of the record year;
• where the revision sits;
• the mismatch in the Strategy indicator;
• a protocol for the 2026 edition: PROTOCOL.md, registered by commit 844d5a3 on 4 October 2026, with dated clarifications in its §12. That edition is expected around September 2027, by inference: INE has not scheduled it.