Spain's vote from abroad and the 'grandchildren's law': 430,000 more voters who barely move seats
The roll of Spaniards who vote from abroad has grown by about 430,000 since the 2023 general election, three in four of them in the Americas, mainly because of the so-called 'grandchildren's law'. In 2023 the vote from abroad decided a seat in Madrid. In the simulations of our forecast, the new voters change at least one seat in 8 out of 100 scenarios, and the Supreme Court's interim suspension of the vote of 163,020 people who became citizens under the law barely alters the result.
21 October 2026 · Miguel · Leer en español
The roll of Spaniards who live abroad and can vote in general elections, known as the CERA, grew from 2,333,056 voters in July 2023 to 2,762,829 in September 2026: 429,773 more, up 18.4%. It is the largest increase between two general elections since at least 2008, and it has one main cause: the so-called “grandchildren’s law”, the eighth additional provision of the 2022 Democratic Memory Law, which let the children and grandchildren of Spaniards who emigrated apply for citizenship.
Who are these new voters, where do they vote, how does the vote from abroad lean and how much could it change the result on November 29? For the last question we reran the simulations of our forecast with different rolls. And there is one more factor: the Supreme Court has suspended, as an interim measure, the vote of 163,020 people who became citizens under this law.
How many
From January to October 2022, before the law, the CERA grew by about 4,131 voters a month. Since the 2023 general election it has grown by about 11,310 a month. Had it kept the earlier pace, it would have about 272,799 fewer voters today: a reasonable estimate of the law’s effect on the roll.
The Foreign Ministry’s figures point the same way. By August 31, 2026, 1,317,108 citizenship applications had been filed under the law, 600,684 had been approved and 369,017 had been registered. Not everyone registered votes: some are minors, and some have not yet registered as residents with their consulate.
73.3% of the new voters live in the Americas. Argentina, with more than 500,000 voters, is by far the country with the most Spaniards entitled to vote, and the one that has added the most. Mexico and the United States have grown by more than 25%. Even so, the new voters are not very different from those already on the roll: 61.7% live in Latin America, against 53.8% of the 2023 roll abroad.
Where they vote
Each CERA voter votes in the district where they are registered, usually that of their last address in Spain or that of their parents. That is why the vote from abroad weighs very differently from one province to another:
Source: INE, Electoral Census Office. Roll for the November 29, 2026 general election.
In Ourense, almost one voter in three (30.8%) lives outside Spain; in Lugo, more than one in five. Galicia, Asturias, León, Zamora and the Canary Islands were the heartlands of 20th-century emigration to the Americas. But the new voters have not mostly registered there: the CERA of the Galician provinces grew by between 8% and 11%, and Madrid’s by 32.5%. Madrid has 122,370 new voters abroad, more than any other district.
| District | CERA 2023 | CERA 2026 | New | % of the 2026 roll |
|---|---|---|---|---|
| Madrid | 376,504 | 498,874 | +122,370 (32.5%) | 9% |
| Barcelona | 217,682 | 265,687 | +48,005 (22.1%) | 6% |
| Asturias | 123,430 | 143,588 | +20,158 (16.3%) | 14.7% |
| Santa Cruz de Tenerife | 107,213 | 125,100 | +17,887 (16.7%) | 13.5% |
| Coruña (A) | 163,367 | 178,478 | +15,111 (9.2%) | 16.1% |
| Pontevedra | 136,739 | 150,148 | +13,409 (9.8%) | 16.2% |
| Málaga | 60,853 | 74,033 | +13,180 (21.7%) | 5.7% |
| Valencia/València | 70,391 | 83,234 | +12,843 (18.2%) | 4% |
| Palmas (Las) | 55,641 | 66,235 | +10,594 (19%) | 7.1% |
| Alicante/Alacant | 45,301 | 55,464 | +10,163 (22.4%) | 4.1% |
| Balears (Illes) | 32,840 | 42,115 | +9,275 (28.2%) | 4.8% |
| Granada | 53,300 | 62,425 | +9,125 (17.1%) | 7.9% |
The 12 districts where the CERA grew the most between the 2023 and 2026 general elections. Source: INE, Electoral Census Office.
How the vote from abroad leans
Voting from abroad has never been easy. Between 2011 and 2022, Spaniards abroad had to request their vote before each election (the “voto rogado”), and CERA turnout plunged from 31.7% in 2008 to 5% in 2011. In 2023, with that requirement gone, it rose to 9%: about 208,431 voters.
In 2023 the vote from abroad looked fairly similar to residents’ vote: the conservative People’s Party (PP) got the same share, the Socialists (PSOE) and the hard-right Vox somewhat less and the left-wing Sumar somewhat more. But it was decisive in one district: in Madrid, the CERA count gave the PP the last seat, which had gone to the PSOE in the residents’ count.
Do those living in the Americas vote like those living in Europe? There are no results by country: CERA votes are counted by province. But each province has a different mix. In Lugo or Asturias, more than seven in 10 CERA voters live in Latin America; in Córdoba, fewer than one in four. Comparing each province’s vote from abroad with that of its residents reveals a clear pattern:
PP
PSOE
Vox
Frente Amplio
- Where a larger share of the CERA lives in Latin America, the vote from abroad leans more toward the PP and Vox than residents of the same province: about 3.4 points more for the PP (between -0.3 and 7, with 90% confidence) and 4.8 more for Vox.
- Where it lives mostly in Europe, it leans more toward Sumar (9.3 points more) and less toward the PP and Vox.
This is an ecological estimate: it compares provinces, not people, and may reflect other differences between them. With that caveat, it suggests the new voters, slightly more Latin American than those already on the roll, would lean a little further right than the 2023 CERA. But the difference in composition is small (eight points more in Latin America), and so is the effect.
The Supreme Court and the 163,020
In early September 2026, at the request of Vox and the association Iustitia Europa, the Supreme Court’s administrative chamber suspended, as an interim measure, the electoral effects of the CERA registration of people who became citizens under the grandchildren’s law. The suspension does not apply to those who show, with a certificate from their consular registry, that they descend from Spaniards who went into exile for political reasons, and it does not strip anyone of citizenship. On October 12, the court rejected an appeal by the State Attorney’s Office, backed by the public prosecutor. According to the government’s figures given to the court, 163,020 people are affected.
On October 21, the Constitutional Court declined, by ten votes to two, to admit the Ombudsman’s appeal, as premature: ordinary judicial remedies have not been exhausted. The suspension stands while the Supreme Court rules on the merits, and the Central Electoral Board has scheduled the first mailing of voting documents to the CERA for October 24–30, with a warning about the possible suspension. We will update this analysis if things change.
How much does it change the result?
We reran the 20,000 simulations of our October 21 forecast with five different rolls abroad. Everything else (polls, residents’ turnout, territorial errors) is identical in each simulation, so the differences are due only to the CERA.
| Scenario | Expected CERA votes | Simulations with at least one different seat | PP | PSOE | Vox |
|---|---|---|---|---|---|
| Without growth since 2023 (reference) | 214,011 | — | — | — | — |
| With the full 2026 roll abroad (the forecast) | 257,352 | 8% | +0.02 | 0.00 | −0.02 |
| With the 163,020 votes suspended | 241,662 | 5.2% | +0.01 | 0.00 | −0.01 |
| Bound: every new vote to the PP | 257,352 | 39.6% | +0.50 | −0.21 | −0.16 |
| Bound: every new vote to the PSOE | 257,352 | 36.9% | −0.23 | +0.45 | −0.13 |
Seats: average difference from the reference scenario, across all simulations. CERA votes: expected valid votes from abroad across Spain.
- The growth of the roll abroad adds about 43,341 votes across Spain, because CERA turnout is low. It changes at least one seat in 8 out of 100 simulations, almost always just one, and in different districts depending on the simulation: only in Madrid (1.8%) and in Barcelona (0.9%) does it happen in more than 5 out of 1,000. On average, it is less than a tenth of a seat for any party.
- Suspending the 163,020 votes removes about 15,690 votes and changes at least one seat relative to the forecast in 2.9 out of 100 simulations.
- Not even the extreme case changes much. If every new vote went to a single party, which happens in no group of voters, the PP would gain 0.50 seats on average, or the PSOE 0.45. That is half a seat on average in the most favorable case for either. At least one seat would change in about 4 out of 10 simulations, mostly in Madrid.
The reason is simple: even if the roll abroad grows by hundreds of thousands, fewer than one in 10 vote, and their votes are spread across 52 districts. Only in the largest, such as Madrid, where much of the growth is concentrated, can they decide a last seat, as happened in 2023.
Since October 21, our forecast removes the 163,020 suspended votes from the roll abroad for as long as the suspension lasts. As the table shows, the difference is far smaller than the uncertainty in the rest of the model. The full forecast under both scenarios, with and without them, is in this analysis.
What this analysis does not say
- We do not know how the grandchildren vote. There is no survey of this group and no results by country of residence. The scenarios assume they vote like the rest of the CERA in their province, and the bounds show the largest possible effect.
- Turnout is unknown. We use each province’s 2023 turnout. Voting from abroad requires receiving the documents and mailing them back or dropping them at a consulate, and the new voters will be doing it for the first time.
- The provincial split of the suspended votes is an assumption. It has not been published: we distribute it like the growth of the CERA in the Americas since 2023.
- The pattern by continent is ecological. It compares provinces, not people.
How we calculated it
- Roll. CERA voters by month, province of registration and country of residence (INE, Electoral Census Office), from January 2022 to September 2026, and the rolls of the 2023 and 2026 general elections by district. Latin America: the Americas excluding the United States and Canada.
- Vote. CERA results by district from 2008 to 2023, and residents’ results (official total minus the CERA). The 2023 Madrid seat comes from allocating seats with the D’Hondt method with and without the CERA votes.
- Pattern by continent. Regression weighted by CERA votes, across the 52 districts, of the difference between the vote from abroad and residents’ vote on the share of the CERA living in Latin America; 90% intervals by bootstrapping provinces.
- Simulations. Those of the October 21 forecast, with the same seed and five rolls abroad. CERA turnout: each province’s 2023 turnout, with national and provincial error; vote split: each province’s 2023 CERA split, shifted by each party’s national change.
- Code.
articulo_nietos.py, in the altnalysis repository.
External sources. Foreign Ministry figures as of August 31, 2026: Europa Press, via PressDigital. Interim suspension and people affected: EFE, via Diario de León and Iberoeconomía. Situation as of October 21: La Región.