Spain's 2023 election census doesn't add up: 590,000 more votes than voters

Spain's 2023 election census doesn't add up: 590,000 more votes than voters

Check the official figures for the general election of 23 July 2023: 24,743,612 votes for parties, 10,402,450 abstentions, 261,078 invalid ballots and 199,392 blank ballots. They add up to 35,606,532. The census of residents in Spain entitled to vote was 35,016,000. The subtraction leaves almost 590,000 extra votes, a discrepancy that spread within days of the count and has never received an official explanation that fully resolves it. The figure has been circulating for years without anyone completely debunking it.

What are the CER and CERA censuses?

Spain manages two separate electoral censuses. The CER includes residents in national territory entitled to vote; the CERA includes Spaniards living abroad. The former was around 35 million voters in spring 2023; the latter, about 2.3 million. The alarm arose when adding turnout, abstention, invalid and blank ballots and finding that the result exceeded the ceiling of the national census. From there came the fraud thesis.

The objection: invalid and blank ballots were already counted

The most repeated argument against the discrepancy is arithmetic. Invalid and blank votes are already part of the votes cast, so adding them again duplicates part of the total. Abstention, moreover, is not counted vote by vote: it is obtained by subtraction, subtracting the votes cast from the electorate called to vote. With that adjustment, much of the supposed excess disappears. What remains, according to the most repeated calculations on the critical side, is a margin of around 140,000 votes without a document to explain it.

Why does the census rise between March and July?

The second front is the municipal register. The census used on election night is not the one closed on 1 May, but the one in force with later additions: those who turned 18 before 23 July and do not yet appear in the INE's monthly figures. Between March and July, four months of new voters are added. The other half of the equation is the deceased, who in some counts were not removed. The official census as of 1 June stood at 35,032,817; the figure used for 23J, according to those who defend this line, is 35,141,122. None of that appears in a bulletin that anyone can consult.

PSOE multiplies its votes between 28M and 23J

Here the numbers are cited again and again. In the Community of Madrid, the PSOE went from 609,718 votes in the regional elections of 28 May 2023 to 993,870 in the general election less than two months later. In the Valencian Community, from 691,861 in the regional elections to 834,758 in the province of Valencia alone. These are increases of around 60% in territories voted in with weeks of difference and without a change of national leadership. The comparison has become the most cited argument to maintain that something does not add up.

The thesis of CERA votes slipped into mailbags

The hypothesis with the most traction holds that votes from the overseas census were introduced into postal vote bags. The reasoning: if the CERA votes at around 1%, a procedure that requires requesting and sending it from the consulate, there is a huge margin between the votes actually cast and those that fit in the census. It would be enough to inflate the postal vote campaign to justify the entry of an extra package. The date, 23 July in the middle of the holiday season, would serve as a cover. No documentary evidence supports the manoeuvre: it is a hypothesis, not a proven fact.

The overseas voter and the envelope that never arrives

Those who live abroad tell it without embellishment: in one reported case, the envelope with the documentation arrived 36 hours before the vote. With the postal vote deadline already closed, anyone far from the consulate is left unable to exercise their right. In that account, the documentation only appeared in two of the last calls to election. The regulations changed to prevent CERA voters from voting in person at their Spanish polling station, which, critics argue, eliminates the only check that the vote cast has not been impersonated.

Segovia: a count that ends up in court

What has come out of the courts is another matter. The presiding magistrate of the Provincial Court of Segovia and of the Provincial Electoral Board, Ignacio Pando Echevarría, resigned after being accused of electoral crimes allegedly committed during the 23J count in that province. The Public Prosecutor's Office ruled that he is not protected by parliamentary immunity and the case fell to the Court of First Instance and Instruction number 1 of Segovia, in preliminary proceedings 104/2024. The accusation points to the procedure, not to a swing in votes.

In parallel, accusations circulated about the counting software: a foreign television network pointed to Indra for alleged systematic electoral fraud and a statement spread on social media claimed that there were ballot boxes arriving at IFEMA and employees who could talk about the matter. None of that has reached the courts as evidence.

The shadow of the postal vote

The volume of postal voting on 23J, close to three million according to the figures handled, fuels all the theories. An official linked to the postal network, Leire Díaz, has been accused on social media of prior rigging of the postal vote in a municipality in Cantabria, an accusation that does not appear in any judicial resolution. The coincidence of that volume with the count of the overseas census, which is counted separately, is what joins the pieces of the conspiracy narrative.

With these ingredients, the question is no longer just where 590,000 votes come from. It is how many times a count that no one checked minute by minute has been taken for granted, and whether the discrepancy is a miscalculation, as the technical line maintains, or something more. There is still no official answer that closes it.