Spain's election fraud claims persist despite no recount

From April 2019 to July 2023: five years of fraud suspicions without a single recount

Six parties. Six results. All multiples of three. The calculation that circulated after the general elections of 28 April 2019 claimed that the probability of this coincidence occurring by pure chance was 0.13553%, that is, one in more than seven hundred. From there to the headline "fraud confirmed" was a single step. What never arrived was an official recount of the tallies that would disprove the suspicion, nor one that would confirm it. Five years later, the theory is still alive, and it no longer needs to prove anything.

The 0.13553% that became a verdict

The arithmetic argument was repeated ad nauseam in videos, tweets and pages: if the six major parties obtained seat counts that were multiples of three, coincidence was ruled out. One analysis raised the finding to a 99.86% probability of "artificial creation" of results, a figure that was spread as "fraud proven at 99.9%". The packaging was impeccable: mathematical language, two decimal places and a conclusion that sounded like an expert report.

The trap is old. Choosing the pattern that fits after the fact—multiples of three, two, or any other—and then calculating its probability turns any coincidence into alleged proof. Some argue that these exercises, with all their margin, point to an anomaly that no one has wanted to explain. Against that weighs a stubborn fact: none of these calculations has ever been accompanied by a recounted municipal tally to back them up. The doubt, however, did its job: it gave a language to a distrust that already existed.

The Catalan company, Soros and the Venezuelan algorithm

Before voting, there was already a narrative. An article on 23 April 2019 cast the shadow of fraud over the company responsible for processing the results: Catalan, it said, and "linked to Soros". No contract, no audit, no documentary evidence accompanied the accusation. The aroma was enough.

Then came the videos. One of the most widely circulated claimed to have decoded the algorithm with which the PSOE would have hacked the elections, and relied on a testimony with pedigree: Venezuela's head of electoral processes, Ana Mercedes Díaz, would have anticipated months earlier how the Spanish fraud would occur. The narrative hooked two antiestéticars at once: blind distrust of software and the ghost of Venezuelan fraud exported to Europe.

In October 2019 a new ingredient appeared. The company designated for the count would have been charged in an open case, according to the material circulated. "Charged" is a word that in Spain sounds like a conviction and legally it is not, but in certain channels it is enough that it sounds like one.

Vox, the 40,000 invalid ballots and the request that did not prosper

On 1 May 2019, Vox requested a general review of all ballot boxes in Spain. It did so after confirming, according to its version, a pool of tens of thousands of invalid votes, including 40,000 ballots in the Valencian Community. Its legal services asked the Central Electoral Commission for a full check of the count. The request remained where such things usually remain: in the air.

The detail matters because it marks a pattern that would repeat at every election. The complaint of irregularities is filed, leaked to sympathetic media, turned into a video and dies in an office without anyone counting a single ballot before a notary.

The recount that never happens: tallies, INDRA and shelved complaints

The technical core of the suspicion soon traveled from the parties to the procedure. The most repeated complaint: the Provincial Electoral Commissions should recount the tallies of each municipality in their province within five days, and there are hundreds of private observers willing to appear as witnesses. These report that, in practice, many tallies were not recounted and the figures provided by INDRA, the technology company that processes the data, were taken as good. The system, in theory, does contemplate review. In practice, citizen oversight crashes against the door.

To this is added a legal problem that repeats in every complaint: the files would be shelved because, "legally", only political parties can file the claim. Translation: the citizen who sees an irregular polling station has no door to knock on. The paradox was summarized in a message on 17 November 2019 with a phrase that became a banner: in Spain, tallies are not counted.

Transparent Elections and the figure of 80 deputies

In January 2020, after the 10-N elections, a platform called Transparent Elections published what it presented as "real counts" of the votes, without INDRA and without Tezanos's polls. Its general secretary, Ignacio Vargas, stated that, according to its data, Pedro Sánchez would not have obtained more than 80 or 90 deputies. The difference with the official result was so overwhelming that, instead of settling the controversy, it multiplied it.

The underlying problem is the same that has dragged the entire dispute since 2019: no one has been able to compare those alternative counts with an audited scrutiny. Against the official figure, the counter-count. Against the counter-count, the absence of a mechanism to determine which one is valid.

The 23-J elections: the gap between Congress and Senate

After the general elections of 23 July 2023, a senior police official would have warned presenter Javier Cárdenas of an anomalous difference between votes for Congress and the Senate, greater than any seen before: "people are not so crazy as to vote for the PP for the Senate and for the PSOE for Congress". In Valencia, according to the videos released, a blank, sealed and signed tally appeared. In another recount, amid the tumult, a box with 580 votes appeared out of nowhere. In Cantabria, towns with "implausible" votes for Junts were reported, and always with the same bias of error.

None of these pieces has been taken to a court to validate them. And there is the knot: evidence is accumulated without chain of custody and zero rulings that turn suspicion into res judicata.

The overseas vote and the grandchildren law

Then the focus shifted to the census. Four members of the Central Electoral Commission would have warned of irregularities in the census of the so-called grandchildren law, the mechanism that grants nationality to descendants of Spaniards. More than 600,000 requests would have been approved and more than half would already be registered. A socialist leader in Argentina was pointed out, according to the publications that circulated, for collecting passports of deceased people and for boasting that the overseas vote would "sustain" Sánchez.

The suspicion spread to the vote of those who do not reside in Spain. Some argue that those who do not live here should not vote, because there are regional governments that could depend on voters who do not set foot in the territory. The argument has traction, and also a counterweight: any change in overseas voting would require reforming the system, something that no party with a majority has wanted to open.

The pattern that repeats election after election

Each electoral cycle leaves the same sequence: a striking calculation, viral videos, the mention of a company or platform, a formal request and a silent shelving. From 28-A to 23-J, four years passed, dozens of tweets and not a single independent audit of tallies that would close the matter one way or another.

Defenders of the fraud thesis point out that the insistence on requesting a review is not a whim, but the only way to know. Critics respond that, by dint of repeating the suspicion without proving it, the word "pucherazo" has ceased to describe a crime and become a way of being in politics. Both things can be true at the same time, and there the mess remains.

If in five years not a single ballot box has been recounted before a notary, what would it take for the next election to begin and end without anyone shouting fraud?