In these days, numerous analyses will proliferate about who Spaniards will vote for on November 29, and they will be made on a board that is still moving. For example, it seems that Aliança Catalana will not run in the general elections and CUP’s membership has endorsed not doing so. Podemos will appear on ERC’s lists in Catalonia, while the Broad Front proposed Ada Colau as a candidate and continues inviting Podemos, even if that implies relinquishing Colau’s leadership. The deadline to register coalitions expires on October 16 and the deadline to present lists, ten days later.
Before asking ourselves who each person will vote for, it is worth asking when voters decide. And there, the electoral advance matters in itself.
An advance is not just any election
We tend to treat early elections simply as elections that arrive earlier than planned, but they are not. Who calls them chooses the moment and thus limits the margin of maneuver of the rest. Moreover, they affect the voters themselves. These elections are becoming increasingly frequent: in Europe, about 45% of national elections held between 2020 and 2024 were advanced, compared with 11% at the start of the century. Spain is no exception: although for different reasons, all general elections called since 2019 have arrived before the legislature ran out.
“In Europe, around 45% of national elections held between 2020 and 2024 were advanced”
In a recent article published in the academic journal Electoral Studies, Klara Müller and I studied what happens to the undecided voters right after an announcement of an advanced election call. We measured indecision by the change in the proportion of people who respond “I don’t know” when asked whom they would vote for.
What we found in Germany and the United Kingdom
In Germany, Olaf Scholz announced in November 2024 that he would face a confidence vote in the midst of collecting data for the German Electoral Study (GLES), which allowed us to compare those who answered beforehand with those who did so afterward, in a natural experiment. In the United Kingdom, Rishi Sunak called elections in May 2024, just between two consecutive waves of the online panel of the British Electoral Study (BESIP).
What we found was that, in the United Kingdom, the share of those who did not know whom to vote for fell from 17.3% to 12.8% in a matter of days, a relative reduction of 26%. In Germany, among those who did not identify with any party, the probability of declaring indecision dropped by 6.5 points, a 21% decrease.
“In the United Kingdom, the share of those who did not know whom to vote for fell from 17.3% to 12.8% in a matter of days, a relative reduction of 26%”
This does not happen before any election: we repeat the same analysis for equivalent periods prior to the German general elections of 2009, 2013, 2017 and 2021, and to the British ones of 2015, all held on the dates planned (i.e., not advanced). In Germany there was no reduction in vote uncertainty and in the United Kingdom it was small and gradual. In other words, electoral advances act like a kind of alarm clock: they push voters to decide earlier.
And in Spain?
If the same mechanism applies here, the October polls should show fewer undecideds than the late-September ones, and that drop should not be read as a success of any pre-campaign, but as a natural effect of the early call.
“Early elections act like a kind of alarm clock: they push voters to decide earlier”
The initial data, however, do not show it yet. The 40dB barometer gives us a few clues. In its round from September 25 to 28, with 2,000 interviews, 8.4% answered “I don’t know” to what party they would vote; in the flash survey of October 7 and 8, with 800 interviews, it was 9.1%. The difference falls within the margin of error, but the anticipated drop has not appeared yet.
Notably, though, the share saying they will not vote falls from 5.5% to 4.1%. And those who say they will vote “certainly” (a 10 on a 0 to 10 scale) rise from 56.7% to 66.7%, although the question changed: in September it was hypothetical and in October it already referred to November 29. A possible interpretation is that the advance has mobilized abstainers who now want to vote, but still do not know for whom.
In Spain there is a key reason to expect a smaller or later effect. In Germany and the United Kingdom, the call fixed a date to a known lineup of parties. In Spain, to the left of the PSOE, that lineup remains open and, in Catalonia, voters from Aliança Catalana and CUP will have to look for another ballot. The very 40dB survey gives another hint: if one asks assuming Frente Amplio and Podemos run together, the “I don’t know” falls from 9.1% to 7.4%. It is hard to decide on an option that does not yet exist.
“In Germany and the United Kingdom, the call fixed a date to a known lineup of parties. In Spain, to the left of the PSOE, that lineup remains open”
My hypothesis is that, with a fragmented party system, Spain resembles the German case somewhat more than the British one, where the winner-takes-all system and a comfortable Labour advantage left little room for doubt about the outcome. In Germany, the reduction in uncertainty about the vote concentrated among those who did not identify with any party, and in the United Kingdom the effect was more uniform across the electorate.
How to read the polls in the coming weeks
First, you must compare each pollster with itself, because the percentage of undecideds depends on how and to whom you ask: in September, the “I don’t know” stood at 8.4% for 40dB and 14.3% for the CIS. And second, it matters when it changes: a sharp drop right after the consolidation of candidacies, before the pre-campaign and the campaign begin, would point to the calendar; a gradual drop later, to campaign effects.
“A sharp drop right after the consolidation of candidacies, before the pre-campaign and the campaign begin, would point to the calendar; a gradual drop later, to campaign effects”
Declaring a voting intention earlier does not guarantee you will keep it or even go to the polls. Our study measures the starting point, not the finish line; but the outcome is clearly key. When, as Alma Ruiz shows in these pages, half a point of provincial vote could swing several seats, the pace at which each group of voters makes up their minds could be almost as important as the direction of their vote.
