29N indecisives: how many are left, when they will decide, and who can win their votes

The last week of 29N begins with around 8% of respondents who do not declare a voting option in 40dB’s tracking. The precedent from 2023 shows that a quarter of voters decided in the final stretch. The left has more previous support to recover, but that mobilization is not guaranteed.

fotonoticia 20261116135359 1920

fotonoticia 20261116135359 1920

Add DEMÓCRATA to Google

Ask FREN

Published

Last updated

6 minutes

Most read

Elecciones 29N · Encuestas

La media de las encuestas

Media a 22 de noviembre de 2026 · 20 institutos

  • PP31,0 %
  • PSOE28,3 %
  • Vox18,7 %
  • Frente Amplio6,0 %
  • ERC2,6 %
01020304050ene 2024jul 2024ene 2025jul 2025ene 2026jul 2026

Media del último estudio de cada instituto en 30 días y su evolución semanal desde el 23J. Partidos con al menos un 2 %.

Todas las encuestas

Elecciones generales del 29 de noviembre · Demócrata

The general elections of 29 November enter their final week with part of the electorate still without a stated decision. The 40dB. tracking has recorded a reduction in undecided voters from 11% to 8% since the start of its interviews, on 10 November, according to the data published by EL PAÍS and provided for this analysis.

The drop indicates that the campaign is clearing up doubts, but it does not, by itself, reveal who is gaining those supporters. Stopping the response “I don’t know” can mean choosing a party, deciding to abstain, or communicating a preference that was previously kept hidden.

The precedent from the last general elections shows how much this stretch can matter. A 15.9% of those who voted in 2023 decided during the final week of the campaign, a 3.3% on the day of reflection, and a 6.8% on the same day as the elections. Together they add up to 26%: approximately one in four voters.

How many undecided voters remain for 29N

The most recent reference from the tracking collected in the starting information places indecision around 8%. The analysis that groups several polls from October and November places it close to 9%. These are measures from different time periods and should not be presented as two contradictory results from the same day.

There is another decisive clarification: the study includes as undecided both those who answer that they don’t know what they will vote for and those who prefer not to respond. Therefore, that percentage does not exactly correspond to people who are available to change their minds.

It also cannot be automatically converted into a closed number of millions of votes. To do that, it would be necessary to know the universe represented by each poll, the turnout of that group, and how many will ultimately end up choosing a candidate.

The useful figure for tracking the campaign is the evolution within the same series: from 11% to 8%, three points fewer responses without a declared option. That does not mean that three points of the electorate have moved to the PSOE, the PP, or any other political formation.

When the vote is decided: the campaign does not end for the voter on Friday

The week that begins on Monday 23 has three different moments: the final days of the campaign, the day of reflection on Saturday 28, and the voting on Sunday 29.

Historical data make it possible to measure the weight of each stretch, but not to allocate it among Monday, Tuesday, or Wednesday. There is no basis to claim that a specific percentage of today’s undecided voters will be resolved each day.

The 2023 decision calendar offers a benchmark:

  • Last week of the campaign: 15.9% of voters.
  • Day of reflection: 3.3%.
  • Election day: 6.8%.

It is behavior observed in the previous general elections, not a prediction for 29N. The denominator also changes: these percentages refer to those who ultimately voted; the current 8% comes from responses to a survey before the elections.

When the vote is decided: the 2023 benchmark
Moment of the decision Voters Equivalent 29N calendar
Long before the campaign 61.0% Before the start of the campaign
At the beginning of the campaign 12.5% First days
Last week of the campaign 15.9% Final stretch up to Friday 27
Day of reflection 3.3% Saturday 28 November
Election day 6.8% Sunday 29 November
Last week + reflection + elections 26.0% Historical data, not a forecast

Source: CIS, study 3420, question 7a. Percentages among those who said they had voted in 2023 and identified their option, including blank or null vote. The published categories add up to 99.5% due to rounding and responses not included. The correspondence with the 2026 dates is indicative: it does not forecast daily decisions.

The political consequence is concrete: the end of the campaign does not close the electoral decision. On Saturday and Sunday preferences, doubts about which candidacy to choose, and decisions about whether to go vote can still be resolved.

Why the 8% of undecided voters does not contradict the 26% that decided at the end

A person can answer today that they will vote for a party and change their option afterward. Another can maintain their preference for weeks, but remember after the elections that they made the definitive decision during the last few days.

That is why stating an intention to vote and considering it closed are not the same thing. The “doesn’t know” or “doesn’t answer” response pool does not capture all the uncertainty of the campaign.

There are also convinced voters about their party who doubt whether they will go to the polling station. And people who are sure they will participate keep comparing two candidacies. These are different decisions and they can be resolved at different times.

Research on Spanish campaigns distinguishes between reinforcing a preference, activating a prior inclination, and changing options. Polls can provide information for that decision, but its effect depends on the competition between parties and the electoral context; there is no automatic direction of late voting.

Who it can benefit: the left has more previous votes to recover

The 40dB. cross-checks, published by EL PAÍS, place the greatest doubts among those who remember having voted for Sumar in 2023: 7.8% still do not declare an option. Among former PSOE voters, the percentage is 4.8%; among those of the PP, 3.5%; and among Vox, 1.2%.

These percentages describe indecision within each previous electorate. They do not mean that Sumar represents 7.8% of all undecided people, nor that those votes will necessarily return to its political space.

For Sánchez, the reading is favorable as an opportunity, but demanding as a task: he has more room than the right to recover previous support still without definition. For that room to translate into an improvement, those people must turn out to vote and choose the PSOE or a candidacy compatible with a parliamentary majority.

If they abstain, the left’s greater indecision becomes a weakness. If they choose another progressive formation, they can strengthen that block without directly improving the PSOE’s position versus the PP.

Undecided people according to their declared vote in 2023
Previous electorate Undecided within the group
Sumar 7,8%
PSOE 4,8%
PP 3,5%
Vox 1,2%
Other parties 8,1%

Source: analysis by 40dB. published by EL PAÍS, with surveys from October and November up to day 20. Undecided people: answers "I don’t know" or "I prefer not to answer." The percentages correspond to each previous electorate, not to the set of undecided people as a whole. They do not identify which party they will ultimately vote for.

The PP can also win with undecided voters

The lowest indecision among former PP and Vox voters indicates a greater definition of their stated preference. It does not demonstrate that all of them will turn out or that they will keep the same party.

Feijóo may benefit from voters who are looking for a change of government, from decisions between PP and Vox, or from voters located outside those previous electorates. The origin of an undecided voter does not determine their destination.

The ideological cross-check reinforces that caution. Indecision reaches 15.2% among those who place themselves in the center, versus 6.4% among those who place themselves on the left and 3% on the right. Among those who do not declare an ideological position, it reaches 38.6%.

It cannot be inferred from those data that the majority of undecided voters is centrist: to know their composition, one also has to consider how many people make up each group. What they do show is that the undefined vote overflows the confrontation between two perfectly delimited electorates.

What would have to happen for the last week to change the result

For undecided voters to appreciably alter the distance between parties, two movements must coincide: that a sufficient part ends up voting and that their support is distributed in an uneven way.

A purely illustrative example makes it possible to understand this. If 100,000 new voters split 50% between two candidacies, both grow without modifying their distance in votes. If they split 60,000 versus 40,000, the difference changes by 20,000.

The decisive factor is not only how many people mobilize, but who obtains the net advantage and where those votes are. In a constituency with the last seat being contested, a small variation can change representation; in another, a larger increase may not change any MP.

There is one final limit for interpreting the upcoming polls: their estimates may already incorporate hypotheses about turnout and the destination of undecided voters. It does not correspond to add afterwards that 8% to the published results as if it were a pool of votes that the models had left completely out of account.