Spain General election29 Nov 2026in 8 days PP + Vox majority 99% 2023 investiture bloc 1% Calendar Track record Research
Research · Spain
EN ES

From today to polling day in Spain

Abstract. We test the British pull-back towards the previous result on Spanish polls at six distances up to a year before the general elections of 2011 to 2023. It did not beat the polls as they stood at any distance, because change between elections has been large and has lasted. The polling-day forecast therefore keeps today's shares and widens the ranges by a drift fitted to how far polls have missed at each distance.

Polling Forecast · updated 9 November 2026 · 9 min read · Spain, election day, polling horizon, drift, pull-back, early election

Summary

Everything else in the Spain section answers the question "what if the election were held today?". The next general election is not today: at the latest it is due on 29 November 2026, 8 days after the most recent poll. This paper describes how we get from today to polling day.

In Britain we found that the older parties finish much closer to their previous result than their mid-term polls suggest, and we pull them part of the way back (From today to polling day: how far British polls move before a general election). We ran the same test on Spanish polls, at six fixed distances from one month to one year before the five general elections from 2011 to 2023 (November 2019, a re-run held six months after April, is left out). It failed. At no distance did pulling parties back towards their previous result beat the polls as they stood, because in Spain the change between elections has been large and has mostly lasted. So our polling-day forecast keeps today's shares as they stand and adds only uncertainty, which grows with the time left: polls this far out have missed each party by about 0.4 points by the time votes were cast.

What has changed since this paper was written

Updated 9 November 2026. The rest of the paper is as first published, except where a sentence below says it was corrected.

  • An early election (5 October). Pedro Sánchez dissolved the Cortes; the general election is on Sunday 29 November 2026. The forecast, the election-day model and the days-to-go figures use that date, and polls may not be published from 00:00 on 24 November (Madrid time) until the polls close.
  • Only elections from 2000 shape the forecast (28 September). Earlier elections stay in the long view and on the history pages but no longer feed any test, correction or error size. In Spain nothing before 2000 was used, so nothing changed.

The question: today is not polling day

A poll taken a year before an election records how people say they would vote at that moment, before a campaign and before the lists are known. The literature on the timeline of elections finds that polls become steadily more informative as polling day approaches, and that early polls carry movement that has gone by the time votes are cast (Wlezien et al., 2013; Erikson and Wlezien, 2012). Whether that holds in a given country is an empirical question. The error at final polls (Measuring Spanish polling error, 2004–2023) answers a narrower one: how wrong were the last polls of a campaign? The polling-day forecast needs its own evidence, and it is kept apart from the "today" forecast on the site.

The evidence: polls at fixed distances before five elections

Source

The evidence is every national voting-intention poll in our series for the twelve months before each general election from 2011 to 2023: about 1,200 polls, most of them taken from Wikipedia's per-election tables, with the CIS's studies from its own site and a share checked against the pages the tables cite (The Spanish data: results, maps, statistics, deputies and polls). The series does not go back further than 2011, and it stops at one year before each election. November 2019 is left out: it followed the April 2019 election by six months, so its "previous result" was a recent snapshot of the same electorate, and the point a year before it fell before the April election.

Horizon points

For each of the five elections we average the polls at six fixed distances before polling day: 30, 61, 91, 182, 273 and 365 days. At distance \(d\) the window is the 60 days centred on the date \(d\) days before the election. Within it each pollster's polls are averaged, and the pollsters' averages are then averaged with equal weight. Each record, a horizon point, holds the election, the distance, the party, the poll average \(\bar{y}_{p,t}(d)\), the party's result at the previous general election \(v_{p,t-1}\) and its result at this one \(v_{p,t}\). The left of the PSOE is measured as one lineage, as in the rest of the Spanish record. A party with no previous result is left out of the test.

That gives 282 horizon points.

The prior

The alternative to a poll taken months out is to assume nothing has changed since the last election. We call the previous result the prior. A forecast for polling day can then be framed as a choice of how far to move from the prior towards today's polls, and the data can say how far that has been worth moving at each distance.

What the data show

"Polls' miss" is the average absolute difference between the poll average at that distance and the result, per party. "Previous result's miss" is the same for the prior. "Pull-back's miss" is the model described below, each election predicted from a curve fitted to the other four. "RMSE" is the root-mean-square miss of the polls as they stand, which sets the uncertainty.

Days before polling dayElectionsPolls' missPrevious result's missPull-back's missRMSE
3050.883.130.881.44
6150.993.130.991.66
9151.213.131.212.00
18251.403.151.422.36
27351.313.151.342.38
36551.493.151.522.97
Live Congreso: polls against the result by distance from polling day, each election predicted from the others Computed from the database, polls to 19 November 2026.

Two findings stand out.

The polls beat the previous result at every distance. The previous result missed by about 3.1 points per party. The polls missed by under a point at one month and by about one and a half at a year. Even a year out, the polls were twice as close to the result as the last election was.

Change between elections has been large, and it has lasted. Spain's party system changed more between 2011 and 2019 than at any time since the early 1980s. Podemos and Ciudadanos went from no seats to 109 between them in 2015; Vox, on 0.2% in 2016, won 10.3% in April 2019 and 15.1% in November; Ciudadanos fell from 15.9% to 6.8% in six months and did not stand in 2023. A mid-term poll in that setting was a better guide than the last result, because the mid-term movement was the new party system forming.

Election by election, the polls' miss at each distance was:

Days out 2011 2015 2016 April 2019 2023
30 0.83 0.76 1.12 0.89 0.80
61 0.93 0.80 1.12 0.96 1.11
91 0.98 0.98 1.21 1.22 1.65
182 1.08 0.76 1.27 2.64 1.26
273 0.77 0.58 1.23 2.68 1.29
365 0.42 1.03 0.94 3.72 1.35

Table: Polls' average miss per party at each distance and election, points.

April 2019 stands out. A year before it the polls missed by 3.7 points per party, because the year changed the contest: Pedro Sánchez became Prime Minister through a motion of censure in June 2018, and Vox broke through at the Andalusian election of December 2018. The other four elections look alike, with misses between about half a point and one and a half at every distance. In 2011 the polls a year out were closer than those a month out, which is what a settled election looks like: the PP's lead over the PSOE was large and stable from a year before.

The pull-back curve and its test

Form

We tested the same model as in Britain. The projected polling-day share is a weighted average of today's share and the previous result:

\[ \hat{v}_p(d) = v^{\,\mathrm{prior}}_{p} + w(d)\,\bigl(y_p - v^{\,\mathrm{prior}}_{p}\bigr), \qquad w(d) = f + (1 - f)\,e^{-d/\tau}, \tag{1} \]

where \(y_p\) is today's share, \(d\) the days to polling day, \(f\) the floor and \(\tau\) a time constant in days. The weight is 1 on polling day and falls towards \(f\) with distance. As in Germany, the curve is fitted to every party with a previous result.

Fitting

The two parameters are chosen by least squares over all horizon points, searching \(f\) from 0 to 1 in steps of 0.01 and \(\tau\) from 5 to 600 days in steps of 5:

\[ \mathrm{SSE}(f, \tau) = \sum_{t}\sum_{d}\sum_{p} \Bigl( v_{p,t} - v_{p,t-1} - w(d; f, \tau)\,\bigl(\bar{y}_{p,t}(d) - v_{p,t-1}\bigr) \Bigr)^2 . \tag{2} \]

The best fit on all five elections is \(f = 0.91\) and \(\tau = 600\) days, the longest time constant searched. That curve counts almost all of the change since the last election at one month out and 96% of it at a year. It is barely a pull-back at all.

The test

Each election is predicted with the curve fitted to the other four only (Stone, 1974), and the pull-back is kept only if, averaged over every distance, it beats the polls as they stand by more than 0.01 points per party. It does not. At one, two and three months it made no difference, and at six, nine and twelve months it was worse by two or three hundredths of a point. The site therefore sets the floor to 1: the polling-day shares are today's shares. The fitted curve is shown on the election-day page, labelled as not used.

Drift: uncertainty that grows with the time left

The pull-back fails, but the uncertainty it was meant to address is real. A poll a year out misses by about twice as much as a poll a month out. The polling-day forecast keeps that uncertainty and calls it drift.

The drift curve

Drift is measured by the root-mean-square miss of the polls as they stand at each distance, the RMSE column above. A logarithmic curve fits it:

\[ \delta(d) = a + b\,\ln\!\left(\frac{d}{30}\right), \qquad a = 1.35,\; b = 0.56, \tag{3} \]

fitted by ordinary least squares over the six distances. At 8 days it is 0.4 points. Under 30 days the drift is scaled down in proportion to the days left, so that it reaches zero on polling day, and it is never less than half a point.

How the drift is applied

The drift is a figure for a party of middling size. In the simulation it is scaled by the square root of the party's share relative to 15%, so that large parties get more and small parties less:

\[ \sigma_p(d) = \sqrt{\sigma_p^2 + \left(\delta(d)\sqrt{\max(S_p, 1)/15}\right)^2}, \tag{4} \]

where \(\sigma_p\) is the party's final-poll spread from Simulating Spanish polling error: where the chances come from and \(S_p\) its share today. A party on 30% gets about 1.4 times the drift, a party on 2% about 0.4 times. The square root is a simple way to let the drift grow with a party's size, since a party on 2% cannot plausibly move by as many points as one on 30%; it was chosen, not fitted.

Polling-day ranges

The polling-day forecast runs the same simulation as the "today" forecast, with 1,000 runs and equation (4) in place of each party's final-poll spread. Because the floor is 1, the centre is today's shares, and the whole difference between the two forecasts is the width.

PartySeats today80% range today80% range, polling dayWins a seat, polling day
PP130116–142116–142over 99%
PSOE115101–127100–127over 99%
Vox6759–7859–78over 99%
ERC118–137–14over 99%
EH Bildu86–106–10over 99%
Sumar75–145–14over 99%
Junts51–81–895%
PNV42–62–699%
BNG22–52–696%
CC10–20–371%
SALF00–00–07%
Podemos00–00–08%
Live Today against polling day (8 days away at most) Computed from the database, polls to 19 November 2026.

The ranges for the large parties are much wider. For the small and regional parties the chance of winning a seat falls, because a wider national spread makes it more likely that a party slips below the threshold of exclusion in the provinces where it now wins. The combinations move the same way:

CombinationTodayPolling day
PP + Vox99%99%
PP + Vox + UPN + CC99%99%
PSOE + Sumar + Podemos0%0%
Sánchez's 2023 investiture bloc1%1%
PP + PSOEover 99%over 99%
Live Chance of 176 seats or more, today and on polling day Computed from the database, polls to 19 November 2026.

A combination that reaches 176 in almost every "today" run reaches it in fewer polling-day runs; one that almost never reaches it today reaches it in a few.

The date, and an early election

A polling-day forecast needs a date. The Cortes are elected for four years (Constitution, art. 68.4), and the site uses 29 November 2026 as the latest date for the next election. That is not a prediction that the election will be held then. It is the one date fixed by law, and the cautious choice: the longest horizon gives the widest ranges.

An earlier election is possible at any time after the first year of the Cortes, on the Prime Minister's proposal (art. 115). Two of the last four general elections were called early. An early election shortens the horizon and narrows the drift: at six months it would be about 2.4 points, at three months about 2.0. The distance is counted from the date of the latest poll in the average, not from the day the page is viewed.

Limitations

Five elections, one year. The test rests on five elections and on distances up to a year. The next election is less than a year away, so the forecast stays inside the tested range, but a longer horizon would rest on extrapolation.

The drift includes the final-poll error. The RMSE at each distance already contains the error that remains at the end of a campaign, and equation (4) adds the final-poll spread again. At a month or two out this makes the ranges somewhat too wide. We prefer that to ranges that are too narrow, but the choice is not calibrated.

Independent drifts. Drift is drawn independently for each party. In practice one party's gain is usually a particular rival's loss.

No fundamentals. The model has no economic data, no leader ratings and no regional elections. Each would add parameters that five elections cannot estimate.

New parties. A party that forms after today's polls has no share to carry forward. Spanish politics produced Podemos in 2014, Vox's breakthrough in 2018 and SALF in 2024; the forecast cannot anticipate the next one.

References

Constitución Española (1978), articles 68 and 115.

Erikson, R. S. and Wlezien, C. (2012). The Timeline of Presidential Elections: How Campaigns Do (and Do Not) Matter. Chicago: University of Chicago Press.

Stone, M. (1974). Cross-validatory choice and assessment of statistical predictions. Journal of the Royal Statistical Society: Series B, 36(2), 111–147.

Wlezien, C., Jennings, W., Fisher, S., Ford, R. and Pickup, M. (2013). Polls and the vote in Britain. Political Studies, 61(S1), 66–91.

How to cite

Polling Forecast (2026). From today to polling day in Spain. Retrieved 21 November 2026 from https://www.pollingforecast.com/research/spain-election-day. Live figures change as new polls arrive; the date above is when the text was last revised.

You may also like