29N polls: votes, differences and possible agreements

Compare the estimates

One column per independent study: EL PAÍS and SER are one 40dB. poll. Sumar / Frente Amplio retains each source’s definition; a registered coalition or automatically additive electorates are not assumed.

Estimated vote · CIS · estudio 3582

Interviews: October 1, 2026 – October 6, 2026 · n = 5,025 · CATI

Shared zero-based scale: 0–40% of votes.

Interviews span the period before and during the election call; not all follow the announcement. Publishes valid-vote estimates, without seats. National shares are not converted into seats.

Votuo’s reading

The models disagree on first place. 40dB., GAD3 and Sigma Dos put PP ahead of PSOE; CIS puts PSOE ahead of PP. PP’s lead over PSOE ranges from +6.6 points in Sigma Dos to −11.8 in CIS. This contrast remains visible: it is neither a change in the electorate between two dates nor a confidence interval.

The clearest agreement concerns Vox: all four estimates place it at 17.1–19.1%, above its 12.4% in 2023. PP is below its 2023 share of 33.1% in all four. PSOE is below 2023’s 31.7% in the three private polls and above it in CIS. These comparisons use an election baseline; they do not establish movement since last week.

In the two usable seat allocations, PP–Vox totals 198 or 203 seats and PSOE–Sumar–Podemos 119 or 114. The first sum would exceed 176 if realised; it does not establish an agreement. CIS cannot establish a different parliamentary majority without seat projections. Sumar and Podemos are not separately compared with 2023, when they shared a list.

2023 baseline: rounded election results published in the 40dB. report, p. 5. 40dB. ↗

What support would add up to

Hypothetical sums, not announced agreements. 176 seats allow investiture at the first vote. Below that, abstentions may permit a simple majority at the second vote. Political feasibility and a shared programme require negotiation.

CIS publishes no seats. GAD3 is excluded from support calculations because its text and chart disagree. Only complete central allocations of 350 seats are used; independent ranges are not added and national percentages are not proportionally converted into seats.

Which measures would enter negotiations

A poll does not measure future legislation. These are conditional implications of published positions, not shared 29N manifestos or assessments of their outcomes. An agreement needs votes, funding, legislative procedures and respect for devolved powers and the European framework.

How this overview was compiled

Cutoff: publications from 1–14 October 2026 located and checked against primary sources. The most recent reviewed national Congress poll from each institute is included. This first overview covers four institutes, not an exhaustive inventory of every poll. Two publishers of one study are not counted as two samples.

We reproduce modelled vote estimates, not raw voting intention, and preserve published groupings. Pollsters are neither weighted nor adjusted, and CIS is not excluded for disagreeing. An average would mix methodological choices not yet validated here. Missing data are labelled, not replaced by zero or normalised to 100%.

Small differences and differences between institutes do not themselves establish a time trend. Sampling error does not cover all uncertainty from turnout, nonresponse, weighting or seat conversion. No winning probabilities are published; party ranges are not treated as a joint interval for a coalition.

ESOMAR / WAPOR 2026 · AAPOR

Methodology and sources