With nine days until the general elections on 29 November, the PSOE narrows the gap with the PP without altering the majority that the polls attribute to the right. The latest edition of The survey of surveys places the conservatives at 128 seats, the socialists at 111 and Vox at 65. Calculated before rounding the seats of each party, the sum of PP and Vox reaches 194 representatives, compared with 195 on Monday. The absolute majority in Congress is 176.
The move is more visible in some polls than in the average. The two daily tracking reports published this Wednesday agree on placing the PP first and the PSOE second. From there, they draw two very different seat distributions. In 40dB., the PP obtains 133 seats and Vox 55. In Sondaxe, the conservatives drop to 121 and Vox rises to 71. Between the two studies there are 16 seats of difference for Vox, but only four in the sum of the right: 188 in the first and 192 in the second, including the UPN-PP candidacy in the popular figure of both.
What changes since Monday: the PSOE cuts the PP’s lead to 3.3 points
The PP goes from 30.9% to 30.8% of the estimated vote. The PSOE moves from 27.3% to 27.5% and Vox from 19.0% to 19.1%. With the values of each institute before rounding, the gap between the first two falls from 3.5 to 3.3 points. PP and Vox together keep 49.9% of the vote in the two editions.
The comparison includes a novelty. To the 18 firms from Monday, InvyMark for laSexta is added, which places the PP at 32.2%, the PSOE at 28.2% and Vox at 19.1%. Its inclusion makes it possible to check whether the socialist improvement depends on having expanded the sample. The result hardly changes: taking only the same 18 firms from both days, the PSOE also reaches 27.5% and the PP stands at 30.8%. The difference is seen in Vox, which records 19.0% with the shared firms and 19.1% when including InvyMark.
| Party or indicator | Nov. 16 | Nov. 18 | Change* |
|---|---|---|---|
| PP · vote | 30.9% | 30.8% | −0.1 points |
| PSOE · vote | 27.3% | 27.5% | +0.2 points |
| Vox · vote | 19.0% | 19.1% | +0.1 points |
| PP advantage over PSOE | 3.5 points | 3.3 points | −0.2 points |
| PP + Vox · vote | 49.9% | 49.9% | No change |
| PP · seats | 129 | 128 | −1 |
| PSOE · seats | 110 | 111 | +1 |
| Vox · seats | 66 | 65 | −1 |
| PP + Vox · seats | 195 | 194 | −1 |
| Margin over the absolute majority | 19 seats | 18 seats | −1 |
Source: editorial calculation by Demócrata. Vote: 18 institutes on Monday and 19 on Wednesday. Seats: 17 in both cuts. *Change between the values shown. The advantage, the sum of parties and the margin over 176 are calculated before rounding. The PP category includes UPN-PP when the institute publishes it grouped together.
Why 40dB. gives Vox 55 seats and Sondaxe reaches 71
The difference begins in the vote estimate. 40dB. places Vox at 17% and Sondaxe, at 19.4%. They also disagree about the strength of the PP: the tracking for EL PAÍS and the Cadena SER attributes 32.2%, compared with the 29.0% of the study for La Voz de Galicia. In both surveys, much of the distance between their predictions for Vox is offset by the PP estimate in the opposite direction.
The PSOE also gets better numbers in 40dB.: 30.7% and 122 seats, versus Sondaxe’s 27.0% and 111 representatives**. The interview periods overlap. The first worked between 13 and 17 November, with 2,000 interviews; the second, between 12 and 16, with 1,000. Dates alone, hardly explain such a distance.
Not all studies place the right on the same side of the majority, either. Data10 for OKDiario gives 202 seats, with 131 for the PP and 71 for Vox. At the other end of the updated deliveries, SW Demoscopia places the PP at 118 and Vox at 54: 172 between the two, four below the absolute majority. The average of 194 also sums up that discrepancy.
| Institute and outlet | Fieldwork or release | Sample | PP | PSOE | Vox |
|---|---|---|---|---|---|
| 40dB. · EL PAÍS y Cadena SER | Fieldwork: 13–17 Nov. | 2,000 | 32.2% · 133 | 30.7% · 122 | 17.0% · 55 |
| Sondaxe · La Voz de Galicia | Fieldwork: 12–16 Nov. | 1,000 | 29.0% · 121 | 27.0% · 111 | 19.4% · 71 |
| Sigma Dos · El Mundo | Fieldwork: 11–16 Nov. | 3,759 | 30.7% · 123–129 | 26.4% · 108–114 | 19.8% · 62–67 |
| SW Demoscopia · La Séptima | Release: 17 Nov. | 9,500 | 28.9% · 118 | 31.9% · 134 | 16.5% · 54 |
| Data10 · OKDiario | Fieldwork: 11–12 Nov. | 1,500 | 31.7% · 131 | 25.5% · 104 | 20.4% · 71 |
| InvyMark · laSexta | Fieldwork: 9–13 Nov. | No data | 32.2% | 28.2% | 19.1% |
Each cell indicates vote percentage and, when published, central estimate or range of seats. InvyMark provides only vote. Sources: studies by the institutes for the indicated media. Consultation cutoff: morning of 18 November 2026, peninsular time.
Compared with 23J, Vox almost doubles its representation
The change over two days is small compared with what the comparison with the last general elections shows. The average leaves the PP at 128 seats, nine fewer than the 137 it obtained separately on 23J. The PSOE appears with 111, ten below its 121 then. Vox stands at 65, versus the 33 it achieved in the ballot boxes. It is the party that most alters the distribution among the three main forces with respect to 2023.
Thus, the PSOE’s advance from Monday and the right’s parliamentary advantage coexist in the same snapshot. The socialists recover two tenths in the average and a rounded-off seat. For PP and Vox, the change at the margin over the absolute majority is from 19 to 18 seats.
How DEMÓCRATA calculates the survey from the surveys
To produce La encuesta de las encuestas, DEMÓCRATA uses a simple arithmetic mean of the latest measurement collected from each institute, with the same weight for all and including the CIS. Each new delivery of a tracking replaces the previous one from that firm. The calculation for 18 November brings together 19 institutes for the vote and 17 for seats. InvyMark and More in Common contribute only percentages.
In addition to the six firms in the table, the calculation also incorporates NC Report, IMOP, SocioMétrica, DYM, GAD3, Ateneo del Dato, Winston, Celeste-Tel, Target Point, Hamalgama Métrica, More in Common, YouGov and the CIS. For seats, the central estimate published by each institute is used, or the mid-point when it only offers a range. Each total is rounded at the end. That’s why the PP’s 128 and Vox’s 65, shown separately, add up to 194 when calculating the joint average before rounding.