Poland Parliamentary electionby 11 Nov 2027in 355 days Government majority 10% Right-wing majority 86% Calendar Track record Research
Research · Poland
EN PL

Forecasting the Sejm from imperfect polls: an overview

Abstract. The Poland section forecasts the next Sejm election with the rule used for Great Britain: take half the average past miss off today's polls, but only where that would have improved past forecasts made without knowing the result. Polish final polls fail that test and are taken as published, with the don't-knows removed. The seats depend on the 5% and 8% thresholds as much as on the votes, and on a line-up the parties have not yet agreed, so the forecast states its assumption and shows three others; our calculation reproduces every Sejm election since 2005. Because the President can veto laws, the forecast reports 276 seats as well as 231. This paper maps the other Poland papers.

Polling Forecast · updated 9 November 2026 · 9 min read · Poland, Sejm, parliamentary election, polling error, leave-one-out, thresholds, D'Hondt, okręgi, veto, overview

Summary

This paper is the way into the Poland section of Polling Forecast. It says what the section forecasts, why Polish polls are not corrected, what decides seats in the Sejm, why the forecast has to assume who runs with whom, why 276 seats matter as much as 231, and how the uncertainty is measured. The last section is a guide to the other Poland papers.

The short version is this. Polish final polls, taken a few days before the vote, have missed each list by about 1.7 points on average at the seven Sejm elections since 2001, with large misses in 2001 and 2005 and smaller ones since. Some lists were missed the same way more than once (the lists that became Trzecia Droga were understated at five of seven elections), but the directions changed too often for a correction to help: taking off half the average past miss, with each election predicted from the others, was closer at three elections and further out at four, and further out on average. So the forecast takes the polls as they are. The seats then depend on the thresholds as much as on the votes: a party list needs 5% of the national vote and a coalition list 8%, and several parties poll close to those lines. Who runs with whom is not yet known, so the forecast states its assumption and shows three others. Our seat calculation reproduces every Sejm election since 2005 exactly.

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.

  • 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 Poland the 1993 to 1997 seat check is still shown but no longer counts.
  • Our figures beside the polls (26 October). The correction is off in Poland, so our projected shares are the poll average, and the pages say so rather than repeating the figures.
  • Pollster grades rescaled (13 October) (Rating Polish pollsters: accuracy, the lead and openness).
  • When the polls missed (2 November): Ten points clear, then second: how the polls missed Duda's first-round win in 2015.
  • The map (November). The map can be coloured by any census figure; each census figure also has its own page (census maps). It can also replay a past polling miss ("The polls miss again") or take a vote you set.

What the Poland section forecasts

The forecast is for the next election to the Sejm. The Sejm first sat on 13 November 2023 and sits for four years; the election must be held on a non-working day in the last 30 days of the term, and the President sets the date. The latest possible date is 11 November 2027, a public holiday, which the site uses. The Sejm can shorten its own term with 307 votes.

The Sejm has 460 seats, elected in 41 okręgi of 7 to 20 seats each. A majority is 231. The Senate, elected the same day in 100 single-seat districts, can amend or delay laws, and the Sejm overrides it by an absolute majority; the site forecasts the Sejm only.

The section answers five questions, each on its own page:

  • How many seats would each list win if the election were held today? (The home page and the okręgi and regions pages.)
  • Which lists clear the threshold, and what happens if parties run together?
  • What is the chance that the governing coalition, or the parties to its right, reach 231, and 276?
  • Which lists win seats in each okręg, and which last seats are close?
  • How different could the result be on polling day? (The election-day page.)

The headline, as of 19 November 2026, is below, under our default line-up: every party on its own list. It is built from the latest poll of each of 8 pollsters.

ListPollsThresholdSeats80% range2023Clears the threshold
KO32.65%191168–206157over 99%
PiS22.05%131117–143194over 99%
Konfederacja14.05%7359–8518over 99%
KKP8.85%4127–49–over 99%
Lewica7.45%240–432686%
Rozwój Plus4.75%00–15–38%
Razem4.35%00–10–21%
PSL3.45%00–0–1%
Polska 20501.85%00–0–0%
Live Our projection by list (every party on its own list) Computed from the database, polls to 19 November 2026.

On these figures the largest list would be KO, with 191 seats. The governing coalition's parties would have 215 seats and reach 231 in 10% of simulated elections; PiS, Rozwój Plus, Konfederacja and Braun's KKP together would have 245 and reach 231 in 86%.

One rule for every country

The British forecast starts from a finding: at every general election from 2015 to 2024, England's final polls missed in fairly steady directions, and a correction built from those misses would have brought the forecast closer to the result (see Forecasting a British general election when the polls are wrong: an overview and Correcting the raw answers: the method behind our UK projection). The rule we took from that work is the same in every country.

For each party \(p\) and past election \(t\), let \(e_{p,t}\) be the average of the final polls minus the result, in points. The correction takes off half of the average past miss:

\[ \hat{v}_p = y_p - \lambda\,\bar{e}_p, \qquad \bar{e}_p = \frac{1}{T}\sum_{t=1}^{T} e_{p,t}, \qquad \lambda = \tfrac{1}{2}, \tag{1} \]

where \(y_p\) is today's poll average, followed by a rescale so the shares keep the same total. A party is corrected only if it has been measured at two or more elections. The rule is kept only if, with each past election predicted from the others alone (Stone, 1974), its average miss per party beats the uncorrected polls by more than 0.01 points. In Italy and the Netherlands it does. In Germany, Spain and Poland it does not.

Why Polish polls are not corrected

The table below is the evidence: the average of the final polls minus the result, at each Sejm election since 2001. Lists that merged, split or changed name are measured by their line to the five lists of 2023: PiS, KO, Trzecia Droga, Lewica and Konfederacja.

Party2001200520072011201520192023Average
KO+1.2+6.5−0.6+1.7−1.2−1.8−0.9+0.7
Lewica+6.8−5.6+0.10.0−0.2+1.2+1.2+0.5
PiS−1.3+2.8−0.1+0.1+0.3+2.6+1.5+0.8
TD (PSL + Polska 2050; earlier PSL and Kukiz'15)+0.2−2.1−1.3−1.4+0.4−1.6−3.7−1.4
Konfederacja––––−0.1−1.0+2.0+0.3
Live Sejm: final-poll average minus result, by 2023 lineage (points) Computed from the database, polls to 19 November 2026.

Some patterns look steady. The lists that became Trzecia Droga (the PSL, joined by Kukiz'15 in 2015 and 2019 and by Polska 2050 in 2023) were understated at five of seven elections, by 3.7 points in 2023. PiS was overstated at five. But the largest misses change sign: the left was overstated by 6.8 points in 2001 and understated by 5.6 in 2005; KO's line was overstated by 6.5 in 2005 and understated in 2015, 2019 and 2023.

The test:

ElectionPolls' average missCorrected, leave one outVerdict
20012.372.69further out
20054.274.31further out
20070.530.47closer
20110.810.70closer
20150.430.73further out
20191.631.65further out
20231.861.81closer
Average1.701.77
Live Correcting Polish polls, each election left out in turn Computed from the database, polls to 19 November 2026.

Correcting was closer at 3 of the seven elections and further out at 4, and further out on average: 1.77 points per party against 1.70 for the polls as published. That fails the rule, so the correction is off, and our projection takes the polls as they are. Reading Polish polls: why they are not corrected sets out the test.

Don't-knows and the election silence

Most Polish pollsters publish a share of respondents who don't know how they will vote inside the headline figures, typically five to twelve points. The forecast takes it out and rescales the parties to 100 before averaging. The final polls are taken three to five days before the vote: Polish law forbids publishing polls from midnight at the start of the day before polling day until voting ends (Kodeks wyborczy, art. 115).

What decides the seats

Thresholds

Only lists with 5% of the valid national vote take part in the share-out, or 8% for a coalition committee of several parties (Kodeks wyborczy, arts. 196 and 197). National-minority committees are exempt. A list under its threshold wins nothing in any okręg, and its votes are lost, which raises everyone else's seats. In 2015 the Zjednoczona Lewica coalition won 7.55% and no seats.

Today several parties poll near the lines: Rozwój Plus, the club Mateusz Morawiecki formed in July 2026 with 40 MPs who left PiS, just above 5%; Razem just under; the PSL and Polska 2050 well under. Line-ups and thresholds: 5%, 8% and the Trzecia Droga question sets out what that means.

D'Hondt in 41 okręgi

In each okręg the qualifying lists share the seats by D'Hondt. With 7 to 20 seats per okręg, the largest lists win a little more than their share: in 2023 PiS won 42.2% of the seats with 35.4% of the vote. From votes to seats: the Kodeks wyborczy gives the details and the check against every election since 2005.

The line-up

Parties decide who runs with whom only when the lists are registered, a few weeks before the vote. A joint list adds its parties' votes but needs 8% instead of 5%. The forecast treats the line-up as an explicit assumption and shows four:

Line-upGoverning coalitionPiS + R+ + Konf. + KKPLists under their threshold
Every party on its own list (our projection)215245Rozwój Plus 4.7, Razem 4.3, PSL 3.4, Polska 2050 1.8
PSL and Polska 2050 together again as Trzecia Droga215245TD 5.2, Rozwój Plus 4.7, Razem 4.3
Two government lists: KO with PSL and Polska 2050; Lewica alone232228Rozwój Plus 4.7, Razem 4.3
One joint list of the governing coalition244216Rozwój Plus 4.7, Razem 4.3
Live The same shares under each line-up: seats of the governing coalition and of the parties to its right Computed from the database, polls to 19 November 2026.

Our default is every party on its own list, because no joint list has been agreed. With today's shares, a single list of the whole governing coalition would win more seats for it than any other line-up, and every party on its own list the fewest, because the PSL and Polska 2050 fall below 5% and their votes are lost.

231 and 276

231 is a majority. But President Karol Nawrocki, elected in 2025 with the support of PiS, had vetoed more than twenty of the government's laws by February 2026, and a veto is overridden only by three fifths of the votes cast: 276 when all 460 vote. The governing coalition's clubs hold 239 seats today. The President's veto and the 276 line explains why the forecast reports both lines.

The forecast, step by step

Step 1: the average. The latest poll of each pollster whose fieldwork ended within 21 days of the newest poll, with the don't-knows taken out, equally weighted.

Step 2: our reading. The polls as they are, because the correction fails its test.

Step 3: okręgi. Each 2023 list's vote in each okręg, scaled by the national change of the parties that came from it (proportional swing), and shared among those parties in their national ratio.

Step 4: seats. The thresholds on the national vote under the line-up, then D'Hondt in each okręg.

Step 5: simulation. 2,000 simulated elections, each party moved by its typical final-poll miss nationally and a little more in each okręg (Simulating Polish polling error: where the chances come from).

Step 6: replays. Today's polls replayed with each past election's miss (If we're wrong: replaying past Polish polling misses on today's polls).

Step 7: polling day. Part of the way back towards the 2023 lists' results, and wider ranges for the time left (From today to polling day in Poland).

The uncertainty

The simulation gives the seat ranges, the chance that each list clears its threshold, and the chance that each combination reaches 231 and 276. The widest uncertainty is at the thresholds: a list near 5% wins either a dozen seats or none, and its lost votes move everyone else's seats. Beside the simulation, 41 okręgi change seats under at least one past miss replayed on today's polls. On polling day, 355 days after the latest poll at most, polls this far out have missed each party by about 4.1 points.

What we do not claim

We do not claim to know the line-up. The default is an assumption, stated on every page, and the other three are shown beside it.

We do not claim that the polls are unbiased. The test says only that a correction built from past misses would not have helped on average.

We do not claim okręg-level precision. Proportional swing from 2023 ignores candidates, and the new parties' votes are spread in the pattern of the lists they came from.

We do not forecast the Senate, the presidency or the government. A majority in the Sejm may still meet a presidential veto.

We do not claim that the seats per okręg will be the same at the next election. A bill passed in September 2026 would let the electoral commission reset them each year; it awaits the President's decision.

A guide to the papers

The Poland library has twelve papers. This one, Forecasting the Sejm from imperfect polls: an overview, is the introduction.

Measuring Polish polling error, 2001–2023 sets out the evidence: the final polls at the Sejm elections from 2001 to 2023, the don't-knows, the error list by list, the gap between the two largest lists, each pollster's record and the long view.

Reading Polish polls: why they are not corrected is the core methods paper: the correction rule, the leave-one-out test that switches it off, and the seat calculation.

From votes to seats: the Kodeks wyborczy covers the Kodeks wyborczy: the okręgi, D'Hondt, the thresholds, the German minority, the check against every election since 2005, and how okręg shares are projected from national polls.

Line-ups and thresholds: 5%, 8% and the Trzecia Droga question covers the line-ups and thresholds: 5%, 8%, the minority exemption and the Trzecia Droga question.

The President's veto and the 276 line covers the President's veto and the 276 line.

Simulating Polish polling error: where the chances come from explains where the chances come from.

From today to polling day in Poland asks how far Polish polls move before polling day, at distances up to three years before the elections of 2011 to 2023.

If we're wrong: replaying past Polish polling misses on today's polls describes the readings of today's polls: as they stand and with each past miss replayed.

Rating Polish pollsters: accuracy, the lead and openness describes the pollster ratings.

The Polish data: results, maps, statistics, MPs and polls documents where every number comes from.

What we don't know yet: limitations of the Polish forecast collects every known weakness in one place.

References

Jennings, W. and Wlezien, C. (2018). Election polling errors across time and space. Nature Human Behaviour, 2(4), 276–283.

Konstytucja Rzeczypospolitej Polskiej, articles 98 and 122.

Kodeks wyborczy (Dz.U. 2025 poz. 365), articles 115, 196, 197 and 232.

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

How to cite

Polling Forecast (2026). Forecasting the Sejm from imperfect polls: an overview. Retrieved 21 November 2026 from https://www.pollingforecast.com/research/poland-overview. Live figures change as new polls arrive; the date above is when the text was last revised.

You may also like