Reading Italian polls: the correction and its test
Abstract. The core methods paper for Italy. Taking off half the average past miss, with each election predicted from the others, was closer on average at the four Camera elections since 2008, by a small margin, so Italian polls are corrected. The gain comes from 2013 and 2018, when the M5S and the PD were missed most; the correction made 2008 and 2022 slightly worse. We set out what it does to today's polls, why there is no raw-figure test in Italy, and the seat calculation that reproduces 2022.
Summary
This paper sets out how Polling Forecast reads Italian opinion polls for the next Camera election, and why, unlike in Germany and Spain, it corrects them. Final polls at the four Camera elections since 2008 missed by about two points per list, and two lists were missed the same way time after time: the Five Star Movement (M5S) understated at every election it contested, the PD overstated at three of four. Taking off half the average past miss, with each election predicted from the others, was closer on average, by a small margin, so the correction is on. The seats then depend on the electoral law (the new law since 8 October 2026, as the note below explains) and on who stands with whom; our seat calculation reproduces the 2022 Camera exactly, and the line-up is an explicit assumption (Who stands with whom: the coalition line-up assumption).
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.
The new electoral law is in force (8 October). The Chamber gave the new law its final approval on 8 October 2026, by 227 votes to 164, and the Italian forecast has used it everywhere since. As the forecast applies it to national shares:
Seats are shared in proportion to the vote, by Hare quota and largest remainder, among coalitions on 10% or more and lone lists on 3% or more. Inside a coalition, lists on 3% or more, plus the best list below 3%, share its seats; every list's votes count towards the coalition's total.
The bonus. The coalition or list that comes first with at least 42% of the valid votes gets 70 extra deputies and 35 extra senators, up to caps of 220 deputies and 113 senators.
The Senate is worked out the same way on national shares (the law allocates its seats by region). Seats abroad, Valle d'Aosta and Trentino-Alto Adige are shared in proportion to the national vote, an approximation.
The collegi no longer elect anyone. The maps and collegio pages show who leads each former collegio.
Where this paper describes the Rosatellum, it describes the law under which the 2022 Parliament was elected: it remains the basis of the check against 2022 and of the by-elections in this Parliament. /it/new-law shows the seats under both laws side by side.
- 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 Italy the tests were already on elections since 2001, so nothing changed.
- Our figures beside the polls (26 October). The polls page shows our projection beside the poll average where they differ; the hub shows the coalitions.
The electoral system
Each voter has one ballot and one vote, which counts twice: for a list, in the proportional count, and for that list's coalition candidate in the voter's collegio. A vote for the candidate alone is shared among the lists supporting them. Split tickets are not allowed. The rules are in the Testo unico for the Camera (DPR 361/1957) as amended by legge 165/2017, the "Rosatellum", and adapted to 400 seats in 2019 and 2020.
- Collegi uninominali. 147 seats, one per collegio, won by the candidate with the most votes.
- Proportional seats. 245 seats shared nationally, by largest remainders, among coalitions with at least 10% of the vote (with at least one list at 3%) and lists with at least 3%; each coalition's seats are then shared among its lists with 3%. A list in a coalition with between 1% and 3% adds its votes to the coalition but wins no seats. Lists of linguistic minorities qualify with 20% of the vote in their region, or by winning two collegi.
- Abroad. 8 seats for Italians resident abroad, elected separately.
A majority is 201.
The data
Results by collegio for 2022 and 2018 come from the Ministry of the Interior's figures, through an unmodified public copy because the Ministry's own sites refused our connections; collegio boundaries and statistics come from ISTAT; seats and deputies from the Camera's open data. Polls come from pollsters' and outlets' releases where the figures could be read, and otherwise from Wikipedia's tables, graded C and marked on the site. The Italian data: results, maps, statistics, deputies and polls gives every source.
The seat calculation
Our implementation of the law, given the official 2022 list votes and the 2022 line-up, reproduces the official result: the winner of every one of the 147 collegi, every list's proportional seats, and every coalition's total.
| Coalition | Collegi, ours | Collegi, official | Seats, ours | Seats, official |
|---|---|---|---|---|
| Centre-right | 121 | 121 | 237 | 237 |
| Centre-left | 13 | 13 | 85 | 85 |
| M5S | 10 | 10 | 52 | 52 |
| Azione–Italia Viva | 0 | 0 | 21 | 21 |
| SVP | 2 | 2 | 3 | 3 |
| Others | 1 | 1 | 2 | 2 |
| Collegi with the same winner | 147 | 147 |
| List | Our calculation | Official |
|---|---|---|
| FdI | 69 | 69 |
| PD | 57 | 57 |
| M5S | 41 | 41 |
| Lega | 23 | 23 |
| FI | 22 | 22 |
| Azione–Italia Viva | 21 | 21 |
| AVS | 11 | 11 |
| SVP | 1 | 1 |
| Total | 245 | 245 |
From votes to seats: the Rosatellum gives the details, and what is not modelled: the distribution of seats to circoscrizioni and collegi plurinominali, and the seats abroad, which are held at their 2022 result.
How far Italian polls have missed
Final-poll average minus result, in points, list by list:
| Party | 2001 | 2006 | 2008 | 2013 | 2018 | 2022 | Average |
|---|---|---|---|---|---|---|---|
| PD (Ulivo 2006) | −1.4 | +1.0 | −0.2 | +4.7 | +3.9 | +2.4 | +1.7 |
| FI (PdL 2008–13) | −1.7 | −2.2 | +1.5 | −1.8 | +2.4 | −0.4 | −0.4 |
| Lega | +0.5 | −0.4 | −3.0 | +1.1 | −3.8 | +3.3 | −0.4 |
| FdI | – | – | – | −0.1 | +0.4 | −1.2 | −0.3 |
| M5S | – | – | – | −10.4 | −4.8 | −2.0 | −5.7 |
| AVS (SEL, LeU) | – | – | – | +0.9 | +2.3 | 0.0 | +1.0 |
The final polls are taken at least fifteen days before the vote, because the law forbids publishing polls after that (Measuring Italian polling error, 2008–2022). The M5S was understated by 10.4 points in 2013, 4.8 in 2018 and 2.0 in 2022, and in 2013 and 2018 by every pollster. The PD was overstated by between 2.4 and 4.7 points at the last three elections, each time by every pollster. The other lists changed sign: the Lega was understated in 2008 and 2018 and overstated in 2013 and 2022.
The test
Each election was predicted with the correction learned from the other three: half the mean error, for lists measured at two or more elections, followed by a rescale so the shares keep the same total. The lists scored are FdI, the PD, the M5S, the Lega, FI and AVS.
| Election | Polls' average miss | Corrected, leave one out | Verdict |
|---|---|---|---|
| 2001 | 1.19 | 1.42 | further out |
| 2006 | 1.19 | 0.83 | closer |
| 2008 | 1.52 | 2.03 | further out |
| 2013 | 3.14 | 2.70 | closer |
| 2018 | 2.95 | 2.42 | closer |
| 2022 | 1.55 | 1.52 | closer |
| Average | 1.92 | 1.82 |
The correction was closer at 4 elections and further out at 2, and closer on average: 1.82 points per list against 1.92. That beats the uncorrected polls by more than the 0.01 points the rule asks for, so our Italian projection uses the correction.
Election by election, the test shows where the gain comes from.
- 2013 and 2018: closer. Predicting 2013 from the other three, the correction adds about 1.7 points to the M5S and takes about a point off the PD; both moves go the right way, though the M5S's miss that year was ten points. Predicting 2018, the correction learned from 2013 and 2022 adds more to the M5S and takes more off the PD, again the right way.
- 2008: further out. The PD was polled almost exactly in 2008, and the correction learned from the later elections takes nearly two points off it. The M5S did not stand.
- 2022: further out, narrowly. The correction learned from the earlier elections takes part of the miss off the PD but adds too much to the M5S, whose miss had shrunk to two points, and it moves the Lega the wrong way, because the Lega had been understated before and was overstated in 2022.
So the correction wins on the two elections where the M5S and the PD were missed most, and loses on the two where they were missed least. That is a real pattern in four elections, but it is a thin one. If the next election resembled 2022, with the M5S's miss smaller still, the correction would probably make that forecast slightly worse. The code checks the rule each time the forecast is built: after the next election, if the correction no longer beat the polls on average, it would switch off without any change to the method.
What the correction does today
| List | Elections | Average miss | Typical miss (RMS) | Taken off today |
|---|---|---|---|---|
| FdI | 3 | −0.3 | 0.7 | +0.1 |
| PD (Ulivo 2006) | 6 | +1.7 | 2.8 | −0.9 |
| M5S | 3 | −5.7 | 6.7 | +2.9 |
| Lega | 6 | −0.4 | 2.4 | +0.2 |
| FI (PdL 2008–13) | 6 | −0.4 | 1.8 | +0.2 |
| AVS (SEL, LeU) | 3 | +1.0 | 1.4 | −0.5 |
The M5S gains +2.9 points, half its average miss of −5.7. The PD loses about 1.4, AVS about half a point, and FdI, the Lega and FI move by a few tenths. Lists with no record (Futuro Nazionale, Azione, Italia Viva, +Europa and Noi Moderati) are not corrected and move only through the rescale.
| List | Coalition | Polls | Our bet | Seats | Proportional | Collegi | Clears 3% |
|---|---|---|---|---|---|---|---|
| FdI | Centre-right | 27.4 | 27.0 | 94 | 94 | 0 | over 99% |
| PD | Centre-left | 20.5 | 19.2 | 101 | 101 | 0 | over 99% |
| M5S | Centre-left | 12.9 | 15.5 | 82 | 82 | 0 | 97% |
| FI | Centre-right | 7.1 | 7.1 | 25 | 25 | 0 | 99% |
| FN | alone | 6.5 | 6.3 | 23 | 23 | 0 | over 99% |
| Lega | Centre-right | 5.8 | 5.9 | 20 | 20 | 0 | 89% |
| AVS | Centre-left | 6.1 | 5.4 | 29 | 29 | 0 | 96% |
| Azione | alone | 3.1 | 3.0 | 10 | 10 | 0 | 52% |
| IV | Centre-left | 2.4 | 2.4 | 12 | 12 | 0 | 13% |
| +Europa | Centre-left | 1.3 | 1.3 | 0 | 0 | 0 | 0% |
| NM | Centre-right | 1.1 | 1.1 | 4 | 4 | 0 | 0% |
Under our default line-up, the "campo largo", the M5S and the PD are in the same coalition, so the correction moves seats mostly within the centre-left rather than between coalitions: compare "the polls as they stand" with "our projection" in If we're wrong: replaying past Italian polling misses on today's polls. Under the 2022 line-up, with the M5S alone, the same correction moves seats from the centre-left to the M5S.
Raw figures
In Britain the forecast starts from what respondents say before the pollster's adjustments (Why we start from British pollsters' raw answers rather than the published headline), and in Germany and Spain we compared one pollster's raw series with its published figure. No Italian pollster publishes a raw series of voting intentions: the figures published are shares of those expressing a vote, after weighting. SWG publishes the share of respondents who do not express a vote, which we store, but that is not the same thing. So there is no raw-figure test for Italy.
From the average to the seats
- Average. Each pollster's latest poll within 21 days of the newest counts once.
- Correction. Equation (1) of Forecasting the Camera from imperfect polls: an overview, half the average past miss.
- Collegi. Each list's 2022 vote in each collegio is scaled by its national change (proportional swing), with the rules for Azione and Italia Viva and for Futuro Nazionale described in From votes to seats: the Rosatellum; each coalition's candidate gets its lists' votes, and the most votes wins.
- Proportional seats. The Rosatellum's thresholds and largest remainders, nationally.
- Line-up. One of three, our default the "campo largo" (Who stands with whom: the coalition line-up assumption).
- Simulation. 2,000 runs, each list moved by its typical final-poll miss and each collegio by a local factor (Simulating Italian polling error: where the chances come from).
Limitations
- Four elections, and a correction that passes by less than a tenth of a point per list, driven by one list.
- The line-up is an assumption, and it changes the seats more than the polls do.
- Proportional swing from 2022 ignores candidates; a list's collegi are our split of its coalition's wins.
- Most polls this cycle come through Wikipedia's tables until the official register can be read.
- The electoral law may change before the vote. What we don't know yet: limitations of the Italian forecast collects the rest.
References
DPR 30 marzo 1957, n. 361 (Testo unico per l'elezione della Camera dei deputati), as amended by legge 3 novembre 2017, n. 165.
Jennings, W. and Wlezien, C. (2018). Election polling errors across time and space. Nature Human Behaviour, 2(4), 276–283.
Legge 22 febbraio 2000, n. 28, article 8.
Legge 27 maggio 2019, n. 51, and Legge costituzionale 19 ottobre 2020, n. 1.
How to cite
Polling Forecast (2026). Reading Italian polls: the correction and its test. Retrieved 21 November 2026 from https://www.pollingforecast.com/research/italy-method. Live figures change as new polls arrive; the date above is when the text was last revised.



