Whenever in Italy we are called to votewe start talking about exit poll and of Forecast on the final vote immediately after the closure of the polls. This will also be the case in the case of the referendum of 8 and 9 June, which for 13 Italian municipalities also coincides with the administrative elections. But what exactly these forecasts are and, above all, how are they made?
These are surveys conducted with the aim of anticipating thefinal outcome of an election or a referendum, analyzing the trend of the vote also based on different social groups (men and women, young and old). In Italy, however, the law establishes the prohibition to publish These forecasts Before from the closing of the pollsto prevent estimates from affecting those who have not yet expressed their vote.
The most common and used type of survey is the exit poll, which is made directly out of the polling station On a representative sample of the voting population: people are provided with a facsimile of the electoral card, where they will have to report, totally anonymouslythe vote they expressed previously.
At the level of reliability, however, if well structured the exit polls tend to have the same degree of uncertainty of a normal survey: the margin of error, therefore, is about the 2%. The voters part of the sample, however, can decide to lie on the expressed vote, a factor that can lead to generating incorrect estimates.
The types of polls: Exit Poll, Instant Poll and intention poll
There are therefore different types of surveys that can be made to try to anticipate the outcome of an election or a referendum: specifically, we talk about exit poll, Instant Poll And intention poll.
- The exit poll They are a type of surveys made on a sample of voters as soon as they come out of polling station (from English Exitexit, e pollseat or survey).
- The Instant Poll are gods telephone surveys made on the day of the vote.
- The intention poll they are surveys made by phone or live and some conducted days before the vote.
Often following an election we also hear about voting forecasts which, however, are different from the surveys: it is always about estimates on the final result of the election or referendum, but in this case the analyzes are built on the basis of real data that come directly from the seats after it started it bare of the votes.
Italian law, however, establishes the prohibition to reveal the outcome of surveys or voting forecasts before the closing of the polls, Just to prevent estimates on the trend of the vote to influence those who have not yet expressed their own.
In general, the type of survey more widespread And used is that of the exit polls: before the elections, some seats are identified, outside which a sample of voters representative of the total of those with vote rights.
If the voter accepts to participate At the survey, his vote remains anyway anonymous: The sample, in fact, is provided documentation to be completed, including a facsimile of the electoral card, where replicate your vote, and a series of questionnaires relating to personal data, including the sex And age. It is precisely this information that allows experts to carry out analysis on the vote of several categories of citizens, from men to women, from young people to the most mature bands of society.
The task of creating these polls, on the other hand, is entrusted to demoscopic instituteswhich deal with conducting statistical investigations on public opinion. After identifying the seats covered byinvestigationthe institutions send some of their own technicians To interview the sample: outside the seats they come then voters selected To be interviewed, with the aim of building a sample as representative as possible of the voting population.
Among other things, on 8 and 9 June 2025, in addition to the referendum, the ballot is held for the administrative elections in 13 Municipalities Italians (including Matera and Taranto): also in this case, the exit polls will be published immediately after the closure of the polls, with the forecasts of the candidates favored to victory.
The reliability of the polls: if well structured, the 2% margin of error
But how much they are reliable These surveys published before the total bare of the votes? It must be said that, despite being conducted by prestige institutesthe voting forecasts are not always revealed to be completely correct. Several factors can exist negatively The analysis: it is possible that the voter chooses to lie on the vote he expressed, or that he simply decides not to provide any information on who or what he voted. After all, in the exit polls they are the voters who do car-candidan As samples, and they can therefore refuse to participate in the survey. Especially in the past, in fact, the voters belonging to some parties proved to be more reticent In declaring their voting preferences, thus going to affect, albeit indirectly, on the results of the surveys.
In short, the usefulness of polls and electoral projections is still at the center of a Large debateespecially in relation to their precision: however, it must be considered that electoral projections and exit polls are one of the Few cases in which a sample analysis obtains an almost immediate response, with the final results which are revealed after a short time. In other types of surveys, such as those used to measure the listening to a program, the public generally does not worry about checking the actual correctness of estimates. In general, however, if exit polls tend to have the same if well structured degree of uncertainty of a normal survey: the margin of error, therefore, is about the 2%.
These forecasts are then particularly useful in the case of big electionssuch as those political for the renewal of the members of the Parliament (who in Italy are held every 5 years), which take several hours for the full bare votes and, therefore, to announce the definitive results.
In the case of the last political elections of the 2022for example, exit polls have shown one good correspondence to the final results, with the estimates that have undergone only some oscillations contained and therefore managed to provide a forecast quite accurate of the vote.
Clearly, exit poll and voting forecasts do not exist only in Italybut they are measurements that are made all over the world to try to anticipate the elections of an election or study the trend of the vote. In recent years, two surveys on the vote have particularly discussed internationally, having anticipated a result which then turned out to be completely opposite: this is the case of Brexit referendum of 2016 (when the surveys indicated the victory of the “remain” to remain in the European Union, while at the end of the bare of the votes the victory of the “Leave” was declared) and the US presidential elections of 2016, which they gave for winner Hillary Clintonwhich instead came out of it from Donald Trump.
