To main content

Facts about the election study

The Swedish election studies are carried out by Statistics Sweden in cooperation with the Department of Political Science at Göteborg University.

The studies are based on surveys of the Swedish electorate, and they have been conducted around each election to the Riksdag (The Swedish parliament) since 1956. Studies have also been carried out at the time of the referenda in 1957, 1980, 1994 and 2003, and also at the time of each election to the European Parliament since 1995. All studies are based on large samples and they have comparatively low levels of non-response. Hence, the Swedish election studies are, together with the slightly older American election studies, the most comprehensive material for systematic studies of voter behaviour. 

Why are elections of interest?

The main purpose of the election studies is to increase knowledge of status of Swedish democracy. The results bring facts into the debate over participation and democracy in Sweden. The studies are funded by the Riksdag and they are independent of the political parties. The results of the studies are not only used in academic circles, but also by mass media, political parties, organisations, authorities and not least the politically interested public.

How are the election studies carried out?

Data collection is carried out through face-to-face interviews. Since the election to the Riksdag in 1973, the survey consists of two-wave panels, i.e. half of the sample in a study were also part of the previous study, and the other half will be a part of the next study. Furthermore, data collection is conducted both before and after each election. Also, immediately after the election, a short questionnaire is sent to those who are interviewed before the election. Hence, the studies have a small panel that measures voter movements and campaign effects during the weeks before election. During later years the sample size has been between 3 000 and 4 000 persons.

Elections studies carried out at the time of elections to the European Parliament differ slightly in that they do not have a panel approach, i.e. the respondents are only interviewed at the time of one election. Also, the surveys are strictly post-election surveys. The sample size has been approximately 2 700 persons.

The design of the election studies carried out at the time of referenda has varied from time to time. The study carried out at the time of the referenda about the euro in 2003 did not have a panel approach. However, the study was conducted both before and after the day of the referendum, and those who were interviewed before the referendum received a short questionnaire immediately after the election. The sample size in that study was approximately 3 000 persons.

Are the results comparable over time?

The results from the election studies are, to some extent, comparable over time. The wording of some of the questions has changed over the years, but the fundamental questions have remained the same over a very long time. In an effort to make international comparison more comprehensive, part of the election studies have, in recent years, been adapted to the international collaboration Comparative Studies of Electoral Systems.

How are the results published?

The results from the national election studies are published (in Swedish) in General Elections XXXX, Part 4, Special Studies (included in Official Statistics of Sweden). Up to and including the 1994 election study, the results were published in General Elections XXXX, Part 3, Special Studies.

Results from election studies conducted at the time of elections to the European Parliament are published (in Swedish) in Elections to the European Parliament XXXX. Results from studies carried out at the time of referenda are published, by Statistics Sweden, in separate volumes for each study. 

The researchers at the Department of Political Science, Göteborg University, also publish extensive analysis based on the election studies. The data from the studies is archived at the Swedish Social Science Data Archive and is used for analysis by researchers from all parts of the world.

Links: