Facts about the participation survey
Similar participation surveys have also been carried out for the 1980, 1994 and 2003 referenda and for elections to the European Parliament since 1995.The information about electoral participation is collected via the county administrative boards.
Why are elections of interest?
As its main purpose, the participation survey aims to increase the knowledge about the state of Swedish democracy. Electoral participation is a key measure for representative democracy. The results of the survey constitute an essential source for the debate on issues of citizen influence and democracy in Sweden. The results are also useful for scientific research, the mass media, organisations, central governments and the politically-interested general public.
How are the participation survey carried out?
Data are principally collected via the county administrative boards. Statistics Sweden sends lists of persons included in the sample to the boards, which then use the checked electoral registers to note whether or not those in the sample have participated in the election.
The participation survey began as a census and hence included all those eligible to vote. From the 1940s onwards, the surveys have been based on samples. Since the 1964 election, the participation survey uses the same sample as is used in the Labour Force Survey, to give access to data such as employment status, union membership etc. In addition, data from Statistics Sweden’s Total Population Register and Register of Education are used. Since 1976, when foreign citizens were entitled to vote in the elections to the county councils and municipal councils for the first time, Statistics Sweden has also carried out participation surveys in this group.
Since the information on electoral participation is taken from the electoral registers, there is no uncertainty about whether the sample individuals have participated or not. However, there is some uncertainty in the estimations due to the fact that survey is based on a sample of those eligible to vote.
Are the results comparable over time?
The participation surveys of different elections since the 1960s are, to a large extent, comparable. From the 1976 election onwards, every estimation of the participation rate is accompanied by a margin of error. Until the 1985 election, there were an upper age limit of 74 years (i.e. the participation among persons 75 years or older was not included in the estimations). By including an independent additional sample of persons aged 75 years or over from the 1988 election onwards, the participation survey no longer has an upper age limit.
How are the results published?
The results from the participation surveys are published in Statistics Sweden’s Statistical Reports (Series ME). Results from elections before 1994 were published in the series BE.