On the basis of a government decision, Statistics Sweden conducts a party preference survey to shed light on the current situation for political opinions in the country. The survey has been conducted every year in May and November since 1972. The public's attitudes towards the EU have been included since 1992, while attitudes towards the EMU/euro have been a part of this survey since 1997.
A sample of about 9 000 persons are interviewed by telephone for each survey round. Statistically scientific methods and the size of the sample guarantee high quality in the reported estimations.
For each survey round party preferences are presented according to sex, age, civil status, number of children, income, region, socio-economic background, employment status, sector, trade union membership, type of housing, degree of density, level of education and foreign/Swedish background. In addition, estimations are made of election results "if an election were held today" and net flows between surveys.
Changes May 2011
As of May 2011, the results are weighted for party preference for non-response. The weighting is done with consideration to sex and age, education, country of birth and region. However, a comparison with unweighted estimations shows that the weighting has little effect on the majority of the estimations. The changes are nearly exclusively within the margins of error. New tables with weighted results from previous surveys have been created in Statistics Sweden's Statistical Database to compare backwards in time. Previously published unweighted results will also be available for some time ahead on Statistics Sweden's website. On the other hand, the estimations of election results for "if an election had been held today" have not undergone any changes in methods.
As of May 2011 the results of the Party Preference Survey will no longer be published in a Statistical Report, but in a table publication as well as in Statistics Sweden's Statistical Database.
EU and euro preferences in November 2012
If there were to be a referendum on the euro in November 2012, 82 percent would vote no to the euro while 10 percent would vote yes. 8 percent say they do not know how they would vote. The percentage of yes votes has decreased, while the percentage of no votes has increased compared to May 2012. Both changes are statistically significant.
Read the entire press release