The statistical analysis of elections is, on the one hand, the analysis of data available from the electoral bodies (= general electoral statistics) and the analysis of data stemming from the official ballot papers with an imprint indicating the sex and age of the voters (= representative electoral statistics) on the other hand.
The fundamental principle of all electoral statistics is to preserve the secrecy of the ballot.
General electoral statistics mainly cover the number of persons entitled to vote, of voters and non-voters, valid and invalid votes, and the number of votes cast for each nomination broken down by Länder, Bundestag constituencies, towns not attached to an administrative district and rural districts, municipalities and polling districts. Representative electoral statistics serve to provide data in representatively selected polling districts on the sex and age structure of persons eligible to vote and of voters, taking into account their votes for individual nominations. It serves the need for information in many areas of our society because it provides information on the pattern of voting, i.e., the voter turnout and voting in different groups of the population.
The compilation of representative electoral statistics was suspended in the 1994 Bundestag Election by the Law enacted on 28 September 1994 (Federal Law Gazette I, p. 2734) and in the 1998 Bundestag Election by the Law enacted on 25 August 1998 (Federal Law Gazette I, p. 2430).
In keeping with the Law on the general and representative electoral statistics for the election to the German Bundestag and the election of the members to the European Parliament from the Federal Republic of Germany (Law on Electoral Statistics) of 21 May 1999 (Federal Law Gazette I, p. 1023) amended by the Law enacted on 27 April 2013 (Federal Law Gazette I, p. 962), representative electoral statistics have again been compiled since the 2002 Bundestag Election. The Law on Electoral Statistics has been applied in European elections in 1999 for the first time.
The rules set forth in the Law on Electoral Statistics stipulate that the previous practice shall be legally binding; they are a more specific legal basis for a statistical analysis of elections than the provisions on statistics set forth in the Federal Elections Act and the Federal Electoral Regulations or the European Elections Act and the European Electoral Regulations, which were repealed by the Law on Electoral Statistics.
The 2002 Bundestag Election and the 2004 European Election are the first elections in which postal voters were also covered by representative electoral statistics to avoid distortions.
General electoral statistics
General electoral statistics document the election results determined and ascertained by the electoral bodies and the general information obtained in the process.
The results of both Bundestag and European elections can be processed and analysed in detail by this statistics. The focus of electoral statistics is on the distribution of valid votes among parties in the country as a whole and in the individual regions (municipalities, districts, constituencies) on the one hand, and on general studies of the persons entitled to vote, the voter turnout at the polls, polling card voters, non-voters and persons casting an invalid vote on the other.
The picture obtained by analysing this kind of data is complemented by comparative figures from previous elections. For this purpose, the absolute numbers as well as the relative numbers (percentages) of each election year are compared. For clarification, average numbers, tendencies etc. are calculated. Changes from one election to another are established as percentages and percentage points (difference between the percentage figures). Candidates and elected candidates are also analysed statistically, this is done in separate analyses by nominations. These statistics provide information, for example, on age and sex and on the occupational group to which the candidates belong.
Representative electoral statistics
The secrecy of the ballot and thus data protection are guaranteed.
The statistical analysis of elections is restricted by the requirement to preserve the secrecy of the ballot. Any violation of the secrecy of the ballot is precluded. The method employed in ascertaining the votes cast by men and women leaves no room for a violation of the secrecy of the ballot. Although the bodies counting the votes may establish, for instance, how many men and women of each of the six age groups formed voted for a certain party, it is impossible to reach any conclusion on how any individuals have cast their votes because there is a large number of people in each age group of men and women. The ballot paper in the representatively selected polling districts merely has an imprint indicating the sex of the voter and assigning him or her to one of the six age groups and thus contains no personal data for that would be in contradiction with the secrecy of the ballot. The survey is anonymous and intended to serve statistical purposes only. The voters’ registers and the marked ballot papers must not be matched. The polling districts selected for the purposes of representative electoral statistics must have at least 400 persons entitled to vote and the postal ballot districts selected must have a minimum of 400 voters.
How are representative polling districts chosen?
From the approximately 90,000 polling districts (including postal ballot districts), roughly 2,500 polling districts for ballot-box voting and approximately 350 postal ballot districts are selected at random for the purposes of representative electoral statistics of Bundestag and European elections. This ensures that the polling districts chosen are representative of the entire electoral area and of the individual Länder. The districts are selected by the Federal Returning Officer in co-operation with the Land returning officers and the statistical offices of the Länder.
Who analyses the data?
The data collected on the basis of representative electoral statistics are analysed by the statistical offices of the Länder and the Federal Statistical Office. The data gathered from the districts which are contained in the random sample are first extrapolated for each Land to obtain the total number of persons entitled to vote and of voters. The extrapolated figures for each Land are then aggregated to establish the result for the federal territory, which is published for the federal territory and the Länder. The data can also be made available to individual municipalities where representative statistics have been compiled so that they can publish aggregate results. In order to protect the secrecy of the ballot, results for individual polling districts contained in the random sample must not be released.
What data are collected?
The turnout at the polls of male and female persons eligible to vote and male and female voters is counted in the polling districts of the random sample according to the following ten age groups using the voters’ registers:
- under 21
- 21 to 24
- 25 to 29
- 30 to 34
- 35 to 39
- 40 to 44
- 45 to 49
- 50 to 59
- 60 to 69
- 70 and over
The votes cast by men and women for the individual parties is analysed for the following six age groups:
- under 25
- 25 to 34
- 35 to 44
- 45 to 59
- 60 to 69
- 70 and over
The votes cast are counted on the basis of official ballot papers with differentiating imprint (man, woman, age group according to year of birth).
Sections 1 to 8 of the Law on Electoral Statistics (WStatG)
Last update: 1 January 2015