Election research

from Wikipedia, the free encyclopedia

The election research (rarely Psephologie ; from ancient Greek Psephos ψῆφος, 'pebble' - these were used in ancient Greece as ballot) deals with the scientific study of political elections apart. The electoral sociology - also known as empirical election research - is also part of the research field as the electoral system - and electoral research .

Academic election researchers and researchers from commercial institutes contribute to election research. There are close ties between the two. However, both pursue quite different goals, which is reflected in different focus areas. Commercial electoral researchers attach great importance to the collection of suitable data and the description of voting behavior. For example, commercial institutes regularly survey the current political mood of the population, including with the Sunday question , and the mood on certain topics or people. Be with the election days exit polls (exit polls) data for projections collected and forecasts. In comparison, academic electoral researchers place more value on not only describing voting behavior, but also explaining it. Academic election researchers are organized in Germany in the German Society for Election Research (DGfW).

The electoral system research

Electoral system research essentially deals with three bundles of questions: the description and classification of electoral systems, the emergence and development of electoral systems and their components, and the effects of electoral systems. For description and classification, the size of the constituency, the form of the candidacy, the voting process and the voting process are often used as important differentiating criteria. As a result of the further development of existing electoral systems beyond the limits of the classic types of majority and proportional representation and various attempts to typify these subtypes, this field of research has become more confusing. At the same time, reforms to electoral systems in existing democracies such as New Zealand and Japan, as well as the development of the new democracies in Central and Eastern Europe, led to increased research in the field. Classical systems are rarely observed, but mixed types, so-called “mixed member systems”. Individual case analyzes and quantitative-statistical studies have contributed to research into the origin and development of electoral systems. Among other things, they show that party constellations and cultural factors play an important role. Investigations into the effects of electoral systems consider, among other things, their influences on the relationship between voting and mandate shares, the behavior of eligible voters, the behavior of candidates and office holders, e.g. in the formation of a government, the development of party systems and the development and stability of political systems .

The empirical election research

Explanatory approaches and theories

Various approaches and theories are used to explain voting behavior. Sociologically, psychologically and economically oriented approaches play an important role.

The microsociological approach following Paul F. Lazarsfeld and his colleagues ( The People's Choice ) explains voting behavior using the party-political norms with which a person comes into contact in their (primary) environment. People who communicate regularly with people who prefer a particular party should also vote for that party after this introduction. The voting behavior of people who are exposed to different party-political norms cannot be predicted with this approach. They are under "cross-pressures" and tend - so the assumption - to abstain from voting and to alternate.

The macro-sociological approach following Seymour Martin Lipset and Stein Rokkan traces voting behavior back to the positioning of people in social lines of conflict ( cleavage theory ). The socio-economic and religious-confessional lines of conflict are particularly influential in the Federal Republic. For example, workers should vote for left-wing, social-democratic parties, and Christian-religious citizens should vote for Christian-democratic parties. The voting behavior of people who cannot be clearly classified in these social conflicts or who follow contradicting recommendations from their social situation cannot be clearly predicted.

The social psychological approach goes back to a research group led by Angus Campbell at the University of Michigan in Ann Arbor ( The American Voter ). In its simplest form, the approach leads voting behavior back to candidate orientations, dispute orientations (issue orientations) and party identification. The party identification is viewed as stable in the long term and therefore serves as a stabilizing factor that can primarily explain constant voting behavior. Candidate and issue orientations, on the other hand, are variable in the short term and can therefore explain changes in voting behavior well.

The economic approach goes back to the work An Economic Theory of Democracy by Anthony Downs , who applied microeconomic concepts to the analysis of political processes. Downs bases his analysis on a number of cognitive and motivational assumptions and assumes, among other things, that citizens seek to maximize their benefits by making a choice. In a simple two-party system, they therefore choose the party whose election victory and government policy will give them the greatest benefit. Downs continues to develop this idea by applying it to multi-party systems and introducing problems of uncertainty. An important finding in his analysis is the so-called choice paradox. It consists in the fact that significant parts of the electorate regularly participate in elections, although from the microeconomic point of view of each individual, turnout is irrational.


The following methods are used for election research:

Quantitative, standardized surveys are most frequently used because they allow statements to be made about the entire electorate. A representative selection (sample) of the entire electorate is surveyed. The data are evaluated using statistical methods.

  • Qualitative methods are suitable for testing lines of reasoning and communication strategies before making an election. For example, the acceptance of a poster line or a TV spot can be checked in focus groups. Focus groups enable the communication strategy to be tailored to target groups.
  • Network analyzes illuminate the communication networks of certain target groups. They can be used to control communication during elections. Or they examine the surfing behavior of voters within the usual Internet portals .
  • Projections are made on election evenings and are often commissioned by TV stations. You calculate the probable final result on the basis of voting circles that have already been counted.
  • Aggregate data analyzes use actual election results against which statistical relationships are analyzed. An example of the application of aggregate data analysis is voter flow analysis .

Commercial election research

Some commercial polling institutes conduct election research in addition to market research. The clients for this are often the media , political parties , public bodies or interest groups . The first commercial survey institutes in Germany were the Institut für Demoskopie Allensbach , founded in 1947, and the German Institute for Popular Surveys , which emerged from a department of the American military administration established in 1945 and which no longer exists. Those who commission the published polls mostly come from the media. In addition to Allensbach, the research group Wahlen (on behalf of ZDF ), Forsa (for RTL ), Infratest dimap (for ARD ) and TNS-Emnid (for ProSiebenSat.1 Media ) are active.

In Austria, the Institute for Empirical Social Research (IFES) and the SORA Institute for Social Research and Consulting deal with election research on behalf of the ORF . Swiss election research institutes are gfs.bern (for SRG Medien) and Isopublic . YouGov operates internationally .

On the problem of influencing the political system through election polls

Voting decision and participation

In political science, the influence of polls on voting decisions, i.e. the decision which party is elected, as well as on voter turnout is discussed. Since the 1950s, various scholars have formulated theories about how polls affect voting decisions. Some of these theories are contradicting themselves and have rarely been proven empirically.

Theories about voter turnout are the mobilization hypothesis (a head-to-head race in polls leads to a higher voter turnout) or the convenience hypothesis (if the election result is seen as clear, the voter turnout decreases). Suspected influences on the voting decision are, for example, the follower effect (connection to the majority opinion) and the underdog effect (connection to the minority opinion). In addition, there are theories on tactical voting such as the “guillotine effect ”, “ lending votes ” voting and the prevention of absolute majorities.

Because of their presumed influence on the outcome of the elections, the publication of survey results one to two weeks before the election is prohibited in various European countries ( France , Italy , Portugal , Spain , Hungary ). There is no such ban in Germany as it would not be compatible with the freedom of information and press guaranteed in the Basic Law . However, ARD and ZDF forego such surveys in a voluntary self-restraint a week before an election. In the run-up to the 2013 federal election , the ZDF published current polls in the Politbarometer .

The publication of the results of voter surveys after voting on the content of the voting decision is not permitted in Germany before the polling stations are closed (Section 32 (2 ) of the Federal Election Act ) and can be punished as an administrative offense (Section 49a of the Federal Election Act). There is also no ban on surveys in Austria , but neither polls nor projections may be published before the polling stations close . There are also no legal regulations in Switzerland; The Association of Polling Institutes, VSMS, however, urges its members not to publish any new polls in the 10 days before the election, which is also recommended by the Association of Newspaper Publishers.

Political decisions

In political science and the media, the influence of election and political research on political decisions is discussed. It is assumed that the decision-makers orientate themselves less on factual arguments than on the majority opinion presented by the survey institutes when making individual decisions. The decision-makers should try to increase their popularity and chances of being elected. There is no empirical evidence for this thesis.

Inconsistent quality requirements

Commercial election research works mainly on behalf of the media and political parties. These clients usually have lower methodological skills and quality demands than academic elective researchers. Their assignments also usually have to be completed in shorter deadlines. Academic election research assignments to commercial institutes are rare and usually less lucrative, as they require more intensive preparation, higher methodological requirements and higher quality requirements. The demands of science here are higher than those of the day-to-day business of the institutes. The information common in academic research about the survey period, number of respondents, question wording, etc. is seldom included in media publications.

False prognoses

According to electoral researchers, there are clearly false prognoses when a large proportion of the voters are undecided until shortly before the election. A high proportion of alternate voters can make forecasts even more difficult. Further sources of error: strategic considerations by voters at the last minute, methodological errors and deliberately incorrect statements by respondents (e.g. when voting for socially undesirable parties).

The political scientist Hans Rattinger emphasized that changes in parties by two percentage points are "statistically not relevant" due to the expected sampling errors . In addition, the experience of the 2005 Bundestag election showed that the undecided in particular were often only to the detriment of one party in polls. When analyzing survey results, it should also be taken into account that voter participation in elections is actually much lower, "but in surveys 90 to 95 percent tell you how to vote" without actually voting, this phenomenon cannot be statistically factored out. Tests showed that when asked about the popularity of politicians, sympathies are often weighted without respondents even knowing it. "We have done experiments with surveys in which we have built in the names of politicians who do not even exist - nevertheless, 30 to 40 percent of those questioned often find them likeable, for example".

Philosophical criticism of empirical election research

In the run-up to the 2005 Bundestag elections , the philosopher Peter Sloterdijk described the opinion polls as an “extra-parliamentary authority” that operates an “unjustified opinion dictatorship” because of its position in the media and the scope of the reporting. He called for a legal restriction on opinion polls on political issues.


  • Kai Arzheimer, Jürgen W. Falter : Elections . In: Eckhard Jesse, Roland Sturm (Ed.): Democracies of the 21st Century in Comparison . Opladen: Leske and Budrich 2003, pp. 289-312, ISBN 3-810-03732-X .
  • Jürgen W. Falter, Harald Schoen (Hrsg.): Handbuch Wahlforschung . Wiesbaden: VS Verlag für Sozialwissenschaften 2005, ISBN 3-531-13220-2 .
  • Nikolaus Jackob (ed.): Election campaigns in Germany. Case Studies on Campaign Communication 1912–2005 . Wiesbaden: VS Verlag 2007, ISBN 3-531-15161-4 .
  • Dieter Nohlen: Suffrage and the party system . 4th edition, Opladen: Leske and Budrich 2004, ISBN 3-8100-1465-6 .

Web links

Individual evidence

  1. a b Andreas M. Wüst, Philip Stöver 2006: Elections in Europe . In: PVS 47/2006, pp. 289-309.
  2. ^ Henry Kreikenbom, Maxi Stapelfeld 2008: Political Research, Increasing Demand in Times of Social Change . In: Eva Balzer (ed.): Qualitative market research in theory and practice , ISBN 978-3-8349-0244-3 , pp. 469-481.
  3. Harald Schoen 2010: Election Research. In: Irene Gerlach, Eckhard Jesse, Marianne Kneuer, Nikolaus Werz (eds.): Political science in Germany. Publications of the German Society for Political Science (DGfP) , Nomos, Baden-Baden, pp. 223–238.
  4. Andreas Reichelt: Alternating voters in focus. In: Tele Regional Passau 1 (TRP1). Accessed March 17, 2019 (German).
  5. Harald Schoen (2005): Wahlsystemforschung . In: Jürgen W. Falter, Harald Schoen (ed.): Handbuch Wahlforschung , VS Verlag für Sozialwissenschaften, Wiesbaden, pp. 573–607.
  6. Harald Schoen (2005): Wahlsystemforschung . In: Jürgen W. Falter, Harald Schoen (ed.): Handbuch Wahlforschung , VS Verlag für Sozialwissenschaften, Wiesbaden, pp. 573–607.
  7. Harald Schoen (2005): Wahlsystemforschung . In: Jürgen W. Falter, Harald Schoen (ed.): Handbuch Wahlforschung , VS Verlag für Sozialwissenschaften, Wiesbaden, pp. 573–607.
  8. ^ Paul F. Lazarsfeld, Bernard R. Berelson, Hazel Gaudet 1944: The People's Choice. How the Voter Makes up his Mind in a Presidential Campaign, New York: Duell, Sloan and Pearce; Bernard R. Berelson, Paul F. Lazarsfeld, William N. McPhee 1954: Voting. A Study of Opinion Formation in a Presidential Campaign, Chicago, University of Chicago Press.
  9. Further Harald Schoen 2005: Sociological Approaches in Empirical Election Research. In: Jürgen W. Falter, Harald Schoen (eds.): Handbuch Wahlforschung , VS Verlag für Sozialwissenschaften, Wiesbaden, pp. 135–185.
  10. ^ Seymour M. Lipset, Stein Rokkan 1967: Cleavage Structures, Party Systems, and Voter Alignments. An Introduction. In: Seymour M. Lipset, Stein Rokkan (ed.): Party Systems and Voter Alignments: Cross-National Perspectives , Collier, Macmillan, New York / London, pp. 1-64.
  11. See in detail Harald Schoen 2005: Sociological approaches in empirical election research. In: Jürgen W. Falter, Harald Schoen (eds.): Handbuch Wahlforschung , VS Verlag für Sozialwissenschaften, Wiesbaden, pp. 135–185.
  12. ^ Angus Campbell, Gerald Gurin, Warren E. Miller 1954: The Voter Decides, Evanston, IL: Row, Peterson and Company; Angus Campbell, Philip E. Converse, Warren E. Miller, Donald E. Stokes 1960: The American Voter, New York: Wiley.
  13. See Harald Schoen, Cornelia Weins 2005: The social psychological approach to the explanation of voting behavior . In: Jürgen W. Falter, Harald Schoen (ed.): Handbuch Wahlforschung , VS Verlag für Sozialwissenschaften, Wiesbaden, pp. 187–242.
  14. ^ Anthony Downs 1957: An Economic Theory of Democracy, New York, Harper.
  15. See Kai Arzheimer, Annette Schmitt 2005: The economic approach. In: Jürgen W. Falter, Harald Schoen (ed.): Handbuch Wahlforschung , Wiesbaden: VS Verlag für Sozialwissenschaften, pp. 243-303.
  16. Newspaper '20 Minuten ' : Wikipedia predicts SVP election failure from June 11, 2015
  17. Jochen Groß: The prognosis of election results. Approaches and empirical performance . VS Verlag für Sozialwissenschaften, 2010, ISBN 978-3-531-17273-6 , p. 39 ff .
  18. Jürgen Maier, Frank Brettschneider 2009: Effects of survey reporting on voter behavior . In: Nikolaus Jackob u. a .: Social research on the Internet . ISBN 978-3-531-16071-9 , pp. 321-337.
  19. Fight for the hit rate. Tagesspiegel, accessed on September 13, 2009 .
  20. Gregor Daschmann: ZDF breaks unwritten law - does the broadcaster influence the election? Retrieved September 20, 2013 .
  21. ^ Oscar W. Gabriel, Silke I. Keil: Empirical election research in Germany: Criticism and development perspectives. In: Jürgen W. Falter, Harald Schoen (Hrsg.): Handbuch Wahlforschung . Wiesbaden: VS Verlag für Sozialwissenschaften 2005, pp. 613–641.
  22. "Prohibition of prognoses would be no harm" , ( Memento from September 4, 2009 in the Internet Archive ) Netzeitung from September 2, 2009.
  23. Democracy thanks to demoscopy: majority as opinion . FAZ , September 26, 2005, No. 224, p. 37.