What is the science behind exit surveys


Anja Kruke

To person

Dr. phil., born 1972; Head of the Archive of Social Democracy (AdsD) of the Friedrich-Ebert-Stiftung (FES), Godesberger Allee 149, 53175 Bonn. [email protected]

In the beginning there was a bet. When George Gallup publicly announced in the middle of the US presidential election campaign in 1936 that he would be better able to predict the outcome of the election with a direct (oral) survey of 2000 people than the traditional (written) reader survey of the magazine "Literary Digest", this media-effective challenge was one spectacular prelude to the success story of the political polls and their makers. Because Gallup was to be proved right; the survey institute he founded is still one of the most renowned market and opinion research institutes.

What impact this method of determining public opinion would have on politics in its various dimensions of content, institutions and their power structure as well as on the general relationship between politics, electorate and media was not foreseeable at the time. As the opinion polling emerged, how it settled in politics in Europe and the USA and what consequences this had is roughly outlined below. [1]

"American Science": Beginnings in the USA

In the USA, opinion research was developed in the context of advertising psychology and market research, which had been established in the 1920s. The discursive equation of market with democracy or of consumer with voter, which occurred in the development of the methodology and institutionalization, helped popularize the new discipline. [2] It soon established itself as a public method of introspection. Newspapers set up columns for opinion pollers - for example the "New York Times", which Gallup gave twice a week for his latest findings - and at the same time the media in the USA as a whole established themselves as the main clients of political opinion research.

For opinion research as a scientific method, 1937 can be given as the year of birth: This was the year in which the magazine "Public Opinion Quarterly", which is still powerful today, was published for the first time as an organ of empirical social science, in which industry leaders also have their say. However, experimental attempts to make opinion polls useful for politics had already been made - for example in the context of the 1932 presidential election campaign; US President Franklin D. Roosevelt also used it intensively for his New Deal and in particular for the implementation of his policy towards Germany. [3]

The breakthrough came with the Second World War; Both in the field of science and politics, opinion research gained in importance, as it helped to observe soldiers as well as the entire American society and also those of the war opponents and to find conclusions for one's own actions and arguments. [ 4] Research benefited greatly from the brain drain from Europe: Representatives of the Frankfurt Institute for Social Research such as Max Horkheimer and Theodor W. Adorno as well as Austrian social scientists such as Paul F. Lazarsfeld and Marie Jahoda, who first exemplified surveys as a method of empirical social research in the context of their study "The Unemployed von Marienthal" in 1933 , gave US (and British) development additional momentum. [5]

Again it was Gallup who in 1940, together with Saul Forbes Rae, shortly before the USA entered the war, brought out the book "The Pulse of Democracy" in which he described the method as a democratic science per se presented: With their help citizens could enter into a kind of direct dialogue with politics, the political theory of a real egalitarian democracy could thus be put into practice. Gallup and Rae provided the soundtrack to the founding myth of opinion research as a democratic science. [6] In this way, opinion research, called "American science" in liberated Germany, could be propagated and understood both as a democratic and as a democratizing science.

Europe is following suit

But the USA was not the only place where opinion polling developed as a method for (self) observation of society through representative surveys. Between the 1920s and the 1940s there was a worldwide rise, which, only briefly inhibited by National Socialism and the Second World War, resulted in worldwide expansion and networking after 1945. [7] Gallup in particular began immediately after its forecast success in 1936 to expand its company internationally with institutes in Great Britain, Canada, Australia and Sweden.

Opinion research, which is always visible in the media, also exerted great attraction on European scientists who - like the French sociologist Jean Stoetzel or the German student of newspaper studies Elisabeth Noelle - came into contact with the new method during their stay in the USA and set about using it to make them usable in their home countries. Noelle, later Noelle-Neumann, founded the Allensbach Institute for Demoscopy in Germany in 1947; Stoetzel founded France's first survey institute as early as 1938, but the German invasion interrupted development.

And even after the Second World War, it would be until the end of the 1960s for polls to become an everyday item in politics and society in France. Politicians and journalists did not use polls as an explanation of public opinion. The electoral system did not provide for a direct election of the president, as in the USA, and the elites saw no reason to orient themselves to public opinion (understood as a numerical value), so that public polls eke out a shadowy existence. Instead, politics resorted to a system of reporting from the regions that had been established in Napoleonic times and that was considered to be more reliable than interviewing a small number of people. With the establishment of the Fifth French Republic in 1958, the political system and, with its crisis, the political culture changed fundamentally - and with it the idea of ​​"public opinion". This led to the fact that surveys were finally established. [8]

The survey institutes in Great Britain took a similarly long time. Although institutes were founded before 1939 that were used to survey the population during World War II, it was not until the beginning of the 1960s that survey research caught on on a broad front in political parties and the media; before that, political actors in particular had had no interest in it due to different "worldviews". [9]