Gallupundersökningen som ett nytt medium mellan press, politik, reklam och vetenskap, 1941-1955
Av Vesterlund, Eskil
This thesis examines the introduction and impact of opinion polling in Sweden during the mid-twentieth century, focusing on the activities of the Swedish Gallup Institute (Svenska Gallupinstitutet). Established in 1941 as the country’s first polling organization, the institute introduced sample surveys as a new means of representing public opinion and had a near-monopoly on polling until its decline and absorption by AC Nielsen in the mid-1950s. Adopting a media-historical perspective further informed by the history of knowledge, and science and technology studies, the study conceptualizes the Gallup poll as a medium and media boundary object that mediated society across multiple societal domains.Drawing on a wide range of sources, including newspapers, archival materials, and market research reports, the thesis treats polls not as transparent indicators of public opinion but as media historical artefacts shaped by technological, institutional, and political contexts. Through analyses of press coverage, political debates, market research practices, and academic reception, it explores how Gallup polls were produced, circulated, and contested. Attention is paid to both the infrastructure of polling – questionnaires, punched cards, and interviewers – and its various forms of dissemination through newspapers, visualizations, and reports. The study demonstrates how polls functioned as statistical attractions in the press, instruments of political argument, tools for market research, and objects of methodological debate within the social sciences. The analysis of the Gallup poll as a media boundary object thus offers a perspective on the scientization of the social as a non-linear process of multidirectional traffic across the changing boundaries of social science.The thesis shows how the sample surveys of the Swedish Gallup Institute introduced new ways of quantifying and visualizing society, while also provoking negotiations over their legitimacy and meaning. Ultimately, it is argued that the Gallup polls of the 1940s and 1950s did not simply reflect public opinion, but actively reconfigured the epistemological conditions for how society could be known, discussed, and governed in the postwar Swedish welfare state