The study of individual-level public opinion and behavior has flourished in recent decades. But understanding of the dynamics of mass opinion prior to the 1950s has been undermined by the absence of high-quality, individual-level data. Over 450 national opinion polls were conducted from 1935-1945. However, these surveys have not been exploited by the social science community because they are not easily usable. The data often contains numerous miscodings and other errors. In addition, the surveys employed now-discredited quota sampling procedures. The principal investigators recode the datasets and implement and disseminate methods appropriate for analyzing this historical public opinion data. By applying methods that make quota sample data usable, this project opens up a new realm for public opinion research in the U.S. and other countries, such as England, that have used quota-sampling methods. Recovering this early opinion data promises to illuminate critical questions concerning the role of the mass public in the political system. The public opinion polls from this time contain a plethora of valuable items measuring attitudes concerning economic policy, World War II, and racial relations. The principal investigators use the polling data on attitudes toward international intervention to determine how the mass public guided and reacted to foreign policy decisions during the Second World War, thereby expanding the systematic study of public opinion and war to include the most significant international conflict in American history. The principal investigators also use the opinion data on domestic policy as a resource for examining the relationship between the mass public and both the programmatic accomplishments and limitations of the New Deal. Broader Impacts Resulting from the Proposed Activity This project makes available to the social science research community a trove of public opinion data that has largely been ignored in the past. The principal investigators compile and produce readily usable computer files for roughly 450 opinion polls undertaken in the United States from 1935 to 1945. These files are available through the Roper Center in Storrs, CT. However these data are almost impossible to use in their existing condition. The principal investigators recode the individual files, prepare documentation, and make available a series of weights to mitigate the biases resulting from quota-control sampling. The investigators also compile the individual polls into a series of cumulative datasets. These efforts make the data more easily accessible to the larger social science research community. The project also promises to expand the field of political behavior by promoting the study of historical public opinion. It provides to the community of scholars a wealth of individual-level opinion data in the pre-1950 era, allowing researchers to gain new insights into an array of substantively important topics, such as changes in Americans' racial attitudes and public support for the early American welfare state.
