Problem Importance across Time and Space: Updating the ‘Most Important Problem’ Dataset
(with T. Murat Yildirim)
Journal of Elections, Public Opinion and Parties, forthcoming
This article introduces the second release of the Most Important Problem Dataset (MIPD) – a publicly available dataset that codes all available responses to the most important problem question in American public opinion. In the article, we first describe a number of innovations in the new release of the dataset and then demonstrate its usefulness through illustrative cases. In the new MIPD, we switch from providing a single code for the entire response to the various quasi-responses contained in the response, added information on the state in which the interview took place, and subcategories that include – when relevant – information on the preferred levels of policy. These improvements allow for more fine-grained categories and facilitate more accurate inferences at the individual- and aggregate-level. We also extended the time frame through 2020. The new MIPD codes responses from over 850 surveys of more than a million Americans from 1939–2020 into 110 issue categories.