After Donald Trump’s surprise win in the 2016 election, the American Association for Public Opinion Research (AAPOR), a COSSA governing member, convened a committee to review the performance of pre-election polling (AAPOR has convened such committees after the past several elections). The committee, chaired by Courtney Kennedy of the Pew Research Center, recently released An Evaluation of 2016 Election Polls in the U.S., outlining its findings and recommendations.
According to the committee, “While the general public reaction [to the election result] was that ‘the polls failed,’ we found the reality to be more complex – a position held by a number of industry experts… Some polls, indeed, had large, problematic errors, but many polls did not. Critically, the reasons for the polling errors are no longer a mystery.” Overall, the committee found the national-level polls were “generally correct and accurate by historical standards,” but that many of the state-level polls were flawed. According to the report, underestimation of Trump’s support in state polls was due to late changing voter preferences, not sufficiently adjusting for overrepresentation of college graduates, and potentially, voters’ misrepresentation of their support for Trump.
The committee cautions against using the 2016 election as reason to discount the importance of polling and survey research: “Well-designed and rigorously executed surveys are still able to produce valuable, accurate information about the attitudes and experiences of the U.S. public.” To improve the outcomes of polling ahead of future presidential elections, the committee recommends finding ways to direct more resources toward critical state polls, which are often under-funded. The complete report is available on AAPOR’s website.