Racial salience, viability, and the wilder effect: Evaluating polling accuracy for black candidates

This study assesses whether polling discrepancies for black candidates can be explained by an interaction of racial salience in an election and the candidate's electoral strength. We hypothesize that voters will have few nonracial justifications for their lack of support for a black candidate when race is a salient issue and the candidate is electorally viable. As a result, polls will experience more problems with socially desirable response bias in such contexts. We use pre-election polls for the near-complete universe of black US Senate and gubernatorial candidates from 1982 to 2010, statewide polls from the 2008 and 2012 presidential elections, and an original measure of the racialization to test our hypothesis. Our results demonstrate that the racialization of the election leads polls to significantly overestimate only support for state-level black candidates and President Obama in contexts where the candidate has the most electoral strength. In the conclusion, we discuss how the results inform pollsters about potential hazards for social desirability response bias.