This phenomenon, relevant to the study of American government and politics, describes a discrepancy between voter opinion polls and election outcomes, specifically where a candidate belongs to a minority group. It suggests that some voters may express support for the minority candidate in polls to avoid appearing prejudiced, but ultimately vote for a different candidate in the actual election. This results in an underestimation of support for the non-minority candidate in pre-election surveys. A hypothetical scenario involves a polling average suggesting a close race between a Black candidate and a White candidate, yet the White candidate wins by a larger margin than predicted.
Understanding this effect is important for accurately interpreting polling data and analyzing election results within the context of American political behavior. It highlights the potential influence of social desirability bias on voter responses and the complexities of race and prejudice in electoral politics. Historically, its presence has been debated, with some scholars attributing unexpected election outcomes to it, while others point to alternative explanations like late-deciding voters or polling errors. Regardless, its existence serves as a reminder that survey data should be viewed with critical awareness of possible biases.
