A discrepancy between voter opinion polls and election outcomes, particularly in races where a white candidate opposes a non-white candidate, is observed periodically. This phenomenon manifests as voters expressing support for the non-white candidate in polls, yet ultimately casting their ballot for the white candidate. This deviation can stem from various factors, including social desirability bias, where respondents provide answers they perceive as more socially acceptable, or an underestimation of latent prejudice within the electorate. A notable example is the 1982 California gubernatorial election involving Tom Bradley, an African American candidate, where pre-election polls predicted his victory, but he ultimately lost the election.
Understanding this effect is crucial in political science, particularly in the context of American government (AP Gov), as it highlights the complexities of polling accuracy and the potential influence of race on voting behavior. Recognizing that stated preferences may not always align with actual behavior allows for a more nuanced interpretation of election data and a critical assessment of the factors driving voter decisions. Awareness of this potential bias is beneficial for campaigns, pollsters, and political analysts alike, as it encourages more sophisticated polling methodologies and a more cautious approach to interpreting pre-election surveys.
