The relationship between political ideology and judgements of bias in distributional outcomes

We examine judgements of bias in distributional outcomes. Such judgements are often based on imbalance in distributional outcomes, namely, the under- or over-representation of a target group relative to some baseline. Using data from 26 studies (N = 14,925), we test how these judgements of bias vary with the target group's characteristics (traditionally dominant or non-dominant) and the observer's political ideology (liberal or conservative). We find that conservatives set a higher threshold for recognizing bias against traditionally non-dominant targets (women, Black people, immigrants), as compared with liberals. Conversely, liberals set a higher threshold for recognizing bias against traditionally dominant targets (men, white people, native-born citizens), as compared with conservatives. However, these relationships between political ideology and judgements of bias diminish when the targets are unknown or ideologically irrelevant. These findings emphasize the context-dependency of bias judgements and underscore the importance of stimulus sampling and appropriate selection of controls.