Measuring Discrimination By Self-Reported Experience: A Discussion on the French Case

In the scientific literature on the measurement of discrimination, several complementary approaches are developed: the analysis of residual discrimination with multivariate statistical methods; correspondence or audit testings; the measurement of explicit or implicit bias against minorities; the experience self-reported by the respondents (often called perception of discrimination). This latter measurement is often criticized for its subjective dimension, casting doubts on the reliability of the measure itself. The communication aims to address the methodological discussion on the contents of self-reported experience of discrimination and the extent to which it can be used to assess the existence and magnitude of the phenomenon. Based on the Trajectoires et Origines survey data (Ined, INSEE,2008-2009) which has collected data on discrimination with different methods, it will discuss the value and interactions of different indicators of discrimination that have been built in the survey. These indicators are analyzed in conjunction with the repeated experience of othering and stigmatization. The comparison of indicators of discrimination shows that self-reported discrimination under-estimates the extent to which discrimination occurs, as measured with other indicators. There is however a strong correlation between the “situational discrimination” indicator and the more subjective one, suggesting that self-reported experience is backed by concrete facts. Finally, there is a significant and autonomous impact of the Muslim religion on the risk of reporting discrimination that required further analysis to be properly understood. These results argue in favor of discriminations based on an approach combining the different types of measure.