The perception paradox: experiencing and witnessing skin color discrimination and self-rated health in Latin America.

INTRODUCTION: An extensive body of literature has established that perceived discrimination is linked to adverse health outcomes. In Latin America, a region deemed a pigmentocracy due to its pronounced skin color-based stratification, research on the relationship between color discrimination and health remains limited and mixed. This ambiguity may reflect a perception paradox in the context of national ideologies of mestizaje (race-mixture) and racial democracy that purport colorblindness. Latin Americans tend to report high levels of witnessing color discrimination against others (vicarious discrimination) but paradoxically low levels of experiencing discrimination against themselves (personal discrimination). METHODS: Using cross-sectional survey data for Brazil, Mexico, Peru, and Colombia from the Project on Ethnicity and Race in Latin America, I examine the relationship between self-rated health and both personal and vicarious color discrimination. Given that national ideologies have historically privileged class-based explanations of stratification while obscuring the role of skin color, I also examine class discrimination as a comparator. RESULTS: The results show that the observed associations between discrimination and health vary by both the target and the attribute of discrimination. While personal color discrimination does not show a statistically significant association with poor health, vicarious color discrimination is robustly associated with poor health. In contrast, personal class discrimination is consistently associated with poor health, while vicarious class discrimination is not. DISCUSSION: Taken together, these findings suggest that associations between perceived discrimination and health must be understood in light of the Latin American racial order in which mestizaje and racial democracy work to define which explanations of inequality and unfair treatment are deemed legitimate and reportable.