Using latent profile analysis to classify US states into typologies of structural racism.

Structural racism is a fundamental cause of racial health inequities; however, it is a complex construct that is difficult to quantitatively analyze due to its multi-dimensionality. We classified fifty US states into three typologies of structural racism using a latent profile analysis. Five domains of structural racism were included in the analysis: Black-White inequities in educational attainment, employment, homeownership, incarceration, and income. In separate analyses, we specified two, four, and five typologies and describe how states' groupings changed when different numbers of typologies are used. In the three typology model, twenty-seven states were classified as typology 1, sixteen states as typology 2, and seven states as typology 3. On average, type 3 states included Midwestern states and had the largest inequities in all domains except education while type 2 states included Western and Eastern states and had the largest inequities in education. States were relatively consistent in which other states they were grouped with, regardless of the number of typologies used in the analysis. Latent profile analysis may be useful in identifying underlying typologies of structural racism at the state-level, which can help researchers identify which states have similar characteristics across multiple, inter-related domains.