The Evolving Nature of Generalized Prejudice Toward Marginalized Groups in the United States 2004-2020

Prejudices intercorrelate positively and can be modeled as a generalized prejudice (GP) factor that is considered robust and central to postulating that some people are relatively more prejudiced than others (i.e., prejudice is not purely contextual). Although past research documents changes in specific prejudices over time, the field tacitly assumes GP stability/robustness, an untested notion. Using nationally representative American National Election Survey 2004-2020 data (N = 21,998) assessing attitudes toward Black people, illegal immigrants, gay people, and feminists, we discovered that prejudices have become increasingly correlated over time. Initially invariant, from 2012 onward GP became variant and required correlated residuals between prejudices (outside of GP). GP vastly increased its association with political conservatism (approximate to.41 in 2004-2008, approximate to.70 by 2016-2020) but less so with age, sex, and education. Indeed, best fit in 2020 involved a "GP 2.0" factor indicated by specific prejudices and conservatism. Implications regarding the nature of prejudice are discussed.