?
Reassessing Racial Differences in Authoritarianism in the United States: Measurement Invariance and Group Comparisons using the ANES Child-Rearing Values Scale
Measurement invariance is a necessary condition for meaningful comparisons across populations. Previous studies argue that authoritarianism between Blacks and Whites in the United States cannot be compared due to the lack of measurement invariance in the American National Election Studies child-rearing values index. This paper challenges this claim. I analyze ANES data from 1992 to 2016 and identify partial measurement invariance in the 1992–2008 surveys (only one item, respect for elders, displays differential item functioning across groups) and full invariance in the 2012 and 2016 data. It is demonstrated that the child-rearing scale is comparable across Black and White respondents once partial measurement invariance is modeled accordingly, and that group comparisons are methodologically sound. I advanced a hypothesis for the non-invariance of respect for elders and suggest that future work should use measurement models that accommodate differential item functioning and further develop the child-rearing scale.