Linear and logistic fixed effects models are estimated for the responses as to whether the situation in the picture meets the state definition and whether the offense is reportable. Independent variables include: the respondent's age, gender, and race; whether the respondent was born in the Twin Cities, majored in social work, was an intake worker, or worked in Hennepin County. The results show statistically significant impacts of the black baby vignette (compared to no baby or white baby) on the likelihood that respondents agree that the situation depicted in the picture meets the state definition of neglect and is reportable.
Also estimated are the impacts of these racialized beliefs on racial disproportionalities in reported and substantiated child maltreatment rates across counties. There are strong and statistically significant impacts of indices of caseworker racialized beliefs on racial disproporationalities in reported and substantiated maltreatment rates.