Work in Progress

Jesse H. Rhodes. Understanding Severe Backsliding in Electoral Democracy in the American States. (appendices). Although researchers have begun to investigate the causes of backsliding in electoral democracy in the American states, to date this work has focused on assessing how factors influence democratic decline on average. Researchers have paid relatively little attention to the phenomenon of catastrophic decline in the quality of democracy, which has afflicted numerous states. In this paper, I use a most similar systems qualitative research design, along with explicit Bayesian inference, to assess the causes of severe backsliding in electoral democracy. I find that the combination of a period of unified Republican control of government and extremism among Republican elected officials is sufficient to produce severe backsliding.

Douglas R. Rice, Jesse H. Rhodes, and Tatishe M. Nteta. Partisan Bias in Juror Decision Making. Building on our previous work documenting the influence of racial resentment on jurors' behavior in trial involving a black defendant, in this study we examine whether and to what extent partisanship biases the decisions of jurors in a trial featuring a Democratic/Republican defendant. We find that jurors exhibit bias against out-partisans in determinations of guilt, but not in sentencing. 

Douglas R. Rice, Jesse H. Rhodes, and Tatishe M. Nteta. Is Justice (Color)blind? The Language of Law, Implicit Racial Bias, and the Impact of Judicial Selection. Recent research demonstrates conclusively that differential racial treatment is reflected in the language of the law, but we know little about what moderates or exacerbates the expression of this bias. Given the well-documented effects of alternative judicial selection methods on judges’ decision-making, we argue it is plausible that judicial selection mechanisms may influence the expression of differential racial treatment in judicial opinions. More pointedly, because existing research suggests that selection methods (such as non-partisan elections) that increase judges’ exposure to public opinion induce greater departures from the norm of judicial impartiality, there are reasons for concern that – in a racially divided polity where both explicit and implicit animus toward non-whites are widespread – such methods might exacerbate expressions of bias in judicial opinions. In this paper, we leverage the variability in judicial selection and draw on the full text of over 400,000 state appellate opinions appearing between 1955 and 2015 to explore whether and how methods of judicial selection affect the prevalence of racial bias toward African Americans. We employ word embeddings trained from judicial opinions in conjunction with new methods for estimating statistical relationships between concepts in text to evaluate the relationship between the method of judicial selection and the severity of demonstrated racial bias. We find evidence that across the mode of judicial selection that the “language of the law” is indeed rife with implicit racial bias. More specifically, our preliminary results show that this bias is most pronounced in states that have competitive partisan judicial elections.