(iPoster) Convict the Innocent or Acquit the Guilty? Public Preferences for Legal Errors

• In Event: Poster Session: Women, Gender, and Politics Research

Abstract

“For the law holds, that it is better that ten guilty persons escape, than that one innocent suffer.”
(Blackstone, William. Commentaries on the Laws of England. Book IV, Chapter 27.)


A tradeoff about justice has occupied philosophers and legal scholars for centuries: is it worse to punish the innocent or to let the guilty go free? And if so, at what cost in terms of the latter should one avoid the former? While Blackstone (1769) famously argued that "it is better that ten guilty persons escape than that one innocent suffer,” political thinkers have held different priorities on this tradeoff.

At the core of the tradeoff is a debate about perceptions of justice that can shape modern legal systems and determine the standard of proof for conviction. Almost entirely overlooked, however, is public opinion on this matter. This paper theorizes and empirically analyzes how contemporary publics resolve this tradeoff, known as the Blackstone Ratio. In particular, we explain why women, typically both more progressive than men across a wide range of political issues and less punitive than men, are nevertheless more willing than men to tolerate the conviction of an innocent person.

We contend that the gender gap in public attitudes about the tradeoff reflects several factors having to do with the likelihood of victimization and differential severity of crimes, as well as women’s perceptions of their safety. In cases where women are more likely to be the target of crimes and less likely to be perpetrators compared to men, they will be more tolerant of false conviction compared to men (H1). Additionally, to the extent that women are more likely to be the target of violent crimes, they are more likely to tolerate the conviction of the innocent compared to men (H2). Lastly, it is not only the reality but also women’s perceptions that matter. The more women perceive their daily lives as unsafe (H3) and the less they trust the police and justice system compared to men (H4), the more tolerant they are of convicting the innocent compared to men.

Our empirical analysis is two-pronged. We first conduct a time-series cross-section analysis, drawing on five waves of the International Social Survey Programme (ISSP) in 33 countries (1986, 1996, 2000, 2006, and 2016). We then conduct an in-depth analysis of public opinion in the US drawing on the General Social Survey (GSS). Our empirical analysis demonstrates the gender gap in weighing the two types of errors, whereby women are on average more tolerant of convicting the innocent compared to men. Preliminary findings suggest that trust in law-enforcement authorities and perceptions of safety account for variation in the gender gap.

Understanding how citizens weigh these competing forms of injustice is of critical importance in an era of algorithmic governance. As we transition to AI-guided decision-making in bail, sentencing, and policing, these systems must be explicitly calibrated to favor one type of error over another. As we outsource judgment to machines, the question remains: whose definition of justice will be coded into the weights and biases of the algorithm? Our findings suggest that the "public will" regarding algorithmic thresholds is not a monolith, but is fractured by gendered experiences of safety and state authority.

Authors

• Orit Kedar, Hebrew University of Jerusalem

• Devorah S. Manekin, Hebrew University

• Pearl Leff, Hebrew University of Jerusalem