"140 Characters of Justice? The Promise and Perils of Using Social Media to Reveal Lay Punishment Perspectives"

The title of this post is the title of this new paper now available via SSRN authored by Itay Ravid and Rotem Dror. Here is its abstract:

For centuries now, penal theorists have engaged in heated debates about two questions at the heart of criminal legal systems: how can we justify the State’s power to punish individuals, and how can we determine what is the proper level of punishment.  To answer these questions, moral philosophers advanced conversations about several theories, which predominantly held in philosophical silos.  Over the years, calls to better understand community views regarding justifications of punishment and adopting penal law and policies that align with these views got traction, culminating in Paul Robinson’s “Empirical Desert” theory.  Despite its intuitive appeal, there have been criticism of this approach, both questioning its core hypotheses and expressing concerns about its perceived immorality.

At the same time, advancement in social-science methodology provided research tools to empirically deepen our understanding of lay people’s attitudes regarding punishment, mostly through surveys and experimental research designs.  One domain, however, remained untouched by those calling to assess lay intuitions of justice: social media. Such oversight is puzzling in light of social media’s potential to reveal public perceptions without scientific intervention.

This Article bridges this gap and engages with two main questions.  First, a methodological question: whether social media discourse can be used to reflect laypeople’s attitudes about criminal culpability and punishment, and second, a normative question: should it be used for these purposes.

To answer these questions, the Article first explores current scholarship about the promises and challenges of using social media data to study social perceptions.  The Article moves beyond theory, however, and utilizes recent technological developments in the field of Artificial Intelligence (“AI”) and Law and Natural Language Processing (“NLP”) to also explore empirically the potential promise of social media discourse in assessing community views on justice and punishment.

While the findings offer some support for the potentiality of using social media to assess laypeople’s attitudes regarding punishment, they also expose the complex challenges of utilizing such data, particularly for penal law and policy design.  First, due to a host of methodological challenges, and second, due to normative challenges, particularly social media’s polarizing nature and the ambiguity around whose voice is amplified through these platforms.