# The Efficacy of Measuring Judicial Ideal Points: The Mis-Analogy of IRTs

Status: published
Visibility: public-pdf
Authors: with Mathew D. McCubbins and Kristen Renberg
Venue: International Review of Law and Economics
Year: 2021
DOI: 10.1016/j.irle.2021.106020
Canonical URL: https://doi.org/10.1016/j.irle.2021.106020

## Summary

Martin-Quinn-style judicial ideal points are revisited here by focusing on the item parameters that most applications ignore. A meaningful set of Supreme Court cases have discrimination parameters inconsistent with the assumptions of a unidimensional IRT model, and trimming or re-estimating around those cases yields more moderate justice scores and even changes the identity of the Court median in several terms. The broader point is that judicial ideal-point estimates are more fragile than users often assume, especially when they are treated as precise cardinal measures.

## Public Files

- Metadata: https://jyl19.github.io/papers/the-efficacy-of-measuring-judicial-ideal-points-the-mis-analogy-of-irts/metadata.json
- PDF: https://jyl19.github.io/papers/the-efficacy-of-measuring-judicial-ideal-points-the-mis-analogy-of-irts/paper.pdf