Topics

Each year, millions of Americans rely on public defenders to fulfill their Sixth
Amendment right to counsel. Despite being the linchpin of the criminal justice
system, public defense remains both underfunded and understudied. This Article
provides empirical analysis to contribute to a critical question: How should public
defender systems be structured?

Criminal justice advocates, scholars, and the American Bar Association strongly favor
vertical representation in public defense. Under this model, a single public defender
represents a defendant throughout their case, from their initial appearance through
sentencing. The alternative approach—horizontal representation—operates like an
assembly line: Different attorneys handle each stage of a case, from preliminary
hearings to pretrial conferences to trials. The preference for vertical representation
stems from the intuitive belief that continuity of representation improves outcomes
for defendants. However, no prior empirical work has tested this assumption.

Using a natural experiment created by the Defender Association of Philadelphia’s
transition from a fully horizontal representation system to a partially vertical one, we
find no evidence that increasing attorney continuity through a vertical representation
system improves defendant outcomes.

These findings have significant implications for how public defender offices should
allocate their scarce resources. While vertical representation is considered by many to
be the ideal, our results cast doubt on whether the additional resources and logistical
challenges relative to horizontal representation are justified given the current reality
of underfunded public defense. As jurisdictions nationwide grapple with a chronic
lack of resources for public defense, this article provides crucial empirical evidence
to inform decisions about how best to uphold defendants’ Sixth Amendment right
to counsel.

Within the literature on legal scholarship, academics have studied citation practices. For example, scholars have examined which authors, journals, and articles are most cited. But no one has examined which parts of articles scholars cite. Understanding which parts of articles scholars cite is not only intrinsically interesting, but also could inform how authors structure articles. This Note presents the results of a unique, hand-coded dataset of thousands of pinpoint citations. In brief: Authors are more likely to cite the beginning of articles but split their remaining citations roughly evenly. This pattern holds across flagship journals of variously ranked law schools and articles of varying length, but it is less pronounced for self-citation. While cynical explanations—that cite-worthy content is concentrated at the beginning, or authors tend not to thoroughly read the articles they cite—of the data is possible, a better explanation serves as a modest rebuttal to certain criticisms of legal scholarship.