In many American courts, a public defender meets a client for the first time minutes before a plea hearing. Felony caseloads in the hundreds are not unusual, and misdemeanor attorneys often carry far more. The promise of Gideon v. Wainwright, that people who cannot afford a lawyer will be given one, is formally kept. The substance of that representation often thins out.

This is usually framed as a fairness problem, and it is one. It is also an empirical question that criminology can help answer. What actually changes when defenders have time to do the work their professional standards describe? Do outcomes shift, or does the same process simply feel more legitimate?

A growing body of evidence offers answers. It includes national workload studies, natural experiments in attorney assignment, and jurisdictions that have funded defense closer to recommended levels. Together they suggest that defender capacity is not a marginal input. It shapes pretrial detention, conviction patterns, and sentence lengths. It also shapes how the entire system allocates its scarce attention, which makes caseload policy one of the more consequential and least visible levers in criminal justice reform.

For nearly five decades, the dominant benchmark came from the 1973 National Advisory Commission. Its standards held that an attorney should handle no more than 150 felonies, 400 misdemeanors, 200 juvenile cases, or 25 appeals per year. These figures were never derived from rigorous time studies. They reflected the professional consensus of the era, yet they became embedded in court rulings, statutes, and funding formulas. Many offices still exceed them.

The 2023 National Public Defense Workload Study, led by RAND with the American Bar Association and partners, took a different approach. Using structured expert panels and time estimates across case types, it calculated the hours a competent defense actually requires. The findings suggested that the old standards were far too permissive. In serious felony categories, recommended annual caseloads fall to a small fraction of the 1973 figures. The gap reflects how much digital evidence, discovery volume, and forensic science have expanded the work.

Linking caseloads to outcomes is harder, because workload is rarely assigned at random. Still, the available evidence points consistently in one direction. Overburdened attorneys spend less time with clients, file fewer motions, conduct fewer independent investigations, and resolve more cases through early pleas. Client contact is often the first casualty. A defender with hundreds of open files may manage only a brief hallway conversation before a plea is entered.

The underlying mechanism is not mysterious. A plea is the cheapest way to close a file. When time is scarce, defenders triage, and triage tends to favor resolution over contest. This does not mean every plea entered under a heavy caseload is a bad one. It means the decision is shaped less by the strength of the evidence and more by the attorney's available hours.

Takeaway

A caseload is not just a figure on a budget sheet. It is an implicit decision about how many cases will be negotiated rather than tested.

Several jurisdictions offer partial natural experiments. In 2014, New York settled Hurrell-Harring v. State of New York, agreeing to improve defense in five counties through caseload limits, counsel at arraignment, and better access to investigators. Legislation in 2017 extended those reforms statewide, with the state gradually assuming the cost. Washington's Supreme Court adopted enforceable caseload standards in the 2010s, creating one of the clearest regulatory frameworks in the country.

Reporting on New York's reforms points to expanded counsel at first appearance, more consistent client communication, and greater use of support staff. Counsel at arraignment matters because bail decisions happen there. Pretrial detention is among the strongest predictors of conviction and incarceration, partly because detained people plead guilty to get out. Research on early representation programs has generally found lower detention rates when a defender is present at the first hearing.

Philadelphia offers a cleaner comparison. Murder defendants there were assigned either to the public defender's office or to private court-appointed counsel on an effectively random basis. Researchers James Anderson and Paul Heaton found that clients of the better-resourced defender office were roughly a fifth less likely to be convicted of murder and received substantially shorter expected sentences. The appointed attorneys faced low pay and thin support, the same capacity problem in a different form.

The efficiency question cuts both ways. Critics worry that more defense capacity means more motions, more trials, and slower dockets, and some of that does happen. But fewer people held pretrial reduces jail costs, and shorter sentences reduce corrections spending. Better-prepared cases may also prevent wrongful convictions and the costly appeals and post-conviction litigation that follow years later.

Takeaway

Defense funding looks like a cost when viewed in isolation. Its returns appear in jail populations, prison budgets, and appellate dockets, where different agencies pay the bill.

Professional standards describe what defense should involve. That includes early client interviews, independent investigation, thorough review of discovery, motion practice, plea negotiation informed by case strength, and trial preparation when warranted. Under heavy caseloads, these steps become aspirational. With manageable workloads, they become routine, and that changes how the adversarial system functions.

Investigation is the clearest example. A defender with time can locate witnesses while memories are fresh, secure surveillance footage before it is overwritten, and consult experts on forensic evidence. Motion practice follows from investigation. Suppression motions test whether police conduct met constitutional requirements. When such motions are rarely filed, the exclusionary rule loses much of its practical force. Police and prosecutors also receive less feedback about the quality of their cases.

Research on holistic defense shows what additional capacity can do. Anderson, Buenaventura, and Heaton compared the Bronx Defenders, where attorneys work alongside social workers, civil lawyers, and investigators, with traditional representation. Holistic defense reduced the likelihood of a custodial sentence by about 16 percent and expected sentence length by about 24 percent. They found no measurable increase in rearrest. The extra capacity let defenders present clients as whole people and address consequences like housing loss and deportation.

The most significant change may be credibility. Plea bargaining depends on the realistic possibility of trial. When prosecutors know a defender lacks the capacity to try cases, offers can harden. When the defense can credibly go to trial, negotiations track the actual evidence more closely. Manageable caseloads therefore affect not only the clients who go to trial, but the far larger number who never do.

Takeaway

Much of the value of trial readiness lies in cases that never reach trial. The real possibility of a contest is what keeps negotiation honest.

The evidence on defender workloads converges on a modest but important conclusion: capacity changes outcomes. When attorneys have time to meet clients, investigate, and litigate, people spend less time detained and receive shorter sentences. The available evidence suggests this comes without a measurable cost to public safety.

This reframes public defense as infrastructure rather than overhead. An adversarial system assumes two functioning sides. When one side is structurally under-resourced, the system does not become neutral. It shifts from adjudicating cases toward simply processing them.

For policymakers, the question is not only whether defense is worth funding. It is how to measure its contribution across budgets that rarely talk to one another. The answer may determine whether Gideon guarantees a lawyer, or a defense.