The Problems with the National Public Defense Workload Study
By John P. Gross, Clinical Associate Professor of Law & Director of the Public Defender Project, University of Wisconsin School of Law
Two years ago, the RAND Corporation published the “National Public Defense Workload Study”. That study concluded that public defender workloads should be roughly a third of the National Advisory Commission (NAC) on Criminal Justice Standards and Goals, standards that had been in place for 50 years. Shortly after the study’s publication, NAPD published their own guidance on how to interpret the report, which noted that the NAC standards were based on “educated guesses” and the news study’s approach is “more defensible.” Washington State may soon be the first state in the nation to adopt caseload standards based on the recommendations contained in the study.
I have found over the years that discussions about workload studies and caseload limits make public defenders uncomfortable and public defender leadership extremely uncomfortable. And for good reasons. The official ethical guidance from the American Bar Association is that having an excessive caseload is a violation of the Model Rules of Professional Conduct and public defenders who engage in the practice of triage can face professional discipline. It is an open secret that virtually all public defenders engage in the practice of triage and therefore, according to the letter of the law, are behaving unethically. And that was when many public defender organizations were using the NAC standards as guidelines. Then, overnight, a study is published that correctly points out that the NAC standards are arbitrary and outdated and announces new workload standards that are roughly one third of those contained in the NAC.
I suspect that upon first hearing about the new workload standards many public defenders had severe sticker shock. Could current caseloads really be that excessive? The study makes the case over 186 pages that they are. But after looking at the methodology used in the national report and the state reports upon which it is based, I think that the study overestimates the number of hours public defenders needed to handle various types of cases which leads to unreasonably low workload recommendations.
Before I explain why I have come to that conclusion, let me say three things. First, I agree with the study’s conclusions that the NAC standards are arbitrary and unreasonable and should be abandoned. Second, I believe that excessive caseloads are a serious problem, and we should find ways to address them. Third, I know that my opinion about the study will be unpopular with some people in the public defender community, but as I’ll explain, I think it is important to talk about how the application of the study could impact indigent defense systems and the right to counsel more broadly.
Cognitive Bias: Priming and Anchoring
My first concern with the study has to do with two related cognitive biases: priming and anchoring. Priming occurs when exposure to one stimulus impacts how you react to a subsequent stimulus. The first exposure “primes” you to respond in specific way. Anchoring occurs when your judgement is influenced by a reference point or “anchor”. All the participants in the national study received “detailed information on applicable ethical rules and professional standards” as well an overview of the previous studies. They also knew they were convening to create a new set of national workloads at a time when public defender workloads were widely regarded as excessive. Under those circumstances, the participants were likely primed to overestimate the amount of time needed to complete certain tasks. They also would have been anchored to the estimates produced by the previous state workload studies, whose participants would have been similarly primed to overestimate the amount of time needed to complete certain tasks.
One counterargument would be that the Delphi method that was used in the national workload study uses an iterative process aimed at achieving a consensus. That would help address the problem of an individual who was primed to respond in a certain way or who had anchored themselves to a position that turned out to be an outlier. But if the entire panel of participants in the study were influenced by the priming and anchoring effect, the iterative process wouldn’t compensate for those cognitive biases. And to be clear, I’m not suggesting that any of the participants in the study conscious engaged in overestimation. The type of cognitive biases I’m describing operate at a subconscious level.
Another consideration is that the national study, as well as many of the underlying state studies who rely on the Delphi method, do not use any existing data on the amount of time that public defenders are currently spending on cases. The study correctly points out that using existing data has the potential to simply reinforce the status quo. And since there is already widespread agreement that public defenders struggle with excessive caseloads, using existing data would just tell us where we are and not where we need to be. While that is true, inviting experts to estimate how much time public defenders should be spending on cases is speculative. To be clear, the people who are being asked to do the speculation are experts, but what we are left with is still an educated guess, even if it is a highly educated guess.
One can see the impact that priming and anchoring could have on a workload study by comparing two of the statewide studies used in the national study. Only one state, New Mexico, had two separate studies that were part of the national study. Those two New Mexico studies came to two very different conclusions about workloads based, I believe in large part, upon their methodology.
The first study, conducted in 2007, used time tracking by public defenders to arrive at an average amount of time they spent on certain case types. Then that average was shown to a panel of experts who were free to adjust the time needed to render effective assistance on that type of case. While the expert panel made some recommendations to increase the amount of time spent on certain case types, those increases were relatively modest, suggesting that the average number of hours reported might have served as an anchor. For example, the time study found that based on current caseloads attorneys could handle 174 felonies and 550 misdemeanors. The “quality adjusted” caseloads resulted in 144 felonies and 414 misdemeanors. The study recommended adding 41 full-time attorney positions, an increase of 24%.
Fast forward to 2022 and a New Mexico workload report that relied on the Delphi method. That report recommended adding 602 full-time attorney positions, an increase of over 200%. The 2007 study that relied on time tracking with subsequent adjustments to ensure adequate representation produced significantly higher workload recommendations than the 2022 study that uses the Delphi method which relies exclusively on expert opinion.
Cases That “Should” Go to Trial
The various state workload studies that use the Delphi method asked participants what percentage of specific case types “should” go to trial and what percentage “should” be resolved by plea. The studies then had participants estimate the hours needed to perform the various tasks necessary to resolve a case through plea or trial. The final average amount of time needed for a specific case type is then calculated based on the relative percentage of cases that are resolved by plea or trial and the relative number of hours needed to resolve that case type by plea or trial.
Currently, the jury trial rate for felonies in state court is often just 2% to 3% with misdemeanors trial rates often falling below 1%. The state studies using the Delphi method that are part of the national study consistently assert that trial rates for felony and misdemeanor cases “should” be around 30%.
I think it is fair to say that if public defenders had more resources trial rates would increase. Some clients might currently have little faith that their overworked public defenders can adequately defend them and might therefore decide to plead guilty. Public defenders who were given more time to investigate cases and do legal research might uncover viable defenses. But would these additional resources fundamentally change our system of plea bargaining into a system of trials? The forces that drive defendants to plead guilty, threats from prosecutors to file additional charges and the trial penalty imposed by judges, would still be present. Notably, even in the federal system, where the federal defenders have resources that go far beyond what state public defenders typically have, just 2.3% of defendants go to trial and just 0.4% of them are acquitted.
The differences between what trial rates currently are and what panelists in the Delphi studies imagine they should be, leads to significant overestimation of the time necessary to handle identified case types. For example, in the Oregon Project, which used the Delphi method, the report uses a “Low-Level Misdemeanor” trial rate of 31% and a “Complex Misdemeanor” trial rate of 45% which leads to a workload estimate of 22.26 hours for low-level misdemeanors and 36.98 hours for complex misdemeanors. If we use instead a trial rate of 2% for both classes of misdemeanor, which was the actual trial rate the year the study was published, we get a workload estimate of 12.66 hours for low-level misdemeanors and 18.13 hours for complex misdemeanors, drops of 43% and 51% respectively. Which means that if we use the actual trial rate for misdemeanors in Oregon at the time the study was published, public defenders could handle almost twice as many misdemeanors as recommended.
Prevailing Professional Norms and Effective Assistance
While NAPD has recognized that the national workload study claims to accurately estimate the number of hours needed to realistically meet the standard of effective assistance of counsel and are therefore not aspirational, there are significant differences between the standard of practice set by the national study and what courts consider effective assistance of counsel.
Consider the fact that the ABA Standards for Defense Function, which are extensively refenced in the national study, require defense counsel to conduct a thorough investigation and obtain whatever evidence is available through the discovery process prior to recommending that a client accept a plea offer. So, time for investigation and discovery is included in every case, even though courts have consistently found that failure to investigate or file motions can be an acceptable legal strategy.
Because the quality of representation can vary and still be considered adequate, all workload studies are based on a subjective judgment about the quality of representation that should be provided. The problem is that by using what amounts to a set of best practices for criminal defense lawyers, the standard that is set becomes aspirational for public defenders. The national study is based around what criminal defense should look like in the eyes of those who practice it, but the legislatures that fund indigent defense simply do not want representation that meets high standards set by bar associations but only the minimal standards set by courts. And while we would all like the ability to give our clients the type of representation that is consistent with professional norms, the people who pay us only want us to do the bare minimum.
Denial of the Right to Counsel as a Path to Decriminalization
Assuming I’ve convinced you that because of cognitive bias, inflated trial rates and qualitative judgments, the workload recommendations in the report are too low, you might still be wondering: what is wrong with that? If jurisdictions start adopting the recommendations in the report, public defenders will be able to provide excellent representation to their clients while also working fewer hours leading to the kind of work life balance that will help with recruitment and retention. That would happen. But what would happen if those workload standards created thousands of unrepresented defendants?
And that will happen. In states where caseload limits have been imposed and public defenders have refused cases, those indigent defense systems have been unable to find lawyers willing to take on the cases the public defenders refuse to take. You could argue that this isn’t the public defender’s problem, an argument that would be consistent with the ethical guidance requiring public defenders to avoid having an excessive caseload. But I don’t think we should ignore the consequences to potential clients and to the constitutional right to counsel more broadly.
And I don’t think the authors of the report are unaware of the consequences that will follow if public defenders are able to institute the recommended workloads. I think they want to put an end to assembly line justice and mass incarceration. Without lawyers, cases can’t go forward. The criminal legal system will grind to a halt. Legislators will be forced to find a solution. They won’t be able to scale up their indigent defense delivery systems to the levels recommended by the report, it would be prohibitively expensive and there just aren’t enough attorneys available even if they were willing to spend the money. So, if they can’t increase the supply of attorneys, they will have to reduce the demand. That means decriminalization and decarceration.
That could happen. We recently had an encouraging decision from the United States Court of Appeals for the Ninth Circuit who ordered the release of defendants in Oregon who were unrepresented. That being said, a dissenting judge called the decision “reckless and extreme” and said the majority was “complicit in a judicial jailbreak.” And recently an Oregon District Attorney said that Oregon’s public defense crisis was just a “work stoppage” by public defenders. We should expect that type of reaction from judges and prosecutors when defendants go unrepresented.
We should also be mindful of the fact that transformation changes will not happen overnight, and while it is happening a very large number of defendants are going to sit in jail without counsel. I would argue that we should try to keep that number to a minimum by making sure that we represent as many defendants as possible. And that is assuming the best-case scenario. We should also be prepared for the worst-case scenario.
What might happen if public defenders represent just a third of the clients that they currently represent, if the private bar is incapable of picking up the slack, if defendants go unrepresented, and state legislatures refuse to do anything to improve the situation. We would find ourselves in a constitutional crisis. Lawsuits would start to make their way through the courts. One of those lawsuits might make it to the United States Supreme Court. Would the current Court uphold the right to counsel or overturn Gideon?
I can easily imagine the current court overturning Gideon. An originalist interpretation of the 6th Amendment does not support the right to have counsel appointed at government expense. The widespread inability of states to provide counsel to indigent defendants could be viewed as evidence that providing counsel in all criminal cases is simply unworkable. Gideon would be declared a failed experiment.
Another cause for concern is that if these standards are not aspirational but become widely accepted, most public defenders will find themselves unable to meet the standards and will be vulnerable to allegations of professional misconduct for having an excessive caseload. At least one of the authors of the national study has endorsed the disciplining of a public defender who labored under an excessive caseload claiming that it “has sent shivers through public defender offices all over the country”. I see no reason to adopt a set of standards so unrealistic that they place public defenders in constant fear of professional discipline. And such a culture of fear is hardly conducive to the recruitment and retention of people willing to become public defenders.
We should absolutely try to address excessive workloads that dimmish the quality of the representation we can provide and that make a career as a public defender unsustainable. But we should not rush to adopt national standards that could significantly underestimate what is a reasonable workload for public defenders.