District Court Study

D. James Greiner, Cassandra Wolos Pattanayak, and Jonathan Hennessy, “The Limits of Unbundled Legal Assistance: A Randomized Study in a Massachusetts District Court and Prospects for the Future,” 126 Harvard Law Review 901 (2012).

We persuaded entities conducting two civil Gideon pilot programs to randomize which potential clients would receive offers of traditional attorney-client relationships from professional service provider staff attorneys and which would receive only limited (“unbundled”) assistance. In both pilot programs potential clients were defendants in housing eviction proceedings, and both programs were oversubscribed. In this Article, we report the results of one of these two resulting randomized control trials, which we label the “District Court Study,” after the type of the court in which it took place. In the District Court Study, almost all study-eligible eviction defendants received limited assistance in the form of help in filling out answer and discovery request forms, and most also attended an instructional session on the summary eviction process. After receiving this limited assistance, each member of a randomly selected treated group received an offer of a traditional attorney-client relationship from one of the legal services provider’s staff attorneys; each member of the corresponding randomly selected control group received no such offer. We compare outcomes for the treated (offered traditional representation from a service provider staff attorney) group versus the control (no such offer) group on a variety of dimensions, focusing primarily on possession of the unit, financial consequences of the litigation, and measures of court burden.

At least for the clientele involved in this District Court Study, a clientele recruited and chosen by the service provider’s proactive, timely, specific, and selective outreach and intake system, an offer of full representation mattered. Approximately two thirds of defendants in the treated group, versus about one-third of defendants in the control group, retained possession of their units at the end of litigation. Using a highly conservative proxy for financial consequences, treated group defendants received payments or rent waivers worth a net of 9.4 months of rent per case, versus 1.9 months of rent per case in the control group. Both results were statistically significant. Meanwhile, although treated cases did take longer to reach judgment, the offer of representation caused no increase in court burden as measured by other, more salient metrics.

Read more

Unemployment Representation Study

James Greiner and Cassandra Wolos Pattanayak, “Randomized Evaluation in Legal Assistance: What Difference Does Representation (Offer and Actual Use) Make?“, 121 Yale Law Journal 2118 (2011).

We report the results of the first of a series of randomized evaluations of legal assistance programs. This series of evaluations is designed to measure the effect of both an offer of and the actual use of representation, although it was not possible in the first study we report here to measure constructively all effects of actual use. The results of this first evaluation are unexpected, and we caution against both over-generalization and under-generalization.

Specifically, the offers of representation came from a law school clinic, which provided high-quality and well-respected assistance in administrative “appeals” to state administrative law judges (ALJs) of initial rulings regarding eligibility for unemployment benefits (these “appeals” were actually de novo mini-trials). Our randomized evaluation found that the offers of representation from the clinic had no statistically significant effect on the probability that an unemployment claimant would prevail in the “appeal,” but that the offers did delay proceedings by (on average) about two weeks. Actual use of representation (from any source) also delayed the proceeding; we could come to no firm conclusions regarding the effect of actual use of representation (from any source) on the probability that claimants would prevail. Keeping in mind the high-quality and well-respected nature of the representation the law school clinic offered and provided, we explore three possible explanations for our results, each of which has implications for delivery of legal services.

Read more


Social Security Disability

The Study

Field operation

To be eligible for the study, an individual must be an adult seeking to appeal an adverse decision regarding eligibility for disability benefits to an administrative law judge (“ALJ”).  The decision might have been either a denial of a request for reconsideration (under the traditional Social Security Administration (“SSA”) system) or an adverse ruling from a federal reviewing officer (under the new Disability Service Improvement (“DSI”) process).  The applicant might be seeking benefits under either the Social Security Disability Insurance (“SSDI”) program or the Supplemental Security Income (“SSI”) program.

After a thorough screening and intake process, consenting individuals are randomized into one of two groups:

  • Treated group: representation by student advocate in a law school clinic
  • Control group: a self-help packet on disability appeals, as well as referral to other legal services providers, and a copy of their own intake information (to streamline the information-gathering that another legal services provider would need).


Randomization is currently active. When the field operation is over, we will analyze the following outcomes for both groups:

  • Were benefits awarded or denied?
  • If awarded, what amount?
  • Compliance with the randomization: did individuals in the control group obtain representation elsewhere? Did individuals in the treatment group continue with their representation?

The Research Team

Jim Greiner, Faculty Director, The Access to Justice Lab; Professor of Law, Harvard Law School

Federal Court Mediation

Alternative Dispute Resolution (ADR)

Almost every court system in the nation has an ADR program, and for most courts, at least some classes of litigants are compelled to use it before or during formal litigation.

The most common forms of ADR are:

  • Mediation: sessions in which the parties and/or their lawyers meet with a neutral (the “mediator”) who attempts to find common ground.  Mediation format can be caucus or non-caucus, depending on whether (or not) the mediator meets each party outside of the presence of the other.  Mediation can be evaluative or facilitative, depending on whether (or not) the mediator provides her views as to the merit of the dispute.
  • Judicial settlement conference: similar to mediations in the goal and the format, but the neutral is a judge or other decision-maker. Judicial settlement conferences are evaluative.

The Debate Surrounding ADR

Proponents of ADR clam that it furthers social welfare on at least four separate dimensions:

  1. conserving judicial resources,
  2. conserving party resources,
  3. increasing party satisfaction, and
  4. increasing party compliance with the decision or outcome.

Opponents, meanwhile, claim that direct negotiation can achieve these same benefits, and that the high cost of litigation provides a strong incentive for parties to settle on their own. The question, therefore, is whether the presence of the ADR neutral (a mediator, a judge) is really necessary to help the parties do what they could do on their own via direct negotiation.

Who is right? We do not know. There is little reliable evidence of any kind, and no reliable evidence at all from the past 20 years, to assess ADR effectiveness along any of these four dimensions discussed above.

The Study

The present study is the first of any kind to evaluate rigorously all of the dimensions ADR proponents cite in advocating such programs. We conducted a randomized control trial in one setting: civil rights cases brought by inmates in one federal correctional facility in Nevada.

Field operation

After a thorough screening and intake process, consenting individuals were randomized into one of two groups:

  • Treated group: mediation
  • Control group: strong suggestion to negotiate


Randomization has closed, and more than half of the cases have reached final resolution. Data we will analyze:

  • Federal court case records
  • Number of hours spent by state attorneys general per case
  • Case outcomes, in both settlements and court orders
  • Inmate plaintiff surveys
  • Future complaints

The Research Team

Jim Greiner, Faculty Director, The Access to Justice Lab; Professor of Law, Harvard Law School

Intimate Partner Violence

The Problem

Right now, in the United States, some victims of domestic and intimate partner violence (DV/IPV) do not obtain restraining orders against their abusers, not because their facts fail to meet applicable legal standards, but rather because legal services providers lack knowledge needed to allocate scarce resources well, i.e., to triage. Were such triage know-how available, more DV/IPV victims might obtain restraining orders, even without new resources.

What is triage?

Triage is scarcity-compelled choice. The military setting provides a useful guide. The Navy Hospital Corpsman Manual instructs field medical technicians facing emergencies with multiple civilians to divide human beings into three classes:

  • Priority I. Victims who “require immediate life sustaining action.”
  • Priority II. Victims with “injuries where treatment can be delayed for a short time.”
  • Priority III. Those who either “have minor injuries,” or those with “obviously mortal wounds where survival is not expected.”

When resources are too scarce to permit treatment to be given to all, the Manual’s instructions are clear: treat Priority I patients. When you cannot treat everyone, treat those who will live with treatment but die without it. Doing anything else may cost lives. Most critically, the Manual provides instruction on how to identify and separate these classes.

On a day-to-day basis, legal aid attorneys facing severe resource constraints should provide representation first to Priority I victims. Doing anything else may cost lives.

At current resource levels, every representation given to a victim in Priorities II or III (i.e., those for whom representation can make no difference vis-à-vis restraining orders) means that two other victims, perhaps two other Priority I victims, go unrepresented. Like Navy Corpsmen, legal aid attorneys may despise the necessity for triage, but they cannot escape it. Some triage mechanism is unavoidable.

But unlike Navy Corpsmen, legal aid attorneys have no manual to instruct them on how to identify those who will obtain restraining orders (no matter what) and those who will not obtain restraining orders (no matter what) from those who will obtain restraining orders only if legal aid attorneys represent them.

Why is there no such manual? Because no one knows what the manual would say. No one has ever researched how to triage well in the DV/IPV context, nor in any legal context, even though courts and attorneys make triage decisions daily.

The Study

The Intimate Partner Violence Triage Study will provide the first rigorous study of triage in the DV/IPV context. It will do so using a randomized control trial design in partnership with a legal aid organization in northeastern Ohio.

Self-help materials

A key element to this project involves creating self-help materials that are useful and accessible, and that convey the appropriate message to the audience. We have developed several sets of images to accompany the text for these materials.

Stick figure restraining order Reverse geometric figure restraining order Bears restraining orderPrint

Please help us by casting a vote for the image you think works best.

Field operation

After a thorough screening and intake process, a legal aid attorney will make a provisional triage decision assigning each eligible and consenting DV/IPV victim to one of its three service levels: (a) full attorney representation; (b) an instructional telephone call plus a self-help assistance packet; or (c) a self-help packet alone.

Eligible and consenting individuals are then randomized into one of two groups:

  1. Triage by an attorney: the person will receive the service that is decided by the attorney.
  2. Randomized triage: the decision made by the attorney will be ignored, and the person will be provided a randomized level of service as chosen by a computer.


The Research Team will follow each victim to discern whether she obtains a restraining order, and subsequently, whether she seeks refuge at a shelter (a strong indicator of revictimization). Depending on funding and the availability of records, the Research Team will also examine police records (to see whether a subsequent encounter is reported for the victim and her abuser) as well as Medicaid expenditures for each victim. By comparing these outcomes across the two treatment arms, the Intimate Partner Violence Triage Study will assess the effectiveness of the attorney-led triage process in DV/IPV cases.

Unlike many randomized trials, however, the Study will go beyond mere “treated” versus “control” comparisons and provide strong guidance to future service providers and researchers. Specifically, the presence of a fully randomized group (the second treatment arm) will allow the researchers to provide data-driven, rigorous evidence on the following questions critical to the legal scheme constructed to protect DV/IPV victims:

  • What background variables (observable at intake) predict whether a legal aid provider will offer higher as opposed to lower levels of service?
  • What background variables (observable at intake) predict whether a DV/IPV victim can obtain a restraining order on her own?
  • How much does legal representation increase the probability of obtaining a restraining order over a telephone call and/or a self-help assistance packet?
  • How effective is a restraining order in preventing revictimization, forestalling the need for subsequent police calls, and reducing state medical expenditures?


The Research Team

Marie Curry, Managing Attorney, Health, Education, Advocacy & Law Program, Community Legal Aid Services, Inc.

Jim Greiner, Faculty Director, The Access to Justice Lab; Professor of Law at Harvard Law School

Erika Rickard, Associate Director of Field Research, The Access to Justice Lab

David Yokum, U.S. Social & Behavioral Sciences Team

Pre-Trial Release

The Problem

Prison overcrowding and associated issues with state and federal sentencing policy receive ample attention among criminal law scholars and practitioners. Just as important, however, are the problems created by ineffective detention policies before trial. A number of reform efforts are underway to better predict an arrestee’s actual risk of failure to appear (FTA) and recidivism. The most prominent new tool is the actuarial risk assessment mechanism, one of many evidence-based practices now used in criminal and other legal spheres. The theoretical use of such mechanisms has come under fire from various quarters. Yet we do not know in practice how effectively these risk scores mitigate FTA and recidivism.

These instruments, despite their promise, give rise to some concerns worthy of study. Many rely on subjective factors, which require the exercise of judgment to score; their predictive power has not been assessed using data from the jurisdictions in which they are used; most instruments depend on information that, realistically, can only be obtained from an arrestee interview; and, crucially, only two instruments have undergone randomized evaluation in over six decades of use in the United States.

The Study

In partnership with the Laura and John Arnold Foundation (LJAF), the A2J Lab will evaluate the effectiveness of the Public Safety Assessment (PSA) in several jurisdictions across the country. The purpose of the study is to test whether an actuarial risk assessment tool that avoids the need for arrestee interviews produces better pretrial incarceration decisions and associated decreases in FTA and repeat offenses. (Visit our Open Science Framework page for more information.)


A team from Luminosity, led by Marie VanNostrand, developed the PSA. To date, it has been fully implemented, and is fully operational, in a variety of U.S. jurisdictions, including the counties of Mecklenburg, North Carolina and Santa Clara, California, as well as all of Kentucky and New Jersey.

The perceived benefits of the PSA include:

  • Reliance only on objective inputs, i.e., those that can be observed without the exercise of significant subjective decision-making.
  • Calculation using only administrative data, obviating the need for an arrestee interview.
  • Avoidance of factors that are highly correlated with problematic characteristics, such as race, national origin, or economic class.
  • Amenability for use in jurisdictions throughout the United States, regardless of a community’s demographics or its criminal activity levels.

Thflowchart2e PSA’s inputs are, according to Luminosity research, correlated with the risk of FTA or engaging in new criminal activity (e.g., arrestee age and whether he has failed to appear within the last two years). An algorithm transforms the various inputs into two scores on a 1-6 scales, one each for FTA and new criminal activity (NCA), as well as a “Violence Flag,” signifying an elevated risk of committing new violent criminal activity (NVCA). A Decision Making Framework (DMF) is then applied and consists of a set of instructions to transform the two FTA and NCA scores, the Violence Flag, and other factors (such as whether the current charge is for certain domestic violence-related offenses) into a recommendation for a judge making an initial release/detention decision. The translation of scores into recommendations involves fundamental questions of a jurisdiction’s budget, tolerance for risk of arrestee misbehavior, attitudes toward incarceration, current and anticipated jail population, and other jurisdiction-specific factors.  As a result, the DMF’s recommendations are the result of a conversation between the LJAF and an implementing jurisdiction, with the LJAF insisting that the jurisdiction adhere to the principle that the DMF’s recommendation be reached based on a consideration of the PSA scores, not solely the current charge against the arrestee.

Field operation

The Lab is currently engaged to evaluate the PSA’s effectiveness in Dane County, Wisconsin.

The core RCT structure is simple, and all relevant parties in Dane County have informally agreed to it. The LJAF’s team is working with Dane County to implement the PSA, including the construction of a software system that can receive data inputs and produce the basic PSA printout for the judge (known in Dane County as the Commissioner), the prosecutor, and the defense attorney to examine.

Cases will be randomized either to the “treatment” group, in which instance the Clerk’s Office will produce the PSA printout, append it to the case file, and make it available to the prosecuting and defense attorneys as well as to the Commissioner in time for the initial appearance; or to the “control” group, in which instance, the Clerk’s Office will not produce the PSA printout. Randomization will be by case number.


The principal outcomes of interest for this study include the following:

  • Duration between initial appearance and case disposition;
  • Whether the arrestee was convicted of any charge;
  • Whether the arrestee was sentenced to any term of incarceration and the number of months ordered to serve, if any;
  • Failure to appear, both whether the arrestee missed any court date in the case as well as how many FTAs the arrestee exhibited.
  • Whether the arrestee engaged in any new criminal activity and the number of separate incidents (both between initial appearance and case disposition and up to two years after initial appearance);
  • Whether the arrestee engaged in any new violent criminal activity and the number of separate incidents (both between initial appearance and case disposition and up to two years after initial appearance); and
  • Whether the arrestee was released at any point between initial appearance and case disposition, the number of days the arrestee spent incarcerated during that period , and the share of days from initial appearance to case disposition that the arrestee spent incarcerated.

The Research Team

Jim Greiner, Faculty Director, The Access to Justice Lab; Professor of Law, Harvard Law School
Chris Griffin, Research Director, The Access to Justice Lab
Heidi Liu, J.D./Ph.D. candidate, Harvard Law School and Harvard Kennedy School

The Problem of Default

The Problem

In the modern United States, too many lawsuits are decided by default. This is especially true in debt collection cases, where reported default rates frequently range from 60% to 95%.
Default is certainly bad for defendants, but perhaps more importantly, default engenders a system in which the state publicly declares a winner to a dispute without any opportunity to assess relevant facts and apply the law. For this reason, default threatens to undermine faith in the judicial system.

Thus our study seeks to answer the question: what steps can legal services providers take to facilitate defendant attendance in court?

The Study: Part I

The Debt Collection Default Study measures what kinds of mailings from legal services providers to defendants are effective in reducing default rates in debt collection cases. To our knowledge, our study is the first of its kind to evaluate an intervention intended to reduce default rates in civil cases using a randomized control trial.

Designing self-help materials

The goal of our self-help materials was to induce debt collection defendants to (1) open, (2) read, and (3) act upon them. In terms of what action needed to be taken, our materials needed to induce a litigant to fill out three copies of an Answer form (which we would provide), mail two of those copies, receive a notice from the BMC of a scheduled court date, and appear in court on that scheduled date.

We designed interventions consisting of two forms of mailings based on research from other disciplines, including psychology, public health, and adult education.
Cartoon of Blog mailing answer forms.

Field operation

The Research Team received debt collection case information on a weekly basis from court staff at the Boston Municipal Court. After identifying potentially study-eligible cases, we compared the defendant’s address in the court file with that in an online address-checking system, and excluded those with inconsistent addresses.

The Research Team randomized whether each defendant receives a mailing, and if so, what kind. Using different formats and messages, the mailings urge defendants to contest their cases (by filing any necessary paperwork and showing up to court) and provide information/materials useful for such a defense (such as an Answer form).

Treatment group A, the Limited treatment:

  • letter from the legal services provider
  • three copies of a check-box style Answer form
  • business envelopes pre-addressed to the court and the plaintiff’s attorney
  • map to the courthouse
  • and a post-it note appropriate for a wall calendar saying “Go To Court Today!”

Treatment group B, the Maximal treatment:

  • First, a postcard with the signature Blob cartoon, stating: “Dear [Recipient Name], Help is on the way. Look for me!.” Next to “me” was a hand-drawn arrow pointing to an image of Blob.
  • The next day, we mailed the defendant the same manila envelope (with corresponding contents) that those in the “Limited” group received, except that the two business envelopes to the Court and to the Plaintiff’s attorney had stamps.

Control group: no mailing


We tracked two outcomes:

  • Whether the defendant filed an answer, and
  • Whether the defendant attended the first court hearing.

We find no difference in effectiveness as between our two mailings, but that both roughly double the rate at which defendants participate in their lawsuits. Specifically:

Chart comparing answer rates in debt collection cases

As compared to a randomly selected Control group with a 13% answer rate (corresponding to an 87% default rate), our “Limited” intervention group saw a 24% answer rate, and our “Maximal” intervention group saw a 24% answer rate.





The corresponding rates for whether the defendant appeared at the first scheduled court hearing were 7.5% for the Control group, 14% for the Limited group, and 15.3% for the Maximal group.



Differences between the Control versus the Limited and Maximal groups were statistically significant. Differences between the Limited and Maximal groups were not.

What this means

The results of this study shed some light on a few different areas of debate in the legal arena:

  • why people obey the law and engage in official proceedings (and why they don’t)
  • the role of civil legal services providers and the types of services they provide
  • how courts present themselves to and interact with people without lawyers

The Study: Part II

Field operation

This study builds on the smaller pilot study in Boston, and will include multiple legal service providers and multiple court locations.

By randomly varying the format and content of the package, the Research Team will learn what is necessary and cost effective to reduce default rates. Potential areas of exploration include the appearance of the external envelope; the text of the letter; whether the letter includes cartoons and/or other illustrations; the contents of the package (e.g., whether Answer forms, return envelopes, maps to courthouses, and reminder post-it notes are included); and whether the materials are translated, and if so, into what languages.

The Research Team

Jim Greiner, Faculty Director, The Access to Justice Lab; Professor of Law, Harvard Law School

Dalié Jiménez, Associate Professor of Law, University of Connecticut School of Law

Andrea Matthews, Bureau of Consumer Protection

More Information

About this study and previous research
About debt collection and default rates


Financial Distress

The Problem

Cartoon, legal self-help, debt collection, show me proofAs of 2014, more than 77 million people in the U.S. had at least one account reported as “in collection” on their credit reports, owing an average of $5,178 (median $1,349).   Distressed debt results in collection lawsuits, a messy and error-prone credit report, and a potential need for bankruptcy.  In other words, debt problems are legal problems, and an inability to resolve debt problems leads to legal consequences.  What proposals are out there to address the legal aspects distressed debt?  How would we know whether those proposals work?