Randomized Trials Do Not Require Abandoning Merits Screens Or Ethical Duties: A Response to Margaret Monsell
To a hypothetical legal services provider that objects to a proposed randomize study on the grounds that participating would compel it to represent “clients with unmeritorious claims” (see Margaret Monsell post below), we ask the following question: if someone won a lottery and offered to give that provider an extra $100,000 to represent additional cases in a practice area, would the provider turn the money down on the grounds that doing so would compel it to represent clients with unmeritorious cases? In other words, is the match of resources to caseload perfect, or nearly perfect, as matters currently stand, or are there practice areas in which a great many clients who would benefit from legal assistance in fact go unrepresented?
We very much appreciate Margaret Monsell’s taking the time to comment on our paper. One of the most encouraging and informative aspects of our current research is that we have had the opportunity to interact with persons in both the front line and, now, the “backup” aspects of legal services delivery and adjudicatory system improvement.
We write to clarify that that randomized trials are entirely consistent with merits screens and with lawyers’ duties of candor to the court.
We believe that oversubscription of legal services programs is a serious, ongoing, and worsening problem. Survey after survey has show that potential clients with meritorious claims or matters who might benefit from legal assistance go unrepresented. If that is right, then there is no reason for randomized evaluation to require legal services providers to represent clients with unmeritorious claims. All that need happen is that a screening process for merit (perhaps something very close to the screening process that preceded the randomized evaluation) occur before cases are randomized. Thus, a legal services provider’s intake system screens out potential clients with unmeritorious claims, just as it did before the evaluation. Then, and only then, are the remaining cases randomized. Unless the answer to the $100,000 hypothetical question we posed above is that providers in general would have no use for additional money, then in many practice areas there should still be pool of post-screen potential clients larger than provider capacity. Randomization is applied to this still-larger-than-capacity pool.
Meanwhile, if a frivolous case slips through a merit screen applied prior to randomization, as will no-doubt occasionally happen, the representative does the same thing as what he/she would do if a frivolous case slipped through when there was no randomized study: withdraws from representation, or takes other appropriate action. The outcome is then coded as a win or a loss, success or failure, etc., based on the outcome of the case.
We know that this process can work because it has worked, and is working, in other randomized evaluations we are conducting. In a summary eviction evaluation in which we are participating with GBLS, MLAC, Mass IOLTA, and others, the legal services provider (GBLS) applied strong merits screens before sending us cases for randomization. We still had approximately 40% more cases than GBLS had capacity to represent. We never heard from the GBLS attorneys involved that they did not have enough to do (quite the contrary). Similarly, in an SSI/SSDI disability study we are conducting with the Legal Services Center in Jamaica Plain, the Center applies some merits screens before sending us cases to randomize. This study is extremely slow-moving because of the glacial pace of SSA’s hearings process, but there has been no issue regarding the merits screens.
Once again, many thanks to Margaret for taking the time to write.
Jim and Cassandra