Discussion Draft 12/13/19
The AALS Section on Scholarship, Advisory Committee on Law Journal Reform
No one is satisfied with today’s legal publishing. The long-standing tradition of simultaneous submission to student-edited journals has always involved tradeoffs, but the costs of that approach have grown dramatically over the last decade. Where once even top journals faced a relatively manageable task in identifying promising submissions, technological innovation now enables authors to easily submit to hundreds of journals with a few clicks. The result has been enormous practical and even ethical pressures on students and authors. Top journals receive more than 4,000 submissions annually. Selection outcomes are often driven not by merit but by insider knowledge, such as whether an author knows when journals are open to selecting articles or how to “expedite” publication offers to more-preferred journals. Increasingly, top journals are demanding exclusive submission windows, undermining one of the core strengths of the traditional structure. With few clear rules of the road, opportunities for gamesmanship on each “side” are prevalent, and may be mutually reinforcing.
While we believe that legal academia can and should agree on “best practices” to improve how authors and editors conduct themselves, we are realists. No set of idealized norms can succeed in the face of enormous structural pressures. Fundamental reforms are necessary.
Thus, the Section offers two possible paths for reform, each of which can be further tailored. In the simpler path, authors will submit to a small number of journals at a time, and must accept the first offer received. Journals will not extend offers during a “quiet period” of four weeks or so. A more ambitious path involves adoption of a two-round Shapley matching system, better known as the “med school” match. In that path, authors will rank a set of journals from which they would accept offers, and journals will rank those articles that meet their publication threshold. Both paths can be combined with a new peer review pool, as we describe, and additionally AALS Member Schools can adopt and encourage compliance with a set of complementary best practices for authors and editors.
Though we detail the strengths and potential weaknesses of these options in more detail below, we want to emphasize here their overwhelming advantage over the status quo: each would essentially eliminate expedited review. Expedited review is the root cause of nearly all the problems we and other stakeholders have identified with the current approach. It motivates mass submissions and other, even less fortunate, gaming behaviors. It turns many journals into screening editors for journals that are more preferred by authors, greatly increasing both their workloads and frustrations. The time pressures it imposes make meaningful peer review next to impossible. And it systematically rewards authors who are most expert at navigating the system.
An alternative, of course, would be to turn to the exclusive-submission model common in other academic disciplines. We believe that would be too radical a step. It would greatly extend time to acceptance for most authors without alleviating the crushing workload of top-journal editors. Further, many outstanding law journals — although not enough, in our view — already operate under the traditional exclusive-submission/peer-review model of the social sciences. We believe that preserving both paths is important for the discipline.
A. Objectives for an Effective Legal Publishing Regime
Publication in top-ranked journals is the sine qua non for tenure in most academic disciplines, especially in the social sciences. University-level tenure committees at many AALS member schools also have a similar expectation for law-school candidates for promotion.
Whatever one thinks of the tenure process, the priority of elite publications reflects the centrality of the idea of excellence to academic identity. In choosing the works that best exemplify the goals and methods of the field, disciplines define themselves.
For that reason, among others, the Section believes that a properly-functioning process for matching articles with law journals is essential to the continued vitality of legal academia. Article selection and placement is one of the ways in which the academy evolves. Major intellectual currents of the past, be they Legal Realism, Critical Theory, Law and Economics, or the like, all have been reflected in the pages of the most selective journals. Articles in these journals have shaped opinion, spurred debate, and served as models of reasoning and approach for new scholars.
Ideally, journal placement also serves as an important signal to readers outside a discipline. Public debates about what the law “is” or what it allows remain among the most prominent of anything discussed in the world today. As evidenced by the stream of recent legislative enactments and even presidential candidate platforms, legal scholarship continues to be read today by journalists, policy makers, and academics from other fields. These readers often lack the ability to identify skillful and good-faith legal scholarship. In a well-functioning matching system, journal placement can be an important signal of the extent to which a profession views the relevant article as exemplary — not necessarily correct, but reflecting the depth of thought and carefulness of method that the discipline demands.
In addition to its possible contributions to these grand purposes, the signaling value of journal placement can also be useful in small, quotidian ways to members of a discipline. We all agree that the best way to judge the quality of a work of scholarship is to read it. But there are many circumstances in which time does not allow a consumer of scholarship to read it from front to back before deciding on its quality. Which references should the law clerk or the research assistant on a tight deadline scan first? Which resumes, among the hundreds available, should receive the closest attention? What article, among the thousands posted each month, should a busy scholar read to keep up to date on the most important recent developments in her area?
In short, the Section believes that the placement of a given article in a given journal should be a meaningful signal discernible by many potential readers. Readers should know where to find the scholarship that will be most useful to them, and should get an immediate sense of whether legal academics would likely consider that scholarship to be of high quality.
Unfortunately, as we describe in Section I.B. below, legal publishing is falling farther and farther from these goals. Immense structural pressures have greatly reduced editors’ ability to give full and careful attention to submissions, and made opportunities to consult in depth with faculty all but impossible. Under these circumstances, the signaling value of journal placement, whether to other legal academics or to outside readers, is increasingly in doubt.
Of course, signaling is not the only function of a journal. Through strong editing, journals can also contribute to the clarity, reliability, and originality of a work of scholarship. Law journals have historically placed a great deal of emphasis on the transparency of an author’s support for each factual proposition, as well as on exposition of background information that would allow unfamiliar readers to engage with the article in depth. These are both important sources of added value that, while not universal in other disciplines, clearly can be useful to some readers.
In our view law journals remain largely successful on these other fronts. Editors and authors sometimes disagree, however, on how best to strike an appropriate balance between credibility and needless over-documenting, or between helpful background and unnecessary filler. While not the focus of our discussion here, we think there is room in a set of future best practices to set out ways in which publishing partners can better respect each other’s views on these issues.
B. Current Challenges for Legal Publishing
The Section has concluded that current journal submission practices only poorly serve the ends we have just described, while also imposing considerable and unnecessary time, dollar, and even ethical costs on all participants. Our view reflects our own experiences, those reported on-line and in personal conversations, and the results of a survey we conducted in 2018 of law journal staff. Our survey was emailed to approximately 200 general-interest journals, and we received forty-two responses. We welcome readers who believe their home journal has not completed a survey to contact us for a survey link.
As prelude, we briefly summarize the typical submission process. Authors submit simultaneously to as many journals as they choose. Offering journals will usually allow an author who has received an offer a window of time in which to decide whether to accept. During this period, authors usually “expedite” to other journals, asking them to complete their review in time to respond to the exploding offer from the first journal. In some instances, authors will then in turn shop a second or even third offer to journals “farther up the chain.”
Deadlines are usually shorter for journals most desired by authors and for journals whose own offer followed an expedite request. For instance, the Columbia Law Review offers a one-hour window for offers made in response to an expedite request (often referred to in short-hand as “an expedite”). In our survey, the median deadline was one week, although the mean period was shorter for offers extended after an expedite.
Multiple submission is facilitated by a pair of on-line platforms, Scholastica and Expresso. Users can upload an article and accompanying documents to each platform, select destination journals, and submit to all of the selected journals with one additional click. The platforms charge a per-journal fee for this service, currently $6.50 per destination journal at Scholastica, although both offer “hardship waivers” for qualifying authors. Most journals are available only on one of the platforms, and most general-interest journals now appear only on Scholastica. Some journals maintain their own proprietary web sites for submissions and do not appear on either platform.
The platforms also facilitate mass communications with journals. Users can send requests for expedited review to any selected subset of journals to which they have previously submitted. This process can be completed in as few as three clicks on Scholastica, which pre-fills information on the offering journal and its deadline if that offer was made via the platform.
The volume of submissions the platforms enable is staggering. The most selective journals in our survey report receiving upwards of four thousand annual submissions. Although a plurality received fewer than one thousand, the median journal received between one and two thousand.
Even more staggering is the number of requests for expedited review. The mean number of expedite requests reported was 432; one journal received more than 1,500. To repeat, these are typically requests to complete consideration of the article within one week or less.
Competition among journals appears to be quite vigorous. Most journals make about three times as many offers as the number of slots they must fill, although this number is typically higher for “mid-tier” journals: those that receive many expedites but also lose many articles to more-preferred journals. One journal reported making eighty offers to fill its dozen or so slots.
As might be expected from these numbers, journals do not spend much time with the average submission. Most articles are rejected after an initial screening read that lasts a mean of about thirty-nine minutes. Journals reject without ever reading about one-third of articles that come with an expedite deadline, and the mean share of other articles receiving any review at all is a bit lower, 64%.
Journals also appear to rely frequently on proxies for article quality. A majority of journals agree that they look to the author’s employer and track record of publications in evaluating a submission. Journals often prioritize articles for review based simply on which ones have the most urgent deadline, although the perceived quality of the journal extending the prior offer is also a factor for many respondents.
In our view, the modern environment has significantly reduced the correlation between article placement and article quality. Simultaneous submission, expedited review, and the accompanying pressures they create introduce a great deal of randomness into the article-selection process. Further, placement outcomes depend significantly on an author’s familiarity with and willingness to engage in strategies for optimizing the chances of obtaining sustained attention from reviewing editors at the most competitive journals. Authors who do not optimize may still obtain excellent placements, but on average their odds of doing so are smaller. This again diminishes the meaningfulness of the signal sent by placement success, as well as systematically stacking the deck against “outsider” authors or those who will not engage in optimizing strategies.
For example, successful optimization of the expedited review process requires careful timing. Authors must ensure that an initial offer arrives at a time when many of her more-preferred journals are able to undertake expedited review: if she submits too early in a fall or spring submission window, her initial offer may expire before other journals begin to read. If she submits too late, her most-preferred journals may have closed for the season before she secures her initial offer. It’s often difficult for authors to know where this sweet spot resides. Journals are far from uniform in how or even whether they report when they will begin and (especially) finish their reading for a season. On-line platforms typically have an indicator of whether a journal is “accepting submissions,” but this indicator is not consistently updated by journals and does not always reveal whether the journal’s editors are in fact able to respond to expedite requests. The Fall cycle seems especially wide and unpredictable: in our survey, some journals report ending their typical Fall review before other journals report even beginning.
Journals game the timing of expedited review in a number of ways. Again, some offer deadlines so short as to be almost meaningless. Others purport to offer a full week but will extend offers on Friday evenings, on the eve of Spring Break, at the end of a submission season, or otherwise at times when they believe it will be difficult for authors to obtain a response from other journals.
Further compounding the information dilemma for authors, most journals do not acknowledge requests for expedited review, and only intermittently communicate a negative decision within the author’s stated deadline. Thus, an author usually must decide whether to accept a pending offer without knowing what share of her more-preferred journals have considered her submission.
Authors respond with an assortment of gaming behaviors of their own. Because authors perceive that their submission is more likely to get the attention of their preferred journals if they request expedited review, many authors will submit to dozens or even hundreds of journals with which they have no intention of publishing. Many authors have said that they attempt to accumulate multiple offers from journals they consider undesirable in order to make their submission look more appealing to preferred journals. Authors regularly refuse offers from journals, even with no other offer in hand. And some authors have admitted in anonymous on-line fora that they have invented offers or fudged deadlines in order to generate attention or to put pressure on more-preferred journals. Scholastica has redesigned its interface to make offer and deadline fudging more difficult, suggesting that journals believed that this was a serious concern. If so, we doubt that Scholastica’s changes have made this practice impossible.
We should note that authors’ beliefs in the essentiality of an expedite request seems to be overstated. On average, responding journals state that only about 40% of their offers (13 of 31) were in response to an expedite request. The most competitive journals answering our survey report reading all or nearly all submissions, suggesting that expedite requests are often unnecessary to get at least an initial screening read.
Nonetheless, gaming behaviors by both sides of the publishing relationship seem to be mutually reinforcing. We have heard from both editors and authors who said that the aggressive behavior of their counterparts justifies their own behavior: “If the journal is going to play hardball, so am I.” Participants who do not engage in gaming sometimes report perceiving themselves as “suckers” who are failing because of their unwillingness to game, leading them to begin gaming themselves. Both these kinds of ethical unraveling are familiar to scholars of many other informal markets.
We are troubled by gaming for its own sake, not just because it undermines the quality signal of journal placement. Law faculty should model ethics — or, at a minimum, courtesy and professionalism — for their students. Fudging and dishonesty are not behaviors that law professors should engage in when dealing with students, period. While other kinds of gaming are not as obviously problematic, and may be individually rational for each participant, most of them are clearly anti-social in the sense that they impose costs on other actors who are unable to consent to the imposition of those costs. For example, we think the act of submitting an article for review should be understood to imply that the author has at least some minimal willingness to publish in the relevant journal. Submissions induce student editors to expend time and resources based on that understanding of good faith and fair dealing. And offers extended to and held by insincere submitters cannot be extended to others with a more genuine desire to publish in the offering journal.
We are not here to assign blame or to punish. Instead, our point is that the current system creates unavoidable pressure points, and these pressures in turn tend to reward gaming. Reform of some kind is clearly needed.
Student editors agree. Not one of our survey respondents disagreed with our suggestion that professors should abide by a code of best practices, and only a few opposed a similar code of conduct for editors. Nearly every one was ready to participate in efforts to implement significant reforms to the law journal system.
To give a more complete sense of student editor sentiment, we include here a few representative quotes:
“We think the submissions ‘dance’ tends to hurt all journals.”
“Expedite requests are unrealistic for us to honor; especially once we have hundreds of articles in our queue.”
“As a journal that sits just below the top 15 journals (currently ranked #[anon]) we are often used as leverage. Almost all of the articles we considered were immediately expedited to top 15 journals and despite best attempts to ‘sweeten’ the deal for authors, many of them went to other journals ranked higher or slightly higher.”
“We feel like we are at the mercy of short review windows for more competitive journals.”
Our own considered view is that best practices alone are inadequate. Professors already know that it is unethical to lie to student editors. A well-designed system should reward ethical behavior, not leave those who act ethically to feel like they are the suckers.
We also note that the current system is highly incompatible with significant faculty involvement in the article-selection process. Some commentators have urged us to find avenues for increasing the role of peer-review and other expert involvement in article selection. We agree that there should be more peer-reviewed venues, and that student-edited journals can and should incorporate peer review more extensively than they do now. But peer review is essentially impossible on a deadline of one week or less. Best practices will not change that.
In short, the current system has several extremely serious failings, and none of these failings can readily be changed with small tinkering at the margins. The failings are driven by the inevitable pressures of simultaneous submission and expedited review. No reform can offer meaningful change unless it addresses those facts.
II. Moderate Change: Limited Submission with Mandatory Acceptance
In our conversations so far amongst the Section and others, one proposal has emerged as preferable to the status quo for almost all discussants. In this model, the existing structure of article selection would remain in place, but authors would be limited in the number of journals to which they could submit at any one time. For example, authors might be able to submit to no more than ten or no more than twenty journals simultaneously. Advocates of this approach suggest that it would reduce or eliminate requests for expedited review. The argument appears to be that authors would submit to their more-preferred journals first, so that there would be limited need for expediting.
In addition, and crucially, authors would be obligated to accept the first offer they received. We call this limited submission with mandatory acceptance, or LSMA. The Section would not support limited submission without mandatory acceptance. LSMA eliminates the expedite process, which we see as lying at the heart of many current problems. Limited submission alone would not eliminate, and could well compound, much of the strategic behavior associated with expedites.
LSMA alone is in some ways a modest change from the status quo, but it can be used as the building block for additional changes. In particular, we recommend that it be combined with two additional modifications. One of these, the peer-review pool, we detail in the next Section.
A second key add-on would be to prevent journals from extending (and authors from accepting) offers in less than a reasonable time span from the time the submission was received. For discussion purposes we suggest that this reasonable span or “quiet period” be four weeks. In this way, authors’ most-preferred journals would not face the absurd time pressure that characterizes the current submission environment.
The quiet period clock would begin to run at the time a journal received an article. In this way, authors would be able to submit to their less-preferred journals as they received rejections from their most-preferred, without concern that the less-preferred journal would rush its consideration and extend an offer before the quiet period had elapsed for the author’s remaining most-preferred journals. At the same time, journals that expect to be less preferred will receive valuable information about author preferences. Journals may, for instance, wish to prioritize review of submissions they receive at the beginning of the submission period, knowing that there are not other journals with a head start on those manuscripts.
LSMA could also include other modifications, if these other features had widespread support. For example, it could include mandatory blinding of author information. We understand from most journals that have expressed a view that they presently would be opposed to blind review. In a world with less time pressure, however, journals would not need to rely as heavily on cues such as the author’s track record or institution, and so might be more willing to accept blind review.
Although some aspects of LSMA could be automated, others likely require law faculty involvement. It will be desirable to have neutral third parties available to ensure that authors do not reject offers, and that journals do not extend offers before expiration of their quiet period. We also anticipate that some automated features could be gamed (e.g., by submitting an identical manuscript under different titles), and here again the presence of arbiters will discourage such efforts.
Therefore, supervision of the submission process and its outcomes will be in hands of a committee to be initially organized by the Section on Scholarship. The Selection Committee once constituted shall develop procedures for selecting and staffing the Committee and securing resources appropriate to ensuring its continued success. The Committee shall maintain an ombudsman office to review and resolve any disputes about the administration of the system, and shall be empowered to take steps necessary to ensure its continued success. Available sanctions may include temporary suspension from access to the submission system.
Authors should also have the option to provide two additional pieces of information at the time of submission. Authors will have a drop-down menu in which they can choose their preferences about republication rights. A representative menu might include options such as “no preference; author must retain right of first republication; author prefers to retain right of first republication; other (specify in field).” The second added information will report author preferences for publication dates. Journals will be able to screen submissions by these categories.
Offers are binding unless good cause is shown. At the request of either party, the Selection Committee shall determine whether good cause has been shown in any case. The Committee will develop public guidance on what cases constitute good cause. In general, good cause must include circumstances that were unforeseen at the time of the initial match.
Authors and journals may still wish to negotiate over publication terms, including assignment of rights and timing of publication. We recognize that parties may be unable to bargain to agreement on these terms. Authors are expected to disclose any important preferences on rights and timing in advance. The Selection Committee will generally not find good cause to withdraw from a match based on publication timing or rights disagreements if the author did not choose to disclose author’s preferences at the time of submission.
III. Sweeping Changes: Matching
The Section recommends a twice-yearly Shapley matching system to be implemented via on-line platform. Participants make three nested decisions in the matching system. First, they decide which round or rounds to participate in. Then, within each round, they select a set of partners to be ranked. Thus authors would choose a set of journals in which they would be willing to publish, while journals choose a set of articles they would be willing to publish. Finally, each participant ranks their preferences within the set of acceptable options.
The platform host implements a standard matching algorithm designed to maximize the satisfaction of all expressed preferences. The algorithm is constructed as an iterative set of proposals: a student ‘‘proposes’’ to his top-ranked journal, asking for publication. The program accepts the proposal on behalf of the journal if that journal has an open spot and the author is among the list of acceptable candidates, retaining the right to reject it later if a better suitor arrives. If the proposal is rejected, the author proposes to the next journal on her list. If the proposal is accepted, another author repeats the process. If this second author is more preferred by the journal than the first author, the second displaces the first, and the first author proceeds as if rejected. Fig. One (copied from Nagarkar & Janis 2012) illustrates.
Matching assignments are binding. Authors and publishers must commit to being bound by match outcomes in order to participate in the system.
Shapley matching has several well-known properties (see Roth & Sotomayor 1992 for a comprehensive review). Most relevantly, it eliminates any incentives for strategic behavior. Participants have no reason to rank other than according to their true preferences (Nagarkar & Janis 2012). In addition, matching maximizes the satisfaction of those preferences, given the expressed preferences of the matching partners (Gale & Shapley 1962, Roth & Peranson 1999).
Each matching round will extend for ten weeks. Although exact dates can be negotiated, a representative set of dates would be February 7 to April 18 (Spring) and Aug 1 — Oct. 10 (Fall). These dates fairly closely match the current outer bounds of periods in which journals that read seasonally report that they now select articles. Journals would have the option of participating in Spring, Fall, or both.
[An alternative approach would be to split each season into two shorter rounds of five weeks each, with a one-week break in between. The rationale for two rounds in each season is primarily to allow journals to manage how many articles they must screen in any round. Journals that have historically been less attractive to authors are unlikely to want to read and rank several hundred articles that are likely to go to other journals. These journals may thus prefer to read only during the second round of each season. Historically, many journals in this position are “specialty” journals that have selected editors and articles later in each year than general-interest journals.
We acknowledge, however, that there is an important tradeoff in separating the season into two rounds. Of course, if there were only a single round, that round could be longer, perhaps as long as ten to eleven weeks. This would ease the time crunch on journal editors during the matching periods, albeit at the cost of extending the time that matching would likely consume. A longer review period would also make the optional peer-review pool (described below) much more viable. The Section welcomes views on whether on balance it would be better to have only a single round each season. ]
Authors submit their manuscript by the opening date of the round in which they seek a match. Authors post two versions, a “blinded” version and an identifiable one; each journal can choose whether they will view only blinded manuscripts or not. Authors also must enter their rankings by the opening date of each round.
In addition to selecting acceptable partners and rank preferences, authors also will have the option to provide two additional pieces of information. Authors will have a drop-down menu in which they can choose their preferences about republication rights. A representative menu might include options such as “no preference; author must retain right of first republication; author prefers to retain right of first republication; other (specify in field).” The second added information will report author preferences for publication dates. Journals will be able to screen submissions by these categories.
Ideally, authors will have the ability to save their rank and other preferences on the platform server, so that these do not need to be reentered from scratch each submissions season. These saved rankings of course would be confidential, although again we emphasize that there are almost never any strategic advantage to any author or journal from knowing the preferences of other participants.
Participating journals must submit rankings by the closing date of each round. With one key exception, journals cannot see author rankings until after they have submitted their choices (again, although there is no strategic reason counter-parties’ rankings should matter, they may be important psychologically). The exception is that journals will not see in their inbox any submission from authors who have not ranked that journal. In other words, journals do not and cannot read articles they cannot match with. Journals’ rankings are confidential and are not reported except in the case of the matched author-journal pair.
If participating journals find it too difficult to rank every article, the matching system could be simplified. Journals might assign each article a score, and then would be matched with random articles of identical score. With some programming sophistication, matching might improve slightly over random by ensuring a representative distribution of key factors, such as topic area and author gender.
Ideally, the platform will replicate the functionality of existing submission software in its ability to sort on key words and track submissions that have already been evaluated by the journal. Again, journals will be able to select whether to view the blinded or non-blinded versions of their submissions. Journals can sort on author preferences for rights or timing. For example, an author may indicate that she will only accept offers to publish within three months of acceptance, and journals may screen for and automatically mark for rejection authors with timing preferences incompatible with their feasible publication schedule.
An option is available to include a pooled peer-review component to the matching process, as we discuss in more detail below. We anticipate that journals would have to certify that a submission has met some first-level internal screening before the submission could be referred to the peer-review pool. Peer review referrals will likely have be decided relatively early in the matching period.
Supervision of the matching portal and its outcomes will be in hands of a committee to be initially organized by the Section on Scholarship. The Match Committee once constituted shall develop procedures for selecting and staffing the Committee and securing resources appropriate to ensuring its continued success. The Match Committee shall maintain an ombudsman office to review and resolve any disputes about the administration of the system, and shall be empowered to take steps necessary to ensure its continued success. Available sanctions may include temporary suspension from access to the matching system.
Match assignments are binding unless good cause is shown. At the request of either party, the Match Committee shall determine whether good cause has been shown in any case. The Committee will develop public guidance on what cases constitute good cause. In general, good cause must include circumstances that were unforeseen at the time of the initial match.
Authors and journals may still wish to negotiate over publication terms, including assignment of rights and timing of publication. We recognize that parties may be unable to bargain to agreement on these terms. Authors are expected to disclose any important preferences on rights and timing in advance. The Match Committee will generally not find good cause to withdraw from a match based on publication timing or rights disagreements if the author did not choose to disclose author’s preferences at the time of match.
Journals will be free to approach authors directly, as in the traditional “essay” publication model. Journals may also hold individual calls for papers for special issues, such as symposia or book-review issues (journals should be discouraged from holding these during the matching periods). A journal may not, however, individually solicit during the matching period an article or essay posted for match, and authors whose articles or essays are entered in the matching system may not accept a solicitation offer made with respect to any papers currently posted for matching.
We believe the matching proposal we have described is better for almost every stakeholder in every situation, although we recognize that effective communication and outreach will be necessary to convince participants that this is true. Much of this benefit comes from eliminating the pernicious effects of the expedite system. Authors will no longer have to wrestle with whether to submit to journals they would prefer not to publish with, and journals will no longer waste time reading submissions from authors who will not publish with them.
Overall, we expect that matching will dramatically reduce the number of articles most journals will have to evaluate. According to our survey data, journals outside the fifty to seven-five most popular receive under one thousand submissions. This number would likely fall considerably, as many authors will no longer use these journals in order to attract the attention of journals they prefer more highly. Authors who remain in the pool for these journals will be relatively more likely to accept than under the present system, making the editors’ reading time more efficient.
We think the biggest “winners” among journals will be those just outside the fifteen or twenty most author-desired journals. As we described earlier, our survey suggests these journals presently receive thousands of submissions annually, make a relatively large number of offers for each acceptance, and, because the expedite requests they receive are typically on sort deadlines, make most of their decisions under extreme time pressure. Matching would probably not appreciably reduce the volume of submissions these journals receive, and may increase the number of articles to which they ultimately give close attention. In exchange what they would get is time and relief from several attendant pressures. Journals would have the full period of the match to assess the complete pool of submissions and prioritize which to screen and read. They would not be hostage to the competitively-set deadlines of other journals and would not have to police suspected gaming behavior by authors.
Matching may not align with the current screening practices at some journals. For instance, some journals report that nearly all article selection is conducted by one or two editors, typically the editor-in-chief and chief articles editor. That will not be practical under a matching system. Many of the most desired journals screen three thousand or more articles a year. That is, they not only receive thousands of articles in their in-box, but also actually respond to them, often without receiving an expedite request. Other journals can learn from the methods these journals employ. While this will involve some transition costs, we do not see any inherent value in any given screening method, and do not view the likely obsolescence of some as an argument against matching.
Admittedly, matching can make it more difficult for journals to perfectly control the mix of articles they select. A journal might prefer a distribution of con law and contracts articles, or a representative selection of male and female authors. A matching system cannot guarantee the journal won’t be matched with, say, eight tax articles. But we think the present system offers only limited opportunities for these kinds of balance in any event. At best, journals now are highly path-dependent: to achieve balance, they must reject otherwise desirable articles for too closely duplicating some earlier selection. This preference for the early applicant is another kind of hidden bias that savvy “insider” authors are aware of but others are not.
A critical question for the success of the matching regime is whether the most author-preferred journals will participate. We believe that eliminating expedite requests and their attendant deadlines will provide an enormous benefit to this group of journals. Indeed, these journals were the driving force behind a (largely unsuccessful) effort to organize a pledge to require at least one week for all expedite requests. Admittedly, however, these journals will also be losing the large screening subsidy less-desired journals formerly provided them. As we noted, however, we believe most of these journals already were committed to screening most submissions regardless of expedite.
Several journals in this category currently employ “exclusive submission” windows, ostensibly in order to provide for more time in which they can complete lengthy review processes that may include peer review. Exclusive submissions can contribute to the unraveling of a matching system, add to the complexity and administrative burdens for authors, and may undermine the sense of fair play that the matching system aims for.
Thus, if necessary, the Committee will have the authority to impose an exclusivity requirement for use of the matching system. That is, the Committee may choose to bar articles from the matching system if the article is submitted to journals using any other method. This would tend to strongly discourage opt-outs or add-on “exclusive submission windows.”
Comparison of LSMA and Matching
Members of the Section and other commenters generally agree that, if achievable, either LSMA or matching would improve considerably on the status quo. Some prefer one to the other, or think that one is more likely to win general acceptance. We invite robust discussion and debate on whether one is preferable to the other, without losing sight of the key principle that the status quo is not preferable to either.
To help begin that discussion, we offer some preliminary thoughts on the relative advantages of the two systems. We first note, though, that LSMA and matching are quite similar in many respects. LSMA resembles matching in slow motion. Authors would construct a subjective ranking of journals, and one approach to submission strategy would be to simply submit sequentially according to that subjective ranking, twenty journals at a time, until receiving an offer of publication.
One potential advantage of LSMA over matching is that it might economize on reading time for less-preferred journals. While under a matching system a journal cannot know whether it will be competitive for a given submission, with LSMA authors would not need to submit to less-preferred journals once they had found a more-preferred match. We note that this is true only under mandatory acceptance; with limited submissions alone, authors could well strategically submit to their less-preferred journal.
At the same time, matching also likely has some relative advantages over LSMA. First, LSMA takes much more time for matches to complete. Assuming that most authors begin submitting at the journals most authors prefer, and that it is infeasible for most authors to receive responses in fewer than two weeks (or four, under our proposed “quiet period”), journals that most authors would rank one hundredth or lower will not receive their first submissions until ten weeks or more after submission season begins. If most-preferred journals begin reading in February, these less-preferred journals will be obliged to read in April and May, or in the summer. None of these are times journals have historically desired to consider submissions. Alternately, journals could all read at roughly the same times as at present, but less-preferred journals would be limited to articles that had failed to match in earlier submission seasons. This implies that time to acceptance for most authors would be considerably longer than under matching — six months to a year or more, in some cases, rather than ten weeks.
Next, in some instantiations LSMA could worsen, rather than alleviating, the time pressure that is one of the least appealing features of the status quo. With mandatory acceptance, the journal that can extend an offer first is guaranteed of landing their first-choice submission. And we expect that the most-preferred journals would still be likely to receive thousands of submissions. While they would not face any one-day expedites, these journals would have far less time per article, on average, than under the present system. They would in effect have to select their entire set of articles in something like a two-week period — probably quite a bit less for articles for which they expect the most competition — without any signal of quality that might be offered by expedite requests. Again, it is this time pressure that we believe reduces students’ depth of engagement with submissions, encourages reliance on proxies for quality, and makes faculty contributions to the selection process difficult.
Further, we think that this situation would be so undesirable for these most-affected journals that it would be difficult to secure their agreement to participate. Even if it were possible for AALS and its member schools to mandate or coerce participation, we think that it is undesirable for student editors to feel that law faculty have worked systematically to disadvantage them.
The importance of timing also means that LSMA does not eliminate incentives for gaming or premiums for insider knowledge. To the contrary, gaming behavior could well increase under LSMA, likely in ways that the Section cannot now imagine. When timing is of the essence, authors and journals would both have reason to strategize carefully about the signals they send to the other party.
For instance, imagine that Author believes she is very unlikely to receive an offer from her most-preferred journal. She might replace this journal in her list of twenty initial submissions with Journal X, which she ranks twenty-first. By doing so, she hopes to grab the attention of Journal X at a moment when relatively fewer authors are submitting to Journal X. Observing this, Journal X may wish to read author’s submission especially quickly, in order to exploit the chance to land an article it might otherwise not have had a chance to read — or, alternately, the journal might draw a negative signal from the fact author didn’t believe author would receive offers from journals most other authors prefer to X. And so on.
As under the present situation, then, LSMA would result in many matches that occur because of luck. Authors who happen to submit to the right place at the right time, or who hit on a successful strategy, might be more satisfied with their results than others. This reduces the quality of the placement signal, while privileging those in a position to understand how to “work the system” to their best advantage.
Further, as we will describe more below, a match system may be easier to mesh with peer review. Because of time constraints, implementing peer review under an LSMA system may require that peer review be undertaken before the author submits. This gives rise to two modest disadvantages. One is simply that authors must wait longer between when they finish their manuscript and when it can be accepted (or, put another way, an author who wants to submit in August must finish her manuscript in May or June). Another is that journals cannot serve as an initial screening mechanism for the peer review system, likely increasing the volume of reviews.
LSMA with a quiet period mitigates some of these disadvantages but may aggravate others. Quiet periods, as we have said already, reduce time pressure for most-preferred journals, and might enable abbreviated peer review. They also help less-preferred journals to prioritize their review on submissions more likely to result in a match. But these gains do come at the cost of a more complex strategic environment in which both authors and journals send and receive information through their choice of timing. Article selection, we think, should reflect article quality and not skill at submission strategy.
LSMA also avoids a situation in which journals are aware that authors have ranked them according to a strict numerical hierarchy. But under our matching proposal, journals would never see authors’ rankings. Further, an LSMA system is, as we have said, in many ways just matching in slow motion. Under LSMA, journals would similarly be aware that authors must have a subjective hierarchy they have used to select the twenty journals for exclusive review. Moreover, nothing AALS does can prevent third-party ranking organizations, such as the Washington & Lee Law Library, from “ranking” journals in strict numerical order. We think matching’s incremental contribution to any harms from the existence of an explicit hierarchy are therefore modest.
IV. Optional Peer-Review Pool
Peer review is more feasible in the absence of one-week deadlines. LSMA or matching thus can be combined, if there is sufficient interest among editors and faculty, with an optional pooled peer-review system. We would strongly urge member institutions to adopt expanded peer review, but we would be willing to go forward with other recommendations even if they do not. We see two main alternative arrangements for the pool.
Several elements would be common to either approach. A Centralized Peer Review organization would be created for all law review submissions. The CPR would aim to solicit one or two reviews for each article it receives. Peer reviewers would be asked to provide substantive commentary on the piece, suggesting important changes/issues. They would also be asked to rank the article as falling within one of several categories, such as (i) among the very best pieces in its field, (ii) excellent work publishable in all but the most selective journals, (iii) excellent work best suited for specialty-journal audience, (iv) not worthy of publication in present format, and so on. Authors would receive copies of all reviews, with the exception of the ultimate “ranking” recommendation; reviewers would be anonymous.
Peer reviews would be made available to participating law reviews directly by the CPR System one month after manuscripts were submitted to the System. In order to be a participating law review, and thus to access the peer reviews, a law review would need to commit to a policy of (i) giving stronger consideration in the screening stage to articles that have been subject to peer review, and (ii) having a strong preference for ultimately selecting articles that have been subjected to peer review through this system. Initial commitments to this system would need to be secured from a critical mass of top law reviews. Law reviews should have a substantial interest in agreeing to this system, as they would retain control over ultimate decisions, but their decisions on individual manuscripts and their screening of drafts would be better informed because they would include the peer reviews. Law reviews would add a footnote to articles selected through this process, stating that they had been subjected to peer review.
The Centralized Peer Review System would obviously need a strong and very large group of initial peer reviewers. But many faculty could probably be convinced to participate if they were assured that they would only be asked to review a set number of times in a year, had the option of turning down a peer review that came at a bad time, and were given a month to complete the review. Additionally, authors whose own work was submitted to the System could be required or encouraged to serve as peer reviewers in the future.
Reviewers would indicate their areas of expertise, and authors would indicate the subject matter of their article. Articles would be matched to faculty by CPR staff. Conceivably, CPR could also provide an additional screen in which some articles were “desk rejected” and not sent out for peer review. CPR would provide a brief notice of this determination to reviewing journals. CPR staff could be drawn in part from full-time faculty as well as from fellows and visiting assistant professors.
Resubmissions could be permitted. For example, an author could be allowed to re-submit their work for a second round of peer review in a subsequent matching cycle, along with an explanation of (i) how, if at all, they have attempted to respond to the peer reviewers’ concerns, or (ii) refuting/responding to the peer reviewers’ criticisms.
Within the overall framework, there are two alternatives for how articles come to be submitted to CPR. In the “journal referral” model, peer review is conducted during the matching or LSMA period. Each journal could determine whether it would find peer review beneficial. If a critical mass of journals so indicate — the number could be low, such as three or five, or higher, such as ten — then the article is sent to CPR staff for assignment. Referrals would have to be made by week four of the matching period. CPR staff would have one week to find a reviewer, and four weeks to obtain review. Journals would then receive reviews back no later than the beginning of the last week of the matching period.
An alternative approach would allow peer reviews to be author-initiated. In this model, reviews could be even more leisurely, as the review process could be conducted in between the matching periods. With this extra time, authors could be given the opportunity to write a brief responsive memo addressing referees’ concerns. The downside to this approach, however, is that if peer review were advantageous most authors would have little reason to eschew it. This could lead to significant burdens on the system. It would also extend the time between completion of a draft and acceptance for publication.
V. Complementary Best Practices
In addition to the reforms described above, the Section recommends that AALS promulgate best practices for law journals authors and editors. We recommend that adoption of these best practices and training on them be a portion of the site evaluation process at each accredited institution. Specification of the exact contents of the best practices will be deferred to a future document.
In general, best practices should exemplify mutual respect and regard on the part of authors, editors, and audience. Partners in the publication process are not engaged in a game in which each side competes for advantage, but instead in a shared enterprise whose goal is to inform, persuade, and move legal systems towards justice. Any submission system, including those detailed here, will benefit when participants all work together in this spirit.
Best practices should also embody norms evident in other scholarly disciplines. Currently, only a handful of journals that we are aware of have specific policies regarding disclosures of financial support or other potential conflicts of interest. Similarly, we do not know of a significant number of official policies regarding the use of human subjects.
VI. Other Alternatives
In recent years, commentators have offered other thoughtful and potentially helpful reform proposals. It may be useful to say a few words about why the Section opted not to pursue these alternatives.
Some avenues for change would solve the expedited review problem at the cost of essentially eliminating the quality signal of law journal placement. For instance, one commentator suggests that faculty should publish only in their home journals. For the reasons we described in Part I.A., we reject these kinds of solutions.
Other proposals would remove students from the article-selection process, usually to be replaced by peer review and (presumably) exclusive submissions. We don’t believe legal academia is currently structured in a way conducive to widespread peer review. Most viable peer review systems rely in good measure on graduate students who are essentially academics in training, already deeply versed in the relevant literature and in cutting edge theory and methodology. Legal academia now has very few of these individuals. There would also be an enormous transition problem, as a relatively small share of current legal academics have experience with the norms and practices of peer review.
Further, successful peer review journals in other disciplines consume significant financial resources. An effective journal has full-time staff, primary editors who are bought out of some teaching or administrative responsibility (or granted relief as an in-kind contribution by their institution), and a substantial editorial board. It is unclear whether the legal academy is prepared to increase its investment in publishing in this way. An alternate route would be dramatic consolidation of legal publishing, with the same resources spread over considerably fewer journals. That, too, has quite evident tradeoffs.
And, as we have said, there are important virtues of simultaneous submission. We believe legal publishing is more vibrant and more successful if authors can choose between exclusively submitted peer-review venues — with their typically very long editorial processes — and simultaneously submitted journals. Again, we would like to see AALS commit to encouraging and supporting more peer-reviewed journals, so that this option is available to more authors. But in our view, and the views of most of those we’ve heard from, there remains a major role for student-edited journals.
This is a discussion draft. The Section invites responses from any interested readers. One way to respond is to attend the annual meeting of the AALS Section on Scholarship, which this year features a roundtable discussion of this draft. Our meeting will be held Jan. 5 from 3:30 to 5:15 in Maryland Suite B at the Marriott Wardman Park, https://memberaccess.aals.org/eweb/DynamicPage.aspx?webcode=SesDetails&ses_key=4a70cc2c-0b59-4dca-b469-158f4603a5a7.
You can also submit written comments before, during, or after the meeting via email to Brian Galle, chair of the section’s advisory committee on law journal reform. His email is brian.galle at Georgetown.edu. Unless you direct us otherwise, your comments may be read aloud and discussed at the roundtable.
 Of course, readers should keep in mind that respondent editors may those that were particularly motivated to respond and so may not be representative of the experience of all editors.
 We note that if authors are not held to this standard, they risk creating a market for lemons in which students will not be willing to review submissions they suspect of being insincere, there will be fewer sincere high-quality submissions, and so on.
 In extremely rare situations — about one time in one thousand–a journal can improve over pure-preference rankings by systematically omitting lower-preferred candidates. But unless the journal has perfect information about author preferences, this is also a highly risky strategy. Therefore we believe it is accurate to state that the matching system eliminates any realistic gaming incentives.
 That is, some journals in this “upper middle” part of the distribution may presently use strategic behavior, such as short deadline offers at the beginning of spring break, to minimize the number of articles they lose to competitors. In a matching system, journals would not be able to compete on this basis. Some journals may be less successful in competing on alternative grounds, and if so this would likely increase the number of submissions they would have to rank in order to fill their volume.
 If this were the new regime, we are unsure what would happen to less-preferred journals during the transition year the new policy is adopted. Possibly those journals would receive few or zero submissions in the first year of the policy.
 Comparable pressures would exist if acceptances were not mandatory. For example, authors hoping to “shop” an initial offer will want rejections as soon as possible, so that they can quickly swap out the rejecting journal for a new reader. Journals would be incentivized to offer authors guarantees of fast decisions to attract submissions.
 Of course, some journals now offer an exclusive-submission window. This is not comparable to what would happen under LSMA. Exclusive-submission windows typically fall in or very close to the traditional submission season, so that authors face a trade-off when deciding whether to submit exclusively. Thus, the number of exclusive submissions is a small fraction of the total number of submissions journals would receive under LSMA.
 The Section is grateful for initial conceptual work by Dan Schwarcz of the University of Minnesota Law School.