Improving Law School “Transparency”
Monday, August 12, 2013
Author: Jeffrey Evans Stake
Download a PDF of this post
Jeffrey Evans Stake
Given their central role in the training and socialization of lawyers, law schools are uniquely obliged to present themselves honestly to the public and to their prospective students. In recent years, law schools have failed to live up to that simple standard. As Ben Trachtenberg documents in detail, some law schools have lied or deliberately presented misleading information. More broadly, all law schools have followed the ABA mandate to report employment statistics that, until recently, were defined in ways that would have surprised many law school applicants. The ABA has taken steps to improve the flow of information from law schools to the public. It has punished perpetrators and reformed reporting rules. What more should be done?
Prof. Trachtenberg argues for increasing the penalties for individuals involved in deceptive representations. One of the admirable attributes of Trachtenberg’s approach is that it focuses punishment on the key actors. When law schools lie, they set a terrible example for law students and lawyers, undermining the norms of the profession. If law schools, the very places that teach professional responsibility and legal ethics, somehow see fit to lie when the benefits outweigh the costs, law students and lawyers may take to heart the apparent lesson that it is acceptable to deceive in order to advance the interests of one’s self or one’s employer or one’s client. Beyond that, it is not implausible that lying by law schools erodes public confidence in the justice and integrity of the legal system and undermines its ability to coordinate behavior by setting social norms. In any case, the lying in which law schools have engaged is plainly deplorable behavior deserving severe punishment. But the lies of law schools start with individual employees; it is their behavior that needs to be constrained. If society punishes only the schools and leaves the punishment of individuals up to the schools, the penalties that the individuals face, typically loss of job, may be inadequate to create an efficient system of deterrence. Trachtenberg’s approach avoids that limitation by punishing the liars directly and in ways that are not available when schools do the punishing.
On the flip side, Trachtenberg’s proposal avoids punishing the innocent, as can happen when the ABA sanctions schools or when peers kick them in the reputation. After lies at the University of Illinois College of Law came to light, the U.S. News and World Report (USN&WR) peer assessment score for Illinois fell from 3.5 to 3.1. That drop contributed to Illinois’s plummet in overall rank from 23 in 2012 to 35 in 2013 and to 47 in 2014. If the teachers and administrators evaluating Illinois were merely responding to the new LSAT and UGPA information and changing their assessment accordingly, then the new reputation score should not be seen as inflicting punishment on Illinois but rather as putting Illinois where it should have been all along. But if the peers lowered their reputation assessments beyond what was justified by the new statistics, then the new reputation score includes an element of punishment. If that is in fact what happened, the peers have punished the victims, the very students who were deceived into attending Illinois by the false numbers.
Despite these, and other, advantages of Trachtenberg’s approach, I am not yet convinced that punishing lawyers in law school administration will enhance what is often called “transparency,” which I take in this context to be a fair presentation of the facts. I do not doubt that an increase in punishment will diminish the likelihood that any lawyer will make a misleading statement. However, as Trachtenberg points out, many of a law school’s reporting tasks that are now being performed by lawyers could be done by non-lawyers. If the increase in the threat of punishment makes the expected cost substantially greater for lawyers than for non-lawyers, jobs that need not be assigned to a lawyer will become relatively more attractive to non-lawyers. Over time, such tasks will tend to be allocated to non-lawyers whose behavior will not be subject to bar discipline and might even be less likely to conform to desired standards. It is not a certainty that the gains from improving the behavior of lawyers outweigh the losses from shifting some of those tasks to non-lawyers.
There is another problem as well. Subjecting lawyers to personal punishment for failure to alert bar authorities might align their interests in secrecy with the interests of their employer. If employees face punishment, they have more to gain from keeping information from the ABA, which will reduce the chances that they will blow the whistle on their employers. Misreporting by law schools might be more likely to remain hidden if the employees think they could be punished for failing to alert authorities at an earlier date.
Finally, although there is no doubt that some law schools have deliberately lied in the past, many (if not almost all) schools have tried to report honestly. And now that Illinois and Villanova have been punished by the ABA and their reputations have been blackened, the chances of misreporting must have fallen still further. The ABA can increase the odds of truthful reporting by making sure that the standards for reporting are clear and practicable. Recently, the ABA has taken important steps to improve the utility of the numbers being reported by dividing the employment into groups. If that dis-aggregation is adequate for law schools to honestly promote their programs and the lines drawn are workable, schools should be happy to comply and will not need to make up their own categories of employment. With regard to the accuracy of information reported by almost all law schools, tailoring and clarification of the ABA standards will be more important than increasing the personal punishment lawyers face for being involved in misreporting.
My skepticism about the effectiveness of punishing members of the bar should not be interpreted to mean that I condone any ethical lapses of law schools and their employees in their releasing of information. More can and should be done to reduce the harms from law school misreporting. To examine this issue, it might be helpful to divide the harms into three groups. The first kind of harm, perhaps the most important, is the harm to social expectations and professional norms that attends deliberate fabrications and falsifications. I mentioned this above, and will not elaborate any further here.
The second kind of harm is the harm befalling individual students when law schools, whether intentionally or not, publish inaccurate information. For example, a lie or mistake in reporting LSAT and UGPA statistics might have increased the rank of a school in the USN&WR ranking, which might in turn have misled individual students into matriculating at that law school. Those students may have been harmed by the misinformation. What is that harm? I do not believe that the overall educational and employment differences between the law school chosen and the school forsaken in each case would have been large, if indeed they existed at all. This is so because inaccurate information about LSATs and UGPAs in most cases would not make a large change in rank, and there is little difference in the overall quality of schools ranked closely by USN&WR. As far as we can tell, no school has lied on many criteria, and until such a school is identified, we can be reasonably confident that few schools if any have lied their way to a rank far above their true quality. The lies and mistakes that have been made have probably made little difference to the quality of legal education received by deceived students.
But a loss of educational opportunity is only one of the harms students could have suffered from misreported statistics. By increasing the rank of some schools, bad LSAT and UGPA numbers may have caused some students to pay more to attend law school than they would have otherwise paid. For example, a student might have been offered a scholarship of $100,000 by school A (rank 40) and a scholarship of $60,000 by school B (rank 38), where both schools have the same nominal tuition. The student might have then chosen to attend school B, based on its higher rank. But suppose also that school B would have been ranked 42 if USN&WR had known its real numbers. Had this student known that, he would have chosen school A and paid $40,000 less for law school. In such cases, the difference in price represents a clear loss to the student.
Is this realistic? Are students willing to pay a higher price to attend a school with higher rank? Anyone in law school admissions would probably say, yes. And the applicants themselves report the same thing. Brian Broughman and I are analyzing the importance of price compared to rank to students choosing between law schools. In our preliminary analysis, we have found that price and rank both matter, but that rank matters more and cost is only a secondary consideration for most students choosing a law school. Of all the participating applicants in our study who had a choice between two schools and who reported their preference, over sixty percent attended the highest rank school at which they were accepted, compared to only twenty-six percent who chose in favor of the law school with the lowest cost. This sensitivity to rank should come as no surprise. After all, rankings sell magazines and it would make no sense for students to buy the magazines and carry them around to forums and then ignore that information. These preliminary findings are consistent with a June 2012 Kaplan Test Prep survey in which thirty-two percent of the respondents said that rank was the most important factor in their decision and, in response to a separate question, eighty-six percent said that rank was very important or somewhat important. According to another study by Amanda Griffith and Kevin Rask, students are also sensitive to rank when they choose an undergraduate college and this sensitivity is independent of other measures of quality. Given that applicants favor rank over price more often than otherwise, it seems likely that in a few of these cases the higher rank school chosen would have been the lower rank school if it had reported accurate numbers.
For the students in those situations, the financial losses may have been large, due in part to the rankings themselves. Over the last two decades, in order to improve their position in USN&WR, law schools have attempted to attract students with high LSATs and UGPAs by dramatically increasing scholarship grants to the point that many students have been offered a scholarship of $45,000 per year or more. At the same time, many schools offer next to nothing to other students. The result is that any given student may receive wildly differing scholarship offers, and of course tuition itself also varies across schools. According to our preliminary analysis based on applicant reports, some students gave up tens of thousands of dollars in net tuition to buy into a school ranked higher by USN&WR. Twenty-five percent of the participants gave up more than $50,000 to attend a higher rank school, and nine percent sacrificed more than $100,000 to attend a school of higher rank. At the extreme, three percent gave up $20,000 per rank place to go to a law school that USN&WR said was better. Perhaps law schools and the responsible employees should be liable for the direct financial losses of students that acted in reliance on their misrepresentations. An award of damages would shift losses from the deceived students to law schools and their employees, creating an additional disincentive for lying.
When the false number is a published employment rate, a more direct argument might be made that the inaccurate number misled a student. About five percent of the matriculants in our study chose to attend the school with the highest reported employment rate over a law school that was less expensive or higher ranked. Similarly, eight percent of the students in the Kaplan survey said that employment rate was the most important factor, and others respondents probably also considered it important. Perhaps some of these students paid extra to attend a school with a higher employment rate when in fact the school had a lower rate than a school the student chose not to attend. If so, they might deserve compensation for acting to their detriment in reliance on a fabricated employment figure. But it will be hard for them to show damages. The choice of school often makes no big difference to subsequent employment. Some students obtain jobs and are none the worse for having attended a school with a lower employment rate. Some students do not find jobs for a while, but would not have done so after attending the other school either. The students harmed, of course, are those in the middle, who failed to find jobs, but who would have landed jobs if they had gone to the school with the higher true employment rate. This must have occurred at least a few times, but it is hard to prove in any individual case. Proof is especially difficult because employment rates are positively correlated with LSAT medians and thus the school with higher employment would generally also have a more competitive classroom environment for students in it. For that reason, the students who did not obtain employment will be in the difficult position of showing that they would not have ended up lower in the class at the school with higher employment. To put it another way, it is not clear that employment prospects for any given student improve by attending a school with a higher employment rate.
Students may have suffered another injury, however, from misleading employment numbers. As in the case of inaccurate LSAT and UGPA medians, students may have given up scholarship money to attend a school that looks like it has a better employment rate. This loss from paying a higher price for purportedly higher employment odds is easily quantified for those students who can show that they chose their school based on employment numbers. The hard part is showing that they did in fact choose based upon published employment figures. However, if they can show that they turned down a school with higher rank and lower cost for a school with higher published (but not actual) employment, they should have a claim for damages in the amount of the difference in net tuition.
So far, the discussion has been about schools harming students by reporting inaccurate numbers to the ABA, where the discrepancies end up making a difference to student choice, either on the basis of the numbers themselves or on the basis of USN&WR rank. Next, consider the harm done when nearly all schools reported misleading numbers, as the ABA instructed them to do by defining graduates as employed if they had any sort of full-time job, even if it was wholly unrelated to the training received in law school. It is not clear whether any law schools gained a ranking advantage from the ABA’s definition of employment. The reporting system was defective, but all schools were evaluated on that same defective system. As long as employment as defined by the ABA correlated highly with the kind of employment that students cared about, the fact that ABA-defined employment was not law-related employment had no effect on the rankings and little effect on student choice among law schools. If the correlation was insufficient and some schools were ranked too high as a result, the harms discussed above might obtain, but at least some of the blame for that should be placed on the ABA.
Even if no school’s rank was changed by the ABA’s reporting rules, the numbers reported under those rules may have done some harm. Which law school to attend was not the only important choice the applicants faced. They also had to decide whether to go to law school at all. By inflating the employment numbers, the ABA reporting requirements probably led some students to go to law school who otherwise would not have attended. Assuming that some students were enticed to law school by the misleading employment numbers, were they harmed? That is not easy to determine. Of course, they expended time and effort, and money too; law school does impose costs. The question is whether the outcome is worth those costs. A draft of a study presented at the 2013 meeting of the American Law and Economics Association indicates that law school is not a bad deal for the majority of law students. Michael Simkovic and Frank McIntyre estimated that the mean pre-tax lifetime value of a law degree is in the neighborhood of one million dollars. Even at the 25th percentile, the earnings attributable to the law degree justified the expense of law school, compared to not going to law school. They also found that the law degree premium has varied over the years, and that the current premium is within historical norms. It is important to note that the study did not address the cost of forgoing other graduate educational opportunities, and had limited controls for ability. Nevertheless, the study casts a skeptical eye on the claim that law school pays off only for those attending the top 14 schools.
It is possible that for most students law school was not a waste of their money or time, even if they were drawn to school by overly optimistic employment numbers. But it is also worth considering whether inflated employment numbers did in fact cause over-attendance at law schools. In our study, across all participants, Broughman and I did not find that employment numbers had a substantial effect on choice of law school. One might think that this would mean that employment figures could not have drawn many to law school, but that does not follow. Students could pay attention to employment percentages when choosing whether to go to law school but then pay no attention to employment in choosing which law school. For a number of years, many schools published nine-month employment rates above 90 percent. A potential student might not care about the difference between 91 and 94 percent employment, but might be drawn to law training by the fact that both numbers are in the nineties. In addition, there is a group of students that might have cared about employment numbers in both contexts, whether to attend law school and which school to attend. Students with low LSAT or UGPA numbers are often denied admission by schools of high rank. Broughman and I found that students choosing between schools of lower rank were less sensitive to rank and more sensitive to employment rate than students choosing between schools of higher rank. The students choosing between schools of lower rank might also be the students most easily dissuaded from going to law school at all. So, it is possible that published employment numbers made a difference to that decision.
Were large numbers of students influenced in this way? Each of the lines in the two figures shows a change over time, compared to the baseline in 1985. The wavy line is the number of applicants to law schools in the US. Why does it fluctuate so dramatically? The supply of potential applicants to law school does not explain it. From Figure 1, we see that the number of BA degrees conferred by American schools each year has increased steadily, with essentially none of the variation seen in the number of applicants. The correlation between JD applicants and BA degrees from 1985 to 2012 is .07, and from 1999 to 2012 is -.14. Thus, since 1999, the number of actual JD applicants has not tracked the number of potential American applicants. Switching attention from the potential supply of students to the demand for law graduates, note from Figure 1 that the total population has increased fairly steadily, which ought to have increased demand for legal services, but the number of applicants has not increased in the same way. The correlation between JD applicants and US population from 1985 to 2012 is .10, and from 1999 to 2012 is -.03. Of more immediate relevance, observe from Figure 2 that employment rates for law school graduates have a varying relationship with applications to law school depending on the time period chosen. The correlation between application rate and employment rate from 1985 to 2012 is -.37, and from 1995 to 2012 is essentially zero. However, if we limit our view to what has happened since 1999, the correlation from 1999 to 2012 is .28. Of course employment rates only tell a part of the story, and potential applicants to law school should care about salary as well. According to Simkovic and McIntyre, however, the premium to be gained from a law degree is about the same in 2011 as it was in 2006, 2001, and 1996, so the variation in applicants over time seems not to be closely tethered to the potential value of a JD degree in the market.
At the present time, it is unclear how much weight applicants place on various factors when they make their choices between law schools. Underlying that uncertainty is a deeper uncertainty about the value of a law school education. We do not know how much harm a student suffers by going to one law school instead of attending another, or by pursuing law instead of some other profession. But those uncertainties provide no excuse for the failure of law schools to present accurate information to applicants and the public, i.e., to paint a fair picture of themselves. Justice and incentives for accurate reporting would be served by awarding students compensation for the direct financial losses they have suffered as a result of a law school’s publication of inaccurate information.
Download article here (.pdf)
Preferred Citation Format: Jeffrey Evans Stake, Improving Law School “Transparency”, 5 Neb. L. Rev. Bull. 9 (2013), http://lawreviewbulletin.unl.edu/?p=1120 .
0. Robert A. Lucas Chair of Law, Indiana University Maurer School of Law. I thank Brian Broughman, Eric Rasmusen, and Michael Alexeev for helpful comments. And I thank Ben Trachtenberg for inviting me to comment on this important issue. ↑
1. See Ben Trachtenberg, Law School Marketing and Legal Ethics, 81 Neb. L. Rev. 866 (2013).↑
2. I doubt that the Illinois drop in peer assessment can be explained purely as a matter of the actual quality of the University of Illinois College of Law as a research or teaching organization. However, information regarding the Law School Admission Test (“LSAT”) is probably a component of peer evaluations. Michael Alexeev and I have found that peer scores seem to respond to changes in LSAT’s. See Jeffrey Evans Stake & Michael Alexeev, Who Responds to U.S. News & World Report’s Law School Rankings? (Ind. Univ. Mauer Sch. of Law Legal Studies Res. Paper Series, Research Paper No. 55), available at http://ssrn.com/abstract=913427. The median LSAT for Illinois was 167 (tied for 16th with 7 other schools) in 2012, and 163 in 2013 (tied for 38th with 8 other schools). The peer assessment for Illinois was 3.5 (tied for 22nd) in 2012, and 3.1 in 2013 (tied for 39th). (The years in this case are USN&WR years, which are a year ahead of the year of publication.) Since the peer rank did not drop as much as the LSAT median rank, these numbers are ambiguous as to whether peers punished Illinois or merely considered LSAT numbers to be that important. (I am not suggesting that, in this case, the peers responded to the LSAT information published in USN&WR because they do not publish the medians and because the peer assessments dropped before USN&WR published the new LSAT numbers. If peers were responding to the change in LSAT’s, they would have to have been responding to other reports of the degree of cheating).↑
3. Trachtenberg, supra note 2, at 920.↑
4. I am writing here of the facts presented to the ABA. I am less confident about law school honesty in other contexts. For example, a number of law schools have enticed students with renewable scholarships without telling the students that it was impossible for all of them to do well enough to renew their scholarships. This is, in my view, an unethical omission.↑
5. A defendant school might argue that the student got what he paid for, namely rank, and not the qualities that rank is supposed to summarize. However, the student might be able to convince the finder of fact that he was attempting to buy the quality indicated by rank, not mere rank itself.↑
6. Brian Broughman & Jeffrey Evans Stake, Law School Tuition: How Much Does Price Affect Choice of School? (work-in-progress as of July 22, 2013) (available soon on SSRN). Our initial draft is based on student- reported data presented on LawSchoolNumbers.com for students matriculating between 2006 and 2012. At this stage, our findings are only preliminary because we intend to extend the database to include entries from the applicants in 2013.↑
7. Press Release, Kaplan Test Prep, Kaplan Test Prep Survey: Despite an Uncertain Employment Landscape, Law School Applicants Still Consider School Rankings Far More Important than Job Placement Rates When Deciding Where to Apply (June 19, 2012) (available at http://press.kaptest.com/press-releases/kaplan-test-prep-survey-despite-an-uncertain-employment-landscape-law-school-applicants-still-consider-school-rankings-far-more-important-than-job-placement-rates-when-deciding-where-to-apply).↑
8. See Amanda Griffith & Kevin Rask, The Influence of the U.S. News and World Report Collegiate Rankings on the Matriculation Decision of High-Ability Students: 1995-2003, 26 Econ. of Ed. Rev. (Issue 2) 244-255 (2007). Their regression included variables for student-faculty ratio and expenditures per student, and fixed effects for the schools themselves.↑
9. A defendant school might argue that the scholarship offer that the applicant declined would have been lower if that school had not had to compete with the defendant, so the loss to the student was less. Perhaps that argument should be ignored on principle. But even if it is not, it may be possible for a student who gave up a scholarship to find out from the school that offered lower tuition how much the net tuition would have been if the student had not had the offer from the defendant school.↑
10. A student could similarly argue that he relied directly on a reported LSAT or UGPA statistic, as opposed to the USN&WR rank to which it contributed. A student might want to attend a school with low LSAT’s because the competition for first year grades would be easier, or because the closely-ranked school with lower numbers must have other, more-important, factors in its favor. Or a student might want to attend a school with high LSAT’s because it would be easier to learn from classmates, or because employers might assume the student had a higher LSAT, or because the LSAT numbers reflect student decisions, which is a form of crowd-sourced information on the quality of law schools. It is not clear, however, that many students could make a convincing case that they made their decision on the basis of the median LSAT or UGPA, as opposed to the rank of the school, or its reported employment rate. There is a lot of buzz on the web when USN&WR publishes and new ranking and the vast majority of it relates to the ranks, not to the LSAT’s reported by USN&WR at the same time.↑
11. It may be the case that students at other law schools are harmed when a school gains a ranking advantage from reporting inaccurate numbers. The rank of their school might sink a bit, but it seems unlikely that many suffer substantial financial harm.↑
12. It should be noted that the reporting requirements were not necessarily misleading when they were adopted by the ABA for the purpose of accreditation.↑
13. Michael Simkovic & Frank McIntyre, The Economic Value of a Law Degree Presented at the 2013 Annual Meeting of the American Law and Economics Association (April 23, 2013), available at http://ssrn.com/abstract=2270175.↑
14. Id., at slide 39, Fig. 5. The premium reached lower points around 1999 and 2002 and a higher point around 2007.↑