February 09, 2016
The latest unscientific fad among law school watchers is comparing job openings projections for lawyers from the Bureau of Labor Statistics* with the number of students expected to graduate from law school. Frank McIntyre and I tested this method of predicting earnings premiums--the financial benefits of a law degree--using all of the available historical projections from the BLS going back decades. This method of prediction does not perform any better than random chance.** Labor economists--including those working at the BLS--have explicitly stated that BLS projections should not be used to try to value particular courses of study. Instead, higher education should be valued based on earnings premiums.
Bloggers who report changes in BLS projections and compare projected job openings to the number of students entering law school might as well advise prospective law students to make important life decisions by flipping a coin.
Many law graduates won't practice law. Many engineering graduates won't become engineers. Many students in every field end up working jobs that are not directly related to what they studied. They still typically benefit financially from their degrees by using them in other occupations where additional education boosts earnings and likelihood of employment.
And if one's goal really is to practice law even if practicing law is not more lucrative than other opportunities opened by a law degree, then studying law may not be a guarantee, but it still dramatically improves the odds.
* BLS job opening projections--which are essentially worthless as predictors for higher education--should not be confused with BLS occupational employment statistics, which provide useful data about earnings and employment in many occupations, including for lawyers.
** There isn’t even strong evidence that changes in the ratio between BLS projected lawyer job openings and law class size predict changes in the percent of law graduates who will practice law, although the estimates are too noisy to be definitive. Historically, the ratio of BLS projected openings to law graduates (or first year enrollments 3 years prior) has systematically under-predicted by a wide margin the proportion of law graduates practicing law shortly after graduation, although it is clear that a large minority of law graduates do not practice law.
February 9, 2016 in Guest Blogger: Michael Simkovic, Law in Cyberspace, Legal Profession, Ludicrous Hyperbole Watch, Of Academic Interest, Professional Advice, Science, Student Advice, Web/Tech, Weblogs | Permalink
February 02, 2016
Smaller or Larger Law Class Sizes Don’t Predict Changes in Financial Benefits of Law School (Michael Simkovic)
One of the most surprising and controversial findings from Timing Law School was that changes in law school graduating class size do not predict changes in the boost to earnings from a law degree.* Many law professors, administrators, and critics believe that shrinking the supply of law graduates must surely improve their outcomes, because if supply goes down, then price—that is, earnings of law graduates—should go up.
In a new version of Timing Law School, Frank McIntyre and I explore our counterintuitive results more thoroughly. (The new analysis and discussion appear primarily in Part III.C. “Interpreting zero correlation for cohort size and earnings premium” on page 18-22 of the Feb. 1, 2016 draft and in Table 10 on the final page).
Our results of no relationship between class size and earnings premiums were robust to many alternative definitions of cohort size that incorporated changes in the number of law graduates over several years. This raises questions about whether our findings are merely predictive, or should be given a causal interpretation.
We considered several interpretations that could reconcile our results with a supply and demand model and with the data. The most plausible interpretation seemed to be that when law class sizes change, law graduates switch between practicing law and other employment opportunities that are equally financially rewarding. While changes in the number of law graduates might have an impact on the market for entry-level lawyers, such changes are much less likely to have a discernible impact on the much larger market for highly educated labor.
Although law graduates who practice law on average earn more than those who do not, at the margin, those who switch between practicing law and other options seem to have law and non-law opportunities that are similarly lucrative. We found that the proportion of recent law graduates who practice law does decline as class size increases, but earnings premiums remain stable. This is consistent with the broader literature on underemployment, and supports the view of law school as a flexible degree that provides benefits that extend to many occupations. (See here and here).
A related explanation is that relatively recent law school graduates may be reasonably good substitutes for each other for several years, blunting the impact of changes in class size on earnings.
Interpretations that depend on law students and law schools perfectly adjusting class size in anticipation of demand for law graduates or future unemployment seem implausible given the unrealistic degree of foresight this would require. Offsetting changes in the quality of students entering law school—an explanation we proposed in an earlier version of the paper—seems able to explain at most a very small supply effect. Although credentials of entering classes appear to decline with class sizes, these changes in credentials are relatively small even amid dramatic changes in class size, and probably do not predict very large changes in earnings. Imprecision in our estimates is another possibility, although for graduates with more than a few years of experience, our estimates are fairly precise.
Even if there are effects of law class size on law earnings premiums, they are probably not very large and not very long lasting.
* The finding is consistent with mixed results for cohort size in other econometric studies of cohort effects, but nevertheless was contrary to many readers’ intuitions.
January 27, 2016
January 25, 2016
January 22, 2016
"As of 01/15/16, there are 157,319 2016 applications submitted by 25,260 applicants for the 2016–2017 academic year. Applicants are up 2.0% and applications are down 0.3% from 2015–2016." Applicants have consistently been up in the 1-3% range all season, and I still wouldn't be surprised if we finished the year up 3-5%.
January 20, 2016
January 18, 2016
Ian Ayres (Yale) argues that the U.S. News rankings have actually helped low-ranked schools during the decline in applicants over the last few years
He sets out his theory here. Briefly: highly ranked schools have enrolled fewer students during the decline, rather than taking the more students with weaker credentials, in order to maintain their rank in U.S. News; the result was more students with good (but not top-flight) credentials available for lower-ranked schools. To which I say: maybe. The decline in enrollments at top schools has been small, and many have seen declines in their student credentials anyway. But it's an intriguing possibility!
January 13, 2016
...applicants up a little over 1%, applications down not quite 2%. Reasonably, applicants are feeling a bit more confident, so submitting fewer applications. Given the trend towards more applications later in the cycle, I think we can say with confidence that we won't see fewer applicants this year, and may still see a very small increase (I'll be surprised if it's more than 5% in the end).
January 12, 2016
A new study by two economists debunks the "barista with a B.A." myth. They do not look at those who earned a J.D. since the Great Recession, but one suspects it's an even bigger myth in that case--not that any factual information would dissuade the group-polarized crowds that congregate on blogs!
January 11, 2016
...courtesy of Prof. Jerry Organ (St. Thomas/Minnesota). Most, but not all, schools taking large numbers of transfers are clearly doing it for U.S. News reasons: it allows them to reduce the size of their 1L class, whose median credentials count for ranking purposes.