New America Foundation: Let the Sins of Grad PLUS Be Visited Upon IBR

I’ll try to go quickly through the New America Foundation’s (NAF’s) Jason Delisle’s and Alexander Holt’s Washington Post opinion piece from Friday. Reacting to news that the president’s budget forecasts income-based repayment programs (IBR) will cost the government an additional $21.8 billion, the authors argue that “too much” of it is attributable the administration’s changes to IBR, i.e. reducing monthly payments even more and accelerating loan forgiveness to 20 years from 25. Their article has many problems.

One, Delisle and Holt don’t provide evidence that the $21.8 billion comes from the changes to IBR. They’re just conjecturing. My hunch is that the additional costs are mainly attributable to the changes in the budget’s model that don’t anticipate as much job growth as before—or just increased participation in IBR. Without this evidence, the rest of Delisle’s and Holt’s article is just righteous huffing.

Two, the authors use this pretext to slide into their grad-students-are-abusing-IBR claim the NAF has been making for a few years now. This argument is problematic because the problem isn’t IBR so much as the Grad PLUS Loan Program, which the authors understand is unlimited and to their credit have advocated abolishing elsewhere. That’s all fine and good, but if the problem is Grad PLUS, then it’s not IBR, and the authors should focus on that instead. More on this point below.

Three, the grad-students-are-abusing-IBR claim has never been substantiated either. The NAF has always trotted it out in hypotheticals without doing the actual research. How many (and what percentage of) graduate debtors are (a) on IBR and (b) earn high enough incomes that could allow repayment under 25-year or consolidated repayment plans without compromising their living standards? Also, how many grad debtors are on IBR but are not earning enough to repay their loans under the older repayment plans?

These questions are crucial because until they’re answered those of us sitting at the feet of the East Coast think-tank elite can’t weigh how many people unfairly benefit from the changes to IBR against those who do not. If every unfair IBR beneficiary is canceled out by dozens of debtors who will never repay their loans in 25 or 20 years, then it’s safe to say that the changes to IBR are useful and the adverse consequences minimal. (And it’s not like the IBR changes have influenced people’s graduate school enrollment behaviors as law school applicants are still falling.) In the end, Delisle’s and Holt’s arguments are really just revamped versions of welfare queen fear-mongering.

Four, Delisle and Holt do not regain any sympathy with their hypothetical graduate debtor, Robert, who finishes law school with $150,000 in debt and earns $70,000 per year. Here are the problems with Robert:

(a)  For those of us who’ve done the research, Robert’s debt is plausible, but his income is not. Robert earns well over the median salary reported to NALP in 2013 ($62,467). Assuming that all non-reporting graduates are making less than the median, which I believe is fair, Robert is above the top 23 percent in law graduate earnings. He is quite atypical. The true median, which would include graduates working part-time and the 12.3 percent who were unemployed (and matter since we’re talking about debt repayment), is much, much lower. It’s likely many of them will never repay their loans. These people will benefit from the PAYE changes, but the NAF ignores them.

(b)  The authors then fashion out of Robert’s rib a wife, who earns $80,000 per year. With an annual household/family income of $150,000, readers should recognize that this partnership is in the top 10 percent by household income. Is this common for graduate debtors? Probably, but again the authors don’t say.

(c)  Delisle and Holt proceed to criticize IBR for not taking spousal incomes into account, that only 1.9 percent of Robert’s household’s/family’s income is going to his student loans. Are you shocked? Well, the response is, so what? Robert’s wife didn’t sign his master promissory notes any more than she would his gambling debts. If Robert wants to leave work to raise their kids, for example, doesn’t that imply that his wife will essentially assume his debts? Would the NAF say this if Robert were Roberta? How would unmarried Robert feel if he had to tell his bride-to-be that she’d be partly on the hook for his student loans if they got married? Again, what if Robert were Roberta, who would be more likely to take time off to raise children?

(d)  The authors’ hypothetical is only as outrageous as the lopsided assumptions they bake into it. It’s one thing to say that Robert, unusual though he is, benefits more from the Obama administration’s changes to IBR than before. But it’s a rhetorical foul altogether to throw in a wife, whose high earnings Robert largely has no power over, and then blame IBR for the result. Delisle and Holt could just as easily give Robert’s parents multimillion-dollar lottery tickets or 5 percent of Maine’s landmass, but it would still have little relevance for IBR as a policy.

Five, the authors repeat that graduate debtors are the unfair beneficiaries of the administration’s changes to IBR, that they’re half of all IBR participants (unsurprising: they have undergraduate debts too), that they have higher incomes and are less likely to be unemployed than undergrad (or non-grad) debtors. Again, no income data on IBR participants is given, so Delisle’s and Holt’s IBR welfare queens are all speculative. Now, I’m sure some exist, but the NAF needs to show us the bodies and carefully tell us whether they’re worth less than the number of underemployed graduate debtors who won’t be able to repay their loans.

Six, even if they do that, all their talk of IBR’s “loan-forgiveness benefits” is really a problem with Grad PLUS loans, not IBR. As I wrote last week, IBR without Grad PLUS loans would be much more innocuous. It’s one thing for Delisle and Holt to make poor arguments with unrepresentative examples, but I question their credibility if they’re going to attack IBR, which I think we all agree was never crafted with Grad PLUS loans in mind, instead of the loan program itself. Why not attack the problem at its source? What’s so special about IBR, then? Nor does it help that they bait their readers with the $21.8 billion IBR shortfall and then switch it with the changes to IBR without evidence. For all their elegant, mathematical—and probably costly—policy papers, the NAF’s results almost always have zero external validity. Like, if I didn’t know any better, I’d say those folks had some kind of ulterior, partisan motive…

Seven, and finally, at the beginning of their article the authors characterize the federal loan program as “an implicit contract: Students get loans to go to college at reasonable interest rates, with no previous credit history required, but when they graduate, they [and their high-income spouses, apparently] have to pay them back. But that agreement is shifting.”

In their dreams. The “implicit agreement” was that the loans would make debtors more productive workers so they could fill higher-paying jobs that required additional skills. Little of this has turned out to be true: There’s no good evidence that widespread college education is raising our national income, and the government has pretty much reneged on its jobs promises.

As far as contracts go, this one has been drafted in favor of the government. When its underlying assumptions are true, everyone wins, but when they’re not, the government won’t be held accountable for self-serving research, false promises, and reckless lending. Instead, attempts to help the debtors will face resistance by people like Delisle and Holt, who will howl at all the alleged benefits the lucky-duckies are getting—and right now we’re only talking about grad student debt! Consequently, you should expect the endgame for all this unpayable student loan debt to be really, really acrimonious.

Site Update 2015-02-09: Law Graduate Overproduction Page

The update can be found here. (The 2011 edition has been moved here.)

To keep the analysis consistent with previous years, I used the class of 2013 even though data for the class on 2014 are available (and logged by moi). It’s a little problematic given that 2013 was the law graduate high tide, but that’s what happens when law schools enroll people without regard to employment outcomes.

I do not discussed the BLS’s proposed changes to its methodology for measuring occupational replacements. Assuming it’s approved, then for future versions, if the BLS separates annual replacement openings between those created by workers who leave the labor force and workers who move to different occupations, then I’ll use the labor force rate as the measure for “sustainable jobs.” It’s imperfect, but the same can be said of the current methodology.

I’ve also updated the site’s highly popular lawyers per capita by state page to include employed lawyers per capita and idle attorneys using the 2012 employment data. I am waiting on the ABA to update its national lawyer counts for 2014 and 2015. (They do plan on doing that right?)

At this time, I will brag that the Census Bureau’s press relations department cited my work on this topic last August.

Law School Gives Away Free Money to Non-Students (More 509 Errata)

Today’s raspberry goes to Duquesne which reports that 117 percent of its full-time students received a grant.

Free Money at Duquesne

My hunch is that there were only 231 full-time students who received less-than-half-tuition grants last year rather than 331. The typo then compounded within the table since so many of the information reports’ numbers are just calculations and not hand-entered data themselves.

Other notables are Texas A&M (formerly Wesleyan), which amped up its scholarships to the extent that only 6 percent of its students are paying full tuition, way down from 57 percent two years ago. New Hampshire’s final year as a private law school also saw a large reduction in full-time students paying full tuition: 21 percent to 8 percent.

Otherwise, the 2013-14 academic year saw another precipitous drop in the percentage of full-time law students paying full tuition.

Percent Full-Time Law Students Paying Full Tuition

Perhaps this year only one-third of the enrollment at a typical law school is paying full freight. Like, maybe I should consider moving the blog to a more descriptive title.

Moving on to the topic of private law school tuition, which is easier to research because most of the time it’s not possible to distinguish resident from nonresident students at public law schools, we can see the plummeting reliance on full-tuition students—as well as revenue from that source.

Mean No. Full-Time Private Law School Students by Grant Received

Aggregate Revenue From Full-Time Private Law School Students Paying Full-Tuition

(These charts exclude Brigham Young because like most public law schools, we can’t know how many students are paying discounted LDS tuition.)

In all, private legal education lost nearly $450 million in annual full tuition revenue since 2011. It now makes less than in 2002. We keep hearing about how resilient law schools are to closing, but when I see this stuff I think that this just can’t go on forever.

Finally, for those curious about the dispersions of students receiving the median grant and how much that’s worth, I’ve come up with a new way of displaying that information: full-tuition quintiles. It occurred to me that simply showing median discounted tuition isn’t useful because an expensive law school can give a large discount while a cheap law school can discount by very little to get to roughly the same price tag. By treating full tuition as the independent variable (and because law schools tend to stay within the same full-tuition quintiles), I can give you a much better idea of how much the median discount is worth. The following charts use the intra-quintile mean average to give a broad picture of what’s going on.

Percent Private Law School Students Receiving Median Grant by Full Tuition Quintile Mean

Full-Time Private Law School Tuition and Median Discounted Tuition by Tuition Quintile Mean

Interestingly, a lot of the action in recent years appears to be within the third and fourth quintiles, which charge the same median discounted tuition to the same percentage of students—but schools in the fourth quintile are able to charge higher full tuition. Perhaps this is where competition over students is fiercest. In general, though, median discounted tuition in all quintiles ranges between 59 and 65 percent of full tuition. Back in 2002, the range was 82 to 91 percent.

If you’re curious what quintile a law school belonged in last year, just look at the tuition data page and compare to this table:

1 $22,560 $36,662
2 $36,994 $40,486
3 $40,890 $44,465
4 $44,520 $48,190
5 $48,730 $57,838

I believe this will be my last analysis of the information reports because I’ve covered all the topics I usually write on. Happy MLK Jr. Day, Americans.

BLS: One in Four Lawyers to Switch Occupations by 2022

…And if you know what that means, great, because there’re plenty of caveats I have to lay out for everyone else.

Steven Harper inspires me to check up on how things are going with the BLS’s proposed rule-change for estimating occupational replacement rates; there was an update on January 2nd. Apparently, the Employment Projections program released a spreadsheet with experimental 2012-2022 replacement and separations data alongside the numbers from the current methodology. I don’t think it was there before, but the BLS says it was. If so I wish I’d noticed it earlier as it’s quite interesting.

For one, my hunch in my American Lawyer article was correct: Under the new methodology between 2012 and 2022, 339,800 out of 759,800 lawyers would be replaced, and the growth rate, which is what we should be caring about because it’s not zero sum, doesn’t get changed. I like getting the numbers right. (Okay, I was off by a thousand.)

For another, the BLS goes further than I expected by separating the total occupational replacement rate into “labor force exits” and “occupational transfers,” which mean as they sound. Labor force exits are certainly going to include most retirements but also people exiting for parental leave and other, less common personal reasons. The labor force exit rate for lawyers is 17.1 percent, which compares strikingly well with the current methodology’s 16 percent replacement rate.

As for occupational transfers, as this post’s title states, it’s 25.5 percent. That’s the concept I’ve been most concerned about all along. These are lawyers who are leaving the profession for different types of jobs. To be clear, some of these transfers are preferable and some not. It includes lawyers who become judges with lawyers who become retail sales clerks. The interesting comparison—and the best I can give you—is with other occupations that require doctoral or professional degrees, sorted by size and occupational transfer rate.

Animal scientists 19-1011 33.3% 54.3% 22.2% 76.4%
Biochemists and biophysicists 19-1021 28.5% 54.1% 18.0% 72.1%
Medical scientists, except epidemiologists 19-1042 21.1% 51.3% 14.5% 65.9%
Computer and information research scientists 15-1111 15.7% 44.9% 11.2% 56.1%
Clinical, counseling, and school psychologists 19-3031 27.2% 42.1% 24.1% 66.3%
Physicists 19-2012 24.5% 38.8% 19.1% 57.9%
Astronomers 19-2011 24.5% 38.8% 19.1% 57.9%
Judicial law clerks 23-1012 16.2% 35.4% 27.5% 62.9%
Postsecondary teachers, all other 25-1199 15.0% 34.1% 32.9% 67.0%
Health specialties teachers, postsecondary 25-1071 15.0% 34.1% 32.9% 67.0%
Business teachers, postsecondary 25-1011 15.0% 34.1% 32.9% 67.0%
English language and literature teachers, postsecondary 25-1123 15.0% 34.1% 32.9% 67.0%
Education teachers, postsecondary 25-1081 15.0% 34.1% 32.9% 67.0%
Biological science teachers, postsecondary 25-1042 15.0% 34.1% 32.9% 67.0%
Mathematical science teachers, postsecondary 25-1022 15.0% 34.1% 32.9% 67.0%
Psychology teachers, postsecondary 25-1066 15.0% 34.1% 32.9% 67.0%
Engineering teachers, postsecondary 25-1032 15.0% 34.1% 32.9% 67.0%
Computer science teachers, postsecondary 25-1021 15.0% 34.1% 32.9% 67.0%
Communications teachers, postsecondary 25-1122 15.0% 34.1% 32.9% 67.0%
Foreign language and literature teachers, postsecondary 25-1124 15.0% 34.1% 32.9% 67.0%
Philosophy and religion teachers, postsecondary 25-1126 15.0% 34.1% 32.9% 67.0%
History teachers, postsecondary 25-1125 15.0% 34.1% 32.9% 67.0%
Chemistry teachers, postsecondary 25-1052 15.0% 34.1% 32.9% 67.0%
Recreation and fitness studies teachers, postsecondary 25-1193 15.0% 34.1% 32.9% 67.0%
Political science teachers, postsecondary 25-1065 15.0% 34.1% 32.9% 67.0%
Sociology teachers, postsecondary 25-1067 15.0% 34.1% 32.9% 67.0%
Law teachers, postsecondary 25-1112 15.0% 34.1% 32.9% 67.0%
Physics teachers, postsecondary 25-1054 15.0% 34.1% 32.9% 67.0%
Economics teachers, postsecondary 25-1063 15.0% 34.1% 32.9% 67.0%
Criminal justice and law enforcement teachers, postsecondary 25-1111 15.0% 34.1% 32.9% 67.0%
Atmospheric, earth, marine, and space sciences teachers, postsecondary 25-1051 15.0% 34.1% 32.9% 67.0%
Agricultural sciences teachers, postsecondary 25-1041 15.0% 34.1% 32.9% 67.0%
Area, ethnic, and cultural studies teachers, postsecondary 25-1062 15.0% 34.1% 32.9% 67.0%
Social sciences teachers, postsecondary, all other 25-1069 15.0% 34.1% 32.9% 67.0%
Social work teachers, postsecondary 25-1113 15.0% 34.1% 32.9% 67.0%
Architecture teachers, postsecondary 25-1031 15.0% 34.1% 32.9% 67.0%
Anthropology and archeology teachers, postsecondary 25-1061 15.0% 34.1% 32.9% 67.0%
Environmental science teachers, postsecondary 25-1053 15.0% 34.1% 32.9% 67.0%
Geography teachers, postsecondary 25-1064 15.0% 34.1% 32.9% 67.0%
Library science teachers, postsecondary 25-1082 15.0% 34.1% 32.9% 67.0%
Forestry and conservation science teachers, postsecondary 25-1043 15.0% 34.1% 32.9% 67.0%
Lawyers 23-1011 16.0% 25.5% 17.1% 42.6%
Judges, magistrate judges, and magistrates 23-1023 16.0% 25.5% 17.1% 42.6%
Administrative law judges, adjudicators, and hearing officers 23-1021 16.0% 25.5% 17.1% 42.6%
Physical therapists 29-1123 24.6% 23.6% 15.2% 38.8%
Pharmacists 29-1051 23.9% 20.5% 16.8% 37.4%
Audiologists 29-1181 20.7% 20.0% 14.5% 34.5%
Veterinarians 29-1131 32.1% 16.6% 14.5% 31.1%
Physicians and surgeons, all other 29-1069 25.0% 14.4% 14.6% 29.0%
Family and general practitioners 29-1062 25.0% 14.4% 14.6% 29.0%
Internists, general 29-1063 25.0% 14.4% 14.6% 29.0%
Surgeons 29-1067 25.0% 14.4% 14.6% 29.0%
Anesthesiologists 29-1061 25.0% 14.4% 14.6% 29.0%
Pediatricians, general 29-1065 25.0% 14.4% 14.6% 29.0%
Psychiatrists 29-1066 25.0% 14.4% 14.6% 29.0%
Obstetricians and gynecologists 29-1064 25.0% 14.4% 14.6% 29.0%
Optometrists 29-1041 29.0% 14.0% 19.6% 33.6%
Chiropractors 29-1011 19.6% 13.6% 13.1% 26.8%
Dentists, general 29-1021 24.4% 12.7% 15.6% 28.3%
Orthodontists 29-1023 24.4% 12.7% 15.6% 28.3%
Oral and maxillofacial surgeons 29-1022 24.4% 12.7% 15.6% 28.3%
Dentists, all other specialists 29-1029 24.4% 12.7% 15.6% 28.3%
Prosthodontists 29-1024 24.4% 12.7% 15.6% 28.3%
Podiatrists 29-1081 20.7% 12.1% 11.1% 23.3%

I didn’t include it, but some of the occupations at the top with high turnover are quite tiny. There are only 2,700 animal scientists, for example, so my guess is there are still problems with the data. You’ll also note that many of the stats tend to clump together by occupation types, e.g. the 33 postsecondary instructor classes, which all have the same replacement rates, for both the new and old methodologies.

But the real money is in comparisons among the professional occupations, which are generally doctors, dentists, and lawyers. Lawyers’ occupational transfer rate is double doctors’ and dentists’. The medical occupations’ replacement rates under the current methodology are only a few percentage points lower than the total occupational separation rate under the new one, but the same can’t be said for attorneys’.

I’m not sure how reliable these experimental data are, but they do tend to show that there’s more turnover for lawyers than the other professions they’re most often compared to. (Ironically there’s even more turnover for postsecondary law instructors.) And I’m not even getting into job quality. I wouldn’t say this is especially strong evidence of a high turnover rate for lawyers, but it’s another piece that fits in that puzzle.

The BLS (still) says it’ll give us another update early this year, but I hope to write on other topics.

10 Ways to Falsify Law Graduate Employment Doomsayers

I begin 2015’s first substantive post by invoking the right of listicle clickbait.

A loose end from December is Loyola Law School, Los Angeles professor Theodore Seto’s response to my American Lawyer article on the Bureau of Labor Statistics’ proposed change to how it measures the replacement rate for lawyers. For Professor Seto, I have good news and better news.

The good news is that when he writes that he’s flattered that I’d respond to his article, he need not be. Of course I was going to write on the topic for The American Lawyer anyway, but his first article usefully illustrated the kind of thinking I cautioned against when I broke the story in early November.

The better news is that his closing line raises an interesting question worthy of further consideration. He writes:

But we should all remember (myself included) that the best legal counselors, when faced with new evidence, adjust their advice accordingly. They do not simply attack the evidence.

Let’s not discuss whether I was attacking new evidence. Readers can compare my article to Seto’s for themselves. Instead, if I interpret Seto fairly here (and he says I didn’t do that for his other article, so I tread lightly), he’s implying that I or perhaps others make unfalsifiable claims about the future of law graduate employment—that we unfairly dismiss any favorable news about graduates’ prospects because it contradicts our dogmatic positions that law school is a poor decision in probably most circumstances.

If so, he’s incorrect. My beliefs are falsifiable, and because the topic of falsifiability arose on this blog two more times in the last month, I’m inspired to write on it. So, here’s a list of events one could point to (and would probably need to) to predict that things will be better for grads in 2016.

(1)  The absolute number of graduates in the classes of 2014 and 2015 employed in full-time, long-term, bar-passage-required, non-school-funded jobs rises. No one disputes that employment percentages will improve on account of there being fewer graduates, but the best way to show that graduates are finding jobs is … showing that graduates are finding jobs. Similarly, I’d like to see evidence that grads are finding better jobs. That could be the NALP reporting that grads are shifting into lawyer jobs at law firms larger than the 2-10 bracket, though 2013 toed in the right direction.

No. Graduates Employed by Size of Firm (NALP)

You can slag biglaw all you want, but it tends to pay better. Likewise, wage growth in the 25th percentile for law grads is absolutely necessary if anyone wants to convince me that law school is better than going back to college for a more lucrative bachelor’s degree, but technically that’s a slightly different issue.

* Note: At this time I’m not too concerned that the ABA’s decision to give law schools a tenth month to report their graduate employment data will substantially impair any comparisons to previous years.

Continue reading

LSAT Tea-Leaf Reading: December 2014 Edition

And the LSAC said, “Let there be growth!” and it was so. A whole 0.8 percent more people (222 souls) took the LSAT last month than in December 2013. The calendar year total is 100,829, which is about 4,500 fewer administrations than the previous year. The decline for 2013 was about half what it was the year before, so we might be bottoming out.

No. LSAT Takers, 4-Testing Period Moving Sum

It’s the second time since the LSAT plunge began in October 2010 that we’ve seen a year-over-year growth in test-takers. The last one was in February 2014, which was a larger, 1.1 percent increase. There will probably be more declines in the future, but at some point in the next few years it will stabilize. I doubt it’ll start growing though.

Meanwhile, week 1 has come and gone, and I can finally give you a trend path for applicants and applications for fall 2015.

No. Applicants Over App Cycle

No. Applications Over App Cycle

It looks like 5,000 fewer applicants is what we can expect for this year. I say again that these numbers still look too high, and I’m surprised that more people aren’t deterred from law school.

Voodoo Links

I’ve been asking myself recently what standards a topic needs to meet before I’ll post on it, and while there’s no point in giving a list of criteria, one thing I try to look for is information readers won’t find at other sites, particularly more heavily trafficked ones. Consider it my offer to readers. So, I thought to myself that maybe I should occasionally point out instances of good work that teaches me something I didn’t know. (It would certainly give me an opportunity to practice not being a meanie in my writing.) I can think of two examples from the last month+.

Law School Truth Center, “LSAT Scores Do Not and Cannot Predict Bar Exam Scores,” Outside the Law School Scam, November 20, 2014.

I have to thank the normally satirical LSTC for playing it straight and explaining item type theory and equating to me. It didn’t just illuminate how standardized testing works; it spared my readers disputable longitudinal comparisons between law school classes. I especially appreciate the post because until then I had no interest in last summer’s bar exam debacle.

Mitchell D. Weiss, “Don’t Believe the Hype: There’s Still a Student Loan Crisis,”, January 5, 2015.

June 2014 blessed us with a policy paper by Beth Akers and Matthew Chingos of the Brookings Institution ridiculing our assertions that there’s a student loan “crisis” on the horizon. Weiss does us a favor by summarizing many of the Brookings Institution’s policy papers over the last year arguing that the student debt system is fine but it’s those irresponsible kids taking on too much debt that’s the problem. I’m wary of giving Brookings any more attention than it already gets: The last thing we need is more genuflecting to D.C. think tanks, but Weiss’ contribution is a solid analysis for anyone interested in Brookings’ position on higher education and student loan debt.

I hope to get to substantive stuff soon.


Get every new post delivered to your Inbox.

Join 160 other followers