Attorney Overproduction

Legal Profession 1994-2008: 103,200 lawyer jobs for 561,220 ABA grads

“That the current employment rate is not lower [than 88.3%] is further evidence of the resilience and remarkably steady employment market for U.S. law school graduates, even in times of relative economic weakness.” – Selected Findings for the Class of 2009 (NALP)

Reading statements like these, Annie Lowery’s Slate article attributing the law school application drop to the scambloggers and the improving economy, and the Public Interest Institute’s study led me to ponder what government data have to say about overall employment for lawyers. Assuming every ABA law school graduate wants to become an attorney, i.e. highly discounting the “versatile juris doctor” argument, things have been very bad for law students for a long time. True, some people do pursue legal educations for non-attorney career tracks and others pursue second degrees and enter other fields, but this assumption must hold true if law schools are to be taken seriously at all. I culled Bureau of Labor Statistics attorney employment numbers courtesy of Mr. Peabody’s Wayback Machine. Apparently, for at least as long as the Internet has existed, the BLS has been banging the lawyer oversupply gong.

Let’s compare employed attorneys to ABA law school graduate output. I don’t have replacement data for this time period.

While the lines match up somewhat in the early 2000s, the growth in employed attorneys is insufficient to absorb all the graduates. To better illustrate, let’s compare net lawyer job growth to ABA graduate output.

The green line shows the required imputed lawyer attrition rate if we were to employ every graduate. Now, we know not all graduates get jobs (and how), we also know attorneys leave the field voluntarily or otherwise (including recent graduates counted earlier), so the imputed attrition rate is an idealistic calculation. Additionally, unfortunately for this dataset, 2001 and 2003 were bullish hiring years for the profession and the BLS gathers data on even numbered years. Oh well, not my problem. We also know that the quality of jobs has been decreasing: more contract work, deferred start dates, and growth in new solo practitioners trying to make a go of things.

Here’s the table:

YEAR: Employed Lawyers Net Growth ABA Grads Required Attrition
1994 656,000 N/A N/A N/A
~1996 622,000 -34,000 78,981 -112,981
~1998 681,000 59,000 80,034 -21,034
~2000 681,000 0 78,526 -78,526
~2002 695,000 14,000 76,066 -62,066
~2004 735,000 40,000 77,479 -37,479
~2006 761,000 26,000 82,696 -56,696
~2008 759,200 -1,800 87,438 -89,238
Totals: 15.73% 103,200 561,220 -458,020

In sum, the economy added a paltry 103,200 lawyer jobs (barely one percent annual growth) for 561,220 graduates between 1994 and 2008. That’s 5.44 graduates per lawyer position over that time period, sadly excluding the replacement rate. In 2009 and 2010, a total of 87,592 people graduated from an ABA law school, and at the 2010 going forward (44,004), between 2008 and 2018 there will be 439,624 new ABA law graduates to compete for 240,400 lawyer jobs, a rate of 1.83 graduates per job. For an apples-to-apples comparison, it’ll be 439,624 graduates to a net of 98,500 jobs, or 4.46:1, which is better than what we’ve had in the past. Because the system is still growing, the actual ratio will increase. If we want a 1:1 applicant to job ratio, enrollments will have to be reduced by more than 55% as of 2008, and that says nothing graduates from non-ABA law schools. The BLS assumes the economy will be at full employment by 2018, so thankfully we can omit underemployed lawyers.

ABA Awakens?

Which leads me to the ABA Standards Review Committee’s Consumer Information Subcommittee’s proposed change to law school employment disclosure requirements.

Anyone who’s read the chart on page 4 will undoubtedly agree it’s a step-up from today’s law school advertising. Transparency advocates must be licking their lips. My complaints are (a) it idealistically assumes everyone who’s employed has had only one job nine months after graduation, and (b) like all transparency initiatives, it is subject to the whim of the “nine months after graduation” standard, which the ABA collects in place of “employed at graduation” because many employers won’t hire people until they’ve passed a bar exam. Law schools, NALP, and transparentists can’t tell us what will happen to graduates two or more years from graduation, even though it’s critical information applicants need. The data I’ve gathered here are the best proxy we have.

If the standard is adopted, expect law schools to lean hard on the bottleneck and versatile JD arguments, convincing applicants that they’ll be employed in the future because the economy is improving just for them. They’ll essentially sell out recent graduates, but law schools will say that students take on the market risk and that they aren’t responsible for the economy, which is true. However, it’s safe to say that if scambloggers can convince people not to go, the law schools’ own shocking employment data will do quite a bit of damage to their own credibility. 0Ls’ reasonableness will be put to the test, though I personally believe many people will still go to law school out of perceived financial desperation. I doubt transparency will be sufficient to save the legal education system.

Adoption of the ABA proposal is also a subtle moral test of the transparency movement, which I’ve felt has been opaque about the lawyer oversupply problem. I hope transparency advocates (to say nothing of the incredible NALP) aren’t surprised at the results that logically follow from what the government has been saying since I was in middle school: A long-term career as a lawyer is a fantasy for the majority of law students and for the federal government that unquestioningly finances the legal education system.

Legal Sector 2018: 500,000 Law Grads vs. 100,000 Lawyer Jobs


[UPDATE: Through more recent research, I discovered that the BLS does calculate job growth due to replacement needs, projecting that 141,900 lawyers will retire or pass away. Adding that to the net growth figure cited below gives us a total of 240,400 lawyer job openings between 2008 and 2018. Read the Law Graduate Overproduction page for more information.]

Introduction: Harvard Says We Need More Lawyers

Recently, the ABA teamed up with Indiana-Bloomington Professor Bill Henderson to produce a chart of attorney salaries county-by-county. Now, I’m the type who’s wont to read blog comments (and here readers give the ABA a good whoopin’), and one attorney going by the moniker, Joel, said he crunched some of the numbers of attorneys in the U.S. based on 2009 BLS data on his own blog, Legally Sociable, where goes by Sagescape. Now, I use the 2008 data because it includes self-employed lawyers (partners and solos) and it predicts future attorney employment (very bad, more on that below), but I appreciate a more recent number. Sagescape directs us to Harvard’s “Analysis of the Legal Profession and Law Firms,” where we read the following:

The ratio of lawyers/jobs was flat for most of the 20th century and then rose dramatically after 1970, roughly doubling between 1970 and 2000. That increase (of 100% over 35 years) is similar to increases in the share of service and knowledge-based jobs in the US economy since 1970. (Wyatt and Hecker 2007)

The growth of lawyers is constrained by the output of US law schools. [!!]

The number of law schools has grown, but more slowly than the ratio of lawyers/jobs.  In 2007, there were 196 ABA-accredited law schools, compared to 144 in 1970, an increase of 36% (ABA).

The number of law school graduates has grown, faster than the number of law schools, but again, more slowly than the ratio of lawyers/jobs.  There were roughly 44,000 graduates of accredited law schools in 2005, compared to 30,000 in 1975 and 36,000 graduates in 198 [sic], an increase of roughly 50% over 30 years (ABA).  (Numbers for the early 1970s were depressed by Vietnam.)

Wow Harvard, the sting of your lash! Who thought universities were cruelly denying the public affordable legal services by declining to open more law schools!

Harvard bases its research on Wyatt and Hecker, who give us a similar analysis, though in their defense they use data going to 2000.

Lawyers and judges increased one-and-a-half times as a proportion of total employment between 1910 and 2000, with almost all growth coming since 1970. (See chart 5…) Between 1910 and 1970, lawyers and judges grew from 0.29 percent to 0.35 percent of employment (reaching a peak of 0.36 percent in 1940), after which they jumped to 0.71 percent by 2000. Employment grew from 112,000 in 1910 to 272,000 in 1970 and 927,000 in 2000. Stiff licensing requirements (for both individuals and law schools) and other restrictions on supply limited growth through 1970, but as these restrictions weakened or disappeared, the number of law graduates grew. At the same time, demand for lawyers increased, as many more laws were enacted, business activities became more complex, and society became more litigious. Civil rights legislation for minorities, women, and older and disabled persons; laws regarding the environment, employer-employee relations, product safety, and consumer protection; and higher crime and divorce rates all contributed to the growth of lawyers and judges. Several Supreme Court decisions expanded the right to a court-appointed counsel for criminal defendants, which in turn led to increased funds for public-defenders’ offices and a sharp increase in the number of court-appointed defense attorneys. [Page 43]

Here’s Wyatt and Hecker’s “Chart 5”:

So I guess law jobs are growing faster than society is making new lawyers. Well folks, that’s it for the Law School Tuition Bubble. I suppose I’ll just blog about music (I’m still listening to the Softies; help me stop!) or just post my crazy doodles on the Internet.

What’s that you say? My dozen or so consistent readers want a pro forma challenge? A rear-guard action, if you will?

Fine, but only because I like you.

Legal Sector Reality

Intuition tells me to return to Amir Efrati’s 2007 Wall Street Journal article, “Hard Case: Market Wanes for Lawyers.” Efrati’s article is better than David Segal’s 2011 NYT piece because he does more empirical research. This graphic always entrances me, like dangling string in front of a housecat:

A slack in demand appears to be part of the problem. The legal sector, after more than tripling in inflation-adjusted growth between 1970 and 1987, has grown at an average annual inflation-adjusted rate of 1.2% since 1988, or less than half as fast as the broader economy, according to Commerce Department data.

A while back, I tried to find the Commerce Department data Efrati refers to and couldn’t, but after my recent ABA→Legally SociableHarvardWyatt and Hecker sojourn, I did. I don’t know where Efrati gets the 1970 numbers, but I can get the 1977 to 2009 ones and I’ve adjusted everything to inflation using 2005 chained indexes.

Here’s the legal sector’s nominal size in billions of dollars:

As my favorite BigLaw partner, Phil Ken Sebben would say, “Rising profits, enlarging revenue, lots of good firm growth, we’re turgid with cash!”

Are we? Not quite. Here’s the size of the legal sector relative to the economy:

Wait a minute. Efrati’s “tripling” between 1970 and 1987 isn’t even that descriptive. Things were flat from 1977 until 1981, and after 1992, the legal sector’s proportion of the economy fluctuates between 1.44% and 1.55% of the U.S. economy. That’s certainly not the impression Wyatt and Hecker give us.

Let’s expand, update, animate, and interpret from Efrati:

DUUUUUUUUUDE. It would appear the legal sector of 2009 is no larger than it was in 2000, and in the twenty years between 1989 and 2009, the legal sector grew just under ten percent.

This is a very, very different story from the one Harvard told us. Taste the whip Harvard, now bleeeeeeed fo~~r me~.

Let’s say there was fantastic growth from the 1970s to the late 1980s. Obviously it wasn’t continuous. The legal sector went into a deep intra-economic recession during Paul Volcker’s infamous years at the Fed. In 1983 though, fantastic growth resumed. From 1983 to 1990, what I call, “Seven Years of Plenty,” the legal services sector grew at an annualized rate of 5.29% from 58.265 to 83.591 (original). Much of that occurred in 1988 when the sector saw a tremendous 13% surge. Savings & Loan litigation perhaps?

Even though the legal sector resumed growth along with our recent bubbles, the growth rates between the two tell us that the legal sector didn’t really keep up with the economy.

What’s going on here?

Two things happened in “Seven Years of Stagnation.” One, while the rest of the economy grew, the legal sector fell into recession twice. Two, the legal sector stopped growing past 1.44% of the economy. Efrati called this a “slack in demand,” but I see it as something else: Maturity, which the Economic Glossary defines as:

The third stage in the product life cycle, characterized by flattening of sales and decreasing profit margins. Advertising and promotion are used to maintain market share and to prevent the erosion of sales and profits. During this stage, the initial decline of a product begins and many businesses try to “re-invent” their products to prevent the upcoming decline stage. Many times the company finds new uses for an existing product (baking soda as a deodorizer), totally new markets (foreign countries), or a way to enhance the existing product to make it better and to re-start the life cycle. The television has gone through at least two life cycles, first from black and white to color and then from color to high definition (HD) and plasma. Along the way there were enhancements such as remote control, VCRs to complement them, and cable to help with reception.

Unlike some products (VCRs), legal services will always be in demand, but 1.2%-1.5% of the economy is the most it will ever get because legal services are an intermediate good: they’re a catalyst for economic activity and never an engine for growth. The best firms can do is increase advertising, to the extent they’re allowed, and I think this includes an even greater emphasis on credentials. This is something else Bill Henderson writes about. Once the pond stops growing the issue turns to how its resources are distributed. For law firms, demand for those who are perceived to be the very best attorneys increases. I suspect this also explains the birth of the bi-modal starting salary distribution, which also occurred at the time, but BigLaw is not my bailiwick, so read Steve Harper and Jerry Kowalski as a supplement. I recall Stephen Bainbridge coming to this conclusion.

Ironically, the fact that the number of law students per capita declined during the 1990s actually mitigated some of this, though the 2000’s growth towards the 1980 record doesn’t help.

Maturity also implies that legal education’s then tolerable inefficiencies hit a wall, and normally market forces would’ve required adaptation. The financing system and accreditation monopoly prevents that, and we can see the rising tuition as a result of increased competition for high quality students. Perhaps maturity even explains the prevalence of the U.S. News rankings.


Another thing that happens in this stage of the product cycle is a shift from providing new services to increasing productivity in existing ones. Fewer inputs for the same output saves money, attracting clients. For example, I redirect you to the recent coverage on the development of super-effective document review software. What caught my eye there was this line:

[The five CBS studios sued by the Justice Department] examined six million documents at a cost of more than $2.2 million (~$7.4 million in 2011), much of it to pay for a platoon of lawyers and paralegals who worked for months at high hourly rates. But that was in 1978. Now, thanks to advances in artificial intelligence, “e-discovery” software can analyze documents in a fraction of the time for a fraction of the cost. In January, for example, Blackstone Discovery of Palo Alto, Calif., helped analyze 1.5 million documents for less than $100,000.

6.7¢ per document versus $1.23 per document. Which would your client rather pay?

One could argue that with cheaper discovery litigation will come into greater vogue, especially between small plaintiffs against large entities, but any technological advancements will likely reduce the legal sector’s size within the economy and it certainly won’t employ all the lawyers it offsets.

Conclusion: Law Graduates Dashed Upon the Rocks

In fact, we’re not going to need nearly as many lawyers as our law schools are producing. The smoking gun, which sadly was under my nose all this time, is in the erstwhile 2008 BLS data, which linked to a PDF projecting job growth for lawyers. Between 2008 and 2018 net lawyer job growth will be…98,500.

Of this, 50,300 will be legal sector employees (non-equity attorneys); 29,400 will be in government, and a whopping 9,800 will be self-employed, i.e. partners and solos. Obviously we’re talking about net job growth, so we must account for retirees, advancers (e.g. judges, legislators, presidents and governors, law professors (which will decline at some point regardless of what the BLS thinks), etc.), walk-aways, lay-offs, failed practices, and ousted partners. Nevertheless, given that the ABA-accredited law schools produce 44,000 juris doctors per year, to say nothing of the roughly forty non-ABA schools, it should be clear that no sector of the economy will gainfully absorb half a million new lawyers by 2018.

It looks like the LSTB will continue its traditional programming.

Law School Foundation & Accreditation over Time

Every time I go to the Wikipedia’s “List of Law Schools in the United States,” (despite its inaccuracies) my eyes move towards the “Year Founded” field. I did this recently and then thought, “Hey, what’s the relationship b’ween foundation and accreditation?” Fortunately, the ABA keeps track of this, and it’s also in the mountain of pdfs the LSAC maintains.

Why should we care about this?

The ABA is often criticized for allowing too many law schools to open, with citation to the sixteen schools that have been at least provisionally accredited in the last decade. Is it just those bad apples, or does the problem go back to the fifty-four that have been accredited since 1970? Earlier?

Here’s a graph.

The average law school was founded in 1917 and was accredited in 1950. The median law school opened in 1909, and the median accreditation year is 1939.

A few notes on what you’re seeing:
(a)    These are all 199 of the currently operating ABA-accredited law schools. No JAG school, online schools, or non-ABA schools are included.
(b)    More importantly, law schools that have closed or merged with others were not counted. If you’re interested in that stuff, see the Wikipedia above. That number of law schools in question is miniscule nonetheless.

Some observations:
(1)    Law schools are very much industrial era institutions. Fifty-four opened in the 27-year period between 1890 and 1916.
(2)    By the time the ABA began accrediting law schools in 1923, 121 of the existing 199 law schools that would eventually join the ABA were operating.
(3)    Law schools did not open during the Civil War or both World Wars (as far as U.S. involvement was concerned), but they’ve opened uninterruptedly since.
(4)    The gap between existence and accreditation stops shrinking after 1970, which is probably due to state bar authorities mandating ABA law degrees for legal competency requirements in the 1950s. It’s only after about 1960 that law schools open and receive accreditation two to three years later.
(5)    So, from the late 1960s until about 1980, the rate of law school establishment increased. From 1967 to 1977, 23 law schools opened. I attribute this to the Higher Education Act’s student lending programs. The post-Carter/Reagan/Volcker era has seen a steady rate of law school foundation that is still not quite as steep as the Gilded Age’s.
(6)    Two law schools received immaculate accreditation: University of Arkansas at Little Rock and Widener Harrisburg were both accredited before they were founded. Most likely this is due to existing part-time satellite programs.
(7)    Two law schools that have been accredited recently have been operating for decades. Faulkner University opened in 1928 but joined the ABA in 2006, and John Marshall Atlanta opened in 1933 but received accreditation in 2005.

Let me be clear here: Not all law schools were created equal. When we find that 61% of the currently accredited law schools existed when the ABA began accrediting them in 1923, those law schools were a world apart from the Super Law Schools of today. Their class sizes would have been much smaller relative to the population. Some law schools would unashamedly provide a more barebones education than others. Jobs would’ve been easier to come by, though we don’t know if law always and everywhere had a high ROI. Tuition would’ve been cheaper but would’ve required savings rather than debt. University of San Diego Professor Maimon Schwartzschild provides the qualitative analysis:

Until the 1970s, it is fair to say, most law schools were not burdened by excessive intellectual ambition. At all but the elite schools the classroom teaching tended to be unimaginatively doctrinal; faculty scholarship was not a priority; there was very little interdisciplinary academic work. Even some of the more prestigious law schools were not necessarily intellectual hotbeds, but there was an enormous gap in terms of scholarly activity and ambition between the best law schools—Harvard, Yale, Chicago, and a very few others—and all the rest.

In the 1970s and ‘80s that began to change. Most university law schools, and even “free-standing” proprietary law schools, began to encourage faculty scholarship, publication, and a more academic style. Academically and intellectually ambitious faculty were hired at law schools where they would have been out of place, to put it gently, in the past.

Add to this U.S. News’s national unitary ranking system in 1987 along with mispauperous bankruptcy laws, and we get the Super Law Schools of today.

However, arguably the problem isn’t just that the ABA is asleep at the switch (notwithstanding its insistence that antitrust regulations tie its hands behind its back). Rather, critics of Christopher Columbus Langdell’s 1870 pedagogy are right: the ABA’s accreditation standards still enable the elite industrial era law school of 1923, and U.S. News encourages it. Thus, it’s inaccurate to say that the law school problem began in the last decade. Sure, the standards must’ve changed in the last ninety years, but most of the law schools were already in existence when all this began.

Frequently predicting that numerous law schools will have to shut down, I think it’s too late. Even if law schools reduce their local impacts by returning to their more varied non-intellectual forms, there simply is not and will not be demand for 45,000 new nationally accredited juris doctors in the legal market for the foreseeable future. No matter what the “I” is, when the “R” in ROI is zero or just less than what one could earn in traditional working class occupations then the system is too big. Though I believe the ABA’s accreditation privileges will be rescinded as the crunch matures, hopefully new lawyer training systems will reflect 21st century practices instead of 19th century ones.