Answer: We’ll never really know (for reasons I’ll get into), but there are some good indicators that it was late 2009/early 2010. This isn’t widely acknowledged, but demonstrating it requires some careful calculating that’s worth showing.
The question is, when did people start not thinking about going to law school? (If that makes sense.)
Enrolling in law school is a multi-step process that I’ll reduce to two, taking the LSAT and applying. Looking at applicants is easy enough; looking at LSATs is not. Some people take the test more than once, and the LSAC is stingy about releasing data that break down LSATs by how many tests were first-time or repeat administrations. Worse, the previous release of the LSAC’s “Performance of Repeat Test Takers on the Law School Admission Test,” which publicizes this information, stopped at February 2010, tantalizing us with what was to come. Nor did it help that the February 2010 administration was a cliffhanger: There was an 11 percent drop in first-time takers over the previous year! Did the decline continue in June! Tune in next season, viewers!
Well, it’s been three years, so the LSAC has at last televised the resolution (pdf).
It turns out that the number of first through third-plus LSATs do not add up to the total. I listed the discrepancy as “Canceled/Nonstandard.” I don’t know if that’s correct strictly speaking, but it tends to correlate with movements in first-time takers. I have no idea how many people take the LSAT under nonstandard conditions, but the data show a discrepancy, so there it is.
A few observations. One, by February 2013, there was a record low number of first-time LSATs going back to 1995 and certainly earlier. Two, the number of second-time LSATs has trended upward since 2006, which is attributable to the ABA’s decision to require law schools to report repeat takers’ highest LSAT scores and not their averages. This is one more reason to isolate first-time takers.
And what do they tell us? Although there was a substantial drop in February 2010 first-time test takers, the decline was modest the following June. Then it hit the fan in October.
You can also see that repeat takers resisted declines for a few more periods. All administration types only started falling together in June 2011. The real question is why did February 2010 fall more than June 2010? For that I have no answer. I’d need to know the typical demographics of the test takers in each cycle, so the February drop could be mostly an anomaly. The only thing I know is that first-time takers in that period tend to have lower scores, which indicates that they might be older, non-trad types—not because people’s LSAT abilities rapidly turn to mush as people age but because people who would naturally do well on the test either did so already or are employed in other fields. It’s a big ol’ case of selection bias.
All told, we can imagine the number of LSAT takers as a ball being pushed up a hill by multiple forces and resisted by others, e.g. unemployment versus the population of college-educated young people and informal voices telling them not to bother with law school. At some point the forces pushing the ball uphill were countered by other forces in late 2009/early 2010. Those were, I imagine, early scamblogs (Big Debt Small Law), Above the Law (Elie Mystal’s writings in particular), and Mark Greenbaum’s op-ed to the L.A. Times.
If you want a second opinion on the evidence, we can hypothesize that a “signature” of the tipping point would be an unusual number of people who took the LSAT once or twice but who decided not to retake it or apply after all. Mathematically, the LSAT year, June to February, corresponds best to the applicant cycle. We find that in 2007, 2008, and 2009, that on average there were 1.08 first-time LSAT takers per applicant in 2008, 2009, and 2010, respectively. In 2010, the ratio spiked to 1.17 and then fell a little in subsequent years. With 102,948 2009-2010 first-time LSAT takers, and assuming the 1.08 ratio, we get about 94,500 applicants for 2010. Instead, there were only 87,900, implying that about 6,500 people with a first-time LSAT score in hand in 2009-10 chose not to apply to law school. (It doesn’t matter that the ratio accounts for behaviors of repeat takers, it just needs to be measurable, correlative, and logically related.) Repeating the same process for all LSATs to find a ceiling that includes subsequent test takers who mayn’t’ve applied, we get 100,300 expected applicants, a shortfall of 12,400.
So, sometime between receiving their scores for the June 2009 administration and the beginning of the fall 2010 application cycle, somewhere between 6,500 and 12,400 people dropped out between the LSAT cycle and the application cycle. Switching to a calendar year LSAT cycle (to align it better with the application cycle) raises the range to 8,800 and 14,400.
This, second analysis, suggests that the tipping point occurred in late 2009 but the impact on first-time LSATs wasn’t felt until 2010. It could mean more credit for the early scamblogs and Above the Law, and other sources than the Greenbaum op-ed.
Once the peak was reached, the countervailing forces overpowered the upward ones. Unemployment gave way to distrust in law school as a viable long-term career path. Certainly other sources contributed as well, e.g. The New York Times, but those articles were published a year after the Internet set things in motion. Now inertia is setting in and the trough is slowly being reached.
It’ll be another three years until the LSAC updates the number of first-time and repeat takers again. We may find out who contributed to the upswing in February LSATs in 2014.