We have our first data update for the September LSAT administration. Why do we care? Because LSAT registrants determine 95+ percent of how many people will be applying in a given cycle—and the number of applicants is directly related to how competitive a cycle is.
Per LSAC, the 2019 September LSAT administration stands at over 25,000 individuals registered. How does that compare to last year? In 2018, 28,853 individuals took the September LSAT. If this were an LSAT question, we hope you would notice the term shift from "registered" to "took." The number of registrants always overstates the number of test-takers. There is still a month left for people to cancel their registration (though the refund deadline has passed, people will continue to cancel). Further, on test day anywhere from 5-10 percent of registered individuals usually don't show up for their test.
So what does that mean for this September? Let's make some conservative assumptions. Let's say relatively few people cancel between now and September—that our final registration number is 25,000 individuals. Let's then assume 5% of those individuals don't show up. That leaves us with 23,750 test takers, a reduction of about 5,100 from last year.
But the cycle isn't just about one test's volume. We have to look at June and July as well. June we already have full volume for. The June test had 16,441 takers — a reduction of 6,048 from 2018's June administration. Further, there was a reduction in first-time test takers, who are the primary drivers of cycle volume.
But, and this is a big but, July numbers were up tremendously. According to LSAC, approximately 23,000 individuals took the July 2019 LSAT. That's an almost 100% increase over 2018's numbers—for a total increase of 11,500 test takers. The reason should be fairly obvious. The July test offered test-takers a unique opportunity to take the LSAT, see their score, and then choose to keep or cancel it. If they chose to cancel it they would receive a coupon for a free retake between October and April 2020.
Since the July administration includes such a tremendous number of potential test cancellations, the number cannot be taken at face value. We need to account for the number of individuals who will cancel their score—since those individuals might as well not have taken the test.
Spivey Consulting conducted an entirely unscientific poll of Reddit's LSAT forum to gauge test taker plans for their July score, and received 650 responses (which is a pretty healthy survey sample). We want to note that there are obvious flaws with this survey's sample, though. It should be used only as a relative guide—but the results do line up with anecdotal reports of cancellation plans we have seen and heard. You can see a graph reflecting cancellation plans for July below.
As the chart shows, almost a third of test-takers plan to cancel their July scores—with another 7% likely to do so. A further 7% are undecided. Approximately 40% of respondents are either almost certain or probably going to cancel their July LSAT score.
If we are incredibly conservative extrapolating that number to all July test-takers and assume that 20% cancel their July score, our final July test taker number would be 18,400 individuals—4,600 fewer than the reported number of test takers.
In chart form, June/July/September volume is more likely to look like this:
As you can see, based on the assumptions we make, test-taker volume is likely to be effectively down for the first three tests of the cycle, this year compared to last. These assumptions go both ways—the fewer July test-takers cancel, the closer our volume is likely to be. The more who cancel, the less we will have compared to the prior cycle. This also does not account for the number of first-time takers (which was down in June compared to the last cycle; we don't have July or September data). Fewer first-time takers equals fewer applicants.
It's unlikely we will find out the number of July test-takers who elected to cancel their scores—so we'll have to keep a close eye on the number of applicants through the first months of the cycle. Nonetheless, we feel comfortable with the assumptions made to arrive at these numbers, and reasonably comfortable that our conclusion that there are fewer LSAT takers through September this year compared to last—which is good news for applicants. Don't celebrate just yet though; October and November could easily change these numbers.