In this post, I will argue that within a few years there will be a shortage of entry-level lawyers. Law school enrollments have dropped so much that the demand for new lawyers will far exceed the supply of law school graduates. In a relatively short time, we will have gone from an environment where employers received hundreds of resumes for every open position to one where employers might not receive any resumes at all.
Already, some readers strongly disagree. Before I get to my case, I want to get a couple of preliminary things out of the way. First, there is a comments policy at the end of this post. Please read it before you comment. Second, I want this post, and the comments, to focus on the job market, and not whether now is a good time to go to law school. The two issues often have been conflated recently, but they are distinct. Third, as I mentioned in my last post, I think that it is important in conversations on controversial issues to presume that a person with whom you disagree is acting in good faith. I hope that readers will presume that I am acting in good faith, and I will presume that comments are made in good faith. As the comments policy reflects, this is a rebuttable presumption. Fourth, I would welcome thoughtful analysis in the comments that explains why I am wrong. When I made a similar argument in a post last April, not a single comment (in a thread with more than 100 comments) undercut the analysis. Not one. If there is a good case to be made that I am wrong, or that I am missing a relevant piece of information, I want to know about it.
The post will have two components. First, I will explain why the demand for new lawyers will almost certainly exceed the supply of new law graduates within a few years. Second, I will address (a) the Bureau of Labor Statistics’ projections of legal employment and (b) the argument that the legal job market is being affected by structural change, and will explain why each fails to undercut my analysis.
Why There Will Be A Shortage of Entry-Level Lawyers
The supply of entry-level lawyers is relatively easy to predict for currently-enrolled classes. Roughly 88% of students who enroll in law school end up graduating. This part of the analysis draws heavily on a prior post that I made last April. In that prior post, I borrowed an estimate by Steven Freedman that, based on an 88% graduation rate, predicted that 33,791 people will graduate from law school in the class of 2017. This is similar to an estimate made by Deborah Jones Merritt, and this number is not controversial. The number of graduates is unlikely to be exactly 33,791, but this number is a reasonable one to go with for our analysis.
On the demand side, ABA data from this past year shows that 26,653 graduates of the class of 2013 reported receiving full-time Bar Pass Required jobs within nine months of graduation. Let’s assume for a moment that there is no job growth at all between now and 2017, and that the class of 2017 will also report 26,653 full-time Bar Pass Required jobs. That would mean that 78.9% of the class of 2017 would report having full-time Bar Pass Required jobs within nine months of graduation.
The 78.9% number alone does not look very impressive in isolation, but it is. As Jack Crittenden showed in the National Jurist, even in the boom years of the early 2000s, the nine-month full-time Bar Pass Required number ranged from 68.4% to 72.3%. In other words, in the robust job market of the early 2000s, when associate salaries at big firms were rocketing upwards, the nine-month full-time Bar Pass Required number never got close to 78.9%.
Stopping right here, we can say with virtual certainly that the class of 2017 will graduate into the best entry-level job market in recent memory. Remember that the 26,653 number for entry-level jobs assumes no job growth at all. I expect that there will be job growth as the economy continues to improve, making the job prospects for the class of 2017 even stronger. As I note below, the Bureau of Labor statistics predicts growth in jobs for lawyers. In any event, we don’t need any growth for the class of 2017 to be in very good shape. Indeed, the class of 2017 would have the best nine-month percentage in recent memory even if there is a modest decline in the number of jobs.
How do we get from 78.9% to a shortage? The key thing to understand is that the nine-month job data does not provide the whole picture on legal employment. As the name implies, nine-month job data measures employment outcomes nine months after graduation. Many people get jobs more than nine months after graduation. I explained this phenomenon in detail in a prior post. It should be a common sense proposition that jobs for entry-level lawyers come open between January and May of each year. The timing of the bar exam also is a significant factor here. Bar results come typically come out five to seven months after graduation. A substantial number of people fail the bar on the first attempt. Remember, to this point in the post, we are talking about full-time, BAR REQUIRED jobs. To be sure, some people who fail the bar the first time end up in jobs that are categorized as bar required, but many legal employers (especially small firms) are hesitant to hire entry-level lawyers who have not yet passed the bar. If 15% of people fail on the first attempt, and half of these pass on the second attempt, then 7.5% of graduates will pass the bar on the second attempt, getting their results more than nine months after graduation.
In a prior post, I reported on a study that I performed of the longer-term employment outcomes from my school’s classes of 2010 and 2011. That study showed that the nine-month numbers undercounted the ultimate number of full-time, bar pass required jobs by around 50%. Adding 50% to the nine-month bar pass required jobs reported by the class of 2013 would lead to 39,979. Our projected number of 2017 graduates is 33,791. 39,979 is a lot larger than 33,791. Being more conservative and adding 25% to the 2013 nine-month numbers gives you 33,316 – roughly the same as the number of projected graduates. Remember again that not all graduates will pass the bar on the first time, and that we are talking about bar required jobs here. Remember also that in the very strong legal job market of the boom years, the nine month number never got above 72.3%.
An objection could be raised that the bar-required numbers include solo practitioners. Solos have been 2.3% of the last two graduating classes. This is too small a number to change the analysis. Further, the analysis so far has ignored JD advantage and other professional jobs. Some of these jobs are good. Some of them are not so good. Even if we discount the numbers in these categories by 50% or more, we end up with thousands of good jobs, and these categories would more than outweigh the number of solos reflected in the bar pass required numbers.
To me, these numbers are stark. In just a few years, there will not be enough entry level lawyers to meet basic demand. This might be a good thing for a short period of time. Currently underemployed lawyers will benefit, though my alumni data study suggests that there are far fewer underemployed lawyers out there than many people might think. The number of law school graduates, however, is likely to stay low for several years after 2017. I expect the number of entering students this year to decline a bit further from last year. I also expect the number of jobs to increase as the economy continues to improve. At some point in the future I hope to write another post about potential impacts of the lawyer shortage. For now, though, I will simply say that the numbers suggest that supply and demand will get far out of balance, and that this imbalance will be as severe in the other direction as the imbalance between historically large classes and a poor job market that followed the 2008 financial crisis.
Bureau of Labor Statistics Projections
People asserting a more pessimistic view of future legal employment often point to the Bureau of Labor Statistics projections as evidence that the legal job market will continue to be weak. In a post from last year, I explained what the BLS data do and do not show based on my conversations with the people who actually create them. Many people that cite the BLS data appear to have no idea how the projections are developed or what they are intended to show.
In this context, people often use the BLS projections like I used the ABA data above to make a comparison between number of graduates and number of expected jobs. Joe Patrice at Above the Law, for example, did just that in a recent post, writing that “given that the Bureau of Labor Statistics continues to estimate that there will only be 19,650 new law jobs every year this just means fewer people will graduate unemployed.”
This is a serious misuse of the BLS data for several reasons. First, the BLS itself explains (p. 6) that "The BLS employment projections should be understood to be a projection and not a forecast. The distinction involves an emphasis on purpose and results. Projections focus on longer term underlying trends based on a set of assumptions, whereas forecasts focus more on predicting actual outcomes in the near term." Patrice (like many others) is using the BLS data as a forecast, not a projection.
As is appropriate for their purpose, the BLS doesn’t actually give the number of new lawyer jobs that it is projecting for any given year. Rather, it projects the number of new jobs over a ten year period. The 19,650 number that Patrice cites is derived from dividing the 196,500 new lawyer jobs that the BLS projects over the next ten years by 10. (Look at this table. Go to line 23-1010. The 196.5 in the far right column is the source of the 196,500 number. Note, by the way, that using this particular line excludes judicial law clerks (23-1012) and judges (23-0120), so it excludes some legal jobs). To be fair, Patrice accurately describes the projection. But he treats it like a forecast, not a projection.
Second, because they are projections, the BLS numbers are inherently unreliable as forecasts. The BLS projections of legal jobs have been way off in the recent past. As I explained in my first post on this topic, the BLS projections made in 2002, based on an assumed growth rate of 17%, projected far more lawyer jobs for 2012 than there in fact were in 2012. There is nothing at all wrong with this – long term projections are often off by a wide margin. This is why we shouldn’t use projections as forecasts.
Third, there is a serious mismatch between the number of new jobs projected by the BLS and the nine-month numbers reported to the ABA. Recall from the discussion above that the most recent ABA nine month report had 26,653 full time, bar pass required jobs. There is a big difference between 19,650 and 26,653. Think about how these two numbers are generated. The 19,650 is derived from dividing the BLS’s ten year projected growth by ten. There is no relationship, by the way, between the BLS lawyer numbers and having a J.D. or passing the bar. Rather, the BLS numbers are generated by surveys filled out by people in human resources departments, and depend on these people’s discretion in identifying someone as a lawyer. The ABA nine month data, in contrast, is based on surveys of actual graduates. Although, as I explained above, I think that the nine-month data undercounts the total number of law jobs, it presents an accurate picture of employment outcomes nine months after graduation.
In sum, the BLS projections do not in any way undercut my analysis that there will be a shortage of entry level lawyers in just a few years. Indeed, the BLS projections strengthen my case. Why? The BLS is projecting 9.8% growth in legal jobs over the next ten years, or about 1% per year. The BLS, by the way, does not forecast the business cycle (p. 6), and instead focuses on long-term growth trends. So this projected growth is not tied to a forecast that the economy will improve over the next few years.
The Structural Change Argument
The second argument often made by legal job market pessimists is that the legal industry is undergoing structural change driven by improvements in technology and process engineering. There are two forms of the structural change argument. The first form argues that long-term structural change is happening. I agree with this form of the argument. The second form argues that structural change is a cause of the poor job market of the past five years, and that structural change will reduce the future number and type of legal jobs. I am skeptical of this second form of the argument.
I’ve explained before why I am skeptical that structural change had a major impact on the job market in the last few years. Among other things, I saw many of the changes that people point to – e.g., predictive coding and outsourcing – when I entered practice in the late 1990s. Having seen these changes then leads me to be sympathetic to the argument that these changes are happening, but skeptical that they suddenly created massive job losses beginning in 2008. Instead, those job losses were likely caused by the recession that followed the financial crisis. Legal employment is not immune from the business cycle. Nor is perception of the legal economy immune from the business cycle. To me, the argument that structural change is causing, or will cause, massive job losses is exactly the type of argument that gains currency at both ends of the business cycle. In the last boom, many smart people were predicting that technological innovation would lead to increases in productivity that would lead to endless growth. Technology did continue to develop, but a recession came anyway. The idea that structural change caused massive job losses in the past five years strikes me as the same kind of thinking, albeit in the opposite direction.
I’ll admit here that I can’t know for sure that structural change won’t undercut future legal job growth. I saw very little evidence of structural change in my alumni data study, but things can change. I don’t find the arguments for massive job losses due to structural change persuasive, and I haven’t seen any persuasive evidence in support of this form of the argument. I don't think there is any concrete evidence, however, that will prove the impact of structural change over the next decade one way or the other.
Although we can’t prove the future impact of structural change in either direction, I think that there is one major reason why the structural change argument does not undercut my analysis. As I explained in the first part of the post, my analysis is based on an assumption of no job growth. It is not dependent, for example, on returning to the levels of legal hiring that we saw in the last boom. Further, under my analysis, we end up with a shortage even if legal employment rates decline modestly. I am not aware of any serious commentator who predicts that legal employment will go down in the next three years. To be sure, some people suggest that structural change will impede job growth, but that is a different argument. Most observers tend to predict modest growth over the next few years. Again, however, my analysis is based on no growth, and holds even if there is a modest decline.
Comments policy: As I noted at the outset, I welcome criticism. I hope that people making comments will presume good faith on my part, and I will presume good faith in return. I may delete comments that are off-topic or that clearly rebut the presumption of good faith.