Labor Dept.: 32,300 New Lawyer Jobs by 2029, Turnover of 24 Percent

On September 1, 2020, the Bureau of Labor Statistics (BLS) released its employment projections for the next cycle, 2019-2029.

For 2019, the BLS’s Employment Projections program (EP program) estimates that there were 813,900 lawyer positions (as opposed to discrete lawyers) in the United States. This figure includes self-employed attorneys. In 2019, the EP program found 823,900 lawyer positions, a small decline from last year. According to the BLS’s Current Population Survey (CPS), in 2019, 1.24 million people worked as lawyers in the United States. The discrepancy between the CPS and the EP program has existed for some time. In their respective contexts, both figures are correct.

The BLS projects future employment trends in part to help job seekers evaluate career choices, and the projections play an outsized role in the BLS’s Occupational Outlook Handbook. Here is an illustration from various sources that converts the flows of law-school graduates and new lawyers into stocks that can be compared to BLS employment measures since the 1980s.

Click here to read more:

What the 2019-29 Data Show

Analyzing Table 1.10, between 2019 and 2029, the BLS estimates a total 399,000 net lawyer jobs will be created (39,900 in the mean-average year). Of those, 32,300 (8 percent) can be attributed to economic growth over the decade, reaching 846,300 lawyer jobs total in 2029 (4 percent growth). The annual average of occupational openings is lower than the 2018-29 cycle (45,700).

The remaining new lawyer jobs, 366,700, consist of “occupational separations,” separating into 174,000 lawyers leaving the workforce and 193,000 lawyers transferring to different jobs. Like all other occupations, lawyers leaving the labor force may be retiring or temporarily exiting for personal reasons, and lawyers transferring to other occupations may be moving up or down the occupational ladder, e.g. judges or retail clerks. On the other side of the equation, people taking lawyer jobs are new to the profession or reentering after either leaving the workforce or transferring in from a different occupation. The BLS does not track this information directly.

In the 2016-26 cycle, the BLS implemented its new methodology for estimating occupational replacements. The older projections used a “cohort-component” methodology that netted out the differences among workers in five-year age brackets. For example, if in 2016 there were 50,000 lawyers in the 56-65 age cohort, and 60,000 lawyers in the 46-55 age cohort, then it would project that the difference, 10,000 lawyers, would leave the occupation and would require replacement by 2026. Add together the differences among all the age cohorts, and one has the net replacement rate.

The BLS abandoned this methodology because it suffered from sampling errors for small age cohorts and assumed that the only workers who needed replacement were older workers. To replace it, the BLS adopted an “occupational separations” methodology, which uses historical data on workers who have left the occupation, identifies common characteristics among them, regresses those characteristics, generates a probability of separation for the whole occupation, and then applies that to the occupation. Consequently, the EP program now detects lawyers who change occupations before middle age. While the results give a higher number of annual job openings than before—especially as compared to the number of new law-school graduates or bar admits—they also indicate a substantial amount of turnover in the legal profession: 193,000 out of the base 813,900 (24 percent) lawyer jobs.

Here is the BLS’s most-recent link for discussing occupational separations and its new methodology.

Methodological note: It may seem wrong to multiply annual average rates (e.g. occupational transfers) by ten and then compare them to the number of jobs in the base year because one number is people in a single year and the other is people over ten years. I disagree for two reasons. One, the BLS used this methodology when comparing its occupational-replacement methodologies. The spreadsheet in question can be found on the Internet Archive. It’s titled, “Experimental 2012-22 data (XLSX),” and in cell K243 it estimates that according to the new methodology the “Occupational Separation Rate, 2012-2022” for lawyers would be 42.6 percent. That doesn’t mean the BLS is right to engage in that kind of comparison, but it gives me a safe harbor. Two, on an intuitive level the comparison makes sense: As of today, some lawyers will leave the profession over the next decade, though some may return outside the ten-year window. Many won’t leave in the current year, but the EP program gives us the tools to estimate how many will leave over the full time period. Partly the culprit is the BLS’s decision to mix data from the CPS and EP program, which means it’s reasoning about lawyer jobs from flesh-and-blood lawyers. Ideally, the BLS would take a snapshot of all people who worked in lawyer jobs over the last decade. It would contain all 813,900 lawyer jobs today plus everyone who transitioned to different occupations or left the labor force. I believe the result would be lower than the 813,900 (today) plus 366,700 (replacemetns) plus 32,300 (growth), but until we have that information, the BLS estimates are the best we have.

‘Strong Competition’ Continues

As with last cycle, the BLS published its biennial Occupations Outlook Handbook (OOH) along with the new employment projections. I’ve tracked its statements about job prospects for the legal profession going back to the 1990s. It’s always been negative, and the methodology change hasn’t affected this. I’ll print the text in full.

Despite the projected growth in new jobs for lawyers, competition for jobs should continue to be strong because more students are graduating from law school each year than there are jobs available. Some law school graduates who have been unable to find permanent positions turn to temporary staffing firms that place attorneys in short-term jobs. These firms allow companies to hire lawyers as needed and permit beginning lawyers to develop practical experience. Many other law school graduates and licensed lawyers end up finding work in other occupations or industries due to the difficulty in finding jobs with traditional legal employers.

Because of the strong competition, a law school graduate’s willingness to relocate and his or her practical experiences are becoming more important. However, to be licensed in another state, a lawyer may have to take an additional state bar examination.

While many new lawyers are hired each year by law firms, this does not guarantee stable employment in the profession. Newly hired lawyers, known as associates, must either advance within their firm or may be forced to leave, a practice commonly known as “up or out.” Those who leave law firms may find work as in-house counsel with companies, with government agencies, or as self-employed lawyers.


For more than a decade, the EP program has released its data biennially, so I was quite surprised to see employment projections only a year after its last update. I don’t know why the BLS chose to increase the publication rate. It may be that next year’s data will be distorted by the COVID-19 recession, or the EP program is relatively cheap to produce and it’s just plugging numbers gathered for other reasons into projections equations.

Anyway, this year’s projections continue the downward trend in predicted annual lawyer job growth. Two cycles ago, the EP program predicted 74,800 net new jobs, which fell to 50,100 last year, and now 32,300 this year. Note, though, that three cycles ago the figure was 43,800, so this year’s numbers aren’t a total deviation from the norm. I commented last year that not much has changed in the economy to warrant this kind of volatility.

The EP program did not predict how many lawyer jobs would exist in 2019, but it did for 2018 and 2020, 857,700 and 801,800, respectively. The Great Recession (aka the last time everyone called it “worst downturn since the Great Depression”) apparently affected the formula’s outputs considerably. Compared to the 2018 projection, we’re high by 1.5 percent, but we’re low by 5 percent compared to 2020. I won’t guess as to how the COVID-19 depression (the new “worst downturn since the Great Depression”) will affect the EP program’s numbers for next year, if the department chooses to run the numbers. I haven’t mentioned it yet, but 2019 is the first year that saw a decline in lawyer jobs since 2010, and there was no recession last year to speak of. There may be a general decline in legal employment, which might correlate to the ABA’s falling national lawyer counts, but I haven’t explored that yet.

In typical years I warn that lawyer turnover is a central issue that prospective law students should consider when choosing to attend law school. That’s still true for all the reasons I’ve written about, but COVID-19, which will hopefully be gone by next year, much less 2029, proves yet again that the risk of bad employment outcomes still falls on law students. To editorialize beyond the employment projections, the pandemic has all made legal education’s shortcomings a sort of farcical self-jumping shark. Some graduates are demanding diploma privilege rights not because bar exams don’t really correlate to legal ability but because of the disease. Some students are suing their law schools not because tuition is excessive, but because the money isn’t getting them in-person classes. The pandemic was not predictable outside of fuzzy probabilities, but few of these hundred-year floods will ever break in law students’ favor.

My comprehensive explanation of the various measures of law-school grads and lawyers can be found on this page (updated!). It also should contain any links I may have omitted in this post.

Last time, I evaluated the BLS’s separations methodology by comparing lawyers to occupations requiring similar levels of education. I won’t do that again now because I don’t think there’s anything new to learn from it. For this reason, though, readers may want to look at the 2016-26 entry to see my thoughts on that at the time. Speaking of which…

Past Coverage

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s