Real estate
Associates · Global offices
Better than last year52.9%
See figures
2026: 32 more left than joined
2025: 68 more left than joined
36 fewer net departures than Q3 2025
36 fewer net departures than Q3 2025
Mosaic Intelligence | Q3 2026
Q3 2026 global attorney talent market report
The Big Law Attorney Talent Market in Q3 2026
Real estate looked 52.9% better than last Q3, tax 50.0% better and IP 15.1% better. More associates still left than joined in those practices. Across global offices overall, the associate result was 3.2% worse: 1,868 more left than joined, compared with 1,810 last year. Here's how it played out by practice, firm and city.
Big picture
Where the quarter looked better
By 'better,' I mean fewer net departures after counting both who joined and who left. Take bankruptcy: last year, 30 more associates left than joined. This year, it was 24. That's a 20% improvement.
Better than last yearWorse than last year
Tax
Associates · Global offices
Better than last year50.0%
See figures
2026: 19 more left than joined
2025: 38 more left than joined
19 fewer net departures than Q3 2025
19 fewer net departures than Q3 2025
Bankruptcy
Associates · Global offices
Better than last year20.0%
See figures
2026: 24 more left than joined
2025: 30 more left than joined
6 fewer net departures than Q3 2025
6 fewer net departures than Q3 2025
IP
Associates · Global offices
Better than last year15.1%
See figures
2026: 118 more left than joined
2025: 139 more left than joined
21 fewer net departures than Q3 2025
21 fewer net departures than Q3 2025
Year over year
Q3 compared with last year
Overall, the associate result across global offices was 3.2% worse than last Q3. This year, 1,868 more associates left than joined, compared with 1,810 last year. That's 58 more net departures. The partner and counsel results were worse too.
Better than last yearWorse than last yearLargest declines first
Counsel
Worse than last year41.1%2026: 223 more left than joined2025: 158 more left than joined65 more net departures than Q3 2025
Partners
Worse than last year4.2%2026: 248 more left than joined2025: 238 more left than joined10 more net departures than Q3 2025
Associates
Worse than last year3.2%2026: 1,868 more left than joined2025: 1,810 more left than joined58 more net departures than Q3 2025
See how many joined and left
| Group | 2025 joined | 2026 joined | 2025 left | 2026 left |
|---|---|---|---|---|
| Counsel | 453 | 567More joined25.2% | 611 | 790More left29.3% |
| Partners | 669 | 804More joined20.2% | 907 | 1,052More left16.0% |
| Associates | 2,076 | 2,549More joined22.8% | 3,886 | 4,417More left13.7% |
See the first nine months
| Group | How it changed | What changed | 2025 | 2026 |
|---|---|---|---|---|
| Associates | Better than last year1.1% | 44 fewer net departures than January to September 2025 | 3,908 more left than joined | 3,864 more left than joined |
| Partners | Worse than last year8.7% | 66 more net departures than January to September 2025 | 759 more left than joined | 825 more left than joined |
| Counsel | Worse than last year1.2% | 7 more net departures than January to September 2025 | 560 more left than joined | 567 more left than joined |
See who joined and who left
| Group | 2025 joined | 2026 joined | 2025 left | 2026 left |
|---|---|---|---|---|
| Associates | 6,834 | 7,957More joined16.4% | 10,742 | 11,821More left10.0% |
| Partners | 2,417 | 2,689More joined11.3% | 3,176 | 3,514More left10.6% |
| Counsel | 1,473 | 1,744More joined18.4% | 2,033 | 2,311More left13.7% |
The first nine months looked 1.1% better. More global associates still left than joined, but the net result went from 3,908 departures to 3,864 through September 30. That's 44 fewer net departures than in the same period of 2025.
So the quarter was a bit worse, while the first nine months were a bit better. The breakdowns below show how many joined and left in each period.
London went the other way. Last Q3, 44 more associates left than joined. This Q3, 2 more joined than left. That's a swing of 46.
Practice areas
How the practices compare
Real estate and tax had the biggest improvements among the global practices shown here: 52.9% better in real estate and 50.0% better in tax. More associates still left than joined in both. Bankruptcy and IP improved too.
Better than last yearWorse than last yearBiggest improvements, then largest declines
Real estate
Better than last year52.9%2026: 32 more left than joined2025: 68 more left than joined36 fewer net departures than Q3 2025
Tax
Better than last year50.0%2026: 19 more left than joined2025: 38 more left than joined19 fewer net departures than Q3 2025
Bankruptcy
Better than last year20.0%2026: 24 more left than joined2025: 30 more left than joined6 fewer net departures than Q3 2025
Intellectual property
Better than last year15.1%2026: 118 more left than joined2025: 139 more left than joined21 fewer net departures than Q3 2025
Banking
Worse than last year112.8%2026: 100 more left than joined2025: 47 more left than joined53 more net departures than Q3 2025
Litigation
Worse than last year26.0%2026: 871 more left than joined2025: 691 more left than joined180 more net departures than Q3 2025
Corporate
Worse than last year4.6%2026: 618 more left than joined2025: 591 more left than joined27 more net departures than Q3 2025
Labor and employment
Worse than last year4.4%2026: 47 more left than joined2025: 45 more left than joined2 more net departures than Q3 2025
See how many associates joined and left by practice
| Group | 2025 joined | 2026 joined | 2025 left | 2026 left |
|---|---|---|---|---|
| Real estate | 111 | 161More joined45.0% | 179 | 193More left7.8% |
| Tax | 51 | 70More joined37.3% | 89 | 89Same number left0.0% |
| Bankruptcy | 51 | 50Fewer joined2.0% | 81 | 74Fewer left8.6% |
| Intellectual property | 75 | 109More joined45.3% | 214 | 227More left6.1% |
| Banking | 190 | 202More joined6.3% | 237 | 302More left27.4% |
| Litigation | 673 | 638Fewer joined5.2% | 1,364 | 1,509More left10.6% |
| Corporate | 545 | 704More joined29.2% | 1,136 | 1,322More left16.4% |
| Labor and employment | 156 | 215More joined37.8% | 201 | 262More left30.3% |
Real estate and tax looked better
Real estate was 52.9% better than last year: 32 more associates left than joined, down from 68. That's 36 fewer net departures. Tax was 50.0% better, with 19 more leaving than joining instead of 38.
IP and bankruptcy improved too
IP was 15.1% better: 118 more associates left than joined, compared with 139 last year. That's 21 fewer net departures. Bankruptcy was 20.0% better, going from 30 to 24 net departures, a reduction of 6.
Corporate and employment slipped a little
Corporate was 4.6% worse: 618 more associates left than joined, compared with 591 last year. That's 27 more net departures. Employment was 4.4% worse, going from 45 to 47 net departures, an increase of 2.
These are selected practice categories. We use the full firm totals for the overall figures and keep the practice comparisons separate.
Geography
The picture within major US cities
Chicago was 22.7% better and Dallas 5.7% better among the nine cities shown here. More associates still left than joined in both. The other seven cities were worse. Together, the nine cities were 29.3% worse, with 1,418 net departures compared with 1,097 last year, an increase of 321.
Better than last yearWorse than last yearBiggest improvements, then largest declines
Chicago
Better than last year22.7%2026: 92 more left than joined2025: 119 more left than joined27 fewer net departures than Q3 2025
Dallas
Better than last year5.7%2026: 50 more left than joined2025: 53 more left than joined3 fewer net departures than Q3 2025
Atlanta
Worse than last year93.3%2026: 29 more left than joined2025: 15 more left than joined14 more net departures than Q3 2025
Washington, D.C.
Worse than last year72.7%2026: 316 more left than joined2025: 183 more left than joined133 more net departures than Q3 2025
New York
Worse than last year36.2%2026: 617 more left than joined2025: 453 more left than joined164 more net departures than Q3 2025
San Francisco
Worse than last year31.5%2026: 96 more left than joined2025: 73 more left than joined23 more net departures than Q3 2025
Boston
Worse than last year11.9%2026: 66 more left than joined2025: 59 more left than joined7 more net departures than Q3 2025
Los Angeles
Worse than last year8.7%2026: 112 more left than joined2025: 103 more left than joined9 more net departures than Q3 2025
Houston
Worse than last year2.6%2026: 40 more left than joined2025: 39 more left than joined1 more net departure than Q3 2025
See how many associates joined and left in these cities
| Group | 2025 joined | 2026 joined | 2025 left | 2026 left |
|---|---|---|---|---|
| Chicago | 95 | 96More joined1.1% | 214 | 188Fewer left12.1% |
| Dallas | 41 | 46More joined12.2% | 94 | 96More left2.1% |
| Atlanta | 38 | 42More joined10.5% | 53 | 71More left34.0% |
| Washington | 194 | 154Fewer joined20.6% | 377 | 470More left24.7% |
| New York | 406 | 444More joined9.4% | 859 | 1,061More left23.5% |
| San Francisco | 53 | 75More joined41.5% | 126 | 171More left35.7% |
| Boston | 43 | 59More joined37.2% | 102 | 125More left22.5% |
| Los Angeles | 97 | 98More joined1.0% | 200 | 210More left5.0% |
| Houston | 49 | 80More joined63.3% | 88 | 120More left36.4% |
See the nationwide U.S. picture
| Role | Q3 2026: joining vs. leaving |
|---|---|
| Associates | 1,712 more left than joined |
| Partners | 201 more left than joined |
| Counsel | 146 more left than joined |
See who joined and who left
| Group | 2026 joined | 2026 left |
|---|---|---|
| Associates | 1,604 | 3,316 |
| Partners | 611 | 812 |
| Counsel | 466 | 612 |
Chicago and Dallas improved
Chicago was 22.7% better: 92 more associates left than joined, down from 119 last year. That's 27 fewer net departures. Dallas was 5.7% better, with 50 net departures instead of 53, a reduction of 3.
New York was worse than last year
New York was 36.2% worse: 617 more associates left than joined, compared with 453 last Q3. That's 164 more net departures. The figures combine all practices, so they don't tell us which specialty drove the change.
Your reason for moving still matters
San Francisco was 31.5% worse than last year, with 23 more net departures. Boston was 11.9% worse, with 7 more, and Houston was 2.6% worse, with 1 more. A particular team could still make sense for you. That's where your practice and reasons for moving come in.
D.C. needs a little context
D.C. was 72.7% worse: 316 more associates left than joined, compared with 183 last year. That's 133 more net departures. But 2025 was unusual: firms took in a wave of government lawyers amid the second Trump administration's agency changes and DOGE workforce reductions. I keep that in mind when reading this comparison. The data doesn't separate those government moves, so we can't say how much they explain the change.
The nine cities are New York, Washington, Los Angeles, Chicago, San Francisco, Houston, Dallas, Boston and Atlanta. Their current figures match the file covering U.S. offices. Last year's figures come from matching city entries in the earlier data. These nine cities are separate from the nationwide U.S. total and the global figures.
Our areas of concern
Our areas of concern
Banking and litigation both looked worse for associates than last year. Banking had the larger percentage change. Litigation had the bigger increase in net departures.
Better than last yearWorse than last yearLargest declines first
Banking
Worse than last year112.8%2026: 100 more left than joined2025: 47 more left than joined53 more net departures than Q3 2025
Litigation
Worse than last year26.0%2026: 871 more left than joined2025: 691 more left than joined180 more net departures than Q3 2025
See how many associates joined and left in these practices
| Group | 2025 joined | 2026 joined | 2025 left | 2026 left |
|---|---|---|---|---|
| Banking | 190 | 202More joined6.3% | 237 | 302More left27.4% |
| Litigation | 673 | 638Fewer joined5.2% | 1,364 | 1,509More left10.6% |
Banking looked worse than last year
Banking was 112.8% worse: 100 more associates left than joined, compared with 47 last Q3. That's 53 more net departures.
Litigation was worse than last year
Litigation was 26.0% worse: 871 more associates left than joined, compared with 691 last year. That's 180 more net departures. I'd want to get much more specific before applying that to your practice. The data groups white collar, securities, antitrust and commercial disputes together.
Practice figures cover global offices in Q3 2026 and Q3 2025. City figures cover the nine U.S. cities shown here.
Firm deep dive
How the firms compare
Six of these eight firms looked better than last Q3. More associates left than joined at all eight, but six had fewer net departures. These figures include each firm's global offices and all its practices.
Better than last yearWorse than last yearBiggest improvements, then largest declines
Skadden
Better than last year82.6%2026: 8 more left than joined2025: 46 more left than joined38 fewer net departures than Q3 2025
Gibson Dunn
Better than last year63.6%2026: 16 more left than joined2025: 44 more left than joined28 fewer net departures than Q3 2025
DLA Piper
Better than last year42.2%2026: 52 more left than joined2025: 90 more left than joined38 fewer net departures than Q3 2025
Cooley
Better than last year26.7%2026: 22 more left than joined2025: 30 more left than joined8 fewer net departures than Q3 2025
Latham & Watkins
Better than last year14.8%2026: 52 more left than joined2025: 61 more left than joined9 fewer net departures than Q3 2025
Kirkland & Ellis
Better than last year6.1%2026: 62 more left than joined2025: 66 more left than joined4 fewer net departures than Q3 2025
Paul Hastings
Worse than last year740.0%2026: 42 more left than joined2025: 5 more left than joined37 more net departures than Q3 2025
Weil
Worse than last year100.0%2026: 46 more left than joined2025: 23 more left than joined23 more net departures than Q3 2025
See how many associates joined and left at each firm
| Group | 2025 joined | 2026 joined | 2025 left | 2026 left |
|---|---|---|---|---|
| Skadden | 31 | 53More joined71.0% | 77 | 61Fewer left20.8% |
| Gibson Dunn | 30 | 53More joined76.7% | 74 | 69Fewer left6.8% |
| DLA Piper | 45 | 95More joined111.1% | 135 | 147More left8.9% |
| Cooley | 21 | 34More joined61.9% | 51 | 56More left9.8% |
| Latham & Watkins | 77 | 97More joined26.0% | 138 | 149More left8.0% |
| Kirkland & Ellis | 67 | 115More joined71.6% | 133 | 177More left33.1% |
| Paul Hastings | 38 | 28Fewer joined26.3% | 43 | 70More left62.8% |
| Weil | 25 | 22Fewer joined12.0% | 48 | 68More left41.7% |
Explore the full firm comparison
Pick a firm to see how joining and leaving compared with last year. You can switch between associates, partners and counsel, or look at the first nine months instead of Q3. These figures cover global offices.
Showing 12 of 100 firms. Biggest improvements first. Missing comparisons are at the bottom.
| Firm | How it changed | What changed | 2025 | 2026 |
|---|---|---|---|---|
| Steptoe LLP | More joined than left6change from last Q3 | Improved by 6 vs. Q3 2025 | 1 more left than joined | 5 more joined than left |
| Ogletree Deakins | More joined than left4change from last Q3 | Improved by 4 vs. Q3 2025 | 2 more left than joined | 2 more joined than left |
| Holland & Knight LLP | More joined than left17change from last Q3 | Improved by 17 vs. Q3 2025 | 11 more left than joined | 6 more joined than left |
| Blank Rome LLP | More joined than left3change from last Q3 | Improved by 3 vs. Q3 2025 | 2 more left than joined | 1 more joined than left |
| Littler Mendelson PC | More joined than left21change from last Q3 | Improved by 21 vs. Q3 2025 | 15 more left than joined | 6 more joined than left |
| Sheppard | More joined than left17change from last Q3 | Improved by 17 vs. Q3 2025 | 13 more left than joined | 4 more joined than left |
| King & Spalding LLP | More joined than left31change from last Q3 | Improved by 31 vs. Q3 2025 | 26 more left than joined | 5 more joined than left |
| Ballard Spahr LLP | More joined than left10change from last Q3 | Improved by 10 vs. Q3 2025 | 9 more left than joined | 1 more joined than left |
| Dechert LLP | More joined than left23change from last Q3 | Improved by 23 vs. Q3 2025 | 21 more left than joined | 2 more joined than left |
| Katten Muchin Rosenman LLP | More joined than left15change from last Q3 | Improved by 15 vs. Q3 2025 | 14 more left than joined | 1 more joined than left |
| Jackson Lewis PC | More joined than left16change from last Q3 | Improved by 16 vs. Q3 2025 | 15 more left than joined | 1 more joined than left |
| Cozen O'Connor | Better than last year100.0% | 1 more net arrival than Q3 2025 | 1 more joined than left | 2 more joined than left |
See who joined and who left
| Group | 2025 joined | 2026 joined | 2025 left | 2026 left |
|---|---|---|---|---|
| Steptoe LLP | 5 | 11More joined120.0% | 6 | 6Same number left0.0% |
| Ogletree Deakins | 17 | 35More joined105.9% | 19 | 33More left73.7% |
| Holland & Knight LLP | 23 | 39More joined69.6% | 34 | 33Fewer left2.9% |
| Blank Rome LLP | 9 | 13More joined44.4% | 11 | 12More left9.1% |
| Littler Mendelson PC | 22 | 40More joined81.8% | 37 | 34Fewer left8.1% |
| Sheppard | 12 | 21More joined75.0% | 25 | 17Fewer left32.0% |
| King & Spalding LLP | 18 | 38More joined111.1% | 44 | 33Fewer left25.0% |
| Ballard Spahr LLP | 7 | 13More joined85.7% | 16 | 12Fewer left25.0% |
| Dechert LLP | 14 | 28More joined100.0% | 35 | 26Fewer left25.7% |
| Katten Muchin Rosenman LLP | 5 | 14More joined180.0% | 19 | 13Fewer left31.6% |
| Jackson Lewis PC | 22 | 51More joined131.8% | 37 | 50More left35.1% |
| Cozen O'Connor | 12 | 23More joined91.7% | 11 | 21More left90.9% |
Skadden was 82.6% better than last year. This year, 8 more associates left than joined, compared with 46 last Q3.
I'd also put Skadden's result in context. The firm faced intense negative scrutiny in the press last year. My sense is that the press coverage and market conversation have felt quieter this year, which could explain some of this change. Background on Skadden's 2025 scrutiny
Gibson Dunn and DLA Piper also looked better. Gibson was 63.6% better, going from 44 to 16 net departures. DLA was 42.2% better, going from 90 to 52. Cooley, Latham and Kirkland improved too.
Weil was 100.0% worse, going from 23 to 46 net departures. Paul Hastings was 740.0% worse, going from 5 to 42. That percentage is so large because last year's net result was just 5 departures.
These figures stop at September 30. Anything that happened after that will show up in a later period.
These are a few global examples. They haven't been adjusted for firm size or mergers. A small starting number can make a percentage change look very large.
Mosaic's perspective
I'll keep you posted
I'll share another update when there's something meaningful to add. And if you're wondering what this means for your own practice, come to us. We're glad to help you think it through!
Your move
What this means for you
If you're thinking about a move, I'd start with what you want more of and what's missing where you are. The numbers give us some context. Then we need to talk about the actual work and the people you'd be working with.
Tell me what work you want more of
Real estate and tax both cover a lot. Two lawyers with the same title can be doing very different things. I'll want to hear about your matters and what you'd like more exposure to.
Let's get specific about IP and restructuring
In IP, patent work, soft IP and transactions can take you in different directions. In restructuring, debtor and creditor work matter too. Those details help us figure out which teams make sense.
Get to know the people you'd work with
That matters in corporate and employment too. Who would you work with? How much responsibility would you get? Those are things I'd want to understand before you commit to a move.
A new city could still make sense
You might have a very good reason to move even if that city's overall figures got worse. We'll start with your practice and why you want to go there.
About the data
If you'd like the details
The numbers come from SurePoint Legal Insights and cover global offices at selected Am Law 100 firms. Q3 means July 1 through September 30. The first nine months means January 1 through September 30 in each year. We put this report together on October 28, 2026.
Underlying data provided by SurePoint Legal Insights. Analysis and commentary by Mosaic Search Partners.
Which offices are included?
The overall, practice, firm and role figures include global offices, including the U.S. The city section covers nine U.S. cities. We also have a separate file covering U.S. offices, but it only gives us this quarter's nationwide total. It doesn't include nationwide comparisons for last year, the first nine months, practices or firms.
How do we work out better or worse?
We subtract departures from arrivals to get net attorney flow. When the answer is negative, more people left than joined. We compare the change with the size of last year's net result. As a hypothetical example, 20 more leaving than joining instead of 30 means 10 fewer net departures, or 33.3% better. If last year's figure is zero or missing, we show n/a. Percentages are rounded to one decimal.
Why don't the practice figures add up exactly?
In the full source file, the practice categories add up to three less than the firm total in each year. The charts here show selected practices, and we use the full firm totals for the overall figures. There are 99 firm names in the Q3 2026 file and 100 in Q3 2025. We use the names supplied. We haven't included the General Practice category in the opening comparisons because we need more detail on what's in it.
What's grouped together?
Each practice includes different kinds of work. Each firm figure includes multiple offices and teams. The city figures combine practices. That's why I get more specific when we're talking about a particular role or move.
How did we check the city comparison?
All eight movement columns for the nine current U.S. cities match the file covering U.S. offices. Last year's city figures come from the earlier data. Some place names appear more than once without a state or country; we've kept those entries as supplied. Our Q2 report covered U.S. offices only, so its overall totals cover a different geography. Two Q3s also aren't enough to establish a normal seasonal pattern.
Why are the nationwide and nine city totals different?
The U.S. file lists 3,316 associate departures nationwide. The nine cities shown here account for 2,512 of those departures. The rest are in other U.S. locations.
FAQ
A few things you might be wondering
Here's how I'd explain it if we were talking through the report.
What does better or worse mean here?
It's a comparison of the net result after counting both arrivals and departures. Across global offices, 1,868 more associates left than joined in Q3 2026, compared with 1,810 last Q3. That's 58 more net departures, or 3.2% worse.
What would a better result look like?
As a hypothetical example, if 30 more people left than joined last year and it's 20 this year, that's 10 fewer net departures, or 33.3% better. In the actual global IP figures, 118 more associates left than joined instead of 139. That's 15.1% better.
Which practices looked better?
Real estate, tax, bankruptcy and IP looked better among the global practices shown here, with fewer net departures. Corporate and employment slipped a little. Banking and litigation had bigger increases in net departures.
Did the first nine months look better or worse?
A little better. Global associates went from 3,908 net departures to 3,864, which is 44 fewer, or 1.1% better. Q3 on its own was 3.2% worse, with 58 more net departures.
What about litigation?
Litigation was 26.0% worse across global offices. There were 871 more associates leaving than joining, compared with 691 last year, or 180 more net departures. I'd want to get into your particular specialty before drawing conclusions about a move.
Which cities looked better?
Chicago and Dallas. Chicago was 22.7% better, going from 119 to 92 net departures. Dallas was 5.7% better, going from 53 to 50. The other seven U.S. cities shown here were worse.
Does this usually happen in Q3?
More global associates left than joined in both Q3 2025 and Q3 2026. We'd need more years to say whether that's a normal Q3 pattern.
Does better mean fewer people left?
Not necessarily. Both joining and leaving matter. IP departures actually rose from 214 to 227, but arrivals rose from 75 to 109, so the net result was better. In bankruptcy, 7 fewer left, from 81 to 74, and 1 fewer joined, from 51 to 50. That meant 6 fewer net departures, from 30 to 24, or 20.0% better.
Can I compare this with Q2?
The overall totals cover different places. Q2 covered U.S. offices only; this report includes global offices. In the separate U.S. figures, 1,712 more associates left than joined this Q3. We weren't given last year's nationwide figure to compare it with.
See the figures for offices outside the U.S.
Outside the U.S.
This part only covers offices outside the U.S. The global figures above include both U.S. and overseas offices.
| Role | Q3 2026: joining vs. leaving |
|---|---|
| Associates | 156 more left than joined |
| Partners | 47 more left than joined |
| Counsel | 77 more left than joined |
See who joined and who left
| Group | 2026 joined | 2026 left |
|---|---|---|
| Associates | 945 | 1,101 |
| Partners | 193 | 240 |
| Counsel | 101 | 178 |
We took the U.S. totals out of the global totals for Q3 2026 and checked the result against the remaining locations. We don't have a complete U.S. total for Q3 2025, so we can't make the same comparison for last year.