Fifty-two, counted two ways that had never met
The CLSB says 52 people entered the costs lawyer profession in 2025. Counting qualification dates on its register gives 52. Nobody published the second number, and nothing was adjusted to make the two agree.
By Rose Traviss · 29 September 2026
A register tells you who is authorised today. It does not tell you how many joined last year — but where a regulator records the date each person qualified, you can count them, and no regulator publishes the result.
This site does that for three of the small legal professions. The question nobody could answer until now is whether the count is any good.
It turns out there is a way to check. Two regulators publish their own intake figure, in an annual report, in a different document from the register. Set the two against each other and the method either survives or it does not.
Costs lawyers: exactly
The Costs Lawyer Standards Board’s Annual Report 2025 gives, inside a three-column panel, 52 new entrants to the costs law profession in 2025.
The CLSB’s register carries a qualification date against each of its 789 costs lawyers. Counting the ones dated 2025 gives 52.
Nothing was tuned to produce that. The register was collected in September 2026 and the report was published in July 2026; neither was consulted while building the other.
Patent and trade mark attorneys: consistently short, and it should be
IPReg publishes admissions to its registers most years. Counting registration dates on the register itself gives a smaller number every single time.
| Year | IPReg’s published admissions | Counted from today’s register | Gap |
|---|---|---|---|
| 2019 | 173 | 162 | −11 |
| 2020 | 116 | 108 | −8 |
| 2021 | 258 | 230 | −28 |
| 2022 | 169 | 168 | −1 |
| 2023 | 225 | 211 | −14 |
| 2025 | 200 | 197 | −3 |
That is the right direction, and it has to be. A register holds the people who are still on it. Anybody admitted in 2021 who has since retired, transferred, died or been removed is in IPReg’s admissions figure for 2021 and is not on today’s register to be counted.
So the gap is not error. The gap is the attrition since: between 0.6% and 10.9% of each entry cohort is no longer on the register. It does not fall neatly with age — the 2019 cohort has lost 6.4% and the 2022 cohort 0.6% — so what this measures is cohort-specific, not a simple decay curve.
The 2021 cohort has lost the most: 258 admitted, 230 remaining, 10.9% gone. 2021 was also by some way the largest intake in the series.
Why the costs lawyer match is the stronger test
The IPReg comparison confirms that the method behaves as it must. It cannot confirm the level, because any undercount would look like attrition.
The costs lawyer comparison can. The CLSB’s 52 is its own count of people entering in the calendar year; the register’s 52 is a count of qualification dates falling in that year. They are measuring the same thing from opposite ends, in different documents, and they agree to the unit.
What the count is worth
Three of the seven professions here have a date on their register that means “joined”. Two do not, and the difference is treated as load-bearing rather than convenient:
- Certificated enforcement agents. A certificate lasts two years and is re-applied for, so the only date on the register is the current certificate’s. Everybody appears to have joined within two years. Plotted as intake it would show a profession that quadrupled in 2025 and did not exist before 2024.
- Notaries. The Faculty Office’s register carries no date at all. Its Registrar’s annual report does: 51 notaries admitted in calendar 2025. That figure is published on the notaries page as the regulator’s own, labelled as a different measurement, and never drawn on the same line as a computed intake.
The figures
Register counts computed from the CLSB and IPReg registers as collected on 27 September 2026. Published figures read from the CLSB’s Annual Report 2025 and IPReg’s annual reports for 2019 to 2025, each with its source sentence on the reports page.
The costs lawyer register also shows 40 qualifying in 2026 to date and 27 in 2024; the CLSB has not yet published a comparable figure for either year.