Failing progression but passing the rest?

The Office for Students measures three things about a course under condition B3: whether students stay past the first year, whether they finish, and where they are fifteen months after leaving. It calls these continuation, completion and progression, and sets a pass mark for each.

Progression is the one people quote, usually as a 60% pass mark, and it behaves differently from the other two.

Before anything else: being below a pass mark is not the same as being under investigation. Conditions are recorded separately, on the OfS register, and of the 48 providers below the progression line on the first degree slice, six carried a specific B3 condition when we checked it on 26 September 2026. The Method says how to reproduce that count.

Who falls below progression

The pattern holds on both undergraduate slices, and it is stronger on the one where most further education sits.

On first degrees, 30 of the 48 providers below the pass mark are clearing the other two. On other undergraduate provision, which is where foundation degrees, HNCs and HNDs live, it is 27 of 32. These providers recruit students and get them through the course, and fall short only on the destination measure.

The OfS does not judge condition B3 on the headline figure alone. It asks how much of the statistical uncertainty around a provider’s indicator sits below the line, which matters most where cohorts are small. The median behind the 48 below the line is 190 students, against 460 across all 247 providers. Of the 48 below the pass mark, 41 are below it with at least 75% confidence and 34 with at least 90%. On other undergraduate provision the uncertainty bites harder: 14 of the 32 reach 75%.

Progression also comes from a survey, so it carries a response rate. The median across all 247 providers is 52.9%. For the 48 below the pass mark it is 46.3%, so the figures that fail are resting on fewer replies.

Is progression the harder test?

On first degrees, progression’s pass mark sits further below the sector median than continuation’s does.

So progression gives providers more room than continuation does, and still catches nearly twice as many: 19% fall below it, against 11% for continuation and 11% for completion. The providers that fail it also fail it by more, a median of 7.8 points below the line against 2.9 and 3.3.

The other undergraduate slice separates the same way for a different reason. There, 14% fall below on progression and 15% below on continuation, but only 4.5% below on completion. Providers on this slice are losing people in the first year and placing the rest poorly, while the finishing rate holds up.

Continuation and completion are about the course. Progression is about what happens after it, which at most colleges is a different team and a smaller one. At one provider in this dataset, higher education employability is one post covering roughly 500 students.

So a college failing progression alone is being told to fix something its teaching staff do not control.

By provider type

The chart below joins the same slice to the OfS register and splits it by provider type.

A university has about a one in twenty-five chance of sitting below this line. A provider without university title has about one in three. Whatever the measure is picking up, it lands on one part of the sector.

Splitting the 129 again separates further education from private and specialist providers, using charity status as the marker, because the register carries no provider type field. The 84 that are exempt charities, which is most colleges, have 28 below the line. The other 45 have 15. Both work out at 33%, so this is not a further education problem specifically. It tracks with not holding university title.

We cannot tell you from this data why. Cohort size, subject mix and who each type recruits all differ, and any of them could carry it. But 96% of universities clear this measure and a third of everyone else does not.

The OfS is removing this threshold from 2027-28

In June 2026 the OfS published the outcome of its consultation on quality regulation. It will remove the progression indicator and its thresholds from the requirements of condition B3, at paragraph 145a. Minimum requirements will cover continuation and completion only.

The measure itself stays. It moves into the benchmarked TEF indicators, where two more post-study measures join it, at paragraph 166b: what graduates say about using what they learned, and a salary measure from the Longitudinal Education Outcomes dataset.

None of this is in force. A second consultation runs in autumn 2026 and the first assessments under the new scheme are in 2027-28. Until then condition B3 works as described above.

The threshold ends in 2027-28 and the measure continues, with two more post-study measures beside it, all benchmarked, and all sitting with the same team.

Check which one you are failing before you spend anything

Your own three numbers are in the same public dataset, on the OfS student outcomes dashboard. Filter to your provider, then read continuation, completion and progression as three separate numbers.

If continuation or completion is below too, look at the course and student support first. Continuation at a widening participation provider turns on money, caring responsibilities and hours worked as much as on teaching. If progression alone is below, an improvement plan aimed at teaching will report activity without moving the number. Most providers below this line are in that second case.

The split also decides who owns it. In most colleges continuation sits with curriculum and progression with careers and employability.

If progression is the one below, four things are worth checking.

• The subject splits on the dashboard, which show whether one area is carrying the provider-level figure.
• The response rate behind your own number, in the same dataset.
• How your graduates’ jobs are coded, and whether the work they go into counts as high skilled. Several providers already make this argument in their access and participation plans, and one TEF panel accepted a version of it for a land-based college, where many subjects do not lead to traditional degrees.
• Careers and employability staffing per student.

None of those four checks needs a supplier, and between them they tell you whether the number is a measurement problem, a subject-mix problem, or a staffing one.

This data cannot separate a weak local job market from weak provision, because both look identical here.

The coding argument fits what we found, though testing it properly needs subject-level analysis, which this note does not attempt.

Method

Source: the OfS student outcomes dataset, February 2026 release. Filters: taught or registered population; full-time; Overall split; progression and continuation for 2019-2022, completion for 2016-2019, which is the most recent window the dataset carries for each. Pass marks are the published values in the dataset itself. Confidence comes from the dataset’s own BELOW_THRESHOLD field, which the OfS defines as the proportion of the statistical uncertainty distribution falling below the threshold.

Every count, share and median on this page uses one population: the providers with all three indicators present on that slice. That is 247 at first degree and 222 at other undergraduate. The dataset also carries a sector-wide aggregate row, which is not a provider and is excluded. Provider type comes from the OfS register field recording whether a provider holds the right to use university in its title, not from its name; all 247 are matched on the register.

The count of providers carrying a specific B3 condition comes from the same register download, dated 26 September 2026: the Register sheet’s two columns headed “Specific condition of registration imposed (B3)”, counted across the 48 providers below the progression pass mark. Conditions that have been lifted sit on a separate sheet and are not counted. The further education split uses the register’s charity status field, where exempt charities are mostly further education and sixth form corporations, because the register has no provider type field. It is a proxy and it is imperfect: a few specialist higher education institutions sit on the exempt side. Both the register and the conditions in it change, so check the current file rather than this number.

Tell us if this is wrong

If a number here is wrong, or if it does not match what you see from inside a college, email research@acelro.com. We will correct the page and say what changed.

Acelro helps careers and employability teams give every student reviewed, hands-on work toward a target role, and show the evidence student by student. We work with universities and colleges.