Higher education regulation runs on documents you wrote long before

Registration and course approval are assessed on records. Almost none of the useful ones can be produced after the request arrives, which makes their creation an ongoing operational discipline rather than a project.

A higher education provider is judged on whether it can show things: that a course was designed with proper academic input, that outcomes are assessed at the level claimed, that a committee genuinely considered a decision rather than noting it, that a complaint was handled, that a review happened and something changed. Each of those is a statement about the past, and the evidence for it either exists already or does not exist at all.

That asymmetry is the single most important operational fact in this area, and it is understood late by nearly every provider. Effort concentrates on the submission, which is written in a defined period by identified people under pressure. The material that actually determines the outcome was generated, or not generated, in ordinary weeks over the preceding years by people who had no idea it would be examined.

The pattern is visible in what goes wrong. A provider can describe its academic governance clearly and cannot show minutes recording that the academic board weighed a decision rather than receiving it. It can explain how assessment is moderated and cannot produce the moderation records for the period in question. It can state that student feedback informs course review and cannot demonstrate a single change traceable to feedback. In each case the practice may have been perfectly sound and the evidence does not exist.

There is a specific and serious temptation at this point, and it should be named plainly. Records must not be created after the fact and presented as contemporaneous. Reconstructing minutes, backdating a review, or assembling a moderation file after a request is a far more serious matter than the original gap, and it is generally detectable, because documents created in one sitting share characteristics that documents accumulated over three years do not.

The honest response to a gap is to record it, say what has changed, and demonstrate the control operating from now. That is an uncomfortable position and it is recoverable. The alternative is not.

What follows from all of this is a fairly narrow operational discipline. Decide, for each thing the provider will need to be able to show, what record its operation will leave, and make that record a by-product of doing the thing rather than a separate task. A committee that records its reasoning in the minute as a matter of habit produces evidence continuously and at no cost. One that records decisions only produces a list that proves nothing about how they were reached.

The second discipline is retrievability. Records that exist and cannot be found within a reasonable period are close to useless in an assessment, and providers routinely discover that their evidence is spread across shared drives, individual mailboxes, a committee secretary's laptop and a system nobody has administrator access to any more. Knowing where each category lives, and being able to produce a period's worth on request, is a separate problem from having it.

One further point about who this burden falls on. The evidence described here is created by academics, committee secretaries, course coordinators and administrators in the course of doing something else, and almost none of them has been told which of their outputs will later be examined. A provider that has never explained to its own staff what the evidential requirements are is relying on those records existing by accident. A single briefing, saying plainly which artefacts matter and why, is the cheapest intervention available in this whole area and it is rarely given.