The collection year: what each return is for

There is a return every month. Each one sends the whole year again, not just what changed.

Most are routine. Three are not: the December one, the early summer one, and the final one.

• R04, in early December: the main source of next year's 16 to 19 student numbers.
• R10, in early June: the year-end forecast funding claim, with the final claim after R14.
• R14, in the autumn: the final hard close for the year.

Because every return restates the year, a correction made in month nine changes what was reported in month two. That is why earnings move in ways that look inexplicable until you understand the mechanism.

It is also why the discipline of monthly correction pays off. Small, continuous restatements are absorbed without comment. A single large one at year end is a finance conversation.

Each collection period submission contains the full in-year population and funding is recalculated across all periods, so period earnings are restated on every submission.

Deadlines come from the data collection timetable for the funding year. The Academy calendar renders those dates and marks them provisional until they are reconciled against the published timetable.

December is the main source of next year's 16 to 19 student numbers. June gives the year-end forecast for adult skills. The final return closes the year's data.

R04 is different because 16 to 19 funding is lagged: for FE colleges, the students it records are the main basis of next year's student numbers. An error there costs money in a year that has not been budgeted yet.

R10 is different because it is late enough for the year to be known and early enough for something to be done about it. For adult skills it supports the year-end forecast funding claim; the final reconciliation uses the R14 final claim.

R14 is different because it is final. The year's ILR data cannot be amended once it closes, and that data is what the final funding position, the achievement rates and the audit population are taken from. Funding can still be recovered afterwards where the data proves to be wrong.

Each of the three has a dedicated explainer in this Academy with its full drives, role actions, checks, sign-off and post-submission steps, because the specification requires their purpose not to be reduced to a sentence.

The readiness checklist generator produces a return-specific checklist for each of them, scoped by stream and role.

The same five checks every month prevent most of what goes wrong at the three big returns.

• Anyone still in learning past their planned end date?
• Any aim that ended with no outcome?
• Did validation produce anything new?
• Does the funding report match what we expected, and if not, why?
• Has every leaver curriculum told us about actually been recorded?

The point of the routine is that it is boring and it is the same every month. A college that does these five things consistently arrives at R14 with very little to do.

A college that does them only before the big returns arrives at R14 with a year of accumulated correction and no time.

Each check corresponds to a report you can produce from your own data. Track the count for each one period on period: a rising count is a control failure, not a workload problem.