What grant and benefit applications actually ask you for
1,724 of our 2,216 records publish a list of the documents they want — 78%. Once you normalise them out of each country's own vocabulary, the same handful comes up again and again.
The list, by how often it is asked for
| Document | Programmes | Share |
|---|---|---|
| Proof of identity | 616 | 36% |
| Proof of income (payslip) | 274 | 16% |
| Medical evidence | 113 | 7% |
| Proof of address | 87 | 5% |
| Bank account details | 86 | 5% |
| Tax return or tax notice | 72 | 4% |
| Proof of enrolment | 69 | 4% |
| Benefit decision letter | 57 | 3% |
| Tenancy agreement | 54 | 3% |
| Birth certificate | 53 | 3% |
Counted per programme, not per mention: a scheme asking for two kinds of income evidence counts once. Requirements too specific to map onto a common type are excluded rather than lumped together.
Why this matters more than it looks
The list is short, and it is the same list. That means the marginal cost of the second application is far lower than the first — but only if you kept what you gathered. Most people do not, because each application arrives as its own self-contained ordeal with its own checklist, and nothing in the process suggests that the payslip you scanned in March is the same payslip wanted in June.
The expiry trap
Several of these have a validity window that nobody states on the form. Proof of address and proof of income are commonly expected to be under three months old; a tax return, under a year. A rejection for a document that is technically correct but four months old is one of the more demoralising ways to lose a claim, and one of the easiest to avoid once you know the window exists.
What we'd suggest
Gather the top five once, note the date on each, and treat the three-month ones as perishable. That single hour is worth more than any amount of searching for programmes, because it converts every future application from a document hunt into a form.
Every number above is computed, not typed. They come from the 2,216 records behind this site at the moment the page was built, so the post cannot drift away from the data. Where a programme publishes no figure we count it as zero rather than estimate — which makes the totals here understatements, and deliberately so.