What local authority budget data can and cannot tell you about strategic intent

The most common mistake analysts make when they open a local authority's statement of accounts is treating spending totals as a map of what the organisation actually cares about. A children's services directorate absorbs 38% of net expenditure, so children must be the priority — that is the logic, and it costs people real money when they act on it. Commissioners use this reasoning to benchmark councils against each other. Consultants use it to advise on restructuring. Think-tanks use it to rank authorities by apparent ambition. The forty-three-authority longitudinal study we carried out at Aurora Pioneer, tracking published accounts from 2016-17 through 2022-23, kept running into the wreckage that reasoning leaves behind. Spending shares shift for reasons that have almost nothing to do with political will, and the accounts, on their own, cannot tell you which reason applies.

Why the numbers move even when the strategy doesn't ¶

Across our sample, the single largest driver of year-on-year movement in service-level spending shares was demand volatility in statutory services — not policy change, not a new administration, not a deliberate reallocation. When an authority's looked-after-children numbers spike, the children's services budget inflates mechanically. The proportion of total spend attributed to that directorate rises. To an outside reader, this looks like a strategic pivot toward children. It is actually the council doing the minimum the law requires, under pressure, with no discretion involved.

We tested this by cross-referencing spending-share movements against the timing of full council elections and published corporate plan updates in all forty-three authorities. In only eleven cases did a shift of more than four percentage points in a directorate's spending share coincide with both a change in administration and a documented strategic restatement within twelve months. In twenty-six cases the shift had no corresponding strategic document at all. The accounts moved. The strategy, as far as the paper trail showed, did not.

What accounting classifications actually measure ¶

The CIPFA Service Reporting Code of Practice — SeRCOP, in the trade — is what determines how English and Welsh local authorities classify their expenditure, and it is designed for comparability and audit, not for revealing organisational intent. When a council reclassifies its community safety team from 'Housing' to 'Central Services' because a shared-services arrangement changed the team's line-management structure, total spending in both headings shifts. Both shifts are real in an accounting sense. Neither one tells you anything about whether the council considers community safety more or less important than it did the previous year.

Three of the authorities in our study underwent significant SeRCOP reclassification exercises during the study period — two because of internal restructures, one because of a boundary commission review that transferred functions. In each case the published accounts showed double-digit percentage-point swings in at least two service headings. Analysts who picked up those accounts cold, without reading the explanatory foreword in full, produced briefings that described strategic realignment where there was administrative tidying.

The foreword matters more than most analysts give it credit for, but it is also the section that gets written closest to the publication deadline, often by finance officers under pressure, and it is not audited in the same way as the main statements. Treat it as a useful pointer, not as verified evidence.

The accounts moved. The strategy, as far as the paper trail showed, did not.

The ring-fencing problem and what it obscures ¶

A substantial portion of what appears in local authority accounts as discretionary spending is not discretionary at all, because it is funded by ring-fenced grants that arrive with conditions attached. The Better Care Fund, the Household Support Fund during the cost-of-living period, specific Department for Levelling Up capital allocations — all of these flow through council accounts in ways that inflate apparent spending in particular service areas without representing any choice the council made. When you see adult social care spending rise sharply in an authority between 2020-21 and 2022-23, a portion of that rise is the council responding to grant conditions, not expressing a view about the relative importance of adult social care.

Our longitudinal dataset let us strip out identified ring-fenced grant income from service-level spending figures and recalculate spending shares on a net-of-grants basis. For fourteen of the forty-three authorities, this adjustment moved the headline spending share in at least one major service area by more than six percentage points. That is a large enough difference to reverse a comparative ranking between two authorities — which means rankings built on gross figures, as most published benchmarking tools use, can put councils in the wrong order for reasons entirely disconnected from their actual resource allocation choices.

What the accounts can legitimately tell you — and how to use that ¶

None of this means published accounts are useless as evidence. They are exceptionally reliable on certain narrow questions: whether an authority is drawing down reserves, how its debt-servicing costs have moved relative to its revenue budget, whether its collection rates for council tax and non-domestic rates are tracking the national distribution. These are questions about financial position and fiscal behaviour, and the audited accounts answer them well. The mistake is using the same data to answer questions about political priority, which is a different category of question requiring different evidence.

The most productive analytical approach we found in the study combined accounts data with at least two additional source types: scrutiny committee minutes (which record the arguments councillors actually made when budget decisions were contested) and Medium Term Financial Strategy documents (which show what officers projected before the final budget was set, making the gap between projection and outturn visible). Neither source is perfect. Scrutiny minutes are inconsistently detailed across authorities, and MTFS documents vary enormously in technical quality. But the combination gives you something the accounts alone cannot: a record of what was argued about, and what changed between the plan and the result.

In the seven-year dataset, the authorities whose scrutiny minutes showed sustained, contested debate about a specific service area were significantly more likely to show durable spending-share changes in that area two to three years later, compared with authorities where the same service area appeared in accounts but not in committee debates. That correlation does not prove causation, but it is a useful diagnostic: if the debate is not in the minutes, treat the accounts figure with proportionally more scepticism.

The forty-three-authority dataset is available to verified researchers through Aurora Pioneer's data-sharing agreement; contact the team directly with a brief methodology note. The 2022-23 accounts for the full sample are now published, and the 2023-24 cycle will begin closing in autumn, which will extend the longitudinal run to eight years.