The idea behind the Office for Budget Responsibility is sound enough. Governments are often overoptimistic about economic growth forecasts. This leads to excessive borrowing, especially before elections. An independent watchdog like the OBR can offset this deficit bias by providing impartial economic forecasts and assessing if policies align with the government’s debt targets.

In theory, this encourages long-term thinking and fiscal discipline. In practice, events featuring OBR forecasts, such as the recent autumn statement, have turned farcical. While the statement delivered some sensible tax cuts, the overall spectacle showed that volatile forecasts and gameable fiscal rules, even under the OBR’s watch, can produce erratic policy change and pre-election smoke and mirrors.

First, unstable OBR forecasts before the statement implied wildly different policy outlooks. On October 4, the OBR projected a £27 billion annual budget shortfall, indicating major spending cuts or tax hikes were needed to ensure a falling debt-to-GDP ratio by 2028–29. By November 17, with gilt yields having fallen, the OBR then predicted £31 billion of “headroom” against the same government target, allowing the tax cuts. “Had the statement been six weeks earlier, we’d have raised taxes again,” one Conservative told me. Such fickle forecasts hardly provide a solid basis for major tax decisions.

The OBR’s current remit also falls short against government chicanery in “meeting” the debt target. The OBR must take government plans at face value, for example, even when they are implausible. As Tim Leunig, a former government adviser, observed, this flatters the public finances by pencilling in additional revenues and proposed spending cuts that nobody thinks will materialise.

The OBR assumes, for example, that the government will raise fuel duty next year and every year thereafter, even though it hasn’t done so since 2011. If the government stuck to form and froze fuel duty through 2028–29, revenue would be £6.2 billion lower than the official forecast implies. In essence, 43 per cent of the chancellor’s remaining “wiggle room” against his debt target depends on a “planned” tax rise he doesn’t want to deliver.

Assuming stated spending ambitions on health, defence, foreign aid, school pupils and childcare are honoured, the government’s “plans” also imply cuts to other “unprotected” government departments of more than 4 per cent a year from 2025–26 onwards. Again, this won’t happen. Yet it’s these very assumptions that enabled the government to cut certain taxes while still getting the OBR’s sign-off that the debt target would be hit.

To mitigate such gaming, Leunig proposes forcing the chancellor to read an OBR statement that lays bare these issues at the despatch box. However, the real issue surely lies in the fiscal rules the government sets the OBR to police. A “rolling” target to have debt-to-GDP falling five years hence inevitably makes forecasts crucial and allows governments to push off budget tightening until tomorrow, in the hope they will never have to do it.

A more effective rule would have the OBR set an annual government spending cap based on the average tax revenue of the past three years, adjusted for inflation and population, plus the current year’s projections. Any additional borrowing within the year would tighten the belt on spending over the next three, with emergency borrowing for recessions paid back with adjustments over a longer six years.

This simpler approach would guarantee balanced budgets over the business cycle. It would also force governments to raise taxes before they could increase spending. And, crucially, the OBR would then have teeth. The body would exist to assess compliance with real-time rules and expose tangible risks to our debt outlook, rather than obsessing over speculative five-year forecasts and government plans nobody thinks are realistic.