The Fiscal Credibility Test
Evaluate government announcements by long-run commitment, not short-run numbers
Miller's final synthesising framework asks not whether a policy is right in isolation but whether the government making the announcement has the credibility to sustain it. Credibility has two components: the policy is coherent (it solves the stated problem in a sustainable way) and the government has demonstrated it will not change course under political pressure.
The pensions system is her primary case study. People are asked to commit to 30-40 year savings plans based on today's tax treatment of pensions. If the rules change every few years — and they have — rational people respond by deferring or restructuring their savings, which is economically harmful and individually rational. The triple lock on the state pension is similarly a credibility problem: it is a ratchet mechanism that politicians keep committing to because they fear the political cost of removing it, but which is mathematically unsustainable over long time horizons.
Fiscal forecasting is the third axis. Miller is explicit that she finds it alarming that major tax policy decisions are made in a three-week window before a budget, based on a single OBR forecast, when those decisions will affect millions of people for decades. The temporal mismatch between the speed of political decision-making and the long-run effects of tax design is itself a credibility problem — good policy requires thinking 20-30 years out, not optimising for the next OBR assessment.
- Credible policy must be sustainable over the long run — not just arithmetically sound in the short run.
- Repeated tinkering with long-run savings structures (pensions, ISAs) destroys the planning environment that gives those structures value.
- Fiscal rules are only credible if the government has a track record of sustaining them — otherwise they are aspiration statements.
- The timing mismatch between political cycles (2-5 years) and good policy design horizons (20-30 years) is a structural problem requiring structural solutions.
- People need certainty about the rules of long-run savings to make rational decisions — uncertainty is itself a form of policy failure.
- Check whether the policy addresses a structural problem or a short-run fiscal gapPolicies designed to close a budget gap in the next OBR window are fundamentally different from policies designed to fix a structural flaw. The former are likely to be reversed or modified when the gap changes; the latter have independent justification that makes them durable.Pro tipAsk: if the OBR's growth forecast came in 1pp higher, would this policy still be announced? If not, it is fiscal gap-filling, not reform.
- Assess the track record on fiscal commitmentsLook at whether the government has followed through on previous fiscal commitments — spending envelopes, tax rule promises, borrowing targets. A government with a poor track record should have its current announcements discounted accordingly.Pro tipMiller's OBR 5-year chart is a useful historical reference. The pattern of governments always promising debt will fall in five years and never delivering suggests structural over-optimism in fiscal forecasts.WarningTrack record assessment should be government-specific and institution-specific (OBR has a better track record than Treasury forecasts).
- Evaluate the long-run sustainability of any ratchet mechanismsIdentify any features of the announced policy that compound over time — like the triple lock, which locks in the highest of three inflation measures each year. Evaluate whether these mechanisms are arithmetically sustainable over 20-30 year horizons. If not, they represent a credibility problem regardless of current political commitment.Pro tipMiller's approach to the triple lock: link the state pension to a fixed percentage of earnings and add inflation protection for temporary spikes — this is sustainable without the ratchet.WarningRatchet mechanisms that benefit vocal constituencies (like pensioners, who vote at higher rates) are politically durable even when economically unsustainable — plan accordingly.
- Demand long-run plans, not just the next budgetFor any area where you are making a long-run commitment (pension savings, business investment, property purchase), ask what the government's 20-year plan for that policy area looks like. If there is no long-run plan — only a budget-by-budget approach — that uncertainty is a real cost to your decision.Pro tipThe state pension age review is an example of a policy where the long-run plan matters more than the current announcement. Miller argues decisions about moving the retirement age should be made now for the 2040s — precisely to give people certainty to plan.
- Apply appropriate skepticism to forecasts that are central to the planWhen a fiscal plan or policy announcement depends critically on a growth or revenue forecast, check whether that forecast is more optimistic than consensus. A policy that only works if the OBR's growth forecast is correct — and the OBR is typically more optimistic than the IMF or other forecasters — has embedded a credibility risk.Pro tipMiller's formulation: it gives her 'heart palpitations' that major tax policy is designed in a three-week window around a forecast. Treat forecast-dependent policy announcements with the same skepticism.
The state pension triple lock guarantees the pension rises by whichever is highest of inflation, earnings growth or 2.5%. In volatile economic periods, this always locks in a higher base. The result is a ratchet: the state pension has grown relative to working-age incomes over two decades, and pensioners have moved from being the poorest demographic to doing better than the working-age population on average.
Every pre-budget period generates speculation that the Chancellor will change pension tax relief — either the 25% tax-free lump sum or the rate of up-front relief. Miller notes this causes people to move money out of pensions pre-budget and restructure their savings in response to speculation, not policy.
Miller argues that decisions about moving the state pension age to 69 or 70 should be announced now for implementation in the 2040s. The further the change is in the future, the easier it is politically and the more planning time individuals have. Waiting until the change is imminent produces the worst outcome: maximum political difficulty with minimum planning time.
Miller's concern about policy credibility grew from watching the pension tax system change repeatedly and observing how each change, even when individually sensible, added to aggregate uncertainty. Her alarm about fiscal forecasting crystallised around the pre-budget period: she made explicit that it 'gives her heart palpitations' to see tax policy fine-tuned in a three-week window based on a growth forecast. This is not an IFS institutional position — it is her personal reaction to watching how tax policy is actually made versus how it should be made.