The Green Book is HM Treasury's guidance on how to appraise and evaluate public spending, and it governs any proposal that asks for government money — a hospital, a road, a grant scheme, a regeneration project. Its central discipline is uncomfortable and useful: you must set out what problem you are solving, consider a genuine range of options, and value the costs and benefits to society rather than to the organisation making the case.
The five case model
Business cases follow a standard structure, and each case answers a different question:
- Strategic — what is the problem, why must it be addressed now, and how does the intervention fit wider objectives? A weak strategic case usually means the answer was chosen before the question;
- Economic — the analytical core. A long list of options reduced to a shortlist, appraised with costs and benefits valued and discounted to present value. This is where the Green Book's methods live;
- Commercial — can it be procured and contracted, and is the market able to deliver it?
- Financial — is it affordable within budgets, and who pays for what, when?
- Management — can it actually be delivered, with governance, risk management and a plan for evaluating whether it worked?
The cases are developed in stages — a strategic outline case, then an outline business case, then a full one — with the analysis deepening as the proposal firms up. The point of the staging is to stop money being spent developing proposals that fail at the first question.
What the economic case requires
Three principles distinguish Green Book appraisal from ordinary investment analysis.
Social value, not organisational value. Benefits are counted wherever they fall — to users, to the wider public, to the environment — and costs likewise, including those borne by people other than the sponsoring body. Non-market effects are valued where credible methods exist.
A genuine counterfactual. Everything is measured against what would have happened anyway. A scheme that claims credit for outcomes that would have occurred regardless is the most common analytical failure, and it is also the easiest for a reviewer to test.
Discounting and optimism bias. Future costs and benefits are discounted to present value using the Treasury's declining discount rate. And because appraisers systematically underestimate cost and time, an explicit optimism bias adjustment is applied, based on evidence from comparable projects, and reduced only where specific mitigations justify it.
The output depends on the method, and this is where appraisals are most often written to the wrong template. Where benefits can be valued in money, social cost-benefit analysis produces a net present social value and a benefit-cost ratio. Where the main benefits cannot sensibly be monetised, the right method is social cost-effectiveness analysis, which values only the costs and expresses the result against a unit of outcome — a net present unit cost — and produces no BCR at all. Forcing a ratio out of an appraisal that should have been cost-effectiveness analysis is a recognisable error.
Either way the number is supported by sensitivity analysis showing which assumptions the conclusion actually depends on, and either way it informs a judgement rather than making the decision: the Green Book is explicit that options are not chosen simply by ranking benefit-cost ratios, and that fixed BCR thresholds are not an appropriate decision rule. A ratio quoted without sensitivity work, or presented as the answer, tells a reviewer very little.
Where business cases fail
The recurring criticisms are consistent across departments and funders. A shortlist that contains one real option and two straw men. Benefits that are asserted rather than evidenced, or double-counted across categories. Costs excluding lifecycle, maintenance and decommissioning. Optimism bias adjusted downwards without justification. Deliverability treated as an afterthought. And no evaluation plan, which means nobody will ever establish whether the money worked — the failure the Treasury's evaluation guidance exists to address.
There is also a distributional question that has grown in importance: who gains and who loses, and how the effects fall across places and income groups. Appraisals that report only an aggregate net benefit increasingly get sent back.
Why it matters outside government
Green Book discipline reaches well beyond Whitehall. Any organisation bidding for public funding — a local authority, a university, a housing association, a charity, an infrastructure developer — will be assessed against it, and the quality of the economic case frequently determines the outcome between competing bids of similar merit.
It is also a genuinely useful framework for large private investment decisions. The options analysis, the counterfactual, the explicit optimism bias and the requirement to evaluate afterwards all address the failure modes that make corporate capital projects overrun — and most private business cases would be improved by borrowing the discipline, even where the social value dimension does not apply.
Preparing a case that survives review
Start with the problem and the evidence for it, not the solution. Build a long list that includes doing nothing and doing something smaller, and document why options were rejected. Evidence benefits from comparable schemes rather than from ambition. Cost the full lifecycle. Apply optimism bias from the published evidence base and justify any reduction with specific mitigations. Test which assumptions the conclusion turns on, and say so. And write the evaluation plan while the case is being made, because nobody writes one afterwards.
Acumon supports public sector bodies and funded organisations on business cases, cost analysis and the assurance that follows through public sector audit, financial modelling and grant audit work. If a funding bid is being prepared, the options appraisal is where reviewers look first.