Benefit-Cost Ratio

Benefit-Cost Ratio and Its Limits as a Decision Rule

The benefit-cost ratio (BCR) is easy to quote and easily misread as a decision rule. A health programme said to return £3 for every £1 spent sounds worthwhile, and such figures appear in public health investment cases and government appraisal, but a cost-benefit analysis needs more than the ratio to reach a decision. Two programmes can have the same net value and very different ratios, and the option with the highest ratio is not always the one that creates the most value. The ratio is tied to net present value by a simple identity, yet net present value is the rule for choosing between mutually exclusive options, and the treatment of cost savings and transfers moves the ratio without changing net value. Public health returns on investment and the HM Treasury Green Book both use the ratio, under different conventions. The formulas for present values and for both measures are on the cost-benefit analysis page.

What the ratio summarises in a health cost-benefit analysis

A cost-benefit analysis in health care values health gains in money, so that they can be added to other monetised consequences and set against costs. The health gain may be valued through willingness to pay, through a value of statistical life applied to changes in mortality risk, or through the human capital approach, which values changes in productive output. The benefit side of the ratio therefore depends heavily on which valuation method is used and whose preferences it reflects.

The HM Treasury Green Book (2026) describes the benefit-cost ratio as the social return of a proposal, meaning the monetisable social benefits generated for each pound sterling of monetisable social costs. Both quantities are present values in real terms. For health programmes the discounting step matters, because the Green Book discounts effects on health or life at a lower rate of 1.5% rather than its standard rate, on the grounds that the wealth-effect component of the discount rate does not apply to additional years of life. Monetised health gains therefore lose less value with time than monetised resource flows, and long-horizon prevention programmes have higher ratios than they would if every flow were discounted at the standard rate (see discounting health). Where health effects are expressed as quality-adjusted life years, the Green Book refers practitioners to its supplementary guidance on health for the monetary value of a QALY.

How the ratio and net present value are tied together

The ratio and net present value are built from the same two present values, so a simple identity links them. Subtracting one from the ratio gives net present value per pound of cost, as shown below.

$$BCR - 1 = \frac{PVB - PVC}{PVC} = \frac{NPV}{PVC}$$

where $BCR$ is the benefit-cost ratio, $PVB$ is the present value of monetised benefits, $PVC$ is the present value of costs and $NPV$ is the net present value, equal to $PVB - PVC$. When costs are positive, a ratio above one and a positive net present value are the same finding. For a single programme compared with doing nothing, the two measures therefore give the same accept or reject answer.

What the ratio loses is scale. For illustration, a ratio of 3 can describe a local pilot costing £10,000 with a net present value of £20,000 or a national programme with a net present value of £200 million.

Why net present value decides between mutually exclusive options

When only one of several options can be adopted, such as alternative designs of the same screening programme, the aim is to choose the option that creates the greatest net value. That is the option with the highest net present value, and it is not necessarily the option with the highest ratio, because a larger programme can add more net value at a lower average return. The US Office of Management and Budget's Circular A-4 (2003) states directly that the ratio of benefits to costs is not a meaningful indicator of net benefits and should not be used for that purpose. The cost-benefit analysis page includes a worked example in which two options rank differently on the two measures.

The ratio can still give the right answer if it is applied incrementally. Moving from option X to a larger option Y is worthwhile when the extra benefits exceed the extra costs, which the incremental ratio measures.

$$BCR_{Y,X} = \frac{PVB_Y - PVB_X}{PVC_Y - PVC_X}$$

where $BCR_{Y,X}$ is the incremental benefit-cost ratio of Y over X, and $PVB$ and $PVC$ are the present values of benefits and costs of each option. When Y costs more than X, an incremental ratio above one is the same condition as Y having the higher net present value. This mirrors the step from average to incremental ratios in a fully incremental analysis in cost-effectiveness work, and net present value plays the role that net monetary benefit plays there.

A different question arises when several independent programmes compete for a fixed budget. If programmes can be scaled up or down in proportion, funding those with ratios above one in descending order of ratio until the money runs out gives the largest total net present value, much as a league table of cost per QALY does under the same assumptions. When programmes are all or nothing, that ordering can leave part of the budget idle, and the correct choice is the combination of programmes that maximises total net present value within the budget constraint. Ranking by ratio also maximises total net present value only when the ratio's denominator is the constrained budget. When part of the cost falls elsewhere, for example on patients or another sector, programmes should be ranked by net value per pound of the constrained budget, which is the logic of the Green Book's return on public sector cost.

Where cost savings go and the netting problem

Many consequences of a health programme can be classified either as a benefit or as a reduction in cost. Averted hospital treatment, productivity gains and reduced informal care can be added to the numerator, or they can be subtracted from programme costs in the denominator. The choice leaves net present value unchanged, because a pound added to benefits and a pound removed from costs have the same effect on the difference, but it changes the ratio, sometimes by a large amount.

With programme cost $C$, averted treatment costs $S$ and monetised health gain $H$, all in present values, the two classifications give the ratios below.

$$BCR_{gross} = \frac{H + S}{C}, \qquad BCR_{net} = \frac{H}{C - S}$$

where $BCR_{gross}$ treats averted costs as a benefit, $BCR_{net}$ nets them off programme costs, $C$ is the present value of programme costs, $S$ is the present value of averted treatment costs and $H$ is the present value of the monetised health gain. Net present value is $H + S - C$ in both cases. When averted costs equal or exceed programme costs, the denominator of the netted ratio becomes zero or negative and the ratio is undefined or negative, even though the programme saves money and improves health.

The Green Book (2026) recognises this ambiguity. Its example concerns emissions, which can be counted as a benefit when reduced or as a cost when increased; it notes that the choice affects the ratio but not net present social value, and asks practitioners to explain it. It also notes that transfers counted as a cost to the payer and an equal benefit to the recipient cancel in net present social value but push the ratio artificially towards one. The Green Book's return on public sector cost adds a further variant, dividing benefits net of costs outside the public sector by public sector costs, so costs borne by patients and families move from the denominator to the numerator. A ratio is only interpretable alongside a statement of how each consequence was classified, and comparisons across studies that classify differently are not valid.

Worked example of one vaccination programme counted two ways

The figures below are illustrative. A health authority is choosing between two mutually exclusive designs of a vaccination programme, a routine cohort programme (X) and the same programme with a catch-up campaign for older age groups (Y). All figures are present values in £ million, with health gains valued in money.

1. Net present value. For X, 5.0 + 3.0 − 4.0 = 4.0. For Y, 9.0 + 5.0 − 9.0 = 5.0. Y creates £1.0 million more net value.

2. Ratio with savings counted as benefits. For X, (5.0 + 3.0)/4.0 = 2.0. For Y, (9.0 + 5.0)/9.0 = 1.56.

3. Ratio with savings netted off costs. For X, 5.0/(4.0 − 3.0) = 5.0. For Y, 9.0/(9.0 − 5.0) = 2.25. The same programme X now shows £5 of benefit per pound of net cost instead of £2 per pound of cost, with no change in its net value.

4. The incremental comparison. Counting savings as benefits, the move from X to Y gives (14.0 − 8.0)/(9.0 − 4.0) = 1.2. Netting savings off costs, the extra health gain is 9.0 − 5.0 = 4.0 and the extra net cost is (9.0 − 5.0) − (4.0 − 3.0) = 3.0, giving 4.0/3.0 = 1.33. Both incremental ratios exceed one, which agrees with the £1.0 million gain in net present value. X has the higher ratio under either classification, but Y is the better choice.

5. A transfer added to X. Suppose X also pays £2.0 million in incentive payments to families, counted as a cost to the authority and an equal benefit to recipients. The gross ratio becomes (8.0 + 2.0)/(4.0 + 2.0) = 1.67, down from 2.0, while net present value stays at 4.0.

6. A cost-saving variant. If X averted £4.5 million of treatment costs instead of £3.0 million, its net present value would be 5.0 + 4.5 − 4.0 = 5.5 and its gross ratio 9.5/4.0 = 2.38. The netted denominator would be 4.0 − 4.5 = −0.5, giving a ratio of −10 for a programme that both saves money and improves health.

In the base case, each option's ratio depends on accounting choices, and ranking the two designs by ratio selects the wrong one; net present value, or an incremental ratio, identifies Y under every classification. In the cost-saving variant, X becomes the better option, with a net present value of £5.5 million against £5.0 million for Y.

Returns on investment in public health

The standard formula for return on investment divides net benefit by cost, which equals the benefit-cost ratio minus one. A systematic review of public health interventions in high-income countries by Masters and colleagues used this convention, calling the ratio a cost-benefit ratio, and reported a median return on investment of 14.3 and a median benefit-cost ratio of 8.3 across 52 studies. The two medians cannot be reconciled through the identity, because they were taken over different groups of studies, since 34 reported a return on investment and 23 reported a ratio.

A later scoping review by Turner and colleagues examined 118 studies published between 2018 and 2021 that reported a return on investment for a health intervention. In 59% of them the analysis appeared to count only fiscal or tangible cost savings, and some figures labelled as returns on investment appeared to be benefit-cost ratios, dividing benefits rather than net benefits by cost. A reported return of "4 to 1" can therefore mean a ratio of 4 or a ratio of 5, counting only fiscal savings or wider monetised benefits. The review recommends reporting absolute costs and benefits, not only the summary ratio, and stating whom the savings relate to. Social return on investment studies, which monetise wider social outcomes, raise the same classification issues.

How appraisal guidance uses the ratio

The Green Book (2026) asks for net present social value, the benefit-cost ratio and the return on public sector cost for each shortlisted option in a social cost-benefit analysis, and uses them for an initial ranking only. It states that the UK government does not choose projects solely on the basis of which have the highest benefit-cost ratios. Value for money also depends on unmonetisable costs and benefits, risks, public sector cost and distributional effects.

Three further rules in the Green Book limit how the ratio is used. Practitioners should not judge value for money with ratio thresholds, for example by rejecting a proposal because its ratio is below two. A proposal with a ratio below one may still represent value for money if it has significant unmonetisable benefits, although low ratios call for rigorous scrutiny. Summary metrics should only compare options that meet the same objectives, not different proposals with different objectives.

Common misreadings of the benefit-cost ratio

Most errors with the ratio come from asking it to answer a question that only net present value, or an incremental comparison, can answer. The common ones are listed below.

• Ranking mutually exclusive options by ratio. The option with the highest ratio may add less net value than a larger option; the incremental ratio or net present value settles the choice.
• Comparing ratios across studies. Ratios built with different classifications of savings, transfers or patient costs, or from different perspectives, are not on the same scale.
• Reading the ratio as an annual return. The ratio compares totals over the whole horizon; a return per year is the internal rate of return, a different measure.
• Treating unmonetised effects as zero. Health or equity effects left out of the numerator do not disappear; a ratio below one can still describe a worthwhile programme.

The ratio remains a readable summary of social return when net present value and the classification of each consequence are reported beside it. Sensitivity analysis should then show how far the ratio and the ranking move when valuation, discounting and classification choices change.

Sources

• Boardman AE, Greenberg DH, Vining AR, Weimer DL. Cost-Benefit Analysis: Concepts and Practice. 5th ed. Cambridge: Cambridge University Press; 2018.
• Drummond MF, Sculpher MJ, Claxton K, Stoddart GL, Torrance GW. Methods for the Economic Evaluation of Health Care Programmes. 4th ed. Oxford: Oxford University Press; 2015.
• HM Treasury. The Green Book: UK government guidance on appraisal (2026). London: HM Treasury; 2026. Chapters 2, 6 and 8. https://www.gov.uk/government/publications/the-green-book-appraisal-and-evaluation-in-central-government/the-green-book-2026
• Masters R, Anwar E, Collins B, Cookson R, Capewell S. Return on investment of public health interventions: a systematic review. Journal of Epidemiology and Community Health. 2017;71(8):827-834. https://doi.org/10.1136/jech-2016-208141
• Office of Management and Budget. Circular A-4: Regulatory Analysis. Washington, DC: Executive Office of the President; 2003 (reinstated 2025). https://www.whitehouse.gov/wp-content/uploads/2025/08/CircularA-4.pdf
• Turner HC, Hori Y, Revill P, Rattanavipapong W, Arai K, Nonvignon J, Jit M, Teerawattananon Y. Analyses of the return on investment of public health interventions: a scoping review and recommendations for future studies. BMJ Global Health. 2023;8(8):e012798. https://doi.org/10.1136/bmjgh-2023-012798