Infrastructure appraisal: keep economic value separate from the funding plan

The most consequential error in infrastructure appraisal is treating a transfer as though it were a resource cost or a benefit.

A grant, a subsidy, a tax or a levy is a transfer. It moves money from one party to another. It does not consume resources and it does not create them. In a financial appraisal the transfer matters enormously, because it changes who pays and how much. In an economic appraisal it does not appear as a benefit at all, because the economy as a whole is neither richer nor poorer for the money having moved.

Counting a grant as a benefit in an economic appraisal therefore double counts. The value the project creates is already counted through the benefits to users; adding the grant counts a transfer on top of it, and inflates the case. Conversely, counting a grant as a cost in an economic appraisal is wrong in the other direction.

The distinction also clarifies a related question that causes confusion: whether a project 'needs' support. Economically, if the benefits exceed the resource costs, the project is worth doing and the question is how to fund it. If they do not, no amount of funding changes that — a transfer can make the finances work but cannot make the project worth doing.

Double counting in appraisal usually takes one of two forms, and both are common enough to be worth naming.

The first is counting the same benefit twice. Where a project generates a benefit that appears in more than one line — improved accessibility counted both as a user benefit and as a land value uplift, for instance — the value is counted once too many. The discipline is to identify each distinct benefit and count it once, in the line where it most directly arises.

The second is counting a transfer as a benefit, as discussed above. The two errors often appear together: a project counts its grant as a benefit and also counts the activity the grant funds as a benefit, describing the same money as creating value twice.

The remedy in both cases is the same: be explicit about the boundary of the analysis, identify what each line represents, and check that no value appears in two places. An appraisal that states its boundary and its treatment of transfers is straightforward to review; one that does not is difficult to trust even when the arithmetic is correct.

Risk appears in both appraisals, and how it is treated matters as much as the base figures.

In economic appraisal, risk attaches to the benefits and costs themselves — whether the demand materialises, whether construction costs more than expected, whether the benefits are realised as predicted. These are usually handled through sensitivity analysis and scenario testing, showing how the conclusion changes as the assumptions move.

In financial appraisal, risk attaches additionally to the funding — whether committed support is received, on what terms, and on time. And here the distinction from earlier applies directly. Support that is legally secured and contractually committed, with defined terms, can properly be included in the funding case; its counterparty risk should be assessed, which means considering the payer's ability and willingness to perform and what happens if a condition is not met. Support that is speculative — untested eligibility, unpublished terms, a benefit dependent on a future administrative decision — belongs in a sensitivity, because including it in the base case would make the case contingent on an outcome the project does not control.

This is a distinction about evidence, not about whether support is legitimate. A project with secured support and sound economics is in a strong position. A project whose case collapses without support that has not yet been granted has an untested assumption at its centre — and the earlier it is identified as such, the more options remain open.

The practical value of keeping the two appraisals separate is that each answers a question the decision actually needs.

If the question is whether a public body should support a project, an economic appraisal answers it, because the concern is whether the project creates value for the economy. If the question is whether a promoter can deliver and sustain the project, a financial appraisal answers it, because the concern is cash. If the question is whether a project should proceed at all, both matter, and a project that fails either requires attention that is specific to which one failed.

Presenting them separately, rather than as a single combined case, also makes each reviewable on its own terms. A reviewer can check whether the economic case has correctly excluded transfers and counted benefits once, and separately whether the funding case rests on committed support or on hope. Combined into one figure, both checks become impossible.