A lottery would have picked the same projects

The chosen and the rejected projects score the same

You compare what was chosen with what was not. Trafikverket produces a cost-benefit calculation for every named candidate for the national transport plan, both the projects that make it in and the ones that are turned away. If those calculations carried weight in the selection, the chosen projects would score clearly higher than the rejected ones. That is the entire point of calculating: rank the options, fund the better ones. And the plan directs roughly 1,200 billion kronor over twelve years, so the ranking is not an academic exercise.

Stockholms Handelskammare ran the comparison. The projects prioritised into the plan have an average net present value ratio, NNK, of −0.3. The candidates that were deselected: also −0.3. Drawing project names out of a hat would have produced the same portfolio. The calculations are produced, filed as appendices, and then do not decide which projects get built.

The −0.3 itself deserves precision. It is Innovationsföretagen's weighted average for the named investment objects in the plan proposal for 2026–2037, computed from Trafikverket's own impact assessments. It means a krona invested returns monetised societal benefit — travel-time savings, traffic safety, emissions — worth about 70 öre. It is not cash flow, and it excludes maintenance and minor upgrades, roughly half the plan, which are profitable. We walk through the number in what NNK −0.3 actually means.

Someone wrote 450 assessments, then filed them

Someone did the work. Ahead of the 2022 plan revision, about 450 samlade effektbedömningar, consolidated impact assessments, were produced, roughly half of them by consultants. Each one follows Trafikverket's published ASEK guidelines: benefits and costs converted into kronor and discounted over a horizon of 40 to 60 years, object by object. As analysis it is careful, transparent work; comparative studies rate Swedish appraisal methodology and documentation above Norway's. Then the assessments were filed.

Riksrevisionen, Sweden's national audit office, examined the plan in 2023 and concluded that it promises more than it can keep:

Investeringarna bidrar till målen men inte på ett effektivt sätt.

The audit also documented how little room the calculations were given. Before Trafikverket even began prioritising for the 2022 plan, 107 billion kronor had been politically earmarked for new main lines, a third of the entire investment envelope, locked before the first assessment could say anything.

None of this is the analysts' failure. Trafikverket's economists are the one part of the system doing exactly what a rational planning process needs: measuring every candidate with the same yardstick and publishing the result. The failure begins the moment the result goes into an appendix.

Britain runs the same numbers and uses them as a gate

Britain runs roughly the same arithmetic and arrives somewhere else with it. The Department for Transport sorts every scheme into value-for-money categories based on its benefit-cost ratio: a BCR of 2 to 4 counts as high value for money, and schemes below a BCR of about 1.5 struggle to get funded. The calculation is a gate, not an appendix.

The gate holds even for politically loaded projects. HS2, Britain's most criticised railway, carries a BCR of 1.2, roughly NNK +0.2 on the Swedish scale: half a step above the entire Swedish portfolio of named objects. And Sweden's own bar is lower still: Trafikverket's guidance counts a project as profitable from an NNK of 0.1, four tenths above where the portfolio sits. The methods are similar. What differs is what happens after the calculation.

The cheapest reform available

A lottery would have picked the same projects. That is not an argument against calculating; it is an argument against filing. The state already pays for the analysis, about half of it produced by consultants, all of it to a published standard. The marginal cost of letting it influence which projects get built is zero. Where the numbers are allowed to speak, they already work: within the same plan, Trafikverket scales up maintenance and trimming measures by nearly 50 percent, citing precisely their high returns and quick effect. Riksrevisionen's recommendation in 2023 said as much: prioritise the projects with high societal returns.

There is a second, quieter cost to filing. An assessment made once and shelved goes stale: the price side drifts through the years of planning while the benefit side sits untouched in an appendix. By the time a project breaks ground, the calculation meant to justify it can describe a project that no longer exists. But the first step needs no new tools at all. Read the numbers before deciding. A lottery shouldn't be able to keep up.