Additionality Test for Public Intervention
Don't fund what the market would do anyway — government money must add net output, not subsidise profits.
The central question of any government intervention in a market is whether it adds something that would not otherwise exist — economists call this additionality. If the government subsidises an investment that the private sector would have made anyway, it has simply transferred money to shareholders or executives without increasing total output. The policy looks like support for growth; it is actually a windfall.
Vicky Pryce encountered this problem directly inside the Department of Trade and Industry, where sector-specific economists were physically located alongside the industries they were supposed to evaluate. The proximity created capture: the economists' assessments were routinely overridden by ministers and lobbyists, and public money flowed to activities the firms had already planned. She restructured the team so departmental economists reported centrally rather than to their sector — creating an independent check on additionality claims.
The test has four components: (1) Is there a genuine market failure — monopoly, externality, or missing price signal — that prevents the investment from happening privately? (2) Would the firm invest anyway without the subsidy? (3) Does the intervention genuinely change behaviour, or does it allow the firm to pocket the grant while doing what it planned? (4) Are the costs proportionate to the additional output created?
- Government intervention is only justified when the market has a structural failure that prevents an efficient outcome from occurring.
- Additionality means the intervention must change what would otherwise happen — if the firm acts the same with or without the subsidy, the money is wasted.
- Regulatory capture — putting industry representatives inside policymaking departments — systematically corrupts additionality assessments.
- Independence of evaluation is not a procedural nicety; without it, evidence is ignored and spending flows to those with lobbying power.
- The question is always: would this have happened anyway? If yes, do not intervene.
- Identify the market failureState precisely why the market is not producing the desired outcome on its own. Valid reasons: monopoly pricing, positive externalities (skills, R&D spillovers), missing information, or public goods. If no clear market failure exists, stop — there is no case for intervention.Pro tipLobbying by an industry for support is not evidence of market failure; it is evidence of private interest.WarningPoliticians often frame preference for a particular sector as a market failure when it is actually industrial favouritism.
- Apply the counterfactual testAsk: would this investment happen at the same scale and timeline without the subsidy? If the answer is yes, the intervention provides no additionality. Require applicants to demonstrate a genuine funding gap or risk barrier that government uniquely resolves.Pro tipVenture capital absence from a sector is a real funding gap signal — it implies risk that private markets cannot price, which is where public guarantees add genuine value.
- Separate the evaluators from the evaluatedEnsure the economists or analysts assessing the intervention do not report to the same ministry or interest group that benefits from the spending. Cross-departmental or central reporting structures are necessary to preserve independence.Pro tipPryce restructured DTI so sector economists reported centrally to her, explicitly to prevent capture by the industries they assessed.WarningSecondments from industry into government policy roles are a direct capture mechanism — flag these when assessing any evaluation.
- Size the genuine additional outputCalculate what additional GDP, employment, or innovation the intervention produces above the counterfactual. If this cannot be quantified even roughly, the programme cannot be evaluated and should not proceed at scale.Pro tipThe Green Book's welfare analysis methodology provides a standard approach — require applicants to complete it.
- Build in sunset and reviewIndustrial policy that cannot be removed is industrial policy that has been captured. Build explicit review gates that require renewed additionality evidence. If the private sector has not taken over by the review date, ask whether the market failure has been resolved or whether the subsidy is now a permanent transfer.Pro tipLong-term consistency matters for business confidence, but this is different from permanent subsidy — the support architecture can persist while the financial transfer is phased out.WarningPolitically popular programmes are the hardest to wind down regardless of additionality evidence.
When Pryce worked in the Department of Trade and Industry, economists were embedded within sector teams (automotive, aerospace, chemicals) and co-located with industry representatives on secondment. The result was that economic assessments recommending against subsidy were routinely overridden because the evaluators had been captured by proximity. She resolved this by requiring sector economists to report centrally, creating independence.
The UK government offered energy price guarantees and nuclear capacity commitments to attract AI data centres from major US tech firms. These firms were planning to expand AI infrastructure globally regardless of UK policy. The question Pryce raises is whether the concessions provided genuine additionality — attracting investment that would otherwise have gone elsewhere — or simply subsidised investment that would have come to the UK anyway.
The additionality test is a standard HM Treasury framework, codified in the Green Book which guides UK public spending appraisal. Pryce draws attention to how systematically it is ignored in practice — both in her own experience inside DTI and in her assessment of subsequent industrial strategies. The HS2 decision, data centre guarantees to tech giants, and austerity-era enterprise support programmes all fail this test in her analysis.