Topic 6 of 64· Public value foundations

Additionality and Deadweight

Additionality asks whether an intervention caused an outcome that would not otherwise have happened. Deadweight is its mirror: the share of an outcome that would have occurred anyway, even without the programme, grant, or subsidy. Almost every impact claim from a government programme or charity overstates its effect until deadweight is subtracted, which is why UK evaluation guidance treats it as the first and most important adjustment to any headline number.

Why it matters

"We supported 500 businesses to grow" sounds like an achievement, but if 300 of those businesses would have grown anyway — because the local economy was recovering, because they had other funding routes, because they were already on a growth trajectory before the programme started — the programme's true additional contribution is 200, not 500. HM Treasury's Magenta Book and the long-standing HM Treasury/BIS "Additionality Guide" (developed originally for regional development and regeneration programmes, and widely used across UK government evaluation since) formalize deadweight as the starting adjustment in the standard net-impact sequence: gross effect minus deadweight, minus displacement, minus leakage, adjusted for multiplier effects, equals net additional impact. Skipping this step is the single most common way public and social-sector impact claims are inflated, deliberately or not — a grant programme that only measures gross participant outcomes, with no comparison group, cannot distinguish its own effect from what would have happened regardless.

Deadweight is not a fixed percentage; it depends entirely on the counterfactual for the specific population and intervention (see counterfactual analysis). English regional-development evaluations under the previous Regional Development Agencies commonly found deadweight rates in the range of 20–60% depending on the type of business support, which is why credible programme evaluations report a deadweight-adjusted range rather than a single assumed figure, and why funders such as the National Lottery Community Fund and Big Society Capital require grantees to address deadweight explicitly in outcomes reporting rather than reporting gross participant counts.

The maths

The standard net-impact adjustment sequence, as set out in UK evaluation guidance (Magenta Book; HM Treasury/BIS Additionality Guide; ESIF and structural-funds evaluation guidance):

Gross outcome
  − Deadweight       (what would have happened anyway)
  − Displacement     (activity/benefit shifted from elsewhere, not created — see
                       displacement-and-attribution)
  − Leakage          (benefit accruing outside the target group/area)
  × Multiplier        (additional indirect/induced economic activity, where positive)
  = Net additional impact

Deadweight rate as a proportion:

Deadweight rate = outcomes that would have occurred without the intervention
                   / total gross outcomes observed

Net additional outcomes = Gross outcomes × (1 − Deadweight rate)

Worked example

Business-support grant programme: a regional grant scheme reports 500 supported businesses increased employment in the following year, an average of 3 jobs each — a gross claim of 1,500 jobs.

A matched comparison group of similar non-supported businesses (see counterfactual analysis) shows that 40% of the supported businesses' employment growth would have happened anyway, based on how the matched group performed over the same period.

Deadweight rate = 40%
Net additional jobs = 1,500 × (1 − 0.40) = 900 jobs

The programme's honestly reportable achievement is 900 jobs, not 1,500 — a 40% reduction purely from the deadweight adjustment, before displacement or leakage are even considered.

Charity employment programme: a charity places 200 long-term unemployed people into jobs at a cost of £600,000 (£3,000 per placement, gross). National labour-market data shows that, absent any intervention, roughly 15% of a comparable long-term unemployed cohort finds work within the same period through natural job-market churn.

Deadweight rate = 15%
Net additional placements = 200 × (1 − 0.15) = 170
True cost per additional placement = £600,000 / 170 ≈ £3,529

The gross cost-per-placement figure (£3,000) understates the real cost of the charity's additional contribution by roughly 15%.

Software engineering connection

Additionality and deadweight matter directly for anyone building impact-measurement or grants- management software for the public or social sector:

  • Outcomes-reporting systems should capture a comparison or baseline group by design, not just participant outcomes — retrofitting a counterfactual after a system launches without one is far harder than building the capture in from the start (see counterfactual analysis).
  • Dashboards that report only gross participant counts will systematically overstate impact to funders and oversight bodies; where deadweight estimates exist (from evaluation literature or a comparison group), the software should surface the net-of-deadweight figure alongside the gross one, not instead of it.
  • This connects directly to social return on investment, whose SROI ratio is only credible once deadweight (and displacement) have been subtracted from the gross claimed outcomes — an SROI calculator that omits this step will produce inflated ratios that do not survive scrutiny.

Pitfalls

  • Reporting gross outcomes as if they were all additional. This is the single most common impact- measurement error in grant and programme reporting; always ask "would this have happened anyway?" before publishing a headline number.
  • Assuming a single deadweight percentage applies everywhere. Deadweight varies hugely by sector, population, and local economic conditions; use a comparison group or sector-specific evidence rather than reusing a figure from an unrelated evaluation.
  • Confusing deadweight with displacement. Deadweight is about counterfactual outcomes for the same participants; displacement is about effects on other people or places — see displacement and attribution. Conflating the two leads to double-counting or under-counting the adjustment.
  • Self-reported deadweight from participants. Asking beneficiaries "would this have happened without our help?" produces systematically low deadweight estimates (participants tend to credit the programme); an independent comparison group is far more reliable.

Sources

  • HM Treasury, "The Magenta Book: Central Government Guidance on Evaluation" (2020). https://www.gov.uk/government/publications/the-magenta-book
  • HM Treasury / Department for Business, Innovation and Skills, "Additionality Guide: A Standard Approach to Assessing the Additional Impact of Interventions" (3rd edition), originally developed with English Partnerships and the Housing Corporation.
  • European Commission, "Evalsed: The Resource for the Evaluation of Socio-Economic Development" — guidance on deadweight, displacement, and leakage in structural-funds evaluation.
  • National Lottery Community Fund, "Guidance on Outcomes and Impact Reporting." https://www.tnlcommunityfund.org.uk/