Topic 2.3 — Social Return on Investment and Impact Measurement

Why this matters in public value

Funders, commissioners and boards increasingly want a single, comparable number that tells them whether a programme is worth continuing. Social Return on Investment (SROI) promises exactly that: a ratio, expressed as so many currency units of social value created for every unit invested. The promise is seductive because it borrows the legibility of a financial return while claiming to capture what a financial return cannot — reduced loneliness, restored confidence, a young person who stays in education.

The tension a reader faces this year is that the number is usually more confident than the evidence behind it. Pressure to produce a large, defensible ratio can quietly reshape which outcomes get chosen, which financial proxies get used, and which caveats get dropped from the funding application before it goes out the door. Some of the outcomes a social-sector organisation creates — dignity, belonging, cultural continuity — are ones many practitioners believe should never be reduced to a money figure at all, on principle rather than merely on grounds of difficulty.

This is the territory that formal cost-benefit analysis sets aside for itself (see Topic 2.2 — Social Cost-Benefit and Cost-Effectiveness Analysis), which treats monetisation as a solved methodological problem: find the shadow price, apply the discount rate, weigh the distribution. This topic covers the harder case — where monetisation is either not yet possible with any rigour, or where forcing it is itself the contested move. Getting this right matters because the alternative to a disciplined SROI or theory of change exercise is not 'no measurement' — it is an undisciplined one, dressed up with a ratio that looks precise and is not.

Core concepts

A theory of change is the explicit, testable narrative of how and why an organisation's activities are expected to lead to the outcomes it claims. The theory of change is introduced as the discipline that keeps a logic model honest about what it actually claims to cause (see Topic 2.1 — The Public Value Scorecard and Outcomes Frameworks); here, it is the logical backbone the rest of this topic rests on. It states the assumptions a programme depends on — who changes, through what mechanism, over what time, under what conditions — so that they can be examined and, ideally, tested rather than simply asserted. A logic model is the diagrammatic cousin of a theory of change: a chain of inputs, activities, outputs, outcomes and impact, useful for planning and reporting but thinner on the causal why. Logic models and outcomes frameworks generally are covered in full there (see Topic 2.1 — The Public Value Scorecard and Outcomes Frameworks). A valuation exercise built on a weak theory of change inherits that weakness no matter how careful the arithmetic that follows.

Social return on investment is a principles-based method, developed from work originally commissioned by the UK Cabinet Office and now stewarded internationally, for expressing extra-financial value — social, environmental, economic — using financial proxies alongside conventional accounts. Social return on investment typically works through six stages: establishing scope and identifying stakeholders; mapping outcomes through a theory of change; finding evidence for those outcomes and giving them an indicative value; establishing impact by removing what would have happened anyway; calculating the ratio; and reporting, using and embedding the findings. Each stage carries judgement calls that a reader cannot see in the final ratio unless the organisation discloses its working.

Four adjustments sit at the centre of that 'establishing impact' stage, and each one exists to stop an organisation overclaiming credit. Deadweight is the proportion of the outcome that would have happened without the intervention at all. Attribution is the share of the outcome properly credited to other actors — other funders, other services, the person's own effort — rather than to this organisation alone. Displacement is the extent to which a positive outcome for one group came at the expense of a negative outcome for another (a job created that simply moved from a neighbouring town, for instance). Drop-off is the decay of an outcome's value over the years following an intervention, since a benefit rarely persists at full strength indefinitely. Skipping any of these four, or applying them loosely, is the single most common way an SROI ratio becomes larger than the evidence supports.

A financial proxy is a monetary stand-in for an outcome that has no market price of its own — for example, using the cost of a comparable leisure activity as a proxy for the value of increased social confidence. A proxy is not a market price and should never be presented as one; it is an argued, disclosable choice, and different reasonable people can choose different proxies for the same outcome. Social accounting is the broader discipline of communicating an organisation's social and environmental effects to its stakeholders — social accounting is the family of practice SROI belongs to, alongside older multidimensional frameworks such as the triple bottom line (reporting social, environmental and economic performance side by side rather than folding all three into one figure) and certification-based approaches such as B Corporation status, which assesses an organisation's social and environmental performance against a standard rather than producing a bespoke ratio at all.

Indicator-set approaches offer a further alternative to a bespoke ratio. The IRIS+ system maintained by the Global Impact Investing Network (GIIN) — a nonprofit network supporting the practice of impact investing, investment intended to generate measurable social or environmental benefit alongside financial return — lets an investor or investee choose from a shared catalogue of metrics instead of monetising every outcome. This matters most for a social enterprise or nonprofit organization reporting to several funders at once, each of whom may otherwise demand a different bespoke valuation of the same outcome. Finally, impact measurement and management is the umbrella term increasingly used across the social sector and impact investing for the ongoing discipline of setting outcomes, collecting evidence and using it to adjust practice — as distinct from a one-off evaluation exercise produced to satisfy a single funder.

Best practices

Build the theory of change before you build the spreadsheet. An SROI ratio calculated on top of a theory of change that nobody has stress-tested is arithmetic wearing the costume of evidence. Draft the causal narrative first, test it against what frontline staff and participants actually observe, and only then decide what is worth valuing.
Treat the ratio as a communication device, not a verdict. A single figure — 'every pound invested returns four pounds of social value' — is a summary, not the analysis. Present it alongside the assumptions and evidence quality that produced it, so a reader can judge how much weight it deserves.
Publish your financial proxies and their sources. If a proxy cannot survive being shown to the people whose outcomes it represents, it should not be in the model. Naming the source of each proxy — a market price, a government valuation, a comparable survey — lets others check your reasoning instead of trusting your conclusion.
Apply deadweight, attribution, displacement and drop-off even when it lowers the ratio. These deductions exist precisely to prevent an organisation from claiming more credit than it has earned. An SROI exercise that produces a smaller, better-evidenced ratio is more useful than one that produces a larger, unexamined one.
Involve the stakeholders whose outcomes you are valuing, not only the funder commissioning the report. SROI's own principles call for stakeholder involvement in deciding what to measure and how to value it; skipping this step and letting the finance or fundraising team choose proxies alone is the most common shortcut that undermines legitimacy later.
Separate what you can count from what you must evidence. Outputs — sessions run, people attending, meals served — can be counted directly. Outcomes — reduced isolation, improved health, sustained employment — require evidence of change in the people affected, not merely evidence of activity delivered.
Match your standard of evidence to the maturity and stakes of the programme. A pilot with 30 participants does not need, and cannot honestly produce, the same evidential confidence as a national programme entering its fifth year. Frameworks that use graded evidence levels, from a plausible logical case through to independently replicated causal evidence, help set that expectation explicitly rather than by default.
Reserve a full SROI ratio for decisions big enough to warrant the effort. The method is resource-intensive to do well. For most day-to-day reporting, a well-evidenced outcomes narrative against a theory of change delivers more decision-useful information per hour of staff time than a forced ratio.
Seek independent assurance before publishing a ratio you intend to use comparatively. Self-assessed SROI figures are common and can still be useful internally, but a ratio put in front of a funder, investor or the public carries more weight, and more risk if wrong, once it is used to compare organisations or make an allocation decision.
Never compare SROI ratios across organisations that used different scopes, proxies or time horizons. A ratio of seven from one charity and a ratio of three from another says almost nothing about which does more good; it more often reflects which one chose more generous proxies.
Pair the quantitative ratio with qualitative case evidence. Numbers persuade a finance committee; a well-evidenced account of what changed for a real (anonymised) participant persuades everyone else, and it is often the qualitative account that catches an outcome the model missed.
Revisit the theory of change on a fixed cycle, not only when a funder asks. Programmes evolve, and a theory of change written three reorganisations ago will quietly misdirect measurement effort towards outcomes the programme no longer actually pursues.

Questions to discuss with your team

What decision will this exercise actually inform? Before committing staff time to a theory of change refresh or an SROI calculation, be explicit about which decision the output will feed — a funding renewal, an internal resource shift, a public report, or a genuine judgement about whether to keep running the programme. An exercise built to serve one purpose (marketing a strong headline ratio) will make different, and sometimes conflicting, methodological choices from one built to serve another (learning where a programme is underperforming). If nobody can name the decision, the exercise risks becoming a compliance artefact that consumes the very staff time that could otherwise go into service delivery. Naming the decision up front also tells you how much rigour is proportionate to invest.

Whose theory of change is this — the funder's or the people who deliver the work? A theory of change written by a commissioning team in a single afternoon, without frontline staff or participants in the room, tends to describe the programme the funder imagined rather than the one that actually runs. Ask who was in the room when the current theory of change was drafted, and whether frontline staff would recognise it as an honest description of how they believe change happens. If the answer is uncomfortable, that discomfort is itself useful information about where measurement effort is currently misdirected.

Which of our financial proxies would not survive being made public? Walk through every proxy currently used in a live SROI model and ask, one at a time, whether it could be shown to a sceptical journalist, an unimpressed board member or the participant it claims to represent without embarrassment. A proxy chosen because it was the largest defensible figure available, rather than the most representative one, is a sign that the exercise has drifted from measurement towards advocacy. This is worth doing as a standing exercise, not only when a ratio is about to be published.

Are we deducting deadweight and attribution honestly, or only when it is convenient? Deadweight and attribution deductions are the parts of an SROI calculation most likely to be softened under pressure to produce an impressive number. Ask the team preparing the next SROI report to show their deadweight and attribution assumptions explicitly, with the evidence or reasoning behind each, rather than as a single unexplained discount percentage buried in a footnote. If the same team cannot explain why this year's deduction is smaller than last year's, treat that as a flag, not a footnote.

What would we do differently if we could not express this outcome as a ratio at all? Some outcomes your organisation creates may be ones your stakeholders would object to seeing monetised — dignity, cultural continuity, a sense of safety. Discuss which outcomes on your current theory of change fall into that category, and whether your reporting currently pressures those outcomes into a proxy value anyway because the reporting template demands one. Where the answer is yes, discuss what a legitimate non-monetary alternative would look like, and whether your funders would accept it.

How do we compare impact across programmes that measure entirely different things? A portfolio funder or a multi-programme nonprofit board is regularly asked to compare a literacy programme against a housing programme against a befriending service, each with its own theory of change and its own proxies. Discuss honestly whether your organisation currently does this by forcing everything into SROI ratios (which invites exactly the false comparability warned about elsewhere in this topic), or through some other means — a shared outcomes framework, a standardised indicator set, or a portfolio-level narrative that resists false precision. Agree, as a team, which approach you are actually willing to defend under challenge.

In practice: a public value example

The following scenario is a fictional illustration, not an account of any real organisation.

A community wellbeing charity, referred to here only as the Trust, ran a peer-support programme for people recently discharged from hospital after a mental health crisis. Its principal funder — a local health commissioning body — asked for an SROI ratio ahead of a funding renewal decision, having seen the technique used favourably by other applicants that year.

The Trust's evaluation lead began, deliberately, with the theory of change rather than the spreadsheet. Two workshops with peer supporters and eight programme participants produced a causal narrative that differed in one important way from the funding application that had originally created the programme: participants consistently described the most valuable outcome as feeling believed, not merely 'less isolated', the term the original logic model had used. The theory of change was revised to reflect this, which in turn changed which outcomes the valuation exercise needed to evidence.

Valuing 'feeling believed' against a financial proxy proved contested inside the Trust itself. One proxy candidate — the avoided cost of a crisis re-admission — was defensible but risked implying the programme's worth lay only in downstream savings to the health system, understating what participants said mattered to them directly. The team chose to use that proxy for one outcome only, applied a conservative attribution deduction of 40 per cent to reflect the other services participants were also using, and reported a second, non-monetised outcome — self-reported sense of being believed, tracked on a standard wellbeing scale — as a narrative finding alongside the ratio rather than folded into it.

The final report to the commissioner led with the theory of change and the evidence, not the number; the ratio appeared with its assumptions disclosed in an appendix, alongside a plain statement that a portion of the programme's value was reported qualitatively because the Trust judged monetising it further would misrepresent what participants had said. The commissioning body renewed funding, citing the quality of the evidence base as much as the ratio itself — illustrating that a defensible, honestly caveated account can do more persuasive work than a larger, less examined figure.

Four sector lenses

Local government

Local commissioners increasingly build social value into procurement, not only into grant-funded evaluation. In the United Kingdom, the Public Services (Social Value) Act 2012 requires public bodies to consider how the services they commission might improve the economic, social and environmental well-being of the area, which has pushed SROI-adjacent thinking into mainstream contract scoring rather than leaving it as an optional add-on for the social sector alone. The distinctive risk at this level is scale mismatch: a national or standardised proxy value applied uniformly across very different local contexts can misstate value in either direction, and local commissioners are well placed to know when a national proxy simply does not describe their area.

National government

National government sits closer to the formal cost-benefit apparatus (see Topic 2.2 — Social Cost-Benefit and Cost-Effectiveness Analysis), with shadow prices and discount rates set centrally and applied consistently across departments. Where SROI-style thinking earns its place at national level is in programme areas — early years support, homelessness prevention, community cohesion — where a national evaluation function still needs a disciplined way to describe outcomes that resist the kind of rigorous monetisation a full cost-benefit analysis would demand (see Topic 2.2 — Social Cost-Benefit and Cost-Effectiveness Analysis). National bodies are also often the source of the graded evidence standards and proxy libraries that local and social-sector organisations rely on, which makes the quality of that central work disproportionately important.

This is SROI's home territory, and also where the pressure to produce a flattering ratio is most acute, because the ratio can directly affect a grant renewal or a place on a preferred-supplier list (see Topic 3.8 — Social Sector and Nonprofit Management for the funding-model pressures that shape this behaviour). Smaller nonprofits often lack the specialist capacity to run a rigorous SROI exercise at all, which creates a real equity problem: well-resourced organisations produce more polished ratios, not necessarily more real impact, and funders who compare ratios naively end up rewarding measurement capacity rather than outcomes.

Multilateral and international

International development and impact investing bodies have converged on shared indicator sets rather than organisation-by-organisation SROI ratios, precisely to make comparison across very different contexts more honest. The IRIS+ system maintained by the Global Impact Investing Network is the clearest example: rather than asking every impact investment to generate its own bespoke ratio, it offers a shared catalogue of metrics that investors and investees can select from, improving comparability without pretending that a single monetised ratio can meaningfully summarise a portfolio spanning agriculture, health and financial inclusion. The lens-specific risk here is that globally standardised metrics can flatten locally meaningful outcomes that do not map neatly onto the standard catalogue.

Common failure modes

Proxy inflation. Under pressure to produce an impressive ratio, an organisation selects the most generous defensible financial proxy for each outcome rather than the most representative one. The cost is not only a misleading ratio; it is a slow erosion of trust once a funder or journalist compares proxies across a sector and notices the pattern, sometimes called 'SROI shopping'.

Soft deadweight and attribution. Deadweight and attribution deductions are the easiest place to understate what would have happened anyway, because they require an honest, sometimes unflattering, judgement about how much credit the organisation actually deserves. The cost is an inflated ratio that collapses under scrutiny the first time an external evaluator applies stricter assumptions.

A theory of change written after the fact. Some organisations produce a theory of change only once a funder demands one, reverse-engineered from the outcomes the organisation already wanted to claim. This produces a document that looks rigorous but was never used to test or challenge the programme's own assumptions, and it fails the moment frontline staff are asked whether it describes what they actually do.

One-off measurement, never repeated. An SROI exercise commissioned once, to win a specific bid, and never updated, functions as marketing rather than management. The cost is that the organisation learns nothing from it internally, and the figure calcifies into a claim repeated for years after the underlying programme has changed.

False precision. A ratio reported to two decimal places — '£4.32 of social value for every pound invested' — implies a level of arithmetic confidence that the underlying proxies, deductions and evidence rarely support. The cost is reputational: a sophisticated reader who probes one input and finds it soft will discount the whole exercise, not just the input.

Comparing incomparable ratios. Funders and boards sometimes rank programmes or grantees by SROI ratio alone, as though the figure were as standardised as a financial return. Because scope, proxies, time horizon and rigour vary enormously between exercises, this practice systematically rewards optimistic modelling over real impact, and it is one of the most-cited critiques of the method in the academic literature on SROI's use in practice.

Maturity model

Sub-dimension	Initiate	Develop	Standardise	Manage	Orchestrate
Theory of change	No explicit theory exists; outcomes are asserted, not modelled.	A theory of change exists for flagship programmes, drafted by management alone.	Every funded programme has a theory of change, co-produced with frontline staff and reviewed on a fixed cycle.	Theories of change are tested against real evidence and revised when the evidence contradicts them.	Theories of change are shared and jointly refined with peer organisations, funders and participants across a delivery network.
Outcome evidence	Only outputs (activity counts) are recorded.	Some outcome data collected, inconsistently, mostly for funder reporting.	A standard outcome evidence framework is applied across all programmes, matched to programme maturity.	Outcome evidence quality is itself measured and reported, with gaps triggering additional evaluation investment.	Outcome evidence is pooled and benchmarked across organisations using shared indicator sets.
Valuation and proxy governance	No financial proxies used; monetisation not attempted.	Proxies chosen ad hoc, undocumented, by whoever prepares the funding bid.	A documented proxy library exists with disclosed sources; deadweight, attribution, displacement and drop-off are applied consistently.	Proxy choices and deductions are periodically audited or independently assured.	Proxy and valuation practice is coordinated across a sector body, reducing duplicated and inconsistent modelling.
Ratio use and communication	No ratio produced; impact described only in narrative, if at all.	A ratio is produced occasionally, usually to satisfy a specific funder request.	Ratios are produced with full assumptions disclosed and are never used as the sole basis for a decision.	Ratios are tracked over time per programme, with changes investigated rather than simply reported.	Ratios and their assumptions are published transparently and used to inform sector-wide learning, not only single-organisation reporting.
Comparability across a portfolio	Each programme measures impact in an entirely bespoke way.	Some common outcome categories exist but are applied inconsistently.	A shared outcomes or indicator framework is mandated across the portfolio.	Portfolio-level comparisons explicitly account for differences in scope, context and evidence quality rather than ranking on ratio alone.	The organisation contributes to, and draws on, cross-sector standardised metrics (such as IRIS+) to support comparability beyond its own portfolio.

Checklist

A current, frontline-tested theory of change exists for the programme being measured.
Outcomes are distinguished explicitly from outputs in all reporting.
The decision this measurement exercise will inform is named before the exercise begins.
Every financial proxy used has a disclosed source and a documented rationale.
Deadweight, attribution, displacement and drop-off are calculated and shown, not folded into a single undocumented discount.
Stakeholders whose outcomes are being valued were involved in choosing what to measure and how to value it.
At least one outcome that resists monetisation is reported qualitatively rather than forced into a proxy.
Any published ratio is accompanied by its assumptions and evidence quality, not presented alone.
The evidence standard applied matches the programme's maturity and the stakes of the decision.
SROI ratios are not being compared against other organisations' ratios without checking scope and methodology first.
The theory of change and valuation model are scheduled for review, not treated as permanent.
Independent assurance has been sought for any ratio used in external comparison or public reporting.

Key sources

Social Value International, A Guide to Social Return on Investment and Standards and Guidance — the primary methodological reference for SROI's principles, stages and assurance process.
Nesta, Standards of Evidence — a graded framework for matching the strength of evidence to the maturity of an intervention, widely used across the UK social sector and social investment.
Global Impact Investing Network, IRIS+ system — a shared catalogue of impact metrics used to support comparability across impact investments without forcing every investment into a single monetised ratio.
Arvidson, M., Lyon, F., McKay, S. and Moro, D. (2013), 'Valuing the social? The nature and controversies of measuring social return on investment (SROI)', Voluntary Sector Review, Volume 4, Issue 1 — a peer-reviewed critique of SROI's use in practice, including proxy selection and comparability concerns.
UK Public Services (Social Value) Act 2012 (legislation.gov.uk) — the statutory basis for considering social value in UK public procurement, used here as an example of the pattern, not as a claim about how every jurisdiction regulates commissioning.

References

Social return on investment — Wikipedia — https://en.wikipedia.org/wiki/Social_return_on_investment
Theory of change — Wikipedia — https://en.wikipedia.org/wiki/Theory_of_change
Logic model — Wikipedia — https://en.wikipedia.org/wiki/Logic_model
Social enterprise — Wikipedia — https://en.wikipedia.org/wiki/Social_enterprise
Impact investing — Wikipedia — https://en.wikipedia.org/wiki/Impact_investing
Nonprofit organization — Wikipedia — https://en.wikipedia.org/wiki/Nonprofit_organization
Triple bottom line — Wikipedia — https://en.wikipedia.org/wiki/Triple_bottom_line
B Corporation (certification) — Wikipedia — https://en.wikipedia.org/wiki/B_Corporation_(certification)
Social accounting — Wikipedia — https://en.wikipedia.org/wiki/Social_accounting
Global Impact Investing Network — Wikipedia — https://en.wikipedia.org/wiki/Global_Impact_Investing_Network
Public Services (Social Value) Act 2012 — Wikipedia — https://en.wikipedia.org/wiki/Public_Services_(Social_Value)_Act_2012
A Guide to Social Return on Investment — Social Value International — https://www.socialvalueint.org/guide-to-sroi
Standards of Evidence — Nesta — https://www.nesta.org.uk/report/nesta-standards-of-evidence/
IRIS+ system — Global Impact Investing Network — https://iris.thegiin.org/
'Valuing the social? The nature and controversies of measuring social return on investment (SROI)' — Arvidson, Lyon, McKay and Moro, Voluntary Sector Review — https://bristoluniversitypressdigital.com/view/journals/vsr/4/1/article-p3.xml
Public Services (Social Value) Act 2012 — legislation.gov.uk — https://www.legislation.gov.uk/ukpga/2012/3/contents