Social Return on Investment

Social Return on Investment in Health and Social Care

Social return on investment, usually shortened to SROI, adds up the changes a programme brings about for the people it affects, gives each change a money value and sets the total against what the programme cost. The result is reported as a ratio, such as GBP 3 of social value for every GBP 1 invested, together with an account of how each value was reached. In health and social care it is used by charities, social enterprises and public health teams, among others, to show funders what a programme achieves beyond clinical outcomes, for example in loneliness or reduced use of other services. Its arithmetic is close to cost-benefit analysis, but stakeholders help choose and value the outcomes, and the ratio is not built to rank one programme against another.

Origins and the principles behind the method

The UK guide to SROI, written by Nicholls, Lawlor, Neitzert and Goodspeed in 2009 within a measuring social value programme funded by the Office of the Third Sector, describes SROI as a framework for measuring and accounting for a broader concept of value than financial returns alone. The guide credits Jed Emerson and the Roberts Enterprise Development Fund with starting the method, and the new economics foundation and the SROI Network with developing it. It states that SROI was developed from social accounting and cost-benefit analysis, and treats money as a common unit for conveying value.

The 2009 guide set out seven principles, which were involve stakeholders, understand what changes, value the things that matter, only include what is material, do not over-claim, be transparent and verify the result. Social Value International now lists an eighth, be responsive. An evaluative SROI uses outcomes that have already occurred, while a forecast SROI predicts the value a programme will create if it meets its intended outcomes.

Six stages of an SROI analysis

The 2009 guide organises the work in six stages, recorded in an impact map, a table linking inputs, outputs and outcomes for each stakeholder group. The stages run in this order.

1. Establish the scope and identify stakeholders. The analysis sets its boundaries and decides who will be involved.
2. Map outcomes. Working with stakeholders, the analyst builds a theory of change linking inputs, outputs and outcomes.
3. Evidence outcomes and give them a value. Indicators show whether each outcome happened and how long it lasts, and a financial proxy is chosen for each.
4. Establish impact. Change that would have happened anyway, or that is due to others, is removed.
5. Calculate the SROI. Values are projected, discounted and compared with the investment, and sensitivity is tested.
6. Report, use and embed. Findings go back to stakeholders and the report is verified.

Financial proxies are where SROI meets monetary valuation in health economics. The guide notes that valuation has a long tradition in environmental and health economics, and names contingent valuation, which asks people their willingness to pay or to accept compensation, and revealed preference methods such as hedonic pricing. In practice many proxies are service costs, and the guide's own worked example values fewer hospital admissions with figures from the NHS cost book.

Deadweight, attribution, displacement and drop-off

The fourth stage is where SROI deals with the counterfactual. The guide defines deadweight as the amount of outcome that would have happened even if the activity had not taken place, and notes that this is sometimes called the counterfactual. Attribution is the share of the outcome caused by other organisations or people. Displacement is how far the outcome displaced other outcomes, as when crime falls in one borough but rises in the next. Drop-off reduces outcomes in later years and applies only to outcomes lasting more than one year.

The guide notes that the Treasury Green Book's idea of additionality is comparable with impact in SROI, and allows estimates to be rounded to the nearest 10% unless better information exists. For outcome $k$ in the first year the impact is

$$I_{k,1} = Q_k \times P_k \times (1 - d_k) \times (1 - a_k) \times (1 - s_k)$$

where $I_{k,1}$ is the impact of outcome $k$ in year 1 in pounds, $Q_k$ is the quantity of the outcome, $P_k$ is its financial proxy per unit, $d_k$ is deadweight, $a_k$ is attribution and $s_k$ is displacement, each a proportion between 0 and 1. Taking each deduction from what remains follows the guide's worked example, where GBP 772 less 10% deadweight is GBP 694.80 and that less 35% attribution is GBP 451.62.

$$I_{k,t} = I_{k,1} \times (1 - \delta_k)^{t-1}$$

where $I_{k,t}$ is the impact of outcome $k$ in year $t$, zero after the outcome ends, and $\delta_k$ is the annual drop-off, taken as a fixed percentage of the outcome remaining at the end of each year.

Discounting and the ratios reported

Impacts spread over several years are converted to a present value by discounting. The guide discounts the first year's impact as well as later ones, so the sum starts at $t = 1$.

$$PV = \sum_{t=1}^{T} \frac{\sum_{k} I_{k,t}}{(1 + r)^{t}}$$

where $PV$ is the present value of all impacts, $T$ is the last year any outcome lasts, $r$ is the discount rate and the inner sum adds the impacts of every outcome in year $t$. The guide gave the basic public sector rate in the Green Book at the time as 3.5%, while calling discounting controversial because it can encourage short-termism.

$$\text{SROI ratio} = \frac{PV}{C}, \qquad \text{NPV} = PV - C, \qquad \text{net SROI ratio} = \frac{PV - C}{C}$$

where $C$ is the value of the inputs, the investment, and NPV is the net present value. The guide accepts either ratio if the report says which is used, and the two are easily confused, since a net ratio of 0.5 and an SROI ratio of 1.5 describe the same result. An optional payback period divides the investment by the monthly impact.

Worked example with a falls-prevention exercise programme

All figures in this example are illustrative. A community exercise and social programme for older people costs GBP 60,000 for one year. Two outcomes are judged material, fewer falls needing a hospital admission and reduced loneliness, and values are discounted at 3.5%.

1. Fewer admissions after a fall. The programme avoids 12 admissions in the year, valued at GBP 3,000 each. With 25% deadweight and 20% attribution to a GP falls clinic, the impact is 12 × 3,000 = GBP 36,000, then 36,000 × 0.75 = GBP 27,000, then 27,000 × 0.8 = GBP 21,600. It lasts three years with 20% drop-off, giving GBP 21,600, GBP 17,280 and GBP 13,824.

2. Reduced loneliness. Forty participants report less loneliness, valued at GBP 500 each a year. With 30% deadweight and 10% attribution, the impact is 40 × 500 = GBP 20,000, then 20,000 × 0.7 = GBP 14,000, then 14,000 × 0.9 = GBP 12,600. It lasts two years with 50% drop-off, giving GBP 12,600 and GBP 6,300.

3. Discount. Yearly totals of GBP 34,200, GBP 23,580 and GBP 13,824, divided by 1.035, 1.071225 and 1.108718, give GBP 33,043, GBP 22,012 and GBP 12,468, a present value of GBP 67,524 (summed before rounding).

4. Ratios. The SROI ratio is 67,524 ÷ 60,000 = 1.13, about GBP 1.13 of social value per GBP 1 invested. The NPV is 67,524 − 60,000 = GBP 7,524 and the net SROI ratio is 7,524 ÷ 60,000 = 0.13.

5. Sensitivity. The guide recommends finding how far each estimate must move for the ratio to fall to 1. The admissions outcome contributes GBP 49,469 of present value and loneliness GBP 18,055, so the admissions present value must fall to 60,000 − 18,055 = GBP 41,945. Solving 0.8 ÷ (1 − a) = 49,469 ÷ 41,945 gives attribution of about 32% instead of 20%. The same point is reached if the proxy falls to about GBP 2,544, a cut of 15%.

6. Interpretation. Under these assumptions the programme creates slightly more value than it costs, but a modest change to one estimate removes the margin, and the report would need to say so.

What a hospital cost proxy measures

The admission proxy above is an average unit cost, and that choice deserves scrutiny. Arvidson, Lyon, McKay and Moro, in a 2010 Third Sector Research Centre working paper, note that the 2009 guide claims GBP 4,964 for each geriatric assessment avoided, taken from the NHS cost book. They argue, as the guide itself concedes, that this cannot be read as an NHS saving, because short-term savings relate mainly to variable or marginal costs, and that an average cost does not reflect the beneficiary's own valuation either.

In the example, replacing the GBP 3,000 average cost with an illustrative GBP 1,000 cash-releasing saving cuts the admissions outcome to 49,469 ÷ 3 = GBP 16,490. The present value becomes 16,490 + 18,055 = GBP 34,545 and the ratio 34,545 ÷ 60,000 = 0.58. The two figures answer different questions, so the perspective and meaning of each proxy need stating, and the gap between average cost and marginal cost matters here as it does in cost-effectiveness work.

How SROI relates to cost-benefit and cost-utility analysis

Arvidson and colleagues argue that from a technical point of view there is not much difference between classical cost-benefit analysis and SROI. They place the differences in style, such as the stronger emphasis on stakeholders, its use as a management tool, and SROI guidance that does not recommend comparing ratios across activities, whereas cost-benefit analysis is designed for such comparison. The 2009 guide likewise contrasts its bottom-up stakeholder view with Green Book appraisal, which values costs and benefits to the whole of UK society.

Banke-Thomas, Madaj, Charles and van den Broek, in a 2015 systematic review in BMC Public Health, compare SROI with cost-effectiveness analysis, cost-utility analysis and cost-benefit analysis. In their table only SROI uses stakeholder engagement and a theory of change, and SROI uses financial proxies for benefits that cost-benefit analysis would list as hard to monetise. Cost-effectiveness and cost-utility analysis report an incremental cost-effectiveness ratio, in cost-utility analysis per QALY, comparing alternatives within the same domain.

The difference that matters most for decisions is the comparator. Cost-utility analysis compares a programme with a named alternative and weighs the extra cost against the health the money could buy elsewhere, its opportunity cost. An SROI ratio sets value against the programme's own cost after deadweight, so a ratio above 1 shows that estimated value exceeded cost but says nothing about whether the money would have done more in another use. The page on return on investment covers the narrower financial ratio of monetary gains to a specified payer.

What reviews of health applications have found

Banke-Thomas and colleagues found 40 public health SROI studies published to the end of 2014, most often in health promotion and mental health. Reported ratios varied widely, the lowest 1.1 to 1 and the highest 65 to 1, and the authors judged it inappropriate to compare them to find the intervention with the most value for money. They found no consensus on whom to include as beneficiaries, how to account for the counterfactual or what time horizon to use.

Hutchinson and colleagues, in a 2019 systematic review in BMJ Open, screened 868 papers and found only 8 peer-reviewed SROI studies of health and social care programmes. Quality ranged between 38% and 90% on their purpose-designed tool, with weaknesses in justifying stakeholders, reporting sample sizes, undertaking sensitivity analysis and reporting unexpected or negative outcomes. Ashton and colleagues, in a 2020 scoping review of public health interventions, found that 37 of 40 included SROI studies were published in the grey literature.

Limits and common misreadings

Most problems with SROI arise when the ratio travels without the assumptions behind it. Each misreading below has been raised in the guidance or in reviews of health applications.

1. Comparing ratios between organisations. The 2009 guide states that it is not appropriate to compare social return ratios alone, because organisations work with different stakeholders and make different judgements.
2. Reading social value as cash. A ratio of 3 to 1 does not mean a funder recovers GBP 3, since most proxies are values or average costs, not budget savings.
3. Treating the ratio as an allocation rule. Without a comparator, a ratio above 1 cannot show that a programme is the best use of a health budget.
4. Weak counterfactuals. Banke-Thomas and colleagues report that most studies estimated what would have happened anyway by subjective means such as stakeholder consultation, and favour before-and-after or control group designs.
5. Borrowed proxies. A proxy taken from another setting needs the same checks as any benefit transfer, including its population, price year and what it measures.

Sources

• Arvidson M, Lyon F, McKay S, Moro D. The ambitions and challenges of SROI. Third Sector Research Centre Working Paper 49. Birmingham: University of Birmingham; December 2010. https://epapers.bham.ac.uk/id/eprint/788/
• Ashton K, Schröder-Bäck P, Clemens T, Dyakova M, Stielke A, Bellis MA. The social value of investing in public health across the life course: a systematic scoping review. BMC Public Health. 2020;20:597. https://doi.org/10.1186/s12889-020-08685-7
• Banke-Thomas AO, Madaj B, Charles A, van den Broek N. Social Return on Investment (SROI) methodology to account for value for money of public health interventions: a systematic review. BMC Public Health. 2015;15:582. https://doi.org/10.1186/s12889-015-1935-7
• Hutchinson CL, Berndt A, Forsythe D, Gilbert-Hunt S, George S, Ratcliffe J. Valuing the impact of health and social care programs using social return on investment analysis: how have academics advanced the methodology? A systematic review. BMJ Open. 2019;9:e029789. https://doi.org/10.1136/bmjopen-2019-029789
• Nicholls J, Lawlor E, Neitzert E, Goodspeed T. A guide to Social Return on Investment. London: Office of the Third Sector, Cabinet Office; 2009. https://www.bond.org.uk/wp-content/uploads/2022/09/cabinet_office_a_guide_to_social_return_on_investment.pdf
• Social Value International. The Principles of Social Value. Accessed 28 October 2026. https://www.socialvalueint.org/principles