Most SROIs I’ve seen are shallow.

Not because the method is shallow, but because most programs feed it with self-reports, pre-post comparisons, borrowed impact estimates, and global-average proxies. You get a ratio resting on air.

SROI done well is different. It integrates everything upstream — and rewards you when you’ve done a rigorous evaluation first.

Using Hogares Saludables — a housing program in Colombia, we evaluated over four years:

The upstream evidence that fed the SROI:

→ A cluster RCT, pre-registered with AEA (N=1,163). Causal effects on mental health, employment, child health, and debt — at p<0.01. The program’s own effects; no borrowing.

→ A quasi-experimental DiD-PSM on food security, with outcome built via Item Response Theory.

→ An observational study with structural engineers + AI image analysis of 1,035 homes — linking physical quality to welfare.

→ A qualitative module using a method we invented: “Home Biographies” — Most Significant Change + photo-elicitation + body mapping.

→ A three-arm RCT isolating psychosocial activation from the material intervention — surfacing a null and adverse effect cleaner designs miss.

The SROI adds the financial translation: national proxies (WELLBY Colombia, GEIH-DANE 2024, ADRES) + Social Value International adjustments (deadweight, 15% attribution, 25% drop-off, 5% discount) across 3- and 5-year scenarios, with sensitivity analysis. All conservative.

The result, conservative scenario at 3 years:

For every $1 invested, $3.19 in social value is generated. At 5 years: $4.08. Drivers: 79% mental health, 14% employment, plus a catalyst effect as families reinvested on their own.

The methodological point:

SROI isn’t the opposite of academic rigor. It’s what academic rigor produces when you ask it the right final question.

The ratio is only as credible as the evidence feeding it. Feed it with an RCT, and you get a number a CFO can defend. Feed it with a self-report survey, and you get a number nobody should publish.

Most public-private partnerships fund one or the other. They should fund both as parts of the same architecture.

Stop treating SROI as a marketing tool. Treat it as a capstone. Invest in the upstream evidence that makes it worth doing at all.

— 📚 The papers behind the SROI: → Housing quality & welfare: https://lnkd.in/eGUijuGd → RCT main effects (R&R AEJ: Applied): https://lnkd.in/ejD_UH4A → Home Biographies (qualitative, open access): https://lnkd.in/eZ5jiGKr → Three-arm RCT on violence: IDB Working Paper forthcoming

Co-authors: @Albert Saiz (MIT), Marcela Ibáñez (Göttingen), @Natalia Cantet (Augusta), María José Rubiano (Exeter), Gustavo García, Daniela Mejía and team.

📋 RCT registry: https://lnkd.in/ePRgZFMJ 🏛️ @Valor Público 🏢 @Cementos Argos

#SROI #ImpactEvaluation #SocialValueInternational #DevelopmentEconomics