I ran a Monte Carlo simulation calibrated to the actual adjustment factor distributions in the corpus:

→ Deadweight median: 20% → Attribution: 25.5%
→ Drop-off: 21.5%

Applied to non-compliant reports:

Observed median ratio: 3.56:1 Corrected median ratio: 1.92:1

That is an 85% overstatement at the median.

Formal assurance helps — assured reports score 10.4 percentage points higher. But even assured reports reach only 22.9% of the maximum standard.

Full paper: https://lnkd.in/ezDhzyX9

#SROI #SocialReturn #ImpactMeasurement #NonprofitData