Definition
The share of claimed AI program effects that survive a controlled or quasi-experimental attribution test, against all claimed effects entering agency reporting or budget justification.
The Measurement Gap
There is no metric that tells you how much of a change in response time, fire loss, or survival is attributable to an AI program rather than to staffing, station, traffic, weather, demographic, or concurrent protocol changes, which means before-and-after comparison assigns the full effect to the program and the agency reinvests on an unverified attribution.
Formula
IEAS = E_survived / E_claimed ; with the attribution method recorded per claim
Proposed formula, pending validation pilots.
At a Glance
Decision Supported
Whether an AI program should be expanded, sustained, or discontinued, and whether the effect claimed in a budget request is defensible.