Offset-Adjusted Impact Evaluation¶
Impact-evaluation model — instantiates Anticipatory Offset Governance
Judges the intervention on the effect that survives anticipatory offset — specifying the intended net effect up front and evaluating realized outcomes against it after subtracting what targets pre-empted or displaced.
An Offset-Adjusted Impact Evaluation is the scorecard that refuses to credit a gross number. Its defining move is to state, before launch, what the intervention is really meant to achieve — the intended net effect, expressed as the true objective rather than a convenient proxy — and then to evaluate realized outcomes through a net-effect frame that subtracts the anticipatory offset: the displacement, substitution, and behavioral compensation that make a local win overstate the whole-system result. Where the audit measures one slice (the announcement window) and the simulation forecasts, this evaluation is the accounting layer that consumes those inputs and delivers the verdict: not "did the measured proxy move," but "did the intended net effect actually land."
Example¶
A road authority installs speed cameras to reduce injuries. The naïve evaluation counts lower speeds and fewer violations at the camera sites and declares success. The offset-adjusted evaluation starts differently: it fixes the intended net effect up front as fewer injuries across the whole road network, not lower speeds at cameras — closing the door on proxy-worship. Then its net-effect frame subtracts the offsets the cameras provoke: drivers braking only at the camera and accelerating between them (spatial displacement), and the general tendency to drive a little less carefully when a hazard feels managed.[n1]
Measuring injuries network-wide against the pre-stated target, it finds a genuine but smaller net gain — roughly half the site-level headline — because a share of crashes simply moved down the road rather than disappearing. The reported figure is the offset-adjusted net effect, judged against the objective the program actually had.
How it works¶
- Specify the intended net effect first. Write down the true objective and the system boundary before launch, so success cannot be redefined afterward to whatever the data happens to show.
- Assemble the net-effect frame. Lay out every offset that must be subtracted from the gross result — anticipation, spatial and channel displacement, substitution, behavioral compensation.
- Net the consumed evidence. Pull the counterfactuals and window measurements from the simulation, the audit, and the rollout's comparison, and subtract the offsets they quantify.
- Report net against intent. State the surviving net effect relative to the intended-net-effect specification, not the flattering gross proxy.
Tuning parameters¶
- System boundary — how wide to draw the frame for displacement. Too narrow hides the crashes that moved down the road; too wide dilutes the signal into unrelated noise.
- Offset completeness — which offsets to net out. Omitting a hard-to-see channel flatters the result.
- Target strictness — whether the intended effect is the true outcome or a measurable proxy, and how tightly the two are bound.
- Attribution window — how long after the effective date outcomes are counted, trading completeness against timeliness.
- Net vs gross reporting — whether both figures are shown or only the net; showing both makes the offset itself visible.
When it helps, and when it misleads¶
Its strength is refusing the false victory: it keeps a program from being celebrated on a gross number that offset has quietly hollowed out, and it ties the verdict to the objective the intervention actually had rather than to the proxy that was easy to move.
Its failure modes live in the boundary and the accounting. Draw the system too narrowly and displacement escapes the frame — the evaluation flatters; draw it too widely and the effect drowns in noise. Offsets that resist pricing (morale, trust, long-run adaptation) slip through even a careful net. The classic misuse is choosing the boundary or the proxy after seeing results, to manufacture success. The discipline is to pre-register the intended net effect and the system boundary, and to report gross and net side by side so the offset is on the page, not buried.
How it implements the components¶
intended_net_effect_specification— the pre-stated true objective and system boundary the whole verdict is measured against.net_effect_evaluation_frame— the accounting frame that subtracts anticipation, displacement, and substitution from the gross result to leave the net effect.
It nets the evidence; it does not generate it. The offset_counterfactual_model it subtracts comes from Pre-Implementation Response Simulation (ex ante) and Announcement Effect Audit (ex post), and the not-yet-treated comparison it leans on is built by Staggered or Randomized Rollout.
Related¶
- Instantiates: Anticipatory Offset Governance — it delivers the verdict on the intervention's net effect, with offset built into the measure rather than discovered too late.
- Consumes: Announcement Effect Audit and Pre-Implementation Response Simulation supply the counterfactuals it nets; Staggered or Randomized Rollout supplies the comparison.
- Sibling mechanisms: Announcement Effect Audit · Pre-Implementation Response Simulation · Staggered or Randomized Rollout · Anticipatory Offset Dashboard · Strategic Response Red Team
Editorial Notes¶
Form Classification¶
Form family: Analysis, Modeling & Optimization
Rationale: Offset-Adjusted Impact Evaluation operates as an analytical, modeling, inference, comparison, or optimization procedure that derives insight or a solution because it judges the intervention on the effect that survives anticipatory offset — specifying the intended net effect up front and evaluating realized outcomes against it after subtracting what targets pre-empted or displaced.
Independent corroboration: The frozen evidence defines Offset-Adjusted Impact Evaluation as 'Judges the intervention on the effect that survives anticipatory offset — specifying the intended net effect up front and evaluating realized outcomes against it after subtracting what targets pre-empted or displaced', so its operative form is Analysis, Modeling & Optimization.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Economics & Finance
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Economics developed net-effect evaluation that subtracts displacement, substitution, crowd-out, and behavioral offset from an intervention's gross effect.
Related originating lineages:
- Behavioral Economics — Risk-compensation research explains anticipatory changes in target behavior that can undo part of a safety intervention.
- Public Administration & Policy — Program evaluation contributed pre-specified outcomes, implementation monitoring, and policy decisions based on realized rather than announced effect.
- Statistics & Experimental Design — Causal evaluation contributes counterfactual estimation of the net effect that survives adaptation.
Review resolution: Both independent reviews agree on primary origin economics_finance; reconciliation resolves reported_ambiguity, alternate_origin_disagreement. Formative alternate lineages retained: behavioral_economics, public_administration_policy, statistics_experimental_design. The broader reach of later applications is kept separate as domain_reach=multi_domain; origin_mode=cross_disciplinary_synthesis describes the historical relationship among lineages. Confidence is conservatively reconciled to medium, and encyclopedia_synthesis=true preserves the reviewers' boundary judgment.
Attribution caveat: The named evaluation is an encyclopedia synthesis over economic offset behavior and causal policy appraisal.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; medium confidence.
Notes¶
This evaluation is the terminal accounting for the archetype: the red team's list, the simulation's forecast, the audit's window measurement, and the rollout's comparison all flow into it. Because it fixes the intended net effect before launch, it is also the yardstick that later reveals whether the forecasting and monitoring mechanisms were themselves well-calibrated.
[n1] Risk compensation — a safety measure being partly offset by the riskier behavior it enables — is the Peltzman effect (after Sam Peltzman) in road-safety economics. It is one of the offsets a net-effect frame must subtract from a gross safety gain, alongside spatial displacement. ↩