Counterfactual Welfare Comparison¶
Method — instantiates Welfare Analysis and Distributional Effects Assessment
Compares the proposed intervention to credible alternatives and the no-action path, separating actual distributional change from baseline choice.
Counterfactual Welfare Comparison holds the intervention's design and its value weights fixed and varies only the thing it is measured against — the current state, a no-action decline path, and one or more credible alternative policies — over a stated time horizon. Its one defining idea is that a distributional gain or loss is never a property of the proposal alone: it is the proposal minus a baseline, so choosing the baseline can manufacture a loss out of thin air or erase a real one. The method's job is to make the comparison point explicit and to run the conclusion against more than one honest counterfactual, so the winners and losers it reports are the intervention's doing and not an artifact of what it was quietly compared to.
Example¶
A city is weighing rent stabilization on its older apartment stock. Advocates compare the policy to today's rents and see tenants protected; opponents compare it to a projected market path and see landlords stripped of future increases. Both are "right," because they are answering different questions. The method refuses to pick one baseline. It assembles three counterfactuals over a ten-year window: (a) status quo continues unchanged, (b) a no-action path where rents track the regional trend and some tenants are gradually displaced, and © an alternative policy pairing a targeted subsidy with upzoning. It then assigns effects to incumbent tenants, small landlords, prospective renters, and future residents under each counterfactual.
The picture separates cleanly. Against the no-action displacement path, incumbent tenants gain substantially. Against the upzoning alternative, future renters are worse off under stabilization because less new supply arrives. The finding is that the stakeholders were never really disagreeing about the policy — they were disagreeing about the baseline. Naming all three surfaces the genuine tradeoff, incumbents versus future entrants, that a single comparison had been hiding.
How it works¶
- Assemble the comparison set. The proposal, the no-action path, and at least one credible alternative — chosen and written down before results are seen, so the set cannot be pruned to flatter the proposal.
- Fix the horizon. A stable time window over which each counterfactual unfolds; the same loss reads very differently at year two than at year ten.
- Hold the value weights constant. Only the comparison scenario varies. How much a future renter's welfare counts relative to an incumbent's is deliberately left untouched here.
- Attribute per party under each scenario. The intervention's effect on a group is its outcome under the proposal minus its outcome under a given baseline — the additionality move.[n1]
- Report robustness. Flag which conclusions survive every counterfactual and which flip depending on the baseline chosen.
Tuning parameters¶
- Comparison set — how many and which counterfactuals. More baselines expose framing effects but multiply forecasting work.
- No-action optimism — how favorable the "do nothing" path is drawn. A pessimistic no-action path flatters every intervention, so its assumptions carry heavy leverage.
- Time horizon — short versus long window. Long horizons capture delayed distributional shifts but stack forecast uncertainty.
- Terminal handling — how effects at the horizon's edge are treated. Ignoring end-point effects can hide costs that land just past the window.
- Attribution method — how outcome-minus-baseline is estimated per group; the harder the counterfactual is to observe, the more this leans on modeling.
When it helps, and when it misleads¶
Its strength is that it kills baseline manipulation, the failure where a comparison point makes one group's loss look smaller or more inevitable than it is. It also reframes stakeholder deadlock productively: it reveals that many "value" disputes are really disputes about the counterfactual, which is a more tractable argument to have.
Its failure mode is that a counterfactual is itself a forecast, and forecasts can be gamed — a conveniently pessimistic no-action path can rescue almost any proposal, and "no action" is rarely the static thing it pretends to be. The classic misuse is quietly selecting the single baseline under which the proposal looks best and presenting it as the comparison. The guarding discipline is to pre-commit the comparison set before results are known and to carry at least two baselines all the way through, so a conclusion that depends on one flattering counterfactual is exposed as fragile rather than reported as fact.
How it implements the components¶
This method realizes the comparison side of the archetype — pinning down against what effects are measured:
baseline_and_counterfactual_frame— its namesake: it defines, pins, and varies the comparison set so gains and losses are attributable to the intervention.temporal_distribution_window— it sets the horizon over which each counterfactual plays out, so timing of gains and losses is part of the comparison.affected_party_map— it assigns effects to each party under every scenario, keeping the comparison distributional rather than aggregate.
It does not vary the normative weights or run the sensitivity over them (welfare_metric_and_value_weights, uncertainty_and_sensitivity_frame) — varying how the result is weighted on a fixed scenario is Value-Weight Sensitivity Analysis, its nearest twin. This method varies the comparison under fixed weights; the twin varies the weights under a fixed comparison.
Related¶
- Instantiates: Welfare Analysis and Distributional Effects Assessment — it supplies the baseline discipline the whole assessment rests on.
- Sibling mechanisms: Distributional Incidence Matrix · Subgroup Disaggregation Audit · Value-Weight Sensitivity Analysis · Compensation Adequacy Review · Externality and Spillover Inventory · Equity Guardrail Test · Public Reason Disclosure Protocol
Editorial Notes¶
Form Classification¶
Form family: Analysis, Modeling & Optimization
Rationale: Counterfactual Welfare Comparison operates as a computation, comparison, model, or analytic representation used to infer, estimate, or choose because it compares the proposed intervention to credible alternatives and the no-action path, separating actual distributional change from baseline choice.
Independent corroboration: The frozen evidence defines Counterfactual Welfare Comparison as 'Compares the proposed intervention to credible alternatives and the no-action path, separating actual distributional change from baseline choice', 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: Welfare economics cohered comparison of interventions against no-action and credible alternatives while separating total surplus from distributional change.
Related originating lineages:
- Public Administration & Policy — Policy appraisal institutionalizes baseline selection, stakeholder incidence, and distributional-impact comparison.
Review resolution: Both reviewers agree on welfare economics and policy appraisal. Public administration is formative because the mechanism requires credible policy baselines and stakeholder incidence, so cross-disciplinary provenance is more precise than a purely economic lineage.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] Additionality is the share of an observed outcome attributable to an intervention beyond what would have happened anyway under the baseline — a workhorse concept in program evaluation and carbon-offset accounting. A welfare gain that the baseline would have produced regardless is not additional, and crediting it to the proposal is baseline manipulation by another name. ↩