How We Decide

A serious policy should survive serious questions.

The method exists to improve judgement, not to imitate certainty or replace democracy.

01

Define the problem

State what is happening, who is affected, likely causes and what remains uncertain.

02

Establish authority and constraints

Ask why government should act, which institution has power and which rights or legal limits apply.

03

Define outcomes and the counterfactual

Describe the outcome and the likely future without a new policy.

04

Generate real options

Compare genuinely different approaches rather than a preferred policy with a straw man.

05

Build the evidence case

Match evidence to each descriptive, causal, delivery, forecast or value claim.

06

Appraise consequences

Map benefits, costs, opportunity costs, administration and long-term effects.

07

Test distribution and rights

Identify winners, losers, risk bearers and less restrictive alternatives.

08

Test behaviour and misuse

Examine incentives, avoidance, gaming, bureaucracy and hostile use.

09

Test delivery

Identify powers, money, people, technology, procurement, transition and ownership.

10

Test legitimacy and challenge

Separate public opinion, consultation, affected experience and expert scrutiny.

11

Make the judgement

State the preferred option, principal cost, uncertainty and political reasons.

12

Monitor, evaluate and end

Set outcomes, guardrails, review dates and conditions for change or withdrawal.

Proportionate scrutiny: a small reversible decision should not face the same process as a national programme, a serious restriction of rights or constitutional reform.

Evidence is not sovereign

Evidence can show consequences, uncertainty and delivery risks. It cannot settle every value. Elected leaders remain responsible for the judgement.

No automatic score

A benefit-cost ratio, model or public-opinion result cannot mechanically select policy. Rights, legitimacy, distribution and political judgement remain visible.