When this rubric applies
The argument’s conclusion is "we should / should not do X" where X is a policy choice, and the disagreement turns on weighing benefits, costs, and unintended consequences against a stated set of priorities (efficiency, equity, liberty, security, growth, etc.).
Not appropriate when: the disagreement is about whether a policy is legally permitted (use STATUTORY-INTERP or CONSTITUTIONAL-STRUCTURE), or whether an empirical premise of the policy case is true (use SCIENTIFIC-CLAIM).
Criteria
1. Surfaces and weights the tradeoff honestly
weight 30%
Does the argument acknowledge what is given up to get what it wants?
10Names the costs of the preferred policy explicitly; assigns a weight; explains why the gain outweighs the loss.
7Names the major cost; the weighting is implicit but defensible.
4Mentions costs only to dismiss them; treats the tradeoff as a free win.
1Pretends there is no tradeoff at all.
2. Engages opposing costs and benefits
weight 25%
Does the argument take the strongest version of the other side’s case and respond to it?
10States the strongest opposing argument, accurately, and answers it.
7Engages a real but not the strongest version of the opposing argument.
4Strawmans the opposing position and beats the strawman.
1Doesn’t acknowledge the other side has any costs or benefits worth weighing.
3. Empirical grounding of stated effects
weight 25%
When the argument predicts effects, are the predictions backed by evidence or asserted?
10Cites specific data, comparable jurisdictions, or named studies for predicted effects; quantifies where possible.
7Cites real evidence but applies it to a different context without acknowledging the gap.
4Predicts effects with no empirical basis; assumes the policy will work as designed.
1Predicts effects that contradict the available evidence.
4. Consistency with stated values
weight 20%
If the argument leans on a value (liberty, equity, efficiency), does it apply the value consistently across the case?
10The asserted value supports the conclusion in this case AND would support symmetric conclusions in adjacent cases.
7The value supports the conclusion; one adjacent case where it would force a conclusion the speaker resists is not addressed.
4The value is invoked selectively — would force the opposite conclusion in obvious adjacent cases the speaker rejects.
1The "value" is post-hoc cover for the preferred outcome.
Version history
v2February 2026 · raised the weight on "surfaces the tradeoff" from 20% to 30% after pilot runs showed the rubric was too forgiving of one-sided framings
v1July 2025 · initial publication