gabrielmoreira Regulatory Impact Analysis Skill
Before a rule is made, good practice (and often law) requires showing it's justified: what problem it solves, what it costs, what it delivers, and whether a lighter option would do better. This skill produces a structured RIA — honest about uncertainty, comparing real alternatives against the do-nothing baseline.
Educational analytical aid. A formal RIA must follow the jurisdiction's guidance (e.g. US OMB Circular A-4, UK Better Regulation Framework) and use validated data — treat this as a rigorous first draft, not an official filing.
Required Inputs
Ask for these only if they aren't already provided:
- The proposed rule & problem — what's proposed and the market failure / risk / harm it addresses.
- Options — the realistic alternatives (including status quo / non-regulatory approaches), or ask the skill to develop them.
- Impacts & data — expected costs (compliance, admin, indirect) and benefits (safety, health, efficiency), who bears them, any figures available.
- Timeframe & discounting — the horizon and any required discount rate.
Output Format
Regulatory Impact Analysis: [rule]
1. Problem statement & rationale — the specific problem (market failure, externality, risk) and why intervention is needed now. If there's no clear problem, say so.
2. Objectives — what success looks like, in measurable terms.
3. Options considered — including the baseline (do nothing) and non-regulatory alternatives. Describe each.
4. Costs & benefits by option — for each option, the expected costs and benefits (quantified where possible; qualitative where not), over the timeframe. A comparison table:
| Option | Key costs | Key benefits | Net assessment |
|---|
State assumptions, data sources, and uncertainty honestly (ranges, sensitivity).
5. Distributional effects — who gains and who bears the costs (small business, regions, groups); any equity concerns.
6. Recommendation — the preferred option and why it's proportionate — the best net benefit for the burden imposed.
7. Implementation & review — enforcement, compliance burden, and how/when the rule's effect will be evaluated (sunset/review clause).
Quality Checks
- The problem/market-failure is clearly established before any option is recommended
- Options include the do-nothing baseline and at least one non-regulatory or lighter alternative
- Costs and benefits are compared per option, quantified where data allows, with sources
- Uncertainty is stated honestly (ranges/sensitivity), not hidden behind point estimates
- Distributional effects and a proportionality-based recommendation are included
- A review/evaluation mechanism is specified
Anti-Patterns
- Do not assume regulation is the answer — establish the problem and test the baseline first
- Do not present only the preferred option — compare real alternatives
- Do not fabricate precise numbers — use ranges and label assumptions where data is thin
- Do not ignore who bears the cost — distributional/small-business impact matters
- Do not omit proportionality — the benefit must justify the burden imposed
Based On
Regulatory impact analysis practice (OMB Circular A-4 / Better Regulation): problem-first, options vs. baseline, cost-benefit, proportionality.