| name | comparative-policy-research |
| description | Method for researching how other countries have tackled a policy problem — selecting comparators for mechanism rather than prestige, writing per-jurisdiction case studies with the instrument, the evidence of results, the failures, and the transferability caveats for Israel. Load when building the comparative half of a policy evidence base, when asked "how do other countries handle this", or when writing policy/<area>/research/03-comparative.md and its case-study files. |
Comparative policy research
The comparative review answers one question: has anyone actually solved this, and by what mechanism? It is not a tour of admired countries. A comparator earns its place because it tried a specific instrument on a recognisably similar problem and there is evidence of what happened next.
Follows the evidence standards in the policy-methodology skill. Everything there applies here; this skill adds the comparative-specific method.
Outputs
policy/<area>/research/
├── 03-comparative.md the synthesis — read this alone and you have the argument
└── comparative/
├── <country-or-city>.md one case study per comparator
└── ...
The synthesis is the deliverable; the case studies are its evidence. Do not write only a synthesis — an unfootnoted summary of five countries is exactly the artifact this project exists to avoid.
Selecting comparators
Pick 4–6 case studies. Choose on mechanism similarity and evidence availability, not on GDP or reputation.
Aim for a spread across these roles, and say in the synthesis which role each comparator plays:
- The canonical success — the instrument worked, and there is measured evidence that it did.
- The near-neighbour in constraints — similar population size, density, fiscal capacity, political fragmentation, or a comparable coalition system. Israel's structural analogues are more often small dense European states than the US or UK.
- The instructive failure — a country that tried this and it did not work, or worked and was repealed. At least one is mandatory. A comparative review with no failures in it is advocacy.
- The over-cited case — whatever comparator gets name-checked in the domestic debate. Research it precisely so the paper can correct the record, whichever way it falls.
- A non-Western case where relevant — Singapore, South Korea, Chile, Estonia. Avoid the reflex of comparing Israel only to Western Europe.
Record the ones you rejected and why, in a "Comparators considered and rejected" section. Selection is an analytic choice and must be visible.
What each case study must contain
Each file in comparative/ uses this structure. A section with nothing in it says "no evidence found", never nothing.
- Frontmatter —
jurisdiction, instrument, adopted (year), status (in force / amended / repealed), last-verified (date), role (which of the roles above).
- The problem as they framed it — in their terms, not ours. Framing differences are often the finding.
- The instrument, mechanically — the actual law, regulation, agency, tariff or contract clause. Official name in the original language with translation, year, and the operative provisions. Who administers it, how it is enforced, how it is funded. This is the section the policy paper will lift from; vagueness here makes the case study useless.
- Adoption politics — who proposed it, who opposed it, what coalition passed it, how long it took, and whether it required a referendum, a crisis, or a court ruling. The Israeli question is always passability, so the how it got through is as transferable as the policy.
- Evidence of results — measured outcomes with sources and dates, before and after. Distinguish the policy's own advocates' claims from independent evaluation (audit office, national statistics agency, peer-reviewed studies). Where the evidence is contested, present both sides.
- What went wrong — implementation failures, cost overruns, unintended consequences, exemptions that hollowed it out, subsequent amendments that weakened it. Every real policy has this section.
- Transferability to Israel — a judgement, argued. Address at least: population and density differences; fiscal capacity; the coalition/PR system versus their governing system; administrative capacity of the relevant Israeli ministry or municipality; legal-system compatibility (civil-law statute vs. common-law precedent vs. EU-directive-driven); and any Israel-specific factor (security, religious-political constraints, Arabic/Russian/Hebrew multilingualism, periphery-centre gaps).
- Sources — full list with links and access dates.
The synthesis (03-comparative.md)
Structure it as an argument, not a country-by-country recital:
- The comparative question, stated in one paragraph.
- Comparator table — jurisdiction, instrument, year, status, one-line result, transferability verdict (high / partial / low, with the criterion stated).
- Findings across cases — 3–6 numbered findings, each supported by more than one case where possible. These are the transferable claims the paper will actually make.
- What does not transfer, and why — the honest constraint list.
- The instructive failures — its own section, prominently placed.
- Comparators considered and rejected, with reasons.
Rules specific to comparative work
- Never cite a country from memory. Instruments change name, get amended, get repealed. Verify current status and date it: "as of ". Search and fetch; do not recall.
- Use original-language official sources where they are authoritative — the ministry, the statistics agency, the audit office — and quote the original name alongside the English.
- Beware the policy-tourism source. Advocacy NGOs, consultancy white papers and press features recycle each other's claims. Trace every headline number to a primary source or mark it
[UNVERIFIED — verify before publication].
- Report effect sizes, not directions. "Ridership rose" is not a finding; "ridership rose 3% in year one against a 1.2% national trend" is.
- Distinguish the instrument from the context that made it work. Vienna's fare policy sits on a network Israel does not have. Say so explicitly rather than letting the paper imply the policy alone produced the outcome.
- Currency and units are converted — give the original figure, then the ILS equivalent with the conversion date.