thedixitjain AEJ: Policy Workflow Router (aejpol-workflow)
Overview
This is the router. It tells you which aejpol- skill to use at the current stage* of a manuscript aimed at the American Economic Journal: Economic Policy (AEJ: Policy) — the American Economic Association's quarterly journal for the economic analysis OF policy. AEJ: Policy rewards a paper built around a clear policy question whose answer carries a welfare, cost-benefit, or distributional implication, identified credibly (quasi-experimental / RCT / applied theory) and reproducible to AEA standards. The policy question and its counterfactual are the through-line of every skill below.
Default assumption: unless the user says otherwise, treat the target as AEJ: Policy. Operational tells that you are at AEJ: Policy and not a sibling: review is single-blind through the AEA submission system (the manuscript carries the title/byline/affiliations); JEL codes are required; the AEA Data and Code Availability Policy and the AEA Data Editor run a pre-publication reproducibility check on a deposit in the AEA Data and Code Repository (openICPSR); the bar is broad policy interest + credible causal evidence + a welfare/cost-benefit reading, not a field-public-finance bar (J. Public Economics) or an identification-for-its-own-sake bar (AEJ: Applied). Re-verify volatile specifics (submission fee, editors) on the official AEA pages — official AEA pages checked 2026-06-20.
When to trigger
- The user asks "what should I do next?"
- A draft needs its current bottleneck diagnosed
- Work is ping-ponging between identification, the policy framing, exhibits, and the response letter
- An AEJ: Policy decision letter arrived and the user needs to switch into revision mode
Routing table
| Current symptom | Next skill |
|---|---|
| No sharp policy question, or fit vs. JPubE / AEJ:Applied / AER unclear | aejpol-topic-selection |
| Contribution vs. the policy-evaluation literature is fuzzy or undersold | aejpol-literature-positioning |
| Causal design (DID / IV / RDD / RCT) for the policy's effect is shaky | aejpol-identification |
| Need a model/welfare framework to read estimates as a policy effect | aejpol-theory-model |
| Estimate stability, sensitivity, and threats need stress-testing | aejpol-robustness |
| Exhibits are dense; significance asterisks still present | aejpol-tables-figures |
| Prose buries the policy takeaway; abstract/intro do not land | aejpol-writing-style |
| Data/code deposit, README, AEA Data Editor prep, restricted-data path | aejpol-replication-package |
| Want to anticipate the referees' objections before submitting | aejpol-referee-strategy |
| Ready to submit via the AEA system; need a preflight | aejpol-submission |
| Received an R&R; need a response-letter strategy | aejpol-rebuttal |
Default order
aejpol-topic-selection— lock the policy question + the welfare/cost-benefit stakeaejpol-literature-positioning— stake the contribution vs. the policy-evaluation frontieraejpol-identification— credible quasi-experimental / RCT evaluation of the policyaejpol-theory-model— the framework that maps estimates to a welfare/policy objectaejpol-robustness— sensitivity, threats, specification stabilityaejpol-tables-figures— exhibits without significance asterisks; a policy-relevant headline exhibitaejpol-writing-style— translate estimates into a clear policy takeaway (abstract + intro last)aejpol-replication-package— assemble the AEA-compliant depositaejpol-referee-strategy— pre-empt the objections referees will raiseaejpol-submission— AEA system preflightaejpol-rebuttal— after the R&R
aejpol-writing-styleis a late-stage polish; do not rewrite the intro before identification/robustness settle.
Anti-patterns
- Treating AEJ: Policy like J. Public Economics (field-only, no broad-interest policy framing) or AEJ: Applied (identification-driven applied micro with no policy question)
- Deferring
aejpol-replication-packageto acceptance — the AEA Data Editor checks before publication - Letting
aejpol-tables-figurespolish exhibits while identification is still moving - Reporting a clean causal estimate with no welfare / cost-benefit / distributional reading — the AEJ: Policy "so what"
Routing by paper archetype
The bottleneck differs by how the policy is identified. Read the archetype, then enter the chain at the right link.
| Archetype | Likely first bottleneck | Enter at |
|---|---|---|
| reform / law change (staggered) | TWFE bias; modern DID; pre-trends | aejpol-identification |
| eligibility threshold / formula | RDD / bunching design + density test | aejpol-identification |
| field experiment / RCT of a program | estimand + pre-registration + welfare cost | aejpol-topic-selection → aejpol-identification |
| sufficient-statistics / structural welfare | mapping reduced-form estimate to welfare | aejpol-theory-model |
Worked routing example (illustrative)
A user says: "My estimate of a tax-credit expansion on employment is clean, but a referee says the paper 'has no policy point' and the welfare claim is hand-waved." Two AEJ: Policy pushbacks — no welfare/cost-benefit reading and takeaway not framed as a policy lesson — owned by aejpol-theory-model (the welfare mapping) and aejpol-writing-style (the takeaway). Route to aejpol-theory-model to turn the employment elasticity into a marginal-value-of-public-funds / cost-per-job number, then aejpol-writing-style for the policy sentence.
Minimal decision snippet
if decision_letter_arrived: -> aejpol-rebuttal
elif ready_to_submit: -> aejpol-submission
elif anticipating_referees: -> aejpol-referee-strategy
elif exhibits_or_significance: -> aejpol-tables-figures
elif estimate_unstable: -> aejpol-robustness
elif welfare_mapping_missing: -> aejpol-theory-model
elif identification_shaky: -> aejpol-identification
elif claim_or_positioning_fuzzy: -> aejpol-literature-positioning
else: -> aejpol-topic-selection
Source: brycewang-stanford/Awesome-Journal-Skills → AEJ-Economic-Policy-Skills/skills/aejpol-workflow/SKILL.md