/aeja-workflow

name: aeja-workflow
description: Use when deciding which aeja-* sub-skill to invoke next, or when sequencing manuscript work from topic selection through rebuttal for an American Economic Journal: Applied Economics (AEJ: Applied) submission. Routes — it does not replace — the specialized skills.

AEJ: Applied Workflow Router (aeja-workflow)

Overview

This is the router. It tells you **which aeja-* skill to use at the current stage** of a manuscript aimed at the *American Economic Journal: Applied Economics* (AEJ: Applied) — the **American Economic Association's** quarterly journal for **empirical applied microeconomics with credible causal identification**: labor, development, health, education, public, urban, environmental, and household finance. AEJ: Applied is **empirical-first**: a clean research design carrying a substantive economic question, with theory used to *interpret and structure* rather than to lead.

Default assumption: unless the user says otherwise, treat the target as AEJ: Applied. Operational tells that you are at AEJ: Applied and not a sibling: review is **single-blind** (author identities are visible to referees); submission is via the **AEA ScholarOne system** with fee tiers by AEA membership and country-income status; **JEL codes and a 100-word abstract** are required; and the signature stable differentiator is the **AEA Data and Code Availability Policy** — administered by the **AEA Data Editor (Lars Vilhuber)**, accepted papers deposit data + code to the **AEA Data and Code Repository on openICPSR**, and the package is checked for reproducibility before publication. On 2026-06-20 the editor page listed **Raymond Fisman** (Boston University) as Editor, with coeditors Leah Boustan, Marika Cabral, Peter Hull, Xavier Jaravel, and Nicholas Ryan. Check live AEA pages for fees and editors before submission-week advice.

When to trigger

• The user asks "what should I do next?"
• A draft needs its current bottleneck diagnosed
• Work is ping-ponging between identification, framing, writing, and the response letter
• An AEJ: Applied decision letter arrived and the user needs to switch into revision mode

Routing table

| Current symptom | Next skill | |-----------------|------------| | Unsure the paper fits AEJ: Applied vs AER / AEJ: Policy / a field journal | `aeja-topic-selection` | | Contribution vs. the literature is fuzzy or undersold | `aeja-literature-positioning` | | Causal design rests on OLS+controls, TWFE on staggered timing, or a weak IV | `aeja-identification` | | A model is needed to interpret estimates or structure a mechanism | `aeja-theory-model` | | Results not shown robust to specification, sample, inference | `aeja-robustness` | | Exhibits are dense; main result not legible in one table/figure | `aeja-tables-figures` | | Prose buries the design; abstract/intro do not land | `aeja-writing-style` | | Data/code deposit, README, openICPSR prep for the AEA Data Editor | `aeja-replication-package` | | Want to anticipate referee objections before submitting | `aeja-referee-strategy` | | Ready to submit via the AEA portal; need a preflight | `aeja-submission` | | Received an R&R; need a response-letter strategy | `aeja-rebuttal` |

Default order

1. `aeja-topic-selection` — lock the applied-micro question and venue fit 2. `aeja-literature-positioning` — stake the marginal contribution 3. `aeja-identification` — make the data-to-causal-estimate mapping credible 4. `aeja-theory-model` — add only the theory that interprets/structures the estimate 5. `aeja-robustness` — show the headline survives specification, sample, inference 6. `aeja-tables-figures` — make the main result legible in one exhibit 7. `aeja-writing-style` — make the design and estimate land (abstract + intro last) 8. `aeja-replication-package` — assemble the openICPSR package for the AEA Data Editor 9. `aeja-referee-strategy` — pre-empt the objections this design invites 10. `aeja-submission` — AEA portal preflight (single-blind front matter, JEL, fee) 11. `aeja-rebuttal` — after the R&R

> `aeja-writing-style` is a late-stage polish; do not rewrite the intro before identification and robustness settle.

Anti-patterns

• Treating the replication package as an acceptance-time chore — at AEA the Data Editor verifies reproducibility **before** publication, so build it as you go (`aeja-replication-package`)
• Polishing exhibits (`aeja-tables-figures`) while the identification or headline estimate is still moving
• Framing the paper as a policy question (that drifts toward AEJ: Policy) or as broad general-interest (that belongs at AER) instead of method/identification-driven applied micro

Routing by paper archetype

AEJ: Applied spans several applied-micro designs, and the first bottleneck differs by design. Read the archetype, then enter the chain at the right link.

| Archetype | Likely first bottleneck | Enter at | |-----------|-------------------------|----------| | field experiment / RCT (own data) | pre-registration, balance, attrition, estimand | `aeja-identification` → `aeja-robustness` | | quasi-experiment (DID / event study) | staggered-timing bias, parallel-trends credibility | `aeja-identification` | | regression discontinuity | density manipulation, bandwidth, bias-corrected CIs | `aeja-identification` → `aeja-robustness` | | IV / shift-share | first-stage strength, exclusion, exposure design | `aeja-identification` | | descriptive / measurement with a clean question | framing the contribution without a causal overclaim | `aeja-topic-selection` → `aeja-literature-positioning` |

Worked routing example (illustrative)

A user says: "My event study on a staggered state policy looks fine, but a referee says the two-way fixed-effects estimate is contaminated and the parallel-trends story is thin." That is two AEJ: Applied pushbacks — *negative-weighting / contaminating already-treated comparisons under staggered timing* and *pre-trend credibility* — both owned by