Restud workflow

REStud Workflow (restud-workflow)

Overview

This is the router. It does not replace any specialized skill — it tells you which one to use next, and in what order.

Default assumption: unless the user names a different venue, the manuscript targets The Review of Economic Studies (REStud) — a top-5 general-interest journal (OUP, for the Review of Economic Studies Ltd; founded 1933 to advance theoretical and applied economics equally, especially by younger economists). REStud prizes a clean, original contribution — a new model, a new identification strategy, or a striking new empirical fact — presented rigorously and economically. It is not a finance-only stack, not a theory-only stack, and not a working-paper repository.

REStud-specific anchors (carry these through every stage)

• Identity: founded 1933, theory and applied weighted equally; home of the REStud Tour (former May Meetings, annual since 1989) for promising young economists — calibrate ambition to that bar.
• Submission: Editorial Express portal, USD 250 fee (reduced USD 150 if every author is a student / within 6 years of PhD / in a low-or-middle-income economy; rate effective 1 June 2026), double-anonymous review, Harvard author-date references, JEL codes, ~45-page limit; editors desk-reject ~20–23% before refereeing.
• Replication: the Data Editor (Miklós Koren) runs an AEA DCAS pre-publication reproducibility check; packages deposit at Zenodo.
• Editors (2026): Joint Managing Editors include Jan De Loecker, Antonio Penta, Jakub Steiner (verify the full current roster).

When to Use

• The user asks "what should I work on next?"
• The user dumps a draft and you must decide where the bottleneck is
• The user is rotating between theory, empirics, writing, and revision and loses track of the stage
• A referee report has arrived and the work mode must switch from drafting to rebuttal
• The user is preparing a job-market paper for the REStud "May Meetings" / Tour pipeline and wants the manuscript at REStud standard

Routing Map

Current symptom	Next skill
Idea is fuzzy; the one-sentence contribution will not compress	restud-topic-selection
Related-work positioning is thin; closest papers not confronted	restud-literature-positioning
Empirical design is the bottleneck (DID / IV / RDD / RCT / structural)	restud-identification
The paper is theory or has a model whose proofs need an online appendix	restud-theory-model
Main results exist but referee-anticipating checks are missing or fragile	restud-robustness
Tables are oversized / footnote-bloated; figures do not carry the result	restud-tables-figures
Prose is bloated; abstract over length; contribution buried	restud-writing-style
Preparing the data/code deposit or auditing reproducibility	restud-replication-package
Want to anticipate who referees and what they will demand	restud-referee-strategy
Final preflight before clicking submit (format, length, anonymity)	restud-submission
A referee report / R&R arrived and a response letter is needed	restud-rebuttal

Default Sequence

For a typical REStud-track manuscript, prefer this order:

1. restud-topic-selection — fix the one-sentence original contribution before anything else
2. restud-literature-positioning — locate the contribution against the closest 3–5 papers
3. restud-identification — for empirical papers, stress-test the design; if it fails here, no later skill saves the paper
4. restud-theory-model — for theory papers (or empirical papers with a model), build the model and route proofs to the online appendix
5. restud-robustness — anticipate the checks the demanding median REStud referee will require
6. restud-tables-figures — finalize exhibits so each one carries a result
7. restud-writing-style — only now polish prose, abstract, and introduction
8. restud-replication-package — assemble the deposit while results are fresh in code
9. restud-referee-strategy — map the likely referee pool and pre-empt objections
10. restud-submission — final preflight: length, anonymity, format, files
11. restud-rebuttal — after review, revise the manuscript first, then write the letter

restud-identification and restud-theory-model are parallel branches: empirical papers lean on the first, theory papers on the second, and many REStud papers (theory-with-empirics) need both. restud-writing-style is a late polish — do not run it before the contribution and design are stable.

Decision Cues

If the user says...

• "I have data but no story" → restud-topic-selection
• "My closest paper is in QJE 2022 and I don't know what I add" → restud-literature-positioning
• "My DiD has staggered treatment and TWFE" → restud-identification
• "My model has a proposition but no proof section" → restud-theory-model
• "A referee will say the result is driven by X" → restud-robustness
• "My table 3 has 14 columns" → restud-tables-figures
• "My intro is 5 pages and buries the contribution" → restud-writing-style
• "I got a conditional accept; I need the data deposit" → restud-replication-package
• "Who is going to referee this and what will they want?" → restud-referee-strategy
• "I'm submitting tomorrow" → restud-submission
• "I got an R&R with three reports" → restud-rebuttal

Differences vs. Other Top-5 Stacks

REStud overlaps heavily with AER / QJE / JPE / Econometrica in standards, but its center of gravity differs:

• vs AER: AER weights cross-subfield interest and policy salience; REStud equally welcomes a clean theoretical or methodological contribution with no policy hook, and is famously friendly to young scholars — the REStud Tour / May Meetings (annual since 1989) showcase graduating PhD job-market work.
• vs Econometrica: Econometrica tolerates very heavy formal machinery as the contribution itself; REStud wants the economics payoff of the theory to be legible.
• vs JPE: JPE has a price-theory / Chicago tilt; REStud is field-agnostic and European-rooted.

If the paper is finance-specific (JF / JFE / RFS) or theory so abstract it is really a math paper, a different stack fits better.

Anti-Patterns

• Polishing prose (restud-writing-style) before the identification or model is stable
• Treating restud-theory-model as optional for an empirical paper that has a model — REStud referees check that the model's assumptions map to the empirics
• Assembling the replication package only after acceptance — REStud's Data Editor (Miklós Koren) runs an AEA DCAS reproducibility check before publication (Zenodo deposit), so it is part of the production timeline, not an afterthought
• Writing the rebuttal against the old draft instead of the revised manuscript
• Skipping restud-referee-strategy because "the design is obviously fine" — REStud referees are demanding and develop strong papers across rounds

Handoff Contract

Whenever this skill is invoked, end with:

NEXT SKILL: <restud-skill-name>
REASON: <one sentence>
INPUTS NEEDED: <artifacts the next skill needs>