brycewang-stanford Rebuttal & Response Letter (jf-rebuttal)
When to trigger
- A JF revise-and-resubmit (R&R) has arrived and you must respond
- You are deciding whether/how to answer a reject-with-encouragement
- The revision is drafted and you need to map every editor/referee point to a change
A JF R&R is rare — treat it accordingly
With ~5% acceptance and ~33–45% desk rejection (afajof.org editor reports, accessed 2026-05-30), an R&R from JF is precious. The path to acceptance runs through the handling editor (currently the team led by Antoinette Schoar (MIT) — verify the masthead): the editor's letter, not any single referee, defines what acceptance requires. Address the editor's framing first and most fully.
Structure
- To the editor: summarize the main changes and how you met the editor's synthesis of the decision.
- Per-referee, per-point: restate each point, then state exactly what changed and where (section/table/Internet Appendix), quoting revised text where useful.
- Disagreements: handle respectfully with evidence; never ignore a point — explain if you chose not to change something.
JF-specific moves
- Route substantial new robustness to the Internet Appendix (bundled at the end of the same PDF; does not count toward 60 pages; see
jf-internet-appendix) and say so explicitly — this matches JF norms and keeps the body lean. - Keep the body within the 60-page limit after revisions.
- Update disclosure and ensure replication code is ready for the Data Editor under JF's Data and Code Sharing Policy (see
jf-submission); near acceptance the code is verified before publication. - Use the cover-letter channel only for specifics (e.g., a code-exemption request) — JF does not want a generic letter.
Triaging the decision letter
A JF R&R bundles the editor's letter with two or three referee reports of unequal weight. Sort every point before drafting:
| Point type | Weight | Response discipline |
|---|---|---|
| Editor's synthesis / central concern | Highest — defines acceptance | Answer first, most fully, in the body |
| Referee point the editor echoed | High | Treat as a near-condition; change, don't argue |
| Referee point on identification/robustness | High | New tests, mostly routed to the Internet Appendix |
| Referee suggestion the editor did not endorse | Medium | Address respectfully; disagree with evidence if warranted |
| Minor / stylistic | Low | Fix quietly; note in the per-point table |
The recurring JF failure mode: optimizing for the most vocal referee while underweighting the editor — who, not any referee, decides.
Worked vignette — answering an R&R on a DID paper
Illustrative. The editor writes: "I am convinced the effect is real, but the staggered-DID estimator is the open question; please satisfy Referee 2." The disciplined response opens to the editor: "Following your central concern, we re-estimate with Callaway–Sant'Anna; the leverage effect moves from 4.2 to 3.1 pp (illustrative) and holds; Internet Appendix Table IA.VII holds the full comparison." Then a per-referee table logs every point with its change and location. New tables go to the IA so the body stays under 60 pages; code is staged for the Data Editor.
Referee-pushback patterns and the JF-specific fix
| Residual pushback on resubmission | JF-specific fix |
|---|---|
| "You answered R1 but not my main point" (editor) | Open the letter with the editor's concern, resolved in the body |
| "The new robustness bloated the paper" | Route it to the Internet Appendix; cite, don't inline |
| "Where exactly did this change?" | Give section/table/IA location and quote the revised text |
Calibration anchors
- A JF R&R is scarce given the flagship's selectivity; treat it as the project's most valuable asset and over-invest in the editor's framing. How many rounds and what "conditional acceptance" requires vary; confirm from the editor's letter.
Checklist
- Editor's points ranked and addressed first
- Every referee point tabulated with response + location of change
- New robustness placed in the Internet Appendix and cross-referenced
- Body still ≤60 pages after revision
- Disagreements handled with evidence, none ignored
- Replication code / disclosure updated for the Data Editor
Anti-patterns
- Optimizing for one referee while missing the editor's central concern
- Burying or ignoring a referee point
- Adding robustness without telling referees it lives in the Internet Appendix
- Letting the revision push the body past 60 pages
- Forgetting the code-sharing obligation surfaces at acceptance
Output format
【Editor's central concern + how met】...
【All referee points tabulated?】yes / no
【New robustness → Internet Appendix + cited?】yes / no
【Body ≤60 pp?】yes / no
【Code/disclosure ready for Data Editor?】yes / no
【Next step】resubmit via the AFA portal