thedixitjain Rebuttal and Response Letter (jhe-rebuttal)
When to trigger
- A JHE R&R or reject-and-resubmit decision letter has arrived
- You need to triage referee comments into must-do / argue / defer
- You are drafting the response-to-referees document and the revision plan
- Two referees disagree and you need a strategy that satisfies the editor
How to read a JHE decision
The editor's letter is the priority signal: it tells you which referee concerns are binding. At JHE the binding asks cluster around design credibility (selection, concurrent policy, staggered-DiD bias, inference) and honest scope (a coverage/utilization effect is not welfare) before exposition. Because review is single-anonymized and the referees are health economists, expect institution-specific pushback you must answer with evidence, not assertion. With ≥2 reviewers you will sometimes face conflicting requests; the editor's letter, not a majority vote, resolves them — address the editor's stated priorities first and in full.
Response-letter structure
- Opening summary — thank the editor and referees; state the 3–5 substantive changes that define the revision, leading with the identification/selection/scope changes.
- Point-by-point — quote each comment, then respond with: what you did, where in the paper (section/table/page), and the result. Keep one consistent format per point.
- Tone — gracious, concrete, non-defensive. Concede valid points fully; where you disagree, argue with evidence and institutional fact.
- Triage discipline — every comment is done (changed + located), partially done (with reason), or respectfully declined (with an evidence-based argument). Nothing is silently dropped.
- Quote the decisive number. If a referee doubted the design, show the heterogeneity-robust estimate / honest-DID bound / selection bound / placebo explicitly, with its SE and location.
- Reconcile conflicting referees by naming the trade-off and the choice the editor's priorities imply.
Triage table
| Referee ask | Default response |
|---|---|
| Selection / identification doubt (binding per editor) | do it in full; show the new estimate or bound + SE; update the paper |
| Concurrent-policy / institutional confound | add the placebo / timing argument in the institutional section |
| Staggered-DiD or inference fix | re-estimate with CS/SA; wild-cluster bootstrap; report stability |
| "Coverage effect ≠ welfare" | add the demand-and-cost model; rescope the claim honestly |
| Additional health outcome/subgroup | run with MHT; or argue out of scope with a reason |
| Exposition / institutional clarity | revise; cheap and builds goodwill |
| Out-of-scope new paper | respectfully decline with a one-line reason |
Checklist
- Editor's stated priorities identified and addressed first, in full
- Every referee comment answered: done / partial / declined — none ignored
- Each "done" response cites the exact section/table/page and the new result
- Decisive identification/selection/robustness evidence quoted with numbers and SEs
- Conflicting referee requests reconciled by reference to the editor's priorities
- Scope claims rematched to the design where a referee flagged overreach
- Replication package and DAS updated to match the revised exhibits (
jhe-replication-package) - Tone gracious and concrete; disagreements argued with evidence and institutions
Coordinating the rebuttal with the rest of the pack
A response letter is only as strong as the new evidence behind it. When a binding concern requires fresh analysis, route the work to the owning skill before drafting the point: jhe-identification for a design fix, jhe-robustness for a stability check, jhe-theory-model for a welfare rescope. Then update exhibits (jhe-tables-figures) and the package and Data Availability Statement (jhe-replication-package) so the revised paper, the letter, and the deposit all tell the same story. A letter that quotes a number the revised paper does not contain is the fastest way to lose editor trust.
Anti-patterns
- Answering the referees but ignoring the editor's prioritization
- Defensive or dismissive tone; arguing by assertion instead of new evidence
- Claiming a change without saying where in the paper it landed
- A wall of new tables that dodges the actual selection/identification concern
- Splitting the difference between conflicting referees instead of following the editor
- Letting the replication package or DAS drift out of sync with the revised paper
Timing and the resubmission cover note
Elsevier R&Rs do not run on a fixed clock the way some society journals do; do the work properly rather than rushing, but do not let a revision go stale — a multi-year gap invites a fresh referee pool and new objections. Open the resubmission with a short cover note to the editor (separate from the point-by-point) that lists, in three to five bullets, how each binding concern was resolved and where. Editors triage resubmissions from that note; make it carry the decisive numbers and locations so the editor can verify the binding asks were met without re-reading the whole letter.
What "resolved" means at JHE
A JHE revision is judged on whether the design and scope concerns are genuinely settled, not on the count of new tables. "Resolved" means: the selection/identification doubt is answered with an estimate or bound (not more controls); the institutional confound is ruled out with a placebo or timing argument; and any welfare/scope overreach the referee flagged is rematched to what the design identifies. A revision that adds analysis but leaves the binding concern technically un-addressed will not clear, however thick it is.
Worked vignette (illustrative)
Referee 1 calls the staggered Medicaid-expansion DiD biased and the parallel-trends story thin; Referee 2 wants more outcomes; the editor flags Referee 1 as binding. A weak response argues TWFE "should be fine." The JHE response: re-estimate with Callaway–Sant'Anna (new ATT 4.0pp, s.e. 1.1, vs. 4.6pp under TWFE), add event-study leads and a Goodman-Bacon decomposition, run an honest-DID bound, and rule out the concurrent marketplace launch with a placebo — then in the letter quote the new number, cite the exact figure/table, and note the conclusion is unchanged but now robust. Referee 2's extra outcomes are added with MHT where central and declined-with-reason where out of scope, explicitly subordinated to the editor's binding concern.
Output format
【Decision type】R&R / reject-and-resubmit / conditional accept
【Editor priorities】[binding asks, in order]
【Triage】done: [...] | partial: [...] | declined (with reason): [...]
【Decisive evidence】new identification/selection result + number(SE) + location
【Conflicting referees】reconciled via editor priorities? [Y/N]
【Scope discipline】overclaim rescoped to design? [Y/N]
【Package sync】replication + DAS updated to match revision? [Y/N]
【Next step】jhe-submission (resubmit) → re-enter jhe-workflow if further rounds
Handoff boundary
This skill plans and structures the response and the revision; it does not produce the new evidence (that is jhe-identification / jhe-robustness / jhe-theory-model) nor run the resubmission preflight (jhe-submission). When the letter is drafted and every binding concern has its evidence and location, hand off to jhe-submission to resubmit through Editorial Manager.
Source: brycewang-stanford/Awesome-Journal-Skills → Journal-of-Health-Economics-Skills/skills/jhe-rebuttal/SKILL.md