agentsclimarketplace

Qe rebuttal

Skill brycewang-stanford/Awesome-Journal-Skills/Quantitative-Economics-Skills/skills/qe-rebuttal

Use when a Quantitative Economics (QE) decision letter arrives (R&R or major/conditional revision) and a response strategy and letter are needed. Structures the revision and response; it does not redo the analysis (route to the relevant qe-* skill for that).From its SKILL.md

Install
npx -y skills add brycewang-stanford/Awesome-Journal-Skills --skill qe-rebuttal

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

SKILL.md

5.7 KB, ~1.2k tokens by cl100k_base, as published. Nobody here has run it

Rebuttal & Revision (qe-rebuttal)

When to trigger

  • A QE decision letter arrived: revise-and-resubmit, or a conditional/major revision
  • You have referee reports and need a response strategy
  • You are unsure how to prioritize conflicting referee demands
  • The revised manuscript exists and you need to draft the response letter

QE reality: the coeditor's letter is the contract, and the clock runs ~12 months

At QE a handling coeditor owns the decision and synthesizes the referees (often including an Associate Editor). The coeditor's letter ranks the issues — it tells you which referee points are binding and which are optional; read it first and let it set priorities. Revise offers are normally valid for 12 months, so plan the analysis, estimation, and reproducibility work against that window rather than rushing a thin resubmission. Because QE is the empirical/quantitative ES journal, expect referees to probe identification, estimation, robustness, and reproducibility specifically — keep the public-posting and replication obligations in view as you revise.

Triage protocol

  1. Decode the coeditor. A point the coeditor underscores is binding; a point only one referee raised and the coeditor ignored is lower priority (still answer it).
  2. Classify every comment. Tag each: fundamental (must add/redo analysis or estimation), clarification (rewrite/explain), or disagreement (you will push back, politely, with evidence).
  3. Sequence the work. Do the analysis first — route fundamentals to qe-identification-strategy, qe-data-analysis, or qe-tables-figures; update the replication package via qe-replication-and-data-policy; then write the letter.
  4. Settle it with a check, not a fight. If a robustness check, an added moment, or a Monte Carlo settles the point, run it; reserve disagreement for where the referee is genuinely mistaken.

Response-letter structure

  • Opening: thank the coeditor and referees; summarize the main changes in a short paragraph.
  • Point-by-point: quote each comment, then respond — what you did + where in the revised paper it now appears (section/table/figure/page).
  • Disagreements: state the referee's point fairly, then give the evidence-based reason you respectfully differ; offer a compromise (e.g., add to the Supplemental Appendix, ≤25 pages) where possible.
  • Reproducibility: note any changes to data/code so the eventual ES Data Editor check stays clean.
  • Tone: gracious, concrete, never defensive.

Checklist

  • Coeditor's priorities decoded and used to rank responses
  • Every referee comment classified (fundamental / clarification / disagreement)
  • All fundamental requests resolved with actual new analysis/estimation, not promises
  • Each response cites the exact location of the change in the revised paper
  • Disagreements are evidence-based, fair, and offer a compromise
  • Replication package and public posting updated to match the revision
  • Revised manuscript updated before the letter was written
  • No significance asterisks reintroduced; no new over-claiming

Anti-patterns

  • Writing the response letter before the manuscript reflects the changes
  • Promising future work instead of doing the analysis within the ~12-month window
  • Arguing where an added moment / robustness check / Monte Carlo would settle it
  • Letting the replication package drift out of sync with the revised results
  • Defensive or curt tone; treating reports as attacks rather than help

The three QE pushbacks and what actually closes them

These are the recurring quantitative complaints at QE; each is closed by a specific, runnable response — not by prose.

Referee complaintThe closing move
"Structural estimates are not credibly identified"add a sensitivity matrix + Monte Carlo recovery; report untargeted-moment fit
"Numerical accuracy is not validated"report tolerances and a grid-refinement check; show headline stability
"Quantitative results not robust to specification"re-estimate under alternative moments/sub-samples; tabulate the movement

Worked vignette: triaging a QE major-revision letter (illustrative)

Suppose the coeditor underscores Referee 1's identification concern, treats Referee 2's request for a robustness table as binding, and notes Referee 3's framing quibble as optional. Triage: tag identification fundamental (run the sensitivity matrix, route to qe-identification-strategy), tag the robustness table fundamental (re-estimate across three moment sets), tag the framing clarification (a paragraph rewrite). Suppose the headline elasticity moves from 0.42 to 0.39 (s.e. 0.05) across specifications — report that explicitly; a stable number under scrutiny is the strongest rebuttal. (Numbers illustrative.) Only after the manuscript reflects all three do you write the point-by-point letter with exact section/table locations.

Output format

【Decision】R&R / major / conditional
【Coeditor's binding points】[...]
【Comment classification】fundamental: [...] / clarification: [...] / disagreement: [...]
【Analysis routed to】qe-identification-strategy / qe-data-analysis / qe-tables-figures
【Letter status】point-by-point with locations? [Y/N]
【Reproducibility synced】package + posting updated? [Y/N]
【Next step】resubmit via the ES portal (qe-submission for file checks)

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.