/submission-guide

name: submission-guide
argument-hint: "<journal name or submission task>"
description: >-
  This skill covers academic journal submission, referee responses, and revision management. Use when the user is preparing a manuscript for submission, formatting for a specific journal, responding to referees, or managing revisions. Triggers on "submit", "referee", "revision", "R&R", "response letter", "journal", "formatting", "submission", "resubmit", "cover letter", "referee report", "revise and resubmit".

Journal Submission

Reference for the full journal submission lifecycle: pre-submission preparation, journal-specific formatting, referee response strategy, and revision management. Covers conventions for top journals in economics, finance, political science, sociology, marketing, and statistics.

When to Use This Skill

Use when the user is:

• Preparing a manuscript for first submission to an academic journal
• Formatting a paper to meet a specific journal's requirements
• Writing a response letter to referee reports after receiving an R&R
• Managing tracked changes and revision logistics
• Anticipating common referee objections for a particular empirical method
• Writing a cover letter to the editor

Skip when:

• The task is choosing an empirical method (use `causal-inference` or `empirical-playbook` skill)
• The task is structural estimation implementation (use `structural-modeling` skill)
• The task is setting up a replication package (use `reproducible-pipelines` skill)

Pre-Submission Checklist

Complete every item before submitting. Missing any one of these is a common reason for desk rejection or delayed processing.

Manuscript

• [ ] **Title**: Concise, informative, no unnecessary jargon. Under 15 words is ideal.
• [ ] **Abstract**: States the question, method, data, and main finding. Respects journal word limit (typically 100-150 words for econ journals).
• [ ] **JEL codes**: 2-4 codes, primary code first. Check https://www.aeaweb.org/econlit/jelCodes.php for current classification.
• [ ] **Keywords**: 3-6 terms not already in the title.
• [ ] **Introduction**: Clearly states contribution in first two paragraphs. Includes a "roadmap" paragraph at the end.
• [ ] **Literature review**: Positions paper relative to 3-5 closest papers. Explains what this paper does that they do not.
• [ ] **Identification section**: Formal statement of assumptions, not just prose. Numbered assumptions preferred.
• [ ] **Results**: Main results first, robustness second. Do not bury the lead.
• [ ] **Conclusion**: No new results. Discuss limitations honestly. Suggest future work briefly.
• [ ] **References**: Every citation in text appears in references and vice versa. Use a bibliography manager (BibTeX/BibLaTeX).
• [ ] **Anonymization**: Remove all author-identifying information. Check PDF metadata, acknowledgments, file paths in code, dataset names that reveal institution.
• [ ] **Page/word count**: Within journal limits. Many journals have strict limits (e.g., AER Papers & Proceedings: 5 pages).
• [ ] **Spell check and grammar**: Run a final pass. Typos in the abstract signal carelessness.
• [ ] **Agent review**: Run the `journal-referee` agent for an adversarial review and the `econometric-reviewer` agent to audit tables against code output.

Tables

• [ ] **Self-contained**: Each table has a descriptive title and notes explaining all variables, sample, and significance stars.
• [ ] **Significance stars**: Use journal convention. Most econ journals: `* p<0.10, ** p<0.05, *** p<0.01`. Some journals (QJE) discourage stars entirely.
• [ ] **Standard errors**: Report in parentheses below coefficients. State clustering level in notes.
• [ ] **Number of observations**: Report N for every regression. Report N by group for DiD/panel.
• [ ] **R-squared or fit measure**: Report adjusted R-squared, within R-squared for FE models, or pseudo R-squared for nonlinear models.
• [ ] **Decimal places**: 2-3 significant digits. Do not report 8 decimal places from Stata/R output.
• [ ] **Consistent formatting**: Same variable names across all tables. Same order of controls.
• [ ] **No vertical lines**: Use horizontal rules only (booktabs style in LaTeX).

Figures

• [ ] **Vector format**: PDF or EPS for line plots and diagrams. High-resolution PNG (300+ DPI) only for heatmaps or photos.
• [ ] **Readable in grayscale**: Use shapes/patterns in addition to colors. At least 20% of readers print in black and white.
• [ ] **Axis labels**: Clear, with units. Font size readable when figure is scaled to journal column width.
• [ ] **No chartjunk**: Remove gridlines, unnecessary legends, 3D effects, excessive tick marks.
• [ ] **Consistent style**: All figures use the same font, color palette, and line weights.
• [ ] **Source note**: State data source and sample period below each figure.

Appendix and Online Appendix

• [ ] **Appendix**: Proofs, additional tables referenced in the main text, variable definitions.
• [ ] **Online appendix**: Supplementary results that support but are not essential to the main argument.
• [ ] **Cross-references**: Every appendix item is referenced from the main text. No orphan appendix tables.
• [ ] **Separate file**: Some journals require the online appendix as a separate PDF. Check submission guidelines.

Replication Package

• [ ] **Data**: All data files, or clear instructions for obtaining restricted-access data.
• [ ] **Code**: All scripts from raw data to final tables/figures. Master script that runs everything in order.
• [ ] **README**: Describes file structure, software requirements, runtime estimate, expected output.
• [ ] **Seeds**: All random number generator seeds set and documented.
• [ ] **Versions**: Software versions pinned (R/Python/Stata version, package versions).
• [ ] **License**: Data license and code license specified.
• [ ] **Tested**: Run the entire pipeline from scratch on a clean machine or container.

Cover Letter

• [ ] **Editor na