Aejpol Tables Figures

Tables & Figures — Exhibits that Carry the Policy Message (aejpol-tables-figures)

When to trigger

• Tables are dense, over-decorated with significance stars, or hard to read
• The paper lacks one exhibit a reader could take away as the policy result
• Figures show coefficients but not the policy-relevant magnitude or its uncertainty
• You are preparing the final exhibit set for an AEA submission

AEA / AEJ: Policy exhibit norms

• Report standard errors (or confidence intervals), not significance asterisks/boldface. Put SEs in parentheses below estimates; the reader judges significance from the SE/CI. This is the house convention to follow.
• Self-contained. Title and notes let an exhibit be read without the text: sample, units, estimator, clustering level, what is controlled, and what the number means in policy terms.
• Self-contained, not anonymized. Review is single-blind, so exhibits need not hide authorship; keep notes clean and neutral (avoid stray local file paths) for readability, not for blinding.
• Figures are the workhorse for policy communication: event-study plots with CIs, RDD plots with binned means and the fitted discontinuity, dose-response or cost-benefit curves with uncertainty bands. Vector output; ≥300 dpi raster only if unavoidable; readable greyscale.

The headline exhibit (AEJ: Policy-specific)

Every AEJ: Policy paper should have one exhibit a policymaker could screenshot: the policy effect in interpretable units with its welfare/cost-benefit reading where possible. Examples (illustrative formats):

• An event-study figure of the outcome around the reform, with the long-run effect annotated in policy units.
• A cost-benefit / MVPF figure: net cost per unit of outcome across policy variants, with bands.
• An incidence figure: who gains and who pays, by income/region group.

Table craft

• Three-line tables (esttab/booktabs), no vertical rules; align decimals; consistent digits.
• Lead column = the policy-relevant specification, not a kitchen-sink spec.
• Put the policy-relevant magnitude (elasticity, cost-per-X, MVPF) in the paper's units, not only a raw coefficient; add a row translating the coefficient into the policy number where natural.
• Sample size, mean of the dependent variable, and clustering level on every table.

Execution bridge (StatsPAI / Stata MCP)

Generate exhibits from the fitted result, not by retyping numbers (the usual source of body-vs-appendix drift). Full map: execution-with-mcp.

• Tables: etable (multi-model columns) or did_summary_to_latex straight from the result_id — one variable definition, one set of numbers, body and appendix in sync.
• Figures: plot_from_result / enhanced_event_study_plot / event_study_table — axis units and the SE/clustering note baked in.
• Every note names the estimator + clustering (from the result's diagnostics) and states the magnitude in interpretable units.