Tables & Figures (aejmac-tables-figures)

When to trigger

• IRF figures lack confidence bands, units, or a clear horizon axis
• Tables are dense, mislabeled, or report coefficients without the macro quantity of interest
• A model-fit comparison (data vs. simulated moments) is asserted in text but not shown
• Exhibits do not stand alone — the reader needs the body text to decode them

The AEJ: Macro exhibits bar

AEJ: Macro follows AEA house style (applied by the AEA at copyediting; the submitted PDF need not pre-conform, but clean exhibits speed review). Macro adds its own conventions: the impulse-response function is the workhorse figure, and the headline quantity (multiplier, peak response, welfare cost, share of variance) must be readable off the exhibit. Every exhibit should be self-contained: title, axis labels with units and horizon, sample, and the inference object (bands / SE) in the note.

Figure conventions (macro-specific)

• IRFs: plot the response with a shaded confidence band (state the level, e.g., 68% and/or 90%); label the horizon axis in the natural unit (quarters/months/years); mark zero; if sign-restricted, show the identified set / median-target, not a spurious point.
• Fan charts / forecast bands: show the full predictive distribution where the claim is about uncertainty.
• Model-fit overlays: plot data vs. model (targeted and untargeted moments) on the same axes; this is the credibility figure for quantitative work.
• Counterfactual figures: baseline vs. counterfactual paths with uncertainty carried through.
• Variance decompositions / contribution plots: stacked or grouped, with a clear legend.
• Vector output (.eps/.pdf) preferred for final files; keep colors legible in grayscale; no chartjunk.

Table conventions

• Report the economic quantity, not just raw coefficients — e.g., the multiplier, the elasticity, the implied share — with its SE/band.
• AEA tables conventionally use significance stars; AEJ: Macro accepts them, but the standard error / band must be present and must carry the inference (do not let stars substitute for a reported SE). Note significance levels in the table note.
• One decisive number per row; align decimals; units in the column header or note.
• Targeted-vs-untargeted-moment tables: clearly flag which moments were matched.
• Notes carry: sample, frequency, estimator, inference (HAC/cluster), and what an asterisk means.

Main text vs. online appendix

• Main text: the decisive IRFs, the headline table, the model-fit figure. AEJ: Macro articles run ~40 pages including exhibits (检索于 2026-06；以官网为准), so exhibits compete for space.
• Online appendix: full robustness IRFs, additional moments, alternative specifications, derivations.

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.

See a full fitted-result → exhibit chain in the JF execution walkthrough.

Checklist

• Every IRF has a confidence band with its level stated, a labeled horizon axis, and a zero line
• Sign-restricted responses shown as a set/median-target, not a fake point estimate
• Model-fit figure (data vs. model, targeted + untargeted) present for quantitative papers
• Tables report the macro quantity with its SE/band; notes give sample/estimator/inference
• Significance stars (if used) accompany, not replace, reported SEs
• Each exhibit is self-contained (decodable without the body text)
• Decisive exhibits in main text; the rest in the online appendix; figures vector + grayscale-legible

Anti-patterns

• IRFs with no confidence bands, or bands whose level is never stated
• Sign-restricted IRFs drawn as a single line, hiding set-identification
• Tables of raw VAR coefficients with no impulse responses or implied quantity
• Stars reported but standard errors omitted
• A model "matches the data" claimed in prose with no overlay figure
• Exhibits readable only with a magnifying glass after the AEA's two-column typesetting

Output format

【Exhibit inventory】IRFs / fan charts / model-fit / counterfactual / tables
【IRF compliance】bands+level / horizon axis / zero line / set-vs-point? [Y/N]
【Quantity legibility】headline number readable off each exhibit? [Y/N]
【Table compliance】SE/band present; notes complete; stars don't replace SEs? [Y/N]
【Main vs. appendix split】decisive in text, rest in online appendix? [Y/N]
【Next step】aejmac-writing-style