Tables & Figures — Five-Exhibit Budget (aeri-tables-figures)

When to trigger

• The draft has more than five figures-plus-tables
• An exhibit spills past one page, or carries significance asterisks/boldface
• Each exhibit costs 200 words of your text budget and you must choose
• The central result is buried in a dense multi-panel table

The AER: Insights exhibit constraints (检索于 2026-06；以官网为准)

• Maximum five exhibits (figures or tables combined). Manuscripts over the limit are returned unreviewed.
• Each exhibit is limited to one page.
• Each exhibit subtracts 200 words from the main-text allowance (7,000 with zero exhibits → 6,000 with five). So every exhibit must out-earn 200 words of prose.
• AEA house style: report standard errors / confidence sets; no asterisks or boldface to denote significance (an AEA-wide convention). Notes state the estimator, sample, clustering, and units.

Budgeting the five slots

Choose the exhibit set that, together, tells the single-insight story. A common allocation for an empirical short paper:

| Slot | Typical content | |------|-----------------| | 1 | The headline result — the one figure or table a reader remembers (event-study, RD plot, treatment-effect estimate) | | 2 | The central identification diagnostic (pre-trend leads, density, first-stage, balance) | | 3 | A mechanism or key heterogeneity exhibit, if it is part of the single insight | | 4–5 | Held in reserve — a robustness summary or a descriptive table only if load-bearing |

If two exhibits can merge into one well-designed multi-panel page, merge them and reclaim a slot (and 200 words). Everything else goes to the Supplemental Appendix.

Design rules

• The headline exhibit must be self-reading — a reader who sees only it should grasp the insight.
• Figures over tables where a figure communicates the magnitude faster (AER: Insights papers lean visual).
• One page each, legible at print size; no microscopic multi-panel grids spilling over.
• Notes carry the inference since there are no asterisks: report point estimates with SEs / 95% confidence sets in the cell or note.
• Units and signs explicit; the magnitude, not just the sign, is the point.

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

• ] ≤5 exhibits total; each ≤1 page
• ] Word budget reconciled: 7,000 − 200×(#exhibits) ≥ main-text length
• ] Exhibit 1 is the self-reading headline; exhibit 2 the identification diagnostic
• ] No significance asterisks/boldface; SEs / confidence sets shown
• ] Notes state estimator, sample, clustering, units
• ] Non-essential exhibits moved to the Supplemental Appendix
• ] Mergeable exhibits combined to reclaim a slot where sensible

Anti-patterns

• A sixth exhibit (auto-return) or an exhibit that runs to two pages
• Significance asterisks or boldface for significance
• A dense table where a single figure would land the magnitude
• Spending an exhibit slot (and 200 words) on a descriptive table that prose covers
• A headline exhibit that needs the text to be intelligible

Output format

【Exhibit count】__ / 5 (each ≤1 page?)
【Word reconciliation】7000 − 200×__ = __ words available; draft = __ words
【Exhibit 1 (headline)】<self-reading result>
【Exhibit 2 (identification)】<diagnostic>
【Exhibits 3–5】<mechanism / robustness / descriptive — only if load-bearing>
【No-asterisk check】SEs / CIs in notes? [Y/N]
【To the appendix】<exhibits demoted>
【Next step】aeri-writing-style