Imfer Tables Figures

Tables and Figures (imfer-tables-figures)

When to trigger

• A cross-country regression table is dense and the reader cannot find the headline coefficient
• Country/sample coverage is buried — a referee cannot tell which economies and years are in
• Impulse responses, event-study leads, or spillover paths need a figure that a policymaker can read at a glance
• Significance asterisks are still on the tables, or SEs/CIs are missing
• A map or panel-by-country exhibit would communicate the international margin better than a table
• A summary-statistics table omits the cross-country dispersion a referee needs to judge the sample

What IMFER exhibits must do

IMFER exhibits serve two readers at once: a referee checking the econometrics and a policy reader who wants the magnitude and its country scope. So the country coverage must be transparent (which economies, which years, balanced or not), the headline number must be findable, and the uncertainty must be visible as standard errors and confidence intervals — not asterisks. The best IMFER figure is one that could appear in a WEO/GFSR chart pack and survive referee scrutiny: clear axis units, sample noted, mechanism legible.

Exhibit craft

0. The summary-statistics table earns its keep. For a cross-country paper this is not boilerplate: report the dispersion across countries (not just the pooled mean), the coverage by income group, and the key conventions (currency, gross/net) so a referee can judge whether the sample supports the claim before reading a single regression.
1. One headline exhibit. Decide the single table or figure that is the paper; make it self-contained — a reader who sees only it should grasp the result, the sample, and the uncertainty.
2. Make country coverage explicit. State the economies and years in the note; if the panel is unbalanced, say so; a coverage table or map pre-empts the "which countries?" referee question.
3. No asterisks. Report standard errors (or CIs); for the headline, prefer a confidence interval so the policy reader sees the range, not a star.
4. Put magnitudes in policy units. Translate coefficients into % of GDP, basis points, or a number a mission economist would quote.
5. Show dynamics with bands. Spillover paths and IRFs need confidence bands and labeled horizons; event studies need visible pre-trend leads.
6. Self-contained notes. Estimator, sample, clustering, and units in the note — the exhibit should not require the text to be intelligible.

Country panels: the recurring exhibit problems

Cross-country exhibits fail in patterned ways. Heterogeneous scale: a few large economies dominate any pooled scatter — log, normalize by GDP, or facet by income group. Unbalanced coverage masked: a table reporting "N = 1,200" hides that half the cells are one decade; show coverage by country-year. Currency and vintage drift: a figure mixing USD and local-currency series, or two data vintages, misleads; state the convention in the note. Map theatrics: a choropleth that colors countries but encodes no number the argument uses is decoration — prefer it only when the geographic pattern is the finding (contagion, regional clustering). The fix is always the same: make the cross-country comparison the reader needs explicit and quantitative.

Execution bridge (StatsPAI / Stata MCP)

Generate exhibits from the fitted result, not by retyping numbers. Full map: execution-with-mcp. IMFER is international macro/finance; cross-country panels with confounded policy — emphasize identification and clustering.

• Tables: etable (multi-model) or did_summary_to_latex straight from the result_id.
• 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 and states the magnitude in interpretable units.

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