Tables & Figures (msom-tables-figures)

When to trigger

• Numerical-study or regression tables are cluttered or not self-explanatory
• Policy structure / comparative statics need a figure a reviewer can read at a glance
• You are fitting exhibits into the official M&SOM template and the page budget is tight
• Notation in exhibits does not match the text

Page budget reality: exhibits count against the cap

M&SOM enforces a hard 32-page maximum for a new submission that includes everything — references, tables, graphs/figures, and appendices — measured when typeset on the official LaTeX/Word style files (one column, 11-pt font, 1-inch margins). Unlike narrative word-limit journals, this cap is mechanically enforced on the template, so every table and figure trades directly against text. Build in the template from the start; move bulky proofs, extra numerical experiments, and supporting tables to the online supplement (≤ 16 pages).

Analytical exhibits

• Policy / structure figures: plot the optimal policy form (thresholds, base-stock levels, switching curves) so the structural result is visible, not just stated.
• Sensitivity tables/plots: show how the optimal decision and its value move with primitives (demand variability, lead time, cost ratios); report the gap to benchmarks/heuristics.
• Numerical-study tables: state parameter ranges, instance counts, and what each column means; make magnitudes (% improvement, cost of an assumption) the headline.

Empirical exhibits

• Summary statistics at the operational unit of analysis.
• Main results tables with clustered SEs noted, the operational magnitude interpretable, and specifications laid out left-to-right from baseline to robustness.
• Identification figures: event-study/parallel-trends plots, first-stage diagnostics, or RDD density plots.

House style

Use INFORMS author-year citations in captions and notes; keep every exhibit self-contained (a reader should not need the text to parse it); match notation, units, and labels exactly to the manuscript. Number tables and figures as the portal/template requires.

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. M&SOM mixes analytical and empirical OM; the chain below serves the empirical-OM lane, while analytical / queueing / optimization work is outside it.

• Tables: etable (multi-model columns) 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 effect size in interpretable units.

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

Checklist

• Exhibits drafted inside the official LaTeX/Word template; page budget tracked
• Bulky proofs/extra experiments moved to the ≤16-page online supplement
• Analytical: policy-structure and sensitivity exhibits make the insight visible
• Empirical: results tables show clustered SEs and operational magnitudes
• Each exhibit self-contained; notation/units match the text
• INFORMS author-year style in notes/captions

Anti-patterns

• A wall of numerical results with no benchmark column or headline magnitude.
• Stating a threshold/base-stock result in prose but never plotting it.
• Letting exhibits push the manuscript over 32 typeset pages.
• Significance stars with no operational interpretation.

What each exhibit must earn (referee read)

At M&SOM an exhibit that does not advance the operations insight wastes page budget against the 32-page typeset cap. A policy-structure figure earns its page only when the threshold/base-stock/switching curve is visible, not merely stated in prose; a sensitivity exhibit must show which primitive moves the optimal decision and by how much; a numerical-study table needs a benchmark column that makes the % gap the headline; an empirical results table needs an interpretable operational magnitude with clustered SEs, not just stars; and an identification figure should make the design credible rather than duplicate a number already in a table. The default repair for over-budget exhibits is the online supplement — full proofs, extra experiments, and notation tables live there. (Confirm the current page and supplement limits against the journal's author guidelines.)

Worked micro-example (illustrative)

Vignette: a base-stock model for a spare-parts network deciding stocking levels across echelons. Two exhibits carry the paper. Figure 1 plots the optimal base-stock level against lead-time variability, making visible that the level rises convexly — the structural insight a reader grasps at a glance. Table 2 pairs the proposed policy against a fixed safety-stock benchmark, headlining a 17% average holding-cost reduction at equal fill rate (illustrative). Figure plus benchmark-anchored table do more for the contribution than pages of extra parameter sweeps, which belong in the supplement.

Referee-pushback patterns and the venue fix

• "You state a threshold/base-stock result but never plot it." → Add the policy-structure figure; M&SOM reviewers expect to see the structure, not only read it.
• "The numerical table has no benchmark, so I can't judge the gain." → Add the benchmark/heuristic column and make the % improvement the headline number.
• "Exhibits push the paper over the page cap." → Move supporting tables and proofs to the supplement; do not shrink fonts or margins against the fixed template.

Output format

【Template / page budget】on official style file; pages used ...
【Analytical exhibits】policy structure / sensitivity ...
【Empirical exhibits】results / identification ...
【Self-containment & style】INFORMS author-year; notation matched ...
【Supplement】moved to ≤16-page online supplement ...
【Next step】msom-writing-style