Skip to content

latex

Category: review
Field: economics
License: private (curator-owned)
Updated: 2026-05-20
Stages: referee-simulation

Curator-private skill — copy text from 100xOS/shared/skills/review/latex.md.

LaTeX Review Checklist for Economics Papers

Purpose

LaTeX is the standard typesetting system for economics papers. A well-formatted manuscript signals professionalism and reduces friction for referees. Formatting errors -- inconsistent notation, broken cross-references, misaligned tables -- distract from the content and can give a negative impression before the referee reads a single sentence of substance.

This checklist covers the most common LaTeX issues in economics manuscripts.

1. Mathematical Notation

Inconsistent notation is one of the most frequent problems in economics manuscripts. Referees notice.

Check these items:

  • Every variable is defined the first time it appears in an equation, and the definition is consistent throughout the paper.
  • Scalars, vectors, and matrices have visually distinct formatting. Common convention: lowercase italic for scalars ($x$), bold lowercase for vectors ($\mathbf{x}$), bold uppercase for matrices ($\mathbf{X}$).
  • Subscripts and superscripts are used consistently: $Y_{it}$ vs $Y_{i,t}$ -- pick one and stick with it.
  • Operators use \operatorname{} or predefined commands: \log, \exp, \max, \Pr, \E (defined as \DeclareMathOperator{\E}{\mathbb{E}}). Do not write log in italics as if it were a variable.
  • Greek letters are used consistently: if $\beta$ is the treatment effect in Section 3, it should not suddenly become the discount factor in Section 4.
  • Equation numbering is consistent: number all equations, or number only those referenced in the text. Do not number some arbitrarily.
  • Multi-line equations use align or aligned, not eqnarray (which has known spacing issues).
  • Parentheses and brackets scale correctly: use \left( and \right) or \big, \Big, etc. for expressions that extend vertically.
  • Error terms ($\varepsilon$ vs $\epsilon$) are consistent. The economics convention is typically $\varepsilon_{it}$.
  • Expectations, variances, and covariances use consistent notation: $\mathbb{E}[\cdot]$, $\text{Var}(\cdot)$, $\text{Cov}(\cdot, \cdot)$.

Define notation at the top of your document to ensure consistency and make global changes easy:

TeX
\newcommand{\E}{\mathbb{E}}
\newcommand{\Var}{\operatorname{Var}}
\newcommand{\Cov}{\operatorname{Cov}}
\newcommand{\indep}{\perp\!\!\!\perp}
\newcommand{\plim}{\operatorname{plim}}

2. Document Structure and Environments

Check these items:

  • Sections are numbered and follow a logical hierarchy: \section, \subsection, \subsubsection. Do not skip levels (e.g., \section directly to \subsubsection).
  • Theorems, propositions, lemmas, and assumptions use \newtheorem environments, not ad hoc formatting.
  • Proofs use the proof environment (from amsthm), which adds the QED symbol automatically.
  • The appendix uses \appendix command, which switches section numbering to A, B, C automatically.
  • Online appendix (if separate) is clearly labeled and has its own self-contained references section.
  • Title page includes: title, author(s), affiliation(s), date, abstract, JEL codes, keywords, and acknowledgments (often as a footnote on the title).
  • Double-spacing is applied for submission (use \usepackage{setspace} and \doublespacing), but check journal requirements.

3. Tables

Tables are the primary vehicle for results in empirical economics. They must be readable, well-labeled, and self-contained.

Check these items:

  • Every table has a descriptive title (caption) that allows the reader to understand it without reading the text.
  • Table notes (below the table) explain variable definitions, sample restrictions, standard error clustering, and significance levels.
  • Standard errors are in parentheses below coefficients. Do not use square brackets for standard errors (brackets are conventionally reserved for other statistics, e.g., t-statistics or confidence intervals).
  • Significance stars are defined consistently: * p < 0.10, ** p < 0.05, *** p < 0.01. Use the same thresholds throughout the paper. Consider omitting stars entirely and reporting confidence intervals instead.
  • Alignment: decimal-align numbers using the siunitx package (S column type) or dcolumn. Misaligned decimals make tables hard to scan.
  • Number of observations (N) and R-squared are reported for each column.
  • Column headers clearly identify the dependent variable and the specification.
  • Horizontal rules use \toprule, \midrule, and \bottomrule from the booktabs package. Never use vertical rules in economics tables.
  • Tables do not overflow the page margins. Use \resizebox or \adjustbox sparingly -- if you need to shrink a table significantly, it has too many columns.
  • Table placement uses [t] or [htbp] -- never [H] (from the float package), which forces exact placement and creates bad page breaks.
  • Regression tables produced by software (Stata's esttab, R's stargazer, Python's statsmodels) are manually reviewed for formatting consistency with the rest of the paper.
TeX
\begin{table}[t]
\centering
\caption{Effect of Policy on Wages}
\label{tab:main_results}
\begin{tabular}{lccc}
\toprule
                        & (1)       & (2)       & (3) \\
                        & OLS       & IV        & IV \\
\midrule
Treatment               & 0.034***  & 0.052**   & 0.048** \\
                        & (0.008)   & (0.022)   & (0.020) \\
Controls                & No        & No        & Yes \\
Fixed effects           & Year      & Year      & Year + Firm \\
\midrule
Observations            & 125,430   & 125,430   & 125,430 \\
R-squared               & 0.142     & ---       & --- \\
First-stage F           & ---       & 42.3      & 38.7 \\
\bottomrule
\end{tabular}
\begin{tablenotes}
\small
\item \textit{Notes:} Standard errors clustered at the firm level in
parentheses. The dependent variable is log hourly wage. The instrument
is the municipality-level lockdown severity index. * $p<0.10$,
** $p<0.05$, *** $p<0.01$.
\end{tablenotes}
\end{table}

4. Figures

Check these items:

  • Figures are vector format (PDF) when possible, not raster (PNG, JPG). This ensures they look sharp at any resolution.
  • Axes are labeled with units: "Wage (2019 USD)" not just "Wage."
  • Font size in figures is readable at the printed size. A common mistake is generating figures with tiny axis labels.
  • Legends do not obscure data. Place legends outside the plot area or use direct labeling of lines.
  • Color is used meaningfully and the figure is interpretable in grayscale (many referees print in black and white). Use different line styles (solid, dashed, dotted) in addition to colors.
  • Confidence intervals or bands are shown where appropriate (e.g., event-study plots).
  • The caption is informative: "Figure 3: Event-study estimates of the effect of policy reform on firm-level employment. Bars indicate 95% confidence intervals. The dashed vertical line marks the reform date."
  • Figure numbering is sequential and matches references in the text.
  • Subfigures use the subcaption package and are labeled (a), (b), etc.

5. Cross-References

Check these items:

  • All cross-references use \ref{}, \eqref{}, or \autoref{} -- never hardcoded numbers. Hardcoded references break when you reorder sections.
  • Every \label{} has at least one corresponding \ref{}. Unused labels are clutter.
  • Labels follow a consistent naming convention: tab:summary_stats, fig:event_study, eq:main_spec, sec:identification.
  • No "??" appears in the compiled document (a sign of undefined references). Compile twice to resolve forward references.
  • Table and figure references in the text capitalize appropriately: "Table 1" and "Figure 3" (capitalized when used as proper nouns), not "table 1."
  • Equation references use \eqref{} which adds parentheses automatically: "as shown in equation (3)" rather than "as shown in equation 3."

6. Bibliography

Check these items:

  • All citations compile without errors. No "?" in the text from undefined citation keys.
  • Every entry in the .bib file that is cited appears in the bibliography. Every entry in the bibliography is cited in the text (no phantom references).
  • Author names are consistent: if the .bib file has "De Chaisemartin" in one entry and "de Chaisemartin" in another, they will appear as different authors.
  • Journal names are abbreviated consistently (or not abbreviated consistently). Do not mix "American Economic Review" and "Am. Econ. Rev."
  • Working papers include the series name and number: "NBER Working Paper No. 28543."
  • Published papers include volume, issue, and page numbers.
  • DOIs or URLs are included for online-only publications.
  • The bibliography style matches the target journal. Most economics journals use author-year citation (natbib with \citet and \citep).
  • In-text citations use the correct form:
  • Parenthetical: \citep{author2020} produces "(Author, 2020)"
  • Textual: \citet{author2020} produces "Author (2020)"
  • Do not write "Author (2020)" manually -- use \citet.
  • Multiple citations are grouped: \citep{author2020, otherauthor2021} not \citep{author2020}; \citep{otherauthor2021}.

7. Typography and Formatting

Check these items:

  • Hyphens (-), en-dashes (--), and em-dashes (---) are used correctly. En-dashes for ranges: "pages 15--32," "2010--2020." Em-dashes for parenthetical statements.
  • Quotation marks use LaTeX convention: `single' and ``double'', not straight quotes.
  • Percent signs use \% in text and are consistent: "10 percent" or "10\%" throughout, not a mix.
  • Thousands separators are used for large numbers: "125,430" not "125430."
  • Non-breaking spaces (~) are used before references: Table~\ref{tab:main}, Section~\ref{sec:data}.
  • Acronyms are defined on first use: "randomized controlled trial (RCT)" then "RCT" thereafter.
  • The paper compiles without warnings (not just without errors). Overfull hbox warnings indicate text running into margins.

8. Common Package Recommendations

Purpose Package Notes
Table formatting booktabs \toprule, \midrule, \bottomrule
Number alignment siunitx S column type for decimal alignment
Citations natbib \citet, \citep for author-year
Subfigures subcaption Replaces deprecated subfig
Hyperlinks hyperref Load last; makes refs clickable
Colors xcolor For figure-matching colors in text
Spacing setspace \doublespacing for submission
Margins geometry Standard: 1-inch margins
Code listings listings or minted For appendix code display

Pre-Submission Checklist

  • Compile from scratch (delete .aux, .bbl, .log files) and verify clean build
  • No "??" from undefined references or citations
  • No overfull hbox warnings wider than 1pt
  • All figures render correctly (check embedded fonts)
  • Abstract fits within the journal's word limit
  • Page count is within the journal's limit (including appendix if counted)
  • Author information is anonymized if the journal uses double-blind review
  • Supplementary materials (online appendix, replication files) are referenced in the main text