Reviewers in economics and management journals are, almost without exception, trained to scrutinize methodology closely. A paper with an interesting question and clean writing can still be rejected — or sent through multiple rounds of revision — because of avoidable statistical missteps. Here are the errors that show up most often, and how to steer clear of them.
1. Confusing Correlation With Causation
This is the most cited critique in economics reviews for a reason: it’s still common. If your paper makes a causal claim, be explicit about the identification strategy that justifies it — and be equally explicit when a finding is only correlational.
2. Weak or Unjustified Instruments
In instrumental variable designs, a weak first stage or an instrument whose exclusion restriction isn’t credibly argued is one of the fastest ways to draw reviewer skepticism. State the exclusion restriction directly and address obvious violations rather than hoping reviewers won’t ask.
3. p-Hacking and Selective Reporting
Running many specifications and reporting only the significant ones undermines the credibility of a paper’s findings — and increasingly, reviewers and editors are trained to look for signs of this. Pre-registration or transparent reporting of all specifications tested strengthens trust in your results.
4. Ignoring Standard Error Clustering
Failing to cluster standard errors at the appropriate level (e.g., firm, state, or individual, depending on the source of correlation in the data) can understate uncertainty and overstate statistical significance. Reviewers with applied econometrics backgrounds will flag this quickly.
5. Overlooking Multicollinearity
Highly correlated independent variables can produce unstable, hard-to-interpret coefficient estimates. A variance inflation factor (VIF) check is a simple, expected diagnostic in many empirical papers.
6. Small or Unrepresentative Samples Without Acknowledgment
A modest sample size isn’t automatically disqualifying, but failing to discuss its implications for statistical power and generalizability raises red flags. Address this directly in your limitations section rather than letting reviewers discover it unassisted.
7. Missing Robustness Checks
A single specification, however well-executed, rarely convinces a skeptical reviewer. Alternative model specifications, subsample analyses, and placebo tests demonstrate that your finding isn’t an artifact of one particular choice.
8. Misinterpreting Statistical Significance
Statistical significance is not the same as economic or practical significance. A precisely estimated but economically tiny effect should be described as such, not oversold as a major finding.
9. Endogeneity Left Unaddressed
If reverse causality or omitted variable bias is a plausible concern in your setting, address it directly — either through your empirical design or an explicit discussion of why it’s unlikely to drive your results.
10. Poor Table and Figure Construction
Beyond the analysis itself, tables that omit standard errors, sample sizes, or clear variable definitions make it hard for reviewers to evaluate your work — and that friction often gets read as a proxy for sloppier underlying analysis.
Before You Submit
Have a colleague with strong econometrics training review your empirical section specifically, independent of the paper’s broader argument. A second set of methodologically trained eyes catches issues that are easy to miss after months of working with the same data.
For a deeper technical reference on applied econometric methods and common pitfalls, the Royal Economic Society’s resources for researchers offer useful discipline-specific guidance.
Once your empirical strategy is airtight, revisit the journal’s scope and submit your manuscript through the paper submission page.