Writing Style & the AEJ: Micro Intro Arc (aejmic-writing-style)
When to trigger
- The introduction opens with machinery (the model's notation, an estimator) instead of the question
- A reader cannot state your main result after the first page
- The abstract describes activities ("we study…") rather than the result
- Late-stage polish before submission (do this after the model and proofs settle)
The AEJ: Micro intro arc
A theory-first AEA journal rewards an introduction that makes a broad micro audience see the question and the result fast. The arc:
question / puzzle → model + equilibrium concept (in words) → main result (a proposition, stated in words) → what drives it / which assumption is doing the work → contribution & scope → brief roadmap.
-
Open with the economic question, not the model. A non-specialist should grasp what is being asked in the first breath.
-
State the result in words on page one — the proposition's content, not its proof. "We show that X, and this happens because Y."
-
Name the driving force early (which assumption / mechanism delivers the result) — this is the AEJ: Micro reader's payoff and ties to
aejmic-identification.
-
Frame the contribution as a result with scope, including what it does not establish.
- The literature delta belongs here in one or two sentences (full treatment in
aejmic-literature-positioning), not a survey.
AEA house-style conventions
-
Abstract: short and result-bearing; lead with the result, not the activity. (Confirm the current word limit on the AEA author page — volatile; 检索于 2026-06，以官网为准.)
-
JEL codes and keywords required.
-
Statistical presentation: for any empirical/experimental content, report standard errors / confidence sets, not significance asterisks.
-
Notation discipline: define on first use; consistent throughout; avoid clutter.
-
Proofs: key proofs self-contained in the paper (main text or appendix), not exiled to supplementary files.
-
Single-blind note: AEJ: Micro review is single-blind (the referee sees the author), so the manuscript need not be anonymized — but acknowledgments and self-citations should still read professionally, not promotionally.
Checklist
Anti-patterns
- Leading the introduction with the model's notation or the estimator (machinery-first)
- An abstract that lists activities ("we develop a model and run an experiment") with no result
- Burying the proposition's content until Section 4
- A literature survey in the introduction instead of a one-sentence delta
- Promotional self-citation ("our influential paper") rather than neutral phrasing
- Significance asterisks in tables
Worked vignette (illustrative)
A draft opens: "We consider a Bayesian game with type space Θ and a sender who commits to a signal π." Rewrite to the arc: "When can a better-informed seller credibly disclose product quality, and when does it pay to stay vague? We model disclosure as a commitment to a signal and show the seller-optimal policy pools high types — vagueness is profitable precisely because…". The question leads, the result is on page one, the driving force (pooling preserves credibility) is named. See ../../resources/worked-examples/01-introduction.md.
Output format
【Intro opens with】question? [Y/N]
【Result on page one】stated in words + driving force named? [Y/N]
【Contribution】result + explicit scope limit? [Y/N]
【Abstract】leads with result; JEL + keywords? [Y/N]
【House style】no asterisks; notation consistent; self-cites professional? [Y/N]
【Next step】aejmic-replication-package (proofs + code) then aejmic-submission
Supplementary resources
