Writing Style (jet-writing-style)
When to trigger
- Drafting or polishing a JET theory manuscript
- Converting a working paper into JET's elsarticle theorem-proof format
- Tightening notation and statements before submission
JET house style
- Typeset in LaTeX, Elsevier
elsarticle class. Source must be editable .tex; PDF is not
accepted as a source file. Use proper theorem environments (amsthm): Definition, Assumption,
Proposition, Lemma, Theorem, Corollary, Proof.
- Theorem-proof structure, not narrative-empirical. Set up the environment and primitives, state
assumptions, then results, then proofs (heavy ones to an appendix). The reader should be able to read
every formal statement without hunting through prose.
- Notation discipline. Introduce each symbol once, never reuse a symbol for two objects, keep the
body and proofs identical. A notation table helps in long papers.
- References. First submission can use any complete, consistent reference format. JET's proof-stage
style is name-year / author-year;
elsarticle-harv is the safest LaTeX default. References
cited in the abstract must be given in full. Keep unpublished results / personal communications
out of the reference list.
- Abstract / keywords / Highlights. Keep the abstract to 250 words or fewer, supply 1-7
English keywords, and include Highlights only if you can make 3-5 bullets of 85 characters or
fewer each. JEL codes are optional metadata rather than a substitute for clear keywords.
- Generative-AI disclosure belongs in the submission, not the prose; reviewers/editors are barred
from using such tools — write so a human expert can follow every step.
Prose targets
- State each theorem so it is self-contained — readable without the surrounding paragraph.
- Give a one-line intuition before a long proof; do not let intuition substitute for rigor.
- Prefer "We prove that …" over hedged "essentially / roughly" phrasing the theorem does not support.
Notation audit
Before submission, create a notation pass:
- Every primitive is introduced before first use.
- Every assumption label is referenced in at least one theorem or lemma.
- Every theorem statement can be read without searching the previous paragraph.
- Symbols in figures, proofs, and appendices match the body exactly.
- Appendix proof numbering preserves the main theorem dependency order.
Notation consistency is not cosmetic in JET; it is how referees verify the argument.
amsthm scaffold for a JET manuscript
% Body carries statements + a sketch; the appendix carries full proofs with restatements.
\begin{assumption}\label{a:sc}
Each type's utility satisfies strict single crossing in (action, type).
\end{assumption}
\begin{theorem}\label{thm:main}
Under Assumptions \ref{a:sc}--\ref{a:compact}, an optimal mechanism exists and is deterministic.
\end{theorem}
\begin{proof}[Sketch] % ≤ 1 paragraph in the body: name the key step, point to the appendix
The allocation is pinned down by Lemma~\ref{lem:mono}; a duality argument then delivers the
transfers. Full details are in Appendix~B.
\end{proof}
The pattern to copy: the theorem cites its assumptions by label, the sketch names the key
argument in one sentence, and nothing in the body forces a referee to page-flip mid-statement.
Sentence-level register for theory prose
- Statements quantify ("for every", "there exists") rather than gesture ("in general",
"typically").
- The body says what each lemma buys economically ("Lemma 2 rules out pooling at the top"); the
appendix carries the epsilon-management.
- "It is easy to see" is banned wherever a referee could disagree — replace it with the one-line
reason or a pointer to the appendix step.
- Intuition paragraphs are labeled as intuition and are never load-bearing: deleting them must
leave the proof intact.
Hedged-claim rewrite (micro-example)
- Before: "Our results show that ambiguity essentially eliminates the gains from screening."
- After: "Theorem 3 shows that for maxmin bidders with full-support prior sets, the optimal
mechanism is a posted price; screening gains are zero in this class — and only in this class
(Example 4)."
The rewrite replaces "essentially" with the exact class where the claim holds and the example
that bounds it. Run this transformation on every hedge word in the abstract and introduction.
Anti-patterns
- Submitting PDF source instead of
.tex
- Mixed reference styles, or abstract citations abbreviated to "et al."
- Symbol reuse and notation drift between statement and proof
- An empirical-paper voice (data-story framing) imposed on a theory result
Output format
【Format】elsarticle .tex, amsthm environments? [Y/N]
【Notation】one symbol ↔ one object, body=proof? [Y/N]
【References】author-year ready; abstract refs in full; no personal comms? [Y/N]
【Metadata】≤250-word abstract; 1-7 keywords; Highlights compliant if used? [Y/N]
【Disclosures】AI declared at submission? [Y/N]
【Next】jet-tables-figures / jet-submission
Source: brycewang-stanford/Awesome-Journal-Skills → Journal-of-Economic-Theory-Skills/skills/jet-writing-style/SKILL.md
1---2name: jet-writing-style3description: Use when applying the Journal of Economic Theory (JET) house style to a manuscript — the Elsevier elsarticle theorem-proof format with definition/assumption/proposition/theorem/proof structure, disciplined notation, full references in the abstract, 250-word abstract, 1-7 keywords, optional Highlights, and author-year proof-stage references. For drafting and polishing JET prose; it does not check proofs.4---567# Writing Style (jet-writing-style)89## When to trigger1011- Drafting or polishing a JET theory manuscript12- Converting a working paper into JET's elsarticle theorem-proof format13- Tightening notation and statements before submission1415## JET house style1617- **Typeset in LaTeX, Elsevier `elsarticle` class.** Source must be editable `.tex`; PDF is not18 accepted as a source file. Use proper theorem environments (`amsthm`): Definition, Assumption,19 Proposition, Lemma, Theorem, Corollary, Proof.20- **Theorem-proof structure, not narrative-empirical.** Set up the environment and primitives, state21 assumptions, then results, then proofs (heavy ones to an appendix). The reader should be able to read22 every formal statement without hunting through prose.23- **Notation discipline.** Introduce each symbol once, never reuse a symbol for two objects, keep the24 body and proofs identical. A notation table helps in long papers.25- **References.** First submission can use any complete, consistent reference format. JET's proof-stage26 style is **name-year / author-year**; `elsarticle-harv` is the safest LaTeX default. **References27 cited in the abstract must be given in full.** Keep **unpublished results / personal communications28 out of the reference list.**29- **Abstract / keywords / Highlights.** Keep the abstract to **250 words or fewer**, supply **1-730 English keywords**, and include Highlights only if you can make **3-5 bullets** of **85 characters or31 fewer** each. JEL codes are optional metadata rather than a substitute for clear keywords.32- **Generative-AI disclosure** belongs in the submission, not the prose; reviewers/editors are barred33 from using such tools — write so a human expert can follow every step.3435## Prose targets3637- State each theorem so it is **self-contained** — readable without the surrounding paragraph.38- Give a **one-line intuition** before a long proof; do not let intuition substitute for rigor.39- Prefer "We prove that …" over hedged "essentially / roughly" phrasing the theorem does not support.4041## Notation audit4243Before submission, create a notation pass:4445- Every primitive is introduced before first use.46- Every assumption label is referenced in at least one theorem or lemma.47- Every theorem statement can be read without searching the previous paragraph.48- Symbols in figures, proofs, and appendices match the body exactly.49- Appendix proof numbering preserves the main theorem dependency order.5051Notation consistency is not cosmetic in JET; it is how referees verify the argument.5253## amsthm scaffold for a JET manuscript5455```latex56% Body carries statements + a sketch; the appendix carries full proofs with restatements.57\begin{assumption}\label{a:sc}58Each type's utility satisfies strict single crossing in (action, type).59\end{assumption}6061\begin{theorem}\label{thm:main}62Under Assumptions \ref{a:sc}--\ref{a:compact}, an optimal mechanism exists and is deterministic.63\end{theorem}6465\begin{proof}[Sketch] % ≤ 1 paragraph in the body: name the key step, point to the appendix66The allocation is pinned down by Lemma~\ref{lem:mono}; a duality argument then delivers the67transfers. Full details are in Appendix~B.68\end{proof}69```7071The pattern to copy: the theorem cites its assumptions by label, the sketch names the key72argument in one sentence, and nothing in the body forces a referee to page-flip mid-statement.7374## Sentence-level register for theory prose7576- Statements **quantify** ("for every", "there exists") rather than gesture ("in general",77 "typically").78- The body says what each lemma buys economically ("Lemma 2 rules out pooling at the top"); the79 appendix carries the epsilon-management.80- "It is easy to see" is banned wherever a referee could disagree — replace it with the one-line81 reason or a pointer to the appendix step.82- Intuition paragraphs are labeled as intuition and are never load-bearing: deleting them must83 leave the proof intact.8485## Hedged-claim rewrite (micro-example)8687- **Before:** "Our results show that ambiguity essentially eliminates the gains from screening."88- **After:** "Theorem 3 shows that for maxmin bidders with full-support prior sets, the optimal89 mechanism is a posted price; screening gains are zero in this class — and only in this class90 (Example 4)."9192The rewrite replaces "essentially" with the exact class where the claim holds and the example93that bounds it. Run this transformation on every hedge word in the abstract and introduction.9495## Anti-patterns9697- Submitting PDF source instead of `.tex`98- Mixed reference styles, or abstract citations abbreviated to "et al."99- Symbol reuse and notation drift between statement and proof100- An empirical-paper voice (data-story framing) imposed on a theory result101102## Output format103104```105【Format】elsarticle .tex, amsthm environments? [Y/N]106【Notation】one symbol ↔ one object, body=proof? [Y/N]107【References】author-year ready; abstract refs in full; no personal comms? [Y/N]108【Metadata】≤250-word abstract; 1-7 keywords; Highlights compliant if used? [Y/N]109【Disclosures】AI declared at submission? [Y/N]110【Next】jet-tables-figures / jet-submission111```112113---114115**Source:** [`brycewang-stanford/Awesome-Journal-Skills`](https://github.com/brycewang-stanford/Awesome-Journal-Skills) → `Journal-of-Economic-Theory-Skills/skills/jet-writing-style/SKILL.md`