Writing Style (jegeo-writing-style)
When to trigger
- The intro lands for one community but alienates the other — too much formalism for geographers, or too much narrative for economists
- The abstract is over the JEG limit or buries the spatial contribution
- The paper reads as "applied economics with maps" rather than economic geography
- Jargon from one tradition (NEG notation, or geography theory-speak) is unexplained for the other half of the audience
Writing for two audiences without watering down
The hardest craft problem at JEG is that every paragraph is read by an economist and a geographer, and the masthead is drawn from both. The goal is not to dilute to a lowest common denominator but to make each move legible across the bridge:
- Lead with the spatial economic question and its stakes, in plain English, before any model or method. Both communities should agree on what the paper is about by the end of paragraph one.
- Translate, don't drop, the technical apparatus. A formal modeler should state the mechanism in prose before the equations; a conceptual theorist should give a concrete instance before the abstraction.
- Name the mechanism, not just the result. "Cities X because of agglomeration force Y, which weakens at distance" travels; "we find a significant coefficient" does not.
- Respect both vocabularies but gloss them once. If you use "relatedness," "embeddedness," or "spatial equilibrium," define it the first time so the other half of the room stays with you.
The JEG intro arc
- The spatial economic puzzle and why it matters (place does real work here).
- What we knew — from both literatures — and the gap where they meet.
- What you do (data/design/model) in two or three sentences, mechanism-first.
- What you find, as a magnitude and a mechanism, not an asterisk.
- Why it matters for how we understand the economy in space (and, where relevant, for policy — but argued, not asserted).
The abstract (a hard constraint)
- ≤100 words (检索于 2026-06；以官网为准), followed by four keywords and JEL codes. This is tighter than most economics journals — every word fights for space.
- Front-load the spatial contribution; name the place/scale, the mechanism, and the headline magnitude. No literature throat-clearing.
- It must read as economic geography to both communities in the first sentence.
Prose discipline
- Active voice; one idea per paragraph; topic sentences that state the spatial claim.
- Magnitudes in interpretable units (per 10km, per region, percent of the gradient), not just coefficients.
- Keep the 8,000-word main text self-contained; the online appendix supports, it does not carry the argument (检索于 2026-06；以官网为准).
- Double-anonymous review: scrub self-identifying phrasing ("in our earlier work in [city]") from the prose, not just the title page.
Where to put the model and the geography
A structural feature of JEG papers is deciding how much formal apparatus sits in the main 8,000 words versus the online appendix. The house solution that satisfies both communities:
- Keep in the main text the mechanism, the key equation(s) that carry intuition, the identification argument, and the central maps/figures — everything a geographer needs to follow the story and an economist needs to trust the design.
- Push to the online appendix the full derivations, the complete parameter tables, secondary robustness, and method minutiae — support, not argument.
- Never let the appendix carry a load-bearing claim; if a referee must open the appendix to believe the result, the main text has failed its self-containment test.
This keeps the paper readable across the bridge while respecting the length limit — and it is exactly the balance the editors look for.
Checklist
Anti-patterns
- An intro that only an economist (or only a geographer) can follow past page one
- Equations or notation dropped in with no prose statement of the mechanism
- Geography theory-speak or NEG jargon left unglossed for the other half of the audience
- An abstract that spends its 100 words on the literature instead of the contribution
- "We find significant effects" with no magnitude, mechanism, or spatial interpretation
- Leaving traces of authorship in the body text under a double-anonymous policy
Worked vignette (illustrative)
An abstract opens: "Building on the new economic geography literature and recent advances in quantitative spatial models, this paper estimates a multi-region general equilibrium model with CES preferences to..." By word 30 a geographer has tuned out and no contribution has appeared. Rewritten contribution-first: "Where high-speed rail connects a region, manufacturing employment rises 7% — but neighboring unconnected regions lose 3%, so the network redistributes activity rather than creating it. We show this with a quantitative-spatial model disciplined by [data]." Now the place, mechanism, magnitude, and the net-vs-redistribution insight land in two sentences, legible to both communities, well under 100 words, with room for keywords and JEL.
Translating across the bridge — concrete moves
- Replace "we exploit exogenous variation in W_ij" with "we use the fact that two regions' exposure was set by [pre-existing geography], unrelated to later outcomes" — then keep the formal version for the methods reader.
- Replace bare "path dependence" or "lock-in" with a one-clause gloss the first time ("path dependence — early advantages compounding into persistent regional specialization").
- After any equation block, add a sentence: "In words, this says..." so the non-modeling half of the audience stays.
- State every key magnitude in a unit a policymaker or geographer can picture (per 10km, per region, share of the gradient), not only as a coefficient.
Output format
【Journal】Journal of Economic Geography
【Skill】jegeo-writing-style
【Para-1 question】plain-English spatial puzzle? [Y/N]
【Two-audience legibility】mechanism in prose + terms glossed? [Y/N]
【Abstract】≤100 words, contribution-first, 4 keywords + JEL? [Y/N]
【Findings】magnitude + mechanism, no bare significance? [Y/N]
【Anonymity】self-identifying phrasing scrubbed? [Y/N]
【Next skill】jegeo-replication-package