Contribution Framing (jeg-contribution-framing)
When to trigger
- The introduction explains what you did but not why it matters for growth
- You cannot state the contribution in one sentence without hedging
- A co-author asks "what's the one thing a reader remembers?"
The JEG contribution bar
JEG is a specialist outlet, so the contribution must be a genuine advance in understanding growth or dynamic-macro dynamics, not a technique demonstration or a within-field estimate that mentions growth. The journal values both:
- Theoretical contributions — a new mechanism, a more general result, a unification of mechanisms, a resolution of a tension in existing growth models, or a tractable model that delivers a sharp, testable growth prediction.
- Empirical contributions — credible evidence on a growth-relevant causal channel, a new long-run or cross-country dataset that resolves a measurement gap, or a test that discriminates between competing growth theories.
Frame the contribution
- One-sentence claim. "We show that [growth mechanism / determinant] [does what], which [overturns / generalizes / first-quantifies] [the standard account]."
- Type-tag it. New mechanism? More general theorem? New identification? New data? Quantitative magnitude? Be specific; JEG referees discount vague "we contribute to the literature."
- Bound it honestly. State the scope of the claim — which economies, which period, which assumptions. Specialist referees punish overreach harder than modest scope.
- Connect to the growth debate. Say which live question the result moves (sources of cross-country income gaps, the demographic transition, the growth effect of institutions/human capital/trade/finance).
Theory vs empirics calibration
- Theory paper: lead with the mechanism and its generality; the contribution is the result and its reach, not the algebra. Note what comparative statics or testable predictions it yields.
- Empirical paper: lead with the credibly identified magnitude and what it teaches about the mechanism; the contribution is the answer to a growth question, not the regression table.
Anti-patterns
- "We are the first to study X in country Y" with no growth mechanism at stake.
- Claiming a general theorem while the result hinges on a knife-edge assumption.
- Selling a method (a new estimator, a new model class) as the contribution when JEG wants the growth insight it buys.
- Listing three weak contributions instead of one strong one.
Worked vignette — sharpening a persistence claim
A draft says: "We study the long-run effects of medieval trade routes on regional development in Europe." Walked through the rules above (all numbers illustrative):
- One-sentence claim: "We show that a one-standard-deviation increase in medieval market access predicts roughly 14 log points higher regional income per capita today, and that about two-thirds of this gap runs through pre-industrial human-capital accumulation rather than later agglomeration."
- Type-tag: new identification (route-network instrument) + magnitude + mechanism decomposition — three tags that reinforce one claim, not three separate claims.
- Scope bound: 1,100 European subnational regions; an effect local to regions near historical routes; explicitly not a statement about national growth rates.
- Debate moved: whether deep-rooted market institutions or industrial-era agglomeration explain within-Europe income dispersion.
The bare reduced-form persistence coefficient would not have cleared the JEG bar; the mechanism share is what converts a long-run correlation into a growth contribution.
Referee pushback on the claim — and the JEG fix
- "Persistence without mechanism" → add a decomposition or intermediate-period outcomes; a growth referee treats an unexplained long-run correlation as a stylized fact, not a contribution.
- "Magnitude not interpretable" → restate in steady-state income ratios, half-lives of convergence, or years-of-development equivalents — never leave the headline as a bare coefficient.
- "The 'first' is overstated" → bound novelty to the strand, sample, and identification, and cite the nearest prior estimate by name.
- "This is development economics, not growth" → re-anchor the claim to divergence, convergence, or transition dynamics; if that re-anchoring is impossible, the venue is wrong, not the framing.
Output format
【One-sentence contribution】"We show that ..."
【Type】new mechanism / general result / new identification / new data / magnitude
【Scope bound】economies / period / assumptions
【Debate moved】which growth question
【Next skill】jeg-identification-strategy