Referee Strategy & Pre-Mortem (ecj-referee-strategy)

When to trigger

• Identification, model, and robustness are in place and you want to harden the paper before submitting
• You want to anticipate the report rather than react to it
• You sense a likely objection but have not addressed it in the text
• You are choosing how to frame the contribution to survive a demanding general-interest referee

How EJ referees read

EJ referees read the paper as an economic argument for a broad audience and probe two things: is it credible, and is it of broad interest to economists at large? The prototypical EJ referee asks "so what — why should the wider profession care?" as hard as "is this identified?". They are unimpressed by a clean coefficient with no economic meaning or no general relevance, and they will name the alternative mechanism or the over-claimed generality. EJ review is single-blind (referees know who you are; you do not know them; verified 2026-06-20), so do not anonymize. Anticipating these objections is worth more than any extra robustness table, because a single unanswered first-order objection — especially "this is too narrow for EJ" — sinks the paper.

The objection taxonomy (pre-empt each in the text)

1. "Too narrow / not of broad interest." — A result only one subfield would read. Pre-empt: ecj-topic-selection / ecj-literature-positioning — argue the general lesson and external relevance explicitly.
2. "Where is the economics?" — A reduced-form result with no model/mechanism. Pre-empt: ecj-theory-model — state the mechanism and the prediction it tests.
3. "Your mechanism vs. the obvious alternative." — A rival economic story produces the same sign. Pre-empt: ecj-robustness — a discriminating test.
4. "Selection / endogeneity / weak design." — Treatment or sample is endogenous; TWFE on staggered timing. Pre-empt: ecj-identification — defend the assumption, show falsification, use a modern estimator.
5. "What identifies the parameter?" (structural) — Pre-empt: explicit "what identifies what" + untargeted-moment validation.
6. "Over-claimed generality." — the headline asserts external validity the design cannot support. Pre-empt: bound the claim to what the evidence licenses (ecj-robustness).
7. "Hard to read / exposition." — a generalist gets lost. Pre-empt: ecj-writing-style — clear exposition is part of the EJ bar, not optional polish.

Pre-mortem procedure

• Write the referee report you fear, in their voice, listing the three objections most likely to trigger a reject.
• For each, decide: fixable in text now (do it), needs new analysis (do it or scope it), or a genuine limitation (state it honestly and bound its damage).
• Identify the one first-order objection that would sink the paper — at EJ this is frequently "too narrow" or "no mechanism" — and make sure the manuscript answers it before a referee has to ask.
• Suggesting referees / handling COIs: name scholars who know the literature and would engage fairly; note conflicts honestly — the single-blind portal will flag obvious co-author/advisor ties.

Checklist

• "Too narrow for EJ" pre-empted with an explicit broad-interest / general-lesson argument
• "Where is the economics?" answered by an explicit mechanism/model
• The leading rival mechanism named and tested against
• Selection/endogeneity threat defended with a falsification test; modern estimator where needed
• (Structural) parameter identification made transparent
• Generality claims bounded to what the evidence supports
• Exposition checked so a generalist can follow the argument
• The single most dangerous objection identified and pre-empted in the text

Anti-patterns

• Submitting a narrow result and hoping the referee overlooks the broad-interest bar
• Anonymizing for a single-blind journal, or suggesting referees with a conflict the portal will catch
• A robustness section that never tests the leading rival mechanism
• Assuming broad importance instead of demonstrating it
• Hiding a known limitation rather than bounding it (referees find it anyway)
• Treating exposition as cosmetic when at EJ it is part of the decision

Output format

【Feared report】the 3 objections most likely to trigger reject
【First-order threat】the single objection that would sink it (often "too narrow" / "no mechanism")
【Pre-empt plan】per objection: fix-in-text / new-analysis / state-as-limitation
【Broad-interest + rival-mechanism status】addressed? [y/n]
【Exposition pass】a generalist can follow it? [y/n]
【Next】ecj-submission