ecta-referee-strategy

Referee and Co-Editor Strategy (ecta-referee-strategy)

When to trigger

• Before submitting, to find the hole a referee will find first
• When you suspect a proof step or a generality claim is the weak point
• When choosing how to present the contribution to survive a demanding referee pool
• When deciding what to put in the body vs. the Supplemental Material to pre-empt objections

Econometrica's referees are usually experts in exactly your method — often the authors of the results you build on. Reviews are demanding, multi-round, and the co-editor and referees check proofs closely. Assume every step will be read by someone who could have proven it themselves.

How Econometrica review works

• A submission is screened for eligibility (ES membership of at least one author, fee, the 45-page limit) and routed by the Editor (Marina Halac, Yale, 2025–2029 — verify) to a handling Co-Editor from a small board (e.g., Keisuke Hirano, Aureo de Paula, Marciano Siniscalchi, Pete Klenow, Patrick Kline). The Co-Editor's read carries decisive weight and they often pick referees who are the authors of the results you extend.
• Referees verify proofs, not just plausibility. A single genuine gap can sink the paper — this is a different failure mode from applied siblings (AER / QJE / JPE / REStud) where a shaky empirical narrative is the usual killer.
• The acceptance rate is low (~6–9%); outright first-round acceptance is rare and R&R is the common positive outcome, typically followed by conditional acceptance (at which point final files go to the ES Data Editor for a reproducibility check).
• Review times are long and there are typically multiple rounds of demanding revision.

(Verify the current board, acceptance rate, and timeline in the latest Report of the Editors and on the journal site; the procedure above is from official ES sources but evolves.)

Red-team the paper

Attack your own manuscript in the order a referee will:

1. Is the contribution real and general? Could a referee say "this is a special case of [known theorem]" or "this only works for the example"? Pre-empt with an explicit nesting/generality statement (see ecta-topic-selection, ecta-literature-positioning).
2. Is every proof complete? Hunt for "easy to see," unstated regularity conditions, a CLT or fixed-point theorem whose hypotheses you have not verified, a uniqueness claim with no argument. This is the most common rejection cause (see ecta-theory-model).
3. Are the assumptions minimal and defensible? A referee will ask why each is needed and whether a weaker one suffices. Have the answer ready (or weaken the assumption now).
4. Is there finite-sample evidence? Asymptotics with no Monte Carlo invites rejection. Are the designs fair, the competitors real, the size controlled (see ecta-robustness)?
5. Did you ignore the obvious competitor? A referee will name the method you should have compared to. Compare to it.
6. Does the replication package actually reproduce? A non-reproducing package is a credibility problem the ES Data Editor will surface at conditional acceptance — and that includes Monte Carlo / simulation code, which is in-scope under the ES policy (see ecta-replication-package).
7. Is this Econometrica-fit at all? The eligibility/co-editor screen returns "pure applied work with no methodological or identification innovation." If the contribution is an off-the-shelf causal estimate, expect a fast desk rejection (see ecta-topic-selection).

Presentation choices that help

• Put the hard step of the proof in full where a referee can find it; do not hide it.
• A short "Discussion of Assumptions" subsection naming what each buys and what relaxing it would cost defuses the "why this assumption?" line of attack.
• The Supplemental Material with extra results, robustness, and the long proofs lets the body stay within 45 pages and clean while pre-empting "what about X?"
• Acknowledge limitations explicitly; a referee who finds an unstated limitation distrusts the rest.

Suggested vs. excluded referees

• Suggest referees who know the method and have no conflict (no recent coauthorship, advisor / advisee, same department). Suggesting people who cannot fairly evaluate the work helps no one.
• If there is a genuine reason to exclude someone (active dispute, direct competitor on the same result), note it factually.

Checklist

• Generality / nesting pre-empted with an explicit statement
• Every proof step complete; no unstated regularity condition or unverified theorem hypothesis
• Each assumption justified; weaker alternatives considered
• Finite-sample evidence present, fair, and against a real competitor
• Obvious competing method addressed
• Replication package reproduces from a clean checkout
• Limitations stated honestly
• Referee suggestions are qualified and conflict-free

Anti-patterns

• Submitting with a known soft spot in a proof, hoping the referee misses it (they will not)
• Burying the hard step so the referee must reconstruct it — and gets annoyed
• No comparison to the method everyone in the field would reach for
• Suggesting only friendly or unqualified referees
• Treating the co-editor's likely concerns as an afterthought
• Assuming one round of revision; planning as if acceptance is near when an R&R is the realistic best case

Output format

【Most likely rejection cause】proof gap / thin generality / no finite-sample / ignored competitor / ...
【Proof audit】complete / gaps at: [...]
【Generality defense】...
【Finite-sample defense】...
【Competitor addressed】yes/no
【Pre-empt moves】[discussion-of-assumptions, supplemental-material item, ...]
【Referee suggestions】[qualified, conflict-free]
【Next step】ecta-submission