Econometrica Response to Reviewers: How to Write a Revision the Co-Editor Accepts (2026)

State the exact section, page, and proposition, lemma, or equation number for each revision, and name the table or appendix you changed. This is the single most-cited revision failure at Econometrica and across the theory-and-econometrics field.

A referee verifying a proof who has to hunt for your corrected step reads it as evasion. A referee who can turn straight to revised Proposition 2 on page 14 and the new Lemma A.3 in Appendix A finishes the verification faster.

Never write "we have addressed this in the manuscript" without a location. Use the numbering from the revised file, and flag explicitly when a repaired step lives in the appendix rather than the main text, because at Econometrica the appendix proofs are read, not skimmed.

The referees and the co-editor see your tone across every reply, and at re-review they read the new response beside the old reports. A defensive reply to Referee 1 is visible to the co-editor and to Referees 2 and 3. Calibrate.

The pattern that works: concede where the referee is right, do the real work, point to the exact location, and push back only on a request that is genuinely out of scope, with a reason and an alternative the co-editor can accept.

Econometrica is co-editor-led. The editor screens the paper, then routes it to a co-editor who leads the review process, assigns referees, and owns the decision. The co-editor can reject after partial review if the reports surface a theory-contribution or methodological concern the desk screen missed.

External review uses referees drawn from the relevant economics subfield, sometimes including an associate editor, under a confidential model where referee identities are not shared with authors.

The co-editor's central tool is the essential points list. Published co-editor guidance asks referees recommending a revise-and-resubmit to state the critical issues as precisely as possible. Those points are your specification.

SciRev community reports are useful as author-reported planning context, while the society's own documentation sets the formal rule: an offer to revise is typically valid for 12 months from the decision date.

Plan inside that window. Referees verify proofs and identification rather than skim them, and theory-heavy papers extend the clock for exactly that reason.

The defining technical feature is that the bar is the proof, not the prose. When a referee flags a gap, the revision must contain the complete corrected proof or the corrected lemma, not a sketch and not an assurance that the result still holds.

An identification concern is a request to rerun the argument from the moment conditions or the structural model. If it breaks, revise the result rather than concede the point in text and leave the theorem standing.

The second defining feature arrives at acceptance: the Data and Code Availability Policy. After conditional acceptance, the paper undergoes reproducibility checks through the Econometric Society Data Editor workflow before final acceptance.

The package needs a PDF README, raw data, all code, and documentation of every variable, sufficient for the Data Editor workflow to reproduce the empirical, experimental, or simulation results. If the materials do not reproduce, acceptance is held until they do. If required materials are not received within a year of conditional acceptance and no extension is granted, the submission can be returned to the co-editor.

How this compares to the field matters for calibration. At a general-interest economics flagship like the American Economic Review or the Quarterly Journal of Economics, the essential-points discipline and data-policy machinery are similar, but the contribution bar is broader and more empirical.

At a finance flagship like the Journal of Finance, page-limit and submission-fee constraints dominate more of the early process. Econometrica sits at the theory-and-econometrics apex: the proof-verification bar is heavier, the identification scrutiny is sharper, and referees check your corrected algebra, not just your claims about it.

In our pre-submission review work with Econometrica manuscripts, the responses that stall in a second revision round share a small set of recurring weaknesses. These are the same ones referees flag at re-review. In our analysis of Econometrica-targeted resubmissions, each weakness below maps to a specific, named failure pattern in the editorial culture, and each is testable against your own draft response and revised manuscript before you upload.

Answering a proof gap with hand-waving instead of a complete corrected proof. The most common and most expensive pattern in our Econometrica pre-submission reviews is a response that meets a referee's flagged gap in a proof with a paragraph asserting the result still holds, rather than the corrected proposition and lemma. Referees at Econometrica verify the algebra; a sketch fails verification and triggers another round.

Across our Econometrica revision reviews, this mismatch between a flagged proof gap and a non-proof reply is the single strongest predictor of a rejection on revision.

Conceding an identification point in text while leaving the result intact. Because the co-editor scrutinizes identification as hard as the headline estimate, a response that grants a referee's identification concern in prose but does not rerun the statistical analysis or revise the affected result reads as evasive. In our Econometrica pre-submission reviews we routinely find an identification objection answered with a hedge in the Discussion and an unchanged theorem; the correct move is to rederive the argument from the methods and report whether the estimate survives.

Shipping an incomplete or unreproducible replication package. A revision can resolve every essential point and still stall at the Data Editor. In our pre-submission review work with Econometrica manuscripts, the reproducibility failures we see most are a missing README, undocumented variables, hard-coded local paths, or code that does not run end to end. The Data Editor reproduces every empirical, experimental, and simulation result before acceptance, so a package that does not run holds your acceptance hostage.

Treating the response as a private note instead of a co-editor specification. Because the co-editor owns the decision and re-reads your response against the essential points, a defensive or scattered reply that does not map cleanly to those points does lasting damage.

In our Econometrica pre-submission reviews, the responses we flag hardest are the ones that answer referee comments at random, miss an essential point, or push back on a figure or table request without an alternative. The same revision, reorganized around the co-editor's three points with a complete proof and documented supplementary evidence, reads as the work of an author the journal can publish.

Resolve every essential point with real proof, rerun the identification, ship a reproducible package, and write to the co-editor. That four-part discipline is what separates an Econometrica revision that clears one round from one that stalls into a second or a rejection. Scan my Econometrica revision for proof gaps and unaddressed essential points before you resubmit.

Source: Manusights pre-submission reviews of Econometrica-targeted resubmissions, 2025 cohort.

Authors consistently underestimate the proof and replication effort and overestimate the writing. This breakdown is about workload, not the journal's decision clock; for the end-to-end schedule, see the Econometrica under review status guide.

Source: Manusights pre-submission reviews of Econometrica resubmissions, 2025 cohort, last updated June 7, 2026.

A revise-and-resubmit at Econometrica is not a soft acceptance. The paper can still end in rejection after re-review if an essential point is unresolved, a proof is still incomplete, or the identification claim is not repaired.

Papers rarely run past one or two revision rounds unless eventual publication is already likely, so a weak first revision is expensive. In our review work, the most common failure at this stage is a flagged proof gap answered with discussion rather than a complete corrected proof. The second most common is an identification concern conceded in text while the theorem stays intact.

Think twice before you resubmit if any of these are true. The response uses generic "we have addressed this" language with no section, page, or proposition numbers. A referee flagged a proof gap and you answered with a sketch or an assurance.

You granted an identification point in prose but left the result unchanged. The replication package is not yet reproducible end to end, which becomes a hard block at the Data Editor stage and can stall a conditional acceptance. Fixing these before resubmission is what keeps a second round from becoming a rejection.

The official editorial procedures tell you to submit a revised manuscript with a response to the referees and to prepare a replication package for the Data Editor workflow at acceptance. They do not tell you how the revision reads in practice.

The co-editor's essential points, not the full set of referee comments, are the real specification. A flagged proof gap demands a complete corrected proof rather than a sketch. An identification concern is a request to rerun the argument. The Data Editor workflow can block final acceptance if the package does not reproduce.

Those facts change how you write every reply. The patterns above come from our pre-submission reviews of Econometrica revisions, and they are testable against your own draft today.

• Manusights pre-submission reviews of Econometrica-targeted manuscripts (2025 cohort)