PR-A: WooldridgeDiD primary-source review (Wooldridge 2025 EmpEcon) by igerber · Pull Request #484 · igerber/diff-diff

igerber · 2026-05-21T22:37:03Z

Summary

Adds the primary-source methodology review for WooldridgeDiD (ETWFE) at docs/methodology/papers/wooldridge-2025-review.md (single new file, 771 lines).
Covers Wooldridge (2025) Empirical Economics 69(5), 2545-2587 Sections 2 (TWFE foundation), 3 (two-way Mundlak + Theorem 3.1 Mundlak ≡ TWFE / Corollary 3.2 RE ≡ POLS), 4 (staggered identification + Procedure 4.1 cohort imputation), 5 (POLS / extended TWFE + Propositions 5.1 / 5.2 equivalence chain), 6 (event study + pre-trend tests), 7 (aggregation — Eq. 7.2-7.4 cohort-share weights), 8 (heterogeneous cohort-specific trends), 10 (extensions). Cross-links the existing docs/methodology/papers/wooldridge-2023-review.md for nonlinear extensions (logit / Poisson / ASF), which Wooldridge (2025) defers to (p. 2554).
Identifies a new untracked methodology finding: the existing REGISTRY ## WooldridgeDiD (ETWFE) § Aggregations Note documents only the Eqs. 7.2-7.4 simple-overall cell-vs-cohort-share weighting deviation; the same deviation extends to the Eq. 7.6 event-time path via aggregate("event") but is not yet in REGISTRY or TODO. Documented in the review's Deviations § "Untracked findings" subsection; lifting it into REGISTRY + TODO is intentionally deferred to the follow-up PR.
First of two PRs for the WooldridgeDiD (ETWFE) methodology-review-tracker promotion. The follow-up PR (methodology test file + R etwfe parity + implementation fixes if any + tracker promotion consolidation) will land after this paper review.

Methodology references (required if estimator / math changes)

Method name(s): WooldridgeDiD (ETWFE) — OLS / Mundlak / POLS-on-cohort-dummies / aggregation (simple, group, calendar, event-time) / event study / heterogeneous-trends Section 8.
Paper / source link(s):
- Wooldridge, J. M. (2025). Two-way fixed effects, the two-way mundlak regression, and difference-in-differences estimators. Empirical Economics, 69(5), 2545-2587. https://doi.org/10.1007/s00181-025-02807-z
- Wooldridge, J. M. (2023). Simple approaches to nonlinear difference-in-differences with panel data. The Econometrics Journal, 26(3), C31-C66. https://doi.org/10.1093/ectj/utad016 (secondary source; review already on disk).
Any intentional deviations from the source (and why): No new deviations introduced. The review documents an existing library deviation (cell-level n_{g,t} aggregation weights vs the paper's cohort-share ω̂_g weights at Eq. 7.2-7.4; already tracked) and a NEW untracked extension of the same deviation under Eq. 7.6 event-time. Both are tagged in the Deviations section but no REGISTRY / TODO surfaces are modified in this PR; consolidation is deferred to the follow-up PR.

Validation

Tests added/updated: No test changes. This is a documentation-only PR.
Backtest / simulation / notebook evidence (if applicable): N/A — paper review only.

Security / privacy

Confirm no secrets/PII in this PR: Yes

Adds the primary-source methodology review for Wooldridge, J. M. (2025). Two-way fixed effects, the two-way mundlak regression, and difference-in-differences estimators. Empirical Economics, 69(5), 2545-2587. DOI 10.1007/s00181-025-02807-z. Covers Sections 2 (TWFE foundation), 3 (two-way Mundlak + Theorem 3.1 / Corollary 3.2), 4 (staggered identification + Procedure 4.1 cohort imputation), 5 (POLS / extended TWFE + Propositions 5.1 / 5.2 equivalence chain), 6 (event study), 7 (aggregation — Eq. 7.2-7.4 cohort-share weights), 8 (heterogeneous cohort-specific trends), and 10 (extensions). Cross-links the existing 2023 secondary review for nonlinear extensions (logit / Poisson / ASF). Deviations section separates documented deviations (Eqs. 7.2-7.4 simple-overall aggregation already tracked in REGISTRY + TODO; R / Stata small-sample-correction items) from untracked findings (Eq. 7.6 event-time aggregation extension of the simple-overall deviation — new finding from this review, not yet in REGISTRY or TODO). First of two PRs for the WooldridgeDiD (ETWFE) methodology-review- tracker promotion. The second PR (methodology test file + R `etwfe` parity + implementation fixes if any + tracker promotion consolidation) will follow. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

github-actions · 2026-05-21T22:41:16Z

Overall Assessment

⚠️ Needs changes

Executive Summary

P1 Methodology: the new review correctly identifies that Wooldridge 2025 uses cohort-share weights for both overall aggregation (Eq. 7.4) and exposure-time aggregation (Eq. 7.6), but the PR explicitly leaves the Eq. 7.6 event-time deviation out of docs/methodology/REGISTRY.md, so the repo still has an undocumented methodology deviation in its authoritative surface. (link.springer.com)
The code evidence matches the review’s claim: diff_diff/wooldridge.py:L723-L753 stores _gt_weights as raw cell counts, and diff_diff/wooldridge_results.py:L150-L166 reuses those same weights for aggregate("event").
Existing registry coverage is narrower: docs/methodology/REGISTRY.md:L1489-L1500 documents the simple-overall Eq. 7.2-7.4 mismatch, but not the Eq. 7.6 event-time path.
Minor docs issue: docs/methodology/papers/wooldridge-2025-review.md:L769-L771 points readers to WooldridgeResults, but the actual class is diff_diff/wooldridge_results.py:L15-L20 WooldridgeDiDResults.
No performance or security concerns in scope; this is a docs-only diff.

Methodology

P1
Impact: docs/methodology/papers/wooldridge-2025-review.md:L607-L630 says the Eq. 7.6 event-time weighting deviation is a new finding “NOT yet tracked.” That matches both the implementation and the current Registry state: diff_diff/wooldridge.py:L723-L753 builds cell-count weights, diff_diff/wooldridge_results.py:L150-L166 uses them for event aggregation, and docs/methodology/REGISTRY.md:L1489-L1500 only documents the simple-overall Eq. 7.2-7.4 deviation. Under the review rubric, that remains an undocumented methodology deviation, not deferrable TODO-only debt. (link.springer.com)
Concrete fix: add an explicit **Note:** / deviation note under ## WooldridgeDiD (ETWFE) in docs/methodology/REGISTRY.md stating that aggregate("event") currently uses cell-count n_{g,t} weights instead of Wooldridge 2025 Eq. 7.6’s exposure-time cohort-share weights N_g / (N_q + ... + N_{T-e}); update TODO.md:L94-L97 at the same time so the deferred implementation item names the event path too.

Code Quality

No findings in scope.

Performance

No findings in scope.

Maintainability

No findings beyond the documentation cross-reference issue below.

Tech Debt

No additional deferrable tech-debt findings. The blocker above is methodological and needs Registry documentation rather than TODO-only tracking.

Security

No findings in scope.

Documentation/Tests

P3
Impact: docs/methodology/papers/wooldridge-2025-review.md:L769-L771 references a non-existent WooldridgeResults symbol. That will misdirect future reviewers who grep from the paper review into the implementation.
Concrete fix: rename the reference to WooldridgeDiDResults to match diff_diff/wooldridge_results.py:L15-L20.

Path to Approval

Update docs/methodology/REGISTRY.md so the WooldridgeDiD Aggregations note explicitly documents the Eq. 7.6 event-time weighting deviation.
Widen the existing Wooldridge TODO row in TODO.md:L94-L97 (or add a second row) so the tracked follow-up names both the simple and event aggregation paths.
Fix WooldridgeResults → WooldridgeDiDResults in the new review.

CI audit found the Eq. 7.6 event-time aggregation deviation was a real methodology mismatch not yet documented in REGISTRY/TODO. Both `simple` and `event` aggregations reuse the same `_gt_weights` cell-count array, so the deviation form is identical across paths. - REGISTRY § Aggregations Note: extend the cell-vs-cohort-share text to name both Eq. 7.4 (simple-overall) and Eq. 7.6 (event-time) - TODO.md row 95: widen the deferred `weights="cohort_share"` follow-up to cover both paths in one parameter addition - Wooldridge 2025 review: consolidate the prior split (documented vs untracked) into a single Deviations entry covering both paths now that REGISTRY documents both; fix WooldridgeResults reference Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

github-actions · 2026-05-21T23:17:09Z

🔁 AI review rerun (requested by @igerber)

Head SHA: 5e07796aa605078278cbdde5fd445c1d5a039190

Overall Assessment

✅ Looks good

Executive Summary

Re-review result: the prior P1 is resolved. docs/methodology/REGISTRY.md:L1490-L1499 now documents the Wooldridge (2025) Eq. 7.6 event-time weighting deviation, consistent with _gt_weights construction and reuse in diff_diff/wooldridge.py:L721-L753 and diff_diff/wooldridge_results.py:L150-L166.
The deferred implementation work is now properly tracked in TODO.md:L95 across both simple and event aggregation paths, so the remaining difference from Wooldridge (2025) is documented/tracked rather than silent.
The prior WooldridgeResults reference issue appears fixed; the new review now points to WooldridgeDiDResults at docs/methodology/papers/wooldridge-2025-review.md:L754-L762, matching diff_diff/wooldridge_results.py:L15-L20.
One minor P3 docs issue remains: the new review describes accepted REGISTRY deviation labels too narrowly at docs/methodology/papers/wooldridge-2025-review.md:L614-L618, even though this PR’s own mitigation uses a plain **Note:** in docs/methodology/REGISTRY.md:L1495.

Methodology

Severity: P3 informational. Impact: The affected method is WooldridgeDiD (ETWFE) aggregation weighting. The existing cell-count weighting deviation is now explicitly documented in the authoritative Registry for both the simple-overall and event-time paths, and that documentation matches the current implementation. docs/methodology/REGISTRY.md:L1490-L1499, diff_diff/wooldridge.py:L721-L753, diff_diff/wooldridge_results.py:L150-L166. Concrete fix: None required for approval.

Code Quality

No findings.

Performance

No findings.

Maintainability

No findings.

Tech Debt

Severity: P3 informational. Impact: TODO.md:L95 now tracks the deferred cohort-share implementation across both affected aggregation paths, so the remaining gap is properly recorded as tech debt rather than left untracked. Concrete fix: None required for approval.

Security

No findings.

Documentation/Tests

Severity: P3. Impact: The new paper review says documented deviations are recognized via **Note (deviation from R/Stata):** or **Deviation from R:**, but this repo’s review rubric also accepts plain **Note:**, and this PR’s own mitigation relies on plain **Note:** at docs/methodology/REGISTRY.md:L1495. That mismatch could mislead future reviewers about what counts as a documented deviation. docs/methodology/papers/wooldridge-2025-review.md:L614-L618, docs/methodology/REGISTRY.md:L1495. Concrete fix: Reword that sentence to include plain **Note:** or simply say “a labeled note/deviation entry in REGISTRY.md.”
No additional findings. The prior class-name reference issue appears resolved at docs/methodology/papers/wooldridge-2025-review.md:L754-L762, consistent with diff_diff/wooldridge_results.py:L15-L20.

CI R2 P3 (documentation): the review's intro to its Deviations section listed only `**Note (deviation from R/Stata):**` and `**Deviation from R:**` as accepted REGISTRY label forms, but the plain `**Note:**` form is also accepted (and is what the corresponding REGISTRY:1495 aggregation note uses). Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

github-actions · 2026-05-21T23:21:25Z

🔁 AI review rerun (requested by @igerber)

Head SHA: 5943f2225439643f87a9ef44bf8fbf020968d631

Overall Assessment

✅ Looks good

Executive Summary

Prior re-review issue resolved: the new review now explicitly recognizes plain **Note:** as a valid REGISTRY deviation label at docs/methodology/papers/wooldridge-2025-review.md:L614-L616.
The previously untracked Wooldridge aggregation deviation is now properly mitigated for both simple and event: docs/methodology/REGISTRY.md:L1489-L1499 and TODO.md:L95-L95 now match the implementation’s shared _gt_weights path in diff_diff/wooldridge.py:L740-L754 and diff_diff/wooldridge_results.py:L150-L165, while the paper’s target weights are cohort-share based in Eq. 7.4 and Eq. 7.6. (link.springer.com)
No new P1+ issues found in code or methodology tracking.
Two minor P3 docs issues remain in the new review file: one overstates when the paper-vs-implementation weighting difference appears, and one uses weighting= even though the tracked follow-up uses weights=.

Methodology

Severity: P3. Impact: docs/methodology/papers/wooldridge-2025-review.md:L614-L618 overstates when the paper’s cohort-share weights differ from the implementation, saying the gap appears in “imbalanced or staggered panels with heterogeneous cohort sizes.” By inference from Eq. 7.4 and Eq. 7.6, a balanced staggered panel with no missing cells still matches exactly because the implementation’s per-cell counts reduce to N_g within each available cell; unequal cohort sizes alone do not create a mismatch. That wording could mislead future reviewers into treating ordinary balanced staggered designs as methodology deviations when they are not. diff_diff/wooldridge.py:L740-L754, diff_diff/wooldridge_results.py:L150-L165. (link.springer.com) Concrete fix: Reword the sentence to say the gap appears when post-treatment cell counts vary within cohort or exposure horizon, e.g. unbalanced panels, missing cells, or filtered cells.

Code Quality

No findings.

Performance

No findings.

Maintainability

No findings.

Tech Debt

No findings. The deferred cohort-share implementation is now explicitly tracked for both aggregation paths in TODO.md:L95-L95, so the remaining gap is mitigated under the review rules.

Security

No findings.

Documentation/Tests

Severity: P3. Impact: docs/methodology/papers/wooldridge-2025-review.md:L723-L725 proposes a future weighting="cohort_share" / weighting="cell" API, but the tracked follow-up in TODO.md:L95-L95 and TODO.md:L203-L203 uses weights="cohort_share". This is minor, but it can create avoidable churn in the follow-up implementation/docs PR. Concrete fix: Align the review text with the tracked TODO API name, or rename both together if weighting= is the intended public keyword.
No test-gap finding. This PR is documentation-only.

CI R3 P3 (methodology): the prior wording said the cell-vs-cohort-share gap appears in "imbalanced or staggered panels with heterogeneous cohort sizes." Codex correctly notes that balanced staggered panels with uniform within-cohort cell counts still match exactly (cell-count reduces to N_g per cell, recovering the cohort-share form). Rewording to "post-treatment cell counts vary within cohort or exposure horizon" captures the actual divergence trigger (unbalanced panels, missing cells, sample-filter drops). CI R3 P3 (documentation): align the review's `weighting="cohort_share"` prose with the TODO row's tracked API name `weights="cohort_share"`. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

github-actions · 2026-05-21T23:25:47Z

🔁 AI review rerun (requested by @igerber)

Head SHA: 3a6853ebf24254e2b3e6286eb87edaff0c372a0d

Overall Assessment

✅ Looks good

Executive Summary

Prior re-review blockers remain resolved: the Wooldridge aggregation-weight deviation is now documented for both simple and event in docs/methodology/REGISTRY.md:L1489-L1499, and the deferred follow-up is tracked in TODO.md:L95-L95.
That tracking matches the implementation path: _gt_weights is built once in diff_diff/wooldridge.py:L740-L754 and reused by aggregate("event") in diff_diff/wooldridge_results.py:L150-L165.
No new P1+ methodology, inference, assumption-check, default-behavior, or edge-case issues are introduced by this docs/TODO-only PR.
One minor P3 methodology-doc wording issue remains in the new review: it assigns a fixed direction to the cohort-share vs. cell-count weighting gap, but that direction depends on where post-treatment cells are sparse or missing.
No code quality, performance, maintainability, security, or testing blockers.

Methodology

Severity: P3. Impact: docs/methodology/papers/wooldridge-2025-review.md:L616-L618 says the paper’s cohort-share form “down-weights short-exposure cohorts ... relative to the cell-count scheme when those conditions hold.” Given the paper formulas summarized at docs/methodology/papers/wooldridge-2025-review.md:L388-L418 and the implementation’s realized-cell weighting path in diff_diff/wooldridge.py:L740-L754 plus diff_diff/wooldridge_results.py:L150-L165, that direction is not fixed: cell-count weighting can either over- or under-weight a short-exposure cohort depending on which (g,t) cells fall below N_g. Concrete fix: rephrase that sentence to say only that the schemes diverge when realized post-treatment cell counts depart from the cohort-share formulas, without asserting a universal down-weighting direction.

Code Quality

No findings.

Performance

No findings.

Maintainability

No findings.

Tech Debt

Severity: P3-informational. Impact: the previously untracked event-time weighting deviation is now explicitly mitigated in docs/methodology/REGISTRY.md:L1489-L1499 and TODO.md:L95-L95, aligned with the shared _gt_weights path. Concrete fix: none required in this PR.

Security

No findings.

Documentation/Tests

No findings. This is a docs/TODO-only diff; no additional tests are required for approval.

CI R4 P3 (methodology): the prior wording asserted the paper's cohort-share form universally "down-weights short-exposure cohorts relative to the cell-count scheme." That direction isn't universal — it depends on which post-treatment cells are sparse, missing, or filtered. Replace with a direction-agnostic statement and keep the mechanical contrast between Eq. 7.4's exposure-multiplier denominator and the cell-count scheme. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

github-actions · 2026-05-21T23:29:13Z

🔁 AI review rerun (requested by @igerber)

Head SHA: b096838416ddf43bc3d03d94454015cde5012dd8

Overall Assessment

✅ Looks good

Executive Summary

Prior re-review item is resolved: the new review no longer assigns a fixed direction to the cohort-share vs. cell-count weighting gap and now correctly states that the direction depends on where sparse or missing post-treatment cells occur, docs/methodology/papers/wooldridge-2025-review.md:L618-L618.
The previously untracked Wooldridge (2025) event-time weighting deviation is now explicitly documented in docs/methodology/REGISTRY.md:L1490-L1499 and tracked in TODO.md:L95-L95.
That documentation matches the implementation: _gt_weights is built from realized (g,t) cell counts in diff_diff/wooldridge.py:L740-L754 and reused by the event aggregation path in diff_diff/wooldridge_results.py:L88-L165.
No new P1+ methodology, inference, assumption-check, control-group, or NaN/Inf issues are introduced by this docs/TODO-only PR.
One minor P3 documentation nit remains: the Gaps and Uncertainties ordered list skips item 3, docs/methodology/papers/wooldridge-2025-review.md:L725-L729.

Methodology

Severity: P3-informational. Impact: The PR properly mitigates the previously untracked Wooldridge 2025 aggregation-weight deviation by adding a labeled REGISTRY note and a matching TODO entry, and the note matches the real shared-weight code path in diff_diff/wooldridge.py:L740-L754 and diff_diff/wooldridge_results.py:L88-L165. Concrete fix: none required.

Code Quality

No findings.

Performance

No findings.

Maintainability

No findings.

Tech Debt

Severity: P3-informational. Impact: The remaining aggregation-weight implementation gap is now properly tracked in the accepted review-debt ledger at TODO.md:L95-L95, so it is not a blocker for this documentation PR. Concrete fix: none required in this PR.

Security

No findings.

Documentation/Tests

Severity: P3. Impact: The ordered list under ## Gaps and Uncertainties jumps from 2. to 4., which makes the new methodology review read like an incomplete edit and slightly reduces readability, docs/methodology/papers/wooldridge-2025-review.md:L725-L729. Concrete fix: renumber that list so it is contiguous.

CI R5 P3 (documentation): the ordered list jumped from 2 to 4 after the R2 removal of a stale Section 8 coverage gap entry. Renumber 4-14 → 3-13 so the list reads as a complete, contiguous edit. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

github-actions · 2026-05-21T23:32:52Z

🔁 AI review rerun (requested by @igerber)

Head SHA: 27acfc362f4290569c609d79fd964790fc6353f6

Overall Assessment

✅ Looks good

Executive Summary

Re-review scope: the prior documentation nit is resolved; the Gaps and Uncertainties list is now contiguous in docs/methodology/papers/wooldridge-2025-review.md:L725-L749.
The previously untracked Wooldridge (2025) event-time weighting deviation is now explicitly documented in docs/methodology/REGISTRY.md:L1490-L1499 and tracked in TODO.md:L95-L95.
That documentation matches the implementation: realized (g,t) cell-count weights are built in diff_diff/wooldridge.py:L730-L754 and reused by aggregate("event") in diff_diff/wooldridge_results.py:L88-L165.
No new unmitigated P0/P1 methodology, inference, assumption-check, control-group, or NaN/Inf issues are introduced by this docs/TODO-only PR.

Methodology

Severity: P3-informational. Impact: The only load-bearing methodology gap in scope is the known cohort-share vs. cell-count aggregation deviation, and this PR now documents it for both the simple-overall and event-time paths, aligning the paper review, registry, TODO tracking, and code. Concrete fix: none required in this PR. docs/methodology/REGISTRY.md:L1490-L1499, docs/methodology/papers/wooldridge-2025-review.md:L614-L621, TODO.md:L95-L95, diff_diff/wooldridge.py:L730-L754, diff_diff/wooldridge_results.py:L88-L165

Code Quality

No findings.

Performance

No findings.

Maintainability

No findings.

Tech Debt

Severity: P3-informational. Impact: The remaining weights="cohort_share" implementation work is explicitly tracked in the accepted review-debt ledger, so it is non-blocking for this PR. Concrete fix: none required here; follow the tracked TODO item. TODO.md:L95-L95

Security

No findings.

Documentation/Tests

No findings. This is a documentation-only PR, and the prior numbering issue has been corrected in docs/methodology/papers/wooldridge-2025-review.md:L725-L749.

Closes the WooldridgeDiD (ETWFE) methodology-review-tracker promotion in METHODOLOGY_REVIEW.md (In Progress → Complete), following the primary-source review for Wooldridge (2025) merged in PR-A (#484). Adds two paper-driven implementation surfaces and extends R-parity goldens to the nonlinear paths. Implementation: - `aggregate(weights="cohort_share")` on WooldridgeDiDResults implements paper Eqs. 7.4 (simple-overall) and 7.6 (event-time, restricted to k>=0) cohort-share aggregation weights as an opt-in alternative to the default cell-count weighting (matching Stata `jwdid_estat`). Inference fields fail-closed to NaN with UserWarning per paper Section 7.5 conditional-on-shares semantics; raises on `survey_design` (design-consistent totals deferred); raises on `type ∈ {"group","calendar"}` (no paper closed-form); raises on bootstrap fits (no matching bootstrap variant). Closes TODO row 95. - `cohort_trends=True` on `WooldridgeDiD.__init__` adds linear `dg_i · t` cohort-specific trend interactions (paper Section 8 / Eq. 8.1) for the OLS path. Rejects on logit/poisson per paper Section 8 OLS scope; rejects on survey_design pending full-dummy/TSL validation; enforces per-cohort pre-period identification check (≥ 2 observed pre-periods per treated cohort). Auto-routes to full-dummy mode regardless of vcov_type. Closes the PR-A Requirements Checklist heterogeneous-trends gap. Tests: - `tests/test_methodology_wooldridge.py` extended with 6 paper-equation-numbered methodology classes (Theorem 3.1, Proposition 5.1, Section 6 event study, Section 7 aggregation paths, Section 8 heterogeneous trends, Section 10 unbalanced panels) + `TestW2025LibraryDeviations` consolidating 5 surviving deviations. Mirrors the HAD PR #473 precedent. - Two new R-parity surface classes (`TestWooldridgeParityRPoisson`, `TestWooldridgeParityRLogit`) lock the structural surface against R `etwfe(family=...)` log-link goldens. - 209 tests total (60 methodology + 149 R-parity + unit regressions). R Goldens: - `benchmarks/R/generate_wooldridge_golden.R` extended with Poisson + logit DGPs via R `etwfe`; augmented panel CSV retains the same seed-generated `y_pois` + `y_logit` columns for cross-language reproducibility. - `benchmarks/R/requirements.R` pins `etwfe >= 0.5.0`. Tracker promotion: - METHODOLOGY_REVIEW.md L52 status flip with merge date; detail section L583-605 rewritten to the Verified Components / Test Coverage / Corrections Made / Deviations / Outstanding Concerns template mirroring HAD / ContinuousDiD / DCDH. L27 example re-pointed; priority queue items #7-#10 renumbered to #6-#9. - REGISTRY.md `## WooldridgeDiD (ETWFE)` extended with `### Deviations from the paper / from R / library extensions` block consolidating 7 surviving deviations + opt-in notes for cohort_share + cohort_trends + survey rejection + bootstrap cohort_share rejection contracts. - CHANGELOG.md `[Unreleased]` `### Added` documents the new parameters, R-parity extension, and tracker flip. - `docs/methodology/papers/wooldridge-2025-review.md` Requirements Checklist + Gaps & Uncertainties items 1 + 11 marked `**Status:** Closed in PR-B`. - `docs/api/wooldridge_etwfe.rst` updated with weighting-scheme notes alongside the existing aggregation table. Second of two PRs for the WooldridgeDiD methodology-review-tracker promotion. PR-A merged at e416aed (#484). Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

igerber added the ready-for-ci Triggers CI test workflows label May 22, 2026

igerber merged commit e416aed into main May 22, 2026
11 of 12 checks passed

igerber deleted the feature/wooldridge-2025-review-pr-a branch May 22, 2026 11:40

igerber mentioned this pull request May 22, 2026

PR-B: WooldridgeDiD tracker promotion + methodology bundle #486

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

PR-A: WooldridgeDiD primary-source review (Wooldridge 2025 EmpEcon)#484

PR-A: WooldridgeDiD primary-source review (Wooldridge 2025 EmpEcon)#484
igerber merged 6 commits into
mainfrom
feature/wooldridge-2025-review-pr-a

igerber commented May 21, 2026

Uh oh!

github-actions Bot commented May 21, 2026

Uh oh!

github-actions Bot commented May 21, 2026

Uh oh!

github-actions Bot commented May 21, 2026

Uh oh!

github-actions Bot commented May 21, 2026

Uh oh!

github-actions Bot commented May 21, 2026

Uh oh!

github-actions Bot commented May 21, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Conversation

igerber commented May 21, 2026

Summary

Methodology references (required if estimator / math changes)

Validation

Security / privacy

Uh oh!

github-actions Bot commented May 21, 2026

Overall Assessment

Executive Summary

Methodology

Code Quality

Performance

Maintainability

Tech Debt

Security

Documentation/Tests

Path to Approval

Uh oh!

github-actions Bot commented May 21, 2026

Uh oh!

github-actions Bot commented May 21, 2026

Uh oh!

github-actions Bot commented May 21, 2026

Uh oh!

github-actions Bot commented May 21, 2026

Uh oh!

github-actions Bot commented May 21, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant