---
title: "DID replication archive"
format:
  html:
    embed-resources: true
    toc: true
    theme: cosmo
    include-in-header:
      text: |
        <style>body{font-family:system-ui,sans-serif;line-height:1.55}main{max-width:1100px!important}img{max-width:100%;height:auto}details{border:1px solid #ddd;border-radius:6px;margin:12px 0;padding:12px}summary{cursor:pointer;font-weight:600}table{display:block;overflow:auto;font-size:.8rem;max-width:100%}td,th{padding:6px}pre{max-width:100%;overflow:auto}.cell-output-display{overflow:auto}p code{overflow-wrap:anywhere}</style>
---

**22 papers · 335 registered specifications · 271 estimated · 253 homogeneity tests · 52 rejections at 5%.**

[All estimates: JSON](aggregate-results.json) · [CSV](aggregate-results.csv) · [Code and configurations](replication-code.zip) · [Editable template](https://lalten.org/jot/s/svqqvfc15sdryq)

## Expanded estimator archive

**22 papers · 335 specifications · 821 available modern-estimator summaries.**

The modern estimators cover **202 specifications in 17 papers**. The plots below show **808 modern-estimator comparisons**, alongside the original BJS and Sun–Abraham comparisons. Multisynth's 9 cohort and 4 individual estimates remain in the paper reports and downloads; their sparse archive scatters are omitted.

[Expanded scatter data: JSON](expanded-aggregate-results.json) · [CSV](expanded-aggregate-results.csv) · [Estimator coverage](expanded-coverage.json) · [Editable template](https://lalten.org/jot/s/svqqvfc15sdryq)

Both axes use the matching **TWFE standard error**: x = TWFE / SE(TWFE), y = modern estimate / SE(TWFE). CS-DID appears twice: an equal-event-time average and an observed-treated-cell average. Paper reports show adoption timing, estimates and event studies; the downloads retain fit settings and failure diagnostics.

| Paper | Assessed specifications | Expanded report |
|:--|--:|:--|
| 1. Curriculum and Ideology | 103 | [Report](study-01-expanded.html) · [JSON](study-01-expanded.json) |
| 2. Do Powerful Politicians Really Cause Corporate Downsizing? | 21 | [Report](study-02-expanded.html) · [JSON](study-02-expanded.json) |
| 3. Reply: Do Powerful Politicians Really Cause Corporate Downsizing? | 16 | [Report](study-03-expanded.html) · [JSON](study-03-expanded.json) |
| 4. Market Integration Demand and the Growth of Firms | 7 | [Report](study-04-expanded.html) · [JSON](study-04-expanded.json) |
| 5. Do Government Audits Reduce Corruption? | 3 | [Report](study-05-expanded.html) · [JSON](study-05-expanded.json) |
| 6. Railroads of the Raj | 1 | [Report](study-06-expanded.html) · [JSON](study-06-expanded.json) |
| 7. The Arrival of Fast Internet and Employment in Africa | 3 | [Report](study-07-expanded.html) · [JSON](study-07-expanded.json) |
| 8. Copyrights and Creativity | 4 | [Report](study-08-expanded.html) · [JSON](study-08-expanded.json) |
| 9. Vertical Integration Supplier Behavior and Quality Upgrading | 7 | [Report](study-09-expanded.html) · [JSON](study-09-expanded.json) |
| 10. Leadership in Social Movements | 2 | [Report](study-10-expanded.html) · [JSON](study-10-expanded.json) |
| 11. Hall of Mirrors | 1 | [Report](study-11-expanded.html) · [JSON](study-11-expanded.json) |
| 12. The Rise and Fall of Local Elections in China | 4 | [Report](study-12-expanded.html) · [JSON](study-12-expanded.json) |
| 13. Ideology and Performance in Public Organizations | 3 | [Report](study-13-expanded.html) · [JSON](study-13-expanded.json) |
| 14. Reconciliation Narratives | 4 | [Report](study-14-expanded.html) · [JSON](study-14-expanded.json) |
| 15. Retail Pharmacies and Drug Diversion | 2 | [Report](study-15-expanded.html) · [JSON](study-15-expanded.json) |
| 16. The Gift of a Lifetime | 9 | [Report](study-16-expanded.html) · [JSON](study-16-expanded.json) |
| 17. Trading Nontradables | 3 | [Report](study-17-expanded.html) · [JSON](study-17-expanded.json) |
| 18. The Rise of Fiscal Capacity | 14 | [Report](study-18-expanded.html) · [JSON](study-18-expanded.json) |
| 19. The Real State | 26 | [Report](study-19-expanded.html) · [JSON](study-19-expanded.json) |
| 20. Credit Markets Property Rights and the Commons | 1 | [Report](study-20-expanded.html) · [JSON](study-20-expanded.json) |
| 21. Landmines and Spatial Development | 11 | [Report](study-21-expanded.html) · [JSON](study-21-expanded.json) |
| 22. Corruption as a Local Advantage | 90 | [Report](study-22-expanded.html) · [JSON](study-22-expanded.json) |

## Estimator comparisons

Both coordinates use the **same TWFE standard error**: x = TWFE / SE(TWFE), y = modern estimate / SE(TWFE). The diagonal is equality. Color identifies the paper; circles identify static archive specifications and squares identify event-study averages. **Only the existing BJS and Sun–Abraham figures use outer black rings for rejection of cohort homogeneity at p < 0.05.** The new estimator figures have no F-test annotations. Rings use the homoskedastic deviations F-test, without multiple-testing adjustment; unknown tests are not classified as non-rejections.

Each figure opens at **±5**, matching Peter’s capped view. Coordinates outside that range are moved to the boundary and marked with diamonds; capped event-study averages retain their square inside a diamond. Hover for actual values, specification IDs, and (only in the two original regression comparisons) F-test p-values, or select **Full range**. All points are filled here; filling does not encode Peter’s paired-difference significance test.

<!-- CALIBRATION_FIGURES -->

### What is being compared

Each specification uses the archive’s outcome-SD denominator and source controls, weights and fixed effects. All four estimators use a common supported sample. Static archive specifications use the static TWFE coefficient on the horizontal axis; event-study specifications use the dynamic TWFE post-period average. BJS and Sun–Abraham share the configured post-period target. Unbinned event studies and common-sample exclusions differ from Peter’s frozen original comparisons; these plots contain only the new matched reruns.

Twelve Sun–Abraham fits that disagree with the reference saturated design are withheld from the comparisons. The F-test restricts cohort deviations at event times greater than zero; event zero and leads remain unrestricted. Estimate standard errors are cluster-robust, with a two-stage correction for imputation. The F-test is homoskedastic. The [sparse-backend validation](sparse-validation.json) compares the large-panel equations with the imported reference design.

## Additional comparisons

[Credit Markets: CS-DID, SDID, matrix completion, horizontal ridge and multisynth](credit-markets-counterfactuals.html) — additional estimators on the same sample, with pooling settings, results and executable source.

## Papers

| Paper | Registered | Estimated | F-tests | Report / data |
|:--|--:|--:|--:|:--|
| 1. Curriculum and Ideology | 103 | 103 | 103 | [Report](study-01.html) · [JSON](study-01.json) |
| 2. Do Powerful Politicians Really Cause Corporate Downsizing? | 21 | 21 | 9 | [Report](study-02.html) · [JSON](study-02.json) |
| 3. Reply: Do Powerful Politicians Really Cause Corporate Downsizing? | 16 | 16 | 16 | [Report](study-03.html) · [JSON](study-03.json) |
| 4. Market Integration Demand and the Growth of Firms | 7 | 7 | 7 | [Report](study-04.html) · [JSON](study-04.json) |
| 5. Do Government Audits Reduce Corruption? | 3 | 3 | 3 | [Report](study-05.html) · [JSON](study-05.json) |
| 6. Railroads of the Raj | 1 | 1 | 1 | [Report](study-06.html) · [JSON](study-06.json) |
| 7. The Arrival of Fast Internet and Employment in Africa | 3 | 2 | 2 | [Report](study-07.html) · [JSON](study-07.json) |
| 8. Copyrights and Creativity | 4 | 4 | 4 | [Report](study-08.html) · [JSON](study-08.json) |
| 9. Vertical Integration Supplier Behavior and Quality Upgrading | 7 | 7 | 5 | [Report](study-09.html) · [JSON](study-09.json) |
| 10. Leadership in Social Movements | 2 | 2 | 2 | [Report](study-10.html) · [JSON](study-10.json) |
| 11. Hall of Mirrors | 1 | 1 | 1 | [Report](study-11.html) · [JSON](study-11.json) |
| 12. The Rise and Fall of Local Elections in China | 4 | 4 | 4 | [Report](study-12.html) · [JSON](study-12.json) |
| 13. Ideology and Performance in Public Organizations | 3 | 3 | 3 | [Report](study-13.html) · [JSON](study-13.json) |
| 14. Reconciliation Narratives | 4 | 4 | 4 | [Report](study-14.html) · [JSON](study-14.json) |
| 15. Retail Pharmacies and Drug Diversion | 2 | 2 | 2 | [Report](study-15.html) · [JSON](study-15.json) |
| 16. The Gift of a Lifetime | 9 | 9 | 9 | [Report](study-16.html) · [JSON](study-16.json) |
| 17. Trading Nontradables | 3 | 2 | 2 | [Report](study-17.html) · [JSON](study-17.json) |
| 18. The Rise of Fiscal Capacity | 14 | 14 | 10 | [Report](study-18.html) · [JSON](study-18.json) |
| 19. The Real State | 26 | 0 | 0 | [Report](study-19.html) · [JSON](study-19.json) |
| 20. Credit Markets Property Rights and the Commons | 1 | 1 | 1 | [Report](study-20.html) · [JSON](study-20.json) |
| 21. Landmines and Spatial Development | 11 | 8 | 8 | [Report](study-21.html) · [JSON](study-21.json) |
| 22. Corruption as a Local Advantage | 90 | 57 | 57 | [Report](study-22.html) · [JSON](study-22.json) |

## Coverage and unavailable results

Every registered specification is listed, including missing-source and unsupported-design cases. They are excluded from the new-estimate scatterplots, not assigned zero effects or p-values.

```{=html}
<table border="1" class="dataframe">
  <thead>
    <tr style="text-align: right;">
      <th>Paper</th>
      <th>Specification</th>
      <th>Reason</th>
    </tr>
  </thead>
  <tbody>
    <tr>
      <td>The Arrival of Fast Internet and Employment in Africa</td>
      <td>S07-P003</td>
      <td>Imputation is unidentified with the configured first-stage controls.</td>
    </tr>
    <tr>
      <td>Trading Nontradables</td>
      <td>S17-P003</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P001</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P002</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P003</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P004</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P005</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P006</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P007</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P008</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P009</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P010</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P011</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P012</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P013</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P014</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P015</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P016</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P017</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P018</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P019</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P020</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P021</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P022</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P023</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P024</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P025</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>The Real State</td>
      <td>S19-P026</td>
      <td>No common absorbing-treatment sample with observed −1 and untreated counterfactual support.</td>
    </tr>
    <tr>
      <td>Landmines and Spatial Development</td>
      <td>S21-P001</td>
      <td>The included native sample lacks act_proj_d (World Bank project outcome) required by this figure.</td>
    </tr>
    <tr>
      <td>Landmines and Spatial Development</td>
      <td>S21-P002</td>
      <td>The included native sample lacks act_proj_d (World Bank project outcome) required by this figure.</td>
    </tr>
    <tr>
      <td>Landmines and Spatial Development</td>
      <td>S21-P005</td>
      <td>The included native sample lacks firstyear_sha_st (cancellation timing) required by this figure.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P001</td>
      <td>Required field-characteristic controls numwells, first_year, onsh and analog_depth are absent from the supplied bundle.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P002</td>
      <td>Required field-characteristic controls numwells, first_year, onsh and analog_depth are absent from the supplied bundle.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P003</td>
      <td>Required field-characteristic controls numwells, first_year, onsh and analog_depth are absent from the supplied bundle.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P004</td>
      <td>Required field-characteristic controls numwells, first_year, onsh and analog_depth are absent from the supplied bundle.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P005</td>
      <td>Required field-characteristic controls numwells, first_year, onsh and analog_depth are absent from the supplied bundle.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P006</td>
      <td>Required field-characteristic controls numwells, first_year, onsh and analog_depth are absent from the supplied bundle.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P007</td>
      <td>Required field-characteristic controls numwells, first_year, onsh and analog_depth are absent from the supplied bundle.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P008</td>
      <td>Required field-characteristic controls numwells, first_year, onsh and analog_depth are absent from the supplied bundle.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P009</td>
      <td>Required field-characteristic controls numwells, first_year, onsh and analog_depth are absent from the supplied bundle.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P010</td>
      <td>Required field-characteristic controls numwells, first_year, onsh and analog_depth are absent from the supplied bundle.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P011</td>
      <td>Required field-characteristic controls numwells, first_year, onsh and analog_depth are absent from the supplied bundle.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P012</td>
      <td>Required field-characteristic controls numwells, first_year, onsh and analog_depth are absent from the supplied bundle.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P013</td>
      <td>Required field-characteristic controls numwells, first_year, onsh and analog_depth are absent from the supplied bundle.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P014</td>
      <td>Required field-characteristic controls numwells, first_year, onsh and analog_depth are absent from the supplied bundle.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P015</td>
      <td>Required field-characteristic controls numwells, first_year, onsh and analog_depth are absent from the supplied bundle.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P016</td>
      <td>Required field-characteristic controls numwells, first_year, onsh and analog_depth are absent from the supplied bundle.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P017</td>
      <td>Required field-characteristic controls numwells, first_year, onsh and analog_depth are absent from the supplied bundle.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P018</td>
      <td>Required field-characteristic controls numwells, first_year, onsh and analog_depth are absent from the supplied bundle.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P021</td>
      <td>The archive supplies frozen estimates but not this specification’s input sample or required outcome/controls.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P031</td>
      <td>The archive supplies frozen estimates but not this specification’s input sample or required outcome/controls.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P057</td>
      <td>The archive supplies frozen estimates but not this specification’s input sample or required outcome/controls.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P059</td>
      <td>The archive supplies frozen estimates but not this specification’s input sample or required outcome/controls.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P061</td>
      <td>The archive supplies frozen estimates but not this specification’s input sample or required outcome/controls.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P063</td>
      <td>The archive supplies frozen estimates but not this specification’s input sample or required outcome/controls.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P066</td>
      <td>The archive supplies frozen estimates but not this specification’s input sample or required outcome/controls.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P068</td>
      <td>The archive supplies frozen estimates but not this specification’s input sample or required outcome/controls.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P070</td>
      <td>The archive supplies frozen estimates but not this specification’s input sample or required outcome/controls.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P072</td>
      <td>The archive supplies frozen estimates but not this specification’s input sample or required outcome/controls.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P086</td>
      <td>The archive supplies frozen estimates but not this specification’s input sample or required outcome/controls.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P087</td>
      <td>The archive supplies frozen estimates but not this specification’s input sample or required outcome/controls.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P088</td>
      <td>The archive supplies frozen estimates but not this specification’s input sample or required outcome/controls.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P089</td>
      <td>The archive supplies frozen estimates but not this specification’s input sample or required outcome/controls.</td>
    </tr>
    <tr>
      <td>Corruption as a Local Advantage</td>
      <td>S22-P090</td>
      <td>The archive supplies frozen estimates but not this specification’s input sample or required outcome/controls.</td>
    </tr>
  </tbody>
</table>
```
