Replication code: Design-Based Variance Estimation for Modern Heterogeneity-Robust Difference-in-Differences Estimators

Fuente: Zenodo
Salvato in:
Dettagli Bibliografici
Autore principale: Gerber, Isaac
Natura: Recurso digital
Pubblicazione: Zenodo 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866901726773116928
author Gerber, Isaac
author_facet Gerber, Isaac
contents <p>This repository contains the simulation and empirical-illustration code that produces the numerical results in Gerber (2026), <em>Design-Based Variance Estimation for Modern Heterogeneity-Robust Difference-in-Differences Estimators</em> (<a href="https://arxiv.org/abs/2605.04124">arXiv:2605.04124</a>).</p><p>The pipeline runs a 66,000-replication Monte Carlo over four scenarios (unconditional parallel trends with complex survey design, informative sampling with heterogeneous treatment effects, repeated cross-sections, and conditional parallel trends) and an empirical illustration on NHANES data. Each replication evaluates three inference paths: HC1 standard errors, weighted point estimate with PSU-level clustering, and full design-based standard errors with strata, PSU, and finite-population corrections.</p><p>Companion package: <a href="https://doi.org/10.5281/zenodo.19803705">diff-diff v3.3.2</a>.</p>
format Recurso digital
id zenodo_https___doi_org_10_5281_zenodo_20097361
institution Zenodo
language
publishDate 2026
publisher Zenodo
record_format zenodo
spellingShingle Replication code: Design-Based Variance Estimation for Modern Heterogeneity-Robust Difference-in-Differences Estimators
Gerber, Isaac
difference-in-differences
survey statistics
design-based variance
influence functions
Binder linearization
Taylor series linearization
complex survey design
Monte Carlo
replication
Python
<p>This repository contains the simulation and empirical-illustration code that produces the numerical results in Gerber (2026), <em>Design-Based Variance Estimation for Modern Heterogeneity-Robust Difference-in-Differences Estimators</em> (<a href="https://arxiv.org/abs/2605.04124">arXiv:2605.04124</a>).</p><p>The pipeline runs a 66,000-replication Monte Carlo over four scenarios (unconditional parallel trends with complex survey design, informative sampling with heterogeneous treatment effects, repeated cross-sections, and conditional parallel trends) and an empirical illustration on NHANES data. Each replication evaluates three inference paths: HC1 standard errors, weighted point estimate with PSU-level clustering, and full design-based standard errors with strata, PSU, and finite-population corrections.</p><p>Companion package: <a href="https://doi.org/10.5281/zenodo.19803705">diff-diff v3.3.2</a>.</p>
title Replication code: Design-Based Variance Estimation for Modern Heterogeneity-Robust Difference-in-Differences Estimators
topic difference-in-differences
survey statistics
design-based variance
influence functions
Binder linearization
Taylor series linearization
complex survey design
Monte Carlo
replication
Python
url https://doi.org/10.5281/zenodo.20097361