difference-in-differences

This study addresses the vulnerability of conventional difference-in-differences (DID) estimators to bias under post-treatment shocks, which arises from their reliance on the parallel trends assumption. To overcome this limitation, the authors propose a novel inference approach that dispenses with this assumption by constructing a DID-specific predictor based on pre-treatment outcome dynamics and embedding it within a conformal inference framework. This method explicitly models potential post-treatment shocks and leverages pre-treatment information to impose identification constraints, thereby enabling robust causal inference even when parallel trends fail to hold. The proposed procedure substantially enhances the reliability and applicability of DID estimates in settings characterized by non-parallel trends.

Standard parallel-trends tests in Difference-in-Differences (DID) estimation merely fail to reject the null of no pre-treatment differences, but cannot actively confirm the assumption’s validity, thereby limiting causal credibility. This paper proposes the first equivalence-testing framework for DID pre-trend assessment: the null hypothesis posits *substantive* pre-treatment trend divergence between treatment and control groups; rejection of this null provides direct statistical support for the absence of meaningful pre-trends. We construct an asymptotically t-distributed test statistic and associated confidence sets within a two-way fixed-effects setting, ensuring theoretical rigor and empirical feasibility. The method naturally accommodates staggered adoption designs. Empirically, it detects subtle yet systematic pre-trend deviations missed by conventional tests—enhancing robustness, reproducibility, and causal interpretability of estimated treatment effects.

This paper addresses the disconnect between structural nested mean models (SNMMs) and dynamic difference-in-differences (DiD) in estimating time-varying treatment effects. We propose a novel SNMM framework grounded in the parallel trends assumption—departing from the conventional no-unmeasured-confounding assumption. We establish, for the first time, that SNMMs achieve nonparametric identification under parallel trends alone. The framework unifies estimation of dynamic treatment effects, sustained-intervention effects, direct effect decomposition, and optimal dynamic treatment regimes. Additionally, we develop a sensitivity analysis method to assess robustness when parallel trends are violated. Integrating dynamic causal inference with sequential decision-making modeling, our approach is validated through empirical applications—including Medicaid expansion, flood insurance adoption, and temperature impacts on crop yields—demonstrating its validity and robustness in real-world policy and environmental settings.

This paper addresses point estimation and uncertainty quantification for treatment effect paths—such as dynamic effects and event-study designs—in policy evaluation. To overcome the looseness of conventional uniform confidence bands, which ignore correlations among path estimators, we propose two data-driven feasible bound methods. Our novel framework jointly enforces average-effect coverage guarantees and path smoothness constraints, integrating post-selection inference, smooth regularization, Monte Carlo simulation, and robust point estimation. The resulting confidence bands are substantially narrower while maintaining valid coverage, especially under high estimator correlation; our point estimator also demonstrates superior performance across diverse simulation settings. The key contribution is the first systematic incorporation of smoothness priors into path inference, thereby unifying statistical rigor with economic interpretability.

The two-way fixed effects (TWFE) estimator is widely used for causal inference in political science but suffers from sensitivity to heterogeneous treatment effects (HTE) and the parallel trends (PT) assumption, undermining result reliability. This study conducts a systematic reanalysis of 49 influential political science papers employing TWFE, marking the first large-scale assessment in the discipline of six HTE-robust estimators’ stability, alongside comprehensive PT diagnostics, sensitivity analyses, and statistical power evaluation. Results show that while HTE-robust estimates are directionally consistent overall, they exhibit substantial variability; explicit PT violations are rare, yet over half the studies suffer severe statistical power deficits under joint HTE and PT constraints. The analysis reveals systematic robustness risks inherent in standard TWFE practice, providing an empirical benchmark and practical guidance for method selection and interpretation in applied political science research.

This study investigates the validity conditions for identifying causal effects using changes in treatment variables rather than their levels, and examines the relationship between this approach and conventional methods. By developing two non-nested structural models and integrating structural causal modeling with difference-in-differences and two-period fixed-effects regression, the authors theoretically demonstrate that strategies based on treatment changes and treatment levels are generally non-nested but become equivalent under specific conditions. They propose a corresponding overidentification test to assess these conditions. Simulation evidence confirms the favorable finite-sample performance of the proposed method, and an empirical application to cigarette demand estimation supports its practical validity. The work clarifies the fundamental distinctions and connections between these two causal identification strategies, thereby extending the methodological foundations of causal inference.

The parallel trends assumption in panel data causal inference is often unverifiable and fragile. Method: This paper proposes a novel “neighbor comparison group” approach that dynamically selects control units with similar outcome trajectories to the treated group during the pre-treatment period—replacing reliance on global parallel trends with local trajectory matching. It establishes a time-homogeneous analytical framework integrating multi-period treatment timing and pre-treatment matching. Contribution/Results: Under non-nested identification assumptions—including local parallel trends and conditional independence—the method consistently estimates the average treatment effect on the treated (ATT). Theoretical analysis demonstrates robustness of the identification strategy. Empirical applications show that, relative to conventional difference-in-differences, this approach substantially improves the robustness and credibility of policy effect estimates.

Conventional conditional average treatment effect (CATE)-based methods struggle to identify treatment effect heterogeneity when effect modifiers are unobserved or subject to severe measurement error. Method: We propose a variance-comparison inference framework that does not require fully observed covariates. Leveraging the variance difference of potential outcomes as a novel causal identification anchor, we construct a doubly robust and asymptotically linear nonparametric estimator, integrating causal machine learning with a variance-sensitive testing paradigm. Contribution/Results: We establish theoretical consistency and asymptotic normality under weak regularity conditions. In a reanalysis of a randomized controlled trial, our method detects statistically significant heterogeneity in therapeutic hypothermia efficacy. The approach demonstrates robustness across diverse data-generating mechanisms and overcomes the strong reliance of CATE-based methods on high-fidelity covariate measurement.

Traditional causal inference methods struggle to capture how interventions affect the dynamic evolution of time series, such as persistence and transition patterns. This work extends the potential outcomes framework to path space and introduces the Dynamic Average Treatment Effect (DATE) to characterize how causal effects evolve over time. It establishes the first dynamic causal inference framework in path space, develops a dynamic inverse probability weighting estimator suitable for observational data, and reveals that, under sparse treatment regimes, the conditional mean trajectory admits a linear state-space structure. Simulations demonstrate that the proposed method accurately captures dynamic effects that static approaches systematically misestimate. In an empirical application to COVID-19 lockdown policies, the method successfully estimates and decomposes the treatment effects over time.

This study addresses the limitations of traditional control-based causal inference methods—such as matching and difference-in-differences—in settings characterized by pervasive or structurally ambiguous spillover effects, where reliance on uncontaminated control units impedes accurate identification of both average direct and spillover effects. Within the potential outcomes framework, this work provides the first systematic comparison between control-based and prediction-based counterfactual approaches—including interrupted time series and machine learning control—in terms of their identification capabilities. Through simulation and empirical analyses, the authors demonstrate that in environments with widespread interference, prediction-based methods can more reliably estimate certain causal parameters over short horizons, circumventing the stringent assumption of unperturbed units and thereby offering a promising alternative for causal inference under complex interference.