Fixed Effects Don't Always Fix Post-Treatment Bias in Camp Legacy Research

🧭 What Was Reassessed

Pepinsky, Goodman, and Ziller (2024, American Political Science Review, PGZ) reevaluate a recent study on the long-term consequences of concentration camps in Germany and conclude that controlling for contemporary (post-treatment) state heterogeneity yields unbiased estimates of camps' effects on present-day outgroup intolerance.

🔍 How the Reanalysis Approaches the Problem

• Reexamines PGZ's empirical strategy and the role assigned to regional fixed effects.
• Identifies two core problems with PGZ's approach: (a) a mischaracterization of what regional fixed effects capture and (b) reliance on two unrealistic assumptions that can be avoided.
• Advocates for using pre-treatment state fixed effects as a corrective modeling choice.
• Uses targeted simulations to compare camp proximity with spatially correlated noise in this empirical context.

📈 Key Findings

• PGZ's empirical strategy depends on a mistaken interpretation of regional fixed effects and on two avoidable, unrealistic assumptions.
• Replacing post-treatment state controls with pre-treatment state fixed effects removes the need for those assumptions.
• When regional fixed effects are incorporated correctly, the original article's substantive results remain essentially unchanged.
• Simulation evidence shows that, in this specific study, proximity to camps consistently outperforms spatially correlated noise as an explanatory factor.

🤔 Why This Matters

This note advances the methodological debate in legacy studies by clarifying which fixed-effect choices are defensible and by demonstrating that careful specification—not blanket use of contemporary controls—better preserves causal interpretation. The discussion highlights practical modeling choices for researchers working on long-term historical legacies and similar settings where treatment timing and regional heterogeneity interact.