Replication data for “The Long-term Effects of War Exposure on Civic Engagement,” published September 2021, and “Corrections for Barceló, The Long-term Effects of War Exposure on Civic Engagement,” published online July 2023 and February 2024, in the Proceedings of the National Academy of Sciences. The corrections purport to address several issues in a 2021 publication, which we raised with both PNAS and the author.

As we describe in our replication study, the corrections are themselves flawed and continue to lead to misleading findings. Barceló commits three major errors that render his research design inappropriate and his conclusions problematic. First, Barceló again commits a critical coding error that again leads him to attest incorrectly that he is clustering standard errors (SEs) at the provincial level, when in fact he is only using robust SEs, leading to false assertions of statistical significance and instrument strength. Second, Barceló miscodes his measure of South Vietnam. Third, Barceló overfits his model by including a linear transformation of his instrumental variable as a control. When these mistakes are addressed the provocative findings no longer hold.