A Comment on “The Heterogeneous Tax Pass-Through Under Different Vertical Relationships” by Bajo-Buenestado and Borrella-Mas

Abstract

We revisit the heterogeneous tax pass-through analysis in Borrella-Mas (2022a) (B&B). We find that B&B’s estimating equation is misspecified, as it excludes a necessary lower-order interaction between vertical integration and the post-treatment period. Including this term reveals a significant structural price trend for vertically integrated firms in the post-reform period, whose origin is uncertain. A spatial placebo test confirms that this structural shift is present even in never-treated provinces, indicating that it is not a local, province-level pass-through of the regional tax. Without this control, the original specification may conflate this shift with the tax effect. Once the interaction is included, the differential pass-through estimate becomes null for diesel, while for gasoline a significant effect persists—albeit at a reduced magnitude—supporting B&B’s central claim that market structure matters for tax pass-through. These findings highlight the importance of explicitly modelling group-specific time trends when estimating heterogeneous treatment effects in difference-in-differences designs.