Sentenced without evidence? An honest diff-in-diff evaluation of Italy’s anti-depopulation national strategy for inner areas

Abstract

This study investigates the medium-term impact of the National Strategy for Inner Areas (SNAI) on internal emigration dynamics, aiming to assess the policy effectiveness in retaining residents in targeted inner areas. Leveraging Italian Census data (2012-2023 period) and OpenCoesione’s 2025 release of SNAI’s geocoded project records, a Doubly Robust Difference-in-Differences design, complemented with partial identification, was used to provide robust causal estimates under incremental violations of the identifying assumption. Results suggest that the policy did not produce significant effects on municipalities treated between 2017 and 2019. However, even under modest violations of the parallel trend assumption, causal estimates become fragile and inconclusive. This fragility persists even after accounting for the confounding effect of the COVID-19 pandemic, signalling a structural difficulty in achieving identification via Difference-in-Differences with available data. This calls for caution in attributing conclusive demographic effects to the policy in both the short and medium term.