by Aaron Williams Lam
Abstract
While standard diffusion frameworks predict that liberalizing reforms generate positive spillovers in geographically proximate regions, borders are not frictionless membranes. Even observable connectivity may fail to translate into cross-border economic co-movement when institutional frictions govern whether interaction activates transmission. Using Uzbekistan’s 2017 exchange-rate unification as a natural experiment, this paper examines whether reform-driven spillovers were attenuated by the institutional regime of neighboring Turkmenistan. Using annual VIIRS night-time lights across 19 matched border crossing pairs on the Uzbek-Turkmen, Uzbek-Kyrgyz, and Uzbek-Tajik frontiers, I document a large and frontier-specific divergence following the reform – a pattern absent on Kyrgyz and Tajik frontiers sharing identical Uzbek-side conditions. Two main tests rule out observable explanations. First, Turkmen attenuation persists even when restricting to the highest-interface crossings, where market-mediated transmission is most plausible. Second, a benchmark residual test shows Turkmen crossings under-absorbed significantly relative to connectivity-based predictions during the reform period, with partial normalization after 2022 coinciding with bilateral border policy relaxation. The timing of this reversal, which tracks policy rather than any change in geography or trade composition, is difficult to reconcile with non-institutional explanations. The results suggest that geographic proximity and observable infrastructure are necessary but insufficient for cross-border spillovers: institutional permeability, itself partly a bilateral policy parameter rather than a fixed property of the neighboring regime, determines whether transmission mechanisms activate.
Erica Field, faculty advisor
Michelle Connolly, faculty advisor
