To control the transmission of the coronavirus disease (COVID-19), Brazilian local governments have adopted partial lockdown measures in economic sectors, thereby triggering transmission shocks along input-output supply chains. The national internal market and territorial disparities favor the formation of subnational production networks within borders, thus increasing the potential effects of lockdown measures on regional integration production networks. Therefore, this study makes hypothetical simulations of COVID-19 mitigation policy decisions to understand the regional impacts on integration in supply chains, considering both domestic and global value chains. The generalized hypothetical extraction method is applied to a Brazilian interregional input-output model with 68 industries and 27 regions, imputing partial removals on intermediate consumption and final demand. The results suggest that richer subnational areas, mainly São Paulo and Rio de Janeiro, are proportionally more impacted by COVID-19 trade shocks. However, the poorer peripheries are doubly affected, either by the foreign shock, which would damage their economic structure, or by the retraction of the subnational demand from core states. The findings highlight that economic shocks are spatially distributed through different industrial structures, thus stressing the need to avoid "one size fits all" regional policies to mitigate the potential negative effects on exposed regions. KEYWORDS Regional effects. Hypothetical extraction method. Domestic value chains. COVID-19 outbreak. JEL CLASSIFICATION R12. R10.

The rise of global sourcing implies a heterogeneous relationship between buyers and suppliers regarding the liberalisation scenarios in emerging countries. This paper analyses the effect of regional trade agreements (RTAs) on participation in the Global Value Chains of Latin American countries between 1995 and 2015. We combine the framework of gravity equations with the trade in value-added, applying a Pseudo Poisson Maximum Likelihood (PPML) estimator with panel data and fixed effects to deal with endogeneity and heteroscedasticity. Heterogeneous estimations show that the deepest RTAs reinforce the position of lower technology-industry suppliers, driven by an extra-regional strategy of Latin American trade policy. The geography of value chains has little effect on industrial upgrading in the region, reducing the development potential. The study concludes that the region's trade policy could reduce the dependent relationship between distant partners and pay more attention to creating shorter value chains as a strategy to generate local capacities to gain competitiveness in value chains.