Since 2014, the armed conflict in Ukraine's eastern provinces (oblasts) of Donetsk and Luhansk has dealt a heavy blow to people's lives. The conflict has magnified the long-standing problems and created new ones. This study shows that scaling up efforts in the government-controlled areas (GCAs) of Donbas is desirable despite the subdued productivity in the region. This study recommends a decision tree approach to programming recovery in Donbas. Given the looming uncertainties and scenario-sensitivity of optimal policies, the recovery strategy should distinguish contingent policies from no-regret policies. Contingent policies change between the status quo and the reintegration scenarios, and they include interventions to mitigate conflict-related risks, risk-related transfers to address skill-shortages in GCAs, and investments for a contingent infrastructure strategy. By comparison, no-regret policies are desirable regardless of the conflict dynamics. They include the reforms to eliminate regulatory burdens and corruption
policies to open up the housing market
investments to modernize education for jobs and target low-hanging fruits in infrastructure
and efforts to produce better data to address knowledge gaps.