While previous research on policy implementation focused on the constraints and incentives defined by the mandate of the policy itself, this study highlights the additional constraints imposed by externalities arising from other conflicting mandates. The pursuit of each policy goal is constrained by how it affects other mandates with comparable importance, leading to the emergence of balanced efforts and the avoidance of over-compliance. By aggregating location data from 1.1 billion mobile devices in China, this study gauges the intensity of the lockdown policies in cities during China’s 2020 COVID-19 epidemic. The study employs event studies and a staggered difference-in-differences design to exploit the natural experiment of the end of a conflicting mandate against “Zero COVID”, the poverty elimination campaign. I found that cities which needed to balance poverty elimination and pandemic control succeeded in both goals with significantly lower lockdown costs. A reduction of 40 percent in the severity of their COVID-19 lockdown measures is observed, compared to a counterfactual scenario where only the Zero COVID mandate was present.