We study the causal effects of global supply chain disruptions by constructing a new index of real-time port congestion using Automatic Identification System data from container ships and a spatial clustering algorithm. We develop a model with search frictions between producers and retailers that links upstream production slack to downstream supply shortages and captures output and price responses to supply chain shocks. The co-movements of output, prices, spare capacity, and market tightness provide novel identification restrictions. We find that demand and supply shocks drove US disinflation in 2020, while the inflation surge in 2021 was driven mainly by supply chain shocks.