Although unreliable electricity is a substantial burden on producers in India, the long-run effects of electricity shortages have received little attention. In this study, I study the long-run output and welfare costs of electricity shortages by examining the dynamic investment responses of producers. Specifically, I examine two dynamic margins of adjustment: (1) investment in productive capital, and (2) adoption of generators as insurance against these shocks. Using state-level power-sector data and plant-level manufacturing data from India, I document descriptive evidence that electricity shortage shocks depress investment by reducing the utilization of inflexible inputs. Furthermore, while plants that adopt generators mitigate the investment effects of energy shocks, they invest less in productive capital. I rationalize these patterns using a partial equilibrium model of investment dynamics augmented with electricity shortage shocks, costly generator adoption, and interactions between both dynamic margins. I estimate my model for two states - Maharashtra and Punjab - and find that electricity shortages are responsible for long-run value-added losses of 52% and 64% along with producer surplus losses of 44% and 57%. The dynamic losses are 2.5 to 3.5 times as large as the static value-added losses of 14% and 24%. However, these long-run losses are mostly ameliorated under a counterfactual dynamic pricing policy that adjusts electricity prices to prevent shortages. With dynamic prices, the long-run losses are reduced to just 5% and 13% for value added, and 4% and 11% for producer surplus. Finally, generator adoption undoes the ``wait-and-see'' effects of shortage uncertainty, though the former distorts the long-run firm size distribution.