Network externalities can generate chicken-and-egg dynamics that result in socially inefficient outcomes. I study whether direct government provision of network infrastructure yields greater welfare than alternative policies such as subsidies or monopoly rights. I do so in the context of Québec from 2017 to 2023. I first exploit the network’s concentric expansion and implement a difference-in-differences design, finding a two-fold increase in EV adoption in newly connected cities. I then develop a structural model in which consumers choose EV models and subsequently derive utility from charging, thereby capturing the charging network’s topography. Counterfactual simulations show that direct public provision generates higher welfare than subsidies and private monopoly provision, by ensuring broader coverage, greater adoption, and more cost-effective outcomes.