Should broadband subsidies make service cheaper where it exists, or build it where it does not? I evaluate the $14.2 billion Affordable Connectivity Program (ACP) and the $42.45 billion Broadband Equity, Access, and Deployment (BEAD) program within a unified welfare framework. I combine continuous-treatment event studies with a tract-level model of demand, pricing, and entry estimated on 170,419 broadband products across 66,454 census tracts. To make nationwide entry estimation feasible, I bound product-level fixed costs using moment inequalities that avoid solving the full equilibrium game. ACP reduced posted prices by about 10 percent at its peak and shifted demand toward faster plans, while BEAD’s early effects follow its construction timeline. Over a five-year horizon, a 75 percent BEAD fixed-cost subsidy generates $6.80–$7.22 in welfare gains per subsidy dollar, while ACP generates welfare gains roughly equal to program costs. The programs relax different binding constraints, implying a clear sequencing rule: build where deployment is unprofitable, then subsidize affordability where price remains the binding constraint.