This project examines how information frictions shape the effectiveness of Michigan’s Tuition Incentive Program (TIP), historically the largest state aid program in Michigan. TIP provides a first-dollar scholarship to students from low-income backgrounds, identified through Medicaid participation, and is designed to reduce financial barriers to college enrollment. Despite these features, take-up remains far below eligibility, suggesting that information barriers substantially limit its impact. The study uses rich linked administrative data covering the universe of Michigan public high school graduates to analyze how TIP affects college enrollment, and graduation. This work adopts a regression discontinuity design to estimate causal effects, and develops a structural model of Bayesian learning and enrollment choice to quantify how imperfect information moderates these effects. This framework allows for counterfactual simulations that assess how enrollment and take-up would change if informational barriers were removed. This project contributes to the literature on financial aid by explicitly modeling the interaction between aid design and information constraints, highlighting how even generous programs can fail to achieve their objectives when potential recipients do not perceive or process key incentives.