Do Neighbors Shape the Sticker Price? Spatial Competition and State Funding in Cost of Attendance Reporting Over Time

Federal law requires U.S. postsecondary institutions to publish a Cost of Attendance (COA) that serves as the legal ceiling on a student's federal financial aid eligibility, yet off-campus living costs are estimated by institutions at their own discretion, with no standardized methodology or external audit. Using a 14-year panel of 2,311 four-year institutions (2010-2023) and externally benchmarked living costs constructed from HUD, USDA, BLS, KFF, and BEA data, we find that roughly half of institution-years report living costs below the benchmark, and this pattern persists across the full 14-year observation window rather than concentrating in any particular year. Spatial analysis reveals significant positive spatial autocorrelation in reporting gaps, and spatial panel models confirm persistent spatial dependence consistent with strategic interaction among geographically proximate institutions, robust to alternative neighborhood definitions, benchmark specifications, and sample restrictions. Public institutions exhibit spatial dependence nearly twice as strong as private institutions, consistent with more direct competition for in-state students within shared geographic markets. A Bartik shift-share instrumental variables design identifies a causal channel through which state fiscal pressure compresses reported living costs at public institutions ($362 per $1,000 funding decline), while a border discontinuity finds no corresponding cross-sectional jump at state boundaries, indicating that the funding-reporting link operates through within-state temporal shocks rather than persistent cross-state level differences. Together, these findings identify discretion in self-reporting, exercised under competitive pressure and state fiscal stress, as a systematic distortion in the cost information that flows into the federal aid system, with implications for federal standardization of living cost methodology.