What sales-tax receipts can (and cannot) tell us about tourism: evidence from Tennessee's counties

County local-option sales-tax collections are available monthly, whereas official visitor-spending estimates arrive annually with a long lag. We test whether the collections indicate county tourism activity, emphasizing tests that could show them uninformative. Using a balanced panel of Tennessee's 95 counties over 80 months (January 2020–August 2026), validated against published county visitor spending for 2024 and 2025, we find that annual collections correlate with visitor spending at r = 0.91 in logs, but state sales tax does so equally (r = 0.915) and local collections add nothing once it is held constant. Year-over-year growth correlates at r = 0.37, falling to 0.16–0.20 after trimming outliers. Collections are spatially dependent (Moran's I = 0.18), yet few clusters survive false-discovery-rate correction, and neither a fixed-effects model nor a seasonal naive forecast dominates. The series suits the level, ranking, and seasonality of taxable activity, not visitor spending.