When Arrivals Recover but Paid Nights Do Not: Measuring Tourism Recovery under Incomplete Accommodation Coverage in Saint Lucia

Full abstract

Abstract

Visitor arrivals, paid accommodation nights, domestic production and income accruing to residents measure different aspects of tourism recovery. This paper reconciles Saint Lucia's published record and develops a decision protocol for cases in which those measures diverge. It assembles 132 monthly arrival observations for 2015–2025, a common-vintage annual visitor series for 2011–2025, and separately classified incomplete 2026 evidence. In 2025, stay-over arrivals were 0.7% above 2019, while published paid nights were approximately 13.6% lower. Real accommodation-and-food-services gross value added was 9.7% higher, providing contrary evidence to a simple equation of fewer nights with lower domestic production, although industry GVA is not tourism direct value added. An exact symmetric decomposition allocates the approximately 380,000-night decline to an arrival contribution of +18,000 and a published-night-intensity contribution of −398,000; the latter does not identify shorter stays. The paper proves conditional accounting and decision results, executes coverage and survey-selection experiments, and enumerates a hypothetical measurement-cost frontier. Compatible expenditure observations can eliminate separate coverage and yield uncertainty from an income comparison, changing which measurements are worth buying. Current public evidence does not identify a national tourism resident-income growth rate. The contribution is an audited Saint Lucia application and an executable measurement protocol, not a new decomposition theorem or a measured retention coefficient. A staged programme assigns institutional owners, release requirements and review rules while preserving an explicit “unresolved” classification.