RVRS Efficiency and Scale Adjustment Decisions of Tourism Enterprises Under COVID-19

The sudden and recurrent nature of COVID-19 challenges the convexity assumption in data envelopment analysis (DEA) and may bias efficiency estimates and crisis response strategies. To address this bias, this study first introduces the non-convex regular variable returns to scale (RVRS) frontier to analyze Chinese listed tourism enterprises efficiency (2017–2023). We evaluate the efficiency evolution and explore how scale adjustment decisions can mitigate the losses caused by the pandemic. Results indicate that while COVID-19 significantly diminished tourism enterprise efficiency, optimal scale adjustments could reduce losses by up to 69.17%. Heterogeneity analysis reveals that large enterprises and travel agency benefited most from scale adjustment. Notably, approximately one-third of enterprises in 2021 were still operating under increasing returns to scale, suggesting that optimal responses to the pandemic were not uniformly contractionary. This study contributes by demonstrating how the RVRS framework improves measurement accuracy and supports more effective, non-linear crisis management decisions.