Regional hotel performance and benchmarking in the Pearl River Delta: An input and output efficiency analysis

Purpose: This study models cost-efficiency against revenue for hotels in the Pearl River Delta (PRD) – in Guangzhou, Hong Kong and Macau – by considering regional differences and weight restrictions on revenue output. Design/methodology/approach: The authors modified and applied a context-dependent assurance region data envelopment analysis (CAR-DEA) model in assessing the performance of 41 hotels in the PRD. The model considers the relationships among output variables and sets the revenue composition of the hotels as weight restrictions in accounting for the relative importance of different revenue sources. Findings: When assessing the 41 hotels all together, those in Guangzhou outperformed the hotels in other two cities by showing better pure technical efficiency (PTE), while those in Macau had the best scale efficiency (SE). When the assurance region (AR) restriction was imposed, the hotels in Macau outperformed those in the other two cities by showing better SE. When considering regional differences, the Macau hotels ranked first in terms of both the average efficiency score and the overall ranking. All the sample hotels in Guangzhou and half of the sample hotels in Hong Kong and Macau exhibited increasing, constant and decreasing returns to scale. Research limitations/implications: The research results are limited by data quality and the variables included in the models. Practical implications: The study helps hotel practitioners in the PRD better assess their cost-efficiency performance by considering regional differences and operational parameters so as to strategically improve their performance. Originality/value: This study improves upon previous hotel efficiency studies by considering the influence of different operational parameters across different localities. It can be extended to examine the performance of different calibers of hotels, restaurants or tourism entities located in various localities and possessing different operational characteristics.