الگوبرداری صنعت گردشگری اقیانوس آرام آسیا : ترکیب بیزی تحلیل پوششی داده ها(DEA) و مرز تصادفی

This study measures and compares the efficiency of leading tour operator and hotel companies across several Asia Pacific countries. We use an innovative methodology that is based on combining the stochastic frontier and data envelopment analysis in a Bayes framework. We show from the results that Australia, Singapore and South Korea are the most efficient in both their tour operator and hotel industries. We further show that international hotels in the region have a slightly higher efficiency than local hotels. We provide a listing of the most efficient tour operators and hotels in each country and discuss the implications of our findings.

This paper aims to offer more accurate and fresh insights into the level of performance of major tourism operations in the Asia Pacific region. It is well established that tourism is an important economic driver and major source of foreign exchange earnings for many Asia Pacific countries. It also contributes significantly to the employment in these countries, helps in the creation of new businesses, and plays an important role in the revival of regional centers and in achieving environmental sustainability (Wang, 2010). In Australia, for example, tourism contributed $38.9 billion to Australia’s Gross Domestic Product (GDP) in 2006–07, representing around 3.7 per cent of the Australian economy in the last ten years. The industry holds a high share in terms of Australian employment (4.7 per cent, or 482,800 jobs) and Australian exports (10.5 per cent, or $22.4 billion). The industry has also important impact in New Zealand; it contributes $18.6 billion to the economy each year—9% of New Zealand’s gross domestic product. It also makes a large contribution in Thailand, Japan and South Korea and ranks among the three most important industries in Hong Kong, Malaysia, the Philippines, Singapore and Indonesia (IBISWorld, 2009). The tourism industry in most of these countries was generally growing rapidly over the last ten years, until recent negative trends affected its growth. For example, the safety fears changed the environment for travel and tourism and threatened the predicted growth of tourism both globally and locally (Travel and Tourism-Australia, 2009). The recent economic crisis also affected business and travel activities, and resulted in less tourism spending (Global Travel and Tourism, 2009). Some measures that are currently being taken by Asian Pacific governments to assist tourism operations include the improvement of tourism competitiveness, facilitation of travel and investment activities, diversification of the economy, and the promotion of innovative launching initiatives (IBISWorld, 2009). We focus here on the efficiency analysis of tourism operations from both the tour operator and hotel industries. Both these industries play an essential role in the tourism industry of the Asia Pacific region. They also present an excellent context for efficiency analysis as they operate in a high competitive environment. The hotel and tour operator industries have both recently experienced a sharp decline in revenues as a result of the economic crisis and drop in international travel. Hotel occupancy rates, particularly in higher-end hotels declined significantly in the second half of 2008. The competition in the tour operator industry also continues to increase with the fluctuation in flight prices, and the political instabilities in some counties like Thailand and India (Global Travel and Tourism, 2009). The present study introduces a sample that allows for a cross-country comparison. The existing literature is replete with studies that are mostly limited to one single country (e.g. Barros et al., 2010, Barros et al., 2009, Chiang et al., 2004, Hwang and Chang, 2003 and Köksal and Aksu, 2007). All these studies have important merits, but the focus on a single country limits the benchmarking comparison (Assaf & Dwyer, in press). It is more important for governments and tourism operators to evaluate how their hotel industry is performing at the regional level, or at least against their major competitors (Blanke & Chiesa, 2009). The methodology used in the study is also unique and innovates on related studies in the area. For the first time we combine the strength of both the data envelopment analysis (DEA) and stochastic frontier (SF) models in one methodology. Specifically, we use the DEA efficiency measures as priors of efficiency in the stochastic frontier model. These priors are then used to obtain posterior estimates of efficiency using the Bayes’ theorem. We provide below more details about the methodology. The paper proceeds as follows: Section 2 presents the literature review; Section 3 discusses the methodology; Section 4 describes the data; Section 5 presents the results and Section 6 presents the discussions and concluding remarks.

The main objective of this study was to provide a benchmarking analysis of the hotel and tour operator industries in the Asia Pacific region. Both these industries are extremely important for tourism in the Asia Pacific region. Destination managers have an interest in benchmarking as it is strongly related to strategy development (Bosetti et al., 2006, Kozak and Rimmington, 1999 and Peypoch and Sbai, 2011). Our results are useful for the tourism industry, at least in the countries we analyzed. Apart from benchmarking the different countries, we also reported the most efficient hotels and tour operators in each country. These are essential information for competitors as well as for investors in the industry. It is possible that several other factors have also contributed to the efficiency difference between the countries under analysis. Australia, for instance, which has the highest efficient hotel industry and tour operator industry, ranks consistently among the most competitive tourism destinations in the world, and is also one of the highest ranked in the Asia Pacific region (Blanke & Chiesa, 2009). The hotel industry in this country benefits from excellent air transport and tourism infrastructure and a high number of significant tourism attractions. The same applies to other efficient countries (Singapore, South Korea and Malaysia). Generally, all these countries have well established tourism industries and are also consistently ranked among the top destinations in the world in terms of their tourism competitive index (Blanke & Chiesa, 2009). They also attract annually a large share of international tourism arrivals and receipts. On the other hand, several factors could have contributed to the low efficiency in the hotel and tour operator industries in countries such as India, China, and Taiwan. In China, for example, the quick increase in hotel construction has resulted in supply outstripping demand even before the financial crisis. The occupancy rates have been very low, particularly in relation to 4- and 5-star hotels, where affluent, high spending, foreign tourists traditionally stay. The occupancy rate was also low in India over the last few years. The financial crisis and the terrorism threats, particularly in the wake of the terrorist attacks in Mumbai in Taj and Trident hotels in November 2008 have had a major impact on the country (Travel and Tourism-India, 2008). In 2008, for example, the occupancy rate was 70%, 5% lower than 2007. Room rates were also particularly low, and despite price cuts and attractive tariffs in premium and budget hotels, people were wary of taking holidays, due to the tourist attacks in the hotels in Mumbai. In summary, it is possible to relate the efficiency of hotel and tour operator industries to several other factors. A major limitation of this study is the small sample size in some countries. We have also limited number of years. Therefore, the results might have been influenced by recent trends such the current financial crisis. Heterogeneity can also be a problem as we compared hotels and tour operators across various countries. Due to data limitation, we obtained more hotels and tour operators in some countries than others. Though we did not estimate the efficiency for every sample separately, the comparison should still be taken with caution.