Productivity Growth of Taiwanese International Tourist Hotels in a Metafrontier Framework

Abstract

Due to the vigorous development of the tourism industry in Taiwan, the number of international tourist hotels has been on an upward trend. To survive in an increasingly competitive environment, it is important to seek out improvement in performance. This study measures the performance of fifty-four international tourist hotels in Taiwan from 2008 to 2011 by using the metafrontier Malmquist productivity index. This index satisfies the requirement of circularity, is immune to linear programming infeasibility, overcomes the problem of base period dependency, and considers the heterogeneity among hotels. To investigate sources of productivity change, this index can be further decomposed into within-group efficiency change, within-group technical change, and technical leadership change. The empirical results indicate that the use of different technologies by hotels affects the productivity change. Both within-group efficiency and technical changes are the main factors of productivity change. Chain hotels are technology leaders, and independent hotels are followers. By identifying each competitor’s productivity change, operators can refer to appropriate best-practice hotels to improve their operations.

Keywords

data envelopment analysis metafrontier Malmquist productivity index international tourist hotels

In the twenty-first century, the growth of the global service industry has been the fastest (Fatma and Timothy 2005), and within this, the tourism industry has undergone rapid expansion, playing an important role in global economic development (Hung, Shang, and Wang 2010). In Taiwan, governmental policy, such as a two-day weekend policy and the deregulation of overseas tourism from mainland China, has further promoted tourism activities (Hu et al. 2010). This expansion of tourism has increased the demand for hotels and the competitive pressure in the hotel industry. The total number of international tourist hotels (ITHs) in Taiwan has increased from fourteen, with a total of 2,163 rooms, in 1970, to seventy, with a total of 20,382 rooms, in 2011 (Tourism Bureau 2012). In this highly competitive environment, how to survive has become an important issue for hotel operators. In the long run, the performance of a firm is an important determinant of its survival (Anderson, Fok, and Scott 2000). Performance measurement plays a critical role in the management process and is a prime concern for hotel general managers because it not only reflects the business situation in reality, but also gives a competitive advantage for following operations (Phillips 1999). It refers to how well a firm uses its resources to achieve desirable outcomes. If a firm can generate more outcomes than its competitors, it is more likely to dominate the market (Brown and Ragsdale 2002). Hence, hotel operators need more information regarding their performance to maintain or promote their competitiveness in the ITH industry.

To identify and assess performance, productivity measurement approaches are often applied. These can provide a performance comparison with competitors over time and space. With regard to productivity, measurement indexes are mainly divided into a partial factor productivity index and total factor productivity (TFP) index. Although partial factor productivity is simple to compute, it may provide a misleading overall productivity score (Coelli, Rao, and Battese 1998). In contrast, the TFP considers all dimensions of production so that studies now tend to use this approach to measure productivity. The Malmquist productivity index (MPI), which extends from the data envelopment analysis (DEA) model, has become a popular approach to estimate the TFP, because it only relies on quantity information, and does not require price information and an assumption on the optimizing behavior. The MPI can further decompose the productivity change into the components of technical change (TC) and efficiency change (EC) so as to further explore the sources of productivity change. Although there is a wealth of literature on both basic and applied research in conventional MPI introduced by Caves, Christensen, and Diewert (1982) and developed by Färe et al. (1994), the conventional MPI fails to satisfy the requirement of circularity (Balk and Althin 1996). The benchmark technology assumption of conventional MPI based on adjacent periods could cause different productivity change scores. In addition, the conventional MPI computed and decomposed by the linear programming (LP) techniques could cause infeasibility (Pastor and Lovell 2005). Although Berg, Førsund, and Jansen (1992), Shestalova (2003) and Nin et al. (2003) proposed new MPI approaches to solve the aforementioned problems, Berg, Førsund, and Jansen (1992) only satisfied the requirement of circularity, while Shestalova (2003) and Nin et al. (2003) were immune to LP infeasibility. To simultaneously overcome those problems, global MPI has been proposed by Pastor and Lovell (2005). The global MPI provides a single measure, avoids LP infeasibility, and is circular.

Conventional approaches assess performance under the assumption of the same technology. O’Donnell, Rao, and Battese (2008) argued that if firms operating under different technologies were assessed by the same efficiency frontiers, the evaluation results obtained by DEA methods might be inaccurate. To investigate the heterogeneity among groups, Battese and Rao (2002), Battese, Rao, and O’Donnell (2004) and O’Donnell, Rao, and Battese (2008) introduced the concept of metafrontier proposed by Hayami (1969) as well as Hayami and Ruttan (1970) into the efficiency measurement. Oh and Lee (2010) further proposed a metafrontier approach to measure productivity. As this approach is on the basis of the global MPI proposed by Pastor and Lovell (2005), it also can provide a single measure, avoid LP infeasibility, and is circular. It can further decompose MPI into within-group EC (the catching-up effect), within-group TC (the innovation effect), and technical leadership change (the technology leading effect). Although the metafrontier approach has been applied to estimate hotel efficiency, according to the authors’ knowledge, it is not used to evaluate hotel productivity over time. Hence, this study tries to fill this void and further explore the sources of productivity change for ITHs. This study provides in-depth information for hotel operators and policymakers to understand how well hotels perform under two operational styles and determinants of productivity change so that they can make related policies to improve their performance.

The remainder of this study is organized as follows: “Performance Evaluation in Hotel Industry” section reviews the relevant performance literature about the hotel industry; “Method” section introduces the methodology, which is used to evaluate productivity changes over time; “Empirical Implementation” presents data descriptions and empirical results; “Discussion” section discusses the major findings and provides managerial implications; and “Conclusion” section derives the conclusions.

Performance Evaluation in Hotel Industry

Early studies assessed the performance of hotels with various financial indicators, such as a cost-volume-profit index (Coltman 1978; Fay, Rhoads, and Rosenblatt 1971; Jaedicke and Robichek 1964), sales receipts (Van Doren and Gustke 1982), revenue (Baker and Riley 1994; Kimes 1989), and lodging index (Wassenaar and Stafford 1991). However, these financial measures only focus on output performance and fail to reflect the multidimensional nature of a hotel. To overcome the drawback, a considerable number of studies began to apply efficiency measures to evaluate the performance of hotels by simultaneously aggregating multiple inputs and outputs. Some studies introduced parametric approaches to estimate the efficiency of hotels (Anderson et al. 1999; Anderson, Fok, and Scott 2000; Barros 2004, 2006; Chen 2007; Hu et al. 2010; Kim 2011; Y.-H. Lin 2011). However, the most common efficiency approach is DEA, which is a nonparametric LP approach. It constructs an efficiency frontier that shows the minimum resource usage for a firm at a given level of output, or the maximum output expansion at a given level of input. After Morey and Dittman (1995) used DEA to measure the efficiency of hotels, a greater number of studies applied this approach to evaluate the performance of hotels. For a review of related studies, see Assaf, Barros, and Josiassen (2012). Furthermore, DEA has been applied to monitor cross-period performance (productivity change) in various hotel studies based on a Malmquist model (Barros 2005a; Barros and Alves 2004; Hwang and Chang 2003).

Beyond DEA efficiency evaluation, differences in efficiency between chain and independent hotels have been assessed in some studies, such as Hwang and Chang (2003); Chiang, Tsai, and Wang (2004); Sun and Lu (2005); Yang and Lu (2006); Chen (2007); as well as Yu and Lee (2009). These comparisons assumed that chain and independent hotels operated under the same technology. However, due to the differences in their operational styles, it cannot be reasonably assumed that they face the same benchmark technology. Recently, Assaf, Barros, and Josiassen (2012) used the metafrontier approach to evaluate the efficiencies of tourist hotels in which different sizes, classifications, and operational styles could achieve different efficiencies. Chiu, Huang, and Ting (2012) applied the nonradial different systems model to measure the efficiencies of ITHs based on the difference between operational styles. C.-H. Lin, Chiu, and Huang (2012) extended the metafrontier approach to assess the technology gaps of ITHs in productive and service processes according to their operational styles. Fang (2013) calculated the efficiencies of tourist hotels by the metafrontier approach. He separated the hotels into ITHs and standard tourist hotels (STHs) with different group frontiers. Huang et al. (2013) used the nonconvex metafrontier approach to estimate the efficiencies of ITHs and found that operational styles had a significant effect on efficiencies. Although Assaf, Barros, and Josiassen (2012), C.-H. Lin, Chiu, and Huang (2012) and Fang (2013) collected multiperiod data for performance analysis, they assessed efficiency by the metafrontier DEA model without considering the TC and the innovation effect of each group. In this study, we focus on investigating the metaproductivity changes and decompositions of ITHs in Taiwan.¹

The major contribution of this study is to compare the productivity changes among ITHs by using the metafrontier MPI based on recent ITH data. By this more comprehensive performance measurement, we can simultaneously explore the technology leading effect, innovation effect, and catching-up effect in the ITH industry over time and space, and investigate differences in productivity changes between chain and independent hotels. In particular, the innovation effect between chain and independent hotels over time and space has not been investigated in previous research. In the next section, we will discuss the proposed method in more detail.

Method

At first, Färe et al. (1994) constructed the conventional MPI as a geometric mean form of two MPIs as defined by Caves, Christensen, and Diewert (1982) and decomposed this index into two components: EC and TC. However, they neglected the problem that firms may operate under different technologies and face different production frontiers. To overcome this, Oh and Lee (2010) constructed a new approach that applied these concepts of metafrontier and global MPI to derive the MPI.

As Taiwanese ITHs operate in a competitive market, an output-oriented method is more adequate to estimate the MPI (Barros 2005b). To define an output-oriented MPI, it is assumed that an ITH uses a vector of N inputs $x^{t} = (x_{1}^{t}, \dots, x_{N}^{t}) \in R_{+}^{N}$ to produce a vector of M outputs, $y^{t} = (y_{1}^{t}, \dots, y_{M}^{t}) \in R_{+}^{M}$ , in period t, $t = 1, \dots, T$ . The benchmark technology, $P^{t} = {(x^{t}, y^{t}) : x^{t} can produce y^{t}}$ , satisfies the assumptions of convexity and free disposability as described in Färe et al. (1994). The output distance function in period t can be defined as (Shephard 1970)

D_{o}^{t} (x^{t}, y^{t}) = \inf {θ : (x^{t}, y^{t} / θ) \in P^{t}},

where $θ$ is the ratio of the vector of outputs to the production frontier represented by P in period t.

Furthermore, to construct the metafrontier MPI, it is necessary to define three technology sets: the contemporaneous benchmark technology, the intertemporal benchmark technology, and the global benchmark technology (Oh and Lee 2010). If the contemporaneous benchmark technology set of group $U_{l}$ is assumed to be $P_{U_{l}}^{t}$ , $l = 1, \dots, L$ , then the intertemporal and global benchmark technology sets can be defined as $P_{U_{l}}^{I} = c o n v {P_{U_{l}}^{1} \cup \dots \cup P_{U_{l}}^{T}}$ and $P^{G} = c o n v {P_{U_{1}}^{I} \cup \dots \cup P_{U_{L}}^{I}}$ . Then, the metafrontier MPI can be formed as

M P I_{o}^{G} (x^{t + 1}, y^{t + 1}, x^{t}, y^{t}) = \frac{D_{o}^{G} (x^{t + 1}, y^{t + 1})}{D_{o}^{G} (x^{t}, y^{t})},

As the metafrontier MPI is on the basis of one global benchmark technology, it does not need to resort to the geometric mean form. Therefore, $M P I_{}^{G}$ will be greater than unity if productivity is increased between periods t and t + 1.

$M P I_{}^{G}$ can be further decomposed as

\begin{array}{l} M P I_{o}^{G} (x^{t + 1}, y^{t + 1}, x^{t}, y^{t}) \\ = \frac{D_{o}^{t + 1} (x^{t + 1}, y^{t + 1})}{D_{o}^{t} (x^{t}, y^{t})} (\frac{D_{o}^{G} (x^{t + 1}, y^{t + 1})}{D_{o}^{t + 1} (x^{t + 1}, y^{t + 1})} \times \frac{D_{o}^{t} (x^{t}, y^{t})}{D_{o}^{G} (x^{t}, y^{t})}) \\ = \frac{T E_{o}^{t + 1} (x^{t + 1}, y^{t + 1})}{T E_{o}^{t} (x^{t}, y^{t})} (\frac{D_{o}^{I} (x^{t + 1}, y^{t + 1})}{D_{o}^{t + 1} (x^{t + 1}, y^{t + 1})} \times \frac{D_{o}^{t} (x^{t}, y^{t})}{D_{o}^{I} (x^{t}, y^{t})}) (\frac{D_{o}^{G} (x^{t + 1}, y^{t + 1})}{D_{o}^{I} (x^{t + 1}, y^{t + 1})} \times \frac{D_{o}^{I} (x^{t}, y^{t})}{D_{o}^{G} (x^{t}, y^{t})}) \\ = \frac{T E_{o}^{t + 1} (x^{t + 1}, y^{t + 1})}{T E_{o}^{t} (x^{t}, y^{t})} (\frac{D_{o}^{I} (x^{t + 1}, y^{t + 1} / D_{o}^{t + 1} (x^{t + 1}, y^{t + 1}))}{D_{o}^{I} (x^{t}, y^{t} / D_{o}^{t} (x^{t}, y^{t}))}) (\frac{D_{o}^{G} (x^{t + 1}, y^{t + 1} / D_{o}^{I} (x^{t + 1}, y^{t + 1}))}{D_{o}^{G} (x^{t}, y^{t} / D_{o}^{I} (x^{t}, y^{t}))}) \\ = (\frac{T E_{o}^{t + 1} (x^{t + 1}, y^{t + 1})}{T E_{o}^{t} (x^{t}, y^{t})}) \times (\frac{B P G_{o}^{I, t + 1} (x^{t + 1}, y^{t + 1})}{B P G_{o}^{I, t} (x^{t}, y^{t})}) \times (\frac{T G R_{o}^{t + 1} (x^{t + 1}, y^{t + 1})}{T G R_{o}^{t} (x^{t}, y^{t})}) \\ = E C (x^{t + 1}, y^{t + 1}, x^{t}, y^{t}) \times B P C (x^{t + 1}, y^{t + 1}, x^{t}, y^{t}) \times T G C (x^{t + 1}, y^{t + 1}, x^{t}, y^{t}), \end{array}

where the first ratio is EC, which measures the change in technical efficiency (TE) between periods t and t + 1; the second ratio is the best-practice gap change (BPC), which measures the change in the best-practice gap (BPG) that is similar to the TC within a specific group, and BPG is the best-practice gap between $P^{I}$ and $P^{s}$ measured along rays $(x^{s}, y^{s})$ , $s = t, t + 1$ ; and the last ratio is the technology gap ratio change (TGC), which is the change in the technology gap ratio (TGR), and TGR is the technology gap between $P^{G}$ and $P^{I}$ measured along rays $(x^{s}, y^{s})$ , $s = t, t + 1$ . EC is greater than 1 as efficiency increases. BPC will be greater than unity as technology progresses between periods t and t + 1. This represents the fact that the benchmark technology in period t + 1 is closer to the intertemporal benchmark technology than the benchmark technology in period t. Similarly, TGC will be superior to 1 as technical leadership progresses between periods t and t + 1.

Exhibit 1 illustrates the concept of the metafrontier in the Malmquist productivity framework. It exhibits decision making units (DMUs) in two different groups, $U_{1}$ and $U_{2}$ with two outputs, $y_{1}$ and $y_{2}$ during periods 1, 2, and 3. $P_{U_{1}}^{1}$ , $P_{U_{1}}^{2}$ , and $P_{U_{1}}^{3}$ are the contemporaneous benchmark technology sets in periods 1, 2, and 3, respectively; $P_{U_{1}}^{I}$ and $P_{U_{2}}^{I}$ are the intertemporal benchmark technology sets; and $P^{G}$ is the global benchmark technology set. A DMU is observed to produce an output vector A in period 1 and an output vector E in period 2. Using Exhibit 1, metafrontier MPI and its components can be derived as follows:

\begin{array}{l} M P I_{o}^{G} (x^{2}, y^{2}, x^{1}, y^{1}) = \frac{O E / O H}{O A / O D} \\ = \frac{O E / O F}{O A / O B} \times \frac{O F / O G}{O B / O C} \times \frac{O G / O H}{O C / O D} . \end{array}

Exhibit 1:

Concept of the Metafrontier Malmquist Productivity Index.

To compute the metafrontier MPI and its components, six output distance functions in Equation 3 need to be calculated by DEA. For each hotel, $k \in U_{l}$ , the output distance functions can be obtained by solving the following LP problem with a variable returns-to-scale (VRS) technology

{[D_{o}^{s} (x_{k}^{s}, y_{k}^{s})]}^{- 1} = \max_{} φ_{C}^{k, s},

subject to

\sum_{j \in U_{l}}^{} z_{j}^{s} y_{j, m}^{s} \geq φ_{C}^{k, s} y_{k, m}^{s}, m = 1, \dots, M,

\sum_{j \in U_{l}}^{} z_{j}^{s} x_{j, n}^{s} \leq x_{k, n}^{s}, n = 1, \dots, N,

\sum_{j \in U_{l}}^{} z_{j}^{s} = 1,

z_{j}^{s} \geq 0,

where $z_{j}^{s}$ represents the intensity variable, $s = t, t + 1$ .

If ${\hat{φ}}_{C}^{k, s}$ denotes the optimal solution to Equation 4, then the intertemporal distance function, which uses all ITHs over all periods within a specific group $U_{l}$ , can be computed by the model

{[D_{o}^{I} (x_{k}^{s}, y_{k}^{s} / D_{o}^{s} (x_{k}^{s}, y_{k}^{s}))]}^{- 1} = \max_{} φ_{I}^{k, s},

subject to

\sum_{j \in U_{l}, s \in ψ}^{} z_{j}^{s} y_{j, m}^{s} \geq φ_{I}^{k} {\hat{φ}}_{C}^{k, s} y_{k, m}^{s}, m = 1, \dots, M,

\sum_{j \in U_{l}, s \in ψ}^{} z_{j}^{s} x_{j, n}^{s} \leq x_{k, n}^{s}, n = 1, \dots, N,

\sum_{j \in U_{l}, s \in ψ}^{} z_{j}^{s} = 1,

z_{j}^{s} \geq 0, ψ = {1, \dots, T} .

If ${\hat{φ}}_{I}^{k, s}$ denotes the optimal solution to Equation 5, then the global distance function, which uses all ITHs and periods over all groups, can be calculated as

{[D_{o}^{G} (x_{k}^{s}, y_{k}^{s} / D_{o}^{I} (x_{k}^{s}, y_{k}^{s}))]}^{- 1} = \max_{} φ_{G}^{k, s},

subject to

\sum_{j \in U, s \in ψ}^{} z_{j}^{s} y_{j, m}^{s} \geq φ_{G}^{k, s} {\hat{φ}}_{I}^{k, s} {\hat{φ}}_{C}^{k, s} y_{k, m}^{s}, m = 1, \dots, M,

\sum_{j \in U, s \in ψ}^{} z_{j}^{s} x_{j, n}^{s} \leq x_{k, n}^{s}, n = 1, \dots, N,

\sum_{j \in U, s \in ψ}^{} z_{j}^{s} = 1,

z_{j}^{s} \geq 0, U = U_{1} \cup \dots \cup U_{L}, ψ = {1, \dots, T} .

Empirical Implementation

Data Descriptions

The balanced panel data on fifty-four ITHs from 2008 to 2011 in this study were gathered from the Annual Operation of the International Tourist Hotels published by the Taiwan Tourism Bureau.² Regarding inputs, we select the number of guest rooms, the number of employees, the total floor area of food and beverage (F&B), and other expenses. Regarding outputs, we choose room revenues, F&B revenues, and other revenues. In addition, the consumer price index based on 2011 is used to deflate other expenses, room revenues, F&B revenues, and other revenues to remove the effect of inflation.

Due to the differences in hotel operational styles, ITHs will operate under different technologies. In Taiwan, ITHs can be roughly categorized as chain hotels and independent hotels by operational style.³ A hotel joining an international or domestic chain can obtain the advantages of economies of scale, knowledge transfer, and resource sharing, while independent hotels possess higher operational flexibility and independence in decision-making (Chiu, Huang, and Ting 2012). Exhibit 2 depicts the descriptive statistics for the variables divided by chain and independent hotels. It can be seen that for both input and output variables, the values of variables for chain hotels are higher than those for independent hotels. The list of fifty-four observed ITHs is presented in the appendix.

Exhibit 2:

Descriptive Statistics for the Variables.

Variables^a	2008		2009		2010		2011
Variables^a	M	SD	M	SD	M	SD	M	SD
Chain hotels (n = 31)
Number of guest rooms	333.68	181.32	333.68	181.32	334.23	181.44	334.23	181.44
Number of employees	386.68	231.05	375.48	222.57	386.55	232.85	391.23	233.37
Total floor area of F&B	6,116.77	9,255.86	6,114.13	9,257.25	5,965.90	9,214.40	5,835.97	9,259.15
Other expenses	216,420.55	186,891.99	211,126.70	181,111.81	212,159.29	168,741.76	213,413.74	161,544.61
Room revenues	331,771.79	276,136.65	300,331.44	243,495.77	325,434.68	250,290.17	338,978.24	257,489.77
F&B revenues	370,911.12	313,213.78	341,861.33	305,090.12	391,667.16	359,986.35	409,132.21	361,722.63
Other revenues	132,992.74	146,654.39	118,739.15	140,645.70	114,212.55	126,510.14	109,858.91	135,425.16
Independent hotels (n = 23)
Number of guest rooms	263.39	91.04	261.30	89.85	257.61	85.34	252.61	85.13
Number of employees	243.87	149.90	234.83	145.78	230.39	148.80	234.22	151.61
Total floor area of F&B	2,814.74	2,292.39	2,814.74	2,292.39	2,809.57	2,291.64	2,721.22	2,258.81
Other expenses	103,184.45	80,258.27	93,315.32	64,832.71	92,135.23	61,818.10	103,984.66	83,014.39
Room revenues	158,542.47	94,374.63	150,654.23	90,289.58	158,829.41	88,090.47	164,985.76	95,678.02
F&B revenues	174,253.39	175,868.40	155,177.60	150,614.42	176,228.18	188,850.76	183,881.63	188,451.29
Other revenues	41,727.93	71,798.66	41,995.39	77,906.12	38,228.17	59,489.76	43,558.06	71,203.13

Other expenses, room revenues, F&B revenues and other revenues are measured in terms of thousands of New Taiwan dollars.

Empirical Results

Comparison between Chain Hotels and Independent Hotels

Before adopting the metafrontier approach, we must confirm that chain hotels and independent hotels operate under different technologies. For this purpose, we have calculated the values of TGR for chain and independent hotels. It is worth remembering that the TGR measures the proximity of the intertemporal frontier to the global frontier. If similar technologies are used between these two groups, the technology gap between chain hotels’ intertemporal frontier and global frontier will be similar to the technology gap between independent hotels’ intertemporal frontier and global frontier. Hence, to examine whether the technologies between two groups are significantly different, we assess the statistical significance of the differences between the TGRs of these two groups by the Wilcoxon–Mann–Whitney test (Beltrán-Esteve et al. 2014). The null hypothesis argues that the TGRs of these two groups are drawn from the same distribution and have identical technologies, while the alternate hypothesis argues the opposite. The result from the Wilcoxon–Mann–Whitney test shows that the null hypothesis is rejected at the 1 percent significance level, indicating that chain hotels and independent hotels have different technologies.⁴ It is adequate to use metafrontier MPI to estimate the productivity change of ITHs.

To compare the productivity changes between chain and independent hotels, the values of productivity change and its sources are reported in Exhibit 3. As can be seen in Exhibit 3, the hotels operating under different technologies will lead to distinct results. The productivity for chain hotels has deteriorated (–0.5%) on average over the sample period, while independent hotels have improved their productivity (8.9%). For the decomposition of productivity change, the rate of within-group EC for chain hotels decreased (0.7%), while that for independent hotels was enhanced (5.8%), indicating that, on average, chain hotels have moved further away from the contemporaneous frontiers, while independent hotels have gotten closer to the contemporaneous frontiers. The within-group TC for chain hotels declined at a rate of 0.7 percent, whereas that for independent hotels increased at a rate of 3.6 percent, which means that, on average, chain hotels have regressed technically, whereas independent hotels have shown technical progress. The index of technical leadership change for chain hotels was an increase of 0.9 percent, while that for independent hotels decreased by 0.6 percent. This indicates that independent hotels have moved further away from the global frontier. However, for both independent hotels and chain hotels, the change of productivity was mainly driven by within-group EC and TC from 2008 to 2011. An interesting result is that the rate of technical leadership change for the chain hotels is better than that for the independent hotels, indicating that the chain hotels’ intertemporal benchmark technology is closer to the global frontier, which means the chain hotels have improved their intertemporal technology more than the independent hotels. However, this does not mean that the within-group TC for the chain hotels should be better than for the independent hotels. The result of the better within-group TC for the independent hotels indicates that technology progress of independent hotels compared with their intertemporal technology during the study period was better than that of chain hotels compared with their intertemporal technology. In a similar way, the rate of within-group EC for the independent hotel was better than the chain hotel, indicating that the independent hotels improved their efficiencies more quickly than the chain hotels compared with their respective technology.

Exhibit 3:

Productivity Growth between Chain and Independent Hotels.

	MPI	EC	BPC	TGC
Chain hotels
2008-2009	0.951	0.960	0.978	1.013
2009-2010	1.013	1.026	0.988	1.000
2010-2011	1.032	1.008	1.028	0.996
2008-2011	0.995	0.993	0.993	1.009
Independent hotels
2008-2009	0.966	0.999	0.969	0.998
2009-2010	1.084	1.030	1.051	1.001
2010-2011	1.039	1.027	1.018	0.994
2008-2011	1.089	1.058	1.036	0.994

Note. MPI = Malmquist productivity index.

Furthermore, because the TGC is the rate of technical leadership change, it cannot provide a clear conclusion about which is the technology leader and which is the follower (Oh and Lee 2010). To overcome this problem, we use the box-plots for the temporal pattern of TGR with respect to each group to examine this argument (see Exhibit 4). If the value of TGR is lower, then the technology of a specific hotel is farther from the global frontier. When the value of TGR is equal to 1, the hotel lies on the global frontier. As can be seen in Exhibit 4, the values of TGR for most chain hotels over time are close to 1, while those for independent hotels are more dispersed.⁵ This means that the chain hotels’ intertemporal frontier is closer to the global frontier than independent hotels’ intertemporal frontier. In other words, the curvatures of chain hotels’ intertemporal frontier and global frontier are more similar. Furthermore, this shows that chain hotels are the technology leaders, and independent hotels are the followers. In other words, chain hotels provide the better technology type, relative to independent hotels, to generate customer outcomes.

Exhibit 4:

Distribution of Technical Gap Ratios for Chain Hotels (CH) and Independent Hotels (IH).

The above results indicate that the chain hotels continued adopting advanced technology, while the independent hotels caught up to the contemporaneous frontier and shifted technology to intertemporal technology more quickly than chain hotels did.

To thoroughly investigate the differences in patterns between chain and independent hotels, the annual cumulative evolution of productivity with respect to each group is depicted in Exhibit 5. These two groups display similar patterns of productivity change with improved productivity after 2010. As tourism is the lagging indicator, this pattern coincides with the effect of the financial tsunami beginning in 2008. In addition, the cumulative productivity growth in independent hotels is higher than that in chain hotels. The cumulative productivity growth in chain hotels is also lower than 1 over the sample period, indicating that although productivity was improved after 2010, the improvement was not enough to compensate for the productivity deterioration in 2009.

Exhibit 5:

Cumulative Productivity Change in Chain and Independent Hotels.

To further understand the sources of productivity changes, we plot the decomposition of productivity for chain and independent hotels from 2008 to 2011 in Exhibit 6 and Exhibit 7, respectively. Values greater than zero indicate that performance is improved. The sources of productivity change for chain hotels are more complex. In 2009, the technical leadership progress alleviated the productivity deterioration for chain hotels. In 2010, the within-group efficiency improvement resisted the within-group technical and technical leadership regress and facilitated productivity growth. Although the within-group efficiency improvement and within-group technical progress contributed toward the productivity of chain hotels, the positive effects were offset by the slowdown of technical leadership progress, resulting in productivity in 2011 being still lower than that in 2008. Except for the negative relationship with the productivity change and technical leadership change in 2011, the productivity change for independent hotels had a similar pattern with its decompositions, in which the within-group technical progress was the dominant factor facilitating productivity growth before 2010, while the within-group efficiency improvement was the dominant one after 2010.

Exhibit 6:

Average Annual Growth in Productivity and Its Decomposition in Chain Hotels.

Exhibit 7:

Average Annual Growth in Productivity and Its Decomposition in Independent Hotels.

Comparing Individual Hotel Productivity

In terms of individual hotel productivity, Exhibit 8 shows that CH10, CH12, CH22, CH23, and IH01 defined the global frontier in both 2008 and 2011, so MPI = 1. This means that these hotels were best-practice hotels in the ITH industry from 2008 to 2011. In addition, fifteen chain hotels and sixteen independent hotels had higher productivity in 2011 than that in 2008, whereas the remaining twelve chain hotels and six independent hotels had lower productivity in 2011 than that in 2008. Which factor leads to the productivity change? To explore the source of productivity change, various combinations of within-group EC, within-group TC, and technical leadership change are represented as follows:

Hotels with efficiency increase, technology progress, and technical leadership progress: CH03, CH04, CH05, CH08, CH09, CH24, CH31, IH05, IH11, IH14, IH16, IH18, IH19, IH20, and IH21 belong to this set. They are the better performing hotels with the improvement of efficiency, technology progress, and technical leadership progress, resulting in increased productivity. These hotels should not only maintain their operating advantages, but also seek further improvements to catch up with these best-practice hotels in the ITH industry. Among them, CH09, IH18, and IH21 are the three hotels with the maximum productivity growth. CH09, IH18, and IH21 show 52.9 percent, 50.4 percent, and 45.7 percent increases over the sample period, respectively.

Hotels with efficiency increase, technology progress, and technical leadership regress: This set includes CH21, IH03, IH06, IH07, IH08, IH09, IH10, IH12, and IH22. These hotels upgraded their efficiency and technology within their group, but enlarged the technology gap with best-practice hotels in the ITH industry. They should refer to best-practice hotels to improve their technology. Among them, productivity was improved at eight of the nine hotels, meaning that efficiency increase and technology progress dominated technical leadership regress.

Hotels with efficiency increase, technology regress, and technical leadership progress: Six hotels are contained in this set: CH02, CH18, CH20, CH27, CH29, and CH30. These hotels improved their efficiency and narrowed the technology gap with best-practice hotels in the ITH industry, but enlarged the BPG with best-practice chain hotels. They should refer to best-practice chain hotels to improve their technology. Among them, CH18, CH27, and CH29 had improvement in productivity, indicating that the problem of technology regress in these hotels was not too serious.

Hotels with efficiency decrease, technology progress, and technical leadership progress: CH07, CH13, CH25, CH26, and IH15 belong to this set. These hotels reduced the BPGs in their group and the TGR in the ITH industry, but experienced declines in efficiency. They should catch up to other hotels. Among them, CH25, CH26, and IH15 still experienced increases in their productivity, which means that the deterioration in efficiency was dominated by technology progress and technical leadership progress.

Hotels with efficiency increase, technology regress, and technical leadership regress: CH14, CH17, and IH23 are included in this set. These hotels improved their efficiency, but enlarged the BPGs in their group and the TGR in the ITH industry. These hotels should not only improve their technology, but also keep up with the best operational technology achieved by the best-practice hotels. MPI of CH17 was greater than 1, indicating that efficiency increase dominated technology regress and technical leadership regress.

Hotels with efficiency decrease, technology progress, and technical leadership regress: CH15, IH04, IH13, and IH17 are included in this set. These hotels experienced technology progress, but deteriorated their efficiency and enlarged the TGR in the ITH industry. They should improve the management of outputs and emulate best-practice hotels in terms of technology. Among them, CH15 had improved productivity, indicating that technology progress dominated efficiency decrease and technical leadership regress.

Hotels with efficiency decrease, technology regress, and technical leadership progress: CH01, CH06, CH16, CH19, and CH28 belong to this set. These hotels experienced technical leadership progress, but deteriorated in efficiency and enlarged the BPG within their group. They should catch up to other chain hotels and refer to best-practice chain hotels to improve their technology. All of them experienced declines in productivity, which means that technical leadership progress was dominated by efficiency decrease and technology regress. In particular, although CH06 experienced technical leadership progress, it had the largest decline in the sample. It shows a 40.2 percent decrease over the sample period due to 39.6 percent deterioration in efficiency and 6.8 percent regress in technology.

Hotels with efficiency decrease, technology regress, and technical leadership regress: This set contains CH11 and IH02. These hotels experienced declines in efficiency and enlarged the BPGs within their group and the TGR in the ITH industry, resulting in decreased productivity. These hotels should improve their management of outputs and keep up with best-practice hotels within their groups and in the ITH industry.

Exhibit 8:

Cumulated Productivity Growth for Individual Hotels, 2008-2011.

	MPI	EC	BPC	TGC
Chain hotels
CH01	0.863	0.941	0.917	1.000
CH02	0.971	1.056	0.895	1.027
CH03	1.086	1.000	1.009	1.075
CH04	1.191	1.000	1.188	1.003
CH05	1.120	1.000	1.120	1.000
CH06	0.598	0.604	0.932	1.062
CH07	0.996	0.994	1.002	1.000
CH08	1.209	1.029	1.175	1.000
CH09	1.529	1.379	1.109	1.000
CH10	1.000	1.000	1.000	1.000
CH11	0.712	0.761	0.956	0.978
CH12	1.000	1.000	1.000	1.000
CH13	0.857	0.695	1.226	1.005
CH14	0.991	1.106	0.925	0.969
CH15	1.045	0.908	1.167	0.987
CH16	0.873	0.882	0.986	1.004
CH17	1.024	1.209	0.865	0.979
CH18	1.050	1.168	0.894	1.006
CH19	0.813	0.819	0.911	1.090
CH20	0.932	1.000	0.932	1.000
CH21	1.074	1.000	1.082	0.993
CH22	1.000	1.000	1.000	1.000
CH23	1.000	1.000	1.000	1.000
CH24	1.052	1.000	1.000	1.052
CH25	1.049	0.993	1.026	1.030
CH26	1.004	0.999	1.005	1.000
CH27	1.153	1.191	0.968	1.000
CH28	0.861	0.986	0.854	1.023
CH29	1.257	1.438	0.874	1.000
CH30	0.897	1.047	0.850	1.008
CH31	1.056	1.000	1.053	1.003
Average	0.995	0.993	0.993	1.009
Independent hotels
IH01	1.000	1.000	1.000	1.000
IH02	0.892	0.974	0.939	0.976
IH03	1.222	1.190	1.085	0.947
IH04	0.926	0.923	1.047	0.959
IH05	1.119	1.000	1.000	1.119
IH06	1.124	1.111	1.040	0.973
IH07	1.326	1.395	1.089	0.873
IH08	1.193	1.060	1.164	0.966
IH09	1.215	1.219	1.044	0.955
IH10	1.037	1.008	1.081	0.952
IH11	1.244	1.000	1.010	1.232
IH12	0.753	1.000	1.000	0.753
IH13	0.990	0.930	1.072	0.993
IH14	1.003	1.000	1.000	1.003
IH15	1.071	0.983	1.065	1.024
IH16	1.029	1.000	1.000	1.029
IH17	0.962	0.938	1.054	0.973
IH18	1.504	1.454	1.018	1.015
IH19	1.174	1.000	1.000	1.174
IH20	1.024	1.000	1.000	1.024
IH21	1.457	1.231	1.123	1.053
IH22	1.133	1.095	1.048	0.987
IH23	0.954	1.000	0.982	0.971
Average	1.089	1.058	1.036	0.994

Note. MPI = Malmquist productivity index.

These results show that there is still room for improvement in almost all of Taiwanese ITHs to pursue the best-practice procedures in management achieved by the best-practice hotels in the ITH industry.

Discussion

This study proposes an alternative framework to reflect the productivity changes of Taiwanese ITHs. As the data analyzed are recent, we can provide an accurate investigation of the recent industry trends. With the use of the metafrontier MPI, we can also identify the sources of productivity changes in the ITH industry. From the empirical results, it could be found that, on average, chain hotels and independent hotels improved productivity after the financial tsunami. However, over the sample period, chain hotels suffered a slight drop in productivity, while independent hotels had growth in productivity. Chain hotels can obtain the advantages of economies of scale, knowledge transfer, and resource sharing (Chiu, Huang, and Ting 2012), but their shortcomings are lower operational flexibility, whereas independent hotels possess higher operational flexibility and independence in decision-making (Chiu, Huang, and Ting 2012), but their shortcomings are the lack of knowledge transfer, resource sharing, and branding effect. With the shock of the financial tsunami, hotels should have adjusted their operational strategies to mitigate the effects of this event. However, since chain hotels’ activities are restricted, they could not immediately adopt new strategies or technology to resist the deterioration in productivity and improve it. Such characteristics of chain hotels might explain why they had greater productivity deterioration in 2009 and lower productivity growth in 2010 and 2011 than independent hotels did.

Regarding the sources of productivity changes for chain and independent hotels, they are mainly driven by within-group EC and within-group TC. Within-group EC requires the ability to efficiently use existing inputs and technology to generate more outputs. It shows whether a hotel operator more efficiently manages its resources over the sample period. If a hotel experienced efficiency deterioration, its operator should make more effort to control the resource usage. Technical change includes both any investment in improving productivity (Barros 2005a) and sales effects (Lee and Johnson 2012, 2014). Hotels can make investment in new technology or innovation by capital accumulation. Although TC is often ascribed to production issues, technology regress may result from insufficient demand (Lee and Johnson 2012, 2014) in a service system. Within-group TC represents whether this hotel group innovates within the scope of its group in the period under consideration or undergoes demand deterioration. In other words, it investigates whether this hotel group enhances its technology to move toward the best-practice intertemporal technology within its group, or whether revenue fluctuations lead to variations in the output levels affecting TC measures of operations. If a hotel experienced technology regress, its operator should either treat the best-practice intertemporal technology within its group as benchmarks to increase its investment in upgrading its production structure or investigate whether the reason is just demand fluctuations or not. Technical leadership change represents whether a hotel operator innovates by enlarging its scope to the whole industry. In other words, it investigates whether a hotel further enhances the best-practice intertemporal technology within its group to move toward the best-practice global technology in the whole industry. If a hotel experienced technical leadership regress, its operator should adjust its product mix to the production point where the distance between the intertemporal technology frontier and the global technology frontier is closer. Under a decrease in demand resulting from the financial tsunami, hotel managers could not immediately react and adopt appropriate measures to prevent a waste of resources, so the efficiency deteriorated.⁶ In addition, with an unfavorable market outlook, technology regress may result in a decrease in investments or lack of demand. Afterward, the improvement of management of outputs and an increase in investment and demand would facilitate efficiency increases and technology progress. However, because chain hotels possess lower operational flexibility, their technology was improved after 2011.

In addition, the results obtained reveal that chain hotels are the technology leaders, and independent hotels are the followers. They also reveal that the rate of technical leadership change for chain hotels on average is higher than that for independent hotels, which is opposite to the results of within-group EC and within-group TC. If a hotel has more advanced technology than other hotels, it is the technology leader in the ITH industry. However, a technology leader does not mean that it would promote the improvement of technology or enhance the technology more than followers do over the period under consideration. In addition, although leaders’ intertemporal technology frontier is closer to a global one than the followers’, those leaders’ contemporaneous frontiers might still shift inward due to a decrease in demand. Similarly, a hotel with within-group technical progress is not necessarily the leader in this industry and does not necessarily adjust its production in the right direction that makes the production point on the intertemporal technology closer to the point on the global technology. Although chain hotels make less effort to decrease wasting of resources and improve their technology, they have the more advanced technology and adjust their product mix to enhance the production level to move toward the global production level. The advanced technology of chain hotels could result from their operational characteristics. Because the members of chain hotels have advantages, such as innovative reservation systems sourced from the management affiliation (Mandelbaum 1997), as well as sharing of the knowledge and technology, their inherent technologies are superior to independent hotels’. In contrast, because independent hotels cannot benefit from the knowledge spillover and technology transfer, they should spend more time and resources through trial-and-error experiences (Mitsuhashi and Yamaga 2006). Hence, they are followers in inventing new technologies, resulting in greater distance between their intertemporal technology frontier and the global technology frontier. Because more chain hotels adopt advanced technology, they are good benchmarks in the ITH industry. Chain hotels with poor performance may more easily imitate chain hotels that stay in the global frontier to adjust their product mix. However, chain and independent hotels belong to different operational styles. Learning from another operational style is more difficult, because they may adopt different operational technologies. Hence, independent hotels may not learn from the best-practice hotels in the ITH industry. That may be why the rate of technical leadership change for chain hotels is better than that for independent hotels.

Finally, how do hotel operators improve their performance to boost competitiveness? Taking CH11 as an example, its 28.8 percent deterioration in productivity results from a 23.9 percent drop in efficiency, 4.4 percent regress in technology, and 2.2 percent regress in technical leadership. To at least return to the original performance level in 2008, CH11 needs to expand its consumed outputs 28.8 percent at a given level of inputs by enhancing resource utilization, improving its technology and adjusting its product mix to pursue the lead in inventing new technology. Because a decline in efficiency is the main source of the productivity deterioration, the operator of CH11 should first focus on the improvement of efficiency. The operator can increase hotel occupancy rate, F&B revenues, and other revenues by applying marketing and price strategies, appropriate managerial procedure, and so on. Second, to keep up with best-practice chain hotels within its group, the operator should make the investment in new methods, procedures, and techniques or understand the reason for demand fluctuations to improve the operational technology. Finally, to be close to the global frontier, the operator should further adjust the product mix to match the direction of bias. This example does not indicate that hotels with productivity growth do not need to improve their performance. If there is still room for improvement of performance, these hotels still can enhance their relative competitiveness by increasing their productivity until they stand on the global frontier.

Based on the above discussion, the main managerial implications of this finding are as follows. Because the performance of ITHs is influenced by environmental factors, they need to carefully face and handle the problems of changes in the operational environment. Although independent hotels can more quickly react to changes in the operational environment by improving efficiency and adopting new technology or innovation, they still need to make more effort to obtain the lead in inventing new technologies. Although chain hotels take the lead in inventing new technologies, they still need to increase their operational flexibility to quickly adjust the usage of inputs and technology.

Conclusion

The major aim of this study is to provide an alternative framework for the evaluation of productivity in the Taiwanese ITH industry based on a dataset of fifty-four ITHs from 2008 to 2011. The metafrontier MPI is applied to measure the productivity indexes and their decompositions, and to overcome the shortcomings of the conventional MPI, which fails to satisfy the requirement of circularity and does not consider heterogeneity between firms.

The empirical results of productivity change in the ITH industry offer several insights. Roughly speaking, the use of different technologies by hotels will affect productivity change. Both within-group EC and TC are the main factors of productivity change for chain and independent hotels. For global technical innovation, chain hotels are technology leaders, and independent hotels are followers. Because this study cannot cover the whole hotel industry and compare the difference between STHs and ITHs, we hope this field can be investigated in the future. Another possible direction for future research is to incorporate nondiscretionary inputs into the metafrontier MPI to more closely conform to the operational characteristics of the hotel industry. In addition, the issue of the direction of bias needs further investigation in the future. Finally, although DEA-based MPI provides valid information about the resource usage, it does not identify the causes of within-group EC, within-group TC, and technical leadership change. This limitation can be considered in future research.

Footnotes

Appendix

Exhibit A1:

List of Sample Hotels.

Hotel	Code
Chain hotels
Ambassador Hotel Hsinchu	CH01
Ambassador Hotel Kaohsiung	CH02
Ambassador Hotel Taipei	CH03
Caesar Park Hotel Kenting	CH04
Caesar Park Hotel Taipei	CH05
Chateau de Chine Hualien	CH06
Evergreen Laurel Hotel Taichung	CH07
Evergreen Plaza Hotel (Tainan)	CH08
Fleur de Chine Hotel	CH09
Grand Hyatt Taipei Hotel	CH10
Gloria Prince Hotel Taipei	CH11
Hotel Royal Chiao His	CH12
Hotel Royal Chihpen	CH13
Hotel Royal Hsinchu	CH14
Howard Beach Resort Kenting	CH15
Howard Plaza Hotel Kaohsiung	CH16
Howard Prince Hotel Taichung	CH17
Howard Plaza Hotel Taipei	CH18
Kaohsiung Grand Hotel	CH19
Landis Resort Yangmingshan Hotel	CH20
Landis Taipei Hotel	CH21
Regent Taipei Hotel	CH22
Shangri-La’s Far Eastern Plaza Hotel Taipei	CH23
Sheraton Grande Taipei Hotel	CH24
Sherwood Taipei Hotel	CH25
Splendor Kaohsiung Hotel	CH26
Splendor Hotel Taichung	CH27
Taipei Grand Hotel	CH28
Taiwan Chiayi Nice Prince Hotel	CH29
Tayih Landis Hotel	CH30
Westin Taipei Hotel	CH31
Independent hotels
Brother Hotel	IH01
Emperor Hotel	IH02
Formosan Naruwan Hotel and Resort Taitung	IH03
Golden China Hotel	IH04
Grand Hi-Lai Hotel	IH05
Han-Hsien International Hotel	IH06
Hibiscus Resort Hotel	IH07
Hotel Holiday Garden	IH08
Hotel National	IH09
Hotel Riverview Taipei	IH10
Hotel Tainan	IH11
Hualien Farglory Hotel	IH12
Imperial Hotel Taipei	IH13
Lalu Hotel	IH14
Marshal Hotel	IH15
Miramar Garden Taipei	IH16
Parkview Hotel	IH17
Plaza International Hotel	IH18
San Want Hotel	IH19
Santos Hotel	IH20
Taoyuan Hotel	IH21
The Lees Hotel	IH22
United Hotel	IH23

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, or publication of this article.

Funding

The authors received no financial support for the research, authorship, or publication of this article.

Notes

Author Biographies

Ming-Miin Yu, Ph.D, is a professor in the Department of Transportation Science at National Taiwan Ocean University. His research interests include the transportation economics and management, logistics, particularly the productivity and efficiency analysis in transportation and logistics.

Li-Hsueh Chen, Ph.D, is a research assistant in the Department of Transportation Science at National Taiwan Ocean University. Her research interests include the industrial economics, transportation economics, and performance evaluation.

References

Anderson

R. I.

Fish

Xia

Michello

1999. Measuring efficiency in the hotel industry: A stochastic frontier approach. International Journal of Hospitality Management 18 (1): 45-57.

Anderson

R. I.

Fok

Scott

2000. Hotel industry efficiency: An advanced linear programming examination. American Business Review 18 (1): 40-48.

Assaf

Barros

C. P.

Josiassen

2012. Hotel efficiency: A bootstrapped metafrontier approach. International Journal of Hospitality Management 31:621-29.

Baker

Riley

1994. New perspectives on productivity in hotels: Some advances and new directions. International Journal of Hospitality Management 13 (4): 297-311.

Balk

M. B.

Althin

1996. A new transitive productivity index. The Journal of Productivity Analysis 7:19-27.

Barros

C. P.

2004. A stochastic cost frontier in the Portuguese hotel industry. Tourism Economics 10 (2): 177-92.

Barros

C. P.

2005a. Evaluating the efficiency of a small hotel chain with a Malmquist productivity index. International Journal of Tourism Research 7 (3): 173-84.

Barros

C. P.

2005b. Measuring efficiency in the hotel sector. Annals of Tourism Research 32 (2): 456-77.

Barros

C. P.

2006. Analyzing the rate of technical change in the Portuguese hotel industry. Tourism Economics 12 (3): 325-46.

10.

Barros

C. P.

Alves

F. P.

2004. Productivity in the tourism industry. International Advances in Economic Research 10 (3): 215-25.

11.

Battese

G. E.

Rao

D. S. P.

2002. Technology gap, efficiency, and a stochastic metafrontier function. International Journal of Business and Economics 1:87-93.

12.

Battese

G. E.

Rao

D. S. P.

O’Donnell

C. J.

2004. A metafrontier production function for estimation of technical efficiencies and technology gaps for firms operating under different technologies. Journal of Productivity Analysis 21:91-103.

13.

Beltrán-Esteve

Gómez-Limón

J. A.

Picazo-Tadeo

A. J.

Reig-Martinez

2014. A metafrontier directional distance function approach to assessing eco-efficiency. Journal of Productivity Analysis 41:69-83.

14.

Berg

S. A.

Førsund

F. R.

Jansen

E. S.

1992. Malmquist indices of productivity growth during the deregulation of Norwegian banking, 1980-1989. Scandinavian Journal of Economics 94:211-28.

15.

Brown

J. R.

Ragsdale

C. T.

2002. The competitive market efficiency of hotel brands: An application of data envelopment analysis. Journal of Hospitality and Tourism Research 26 (4): 260-332.

16.

Caves

D. W.

Christensen

L. R.

Diewert

W. E.

1982. The economic theory of index numbers and the measurement of input, output, and productivity. Econometrica 50:1393-414.

17.

Chen

2007. Applying the stochastic frontier approach to measure hotel managerial efficiency in Taiwan. Tourism Management 28:696-702.

18.

Chiang

W.-E.

Tsai

M.-H.

Wang

L. S.-M.

2004. A DEA evaluation of Taipei hotels. Annals of Tourism Research 31 (3): 712-15.

19.

Chiu

Y.-H.

Huang

C.-W.

Ting

C.-T.

2012. A non-radial measure of different systems for Taiwanese tourist hotels’ efficiency assessment. Central European Journal of Operations Research 20 (1): 45-63.

20.

Coelli

Rao

D. S. P.

Battese

G. E.

1998. An introduction to efficiency and productivity analysis. Boston: Kluwer Academic.

21.

Coltman

M. M.

1978. Hospitality management accounting. Boston: CBI Publishing Company.

22.

Fang

C.-Y.

2013. Did the financial crisis and policy deregulation for Chinese tourists affect the efficiency and output slacks of Taiwanese hotels? In Quantitative methods in tourism economics, ed. Matias

Á.

Nijkamp

Sarmento

, 287-305. Heidelberg: Springer.

23.

Färe

Grosskopf

Norris

Zhang

1994. Productivity growth, technical progress, and efficiency change in industrialized countries. American Economic Review 84 (1): 66-83.

24.

Fatma

Timothy

N. H.

2005. Patient satisfaction in a preoperative assessment clinic: An analysis using SERVQUAL dimensions. Total Quality Management & Business Excellence 16 (1): 15-30.

25.

Fay

C. T.

Rhoads

R. C.

Rosenblatt

R. L.

1971. Managerial accounting for hospitality service industries. Dubuque: William C. Brown Publishers.

26.

Hayami

1969. Sources of agricultural productivity gap among selected countries. American Journal of Agricultural Economics 51:564-75.

27.

Hayami

Ruttan

V. W.

1970. Agricultural productivity differences among countries. American Economics Review 60:895-911.

28.

Huang

C.-W.

Ting

C.-T.

Lin

C.-H.

Lin

C.-T.

2013. Measuring non-convex metafrontier efficiency in international tourist hotels. Journal of the Operational Research Society 64:250-59.

29.

J.-L.

Chiu

C.-N.

Shieh

H.-S.

Huang

C.-H.

2010. A stochastic cost efficiency analysis of international tourist hotels in Taiwan. International Journal of Hospitality Management 29 (1): 99-107.

30.

Hung

W.-T.

Shang

J.-K.

Wang

F.-C.

2010. Pricing determinants in the hotel industry: Quantile regression analysis. International Journal of Hospitality Management 29:378-84.

31.

Hwang

S.-N.

Chang

T.-Y.

2003. Using data envelopment analysis to measure hotel managerial efficiency change in Taiwan. Tourism Management 24:357-69.

32.

Jaedicke

R. K.

Robichek

1964. Cost-volume-profit analysis under conditions of uncertainty. The Accounting Review 39 (4): 917-26.

33.

Kim

2011. Factor determinants of total factor productivity growth in the Malaysian hotel industry: A stochastic frontier approach. Cornell Hospitality Quarterly 52 (1): 35-47.

34.

Kimes

S. E.

1989. The basics of yield management. Cornell Hotel and Restaurant Administration Quarterly 30 (3): 14-19.

35.

Lee

C.-Y.

Johnson

A. L.

2012. Two-dimensional efficiency decomposition to measure the demand effect in productivity analysis. European Journal of Operational Research 216:584-93.

36.

Lee

C.-Y.

Johnson

A. L.

2014. Proactive data envelopment analysis: Effective production and capacity expansion in stochastic environments. European Journal of Operational Research 232:537-48.

37.

Lin

C.-H.

Chiu

Y.-H.

Huang

C.-W.

2012. Assessment of technology gaps of tourist hotels in productive and service processes. The Service Industries Journal 32 (14): 2329-342.

38.

Lin

Y.-H.

2011. Estimating cost efficiency and the technology gap ratio using the metafrontier approach for Taiwanese international tourist hotels. Cornell Hospitality Quarterly 52 (3): 341-53.

39.

Mandelbaum

1997. Hotel sales-and-marketing management: A snapshot of current practice and technology use. Cornell Hotel and Restaurant Administration Quarterly 38 (6): 46-51.

40.

Mitsuhashi

Yamaga

2006. Market and learning structures for gaining competitive advantage: An empirical study of two perspectives on multiunit-multimarket organizations. Asian Business and Management 5:225-47.

41.

Morey

R. C.

Dittman

D. A.

1995. Evaluating a hotel GM’s performance: A case study in benchmarking. Cornell Hotel Restaurant and Administration Quarterly 36 (5): 30-35.

42.

Nin

Arndt

Hertel

T. W.

Preckel

P. V.

2003. Bridging the gap between partial and total factor productivity measures using directional distance functions. American Journal of Agricultural Economics 85:928-42.

43.

O’Donnell

C. J.

Rao

D. S. P.

Battese

G. E.

2008. Metafrontier frameworks for the study of firm-level efficiencies and technology ratios. Empirical Economics 34:231-55.

44.

D.-H.

Lee

J.-D.

2010. A metafrontier approach for measuring Malmquist productivity index. Empirical Economics 38:47-64.

45.

Pastor

J. T.

Lovell

C. A. K.

2005. A global Malmquist productivity index. Economic Letter 88:266-71.

46.

Phillips

P. A.

1999. Performance measurement systems and hotels: A new conceptual framework. International Journal of Hospitality Management 18 (2): 171-82.

47.

Shephard

R. W.

1970. Theory of cost and production functions. Princeton: Princeton University Press.

48.

Shestalova

2003. Sequential Malmquist indices of productivity growth: An application to OECD industrial activities. Journal of Productivity Analysis 19:211-26.

49.

Sun

W.-M.

2005. Evaluating the performance of the Taiwanese hotel industry using a weight slacks-based measure. Asia-Pacific Journal of Operational Research 22 (4): 487-512.

50.

Tourism Bureau. 2012. Annual operation of the international tourist hotels 2011. Taipei: Tourism Bureau, Ministry of Transportation and Communications.

51.

Van Doren

C. S.

Gustke

L. D.

1982. Spatial analysis of the U.S. lodging industry, 1963-1977. Annals of Tourism Research 9 (4): 543-63.

52.

Wassenaar

D. J.

Stafford

E. R.

1991. The lodging index: An economic indicator for the hotel/motel industry. Journal of Travel Research 30 (1): 18-21.

53.

Yang

W.-M.

2006. Performance benchmarking for Taiwan’s international tourist hotels. INFOR 44 (3): 229-45.

54.

M.-M.

Lee

C.-Y.

2009. Efficiency and effectiveness of service business: Evidence from international tourist hotels in Taiwan. Tourism Management 30:571-80.