1 Introduction
The travel and tourism industry has been a primary economic activity and reliable source of revenue generation for many nations across the globe. The hospitality industry plays an important role in the overall economic growth of an economy (Seetanah 2011; Schubert et al. 2011; Sharma et al. 2022). The performance of firms has been a subject of great interest to researchers over the past few decades. Many studies on firm performance have initially focused on developed markets (Anderson and Reeb 2003), though the focus has recently been shifting to emerging markets (Jaisinghani et al. 2018). In addition, there are varied aspects of firm performance that have been investigated in depth. These include firms’ financial performance (Waddock and Graves 1997), employees’ performance, and the performance of associated companies (Luo and Chung 2005). Recent studies, however, have focused not only on how organizations perform but also on how organizations sustain their performance in the long run (Gschwandtner 2012). This is usually referred to as the theory of profit persistence (Mueller 1977). The theory states that profitable firms can sustain their profitability in the long run by any of several means including the creation of entry and exit barriers (Gschwandtner 2005), formation of ownership concentration (Agostino et al. 2005), utilization of consumers’ inertia (Choi and Wang 2009), development of managerial and political ties (Kotabe et al. 2011), and exploitation of innovation capabilities (Bartoloni and Baussola 2009).
Studies on persistence dynamics, however, have been mostly restricted to firms’ profitability. There are other aspects of performance that need thorough investigation in terms of their persistence over time. One such aspect pertains to the persistence of firms’ efficiency. It is usually contended that firms’ efficiency should be measured along with firms’ performance. This is because efficiency also provides an insight into the extent of inputs consumed to generate a level of output (Charnes et al. 1978). Various authors have studied the efficiency of firms in the context of both developed and emerging economies (Huther 1997). Most of these studies have primarily concentrated on analyzing the factors that impact the efficiency of the firms. However, there is not much work on analyzing the persistence of firms’ efficiency for both advanced and emerging markets.
Anzeige
Among the emerging markets, the Indian hospitality industry is rapidly emerging as the primary growth driver for the overall services sector. However, to the best of the authors’ knowledge, Indian hospitality and tourism firms have never been researched in terms of their efficiency persistence. Such gap becomes more obvious, as India is ranked 7th out of 184 countries, in the year 2019, in terms of travel and tourism’s total contribution to GDP and is ranked 2nd in the world in terms of the number of people employed in the hospitality and tourism industry (Statista 2021). The hospitality and tourism industry in India generated 8 per cent of the total employment opportunities in 2017 by providing direct and indirect employment to 41.6 million people.1 Pillai (2017) highlights the importance of Indian tourism in the region stating that the capital accumulation of tourism in India is substantially higher than the other countries in the region, Asia-Pacific, and world averages. India has attracted the attention of domestic and international tourists due to its unique portfolio of diverse tourism segments, namely, eco-tourism, pilgrimage tourism, medical tourism, rural tourism, heritage tourism, and luxury tourism.2 The Government of India has taken several steps to make India a hub for global tourism including the allocation of a significant proportion of the federal budget to the promotion of the hospitality and tourism sector.
The Indian hospitality industry has witnessed enormous growth during the last decade, and the industry has maintained persistent growth. The industry presently contributes more than USD 23 billion in the form of international tourist receipts, and the figure is expected to exceed USD 27 billion by the year-end of 2020.2 The prime growth drivers for the hospitality and tourism industry in India include better travel connections, enhanced international marketing efforts, and increased focus on medical tourism.3 The increase in tourists’ arrival is expected to provide a growth impetus to hotels in terms of their overall occupancy rates and increased tariffs. Table 1 shows persistent growth in the industry’s direct contribution to the GDP (CAGR 14.05%) and total contribution to the GDP (CAGR 9.72%); the number of foreign tourist arrivals (CAGR 11.51%); and the capital investment outflows in the tourism sector (3.3%).
Table 1
– Economic Indicators of the Indian Tourism Sector (2009–2017)
Year | Direct Contribution to GDP (USD Billions) | Total Contribution to GDP (USD Billions) | Foreign Tourist Arrivals (Millions) | Capital Investments in Tourism Sector (USD Billions) |
|---|---|---|---|---|
2009 | 26.0 | 89.0 | 5.2 | 23.8 |
2010 | 31.0 | 103.0 | 5.8 | 29.9 |
2011 | 31.0 | 105.0 | 6.3 | 33.0 |
2012 | 36.0 | 116.0 | 6.6 | 31.3 |
2013 | 40.1 | 122.1 | 7.0 | 32.2 |
2014 | 41.6 | 126.8 | 7.4 | 35.0 |
2015 | 42.8 | 147.7 | 8.0 | 34.6 |
2016 | 71.7 | 208.5 | 8.8 | 36.6 |
2017 | 76.6 | 239.4 | 10.2 | 38.2 |
CAGR
|
14.1%
|
9.7%
|
11.5%
|
3.6%
|
Such contribution of the hospitality and tourism industry, as in the case of India, in terms of its socio-economic contributions, is critical for the firms operating in the industry to be more efficient and maintain a reasonable level of persistence efficiency. This is important for their competitive advantage and capacity to make sustainable contributions to the growth of the hospitality and tourism industry in India. The present study aims to meet this expectation gap and analyses the persistence of firms’ efficiency for firms operating in the tourism and hospitality industry in India. To study the persistence of firms’ efficiency for firms operating in the tourism and hospital industry in India, the study uses Data Envelopment Analysis (DEA) for the appraisal of the firms’ technical efficiencies; followed by the application of the dynamic panel regression estimation of the efficiency scores over a set of specific variables such as firms’ liquidity, age, size, leverage, capital expenditure intensity, and return on sales. The results convey the positive and significant persistence of the financial efficiency of hospitality and tourism firms in India. The results bring the argument for the creation of entry and exit barriers by firms to generate sustainable competitive advantage. The evidence suggests that Indian hospitality and tourism firms are creating entry and exit barriers, besides possibly exploiting the behavioural inertia of consumers to generate higher revenues. This should be helping certain firms to improve and sustain their financial efficiency. The broad results indicate, furthermore, that emerging market firms behave differently from their counterparts in advanced markets. The current results bear certain critical managerial implications for the Indian hospitality and tourism industry in the context of inefficient markets. It is imperative for hospitality and tourism firms to streamline their financial management policies, this involves controlling costs and devising methods for the enhancement of revenues. This can be accomplished by an in-depth analysis and rationalization of certain discretionary expenses such as marketing costs and costs of training and development. The results suggest the firm in the sector can create certain entry and exit barriers and devise ways of exploiting inertia in consumers’ behaviour that enable them to sustain their efficiency in the long run. The management response to changing consumer needs is required to increase financial efficiency. The study suggests regulators specifically focus on policies that can enhance the competitive dynamics of the industry. For this to happen, the licensing requirements must be eased and the opening the sector to foreign competition. This is besides taking measures to prevent firms from forming cartels and indulging in other restrictive practices.
Anzeige
The study makes several useful contributions to the existing body of knowledge. Given the scarcity of past studies on the persistence of efficiency in the hospitality and tourism industry, the study is the first of its kind that examines the persistence of efficiency for the tourism and hospitality firms operating in India. The study analyses the key factors that impact the efficiency of hospitality and tourism firms, the results of this study shall provide managerial implications for the tourism and hospitality industry. This is because managers can take decisions that can enhance the persistence of their firms’ efficiency. These results provide support to the argument on the creation of entry and exit barriers by firms to generate sustainable competitive advantage. Similarly, policymakers can frame regulations that force firms to adopt practices that increase their efficiency in the long run. The findings of the study shall also help the management, regulators and tho policymaking bodies in emerging economies in the region concerning hospitality and tourism firms to understand the need for the persistence of operating efficiency and to develop strategies to attain sustainable competitive advantage. The study applies two advanced quantitative techniques, i.e., DEA and dynamic panel regression to estimate the persistence of financial efficiency in the tourism and hospitality industry. The use of these techniques aids the estimation of the persistence of firms’ efficiency in addition to the persistence of firm performance. This is a significant methodological contribution in measuring tourism efficiency literature.
The remaining paper is organized as follows. Section 2 presents the literature review. Data and research methodology are described in Sect. 3. Empirical results are presented in Sect. 4. Discussion of the results is presented in Sect. 5. The final section concludes the paper and also presents its limitations and scope for future research.
2 Literature Review
Several research studies have contributed to the literature concerning the measurement of efficiency in the hospitality and tourism industry (Corne 2015). There have also been studies that have analyzed the efficiency of tourism destinations (Fuentes et al. 2012). It has also been contended that there exists a positive association between efficiency and the number of tourists arriving at a particular destination (Hu et al. 2010). However, the focus has now been shifted towards maintaining long-term efficiency and generating persistent revenues (Fuentes et al. 2012; Chaabouni, 2018). Therefore, the present paper seeks to expand the existing literature by significantly contributing towards the measurement of financial efficiency and its persistence. In addition, the study also makes an effort to determine the key factors that affect the efficiency of firms operating in the hospitality and tourism industry.
The theory of competitive environment suggests that it is not possible to sustain any kind of abnormal profits over a long-run period. This is mainly because any form of abnormal profit will lead to an increase in competition and thereby restore the normal profits. If this argument is indeed true, then there should be no persistence of abnormal profits in the long run. This should be especially true for firms that operate in highly competitive industries. However, many studies confirm the existence of positive persistence in profitability and other related measures of firm performance (Gschwandtner 2005; Hirsch and Hartmann 2014). Thus, it is a fact that certain firms can create and maintain superior performance in the long run. The usual arguments proposed for this particular occurrence are related to the creation of entry and exit barriers, and exploitation of behavioural inertia. A natural extension of this logic therefore should pertain to the persistence of firms’ efficiency in the long run. It should also be interesting to check the factors that impact firms’ efficiency. Hence, the following two sub-sections present a review of literature on the persistence of efficiency and factors affecting efficiency, respectively.
2.1 Literature on Persistence of Efficiency
It is mostly argued that there is a noteworthy relationship between profitability and efficiency. However, the nature of this relationship is not clear, and contrary empirical evidence is usually reported in applied studies ( Stierwald 2009). These studies also bring to the fore the need for studying the persistence of efficiency along with analyzing the persistence of profitability. The formation of entry and exit barriers and utilization of production and behavioural inertia, which are usually contended to be causing persistence of profitability, can also cause persistence of efficiency in the firms.
Many authors have analyzed the relationship between profitability and efficiency in the context of advanced markets. For instance, Stierwald (2009) studies the impact of cost-efficiency on the profitability of Australian firms. The author performs the study for a sample of 961 Australian companies operating over the period 1995–2005. The author finds that efficiency has a significant and positive impact on firms’ profitability. The author further finds that the relationship does not vary greatly across different sectors. A few studies have tried to analyze the persistence of certain aspects of efficiency for developed markets. Wang and Huang (2007) analyze the persistence of economic efficiency of Taiwan’s commercial banks. The authors find a moderate persistence of economic efficiency for the sample firms. Similarly, Chen et al. (2009) examine the impact of efficiency on reward for CEOs of US banks over the period 1997–2004. The authors also analyze the persistence of efficiency of CEOs over the study period. The authors found that CEOs’ efficiency has a positive and significant impact on their overall compensation. The authors also find that efficiency scores of CEOs decreased over the study period. However, this subject remains less explored in the context of emerging economies.
2.2 Literature on Factors Affecting Efficiency
In recent years many studies have attempted to identify and analyse the factors that account for the differences in the efficiency of firms. These studies mostly posit that firm-specific factors such as age, size, growth intensity, marketing, and research and development have the largest impacts on firms’ efficiency (Singh and Gaur, 2009; Mia and Soltane 2016). Thus, efficiency is believed to be most affected by internal factors rather than by external factors (Kilic and Okumus 2005). Hence, there is a need to determine the key factors affecting efficiency for different firms operating in different industries.
Several researchers argue that the age and size of a firm are the key determinants of its overall efficiency. The underlying belief is that learning by firms acts as a costless and automatic method of improving efficiency. Sharma et al. (2022) in their study on the persistence of the financial efficiency of the tourism and hospitality industry in China observed that even though creating efficiency in financial operations may require time, but advocated efforts by both new and established entrants as this ultimately provide consistent improved financial performance. Such clarity in long-term strategies is important for sustainable value creation for stakeholders in the tourism and hospitality industry (Sharma et al. 2022; Glancey 1998) posits that older firms enjoy the benefit of reputation which allows them to achieve higher profitability and efficiency. Similarly, Jovanovic (1982) contends that large-sized firms are more efficient than small-sized firms, because of their ability to exploit economies of scale. However, certain authors also cite that age and size are inversely related to firms’ efficiency (Lundvall and Battese 2000). This is mostly because smaller-sized and young firms embrace the latest technologies to outperform the more established firms. Moreover, the organizational inertia in the older firms can also lead to their inefficiency (Sorensen and Stuart, 1999).
Apart from age and size, it is also generally argued that firm-specific variables such as leverage, marketing, growth, and accounting profitability have large impacts on firms’ efficiency (Mok et al. 2007). In addition, it has also been found that the levels of efficiency are also related to location, internationalization, audit service and other management variables (Parte-Esteban and Alberca-Oliver 2015). The proponents of these variables argue that firm-specific variables are the leading factors that cause organizations to be efficient or inefficient. Thus, there is a need for proper identification of the key factors that impact efficiency. This is even more relevant for firms belonging to emerging markets.
3 Data and Methodology
3.1 Data and Source
The data for the present study comprises information on various firm-specific factors for hospitality and tourism firms in India. The data have been collected for all the publicly listed hospitality and tourism firms operating in the country. The Prowess database, maintained by CMIE, has been utilized to obtain the data. The data is of nine years from 2009 to 2017 for all the firms. Indian firms follow an April-to-March fiscal year. The firms having missing data for any year during these years were not considered. The filtering process produced a final sample of 32 Indian firms, resulting in a balanced panel of 288 firm-year observations.
3.2 Estimation Methodology
3.2.1 Data Envelopment Analysis
The methodology applied to measure the financial efficiency in this paper is based on Data Envelopment Analysis (DEA), which is a mathematical programming approach directed towards the construction of frontiers. The study has preferred the use of DEA over other frontier approaches for the reason that there is no need for specifying the parametric functional form which would demand the weights of inputs and outputs and an explicit distributional assumption for the inefficiency terms. The empirical estimation of efficiency in the DEA method dates back to the ideas of Farrell (1957), who sketched further upon the work of Debreu (1951) and Koopmans (1951) to define a simple measure of productive efficiency in order to account for multiple inputs. Following Farrell (1957), Charnes et al. (1978) and Banker et al. (1984) introduced the two most popular DEA models, which later received recognition via the use of the initials of their names only, as the CCR and BCC models, respectively.
Literature reports increasing interest in the use of DEA methodology to measure firms’ efficiency. In addition, several mathematical based programming models have been suggested in the DEA literature to estimate different forms of efficiencies of firms as several researchers have extensively used and modified the basic DEA models for advanced estimation of efficiency (Thanassoulis 2001). In the present study, the directional super-efficiency DEA (DSDEA) model has been used, which was first introduced by Andersen and Petersen (1993) with basic input and output orientation, and was further amended by Ray (2008) using a directional distance function based on Luenberger’s (1992) benefit function. The model, as amended by Ray (2008), has its roots in the procedure proposed by Chambers et al. (1996) which considers simultaneous improvements in inputs and outputs. Sharma et al. (2022) observed summarised the two advantages of the DS-DEA model. The first observation they cited is that it simultaneously accounts for increasing levels of outputs and decreasing levels of inputs. The second advantage of using the DS-DEA model as per Sharma et al. (2022) is that the upper score limit of 1 for efficient firms found in traditional DEA models is relaxed to restore the full information of efficiency. Thus, the DS-DEA model provides a comprehensive measure of efficiency. They explained that in this in this model, the upper bound of 1 for the efficient decision-making units (DMUs), as in the case of basic DEA models, is relaxed to restore the full information on efficiency over the production frontier. To compute the super-efficiency scores, the DMU under assessment is eliminated from the set of reference DMUs. The super-efficiency scores of efficient DMUs will always be greater than or equal to 1. However, since the inefficient DMUs cannot spread beyond the range of production frontier, the technical efficiency scores of those units will remain the same.
To illustrate the empirical estimation of the financial efficiency scores of hospitality and tourism firms using the DSDEA model, it is important to assume that output\( y\in {\mathfrak{R}}_{+}^{S}\) is produced from\( x\in {\mathfrak{R}}_{+}^{M}\) to obtain a feasible input-output bundle\( (x, y)\). The feasible input-output bundles constitute the technology (\( T)\), also referred to as production possibility set, is given as:
$$ T=\left\{\left(x,y\right):x can produce y;x\ge 0;y\ge 0\right\}$$
(1)
In case of multiple-inputs and multiple-outputs (assumed to be freely disposable), the production possibility set (assumed to be convex) can be empirically presented as:
$$ T=\left\{\left(x,y\right):x\ge \sum _{j=1}^{n}{\lambda }_{j}{x}_{j};y\le \sum _{j=1}^{n}{\lambda }_{j}{y}_{j};\sum _{j=1}^{n}{\lambda }_{j}=1;{\lambda }_{j}\ge 0;j=\text{1,2},3,\dots .n\right\}$$
(2)
The directional distance function (Chambers et al. 1996) can be presented as follows:
$$ \overrightarrow{D}\left(x , y;{g}_{x} ,{g}_{y}\right)=Max \left\{\beta : \left(x-\beta {g}_{x} , y+\beta {g}_{y}\right)\in T\right\}$$
(3)
where \( ({g}_{x} ,{g}_{y})\) is a non-zero vector in\( {\mathfrak{R}}_{+}^{M}\times {\mathfrak{R}}_{+}^{S}\). This vector determines the direction in which inputs and outputs have to move. Essentially, \( \beta {g}_{x}\) is subtracted from \( x\), and \( \beta {g}_{y}\) is added to\( y\). Moreover, \( \left(-x ,y\right)\) can be considered for \( {(g}_{x} , {g}_{y})\), and in that case the directional distance function, as proposed by Chambers et al. (1996) and Ray (2008), becomes:
$$ \overrightarrow{D}\left(x , y\right)=\text{max}\beta : \left\{\left(1-\beta )x , (1+\beta )y\right)\right\}\in T $$
(4)
The symbol \( \overrightarrow{D}\) denotes directional distance function. The model seeks to increase the outputs and reduce the inputs simultaneously by the proportion \( \beta \) (Ray 2008). The factor \( \beta \) represents technical inefficiency. By implication, its efficiency equals\( (1-\beta ).\)
For the estimation of super-efficiency scores under the directional distance function based approach, the production possibility set \( T\) for DMU \( k\) can be modified as:
$$ T=\left\{\left(x,y\right):x\ge \sum _{j=1}^{n}{\lambda }_{j}{x}_{j};y\le \sum _{j=1}^{n}{\lambda }_{j}{y}_{j};\sum _{j=1}^{n}{\lambda }_{j}=1;{\lambda }_{j}\ge 0;j=\text{1,2},3,\dots .n,\forall j\ne k\right\}$$
(5)
The super-efficiency score of a DMU\( k\), using the directional distance function based CCR model, can be obtained by solving the following linear programming problem.
$$ Maximize {\beta }_{k}^{Super}$$
$$ Subject to: \sum _{j=1}^{N}{\lambda }_{j}{x}_{ij}+{\beta }_{k}^{Super}{x}_{ik}\le {x}_{ik}$$
$$ \sum _{j=1}^{N}{\lambda }_{j}{y}_{rj}-{\beta }_{k}^{Super}{y}_{rk}\ge {y}_{rk}$$
$$ {\lambda }_{j}\ge 0 ; j\ne k ;{\beta }_{k}^{Super}=free$$
(6)
where \( {TE}_{CRS}^{Super}={\theta }_{k}^{Super}=\left(1-{\beta }_{k}^{Super}\right)\) represents the super-efficiency score measured by the directional distance function based CCR method. If DMU\( k\) is super-efficient, the value of \( {\beta }_{k}^{Super}\) will be \( \le 0\), which implies that the output bundle of the given DMU has to be scaled down while its input bundle has to be scaled up. Between two super-efficient DMUs, the DMU with a lesser (that is, more negative) value of \( {\beta }_{k}^{Super}\) is ranked superior in terms of super-efficiency.
The corresponding optimization problem for the directional distance function based BCC model to calculate the super-efficiency of the DMU \( k\) can be written as follows:
$$ Maximize {\beta }_{k}^{{Super}^{*}}$$
$$ Subject to: \sum _{j=1}^{N}{\lambda }_{j}{x}_{ij}+{\beta }_{k}^{{Super}^{*}}{x}_{ik}\le {x}_{ik}$$
$$ \sum _{j=1}^{N}{\lambda }_{j}{y}_{rj}-{\beta }_{k}^{{Super}^{*}}{y}_{rk}\ge {y}_{rk}$$
$$ \sum _{j=1}^{N}{\lambda }_{j}=1$$
$$ {\lambda }_{j}\ge 0 ; j\ne k ;{\beta }_{k}^{{Super}^{*}}=free$$
(7)
where \( {TE}_{VRS}^{Super}={\varphi }_{k}^{Super}=(1-{\beta }_{k}^{{Super}^{*}})\) represents the super-efficiency score measured by the directional distance function based BCC method. For the super-efficient DMUs, the optimal value of \( {\beta }_{k}^{{Super}^{*}}\)will be negative, which leads to the measured super-efficiency levels to exceed unity.
3.2.2 Dynamic Panel Data Regression
The present dataset combines the characteristics of cross-sectional units (firms) and periods (years). Hence, a panel data technique is most suited to estimate the results. In addition, to estimate the persistence in the efficiency, the dynamic panel model is highly appropriate. Therefore, the present study applies a dynamic-panel regression methodology to estimate the degree of persistence of efficiency for hospitality and tourism firms. The greatest advantage of dynamic panel-based regression technique over traditional static panel regression is its ability to consider the lag of the dependent variable as an explanatory variable in the model. Hence, the baseline equation can be represented as follows:
$$ {{\Psi }}_{\text{i}\text{t}}= {\upalpha }+ {\uplambda }{{\Psi }}_{\text{i}\text{t}-1}+ {\text{Z}{\prime }}_{\text{i}\text{t}}{\uptheta }+{{\varepsilon }}_{\text{i}\text{t}}$$
(8)
where, \( {{\Psi }}_{\text{i}\text{t}}\) represents firms’ efficiency, λ is a scalar quantity and represents the degree of persistence in the dependent variable, and Z′it represents a 1 \( \times \) K vector that consists of all the explanatory and control variables. The coefficients of the independent variables are represented by θ which is a K \( \times \) 1 vector, εit represents the error terms, which are assumed to be following a one-way decomposition as presented below.
$$ {{\varepsilon }}_{\text{i}\text{t}} = {\text{v}}_{\text{i}} + {{\upmu }}_{\text{i}\text{t}}$$
(9)
where \( {\text{v}}_{\text{i}} \) and \( {{\upmu }}_{\text{i}\text{t}}\) are white noise error terms having zero mean and constant variance. It is also assumed that \( {\text{v}}_{\text{i}} \) and \( {{\upmu }}_{\text{i}\text{t}}\) are independent of one another and are also uncorrelated with their past values. In the general setting, the subscript “i” denotes a particular cross-sectional unit (firm), and the subscript “t” denotes a particular period (year). \( {{\Psi }}_{\text{i}\text{t}-1}\), which is the dynamic term and represents the lagged value of the dependent variable that has been introduced as an explanatory variable in the equation. λ denotes the persistence level. It is generally presumed that a λ value lying in the range of 0 and 0.2 indicates low persistence. Similarly, a λ value between 0.2 and 0.5 indicates a moderate level of persistence. Finally, a λ value above 0.5 indicates a high degree of persistence.
The dynamic-panel regression model, described with the help of Eq. (8), suffers from certain estimation issues. The key issue pertains to the presence of autocorrelation that arises because the lag values of the dependent variable have been included among the independent variables. Another issue is related to the presence of the individual, specific effects in the panel setting (Baltagi 2008). To overcome these two issues in the dynamic setting, Arellano and Bond (1991) proposed considering the Generalized Method of Moments (GMM) technique. Bhattacharyya (2012) suggests that the basic principle of such estimation is to use a first difference transformation to eliminate the individual-specific effects and then to consider the dependent variable with two-period lags or more lags as valid instruments. The authors posited that Instrument Variables (IV) can be exploited to correct the bias arising because of considering the lag dependent variable among the regressors. Arellano and Bond (1991) also suggested that the orthogonality condition prevailing between the lag values of the dependent variable and the noise term \( {{\upmu }}_{\text{i}\text{t}}\) can be explored to generate a set of valid instruments. The authors proposed that Eq. (8) can be differenced to generate the following functional form:
$$ {{{\Psi }}_{\text{i}\text{t}}- {\Psi }}_{\text{i}\text{t}-1}={\uplambda }\left({{{\Psi }}_{\text{i}\text{t}-1}- {\Psi }}_{\text{i}\text{t}-2}\right)+ \left({{\text{Z}}^{{\prime }}}_{\text{i}\text{t}}-{{\text{Z}}^{{\prime }}}_{\text{i}\text{t}-1}\right){\uptheta }+({{\upmu }}_{\text{i}\text{t}}-{{\upmu }}_{\text{i}\text{t}-1})$$
(10)
The functional form, as represented by Eq. (10), can be extended for t = 3. In this case, \( {{\Psi }}_{\text{i}\text{t}-2}\) becomes a valid instrument because it is correlated with (\( {{{\Psi }}_{\text{i}\text{t}-1}- {\Psi }}_{\text{i}\text{t}-2}\)) but is not correlated with (\( {{\upmu }}_{\text{i}\text{t}}-{{\upmu }}_{\text{i}\text{t}-1}\)). Similarly, extending the time period up to t = 4 will yield \( {{\Psi }}_{\text{i}\text{t}-3}\) and \( {{\Psi }}_{\text{i}\text{t}-2}\) as valid instruments. Thus, this logic can be extended further to obtain a set of valid instruments that can be represented by (\( {{\Psi }}_{i1, } {{\Psi }}_{i2, }\dots {{\Psi }}_{it-2}\)). Furthermore, all the purely exogenous variables can also be considered valid instruments. Therefore, the overall matrix comprising all the valid instruments can be represented by the following form.
$$ {{\Omega }}_{\text{i}} = \left(\begin{array}{ccc}\left[{{\Psi }}_{\text{i}1}\right], [{{\text{Z}}^{{\prime }}}_{\text{i}1},{{\text{Z}}^{{\prime }}}_{\text{i}2},\dots {\text{Z}{\prime }}_{\text{i}\text{T}}]& & 0\\ & \left[{{\Psi }}_{\text{i}1}, {{\Psi }}_{\text{i}2}\right][{{\text{Z}}^{{\prime }}}_{\text{i}1},{{\text{Z}}^{{\prime }}}_{\text{i}2},\dots {\text{Z}{\prime }}_{\text{i}\text{T}}] & \\ & \ddots & \\ 0& & \left[{{\Psi }}_{\text{i}1}, {{\Psi }}_{\text{i}2},\dots .{{\Psi }}_{\text{i}\text{T}-2}\right],[{\text{Z}{\prime }}_{\text{i}1},{\text{Z}{\prime }}_{\text{i}2},\dots {\text{Z}{\prime }}_{\text{i}\text{T}}] \end{array}\right) $$
(11)
Arellano and Bond (1991) proposed the one-step and two-step techniques for estimating the coefficients of λ and θ. The details of the estimation process are available in Baltagi (2008). Dynamic panel regression techniques are most suitable for datasets consisting of many cross-sectional units and few periods.
3.3 Variables
3.3.1 Input and Output Variables for DEA Analysis
In DEA analysis, an accurate selection of input and output variables is vital to the relevance and effectiveness of the results. Since the present study aims to gauge the financial efficiency of Indian hospitality and tourism firms, a set of financial input and output variables has been selected to accomplish this objective. For this purpose, a detailed survey of various past studies has been done (Morey and Dittman 1995) and accordingly, the variables, compensation to employees, operating and marketing expenses, and net fixed capital, have been selected as input variables, and total revenues have been selected as output variable.
3.3.2 Dependent Variables
The present study utilizes two different measures of efficiency, namely the overall technical efficiency (OTE) scores and the pure technical efficiency (PTE) scores, as the main dependent variables. The OTE measure originates from the CCR model in which the efficiency scores are independent of the scale of production. This implies that the model assumes the frontier to exhibit a constant return-to-scale (CRS). On the other hand, the PTE measure is the end result of the BCC model which assumes that the frontier observes a variable return-to-scale (VRS). PTE, which can be interpreted as management’s capability to convert the inputs into outputs, provides a measure of managerial efficiency and is therefore devoid of scale effect.
3.3.3 Explanatory Variables
The explanatory variables for the present study consist of a host of firm-specific factors. These factors have been considered for all the firms and all the years. The key explanatory variables included in the final model are firm size (SIZE), financial leverage (LEVER), capital expenditure intensity (CAPEX), firm age (AGE), liquidity (LIQ), and return on sales (ROS). SIZE has been operationalized as a natural log of total assets. LEVER is measured as long-term borrowings as a proportion of total assets. CAPEX has been measured as the ratio of total capital expenditure to net sales. AGE has been operationalized as a natural log of the difference between the current year and the incorporation year. LIQ has been measured as the ratio of current assets to current liabilities. Finally, ROS has been measured as operating profit scaled by net sales.
4 Results and Discussion
4.1 Descriptive Statistics and Correlations
The descriptive statistics as well as the frequency distribution of the efficiency scores (OTE and PTE) are presented in Table 2. Table 2 shows that, amongst the 32 firms operating in the hospitality and tourism industry of India, only 5 firms are super-efficient with an overall efficiency score of greater than one. However, as far as pure technical efficiency is concerned, the number of super-efficient firms counts to 11. The mean OTE score of inefficient firms is 0.881, whereas the mean PTE score of inefficient firms is 0.916. This shows that there is a great scope for improvement in terms of the financial efficiency of hospitality and tourism firms operating in India. Furthermore, the minimum and maximum values of the efficiency scores highlight that there is a huge variation amongst these firms in terms of financial efficiency.
Table 2
Descriptive Statistics and Frequency Distribution of OTE and PTE Scores
2009 | 2010 | 2011 | 2012 | 2013 | 2014 | 2015 | 2016 | 2017 | Overall | |
|---|---|---|---|---|---|---|---|---|---|---|
Panel A – OTE
| ||||||||||
Score < 1 (Inefficient)
| 29 | 27 | 27 | 25 | 26 | 26 | 26 | 29 | 28 | 27 |
Score > 1
(Super-efficient)
| 3 | 5 | 5 | 7 | 6 | 6 | 6 | 3 | 4 | 5 |
Total
| 32 | 32 | 32 | 32 | 32 | 32 | 32 | 32 | 32 | 32 |
Mean*
| 0.869 | 0.898 | 0.897 | 0.907 | 0.880 | 0.881 | 0.866 | 0.872 | 0.860 | 0.881 |
Minimum
| 0.706 | 0.738 | 0.792 | 0.790 | 0.802 | 0.769 | 0.778 | 0.575 | 0.510 | 0.510 |
Maximum
| 1.672 | 1.607 | 1.360 | 1.572 | 1.444 | 1.368 | 1.687 | 1.475 | 1.469 | 1.687 |
Panel B – PTE
| ||||||||||
Score < 1 (Inefficient)
| 23 | 19 | 19 | 19 | 21 | 21 | 21 | 23 | 24 | 21 |
Score > 1
(Super-efficient)
| 9 | 13 | 13 | 13 | 11 | 11 | 11 | 9 | 8 | 11 |
Total
| 32 | 32 | 32 | 32 | 32 | 32 | 32 | 32 | 32 | 32 |
Mean*
| 0.912 | 0.932 | 0.925 | 0.930 | 0.907 | 0.905 | 0.902 | 0.917 | 0.914 | 0.916 |
Minimum
| 0.715 | 0.810 | 0.798 | 0.862 | 0.829 | 0.788 | 0.791 | 0.641 | 0.606 | 0.606 |
Maximum
| 2.200 | 1.977 | 1.556 | 1.574 | 1.462 | 1.399 | 1.704 | 1.479 | 1.472 | 2.200 |
The descriptive statistics and correlations for all the independent variables are presented in Table 3. Panel A of Table 3 presents the descriptive statistics for the Indian firms. The table clearly shows that there is huge variation within different variables in terms of their average, maximum, and minimum values. Panel B of Table 3 reports the coefficient of correlation among different pairs of independent variables. The table highlights that none of the correlations is higher than 0.30. Thus, the current dataset does not suffer from the problem of severe multicollinearity.
Table 3
Descriptive Statistics and Correlations – Independent Variables
Panel A – Descriptive Statistics | ||||||
|---|---|---|---|---|---|---|
AGE
|
SIZE
|
CAPEX
|
ROS
|
LIQ
|
LEVER
| |
Mean
| 3.501 | 7.512 | 0.153 | 0.265 | 1.532 | 0.229 |
Median
| 3.434 | 7.469 | 0.089 | 0.262 | 1.012 | 0.218 |
Maximum
| 4.745 | 11.231 | 1.827 | 0.981 | 11.200 | 0.888 |
Minimum
| 2.303 | 4.185 | 0.003 | -0.594 | 0.077 | 0.000 |
Std. Dev.
| 0.470 | 1.820 | 0.211 | 0.159 | 1.484 | 0.189 |
N
| 288 | 288 | 288 | 288 | 288 | 288 |
Panel B – Correlations
| ||||||
AGE
|
SIZE
|
CAPEX
|
ROS
|
LIQ
|
LEVER
| |
AGE
| 1.000 | |||||
SIZE
| 0.301 | 1.000 | ||||
CAPEX
| -0.097 | 0.118 | 1.000 | |||
ROS
| 0.107 | 0.167 | 0.240 | 1.000 | ||
LIQ
| -0.058 | 0.171 | 0.280 | 0.019 | 1.000 | |
LEVER
| -0.067 | -0.121 | -0.047 | 0.081 | -0.186 | 1.000 |
4.2 Results of Stationarity
The regression methodology is based on the underlying assumption of the stationarity of the dataset. In the case of non-stationarity, the regression estimates can be biased and unreliable. Hence, it is vital to test for the stationarity of the dataset before interpreting the regression results. Since the current dataset is in a panel format, the panel unit-root test of Levin, Lin and Chu has been deployed to test for the stationarity of all the variables. The results of the unit root test are presented in Table 4. The test has the null hypothesis of non-stationarity. Hence, rejection of the null hypothesis implies that the underlying series are stationary. Table 4 confirms that the dataset is stationarity, and hence the regression results can be relied upon.
Table 4
Results of Test of Panel Stationarity
Variable | Levin, Lin & Chu t | p-value |
|---|---|---|
OTE | -5.9513*** | 0.0000 |
PTE | -4.7881*** | 0.0000 |
AGE | -72.7068*** | 0.0000 |
SIZE | -8.1643*** | 0.0000 |
CAPEX | -12.0486*** | 0.0000 |
ROS | -7.8984*** | 0.0000 |
LIQ | -22.0763*** | 0.0000 |
LEVER | -635.8640*** | 0.0000 |
4.3 Dynamic Panel Results
The dynamic panel-based regression results are presented in Tables 5 and 6. Panel A of Table 5 presents the results by considering OTE as the main dependent variable. The table highlights a few interesting findings. It is noticed that the coefficient of the lagged value of OTE is around 0.67. This indicates a high degree of persistence in the efficiency of the firms operating in the hospitality and tourism industry in India. The degree of performance persistence is statistically significant with p < 0.01. Table 5 further highlights that the coefficient of AGE is negative and significant at the 1 per cent level. This shows that older firms are less efficient as compared to younger firms. However, the coefficient of SIZE is positive and significant, thereby indicating superior efficiency for the larger firms. This finding also lends support to the economies of scale argument. It is also observed that higher values of CAPEX, ROS, LIQ, and LEVER are inversely related to firms’ efficiency.
Table 5
Dynamic Panel Regression Results – Dependent Variable – OTE
Panel A: Coefficient Estimation | ||||
|---|---|---|---|---|
Variable
|
Coefficient
|
Std. Error
|
t-Statistic
|
p-value
|
\( {OTE}_{t-1}\)
| 0.6712*** | 0.0192 | 34.9852 | 0.0000 |
\( {AGE}_{t-1}\)
| -0.3013*** | 0.0155 | -19.4815 | 0.0000 |
\( {SIZE}_{t-1}\)
| 0.0377*** | 0.0104 | 3.6369 | 0.0003 |
\( {CAPEX}_{t-1}\)
| -0.0432*** | 0.0097 | -4.4646 | 0.0000 |
\( {ROS}_{t-1}\)
| -0.0239** | 0.0113 | -2.1207 | 0.0351 |
\( {LIQ}_{t-1}\)
| -0.0044*** | 0.0011 | -4.1744 | 0.0000 |
\( {LEVER}_{t-1}\)
| -0.0342** | 0.0162 | -2.1162 | 0.0355 |
J-statistic | 30.7286 | -- | -- | 0.1982 |
Panel B: Residual Diagnosis
| ||||
Test order
|
m-Statistic
|
Rho
|
SE(rho)
|
p-value
|
AR(1) | -2.2992** | -0.8159 | 0.3548 | 0.0215 |
AR(2) | 1.6184 | 0.2015 | 0.1245 | 0.1056 |
Table 6
Dynamic Panel Regression Results – Dependent Variable – PTE
Panel A: Coefficient Estimation | ||||
|---|---|---|---|---|
Variable
|
Coefficient
|
Std. Error
|
t-Statistic
|
p-value
|
\( {PTE}_{t-1}\)
| 0.5640*** | 0.0168 | 33.5768 | 0.0000 |
\( {AGE}_{t-1}\)
| -0.1967*** | 0.0480 | -4.0982 | 0.0001 |
\( {SIZE}_{t-1}\)
| 0.0020 | 0.0142 | 0.1434 | 0.8861 |
\( {CAPEX}_{t-1}\)
| 0.0158* | 0.0082 | 1.9160 | 0.0567 |
\( {ROS}_{t-1}\)
| -0.0399* | 0.0223 | -1.7897 | 0.0749 |
\( {LIQ}_{t-1}\)
| -0.0045*** | 0.0011 | -4.1350 | 0.0001 |
\( {LEVER}_{t-1}\)
| -0.0106 | 0.0214 | -0.4955 | 0.6208 |
J-statistic | 26.1349 | -- | -- | 0.4004 |
Panel B: Residual Diagnosis
| ||||
Test order
|
m-Statistic
|
Rho
|
SE(rho)
|
p-value
|
AR(1) | -2.4315** | -0.7182 | 0.2954 | 0.0150 |
AR(2) | -0.5688 | -0.0782 | 0.1375 | 0.5695 |
The overall model fit of the Arellano and Bond (1991) estimation methodology is tested by analyzing the J-statistics. The model fit is established when the J-statistics are not significant at the conventional levels. Table 5 clearly shows that the J-statistics have a corresponding p-value of more than 19 per cent. Hence, it can be concluded that the instruments deployed are valid and a good model fit has been obtained. Panel B of Table 5 presents the results of the residual diagnosis test. The Arellano and Bond (1991) technique implicitly assume that the error term observes an AR (1) pattern but not an AR (2) pattern. The test of the residuals clearly shows that the assumptions of the technique are fulfilled. Thus, it can be inferred that the model is valid, and the results are robust.
Table 6 presents the results with PTE as the dependent variable. Panel A of the table reports the coefficient estimates, the standard errors, and the corresponding p-values. The table highlights that the lag value of PTE is positive and significant with p < 0.01. In addition, the coefficient of the lagged PTE variable is around 0.56. This further shows that there is a moderate to high level of persistence of efficiency in the hospitality and tourism industry in India. Furthermore, the results of other independent variables are qualitatively similar to those obtained with OTE as the dependent variable. The only noteworthy difference is that the variable SIZE is insignificant at the traditional levels. The table also shows that the J-statistics are insignificant and thus indicate a good model fit. The results of the residual diagnosis, as reported in Panel B of Table 6, further indicate that the assumptions underlying the dynamic panel are valid.
5 Discussion
The results of the present study convey the positive and significant persistence of the financial efficiency of hospitality and tourism firms in India. The results also show that the level of persistence is quite high for both OTE and PTE scores. These results are in sharp contrast to some of the previous studies that have found moderate to low levels of persistence in the efficiency (Chen et al. 2009; Galán and Pollitt 2014). The findings are also in contrast to some of the previous studies that have found a low level of persistence in certain other measures of firm performance such as profitability and market share (Gschwandtner and Hauser 2016).
These results lend support to the argument about the creation of entry and exit barriers by firms to generate sustainable competitive advantage. Thus, the empirical evidence suggests that Indian hospitality and tourism firms are creating entry and exit barriers. The evidence also shows that Indian firms could be exploiting the behavioural inertia of consumers to generate higher revenues; and that this, in turn, is helping certain firms to improve and sustain their financial efficiency. The high degree of persistence of financial efficiency also indicates the pursuance of prudent financial policies by a segment of hospitality and tourism firms. The broad results indicate, furthermore, that emerging market firms behave differently from their counterparts in advanced markets. This shows that evidence on advanced markets may not be valid for firms from emerging markets.
The current results bear certain critical managerial implications for the Indian hospitality and tourism industry. Since the study reveals that the majority of firms are in inefficient regions, it is imperative for hospitality and tourism firms to streamline their financial management policies. This should specifically involve controlling costs and devising methods for the enhancement of revenues. This can be accomplished by an in-depth analysis and rationalization of certain discretionary expenses such as marketing costs and costs of training and development. Furthermore, the results on positive persistence of efficiency highlight that the benefits of increased efficiency can be sustained in the long run. Hence, hospitality and tourism firms should focus on concentrated efforts for increasing and sustaining financial efficiency. This can be achieved in multiple ways. However, the most common modalities for achieving this are through the creation of entry and exit barriers and devising ways of exploiting inertia in consumers’ behaviour. Furthermore, there should be binding incentives for managers to explicitly increase financial efficiency by focusing on higher quality standards and by swiftly responding to changing consumers’ needs.
The findings of the present study also have certain useful implications for regulatory authorities. The findings of the high level of persistence of efficiency reveal that the intensity of competition is quite low in the hospitality and tourism industry in India. Thus, regulators should specifically focus on policies that can enhance the competitive dynamics of the industry. This can be achieved by allowing additional players to operate in the industry. For this to happen, the licensing requirements must be eased. Besides this, the opening of the sector to foreign competition can also go a long way in making the industry more competitive. Moreover, efforts should also be made to promote healthy competition by preventing firms from forming cartels and indulging in other restrictive practices.
6 Conclusions
Given the increasing role of the hospitality industry in the overall economic growth of an economy (Seetanah 2011; Schubert et al. 2011; Sharma et al. 2022), the nature and degree of persistence of the financial efficiency of firms are important considerations in developing a fundamental understanding of legitimate long-term industry growth. The present study analysed the persistence of financial efficiency of firms operating in the hospitality and tourism industry in India. The analysis was carried out for 32 publicly listed firms in the Indian hospitality and tourism industry. The time frame of 2009 to 2017 was considered to estimate the results. The study uses a combination of advanced techniques of DEA and dynamic panel regression analysis, a significant methodological contribution in measuring tourism efficiency. The results confirm significant and positive persistence of firms’ financial efficiency. The results also hint at a moderate to the high level of persistence of efficiency. These results indicate that there is low intensity of competition in the Indian hospitality and tourism sector. The results also signify that Indian hospitality and tourism firms can create certain entry and exit barriers that enable them to sustain their efficiency in the long run. The finding during the study can be read for both, new entrants as well established firms in the sector in the context of inefficient markets. As the results suggest a different response from the firm in the inefficient markets vis-a-vis their counterparts in the developed markets. The study has implications for the policymaker in the sector. The sector needs measures to encourage and protect the competitive dynamics of the industry. The easing up of the licencing, allowing the foreign and local entrants, besides measures against cartelisation. Such measures may make it imperative for the management of the firms in the sector to streamline their financial management policies. An in-depth analysis and rationalization of expenses such as marketing costs and costs of training and development are required to control costs and devise methods for the enhancement of revenues.
There are a few limitations of the current study that need to be mentioned. Although the study is conducted on an emerging economy with high growth in the tourism and hospitality sector, the results of the present work cannot be generalized to other industries due to the limitation of the sample However, there is tremendous scope for gaining more insights by replicating the study for other emerging economies which are experiencing similar growth patterns in the hospitality and tourism industry. The study has also considered one combination of inputs and outputs to generate the efficiency scores. However, future studies can extend the scope of this work by exploring more variables for estimating the efficiency of firms. In addition, adding qualitative inputs and outputs variables in the model, such as quality of services and customer satisfaction, can further increase the depth of the research.
Declarations
Compliance with ethical standards
When carrying out this research project accepted principles of ethical and professional conduct have been followed. The project does not involve human participants and/or animals. The data used in this study is publicly available and there are no restrictions on using the public documents analysed in this study.
Conflict of interest
The research has not to be funded through research grants and financial assistance has not been obtained from any source. The authors declare that they have no conflict of interest.
Consent to Publish
The authors provide consent to publish their details in case the article is accepted.
Open AccessThis article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.