Total factor productivity and regional difference of tourism industry of listed companies

Tourism industry has become a breakthrough point for China’s economic growth. As of 2019, the tourism economy has been maintaining a faster growth than the GDP growth rate, the entry and exit market has been developing steadily, the structural reform on the supply side has been effective, and the per capita income of China’s residents has grown significantly. In this context, this paper selects fifteen listed tourism companies in China as samples, uses the Malmquist index method in DEA to measure the total factor productivity of these fifteen companies in China as samples, analyzes their resource allocation efficiency, and tests the robustness of the empirical results. Three conclusions are obtained: from the overall perspective, the total factor productivity of fifteen listed companies as a whole shows an upward trend during the 5-year period, mainly due to the technological progress; from the year perspective, the Malmquist productivity index of China’s tourism industry shows a turnaround change during the 5-year period, and the operational efficiency shows regional differences; the fluctuations of various listed tourism companies in China are influenced by the environment and policies.


Introduction
With China's characteristic socialism into a new era, China took a new historical era.The development of tourism industry is facing the increasing quality of people travel temporarily behind the contradiction between supply and tourism demand, tourism consumption increasingly rich, tourism industry has become a breakthrough point of economic growth in China (Zhang and Lin 2018).Apart from 2020, when the tourism industry suffered a big blow due to the epidemic, the tourism economy continued to maintain a high growth rate higher than GDP until 2019.In 2017, the domestic tourism market grew rapidly, the domestic and foreign markets developed steadily, and the supply-side structural reform achieved remarkable results.Domestic tourists made 5.01 billion trips, up 12.8% over the same period last year.The number of inbound and outbound trips reached 270 million, up 3.7% year on year; Total tourism revenue reached 5.40 trillion yuan, up 15.1 percent.According to preliminary estimates, the overall contribution of tourism to China's GDP was 9.13 trillion yuan, accounting for 11.04% of the country's total GDP.28.25 million people were directly employed by tourism, and 79.9 million were directly and indirectly employed by tourism, accounting for 10.28 percent of the total employed population in China.It further shows that the per capita income of Chinese residents has increased dramatically, the income structure is becoming more reasonable, and the further narrowing of the national income gap makes the middle and low income groups also become the main force of tourism consumption.Therefore, the base of purchaser groups is increasing, and the level of client groups is constantly improving and developing.In addition, while the market of traditional tourism products is expanding and improving, leisure tourism is rising rapidly, medical tourism, prize tourism, competition tourism, coastal tourism, cruise tourism, adventure tourism and other emerging industries are developing rapidly, and consumer groups are larger.Under the common influence of the above industries, China's tourism trend should continue to grow and develop.
Therefore, the motivation of this study is to explore the total factor productivity of listed companies in China's tourism industry from the theoretical and empirical perspectives, and to clarify the future direction of reform and development of tourism industry by further analyzing and discussing the problems that exist in the tourism industry.
The contribution of this paper is to construct a system of input-output indicators for listed tourism companies by reading relevant literature, and to process the input and output data of listed tourism companies in China using DEAP software to analyse the patterns of changes in Malmquist productivity of tourism companies and the reasons for their changes, and to put forward several feasible suggestions that can promote the development of the tourism industry in China.The rest of the paper is organized as follows.Section 2 reviews the domestic and international studies on the issue of total factor productivity in tourism, Sect. 3 introduces the chosen model and data processing, and Sect. 4 interprets the empirical results from the overall perspective, the firm perspective, and the year perspective, respectively.Section 5 concludes the whole paper.

Research on total factor productivity measurement
The study of total factor productivity began with the British economist Adam.Smith.Solow (1957) published "Technological Change and the Aggregate Production Function", which is considered to be the first study of total factor productivity, and which took assets and labour as the two major input factors and created a function in which the growth in total output was reduced by the growth in output due to capital and labour.The remaining value is considered to be the contribution of technological progress to output.In terms of theoretical and practical discussions on productivity measurement, scholars have expanded their research on the measurement of total factor productivity.Aurélie et al. (2015) find that the contribution of total factor productivity and capital growth varied significantly across countries during the global financial crisis, reflecting the diversity of the crisis, and that measuring total factor productivity is a challenging issue.Li et al. (2020) calculated China's tourism GTEP from 2007 to 2018 using a data envelopment analysis (DEA) model and ML index measures found that pure technical efficiency has been greater than 1 since 2014, indicating that China's tourism industry has entered a phase of change and promotion.At the same time, there are significant spatial differences in China's tourism GTFP, with the overall pattern of the strongest in the east and weakest in the west remaining unchanged.Abou Hamia (2020) argues that total factor productivity captures the growth dynamics of factors of production such as technology, management and knowledge and their contribution to economic development, and that the contribution of technological progress will be greater in the long run compared to technical efficiency.

A study on the factors influencing total factor
productivity in tourism Ted (2003) found that tourism efficiency in cities is driven by multiple factors, with economic level, resource endowment and transport conditions having different effects on diverse cities and levels of development.Barros (2006) found that capital, labour, sales and M&A activities were the main factors affecting their efficiency.Köksal (2007) used data envelopment analysis to study the differences in efficiency between different types of efficiency differences between different types of travel agents and found that inefficiencies could be solved by adjusting the input structure.Peypoch (2010) conducted a comparative analysis of Moroccan hotels and identified the importance of technological change in improving productivity.Assaf (2011) explores the relationship between tourism industry agglomeration and total factor productivity in tourism.The results found that there is a significant positive relationship between industry agglomeration and total factor productivity, and the relationship is mainly enhanced through technical efficiency.This is mainly through technical efficiency.Levenko (2019) identified purely technical inefficiencies and strong regional imbalances as issues that need to be addressed in the tourism development process, with significant room for improvement in capital and inputs, and technological innovation being crucial.Martín et al. (2017) concluded that regional financial conditions and the size of tourism resources have a positive contribution to tourism efficiency, and that the total number of travel agencies shows a negative correlation with tourism efficiency.Song (2017) analysed the development of TFP in China's tourism industry and took the attitude that tourism TFP as a whole is characterised by rough development, but that the growth of TFP in the national tourism industry is mainly due to the improvement of technical efficiency.Ramírez-Hurtado (2017) incorporates environmental factors such as carbon emissions, energy consumption and sulfur dioxide emissions into the indicator system of total factor production in tourism and held the view that these both would contribute to the sustainable development of tourism.Kularatne (2019) measured the extent to which environmental measures impact on efficiency gains in tourism and found that energy efficiency and waste management measures can have a positive impact on total efficiency.Yin and Li (2021) concluded that science and technology innovation had a significant propulsive effect on total factor productivity in tourism, and that lagging industrial structure upgrading had a favorable impact on total factor productivity in tourism.Wang (2021) found that total factor productivity in China's tourism industry fluctuates considerably, showing alternating "waves and valleys"; the main driver of tourism revenue growth is the increase in tourism resource inputs.

Research methodology on total factor productivity in tourism
Scholars have used various research methods such as Data Envelopment Analysis (DEA) and Stochastic Frontier Approach (SFA) in their studies, rooted in the efficiency of the tourism industry, to conduct in-depth studies on the sub-sectors and build models of the influencing factors.Assaf (2011) explores the relationship between tourism industry agglomeration and total factor productivity in tourism.The results found that there is a significant positive relationship between industry agglomeration and total factor productivity, and the relationship is mainly enhanced through technical efficiency.This is mainly through technical efficiency.Paço (2015) explains the mechanism of the relationship between ICT and efficiency in the hospitality industry through the DEA-MPI approach.Moutinho et al. (2015) found that total factor productivity is reduced when the strength of the environmental regime is weak due to firms' low incentives for firms to innovate technologically, resulting in lower.Amado (2017) correctly assessed the impact of privatisation on the efficiency of hotels in Portugal using a DEA model.Ren et al. (2017) measured the static and dynamic measures of China's listed tourism companies based on the BCC model in the data envelopment technique and the Malmquist index method, and studied the management efficiency of China's listed tourism companies using dual inputs and dual outputs, and found that the total factor productivity of China's listed tourism companies was mainly attributed to the improvement in the level of technical efficiency, while technological progress contributed little.George Assaf and Tsionas (2018) uses a Bayesian approach to estimate total factor productivity and make cross-country comparisons, taking into account the heterogeneity of tourism destinations and the potential endogeneity of inputs.Chatzimichael (2018) uses a stochastic frontier approach to measure total factor productivity in the hotel industry in 25 European countries and makes cross-country comparisons.Qu et al. (2020) used the three-stage DEA-Malmquist index method to measure the efficiency of China's tourism industry and found that tourism input efficiency was underestimated in the western region and overestimated in the eastern and central regions.External environmental factors had a significant influence on total factor productivity in tourism.
3 Model framework

Model specification
Malmquist productivity index studies generally use the non-parametric DEA-Malmquist index method proposed by Fare et al. (1994) as follows: M 0 in Eq. ( 1) refers to the Malmquist productivity index, with (x t+1 , y t+1 ) and ( x t , y t ) denoting inputs and outputs in periods t + 1 and t, respectively; and denotes the distance functions in periods t and t + 1, and respectively, when using period t as a reference.When taking period t as reference, we get: According to the above analysis process, if we take period t + 1 as the reference, we can get: Based on the geometric mean of the two Malmquist productivity indices above, the change in productivity from period t to t + 1 can be measured as: (1) (2) 4) refers to the technological progress index, PC is the pure technical efficiency index and SC is the scale efficiency index.

Index selection
The operational efficiency of listed companies in tourism industry mainly reflects the relationship between inputs and outputs, and the operational efficiency index system is mainly divided into two categories of input and output indicators.In this paper, we consider the relative independence among input indicators, the correlation among output indicators, the importance of output elements, the relevance of indicators, and the availability and accuracy of data to select indicators.In constructing the index system for evaluating the operational efficiency of listed companies in the tourism industry, the selection of input and output indicators in this paper is mainly based on the Study on Total Factor Productivity of Listed Companies in China's Tourism Industry, while the relevant descriptions in Geng (2012) and other literature are also referred to "total operating revenue" and "net profit" are used as output indicators, and "total assets", "selling costs", "management costs" and "Finance costs" are used as input indicators.Among them, total assets represent the economic scale of the listed company, management expenses reflect the management status of employees, selling expenses represent the expenses incurred in the process of providing services, and financial expenses reflect the expenses incurred by the enterprise to raise funds for production and operation.The total operating revenue and net profit indicators can reflect the scale of output and economic efficiency of the enterprise, and the final determination of the listed company's tourism inputoutput indicator system is as follows (Table 1).

Data acquisition and processing
The data in this paper are obtained from the annual reports of individual companies in the Guotaian database (CSMAR), Flush website, and Juchao information website, and the study period is 5 years.The data processing software used is DEAP2.1.Fifteen listed companies in the tourism industry in China are selected for this study: Xi'an Tourism, Lijiang Tourism, Zhongxin Tourism, Jiuhua Tourism, Beijing Culture, CITS, Jinjiang, CYT, Changbai Mountain, Zhangjiajie, Emeishan A, CITS United, Jinling Hotel, Tibet Tourism, and Beibuwan Tourism.
The reasons for the selection are: (1) the selected listed tourism companies include most of the more influential companies in China, which can basically reflect the overall situation of China's tourism industry; (2) the scope of operation of the selected tourism companies is mostly nationwide, and some of the scope of operation includes overseas, the external business environment they face is generally not very different, therefore, the comparability between individual tourism companies is relatively strong; (3) the data interval is 2016-2020, which helps to comprehensively examine the relevant development of the entire tourism industry in the last 5 years; (4) All data are collected from the publicly available annual reports of listed companies.As the financial reports are prepared in accordance with uniform accounting standards and are subject to audit by social intermediaries, the indicators are more comparable in terms of data calibre.
Among them, since the DEA model requires positive data for inputs and outputs, and there is a negative case of net profit in output indicators, so the data need to be pre-processed by dimensionless processing method, as follows.
The a and b values are the maximum and minimum values of the indices respectively.x represents the standard value of each raw value after processing in a dimensionless way.The dimensionless treatment excludes the case where the optimal solution is zero, maps the data to the (0, 1) dimensionless interval, does not change the relative relationship between the decision units and does not affect the evaluation results.

Analysis of the results of changes in the Malmquist index 4.1.1 Malmquist index visual analysis
By analyzing the calculation results in Tables 2 and 3, the following basic judgments can be drawn for the fifteen listed tourism companies in China during the period 2016-2020.
(1) From an overall perspective, Overall, the total factor productivity of the 15 listed tourism companies showed an overall upward trend over the 5-year period, as demonstrated by an annual aver- age Malmquist productivity index of 1.032 for all companies over the 5-year period, or an average increase of 3.2%.The main reason for this is technological progress, as demonstrated by the fact that the annual average technological change index for all companies over the 5-year period was 1.042, representing the largest increase of 4.2% on average.This compares with an index of change in annual average technical efficiency over the 5-year period of 0.990, which represents an average decrease of 1%, an index of change in pure technical efficiency, which represents an average decrease of 1.3%, and an index of change in scale efficiency, which represents an increase of 3%.
(2) From the company's point of view, most of the companies show an upward trend in total factor productivity, among which, Beitou Travel Bay and CITS United have the most obvious upward trend with 1.520 and 1.222 respectively.The worst performing company is Beijing Culture, with a Malmquist productivity index of only 0.940.The tourism indices of Changbai Mountain and Zhongxin are also relatively low, with values of 0.942 and 0.951, respectively.From the perspective of the reasons, they are mainly due to slow technological progress.(3) From a year perspective, the Malmquist productivity index for tourism companies shows a radical changes, as shown in Fig. 1.The Malmquist productivity index has increased significantly through 2019, with the Malmquist productivity index reaching a 5-year high of 1.243 in 2018-2019.2020 saw a decline to 0.925, the lowest value in the last five years, indicating that the impact of the epidemic and other factors on tourism is indeed significant.

Malmquist productivity index ranking analysis
Based on further analysis of the DEA-Malmquist analysis and index decomposition of the operating data of the above listed tourism companies, it can be found that most of the companies have fluctuations in their development and operation, and the fluctuations are more dramatic.For example, Beijing Culture, whose Malmquist during 2016-2017 index ranked first with a Malmquist index of 1.111, which has high productivity.The index decomposition found that its technological progress is the main reason for its extremely high productivity.

Analysis of the results of changes in the Malmquist index
The above analysis shows that the Malmquist productivity index for our listed tourism companies shows an upward trend over the 5-year period 2016-2020.The main reason for the rise in the productivity index is technological progress (the technological change indicator), with the average technological change index rising by 4.2% over the 5-year period.The technological progress was particularly strong in 2018-2019, with an average technological change index of 1.180 for the fifteen tourism companies, an average increase of 18%.The Technical Efficiency Change Index, on the other hand, fell by an average of 1% over the 5-year period and was the worst performer of the four indicators.The index of change in scale efficiency and the index of change in technical efficiency remained at around 1.000, rising but not as much as the index of technological progress on the Malmquist productivity index.
And the indicators of technological change, on the other hand, mainly refer to intangible factors other than input factors that have an impact on output, generally considered as technological progress, organisational innovation, etc., and can of course also include the external environment, such as the market environment, different ways of thinking, policy guidance.We can therefore consider the market environment and changes in the mindset of travellers.For one thing, the tourism market is expanding rapidly, China's GDP per capita is growing rapidly, and the disposable income of residents is also rising, so more and more residents are choosing tourism as a basic part of their lives.For another, the level of population ageing is increasing, and the tourism consumer population no longer contains only young people, but gradually develops in the direction of multi-level ageing.The explosive growth of the tourism population has also provided a broader market for the tourism industry and has contributed to the development of the tourism industry and the surrounding industries in a radial manner.Finally, a series of policy guidelines on the development and orientation of the tourism industry have been issued, not only to clarify the direction and training measures for the tourism industry, but also to allow more tourists to enjoy safer, more convenient and modern services, providing guidance for the development of the tourism industry and tourism companies.

Analysis of Malmquist index operational efficiency results for four economic regions
The above-mentioned companies were collected and organized according to the criteria of the four major economic regions in China.The four major economic regions are the eastern region (Nine regions regions including Beijing, Tianjin and Hebei), the central region (six regions including Hunan and Anhui), the western region (Six regions including Chongqing and Sichuan) and the northeastern region (Three regions including Liaoning, Jilin and Heilongjiang).The analysis in Table 5 shows that there are significant differences in the operational efficiency of the tourism industry in the four major regions, enabling the following judgments to be made: (1) The 5-year average Malmquist index for the tourism industry of listed companies in the four economic regions is greater than 1, except for the northeast region, indicating that the level of operational efficiency of the tourism industry in the four major economic regions has increased over the past 5 years.(2) Although the efficiency of tourism operations in the eastern, central and western regions all increased from 2016 to 2020, their development was very unstable and did not show a solid upward trend.In the case of the Central Region, its operational efficiency was more prominent in 2017-2018, when the index reached 1.3000, but then trended downwards in 2019, with an index value of just 0.947.(3) Among the four regions, the western region has the best tourism operating efficiency, with an average Malmquist of 1.114.The eastern and central regions have comparable levels of development, with an average Malmquist index difference of only 0.036, or 1.038 and 1.074 respectively.The northeastern region has the worst development, with an average Malmquist of less than 1, or 0.898.
In terms of economic base and policy, the gap between the east and the mid-west and other regions is mainly due to the fact that the eastern region has always had a higher level of economic development than other regions, and therefore receives better resources and economic support in the development of tourism than other regions.At the same time, as the efficiency of the use of existing tourism resources reaches a certain level, the marginal benefits decline and a bottleneck is reached.Excess resource input is therefore instead counter-productive to tourism productivity.On the other hand, compared to other regions, the eastern region is under greater pressure from international convergence and external competition, and this double pressure has led to an urgent need for the tourism industry to face industrial structural transformation.The central and western region, have a weak foundation, and although they are blessed with a number of unique landscapes and ecological environments, their high quality tourism resources have not been effectively explored and utilised due to long-standing constraints such as transportation closures.However, with the support of a series of favourable policies such as the "Rise of Central China" and "Western Development", and with the transfer of tourism resources and tourism technology to the west, the central and western regions have more obvious advantages in tourism.

Robustness tests and further analysis
In this paper, we choose to do a robustness check on the research results from the variables, using "R&D costs" instead of "finance costs" as an input indicator to measure the new total factor productivity, leaving the design of the estimation method and other variables unchanged and following the previous.The empirical test was repeated in accordance with the previous operation, and the dynamic trend of the total factor productivity index was obtained, which is more consistent with the previous results (the results are shown in Table 6).The results show that the total factor productivity of listed companies in the Chinese tourism industry shows an upward trend during the observation period, with a technical efficiency index of 0.910 and a technological progress index of 1.062, indicating that the increase in the total factor productivity index is mainly influenced by technological progress, and that the change in the technological progress index is the main reason for the change in the total factor productivity index.

Conclusion
Based on the previous empirical analysis, the following research conclusions can be drawn for the situation related to the total factor productivity of listed tourism companies in China since 2016.First, from an overall perspective, the Malmquist Productivity Index for our listed tourism companies is generally on an upward trend, with the Malmquist Productivity Index rising by an average of 3.2% over the 5-year period.In terms of causes, the rise in the technical change index was the most significant contributor to the rise in total factor productivity, with the average annual technical change index for all companies rising by an average of 4.2%.The index of change in pure technical efficiency and the index of change in scale efficiency remained broadly unchanged overall, at around 1.000.The technical efficiency index of change fell by an average of 1% over the 5-year period, with room to rise.In the current era of big data, big data technology can be put to good use to deeply mine and correlate the massive amount of tourism-related data and build a comprehensive information service platform for different groups in order to provide accurate services to the government, tourists and others.
Secondly, from a yearly perspective, the Malmquist productivity index for China's tourism industry has shown a turnaround in the 5-year period, with the highest Malmquist productivity index value of 1.243 and the lowest value of 0.925 in the last 5 years in addition, there are regional differences in operational efficiency, with the worst development in the northeast, where the average Malmquist is less than 1, at 0.898 for different strategies are needed to improve the total factor productivity of regional tourism.The eastern region needs to further align itself with international tourism development, introduce advanced tourism business management experience and take the lead in completing the 'wisdom' and 'intelligence' of tourism businesses.For the central and western regions, they should make full use of existing policy advantages, consider the development of urban tourism in a holistic manner, consider the development of tourism in the long term, and develop strategic plans that combine tourism development and destination construction to deepen tourism industry integration.Thirdly, the fluctuations of various listed tourism companies in China are strongly influenced by the environment and policies.For example, in 2019-2020, most regions and companies experienced a significant decrease in efficiency, as in the case of the Northeast region where the operational efficiency index fell to 0.547 in 2020, which is related to the epidemic outbreak that occurred in 2020.Therefore, all types of tourism companies should develop risk contingency plans and adjust their business strategies in a timely manner to improve the ability of tourism companies to adapt to the environment.In addition, the government should continue to play a leading role in promoting the longterm sustainability of tourism and the transformation of the growth pattern from a crude to an intensive model.
The conclusions of this study are broadly consistent with those of some of the literature content, mainly due to the selection of sample companies and the selection of input and output indicators, as well as the selection of roughly the same number of years for analysis.This shows that the research analysis in this paper has a certain degree of reliability and scientific validity.At the same time, there are some shortcomings in this paper.Firstly, in the analysis of the empirical research on the total factor productivity of listed tourism companies in China, the method of collecting input and output data indicators and the time of selection in this paper are limited, which has some influence on the reliability and rigour of the research results.Secondly, some more in-depth insights are lacking in the analysis of the causes of the changes in the Malmquist Index of listed tourism companies in China.In view of the above shortcomings, future research could further deepen the analytical system of tourism factor productivity measurement.Although there is currently no uniform definition of the indicator system for measuring tourism total factor productivity in academia, it is still in the process of continuous improvement, and the selection of the indicator system can be better improved based on different analyses.In addition, further analysis can be done on the mechanism of the spatial and temporal evolution pattern of tourism total factor productivity, and the factors affecting the total factor survival rate of tourism in the four regions can be explored from a quantitative perspective.

Table 1
Input-output index system of listed tourism companies

Table 2
Total factor productivity change index of listed tourism companies and its decomposition

Table 3
Total factor productivity change index of listed tourism companies and its decomposition Chart of changes in the Total Factor Productivity Index of listed tourism companies

Table 4
Malmquist productivity index ranking of listed tourism companies in China

Table 5
Changes in the efficiency index of tourism operations of listed companies in the four major economic regions

Table 6
Total factor productivity change indices and their decomposition for listed tourism companies