Tourism industry has fuzzy boundaries and highly complex integration and intersection. Tourism statistics is a recognized worldwide statistical problem. Being able to participate in tourism activities, or even engage in tourism work or research, does not necessarily mean that one understands tourism statistics. Tourism activities reflect the emotional management of micro-individuals, while tourism statistics reflect the scientific rationality of the macro-society. Without mastering the connotation, data sources, statistical caliber, and estimation methods of tourism statistical indicators, one will use individual perceptual experience to "take it for granted" in tourism statistics, which often leads to a thousand miles of error. For this reason, it is necessary to further explain the concepts, estimation methods, and regional scales contained in the Tourism Statistical Survey System (hereinafter referred to as the "System").

1. Who is a tourist and how to measure tourists can and can only follow the provisions of the "Regulations"

1. Tourism is an economic activity in a broad sense and cannot be counted in a narrow sense.

Some people think that sightseeing and leisure vacations are tourism. Some scenic spot operators even think that buying tickets to enter scenic spots is tourism, while travel agency practitioners think that buying various products from travel agencies is tourism. This is too narrow. In fact, in tourism statistics, tourists cover a wider range. The "2008 International Tourism Statistics Recommendations" compiled by the United Nations World Tourism Organization states that "tourists are those who travel outside their usual environment for less than one year for the purpose of vacation, leisure and entertainment, visiting relatives and friends, education and training, health and medical treatment, religion, etc." The "National Tourism and Related Industries Statistical Classification (2018)" issued by the National Bureau of Statistics uses the above definition. The "Tourism Statistical Survey System" further clarifies the four technical requirements for the identification of tourists, namely leaving the usual environment, the time and space constraints of 10 kilometers and 6 hours, the motivation for travel, and the lack of an employment relationship, and points out that the "commuting" behavior of regular round-trips between two usual environments for work or study is not within the scope of tourism statistics. Some people take it for granted that they are "counted as one person every day" just because their commute is more than 10 kilometers and 6 hours. Others believe that business trips and visiting relatives in hometowns are not tourism according to their own understanding. These are all cognitive deviations that put personal judgment above international common rules and statistical science.

(II) One “trip” generates multiple “receptions”, which cannot be simply added up vertically

Tourism data is divided into two statistical calibers: "travel" and "reception". The annual domestic tourist number at the national level belongs to the "travel" statistics, and the statistical perspective is a complete recreational behavior, that is, one trip from leaving the usual environment to returning to the usual environment is counted as one person, but during this period, multiple tourist attractions may be visited, or even multiple provinces or cities may be visited. The national holiday tourism statistics belong to the "reception" caliber, and one tourist traveling to one province during the holiday is counted as one person. Local domestic tourists and holiday tourism are both "reception" statistics, and the statistical perspective is the division of tourism activities in regional space, that is, if a tourist visits a province or prefecture-level city, the province or prefecture-level city can be counted as one person. A complete tourist trip may form multiple regional receptions, which cannot be simply compared with the "travel" statistics, and even the reception data at different regional levels cannot be simply compared. The total number of provincial tourism receptions must be greater than the national number of tourists, and "vertical non-additivity" is not necessarily an abnormal phenomenon. Some people on the Internet directly compare the total number of domestic tourists in all provinces with the national number of tourists, which obviously violates data ethics. Of course, the sum of the number of tourists received by each province should not be much greater than the number of tourists in the country. The average number of provinces visited by tourists during short holidays generally does not exceed 1.1, and during long holidays generally does not exceed 1.3. In theory, the difference between the sum of the number of tourists received by each province and the number of tourists in the country should not exceed 30%.

3. National tourism revenue is different from provincial tourism revenue. In theory, the former should be greater than the latter.

Domestic tourism revenue at the national level is the total cost of a complete trip for tourists, including all tourism expenses paid by tourists during the entire travel process, including travel, sightseeing, accommodation, food, shopping, entertainment, and purchase of souvenirs and gifts for relatives, friends and family. In local statistics, provinces only count the expenses of receiving tourists in the province, and provincial "tourism revenue" only includes various expenses incurred by tourists in the province, excluding most of the transportation expenses for tourists' pre-trip preparations and from home to the province. In theory, the total provincial tourism revenue should be less than the national level domestic tourism total revenue statistics.

2. Tourism statistics are not just about “counting people”; data at different regional scales need to be connected

1. The number of tourists is not the number of times they participate in tourism projects, but a complete tourism activity of tourists at the destination.

The number of tourists is the sum of the number of people participating in tourism activities. The "Regulations" also provides the statistical caliber of the number of tourists. "The number of domestic tourists is counted as one person per trip." The number of tourists is a complete tourism activity of tourists at the tourist destination. For local tourism statistics, the number of tourists is the entire process from entering the destination to leaving the destination. It is not a simple sum of tourists in multiple tourist facilities. When counting the number of tourists, tourists who use multiple tourist facilities should be excluded to avoid duplicate statistics. Some districts and counties in a western province use the summary data of scenic spot ticket sales, hotel accommodation and ticketing of typical tourist projects as the number of local tourists received. This unscientific practice is wrong and does not comply with the operational provisions of the "Regulations" and will cause statistical deviations. Professor Bao Jigang's article also mentioned this situation, "If a tourist visits 10 attractions in a destination, he will be counted 10 times, plus one more count for accommodation, so one tourist becomes 11 people. This is mainly because local governments do not have a good grasp of the connotation of tourism statistics and do not fully understand the technical definition of tourism statistics, which leads to the deviation of statistical design and statistical methods from the regulations of the "System". At present, most places strictly implement the "Tourism Statistics Survey System" plan. Even if there are individual irregularities in operation, they are very rare and will not affect the overall tourism data.

2. The regional scale view of tourism statistics has long existed, and longitudinal data can be connected

At present, the regional statistical scale of domestic tourism statistics is clear, and the statistical calibers of different levels at the national, provincial, municipal and county levels are clear. According to the administrative division, it is divided into a four-level tourism statistical system of "national-W-level-municipal-county level". In addition to the national level, which belongs to the statistics of travel caliber, the three levels of province, city and county all belong to the statistical scope of local reception of domestic tourism. The lower-level data is the basis for the upper-level calculation, and the upper-level data is summarized and calculated by the lower-level data. The provincial and municipal data are basically connected with the city and county summary data respectively. Tourists are highly statistical objects. Considering that tourists have visited multiple urban areas or counties in the province within a certain period of time, scientific de-duplication technology is used for the data when calculating. The number of tourists received in the province is the total number of local and foreign tourists who meet the technical definition of tourists. Tourists travel once in the province, which is 1 person, regardless of how many cities they visit, how many scenic spots they visit, how many days they stay, and how many tourism projects they experience; after the provincial government summarizes the city data, it should be divided by the number of cities where tourists visit on average (see Figure 1). Similarly, when the city summarizes the county data, it should also be divided by the number of districts and counties where tourists visit on average. The number of tourists received by the county is the total number of local and foreign tourists that meet the technical definition of tourists. There may be a tourist visiting multiple tourism elements, such as tourist attractions, accommodation units or tourism projects, etc. Different tourism elements are counted as one person each, and the number of tourism elements needs to be used to remove duplicates. Overall, the connection between county-city-provincial tourism statistical data is relatively smooth.

Figure 1: Provincial domestic tourism reception estimation process

(III) The attraction survey helps local tourism reception to “collect all that should be collected” When the “System” was revised in 2017, the “tourist attractions” in the sampling survey of local domestic tourism reception were changed to “tourist attractions”. This is a broad concept, which is the sum of all factors that attract tourists to the destination, including closed scenic spots (parks) and open tourist resources such as farmhouses, historical and cultural blocks, and iconic buildings. The original tourism statistics that took tourist attractions as the survey object did not include tourists in open attractions in the statistics, resulting in overall omissions in the survey and smaller estimated data. For example, suppose that a place has 6,000 overnight tourists from the public security department, and the overnight rate of tourists obtained from the original "tourist attractions" is 30%, then the total number of tourists received by the place is 6,000/0.3=20,000; but if it is enlarged to "tourist attractions", the base of local leisure tourists will increase, and the proportion of tourists who go to open attractions such as suburban rural tours and farmhouses for one-day round trips will increase, which will inevitably lower the overnight rate. Assuming that the overnight rate obtained from the survey is 15% in this case, the total number of tourists received by the place is 6,000/0.15=40,000, which is twice the result of the previous calculation. Taking tourist attractions as the basic data source, the gap between the estimated data and the actual situation is significantly smaller, which is helpful for local tourism reception to be fully included, without omissions, and truly reflect the local tourism development.

III. Key points of the revision of the Tourism Statistics Survey System

(I) Carry out system revision around the situation of statistical supervision On May 14, the 11th Statistical Supervision Group of the National Bureau of Statistics reported the situation of statistical supervision to the Ministry of Culture and Tourism. The supervision group found that the statutory statistical duties of the tourism department were not fully fulfilled, the responsibility for preventing and punishing statistical fraud and falsification was not strict enough, and the quality of statistical data needed to be further improved. And it is required to further perform the statutory statistical duties and do a good job in cultural and tourism statistics in accordance with laws and regulations. Further establish and improve the responsibility and accountability system for preventing and punishing statistical fraud and falsification, and build a line of defense that dares not, cannot, and does not want to falsify. Further strengthen the management of the whole process of statistical data and effectively improve the quality of cultural and tourism statistical data. It is recommended to strengthen statistical verification in the "System", strengthen the management of provincial tourism statistics, and put forward the requirement for provincial tourism authorities to verify downward. Establish a source data quality control mechanism. Tourism statistical agencies at all levels should formulate measures and methods to control the quality of source data to ensure that the data is objective and true; establish a statistical data verification mechanism. Tourism statistical agencies at all levels should conscientiously implement the "Tourism Statistical Survey System", do a good job in statistical basic work, and establish a database and statistical work ledger. The higher-level competent departments should adopt the method of "double random, one public" to verify the statistical data and statistical work of the lower-level departments.

(II) Introducing indicators to measure the quality of tourism development The tourism industry is currently in a stage of high-quality development. Administrative departments’ macroeconomic decisions require more timely, accurate and comprehensive statistical data, and the public also needs matching data to intuitively feel the quality of tourism development. Tourism statistics should reflect the characteristics of high-quality development, but the current reports and indicators in the "System" mainly reflect the scale and structure of the tourism industry, mainly total speed indicators, and indicators reflecting quality are not obvious. Therefore, it is necessary to construct statistical indicators that adapt to high-quality development at this stage to provide data support for the research and judgment of tourism industry reform, development and quality. It is recommended to add a domestic tourism service quality report to the "System", add a comprehensive index of domestic tourism service quality that reflects service levels, and add quality indicators such as overnight rates and average number of travel days per capita to comprehensively reflect the connotation of high-quality development.

(III) Standardizing the application of big data statistics In recent years, the use of big data for passenger flow and consumption statistics has become an important trend. The application of big data has made tourism statistics more timely and accurate. Various places are also competing to use big data statistics to replace traditional statistical data, but some places have not eliminated big data according to the technical definition of tourists, resulting in data distortion.

In order to improve the quality of tourism statistics, it is necessary to standardize the application standards of big data statistics. It is recommended to add requirements for the application of relevant data statistics in the "Regulations", and put forward normative requirements for the identification of tourists' usual environment, passenger flow monitoring algorithms, the combination of big data and traditional surveys, and the protection of tourists' privacy. Authors: Yang Suzhen, Ma Yiliang