B.2.2. Determining the national economy involved in some specific transactions of visitors
4.24. In general, determining which part of an international visitor’s expenditure corresponds to inbound/outbound tourism expenditure, which part corresponds to domestic tourism expenditure, and which part should be excluded is not a very complex process: visitors themselves can often easily identify where their expenditures were made and the country of residence of the sellers or providers.
4.25. For a few areas of expenditure, however, such determinations are less straightforward, in particular for the following three :
4.26. Identifying the provider of air transportation and its country of residence and relating it to an international visitor, can be problematic.
4.27. This can be attributed to both the multiplicity of arrangements that exist among air service providers and the existence of transnational companies for air (and rail) transportation.
4.28. An international visitor may book air travel on more than one airline. These airlines can be resident companies in the country of origin, or in the country of destination, or even in a third country or in two or all three of them. If a separate ticket is issued for each segment using a different carrier, it is easy to determine their respective country or countries of residence and the amounts paid to each. However, it is also common for one airline (the validating or plating carrier) to sell a single ticket for travel on many other resident or non‑resident airlines (for segments of the journey that are not served by the validating airline). This practice is called interlining. Where there are no interlining arrangements between companies, separate tickets have to be issued.
4.29. It is not easy to distribute the total amount paid for a ticket among the different legs of the journey and the different airlines involved: this amount usually reflects different price categories, different taxes levied by the countries of origin and/or destination and, in some cases, a commission for the travel arranger (formal travel agent or online operator), whose country of residence it can also be difficult to determine. Survey respondents cannot be expected to sort all this out, nor do they have the detailed information necessary to do so.
4.30. Another complicating factor is that airlines operate in certain segments under code‑share agreements. A code‑share flight is a commercial flight operated by one airline (the operating carrier) but marketed by others (the marketing carriers). The operating carrier provides the service to the visitor and pays the marketing carrier a certain share of the revenues earned, based on complex formulas and bilateral arrangements specific to each segment. What needs to be identified for tourism purposes is the country of residence of the operating carrier: the one providing the service to the visitor. For Balance‑of‑Payments purposes, it becomes necessary also to identify the countries of residence of the marketing carriers involved (as they are remunerated by the operating carrier). Here again, the visitors surveyed cannot be expected to make those identifications.
4.31. Additionally, airlines and railway companies might operate as multi‑territory enterprises, defined as enterprises that “operate as a seamless operation over more than one economic territory” (see BMP6, paras. 4.41-4.44). Although such an enterprise has substantial activity in more than one economic territory, it is run as an indivisible operation with no separate accounts or decisions, so that no separate branches can be identified” (ibid.).
4.32. Fortunately, compilers of tourism statistics are not the only compilers interested in these issues: National Accounts and Balance‑of‑Payments compilers deal with similar questions and face precisely the same difficulties, which means that cooperation in this field is essential.
4.33. For issues relevant to their countries, many Balance‑of‑Payments compilers have developed methods of estimation based on information collected from railway companies or airlines. It is important that compilers of tourism statistics understand and participate in the estimation procedure so that they can apply the results properly. The data derived from these procedures, however, may not be sufficiently detailed for the purpose of making specific adjustments to the information collected from surveys of visitors and assigning them to the different categories of tourism expenditure, because, the Balance of Payments is concerned only with global data. This distribution will possibly require some types of adjustments in the data and the final presentation within the table recommended by the TSA (if relevant). It is also to be remembered that tourism statistics (which, basically, are compiled for the purpose of setting up a TSA) use the net valuation principle for reservation services (see IRTS 2008, paras. 6.46-6.54), which is not the case. Either in the Balance of Payments or, necessarily, in National Accounts (see TSA: RMF 2008, paras. 3.21-3.24).
Imputation in the Statistics Canada International Travel Survey Program
In the International Travel Survey (ITS), missing transportation fares and/or total travel expenses are imputed when the other fields of the questionnaire are valid. The imputed values for such a questionnaire are calculated from the mean of the corresponding fields of the other questionnaires sharing some identical key characteristics with the given questionnaire.
Target populations (American, overseas and Canadian international travellers) are partitioned into Port Factor Groups (PFGs), based on selected traveller characteristics, such as country of residence, mode of entry and duration of stay. Total imputation (i.e., imputation of complete questionnaires) is carried out for all PFGs or strata that are outside the scope of questionnaire distribution. There are 120 Canadian and American PFGs for which Statistics Canada never receives questionnaires. These imputed questionnaires accounted for only 4.4 percent of all United States travellers to Canada and 1.2 percent of Canadian residents travelling outside Canada.
Imputation of questionnaires is required only for Canadian and United States travellers. Total imputation is also performed for any in‑scope PFG for which an insufficient number of questionnaires have been received for the quarter. In these instances, all the questionnaires from the same quarter of the previous year that belong to the PFG are brought forward and added to the sample of that PFG for the reference quarter. If necessary, additional total imputation is also performed for United States car travellers by States of origin to meet minimum requirements (combination of minimum number of questionnaires and maximum weight) based on the frontier counts.
Source: Statistics Canada.
4.34. A package tour consists of a “tourism product” developed and provided by a tour operator, which sells it directly or through travel agencies to would‑be travellers (see para. 4.52). Travellers on package tours receive a combination of trip‑related products and tourism services, including international and national transportation, accommodation, food and beverage services, sightseeing and entertainment. Package characteristics vary. Some are “off the shelf”, while others are customized in response to specific traveller requests through combining elements previously negotiated and selected by the packager (and purchased from designated providers) who anticipated a demand for them and often places itself at risk (IRTS 2008, 6.46-6.54; and TSA: RMF 2008, paras. 3.21-3.24).
4.35. Some package tours are limited to one or more domestic destinations, i.e., to places within the economic territory of the packager’s country of residence. Many, however, include travel to foreign destinations, i.e., destinations outside the packager’s economic territory, or a combination of domestic and foreign travel.
4.36. The statistical treatment of package tours is complicated. Package tours usually involve a number of different products and services which are supplied by different providers. In addition, these companies might reside in various countries: some in the country of the purchaser, and some in the destination country or even in a third country. Tourism statistics must take this constellation into account if the extent of the involvement of different economies is to be correctly determined.
4.37. Usually, if the package entails domestic travel, it will be assumed that all goods and services included in the package have been provided by resident producers. However, this assumption may need to be reviewed in the future because of the call for open skies in the context of increasingly globalized economies.
4.38. A packager might sell its product not only to residents of the economy within which it operates, but also directly to non‑residents, through a travel agency (which may or may not reside in the economy of reference) or associated packagers or travel agencies operating in other countries, or even directly to would‑be travellers over the Internet.
4.39. Visitors usually receive information only about the total value of a package and about its components. They will usually not know where the providers reside or how much each service purchased within the package might cost, let alone how those costs might be apportioned among product packagers and marketers.
4.40. As a consequence, the total cost of packages reported by visitors needs to be broken down by analysts using information collected directly from the packagers themselves or from the travel agencies that sell packages to the public. There are two approaches available to countries in this regard. One would be to obtain information by means of (modules added to) supply surveys, as explained in the example below. The alternative would be to obtain information by means of a “case study” sample of selected key travel agent establishments. In either case, compilers will need to be well prepared if they seek to earn the trust and secure the collaboration of data providers.
4.41. A challenge that needs to be considered irrespective of the approach is that of differentiating between wholesalers and retailers, who may either purchase pre‑packaged tours or package them themselves, and, between their respective margins on the various products. As tourism statistics encompass matters involving only direct contact with visitors, the retailers’ margin but not the wholesalers’ should be included.
Box IV.2 T
reatment of package tours in the Balance of Payments: example of Austria
In Austria, the main source for collecting data on expenditure on goods and services for outbound trips is the quarterly sample survey dealing with holiday and business trips. The outcome of this survey encompasses the expenditure for outbound travel.
However, the amount of expenditure also includes components that must be allocated to domestic tourism (e.g., the travel agency fee for services rendered). Thus, owing to conceptual and methodological factors, these arises a discrepancy with respect to the Balance‑of‑Payments Travel item (T-BoP), which covers the collection of expenditure abroad. Therefore, these domestic components have to be isolated in order for these to be consistency with the recommendations of the International Monetary Fund (IMF) as the physical travel destination may not coincide with the actual monetary flow, as in the case where tour operators located in third countries are involved. Indeed, in the many cases where tour operators located in Germany or the United Kingdom organize package tours for Austrian residents, the monetary flows do not coincide with the physical travel destinations but rather with the official location of the tour operators. According to IMF recommendations, passenger transport has to be accounted for separately and is not part of the Travel item in the Balance of Payments as conceived in its narrower sense. Therefore, as package products in many cases include a transportation item, it is necessary to separate these products into their components to at least isolate the international transport item.
Given that the sample survey is demand‑related, necessary adjustments relating to package products need to be made namely:
As these operations cannot be accomplished by using only the information from the sample survey, additional information from business providers is needed. This supply‑oriented information is used to establish a disaggregation model so as to enable the expenditure amount of package tours to be separated into sub‑aggregates, which can then be adjusted according to T-BoP requirements.
It should be emphasized that this method is adapted to compilation conditions in Austria and should therefore be viewed as one input into further discussion on the issue of package tours.
More information on how Statistics Austria treats package tours within the travel item of the Balance of Payments can be found in the Eurostat Methodological Manual for Tourism Statistics, version 1.2.
Source: Statistics Austria (2013).
Statistics on products in the services sector: example of Spain
The form of the organization of a trip is a characteristic of the trip which is usually collected in demand surveys and enables an evaluation of the volume and importance of the package tours. In Spain, this variable is included in household surveys (La encuesta de monumentos turísticos de los Españoles (FAMILITUR)) and in frontier surveys (La encuesta de movimientos turísticos en fronteras (FRONTUR) and La encuesta de gasto turístico (EGATUR)).
The information related to package tours, collected from the demand side, is highly useful. However, the supply survey approach, where this information is directly provided by the companies that develop these products, should not be neglected.
Information from the supply side is best for elaborating various elements of great significance in characterizing tour packages, such as their composition (services included) and their marketing circuit.
In order to obtain this information, Spain’s National Statistics Institute has developed a specific module for travel agencies through on which certain data related to package tours are collected. This survey is included in Statistics on Products in the Service Sector, which is part of the Annual Services Survey.
For travel agency and tour operator activities, the available information is divided into three sections:
The data derived from this survey can be applied to some of the tasks associated with the TSA framework (e.g., the unbundling of package tours).
4.42. Visitors often purchase services for and during their trips through intermediaries, most commonly, travel agencies (physical or Internet‑based) but, often, also including reservation services (e.g., for shows and hotels). The transaction is usually conducted on a commission or fee basis (either explicitly, with individually invoices, or implicitly, as part of the total price charged).
4.43. For the sake of homogeneity, and because reservation services do benefit users, it is recommended that, as with intermediated services, in all cases, visitors who have used reservation services should be recorded as having done so (see IRTS 2008, paras. 6.46-6.54; and TSA: RMF2008, paras. 3.21-3.24).
4.44. Besides the difficult issue of valuing reservation services, there is the challenge of determining the country of residence of such providers, so that consumption of their services can be assigned to the proper category of tourist expenditure.
4.45. Usually, the adjustment will be carried out only in respect of reservation services provided by travel agencies for international air transportation, packages and cruises. It is assumed that when those services are purchased in the traveller’s country of residence, the travel agency resides in that 250 country and that the value of its services represent a fixed percentage of the sale value based on the fees currently charged by local travel agencies.
4.46. It should be noted that, if available, the information for making such adjustments will be only of a global nature, so that, usually, the type of adjustments discussed cannot be made on an individual basis. Rather, adjustments will tend to be made on an aggregated basis, possibly to all relevant visits and for data over a given period (e.g., a full year’s data).
4.47. The increasing use of the Internet to book travel, accommodation and events, using both resident and non‑resident intermediaries, poses new measurement difficulties. A possible solution is the use of credit card payment databases, as credit cards are usually used for such transactions (see paras 4.65-4.66), although this source will not always provide the level of detail required for goods and services and industries.
 World Tourism Organization (2004c), “Clarifying the treatment of travel agency, tour operator, travel agency services and package tours in SNA, Balance of Payments and TSA and their mutual relationship”, Enzo Paci Papers on Measuring the Economic Significance of Tourism, vol. 4, pp. 151-175 (online), available at: http://statistics.unwto.org/sites/all/files/docpdf/clarifying.pdf (30-05-2014).