3.29. At what point in time should the traveller entering a country be counted? Upon entry, exit or at both points in time? For tourism statistics, the requirements of immigration laws usually ensure the availability of count data from the time both of entry into and of exit from the country. How the question whether to measure at entry or at exit is answered will affect the kind of information that can be obtained (para. 3.24). It is essential to be aware of the methodological issue arising when assigning characteristics of trips or visitors to a certain period of reference (see sect. C.2.1).
3.30. The general practice is to count inbound travellers at entry, that is, when the trip begins, but some of their characteristics, mostly related to the stay itself and expenditure during the trip, can usually be collected accurately only when the stay ends, i.e., upon departure (see para. 3.65). These characteristics might also be collected, although this occurs less frequently, during the trip (see sect. C.2.2.2) or after the trip (see, for example, Box III.19). Countries should have a clear awareness of when this information, e.g., characteristics of a trip, is collected and to which population it refers, whether “arrivals” or “departures”. Reliable information on expenditure can be obtained only upon departure or after the trip. If, however, a country seeks to obtain expenditure information on arrival figures, it will need to work with certain assumptions to calculate these data.
3.31. Similarly, outbound travellers should be counted at departure, that is, when the trip begins, but some of their characteristics, mostly related to the stay itself, can usually be collected accurately only at, or after return, that is, upon arrival or later.
3.32. As part of their border control procedures, some countries measure both arrivals and departures of non‑nationals and reconcile the flows over a given period of time using matching procedures. The matched records are those used in tourism statistics to measure international tourism flows. They often consist of a form divided into, in two parts, with the second having to be turned in upon departure. Computerized systems with electronic passport reading upon both arrival and departure are also used. However, it should be recognized that these border‑control procedures are not designed primarily for tourism purposes. Rather, they are usually required for specific immigration controls‑related purposes as a means of identifying illegal immigrants, whose stay exceeds its authorized length. With a regular and efficient systematization process in place, entry/departure (E/D) cards used in such countries to identify the visitor’s length of stay, obtained by checking dates of entry and departure, should in principle provide the best possible estimate. The drawback is that the final corrected data are available only after all tourists authorized for a specific period have left.
3.33. Observation of the flows of international visitors and other international travellers inevitably requires the cooperation of various entities and institutions, usually including immigration authorities, entities responsible for road border traffic, state security forces and authorities responsible for the administration of ports, cruise terminals, airports, land terminals, and various other means of access by passenger traffic to the country. All of them should provide support in the design of the observation procedure and act as sources of information to serve as checks and controls for the final data.
3.34. The cooperation of the private sector may also be requested, in particular passenger transport companies (for land, air and water). The cooperation of different specialized organizations and bodies might also be required for particular subsets of visitors (for instance, those for whom the main purpose of a trip is education and training or health and medical care).
3.35. In the case of foreign students enrolled in national education programmes, or of residents enrolled in such programmes in foreign countries, it might be necessary to elicit cooperation from such institutions as the ministry of education, the ministry of foreign affairs or the boards of universities with established programmes for foreign students or exchange programmes with other academic centres abroad (the European Union Erasmus Programme, for instance).
3.36. The completeness and quality of arrival data are therefore closely linked to efficient inter‑agency cooperation (see chap. I, sect. D, for a discussion of various ways to achieve such cooperation).
C.1.1. Typology of the different modes of transport used to arrive in or depart from a country
3.37. Countries need to identify with precision the different modes of transport used to arrive in or depart from a country and their intensity of use, since each mode will require a different observation methodology for the measurement of inbound traveller and visitor flows (see IRTS 2008, paras. 3.32-3.34).
3.38. The following typology has been established:
Air transport may consist of the following modalities:
- Public transport operated on a regular basis or through charter flights operated by regular airlines or specific charter operators. These usually operate at a relatively small number of (international) airports, and their immigration procedures are usually well under control.
- Private transport might be provided by specific for‑hire operators, usually serving businesses. Businesses and some individuals might also own their own aircraft. In some countries, private and commercial aircraft operate out of the same airports; in others, all or some private planes operate from different airports, where immigration procedures might be more ad hoc.
In the case of air transport, there is usually an established control, perhaps with some exceptions, as in the case of movements within zones that have abolished their internal borders (as is the case for European countries belonging to the Schengen zone).
Arriving passengers might be in transit on an international journey, entailing either a stop on a continuing flight or a change of planes or even airports. IRTS 2008 (para. 2.61) defines and discusses the treatment of such passengers (transit passengers). Only passengers entering the legal and economic territory should be counted as visitors.
The modes of transport used to cross land borders include the following:
- Railways. Operators might be from one of the two countries involved or be managed as a multi‑territorial enterprise. Immigration controls are often conducted on board the trains.
- Other public transport by land. Public transport by land may take different forms: buses, taxis, mototaxis, etc. Bus transport is usually provided by well‑organized businesses, and authorized operators might be numerous. Taxis and mototaxis are often operated on a more independent or informal basis with little or no organization. Some operators engage in long‑haul transport, with formal organizations; others (sometimes the most important in terms of numbers of persons transported) operate locally, as within a zone of free movements, and are subject to few if any controls.
- Private transport by land. Land borders can also be crossed by private cars (owned or rented), by freight vehicles that carry passengers as well as crew and by other vehicle such as bicycles, motorcycles and animal‑drawn carts, which are often utilized by travellers within their usual environments who routinely cross the border and are mainly engaged in shuttle trade or other personal activities.
- Pedestrians. People living or working close to a border point might simply walk across it. Nomads might also fall in this category.
Flows at land borders are undoubtedly the most difficult ones to measure, and the degree to which they are controlled may vary from country to country and from post to post. Some countries might control all border crossers while, others might not control any (as is the case for Schengen‑type borders). There is a wide range of intermediate approaches.
C.1.1.3. Sea and river
Public and private modes of transport are also used for the crossing of borders by water:
- Public passenger transport, including ferries and cruises, is usually provided by established businesses, with boats arriving at organized moorings. Their control by immigration and port authorities is usually strictly organized, with some exceptions.
- Private passenger transport (including yachts, sailboats and canoes) is often provided more informally. Boats are usually required to call at authorized moorings only and to report to the local port authority (although they do not always do so).
- Freight ships, though with less and less frequency, might also carry passengers, who should usually be counted as visitors, as well as crew, who are not counted as visitors since they are deemed to be within their usual environment.
Landings can be controlled more easily if they occur at organized ports. Otherwise, the situation is very much similar to that associated with land border crossings.
C.1.2. Complexity in the measurement of flows
3.39. The following examples illustrate the need for compilers to develop appropriate observation methods:
- Countries belonging to a zone within which controls for all movements have been abolished
The typical case here is that of European countries belonging to the Schengen zone. Controls have been abolished for all travellers within the zone, remaining in place only at its borders with the rest of the world. Specific statistical procedures, automatic counts in most cases, need to be developed, coupled or not with survey procedures, since no administrative procedures are in place (see para. 3.92 and Box III.15).
- Countries with special land border zones but with border controls usually in place in other non‑land borders (air)
In many countries, special land border zones implicitly allow the free movement of persons, with little or no control or counting by immigration authorities for populations living on either side of the border. In some but not all cases, persons exempt from control need to hold a special permit. All such flows (whether of tourism or not) and their corresponding expenditures should theoretically be taken into consideration when estimating the “travel” item for balance‑of‑payments purposes. Most such persons are taking trips within their usual environment, and should not be included in tourism statistics. For analytical purposes, the measurement of travellers that are not visitors might also be of interest to National and Regional Tourism Administrations. On the other hand, if the movements are too frequent (i.e., the same persons move across a border several times a day), their numbers will tend to lack significance. Countries should lump such movements together so as to obtain a unique count of persons.
- Movements of persons by land between two non‑contiguous parts of the national territory
In certain cases, going from one part of a territory to another (a domestic trip) involves as it does for Oman, the Russian Federation (Kaliningrad), Malaysia and the United States of America Alaska transiting through the territory of another country.
Based on the established rules, as the origin and destination of the trip are both part of the same economic territory, such a trip undertaken by a resident is considered a domestic trip, a leg (fraction) of which is considered an outbound trip (for the purpose of transit). From the standpoint of the country of transit, on the other hand, the same leg of the trip should be considered an inbound trip (also for the purpose of transit). As discussed in Chapter IV, expenditure by a resident of country A during transit through country B falls under the rubric of outbound tourism consumption (from the perspective of country A) as long as it takes the form of a resident‑to‑non‑resident transaction.
Cruises constitute a particular case. Cruise ships, first of all, are usually enormous vessels that need to be received at special moorings, where passengers can disembark safely and the ship can, inter alia, be supplied with water, electric power and food and can have easy access to fuel and other bunker. This arrangement facilitates control of the movement of passengers. Additionally, cruise passengers will have specific plans: arriving passengers might stay on board or disembark to visit the country. They will disembark when ending their cruise or in order to board another cruise ship (under what is referred to as the hop‑on, hop‑off system). Passengers from the country visited, or from other countries, might be embarking at the stop in question either to initiate or to hop on their next cruise. For the purpose of tourism statistics, each of these situations should be identified and given specific treatment.
- The case of international cruises calling at different ports in the same country
The difficulty in this case lies in the fact that (a) if the ship stays in continental waters, it is deemed to be in the country’s economic territory, whereas (b) once it leaves continental waters, it is deemed to have left that territory. Under scenario (a), clearly, arrivals at different ports would be considered part of a single trip. However, in theory, under scenario (b), the ship (and passengers) will have left the country (territory) before coming back to it. However, for practical and common‑sense reasons, it is recommended that both cases be treated in tourism statistics as a single trip. (If data are based on interviewing passengers on shore, it is unlikely that they would know whether they had left continental waters or not. Also, from the point of view of measuring visitors, it is somewhat establishing whether or not the ship crossed the line between continental and international waters is an exercise that is somewhat academic). Such treatment is akin to treating a “normal” inbound visitor who visits a number of local destinations in the destination country as taking one trip to that country, but making a number of visits within it. However, in the case where a cruise ship calls at a port in country A, goes on to visit a port in country B and then travels back to another port in country A, that ship (and it’s passengers) should be treated as having made two separate trips to country A.
- Case of international cruises passing through waterways or by the coastline of a country
Passengers on cruise ships (a) that do not call in at a port or (b) that stay in waterways, e.g., the Suez Canal, shall not be considered visitors to the country of reference. Although they enter the legal and economic territory, there is no “stay” involved and hence no “tourism visit” (see IRTS 2008, para. 2.33).
C.1.3. Main sources for the measurement of global flows
3.40. The main sources available for the observation and measurement of international traveller flows are the following:
- Official administrative sources, including immigration records based, although not necessarily, on entry/departure (E/D) cards, manifests (sea and river) and advanced passenger information (API) (mainly air)
Data provided by such administrative sources are produced on a continuous basis and are usually aggregated monthly
- Complementary sources: Airlines, bus companies, ferries, etc., for information on passengers transported, embarking or disembarking, passengers arriving and departing at airports, manual or automatic counts at the borders (land borders), counts at toll booths in the vicinity of the border, etc.
Some of these sources are not publicly accessible. If information is published, frequency of publication may not always be monthly
- Specific sample survey sources. In countries where no such administrative sources exist, or where the data provided are too global to be of much use, sample survey sources are the only possible alternative for estimating the universe of arrivals by non‑resident travellers and departures by resident travellers. In some countries where administrative data are available, sample surveys are conducted (phase two) (see sect. C.2 below) to provide additional information on visitors and trips that is not available through the administrative source
Mirror data, mainly from neighbouring countries (inbound data derived from outbound data gathered by partner countries), may serve as a source, although the differences between the definitions of inbound and outbound trips should be kept in mind (IRTS 2008, para. 2.39). For example, in the case where an outbound visitor is visiting more than one country on his or her trip, possibly only the main destination country is recorded in the departure statistics
3.41. Before moving on to the description of official and complementary administrative sources, it is worthwhile examining Figure III.1 (a) and (b), which illustrates the complexity of the process required to create the universe of international traveller arrivals.
C.1.3.1. Official administrative sources
3.42. The three main official information sources used to measure inbound tourism flows – border controls, manifests collected by port authorities, and advanced passenger information – are reviewed below.18
3.43. As a result of the current activity of border control authorities, in most countries, reports are generated for all individuals crossing borders, whether nationals or non‑nationals, residents or non‑residents. The exceptions are, as noted earlier, countries belonging to regional groupings such as the European Union and those where frequent travellers (e.g., workers crossing a border on a daily basis) are not reported. However, where they exist, these individual reports can be extremely important as a basis for the measurement of inbound visitor and traveller flows.
3.44. Nevertheless, before using these reports as a basic source of information, for some or all cases of border crossings, it is recommended that there be an overall review which focuses on and clarifies important issues. Some of those issues are discussed directly below:
- While not all border crossings need to be observed using the same instrument, the geographical coverage of such operations does need to be established. Which border crossing points and what types of border crossing are covered by border control operations? Does this control apply to air passengers only, or to other types of arrival as well (e.g., by sea, land and river)? This issue is particularly relevant for countries with extensive land borders or borders delimited by rivers, where geography makes crossing the border an easy matter or where border controls are absent at some crossings perhaps owing to their isolated location. Border control authorities will usually have an estimate of what is beyond their current control procedure, but this estimation might need to be permanently monitored so that changes in behaviour over time can be detected.
- Which persons are covered? Are there specific conditions that exclude some persons from border controls? What is the status of, e.g., refugees and border workers? In many countries, nationals are often exempted from border controls or detailed reporting requirements (and from completing an E/D card where such cards exist). For the purposes of analysing tourism, however, nationals and non‑nationals who are also residents should be excluded from the flow of inbound visitors to be measured, while non‑resident nationals should be included. Frequent border crossers might have special permits, might be excluded from the controls altogether, or might be covered by a global figure. Such travellers would not be included as visitors because they are frequent border crossers. Finally, certain types of border crossing might be subject to less cumbersome procedures (for instance, at private airports, or at land borders for nationals of the border countries).
- Temporal coverage of the flows. Are the controls of uniform intensity irrespective of the day or hour of the day? If not, the scope of coverage of less stringent controls should be established and periodically updated.
- Actual content of the data collected. This pertains to the form of the database and access of tourism analysts to detailed microdata so as to permit, e.g., debugging and correction of invalid codes. In general, countries should not expect border control operations to provide all the information needed to measure traveller and visitor flows, and to observe all needed variables. This issue has to be further investigated, particularly in those countries that use this source only and do not determine whether other information that is needed to measure tourism (such as country of residence) exists elsewhere. In the best of cases, the data provided will be sufficient to frame a border survey (see sect. C.2.2.1 below) covering characteristics of interest. Additionally, neither all the controls in a given country, nor, the questions asked, will be the same at all border points (questions at land borders, for instance, might be kept to a basic minimum because of the time constraint).
- Quality of the data collected. There are various repeated inconsistencies in the information taken from administrative sources which stem from the latter’s specific functions. The main interest of border control authorities, for instance, is controlling the flows of non‑nationals; other data that are of less direct interest to them are not always adequately collected (e.g., on a national’s country of residence, origin or destination, and purpose of trip in detail); their concern is that the declared purpose be consistent with the type of visa, for instance, or with controls at the declared place of stay. Revisions, checks and controls are needed to make the information most usable for tourism analysis.
3.45. Some countries record only minimal data from travellers, which should be used as a first source for tourism purposes. Where countries also use E/D cards, the available data are usually considerably more extensive so as to include, e.g., demographics of travellers, length of stay and main destination in country, and this being the case, should be used to complement the basic administrative data (see Table III.1).
3.46. Data collected by migration authorities through direct capture (optical or machine reading of passports), through direct questions to passengers, or by means of E/D cards, might be more limited than what UNWTO is proposing, particularly with respect to issues not strictly related to border control per se. While it is acknowledged that the border control authorities of more and more countries’ are moving towards electronic data capture and ceasing to use E/D cards, there are still many countries that continue to capture data through this source. The important issue is what data are collected, and not the means of collection.
(a) Guidelines for the creation of a statistical universe for non‑resident visitors ial
(b) Guidelines for implementing a border survey
Common information items on entry/departure cards
Usefulness for tourism statistics
Essential for measuring number of visitors’ trips by month
Not useful for tourism statistics
Not useful for tourism statistics
Useful for designing and stratifying surveys; very useful for marketing purposes
Useful for designing and stratifying surveys; very useful for marketing purposes
Date of birth
Useful for designing and stratifying surveys; very useful for marketing purposes
Place of birth
Useful for tourism analysis, e.g., travel propensity due to links to another country
Useful for designing and stratifying surveys; useful in conjunction with “current country of residence”
Very useful for marketing purposes
Current country of residence
Essential for identifying person as a visitor, and data on country of origin; very useful for marketing purposes
Address in visited country
Useful for designing and stratifying surveys; useful for regional tourism statistics, along with “port of entry” in multiple destination countries
Not useful for tourism statistics
Place of issue
Not useful for tourism statistics
Date of issue
Not useful for tourism statistics
Type of passport
Not useful for tourism statistics
Type of visa
Useful for identifying certain categories of border crossers and determining which of them are non‑visitors
Port of entry
Very useful; useful for stratification of universe
Mode of transport
Very useful; useful for marketing purposes and for stratification of universe)
Flight number or name of ship
Essential for identifying residency of transportation company (particularly for international transport) for TSA, whether airfare is part of inbound, outbound or domestic tourism expenditurea
Essential for identifying residency of transportation company (particularly for international transport) for TSA; useful for cross‑checking with other sourcesa
Intended length of stay
Very useful as initial indication of actual length of stay; needs to be confirmed
Of some use as initial indication of actual accommodation; needs to be confirmed
Purpose of travel
Essential for identifying types of visitors’ trips
(a) Also useful for determining whether there is an economic effect in the country of reference, and whether airfare is part of inbound, outbound or domestic tourism expenditure, although code sharing and the possibility of using various airlines on the same journey reduce usefulness.
3.47. As mentioned earlier (see para. 2.9 above), travellers’ trips should be classified according to country of residence of the traveller, which is usually approximated using his or her current home address. If the traveller’s country of residence is not determined, the migration statistics will not provide a basic count of non‑residents entering the country of reference, or a sampling frame for further observation.
3.48. UNWTO proposes a set of questions that can be used on Entry/Departure cards. The questions seek to elicit both information useful for border control purposes and the minimum additional information for tourism statistics purposes.
3.49. The usefulness for tourism analysis of questions related to a visitor’s address in the country visited, intended length of stay, accommodation and purpose of visit is usually limited for two reasons:
- The answers might be biased by the fact that it is the migration authorities that are asking the questions. In some countries, migration authorities apply conditions for entry based, for instance, on the consistency of a traveller’s declaration of purpose of visit or place of stay with the type of visa presented. For an answer of “recreation” for purpose of visit, in the case of a tourist visa, for example, would be acceptable. On the other hand, the answer “attending a meeting”, might prompt an officer to request corroboration (e.g., an invitation). As regards place of stay, if it is known that the officer will not ask for a reservation voucher, a traveller may give the name of a well‑known hotel instead of responding that his or her place of accommodation has not yet been decided. Such answers might not necessarily reveal a visitor’s actual characteristics and to provide useful information, would need to be verified.
- The circumstances in which the information is collected, do not permit the respondent to enter into much detail so as to better delineate some important dimensions, such as the expected duration of a (student’s) course, the specific form of accommodation (e.g., whether fully owned or a timeshare) or the circumstances surrounding the trip’s purpose (according to the rules previously stated in chap. II, sect. C.1.1). Accordingly, information collected using this instrument exclusively might need to be adjusted for tourism purposes.
3.50. It would be especially helpful if an agreement could be reached with immigration authorities regarding access to their entry/departure databases so that, once individual records were duly rendered anonymous, their records could be reviewed for consistency, subjected to statistical debugging and used to estimate the average length of stay of various groups, broken down by country of residence. A far more precise estimate of this basic characteristic for use in the analysis of visitor behaviour can be obtained by using migration records (since they represent the actual population) than by carrying out surveys (except in countries that use sufficiently large samples for their border surveys and cover every month of the year). It may be preferable for such access to the immigration authority’s databases to be given to an NSO rather than to an NTA. Since the NSO knows how to deal with such sources, is usually subject to legal constraints on manipulating and producing data, and is generally perceived as being more professional in handling such issues as confidentiality of data than other agencies, (possibly) including the NTA. Also, the NSO may seek a wider range of data from the databases, such as on migration movements, than are required, by the NTA. And from the perspective of the NTA as well, it may be seen as preferable for the official professional statistical body to have the responsibility for producing the data required according to well‑established statistical procedures, so as to ensure good‑quality and timely information.
3.51. A manifest is an instrument used mainly in maritime transportation. In the case of cruise ships, ferries, yachts and all type of recreational vessels, the captain is usually requested to provide the port authority with a list of passengers and crew on board, indicating name, surname, nationality, passport number and any additional information authorities might decide to request, often for the purpose of paying a head tax. That is usually all the information provided.
3.52. Arriving passengers who stay on board are usually considered non‑resident visitors. In the case of passengers disembarking, however, additional information will need to be collected (e.g., through use of an E/D card or by way of a dockside survey), since some might be residents and others non‑residents of the country of reference (IRTS 2008, para. 2.63). Crews on public modes of transport should be excluded from the visitor category (IRTS 2008, para. 2.62).
Advanced Passenger Information (API)
3.53. Advanced Passenger Information (API) consists of a set of data generated on passengers arriving by air. It is constructed from the “digital prints” left by travellers when booking, paying for and engaging in travel. These databases are managed by private businesses which sell the product, in a variety of forms, to migration authorities, to provide them with clear, detailed and standardized records for each incoming passenger ahead of arrival. Anonymous information on incoming travellers from these databases can also be provided to tourism authorities.
3.54. The great benefit of this source is the coverage, reliability and timeliness of the data provided. Information on country of residence is not available from this source, but it can be approximated based on “country of origin” (for round as well as one‑way trips).
C.1.3.2. Complementary sources
3.55. Passengers are counted by various complementary sources, which can provide information on personal characteristics and/or characteristics associated with the trip. Such sources include airlines, international bus companies, railways and highway toll booths (for automatic vehicle counts at or near land borders).
3.56. Airlines and airports systematically generate data on air passenger flows, including the number of passengers on each flight, which are sometimes disaggregated by nationality, port of embarkation, etc.). This information can be used to check immigration data (available classified by airline, day and flight) or to substitute for counts where border controls are not in place (e.g., the Schengen zone). Data on passenger’s country of residence, however, are not usually gathered by these sources. Care should be taken, however, to exclude passengers in transit on international flights that do not necessarily enter the economic territory of the country of disembarkation.
3.57. Buses crossing international borders are also often required to present manifests in some form, providing information similar to ship manifests, including passenger identification, type of passport and passport number, so as to permit classification by nationality.
3.58. Sometimes, railway companies can also provide information on their passengers when they cross international land borders, but this is confined to flows between stations.
3.59. It should be noted, however, that while these transport businesses may be required to provide the data discussed above to the official authorities, some may be reluctant to make the data (or at least, some of it) available to other bodies, considering it to be “commercial‑in‑confidence”.
3.60. At busy land borders, authorities might set up an automatic count of vehicles, identifying the type of vehicle (light passenger car, bus, truck, etc.) and/or its licence plate (to determine the country of origin see Box III.9 for the case of the border between the United States of America). The number of non‑residents crossing the border can be estimated by assigning an average number of passengers per vehicle (often done manually) and assigning residence according to the licence plate of each vehicle.
3.61. As an alternative to automatic vehicle counting at the border, which requires the purchase and maintenance of special equipment, countries with toll booths located near border posts can refer to toll booth counts of vehicle flows. They might also collect additional information to distinguish residents from non‑residents, supplementing the vehicle counts, where necessary, with the observation of licence plates and number of persons per vehicle, so as to better approximate visitor flows.
3.62. Compilers should be aware, finally, that E/D cards, while usually including detailed questions, still do not provide sufficient information to identify certain classes of travellers precisely. For the purpose of tourism statistics, frequent border crossers, border workers and long‑term students and patients, for instance, should be considered “other travellers” rather than visitors (i.e., they should be regarded as being within their usual environment: the place of study or medical treatment).
 Since border‑control procedures apply also to nationals the same source can be used to obtain data on outbound tourism (see sect. E); hence, what follows applies by analogy to outbound tourism as well. However, the data should be based on country of residence rather than on country of nationality.
 It is to be noted that crew are not considered as visitors.
 World Tourism Organization (2005a), Tourism as an international traded service, (online), available at: http://statistics.unwto.org/sites/all/files/docpdf/border.pdf (30-05-2014).