Definitions, data sources and caveats

Tourism data are collected and analysed by organisations such as the World TourismOrganisation (WTO/OMT) and the World Travel and Tourism Council (WTTC), while crosssectoraldata on trade flows are collected by the United Nations Conference on Trade andDevelopment (UNCTAD) and Balance of Payments data are produced by the InternationalMonetary Fund (IMF). The most limited definition of tourist services is that of the World TradeOrganization (WTO), which includes only hotels and restaurants, travel agencies and guides.This definition, based on producing sectors, is designed to cover services provided internationallywhich may be regulated (and therefore which need to come under the General Agreement onTrade in Services - GATS) and which are not included in other categories. While these are amajor component of tourism, and for some purposes the question of what is regulatedinternationally will be important, the WTO definition clearly excludes a great many activitieswhich are commonly purchased or used by ‘tourists’ and includes some (restaurant meals or hotelconference facilities for non-visitors, for example) which are not. An alternative approach is tomeasure the activities of particular types of people. This is the basis of the WTO/OMT definition,of spending by people away from their homes. It corresponds to the normal balance of paymentsdefinition, of spending by tourists, and to what is measured by most national surveys of tourists.National sources give more detailed data according to their different specialisations. For somepurposes, other types of data (regional data, information about particular activities or data fromtypes of supplier) can supplement these. The international and macro-economic data allow us tocompare countries and to compare tourism to other exports or other sectors. The more specificdata help us to understand the details of its impact. The international tourism data are built upfrom information on movements of people and some country data on their spending. Internationalbalance of payments data, although some use the same surveys, are based on what is spent, notwho spends it. Since data on services tend to be less accurate than those on trade in goods (whichBox 1: The growing significance of tourism to developing countries• ‘Since the 1950s developing countries have received increasing numbers of international tourists,mainly from developed countries. International tourist arrivals have grown significantly faster indeveloping countries than they have in the EU or OECD countries. Developing countries had 292.6million international arrivals in 2000, an increase since 1990 of nearly 95%. The subgroup of LeastDeveloped Countries (LDCs) had 5.1 million international arrivals in 2000. They achieved anincrease of nearly 75% in the decade. This performance by developing countries compares veryfavourably with the growth of tourism to countries of the OECD and the EU, which achieved around40% growth.’ See Appendix, Table A1.• ‘Over the last ten years there has been a higher rate of growth in the absolute value of tourismexpenditure as recorded in the national accounts in developing countries than in developed countries.The absolute earnings of developing countries grew by 133% between 1990 and 2000 and in theLDCs by 154%, this compares with 64% for OECD countries and 49% for EU countries’. SeeAppendix, Table A2.• ‘The developing countries and particularly the LDCs secured a larger increase in the income perinternational arrival between 1990 and 2000 than did the OECD or the EU. The LDCs secured anincrease of 45% between 1990 and 2000 and the developing countries nearly 20%, this compareswith 18% for OECD countries and 7.8% for the EU.’ See Appendix, Table A3.• ‘In developing countries the export value of tourism grew by 154% [between 1990 and 2000] secondonly to the growth in the manufacturing sector.’ See Appendix, Table A4.Source: WTO/OMT 2002:26-29. See Annex 1 for statistical background information.8can be ‘seen’ crossing the border), tourism and other services data in balance of paymentsinformation are often less carefully estimated and checked than in those sources that are dedicatedto tourism. Balance of payments data also, of course, exclude data on domestic tourists (as doother sources of tourism data – see Box 2), an increasingly important part of the sector in manydeveloping countries.The World Tourism Organisation defines tourism as ‘the activities of persons travelling to andstaying in places outside their usual environment for not more than one consecutive year forleisure, business and other purposes’. This broad definition of tourism is then broken down intosix categories1 according to the purpose of the trip:1. Leisure, recreation and holidays2. Visiting friends and relatives3. Business and professional4. Health treatment5. Religion/pilgrimages6. OtherThose that engage in tourism – i.e. ‘tourists’ – can also be divided into international tourists anddomestic tourists and into overnight tourists and same-day visitors. Official statistics tend to focuson ‘international visitors’. Each classification may be useful for looking at the effects of tourismor at expectations for how it will grow. Business travel, for example, will respond to some factorswhich are different from those driving leisure travel. Within the conventional understanding oftourism, as leisure trips of days or weeks, there are a range of possible activities, with potentiallydifferent impacts on the country by creating different demands on it. Other distinctions possibleare between ‘high quality’ (in the sense of high spending tourists) and low or between ‘highimpact’ (in either the sense of integrated into the non-tourism activities of the country or in aspecifically poverty or environmental sense) and low. Much of the basic infrastructure of all these(accommodation, food, and travel) are the same and therefore a high proportion of the effects, areconstant, but not all.If we are studying ‘spending or activity by tourists’, not just ‘production in the hotel and othertourism sectors’, we need to know absolute numbers of tourists, not just estimates of spending,partly because for many supplier industries information on numbers will be the only firminformation on which to base estimates of tourism’s impact, but also because concern about theenvironmental and other non-economic impact is often based on the quantity of tourists, not thevalue of services. We also need to know from which countries they come, because surveys showthat different nationalities use different services with different impacts.The WTO/OMT has done a great deal of work on the refinement of the technical standards usedto collect and report the data. However, not all countries are members of the WTO/OMT and thereliability of the data varies from country to country. The most basic statistic is internationalarrivals. But some countries do not report international arrivals figures and may have no effectivemechanism in place for collecting them. Where the figures do exist, they have a number oflimitations in assessing the potential significance of tourism for pro-poor growth:• By definition these do not reflect the strength and significance of domestic tourism • National data have limited utility in assessing the importance of tourism in differentlocalities within a country. At the local level there is no standard mechanism available forthe collection of data that can inform decision making – hotel occupancy figures aresometimes used as a proxy indicator but are commercially sensitive, time-consuming tocollect and under reporting can have tax advantages for accommodation owners.• Visitor arrival figures focus on volume rather than value. It is the amount of time eachtourist spends in any location and the amount of money he or she spends onaccommodation, transport and other goods and services as well as the overall volume oftourists that determines the economic impact that will be felt. Even where figures onaverage daily spend and in-country tourism expenditure are calculated however, they donot generally provide any information about the potential for pro-poor tourism unless thisis specifically analysed (e.g. by focusing on expenditure on goods that are, or could be,produced locally).• Figures on visitor numbers often say little about the potential of the industry for pro-poordevelopment. In West Humla, Nepal, for example very small numbers of tourists have asignificant impact on local livelihoods.2• The form of tourism and the extent of access to tourists by local people as well as thevolume of tourists define opportunities for pro-poor tourism. If tourists arrive in adestination or at an attraction in coaches in the evening, drive straight to the hotel and eatdinner without venturing out, and in the morning pack and depart by coach to visit acultural monument before driving on to their next overnight stop, there are very fewopportunities for poor producers to sell to tourists – what ever volumes they arrive in.Beyond visitor arrival statistics, a Tourism Satellite Accounting (TSA)3 system has been developedby the World Tourism Organization and the World Travel and Tourism Council4 to estimate theeconomic value of tourism at the national level. Satellite accounts involve the ‘rearrangement ofexisting information found in the national accounts’ in order to estimate the significance ofsectors such as tourism that are not accounted for separately. Including the sectors from whichtourists or their suppliers buy inputs is logical in examining the nature and tourism. But it is notlegitimate then to use the data to compare the size of tourism to other sectors, because all sectorsbuy from others, and therefore aggregating sectoral accounts is double- (or multiple) counting.And there are limitations of this system in determining the potential of tourism for pro-poorgrowth. Aside from the fact that the development and reliability of satellite accounts is heavilyreliant upon the quality of the data available, as with visitor arrivals statistics the analysis isgenerally carried out at the national level and again is not disaggregated at the destination level.