Alfred Franz, Vienna April, 2004
Ref.: 7th Int. Forum on Tourism Statistics
(Stockholm, 9-11 June, 2004):
Theme 1: Tourism Satellite Account
TSA Statistics – Dealing with an Onion ?
An “Onion”? Planting an onion in a bed and waiting for the shoots which might possibly arise after some time: this kind of “vegetal” comparison would certainly not apply to the situation of a newly established TSA. The meaning of this symboliccomparison is rather - “onion” as the ingredient of tasty cuisine, here used as a metaphor of the decisions that will possibly be taken there: the existence of a sequential order (instead of chaotic mixture); convergence towards a “core” (when progressing successively); unclear preference per se (with a view to various uses). Transposed into the context of statistics, one might specifyfurther: there is an overall scope and an internal structure – both represented in aggregation; and there are degrees of freedom – represented in a variation of choice. In terms of operational implementation further notions can be added: building blocks (in analogy to the “skins” of an onion); “top down” vs “ bottom up” procedures (instead of peeling the onion), equally proceeding from the outside to the inside and vice versa. Differentiation of importance (rather than uniformity), marginal vs central role, etc are categories of preference which can be applied there, too, but cannot be found in the “onion” itself, because the latter is a concept of formal understanding, therefore. - Young or fully grown, hot or sweet, black or white doesn’t change this basic pattern, but may be important for what to cook exactly in the kitchen of official statistics. If a concentration of problems happens to be found, or if there is an outright failure of methodology, this could still be dealt with in terms of “onion”, like removing a rotten or too hot a skin, or a young shoot. For the remaining rest, it may usually be more realistic to expect a range of overlapping, not immediately clear applications, like the possible use of foodstuff for a variety of cooking recipes: for goods that can be stored, a clear-cut choice of uses at the beginning is not sotypical. However, it is not necessary to overstrain this comparison here: it is used only to support understanding by some every day associations.
1INTroduction
While there is general agreement on the importance of having an “official” measure of the significance of tourism in the economy (Cooper-Wilson, 2002, p. 61; Fleetwood 2002, p. 131, just taken as examples)there is still much less agreement on how this is to be achieved in practical terms. Such differences are no surprise in view of a such a polymorphic, multifaceted body of information as is the case with a “Tourism Satellite Account” (TSA), by its very nature allowing for several approaches of measurement and accordingly a certain variation of the outcome (Franz, 2001, and Franz-Laimer, 2000). In addition to the methodological options of measurement there are usually limitations of practical compilation of a TSA, too, in their turn resulting into variation of how, and how far, the various TSA are actuallyimplemented . Decisions of this kind determine the actual shape of the TSA of a given country and,therefore,have to be taken whenever the introduction of a TSA is envisaged. Accordingly, there is good reasonfor such an investigation, having as its target the explorationof the range of the respective options in a more systematic way.
Against this background of variation in the reality, as indicated above, the present text basicallydeals with the TSA in the understanding of the Recommended Methodological Framework (WTO et al., 2001).However, while starting from the scope and the structures of TSA as put forward by the RMF, these premises are accordingly not taken as an automatically uniform, thoroughly equivalent canon, but reviewed in the perspective of implementation and use. Deepening or coarsening, emphasis or curtailment, exclusions and inclusions are considered as suggested by the needs and possibilities of practical compilation, whereas analysis as such is left out.
For the aim of this investigation a solid overview of the state of the art of the compilation of TSA would have been most useful, but has not been achieved so far.[1] Therefore, the arguments are mostly derived from current methodological literature,but only sporadically from actual country practice. Carefully considering the range of guiding concepts as well as of unavoidable determinants, and even general requirements of good practice in National Accounts (NA), a good deal of the options of implementation can still be explored, which might be worthwhile to take into account when aiming at a TSA.
Accordingly, this text continueswith a little recapitulation of more general “Preliminaries of Concept” (Chapter 2). Then the main features of the RMF are reviewed in some detail, as the central, unavoidable “Basic Reference” in this context, supported by a few diagrams for illustration (Chapter 3). Similarly the “Determinants” governing the specifications at issue are revealed and systematically considered (Chapter 4). This diagnosisis the basis fordescribingthe minimum content as well asa set ofmore or less concreteother criteria,andto point out the room of manoeuvre left for actual ad hoc decision (Chapter 5). Finally, a couple of “Conclusions” are summarized (Chapter 6).- Chapters 2 - 4 recall sine ira et studio what can be recognized from existing basic reference, thus exploring the room of possible decision, whereas Chapter 5 tries to draw practical conclusions.- It may be emphasised at the beginning that,within the given limits,onlyan idea of the principles rather than an exhaustive discussion can be attempted .[2]
2PRELIMINARIES OF CONCEPT
What are the indispensables of a “Satellite Account” (SA)? SNA tellsus a lot about SA, though it ismore moderate and flexible than particularly positive in respect of basic principles (UN et al., 1995, Chapter XXI). An SA is a “system” (i.e. with an inner “systematic” composition of its compartments); capable of being represented in “accounts” (combining incomings and outgoings in a balancing manner); resting on basic notions underlying the data assembled there (units of observation, classification and presentation); being in a way related to a “parent” system (like SNA). In a way, introducing additional breakdown SA are often “intensifications” rather than extensions of the SNA, at the same time remaining open for further elaboration, but always on tow of the central system.
Thus, an understanding of any SA as a convention of minimum standards may be more appropriate than the expectation of strict guidance being imposed.Even with these more limited demands,the usual requirements of sources remain as well as those upon techniques of compilation and levels of compatibility etc to be achieved. This on SA in general.- There are further self-evident preliminaries for a TSA proper, which are mainly represented by the concepts of “Tourism as the central notion of interest. At its fringes, however, even tourism is subject to variation and sophisticated refinement of definition. For the latter respect some further discussion can be found in Section 4.2 (“Core”).
3“RMF”: THE BASIC REFERENCE
The Recommended Methodological Framework (RMF, for short)is here briefly analysedas a framework,and with its basic structure as well - both together the central reference for the points at issue. (No decision is yet anticipated as to the actual contents.) (RMF 1.19) The shape and features circumscribed by this “body of information” is in utmost condensation exhibited in Fig.1.
3.1Scope
As to scope, this is the outer delimitation as demarcated by the content represented by the 10 Tables advanced by the RMF itself. (RMF 4.7) However, these are put into a more relative perspective by the respective authorities themselves, which point to a sort of “official“core of this system. As to its scope, this appears to be defined indirectly by reference to the sharein output accounted for by visitors’ consumption. That way the scope - the content of the “core”- reduces immediately to what is circumscribed in monetary terms as the substance of Tables 1 to 6.[3][4](WTO 2000, Add.1/3.16 in particular/)
3.2Structure
Apart from the structure determined by the Tables themselves a more intrinsic, elementary structure is found in terms of the various “building blocks “ (BB) of those Tables. They represent the horizontal structure of this body of information, while each of them is further broken down in terms of classification – the vertical dimension.[5] An inventory of these BB is a convenient means to find out the”netcontent”[6], and accordingly the data requirement, of the system: Fig. 2. As a matter of fact,with regard to the subject covered by the individual BB this is not always as homogeneous as explicitlysuggested by its heading – implicit content.
3.3Context
Beyond the above features for actual compilation, further requirements arise. First, a number of “extensions” are nearby, each in nuce provided already by the parent system:
- short term TSA
- regional TSA (and even supranational aggregations)
- indicators[7]
- strictly institutional perspective (in contrast to a strictly functional one)
- constant price versions.
In a way these are at least in part onlythe applications of additional breakdown – i.e. “intensifications” rather than“extensions” of scope, therefore, and thus applicable to the “core”, too.At least they can have the same “scope”.
This similarly applies to Worksheets prepared during more preliminary stages of TSA work, too, but there is also an important difference .[8]They must necessarily be elaborated for more complex topics like package tours or margins (if goods are alsoincluded), transactions in kind, and others, and insofar more closely related
to the “core”. This kind of evaluation - integrated but not directly appearing in the “core” - is its “background”.- ” The systematic establishment of links to the parent system (SNA, ESA..),[9]even to Labour Accounts (LA) & Social Accounting Matrices (SAM) may be recognized as unconditional “extensions.[10][11] Altogether additions of this kind on the level of preliminary basic work or on subsequent further elaboration may be subsumed under the term “ambiance”.
That way, a whole corona of additions can be summarized beyond the scope of the tabular presentations of RMF proper, but still wholly “TSA”. Envisaging, on the other hand, “net content”, and “very core” a major conclusion may be drawn at this stage, viz. that TSA is not 1:1 tantamount to the RMF’s Tabulations but may turn out smaller in content and at the same time greater as well:
TSA ≤ / ≥ RMFThis situation is illustrated in Fig. 3, which tentatively addresses the various reference points of deductions/omissions and elaborations/additions when taking a choice of a concrete TSA. For these decisions, further determinants are needed, which cannot be derived from descriptive criteria as the above ones alone.
4WHICH DETERMINANTS ?
4.1In General
A variety of determinants can be put forward usually guiding the design of exercises of this kind - the “skinning of the onion“. It would be nice to have an operational inventory of determinants, a readycheck list for application to the given circumstances and needs, so that ambiguities of design, choices between different options, etc would not arise at all. This is not the case, however: Even if quite clear basic determinants can easily be found: these are mainly represented by the given statistical equipment of a country, in terms of legal as well as actual instruments (surveys, registers, staff)on the one hand, and the existing methodological frameworks (RMF, SNA/ESA, related standard classifications) on the other hand. As a matter of fact, for the actual specification of a TSA both these references are by no means alone sufficient for decision. On the contrary, in combination they leave open a wide range of possible actual applications: the actual equipment does not normally allow 100 p.c. implementation of the RMF; nor is the latter a thoroughly cogent reference system; and even more complicating: there are options of change, additions, development in each country. With this finding we don’t mean to justify nihilism but to emphasize that further criteria must be taken into account, too, whether implicit or explicit ones, and not always immediately found in the mentioned reference systems.
Altogether the determinants/criteria divide into several categories (classes), according to their nature, degree of urgency, applicability, consequences: theoretical (ideal) vs practical (feasible); strict (cogent) vs soft (negotiable); principled vs economic; delege lata vs de lege ferenda; etc. These are examples of such distinctions but on a fairly general level. They may be useful to further understanding of what is going on, and which kind of decision is to be envisaged but otherwise they are either trivial, or giving fairly broad indication only, and are thus mostly not really operational at all. Even worse, on this general level the determinants themselvesmostly tend to merge into those categories of distinctions. What is needed is more concrete and positive guidance of what should actuallybe done . For this additional topical points of reference are needed.
4.2 “Core“ revisited
As already mentioned above, the idea of a “core“ of a TSA has godfathered its development from the beginning.(WTO, 2002, p.6).[12] Whether this is tantamount to ”minimum content“and how this would have to be defined otherwise: this is the point at issue here, with a view to a possibly most important determinant. The answer is found when considering which elements must not be left out, because they are still indispensable for calculating the “contribution to“, or the “share in“ the economy accounted for by tourism. Even in a narrow understanding, these indicators cannot be figured out in any other way than by connecting demand side data with supply side data, the statistical equivalent of economic circulation. Such connection in the first instance requires a symmetry of classification (the vertical dimension; see above, Fig. 2). The scope of what is covered by classification in terms of characteristicity largely follows from the standards of our understanding of “what is tourism“ but is unavoidably also conventional, at least to a certain degree. Next, such connection needs specification of the scope of demand. The latter point is not as easy as often anticipated by taking touristic consumption as an equivalent of touristic demand. In terms of „causal“ generation by tourism there may be important segments of final demand outside consumption of tourists – gross fixed capital formation of characteristic suppliers, and others (like government), too; and related current consumption expenditure of government. Even worse: touristic consumption in its turn systematically decomposes into further major “Building Blocks“ (BB, as pointed out in Fig. 2). In other words: some conventions are needed to specify demand 1 However, all in allit is clear that the notion of “core“ is can be used in a very narrow (if traditional) understanding, still without violating the idea of representing the essence of TSA:
Touristic Consumption Expenditure + Characteristically correspondent Supply == „Core“
Accordingly, such a “minimum“ can be positively described, but cannot as such be found, like in the search for a logical “minimum“ alone, but by convention only. In terms of the RMF, Tables No 1 – 6 comprise all elements needed, thus demarcating the scope of such more or less conventional “core“.
Within a certain range, at least, there cannotbe a “once for all” but only an ad hoc “minimum“, and there is no canonic definition of a “core“. Variation of this kind becomes even more true when the criterion of what is indispensably necessary is further applied, viz. to steps preliminarily needed when figuring out the data of the above described Demand – Use connection.The above mentioned so-called “Worksheets“(WS) are important in zhis context, as they are inevitably needed for the transformation of packaged tours, probably the most characteristic tourism product, but usually a quite complex piece of information; similar applies to WS on trade margins, provided that goods consumed by tourists are alsoincluded . And there may be other WS, too, depending on the preliminary decisions on the content of the “core“ in terms of Tables 1 – 6 (e.g. on transactions in kind). Such WS are not only needed for further elaborating the TSA but are analytically interesting in their own right, and they are useful to demonstrate the inner coherence of the data. In operational terms it, therefore, seems to be appropriate to include them in the notion of “core“, and to use this term in this more comprehensive meaning, alsocomprising its ”background“ .
That way the “core“ includes a set of Standard Tables (or their approximations), together tending to a (however defined) minimum, each suited for presentation and together for integrated analysis; and it also includes a set of typically needed preliminary evaluations, which are often quite interesting analytically. The former appears to be determined by convention (e.g. tradition), whereas the latter by subsequent necessity. Necessity is also the specification of the “core“ in time and space, and other features, like the solution of particular methodological problems involved here and there. In this sense, the term will be used in the rest of this text. Indeed, the user always presupposes some “core“ of the “onion“ but s/he would also benefit from some greater knowledge because onions do’nt not know, and cannot tell him, what to use, and what to throw away.
4.3 Other criteria
A first category is, of course, the urgency of need of the individual segments of the system (usefulness or importance are similar notions). Irrespective of their operationalization, it may still be useful to introduce further distinctions for the right signals: juridical; political; organisational; commercial; analytical. However, these signals come from outside and cannot be judged upon by a statistician, except if there is a need for some super-ordinate statistics of which the former is a part or pre-requisite.
A category more closely related to statistical work itself is feasibility. This is dependent on the strength of the statistical data basis, the ability of the staff, complicated circumstances which prevail in the tourism system itself (less pronounced regionality/seasonality, mixing up with non-touristic activities, etc), and the like.- Limitation of resources (cost) of a TSA is an obvious overall determinant, enforcing the formulation of clear preferences, and usually best found in cooperation with the statisticians.- Altogether, it is obvious that such criteria are highly interrelated if not interchangeable; therefore often of questionable operational use, and all the more far from straight application in terms of a ready “cooking recipe“. On the contrary, there will always be some interaction with “urgency“, too, and therefore large room for action/decision, which is found in a dialogue procedure, and in the end political (even if actually taken by statistician).[13] It is finally up to the statistician to defend the methodological necessities of the TSA, which are pervasive, rather than to engage for a certain shape of the system itself.