A Class-Society of National Accounts Compilers ? Considerations on Certain Requirements and Some Consequences of the Transition to the New Systems

Recent changes in the “regime“ of international National Accounting (NA) Standards entail clear changes to traditional practice, self understanding and societal positioning of this kind of statistical work. Since the ability to follow and/or the legal impact of the new system is not evenly distributed over countries there is danger of segregation of the former NA family into a sort of “class society“. Surprisingly, neither on the side of government nor of scientific circles this problem has found sufficient attention so far. In the present text the implications and some consequential necessities are considered from different points of view, viz. the suitability of countries, the legalistic requirements and the operational use of the new instruments. Zusammenfassung: Die jüngsten Neuerungen im Regime der internationalen Systeme der VGR bedeuten markanten Wechsel, was die bisherigen Praktiken, das Selbstverständnis und die gesellschaftliche Bedeutung dieser Art von Statistik anbelangt. Die Konsequenzen hinsichtlich der Voraussetzungen wie die gesetzlichen Verbindlichkeit zur Durchführung dieser Neuerungen sind allerdings keineswegs gleich-verteilt, sodaß die Gefahr einer Auseinanderentwicklung der früheren Gesamtrechnungsfamilie in eine Art Klassengesellschaft nicht von der Hand zu weisen ist. Erstaunlicherweise haben diese Fragen bis jetzt weder auf politischer noch auf akademischer Seite größere Beachtung gefunden. Im gegenwärtigen Text werden diese Fragen vom Standpunkt der prinzipiellen Befähigung zur Umsetzung der neuen Systeme, ihrer gesetzlichen Verbindlichkeit wie ihrer operativen Bedeutung untersucht.


Introduction
In the context of the National Accounts (NA) the term "transition" will assume a much broader meaning than it used to have: many more countries than the commonly socalled "transition countries" see themselves in severe difficulties in adapting to the requirements of the new '93 SNA or the related European System of Accounts ('95 ESA).With regard to the latter, additional elements are effective because an unprecedented legally binding character is involved on the one hand; and on the other hand NA results are increasingly used as parameters of political decision and even more, of automatic consequences.While the increasing requirements of the new system seem to apply more or less to the whole UN membership, this is not the case with the latter circumstances, because these obviously apply to the EC membership only.In other words, there are countries that produce accounts with increasing demand on quality and detail, but otherwise in the same conditions of responsibility as before; and other countries that are confronted with additional needs and consequences of their compilation, which is essentially execution of law, subject to audit and burdened by farreaching economic and financial consequences.These basic divergencies may easily become the beginning of a tendency of segregation, so to say resulting into a "two-classsociety" of NA producers with an increasingly diminishing common denominator.
However, the issue addressed first (namely the considerable methodological requirements of the new system and the varied possibilities of countries to come to grips with these requirements) may also end in a sort of "two-class-society", namely the countries which are well suited to introduce the new instruments quickly; and the countries that are not, and that will require longer periods of transition.The latter are by no means identical with the usual transition countries: another, more general situation which appears to result in a "two-class-society"!So far, little attention has been paid to these imminent developments, although any efforts would be justified to avoid such undue segregation; and to maintain coherence within the whole NA family.So far, unfortunately, remedy is not in sight for the various aspects of the problems addressed.However, sometimes diagnosis and recognition of a problem is already a first step towards resolution!Subsequently, the ability of countries to introduce the new system; the differentiated status of the standards as regards their legal character; and the implications of the use of NA for operational purposes will be addressed in turn.The first point may be called the suitability issue; the second point the legalisation issue; and the third point the operational issue.

The Suitability Issue
It is one of the explicit aims of the new systems (SNA'93;ESA'95) to result into significantly extended, deepened and intensified reporting.Of course, in the first instance it is primarily a matter of resources to meet these requirements in time and in sufficient quality.However, countries also have strong traditions as regards their "NA culture", their expertise, their internal equipment of software and labour etc., and last but not least, their data basis.In the latter respect it is but surprising that the systems, at least the SNA, aim at a largely comparable and interchangeable superstructure but almost completely ignore the often fundamental differences in the substructure of the statistical data basis.While this has been fully recognised in a region like the EU or the EEA, even in these cases the actual completion of such intended new foundation my take years if not decades, so that the difficulty of the topical situation is not so different throughout, and will still stay so for some time.A couple of new, in some cases quite heavy instruments like new classifications, registers, systems of censuses and surveys will take time to be completed, some more time to be fully evaluated and another couple of years, to be "digested" in the NA.This is all the more problematic where traditional accounts have so far not covered the whole scope, so that appropriate new tools have to be developed first.Some such tools could be devised in a way which may not only be useful for a certain country but could be easily transferred to others, so that precious time could be saved which would be badly needed for conceptual development and experimentation.Such tools might, e.g., apply to the classifications part of the whole system, providing integrated evaluation of basic data, basically structured according to instruments like ISIC, or HS.This, however, is only one such candidate field.Others may be found, e.g., commodity flow type evaluations; programs for reconciliation of accounts, etc. Isn't it much more due to tradition rather than actual "political need" that our basic surveys are still not less different than our evaluation tools used for NA and other purposes?Couldn't significant inputs be saved for more meaningful purposes if joined together with what countries are now going to do in this field?A situation of duplication, separation and unnecessary differentiation is the reality and perhaps more so in the more developed than in the transition countries!There is no ready answer to the question how the various parallel initiatives could be bundled together, possibly by means of concentration of such work in those places wherever the best ability exists.There is also no ready answer to the problem which role international organisations should play.But it is obvious anyhow that so far awareness of the problem, and therefore assistance, training facilities etc. were not sufficient, often too theoretical and basic rather than practical and application-oriented.
However, some first steps could easily be suggested: e.g. that this concern is put on the permanent agenda of the competent international fora, like the UN Statistical Commission and/or the Conference of European Statisticians (CES) for the next years to come; that a special task-force is built to watch and systematise what is going on in this respect in the various countries; and that a sort of clearing house would be established of information on the availability of such tools whether presently or in the near future.In this endeavour the countries may tie in on various levels, either as contributors, or possible users, or with consulting expertise, etc.In the end, a strong element of convergence, economisation and speeding up of the whole transition procedure could be expected.

The Legalisation Issue
While the status of the SNA'93 is still "recommendation" there is at least one region in the world where this is no longer the case i.e. the EU.There the standards are assuming a legally binding character, at least as regards definitions, classifications and even the concepts of compilation.While this may at the beginning not mean too much with regard to differentiation, in the long run a couple of consequential tendencies are more or less likely: -organic development of integration and change may be hampered, for procedural reasons; -the implementation of legal norms may become a target predominating over scientific interpretation; -formal ("cannonistic", "dogmatic", "orthodox") understanding is promoted; thinking in alternatives is discouraged; -casuistic tendencies may creep in as to solve borderline cases and ambiguities; -flexibility may get lost in cases of newly emerging phenomena.
Admittedly, for specific reasons this situation may still seem a desirable or necessary alternative or at least not to be too embarrassing for a certain region, like the EU.But for the sake of utmost comparability this situation can easily become another source of separation and segregation in that other countries are not willing to follow the rules and standards of such "legalisation core".It may, however, also mean that, for the sake of convenience, other countries will follow.Neither development would seem a priori desirable, however!In the first case the consequence would be segregation, which cannot be an attractive target, because of its inherent loss of comparability.In the second case, the experimenting ("laboratory") character of NA work would get lost even more.
Therefore again a sort of remedy is needed by which both disadvantages might be avoided.Again without attempting even to think of a ready answer, first of all the attention of the international fora should be drawn to this particular issue, by putting it on the agenda for repeated if not permanent discussion.Experience in this respect might be regularly collected of countries inside as well as countries outside the legalisation core.A regular link to the scientific (academic) level might be established to get feed back on the suitability of the various concepts with a view to theoretical needs.The extension of the common denominator by voluntary (non-obligatory) adoption of the legalistic standards may be promoted, but not without trying to identify and maintain elements of flexibility felt necessary in the wider context.In particular, the anticipated effects of ad hoc solutions by means of legalistic measures, and the consequences of alternative solutions may be kept under study (possibly to be done by volunteering countries rather than the Secretariats).A well-known example of hte practical importance of such legally determined rules is the delineation of the public sector.

The Operationality Issue
This point is closely related to the previous one, because normally the reason of legal determination of NA standards is mainly the importance of the figures for operational uses, like payments (contributions) to a central or supranational level; or -the other way round -of payments to the individual member countries to meet special financing needs, etc.Although not completely new, this increasing and often extraordinary importance of NA figures puts the compilers into a possibly difficult situation, in that --they may face pressure from the national side not to produce figures "unduly" burdening the countries by financial consequences (excessive payments; too little receipts...); -they may face opposite pressure from the international (supranational) level (or even from other countries).
In such situations there is a danger that --the scientific argument is not recognised by those interested in but staying outside compilation; -the behaviour of the compilers (if unintentionally) my tend to "discount" the expected problems otherwise encountered; -"failure" of the NA figures (which cannot always be avoided; cf. the revision problem) may attract undue attention of the public, that way again possibly affecting the behaviour of the accountants.
Possibly, further aspects would also have to be taken into account, like (in extreme cases) due to the immoral compliance of accountants, etc.Also possible consequences on the behaviour would have to be considered due to the rigorous monitoring regime, with on the spot audits etc., i.e. arrangements which are altogether able to change the usual NA working environment.
To ask such questions should not be confused with blunt opposition: It by no means implies being against the new obligations at all: it is just a matter of being highly concerned by a fundamentally new feature of NA uses and related NA work.Since these new features do not equally apply to all countries, and may even not affect all applicable countries similarly, there is an obvious interest to keep this new situation in mind and to draw the right conclusions from experience made.And these tendencies are even going to increase.Therefore, it would be all the more necessary to well consider: -how to find right procedures and international arrangements under such circumstances; -how to protect the compilers against undue interventions; -how to provide a suitable theoretical framework for this delicate situation.
At the same time, there are still problems of the possible degree of accuracy and completeness that should be achieved eventually: "Nobody really knows true GDP, nobody has ever seen it!"In order to achieve a generally satisfying procedure much more attention would be needed also from the part of science, whether theoretical economists, or analysts, auditing and financial experts, or maybe even lawyers and psychologists.There is also the obvious danger that general attention will more and more concentrate on the NA outcome of the operational ("parametric") kind whereas its other role may be under-valued, which consists in providing a comprehensive fund of coherent data appropriate for the needs of manifold analyses ("multipurpose").
In any case, it would not seem right to disregard or ignore this situation, if indeed there are different groups of countries working under significantly differing circumstances.
Similar to the previous proposals, the very first step in the right direction would be recognition of the problem.Again, the first steps would obviously be candid information from those which are immediately involved, whether on the national or the supranational etc. level.Widespread discussion and coordination of diagnosis and possible further steps would already represent important progress in this particular field, which is probably the most complicated one.