ALBERT

Notes: Within the last two decades, national accountants wondered whether they were able to solve problems that in many respects were of the squaring the circle type: how could they embellish their representation of the economy without having their instrument grow heavier and less flexible? How could they elaborate synthetic frameworks in fields of collective concern that could both be specific and insure a dialogue with macroeconomics? How to keep consistent with the central framework and at the same time experiment with some freedom and flexibility. Satellite accounts are an answer to these questions. Much experience has been accumulated on satellite accounts in both national statistical offices and in international organisations. This paper provides a survey of this work. After an introduction (Part 1) devoted to stating the question, we try to propose a classification for the different types of satellite accounts (Part 2). Then we show their common characteristics (Part 3). However, this exercise would still remain abstract if we did not present samples of currently produced accounts. That is what we do in Part 4 with four examples. These examples have purposely been chosen in fields where many countries and international organizations have heavy investments, thus insuring the specific results will be obvious.

Notes: National accounts are a powerful means of coordinating different statistical systems. The better their classifications are adapted to the basic statistics or the information blocks one wishes to use, the better the national accounts play their part. This statement explains why, taking the opportunity of revising the whole system, French national accountants tried to improve the concordance between financial operation tables and monetary statistics. Other reasons leading to this attempt can be found in the dissatisfaction of users having to face different and inconsistent financial information such as the monetary statistics on one hand and the financial aggregates of the national accounts on the other; and even more reasons appear in the organizational field since those two statistical systems are issued by two neighbour services of the Banque de France, often depending on the same sources. Further, many propitious factors are converging at the same time: the French financial system is undergoing profound transformations originating as much in the behaviour of economic agents as in the law, and the statistical operations have to adapt to these changes. The national accounts will in the near future include balance sheets in which financial asset holdings are directly comparable to the money supply aggregates. In its first part our paper sets forth the detailed reasons for our attempts, the conditions in which it took place and the present results. We have reached a much better degree of consistency between the two systems, even if the final scheme has not yet been adopted in either the monetary field or in the field of national accounts. But an important question remains open about the durability of the harmonization: we think that it could be relatively uncertain because of the differences in the goals pursued by the two systems and the constraints which they face. That is why in the second part of the paper we tried to review the way such a pragmatic undertaking as ours could call into question the way in which financial operations are described in the system of national accounts. If one agrees with the present boundary between the real and the financial sphere, the articulation must remain somewhat elementary. But if one wants to revise the usual so-called dichotomy between financial and non financial phenomena, we think that a complete rebuilding of the conceptual framework of the accounts has to be done; this would necessitate a considerable amount of theoretical and practical work.