Journal of the Statistical and Social Inquiry Society of Ireland, Vol. XXIV, Part IV, 1981/82, pp.175-216 THE FLOW OF FUNDS 1972-1977 Patrick Honohant with the assistance of John Dunne (Read before the Society } 22 April 1982) 1. INTRODUCTION Corresponding to the existence of financial assets there is the phenomenon of saving and dissaving, of financial surpluses and deficits. We may say that whenever an economic agent spends, over a certain period, more than his income on goods and services, he is dis- saving and is a net issuer of financial assets: he is accumulating financial liabilities. The questions addressed by a flow of funds analysis are: how much financial saving or dis- saving has been undertaken by various classes of agents; and which assets and liabilities have they accumulated over a given period? For Ireland, Brendan Dowling (1973/1974) provided an answer to the first question and also discussed the growth of aggregate holdings of certain financial assets and liabilities; but he did not attempt to break the holdings of financial assets down by class of agent. In the present paper we try to fill this gap. There are basically two ways of doing this. A direct but costly approach would involve gathering data which has not been recorded before. We have not collected any new data, although we do use some data which has not been published. In the main we have relied on existing and known data and have followed the second way of constructing the flow of funds table. This is to adopt a probabilistic framework and to make inferences about the unknown entries in the table from what data we have, together with the information that (a) the net accumulation of a single asset (liability) must sum to zero for the whole economy, and (b) the net accumulation by a class of agents of all assets/liabilities must equal the financial deficit of that class of agents. To some extent our work is in the nature of a methodological exercise — a prototype subject to modifications if it should be decided to update the calculations on a regular basis. In particular, it may be possible to refine the estimates using unpublished infor- mation which may be available as a result of data collection for the purpose of the National Accounts, for example. We decided, in line with the policy of not collecting new data, not to burden the Central Statistics Office with unstructured inquiries designed to elicit such information, but to see how far we could get using available data. We believe that the estimates that we have obtained are usable; they contain previously unknown information. Nevertheless, they do not have the same status as most series pub- lished in the Central Bank Bulletin so far as accuracy is concerned. This is partly because the assumptions which had to be used to fill the gaps are strong ones and partly because these assumptions were applied to data which is sometimes itself of indifferent quality, particularly the capital account of the Balance of Payments. At the same time we believe that we have obtained a logically consistent flow of funds table, at the cost of some definitional innovations which may startle those who are familiar with standard flow of funds methodology. f Patrick Honohan is a Senior Economist at the Central Bank of Ireland, where John Dunne worked as an undergraduate in the Summer of 1980. The views expressed in this paper are not necessarily those held by the Bank. We are indebted to Peter Bacon, Terry Donovan, John FitzGerald, Martin Kenneally, Pat McArdle, our colleagues in the Bank, and an anonymous referee for valuable comments on earlier drafts. The responsibility for remaining errors lies, however, with the authors. 175