Economic statistics before the war were rarely collected for their own sake nor even primarily as a contribution to knowledge. They were assembled more usually by administrators who needed them in the discharge of their business or compiled as a by-product of some administrative function. This was true both of official statistics collected by governments, local and national, and of statistics pre-pared by and for the use of business executives, trade associations, and the boards of public undertakings.1 Kings, princes and govern-ments, whether popularly elected or not, all, from the eariiest times, have had need of money. Customs, levied on goods crossing national boundaries, have for centuries been a principal source of public revenue. Merchandise must be entered for duty at the frontier where the quantity, value and description can be duły noted by the revenue officers. From these returns are compiled statistics of the volume and value of exports and imports of merchandise. Drawback being allowed upon re-exports of imported merchandise, the accounts of customs revenue also yield statistics of imports retained for home consumption. Besides the trade in goods (the visible items), trans-actions in sendces of all kinds cross national boundaries. Nationals in one country borrow from or lend to nationals in another. These transactions are followed in due course by the receipt and payment of interest and sinking fund. During and after wars, government accounts are complicated by subsidies and tributes in all the forms of inter-governmental loans, lease-lend, mutual aid, reparations, indemnities, inter-allied war debts, sterling balances and direct aid. Statistics of exported and imported merchandise collected by H.M. Customs and Excise have generally been regarded as compre-hensive, accurate and reliable, at least sińce 1854 when the conven-tional prices at which goods had previously been valued were re-placed by current market prices. But neither the export and import of services nor international capital transactions have in the past been subject to tax or other impediment. At no time before the outbreak of the second world war in 1939 (except briefly towards the end of the first world war) has any administrator been reąuired, for reasons of fiscal or any other duty, to keep a tally of receipts and payments for services rendered to or by persons resident abroad; of borrowings and lendings nor of interest and sinking fund received or paid. The Board 1 Journal of Royal Statistical Society, "Memorandum on Official Statistics", 1943, pp. 145-66. of Trade before the war prepared estimates from time to time of the net receipts earned by British ships and the net income from overseas investment and other invisible sources. But these were no more than rough approximations of the finał balance of overseas income; there was no official attempt to ascertain separately each side of the account nor to estimate the value and naturę of capital transactions except as the net annual debit or credit on current account. The imposition of exchange control at the outbreak of war in 1939 established a body of officials whose business it was to regulate the purchase and sale of foreign exchange and to arrange for the pur­chase of foreign assets held by British nationals. The proper dis-charge of their duties demanded details of all transactions crossing British frontiers and the ascertainment of the value and disposition of all securities and other assets held abroad by residents in Great Britain. Exchange control and the registration of foreign securities held by owners resident in this country has been continued and is now made permanent by Act of Parliament.1 The varying degrees of exchange restriction soon forced a drasion of the world between the sterling area, the North American continent, O.E.E.C., non-participating countries in Europę, and so on. With each of these areas, a separate account is kept and the transactions can conse-ąuently be separately analysed. The statistics of the balance of payments are now as complete and accurate as the diligence of Treasury officials and Government statisticians can make them. But certainly before the first world war and even after, the expansion of the bare official returns of the volume and value of exports and imports of merchandise into an estimate of the balance of payments, was left to the work of devoted amateurs. Reliance on a very per­sonal enterprise continued even longer. During the second world war, as Mr. Harrod now tells us, Professor Sir Dennis Robertson "was assigned the task of keeping track of the British balance of pay­ments, and this he did without staff, in his own fair hand".2 Private citizens, in their capacity as tax-payers, have a deep and abiding interest in the raising and expenditure of public moneys. One of the principal tasks and privileges of elected assemblies in Western countries has been the scrutiny of the public accounts. Questions of supply in Britain have long been the monopoly of the Commons and the House have taken their responsibilities for the expenditure of public money very seriously indeed. The accounts are audited by officers of the Treasury and reviewed by Select Com- 1 The Exchange Control Act, 1947. 2 Harrod: Life of John Maynard Keynes, p. 530. mittees. Returns of public revenue and expenditure are exhaustive and as accurate as careful book-keeping can make them. The financial statements of national and local governments lack nothing in detail. The problem is one of accounting. In what form should the public accounts be thrown in order to disclose clearly and unambiguously the precise ąuantities which must be known? What in addition to the Exchequer accounts and the estimates of local authorities shall be included in order to take in the whole of the expenditure which, in sum, constitutes the outlay of public authori­ties as that ąuantity has been isolated in the analysis of national income and expenditure? The Exchequer accounts in their present form were designed to assist the oversight by Members of Parliament and others, of the revenue and expenditure of Central Government. All public revenue from tax and other sources was paid into the Exchequer. All pay­ments issued from the Exchequer. All unexpended balances were returned to the Exchequer at the end of the fiscal year and no public department carried unspent funds from one fiscal year to the next. The whole revenue and expenditure of Central Government during the year was thus included in the one account, of the receipts into and issues from the Exchequer. It was asked of public servants that public funds should not be diverted nor spent except for pur-poses expressly authorised by Parliament; and of Chancellors that revenue and expenditure should be exactly balanced with neither deficiency nor surplus except as might be provided by Parliament for the redemption of debt. Detailed audit and the grand tradition of personal integrity assured the first. The return of unexpended bal­ances into the Exchequer at the end of each fiscal year prevented public departments from building up a reserve of their own, free from the direct control of the Treasury. Public expenditure could thus be kept strictly within the bounds authorised by Parliament; and the principle of a precise balance secured that moneys raised in taxes should not exceed the minimum required for public purposes except by the sum ear-marked for the sinking fund. This simple and effective system was devised in the nineteenth century. The activities of Central Government have sińce greatly expanded and the public accounts have become immeasurably more complicated. Receipts and expenditure are authorised by many measures other than the Finance Act. Certain funds such as national insurance (and until 1936, the Road Fund) have an independent existence and there have (at times) been departments, such as the Ministry of Food and the Raw Materials Departments of the Board of Trade, which, being in fact trading undertakings, cannot be ex-pected to return unexpended balances to the Exchequer on March 31st of every year, sińce these balances include their working capital. More important still, the surplus or deficiency in the public accounts is now seized upon as one of the principal instruments for maintain-ing that high and stable level of employment which has emerged as one of the prime objects of contemporary economic policy. It is no longer sufficient that the fiscal budget should balance. It should, on the contrary, be deliberately unbalanced and held in surplus or deficiency as the needs of the times might dictate. There is every indication indeed that in high inflation and deep depression the sur­plus and deficiencies which would be reąuired to restore stability or encourage recovery would amount to a very considerable proportion of the entire public revenue. The Financial Statement laid annually before Parliament on the day the Chancellor opens his budget shows "above the line" ordinary revenue and expenditure raised by and issued from the Exchequer under authority of the (annual) Finance Act. "Below the line" are subjoined receipts and expenditure authorised by other Acts of Parliament. Payments below the line include such items as claims for war damage, loans to local authorities, housing subsidies and refunds of excess profits tax. The sum of these payments now ex-ceeds, more or less considerably, the relatively smali receipts "below the line"—repayments by local authorities, receipts from rents under the Housing Acts and such extraordinary revenues as the gifts from commonwealth countries. The difference emerges as a deficiency below the line, to be deducted from, or added to, the surplus or deficit above the line of ordinary revenue and ordinary expenditure, in order to arrive at the finał balance of the budget overall. Ordinary expenditure, in the first years after the war, exceeded ordinary rev-enue and the budget was in deficit above the line as well as below; but Chancellors sińce 1947 have been aiming at an excess of revenue "above the line", to cover the deficiency below, and their budgets have generally been in surplus. In the "alternative classification" included in the Financial State­ment sińce 1948, the fiscal accounts have been rearranged. The several heads of revenue and expenditure are now divided, not only by the administrative principle of the statutory authority under which the revenue is collected and the expenditure issued, but to correspond as far as possible with the distinction between current and non-current receipts and expenditure. The account in this form now shows "above the line" the surplus on current account accruing to Central Government of revenue over expenditure; and "below the line" that excess of non-revenue payments over non-revenue receipts which is to be financed by the (current) surplus "above the line". But the finał result, the budget surplus or deficit overall above and below the line remains the same in the two classifications. It is only the line which is drawn in a different place.1 The significance of savings and investment for the stability of income and employment was not widely recognised until the discussions following the publication of Lord Keynes' General Theory in 1936. No officials have, nor ever have had, occasion to assemble statistics of personal and business savings for any administrative purpose. Direct tax is levied upon income and not upon expenditure, upon company earnings and sińce 1947, on dividends as such. A concession is allowed upon earned income; but no distinction is drawn, for purposes of liability to tax, between income which is spent and income which is saved. Neither the sums set aside out of income during the year nor withdrawals from previously accumulated savings, can be directly determined. Net saving alone can be measured and that only as the difference between income and the sum of expenditure.2 A ąuantity arrived at by difference automatically accumulates all the errors in the esti-mation of the totals. It is perhaps fortunate for the statisticians that the surpluses of public enterprises, the depreciation allowances and the undistributed profits of companies, all of which can be got from the tax returns and the accounts of public authorities and com­panies, should constitute so large a part of the annual fund of savings. Statisticians who before the war, though unassisted and little en-couraged, were led to attempt their own estimate of that important ąuantity, the sum of the year's capital formation, gross or net, could rely, as for example Mr. Colin Clark did, upon the "fortunate" (sic) arrangement "that the output of finished capital goods for new investment and for meeting depreciation and maintenance consisted of the output of five Census industries".3 Expenditures on capital fixed in buildings and eąuipment were prohibited during the war and for five years and more afterwards by licence and rationed by allocation of raw materials, particularly of steel and timber. The 1 Sir Stafford Cripps: 449 H.C. 56-60, 463 H.C. 2072-^t. Hicks: Problem of Budgetary Reforms; Norris: Accounting Research, Jan. 1950; and Garrett and Reddaway: Accounting Research, Jan. 1951. 2 Kaldor: An Expenditure Tax. 3 Clark: National Income and Outlay, p. 169. The Census of Production was taken only at five-yearly (or even less frequent) intervals and the last one had been in 1935. closer the control, the greater the detail in which applications for licences have to be expressed and the more likely, therefore, that statisticians, called upon to ascertain the investment realised on these programmes, will find the data which they need. But a part of (fixed) capital formation always escapes record in this way, because it reąuires no new buildings nor extensive alterations to existing structures—and there are still variations in stocks to be accounted for and in the value of work in progress. Increase or reduction of stock and work in progress is one of the most variable of factors and the total value held at the end of any one year may outweigh the addition to fixed capital made during the year. Records of capital expenditures have been kept in increasing detail as a result of the annual reviews of investment programmes begun in 1947. The "few but vital" figures of fixed investment and stocks in the distributive and service trades could be supplied, so the (Conservative) Chancellor, Mr. Harold Macmillan, explained in 1956, "without undue trouble" by a sample of firms; but he pointed out that the enąuiry, if it were to yield reliable information, would have to be compulsory "so that the value of the figures willingly supplied by the majority of traders will not be nullified by the inaction of the minority". Estimates of savings have been regularly prepared sińce war-time Chancellors began to rely upon income voluntarily withheld from consumption as one element in their fiscal policies;1 but the figures for private savings by persons and businesses and for alterations in the value of stocks and work in progress are still, as their authors have so modestly remarked, "precariously based and subject to extensive revision".