A FULL explanation of the data and methods used A1 in constructing our index of general business conditions was given in THE REVIEW OF ECONOMIC STATISTICS for January and April i919. That explanation should be consulted by readers of the REViEw who want information concerning the technique or details of the construction of the index. The majority of readers, however, are interested in results rather than in methods or technical details. The explanation which follows is designed to meet the needs of such readers. Although the explanation is non-technical, it is nevertheless complete and it is intended to give a statement sufficient for an understanding of our index. Our index of general business conditions is derived from those series of monthly items of the industrial, commercial, and financial statistics which ordinarily serve as the basis for judgments concerning the fundamental speculative, business, and banking situation. Measurement of business activity, however, is relative to some standard. Isolated items of statistical series, upon which such a measure must be based, can have no significance by themselves. Only by a comparison of items over a period of time can we ascertain their meaning. Items pertaining to widely separated periods cannot, however, be used in their crude form. Each monthly item of bank clearings, pig-iron production, interest rates, and the like, is a composite, the make-up of which depends on the year and season. That is, various elements contribute to make bank clearings for January I920, for instance, the precise total reported. In general the actual items result from the combination of four elements -secular trend, seasonal variation, cyclical fluctuation, and a residual factor. The secular trend is the regular increase or decrease, according to some principle, over the whole period under consideration. For most series it is a growth element, dependent upon population and the development of industry. There is a normal change year after year in a developing or altering industrial society, just as there is a normal change in the physical or mental status of a growing child. As used in this exposition, the expression secular trend may refer either to a statistical series or to one item of a series. When it refers to a series it means the straight line fitted to the data; the measure of growth is given by the slope of the straight line. When the expression refers to one item of a series it designates the vertical distance (or its numerical value) from the curve or line of trend to the zero base line. The seasonal variation is the movement of the items within the year, which we attribute to the round of the seasons. There is a seasonal change in various lines of business activity just as there is a seasonal change in temperature or rainfall. Although an iron-clad system is not to be expected, the movement of the items, to be seasonal, must be systematic year after year. The seasonal variation of an item for any month is given by an index which expresses the normal for that month as a percentage of the monthly average for the year. The cycles are the undulating curves, or numerical values, secured by removing from the actual items the secular trend and the seasonal variation, and expressing the results in terms of comparable units.' The actual figures thus corrected and expressed measure the rhythmic movement of business, the ebb and flow corresponding to depression and prosperity. The residual element includes all sporadic developments which affect individual series, or widespread changes due to momentous occurrences, such as wars or national catastrophes, which affect a number of series simultaneously. Thus pig-iron production may take a sudden slump if a strike occurs, railway gross earnings may drop because of unusual storms or floods; trading on the stock exchange may be greatly affected by a court decision. The continuity of nearly all series was interrupted by the outbreak of the Great War. This irregular or residual element has not been eliminated by our process of correcting the data and hence it is present, along with the cyclical fluctuation, in the statistics used in constructing our Index. In order to adapt the statistical data to our purpose, the securing of a numerical measure of the ebb and flow of business, two problems were solved; first, we devised a method of correcting the statistics for secular trend and seasonal fluctuations; that is, we unraveled the tangle of elements which constitute the fluctuations in fundamental series. In other words, we devised a method whereby the actual series of business statistics now published in our trade and financial journals were set forth in such a form as to be significant and reliable indices of business conditions. Second, having corrected the various series of fundamental statistics for secular trend and seasonal variation, it was necessary to solve another statistical problem. The corrected series have wave movements; for all corrected series the elapsed periods from crest to crest or trough to trough are approximately the same. The times, however, at which crests or troughs of the waves of the several series occur are frequently not the same. Never-

T HE chief defect of the old national banking system was its decentralization resulting in scattered bank reserves. Even though the national banks of the United States possessed enormous aggregate cash reserves yet, before the establishment of the federal reserve system, those reserves were ineffective. There was no method of mobilizing cash for use where and when it was needed. No responsible body was empowered to adopt or carry out discount or other policies to avoid a financial panic or to handle a panic when it occurred. Other countries had economic crises; the United States not only had crises but financial panics as well. This is the reason that under the national banking system the United States was called an international financial nuisance. Previous to the establishment of the federal reserve system country banks were allowed to keep three-fifths of their required reserve of I 5 per cent in reserve city or central reserve city banks. Reserve city banks were allowed to keep half of their required reserve of 25 per cent in banks of central reserve cities. National banks in the central reserve cities, New York, Chicago, and St. Louis, were required to maintain a cash reserve of 25 per cent. Under the reserve system just described the ratio of reserve to net deposits, during the period I909-I4, for all national banks averaged between I9 and 22 per cent. The aggregate lawful money held was between 852 and 904 million dollars and aggregate net deposits ranged between 6,164 and 7,293 million dollars and the ratio of cash to net deposits for all national banks between I2 and I4 per cent (Table I). The inter-bank reserve relations resulted, therefore, in a cash reserve of about I2 per cent appearing as a legal reserve of about 20 per cent. Under the old national banking system, an increase of one dollar in lawful money would be the basis for an average increase of about eight dollars in loans and deposits. During the period I909-I3 the national banks of New York City held upwards of one-third of the specie in the vaults of all national banks of the United States.' In spite of a system which compelled the scattering of reserves and of responsibility for financial policies, New York City was the outstanding and even dominating capital of the national banking system. Because of the fact that New York City was, by force of circumstances, the financial capital and main reserve center of the United States, the weekly statements of the New York Clearing House Banks were rightly considered the most important data available upon which to base a judgment of current credit conditions. Of the items published in the weekly statements of the New York banks the excess reserve was generally considered the most significant single item. An excess reserve in New York meant that banks there could increase their loans and deposits to four times the amount of the excess, or that they could respond to the demands of country banks for cash and still maintain the legal ratio of 25 per cent. A deficiency in reserves was often the signal for a tight money market in New York, erratic fluctuations in call rates, apprehension on the part of country bankers, and a drain of cash to the interior. Under the federal reserve system, however, the figures for reserves and excess reserves of New York banks (or any other group of member banks, for that matter) have lost their former significance. Since June 2I, I9I7 every member bank has been required to maintain reserve balances in the federal reserve bank of its district. Cash in the vaults of member banks does not now count as reserve. In other words, legal reserves of member banks now consist only of deposits with the federal reserve banks. These reserve deposits are required to be not TABLE i. SELECTED ITEMS FROM THE REPORTS OF ALL NATIONAL BANKS IN THE UNITED STATES EACH YEAR FROM I898 TO I9I4 *

Warren M. Persons, Pierson M. Tuttle, Edwin Frickey, Business and Financial Conditions Following the Civil War in the United States, The Review of Economics and Statistics, Vol. 2, Supplement 2 (Jul., 1920), pp. 5-21