HC .3 i SCHOOL OF BUSIilESS and i PUBLIC ilDMIIIISTRlTIOK n ^ CORIJELL CJIIVBIISITY t > r- COLLECTION PRESENTED BY EDMUND EZRA DAY Cornell University Library HC 106.3.H33 Special and advance letters; September 23 3 1924 019 257 553 1^ 'm XI Cornell University Library The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924019257553 SPECIAL AND ADVANCE LETTERS September 2^ — 'December 24.^ Ip2l HARVARD UNIVERSITY ■s- COMMITTEE ON ECONOMIC RESEARCH CAMBRIDGE, MASSACHUSETTS, U. S. A. I92I COPYRIGHT, 1 92 1 BY HARVARD UNIVERSITY /^7f ^ job PRINTED AT THE HARVARD UNIVERSITY PRESS CAMBRIDGE, MASS., U. S. A. TABLE OF CONTENTS ADVANCE LETTERS General Business Conditions Oct. i, Oct. 15, Nov. i, Nov. 15, Dec. i, Dec. 15 SPECIAL LETTERS The Volume of Manufacture Sept. 23, Oct. 22, Nov. 29, Dec. 24 Industrial Employment Oct. 8 The Commodity Price Situation Nov. 8 The Volume of Foreign Trade Dec. 10 ERRORS October 8, 192 1 Page 5, Chart 3. The caption for the drawing in the upper right-hand corner should read Chemicals instead of Chemicals, Oils, Paints, etc. October 22, 1921 Page 3, Chart 2 and page 8, Table B. The adjusted indices here shown for the textile group and for its two components, the cotton mdustry and the woolen and worsted industry, were replaced, in the Special Letter of November 26, 1921, pp. 4-5, by revised figures, as explained on page 4 of that Letter. Page 4, Chart 3. The annual averages, 1905-14, apply to twelve-month periods ending August 31, not to calendar years as plotted. Page 5, Chart 4. The annual averages, 1904-17, apply to twelve-month periods ending June 30, not to calendar years as plotted. Page y. Chart 5 and page 8, Table B. Note the following corrections: Woolen ahd Worsted Goods Carpets and Rros Per cent of normal activity Per cent oj carpet of spindles and looms looms active ig2i January 5p.i 54 February 68.8 50 March 82.2 40 April 94.7 44 May .... . 103.6 46 June 107.I 42 July . . 106.4 49 November i, 1921 Page I, 2d line. Read speculation and banking instead of speculation and business. November 8, 192 1 Page 5, footnote i. This footnote should read as follows: * Expressed as percentages of the average price of the respective commodities for 1898-1914. Similar figures for all the groups of the Bureau of Labor Statistics are: farm products, 3.12; house-furnishing goods, 2.6y, lumber and building materials, 2. 60; food, etc., 2.1 y, all commodities, 2.06; miscellaneous, 1.45; cloths and clothing, 1.39; fuel and lighting, 1.35; chemicals and drugs, 0.45; metals and metal products, —0.02. Similar figures for 27 agricultural and mineral raw products, 71 articles manufactured from them, and the 98 commodities combined, are, respectively, 2.62, 1.74, and 2.22. Page 8. In the last two sections of the table, the prices for the year 1913 for steel beams and alcohol should read .015 and 2.51, respectively, instead of .105 and .251. December 15, 1921 Page 3, nth line from bottom of page. Read unwillingness instead of willingness. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH GENERAL BUSINESS CONDITIONS Advance Letter — October /, ig2l THE INDEX THE favorable movements, recorded in August, of the curves of our Index Chart, representing speculation and banking, have continued in September. These move- ments reflect improved fundamental conditions, and forecast increased business activity and higher commodity prices. The trough of the business depression has, very clearly, been passed. Ciirve B of oiu: Index Chart, representing business, reached the end of its very rapid, dechne in May; the +5 +4 +3 +2 +1 -I -£ -3 -4 1134567 89 10 II 12 I 234 5678910 II U I 23456789 I I9I9 I 1920 I I92I INDEX CHART. The items locating curves A, B, and C all refer to complete calendar monlks and the points are plotted in the middle of the spaces allotted to the months. volume of manufacture, as shown by the index pubUshed in our Special Letter of September 20, halted its abrupt dechne in February; prices of important commodities such as cop- per, tin, lead, rubber and cotton have recovered after a long period of weakness, and of late imemplojTiient in general has sUghtly decreased. These facts do not mean that business is not depressed but rather that we are rising out of the trough and beginning an upward swing. In other words, the depressed condition of business, which has gripped agriculture, mining, and manufacture, is now being followed by a recovery which, although very slow, is clearly registered by most of the available indices. If conditions in Eiurope were more nearly normal, and there were no possibiUty of a railroad strike in the United States, we could confidently count upon continued improvement. There are still a num- ber of disturbing factors in the situation which have made recovery slow and may reverse the present favorable indications. On this point no certain conclusion is possible at this COPYBIGHT, 192 1, BV HABVAED DNIVSKSirX V f< \ k / ^ 5g !>. C^n^ ;in I— 3 ^ r > V X ^ an ff «te =^ ^ A, ^ f ^ J 7 ^ '^'^ h u? p K > k k P -^ "T9' n ^ n k f 0^ A ^ ^ te ss, s** ^ X ^ «^ S / ^ ( S ^ as* n ■N •4 -3 pe pu at on Si 3f d- \ ^ks A B C J — 1 — 1 — 1 — 1 — 1 1 i__i UX Bank Cle\rings Shares Traded on NY. St L vrh \ Price of Induatria.! Stocks Bank Clearings Outside MYOhj _BrQdstreets Price Indices Rate on 4-6Mo5.Comni Paper Rale on 60-90Day Comm. Paper ^ -E U5 me >3S > >^ ■> k V ^ •/ f 1 r 1 1 1 1 1 1 1 1 1 time, but it is clear that the favorable movements recorded by our Index for August and September give considerable assurance that business conditions are on the mend. The bulk of the upward movement of curve A of our chart results from an increase of 5 per cent in the price of industrial stocks and i8 per cent in the niunber of shares traded on the New York Stock Exchange. New York clearings increased 3 per cent, although figures for September have usually been somewhat lower than for August. Clearings were somewhat increased by the payment of the third installment of the income tax on September 15. The continued downward movement of curve C is due to the fact that interest rates on commercial paper decreased at a time when the normal movement of such rates is upward, SPECULATION The price of railroad and industrial stocks advanced during the first part of September. Dow, Jones and Company's index of the price of 20 railroad stocks reached $74.30 on Sep- tember 13 ; declined irregularly to September 20, and has since risen, standing at $74.69 on September 23. The index of industrial stocks reached $71.92 on September 10, and has been lower since. The average for September is 5 per cent above that for August. The average price of 40 bonds has advanced since the first of the month, reaching $79.52 on September 26, a figure 1.6 per cent above the high point of August 4. Advance has been principally in railroad and pubUc utility bonds; that in industrials has been slight. Shares traded on the New York Stock Exchange numbered approximately 13,100,000 during September, an increase of 18 per cent over August. New York clearings were 15,040 million dollars, an advance of 3 per cent over August. A shght decrease usually occurs. Despite recent increases in prices, there is no evidence that outside buying is taking place on any considerable scale, and trading seems still largely confined to professional operators. Better earnings on the part of industrials seem necessary to attract outside investment and speculation, and high interest rates and the losses resulting from depression in business tend to retard any such development in the immediate future. BUSINESS The advance from the trough of the depression which was evident in August has con- tinued through September. The rapid increase in cotton prices culminated on September 10, with spot quoted on the New York market at 21.10 cents per pomid. Since that date the price has fluctuated between 19.50 and 20.50 cents per poimd, except on September 17, when it dropped to 18.60 cents and September 27 and 28, when it reached 21.55. The Liverpool and Manchester markets are reported active. The South seems disposed to sell at present prices. Receipts of cotton have been large, both at the cotton-collecting ports and interior towns. Continued sales at the current price levels wiU greatly facihtate Hquidation and rehabihtate the purchasing power of the South for goods from all sections. In the manu- facturing and distributive end of the industry, doubt exists as to the extent to which cotton goods, at the higher prices necessary to meet the increases in the cost of raw material, can be disposed of to the consumer. A reaction in the large primary markets for cotton goods has followed the price advance. In the wool manufacturing industry the smallest proportion of idle worsted spindles since the beginning of the recovery of that industry was reported for September i, while more wookn spindles were idle on September i than on July i and August i. Further advance in the production of iron and steel after the increase recorded last month is indicated for September. Pig-iron prices are firm; prices for iron and steel prod- ucts are still unsettled. An advance in sheet prices has followed that recently made in wire. Exports of iron and steel last month, however, reached the lowest (estimated) total since January 1909. Improvement has occurred in the non-ferrous metal markets; copper, lead, and tin have advanced in price, and the market for zinc is reported firmer. The price of crude rubber also has advanced. Unemplojonent for August was less than for July. The pay roUs reported to the Bureau of Labor for August were 1.4 per cent larger than for July. Unemployment increased in the automobile, woolen, silk, and cigar manufacturing industries, and the naining of coal; and it decreased in the iron and steel industry, carbuilding and repairing, cotton manu- facturing and finishing, and in the hosiery and imderwear, men's clothing, leather, boots and shoes, and paper making industries. The rapid marketing of wheat is reflected in exports of that product, especially during August. Monthly exports for the last three years, and the pre-war years 1913-14 are given below: Exports or Wheat from the United States (In millions) 1913 Quantity Value 1914 Quantity Value 1919 Quantity Value 1920 Quantity Value 1921 Quantity Value January . . February . March . . . April May June July August . . , September October . November December Total bus. 8 4 5 7 7 6 9 24 12 7 4 6 dollars 8 4 S 7 7 6 9 23 II 7 3 S dollars bus. 10 6 10 17 14 16 6 13 17 14 IS 10 99 9S 148 dollars 24 14 24 41 35 40 14 31 41 33 36 23 bus. 8 5 7 4 II 13 24 28 31 36 26 26 dollars 21 12 17 II 30 38 71 81 89 99 68 62 356 bus. 21 18 IS 18 26 25 2S S9 219 S99 dollars 45 37 28 29 41 40 37 81 The wheat crop for the last two years has been somewhat less than the 1915-19 average. Nevertheless, shipments since Jime 1920 have been large; and those for the first eight months of 192 1 have in each month exceeded those for 1920. Exports for August of this year, which total 59 miUion bushels, are more than twice those of August of a year ago, which exceeded August exports for the preceding decade. The value of exports for the two years is, however, almost exactly the same. Exports for the first eight months of 192 1 equal those of the entire year 191 9 and exceed those of 1913; they are greater than the cor- responding eight months of aU the years given above. Apparently 1921 shipments have been made earlier than in past years. Weekly exports for the week ending September 3 were larger than in the preceding four weeks, but they have fallen off decidedly since that date. As with wheat, the marketing of cotton has been rapid this year. The quantity ex- ported, compared with that shipped in other years, has not been relatively so great as is the case with wheat, but August exports — 49S,ooo bales — although less than in July, exceed those for August for each year of the last decade. Weekly exports have fallen off since the week of September 3. BANKING The strengthening of the position of the federal reserve system, and steady improve- ment in the general banking situation have continued during September. The reserve ratio has increased from 66.2 per cent on September 7 to 68.7 per cent on September 21. Con- tinued increase in gold holdings, and declines in deposits and federal reserve notes in actual circulation have brought about this increase. Although the last item was larger in the first week of the month than in the last week of August, because of holiday demand, a slowing up of the rate of decline maintained durmg the last six months is not discernible. On the other hand, net demand deposits of the reporting member banks seem to have shown an upturn since the latter part of August. The most striking development in the banking situation during September has been the lowering of the rediscount rate from sl to 5 per cent by the New York Federal Reserve Bank on September 22 and by the Boston Federal Reserve Bank on the following day. The Bank of England, which was expected to follow any such action, stiU maintains the 5 1 per cent rate. The lowering of the rediscount rate is in accord with the continued de- crease in interest rates throughout the country. The average monthly rates show a con- tinuous decline to date from August 1920 for 4-6 months commercial paper, and from October 1920 for 60-90 day commercial paper. Present rates are 5 2-5 1 per cent. It is noteworthy that the decreases of the last few months have been contrary to the usual sea- sonal movement. Prior to the foundation of the federal reserve system, July usually showed an increase over June; August over July, and September over August. Thus to date any tendency of the crop-moving season to enhance interest rates has not appeared. It has been reported that funds have flowed into New York from out of town banks dining Sep- tember. Call rates, which stood at 6 per cent during the first three weeks of August, have been lower since; they reached 4I per cent on the twenty-first, and are now 5. An indication of the pressure which has fallen upon the primarily agricultural districts since the beginning of the harvest season is suppHed by the borrowing of the federal reserve banks of the Minneapolis, Dallas, Richmond and Atlanta districts from the other reserve banks. On June 8 the total of such borrowings was 25 million doUars. This figure increased each week, with two exceptions, until it reached 71 million dollars on September 14. The borrowings of the Richmond bank, except for the last two weeks in August, have been the largest, though without a definite trend upward. The Dallas bank, the borrowings of which exceeded the Richmond bank for the weeks just mentioned, shows the largest in- crease between the two dates; its borrowings rose from 2 J to 25 millions of dollars. The highest figmre of the Minneapohs bank was 18 millions on August 24. The Atlanta bank has increased its borrowings from 2 millions on August 3 to 16 millions on September 24. It is the only bank showing an increase after September 14; the decline for Minneapolis has been steady since August 24, while the figure for the Dallas bank shows a decided decrease since August 31. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH Chakles J. BxiLLOCK, Chairman Warben M. Peksons, EdiUir CAMBsiDaE, Massachusetts, U.S.A. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH GENERAL BUSINESS CONDITIONS Advance Letter — October 75, ig2l THE INDEX CONTINUANCE, during September, of the slow improvement in economic conditions which has taken place during recent months is indicated by the very slight upward movement of curve B of our index chart, representing business. Since last May, when the trough of the present depression was reached, its movement has been upward, showing clearly the improvement since that month. The process has, however, been extremely slow. It is clear now, from the curves of oUr chart, that we may expect continued advance, but that the rate of improvement, during the coming few months at least, wiU not be rapid. .- U /* s ^' ^ :^ ^. 1 1 1 1 C- Banking ^ V ■^ r J =: ^ V ^^ r =^ V \ A f t J H> "J U L^ > s, k r \ ^ Tf^ ^ T k f N Sss as^ s^ ^ t \ i t w I r ^ 5:^ u ^ ^ -3 SSSS Dec ssa ,ulc sss itio ssa 1 ■«s_ } k libs S^ \ i \ fNyCituBcink Clearings A 1 Shares Tresded on WystExch LPrice of Industrial Stocks □ fBt^nk Oeari ngs Outside N.Y. City 1 BradstreeVs Price Indices -^ \ -B jsi ie; .s \ ^i ■> - C ■fR ate ate on on 4-C 60 -90 )S.( Da Ion gc orrT ^^ ep iper V ^ B^ ^ Although curve A, representing speculation, has moved upward for the last two months, the recovery has been slight. The very marked decline in curve C, representing banking, reflects dvuring the past three months a decline in interest rates on commercial paper which occurred at a time when their trend is normally upward. Even after the substantial de- crease (I per cent) during the last week of September, rates are still high (s| to 51 per cent) for this period of the economic cycle, and their height operates to retard rapid expansion of speculation or business. The curves of our chart, then, indicate that the betterment in fimdamental conditions which in Jxme was technical only and did not become substantial until August, has been maintained in September; that the improvement during the month has been very slightly more than that which normally takes place at this time of the year; and that the advance noted will continue diuing coming months, although probably at a very moderate rate. These favorable indications may be reversed by vuitoward developments in Europe, or a COFXBIGBT, 193I1 BY HAKVAKD UNIVEESITX raikoad strike in the United States; but, apart from such unfavorable contingencies, assur- ance is given by our index chart that business conditions are slowly righting themselves. In judging the present outlook account must be taken of a new factor which may prove extremely important in accelerating revival. It is evident that the federal government is anxious to do what it can to restore business prosperity; and, while some of its meastu'es may not be effectual, others may bring about conditions, such as easier money and credit, which will appreciably hasten recovery. The pressure of agricultural and other interests is not alone responsible for the policies which seem to have been adopted at Washington; it is, indeed, quite inevitable in such times as these that the assistance of the government should be invoked and rather freely granted. With the agencies now at its command, and with the powerful federal reserve system in operation, it is probable that the government can stimiilate business activity in no slight degree at this juncture. On a falling market, such as we had last winter and spring, stimulants would have had no effect; but now that the bottom has been reached and recovery has begim, it is impossible to doubt that govern- mental action may be effective, at least for the time being. The ultimate results may be good or bad; that question cannot be considered in this letter. The thing of present im- portance is to recognize that we may be entering upon a period of renewed credit expansion which may accelerate materially the recovery of business, which is now proceeding slowly under the influence of forces that normally operate at this phase of a business cycle. Gold imports, if they continue, will supply conditions favorable to the success of the effort to bring about what is frequently, and very inacciurately, referred to as "secondary" inflation. SPECULATION The price of both railroad and industrial stocks has fluctuated only sUghtly during the past few weeks, Dow, Jones and Company's index of the price of 20 railroad stocks reached $74.69 on September 23, the highest point since August 2. It decreased to $73.45 on the twenty-eighth, and has not advanced beyond the figure of the twenty-third since. The index of the price of industrial stocks has not shown a distinct trend since it reached $71.92 on September 10. On October i it stood at $71.68, and has since been lower. The steady increase in bond prices during September culminated on the twenty-sixth, when Dow, Jones and Company's index stood at $79.52. It fell in the next two days to $79.22, and Closing Prices of Liberty and Victory Bonds New York Market Issue First Second (cvs.) Third (reg.) Fourtli Victory (reg.) Rate 3i* per cent 4J per cent 4i per cent 4I per cent 4I per cent 4I per cent Date of maturity 1932-47 1932-47 1927-42 192S 1933-38 1922-23 Date September i September 9 October i $87.40 87.26 t 88.46 89.88 $87.90 87-94 90.82 94.60 $87.84 87-94 90.52 93.00 $91.80 91.88 93-74 95.16 $87.92 88.08 90.90 93-30 $98.80 98.88 99.28 99.20 October 8 Increase September i to October 8 2.8% 7-6% S-9% 3-7% 6.1% 0.4% * Tax exempt. t Low for the month. has since been higher. The average price for September was 1.4 per cent above that for August. The movement in the prices of Liberty bonds which, except for the first 3|'s, has been ahnost continuously upward since September i, is one of the most noteworthy features of the bond market. The table above presents closing prices of various issues on four days during September and October. Large funds held in corporate treasuries, which have not found outlet in business channels, are being used in the purchase of these securities, and such bujong is regarded as a factor of importance in increasing prices, in addition to the general feeling that prices are low. Some of the purchases recorded have been of large amounts. BUSINESS The slight upward movement of curve B is caused by an increase in Bradstreet's price index, which stood at $11.09 on September i and $11.19 on October i. Of the groups re- ported, fruits, textiles, metals, coal and coke, naval stores, building materials, and miscel- laneous products advanced, while breadstuffs, live stock, provisions, hides and leather, oils, and chemicals and drugs declined. Outside bank clearings for September were 13,210 millions of doUars, an increase of 3 per cent over the amount for August. This is merely the usual September increase. There are various indications of business improvement during September and thus far in October. The spot price of middling upland cotton on the New York market, which reached its highest, 21.55 cents per pound, on September 27 and 28, has been lower since those dates, but has not again fallen below 19.5 cents. The disturbance wrought in the market for cotton goods by the rapid increase in. the price of the raw material during the first ten days of the month, and which prevented many mills from quoting prices for several days, has been largely rectified, and new prices have been quoted on several branded lines. Some lines, however, were still heM "at value" at the end of the month. Increase in the prices of cotton goods has not been rapid, and sales at the higher levels were comparatively small dviring the first three weeks of the month. The production of pig iron in September was 985,529 tons, an increase of 3.3 per cent over August and of 14 per cent over July. Eighty-two furnaces were in blast on October i, in contrast to seventy on September i. Further increase for October is indicated. Unfilled orders of the United States Steel Corporation, which totaled 4,560,000 tons on September 30, showed a gain over August which, although small — 0.7 per cent — is significant because the first recorded since July 31, 1920. Steel ingot production rose 3.2 per cent. The increased activity in steel wire, sheets, tin plate and pipe continues, without corre- sponding improvement in rails and steel plate, shapes, and bars. The price decline in the last three products appears at least to have been checked. The production of bitxmiinous coal for the month was approximately 35,105,000 tons, an increase of 1.6 per cent over the August figiure. The rapid marketing of wheat has brought a sharp decHne in price since September 9. On that date the price of no. 2 red wheat, spot, on the New York market was $1.47^ per bushel. On October i the price had fallen to $1.28^ and on the third reached $1.15. It has been higher since. Much less liqmdation has resulted from the large September grain movement into Minneapolis than nomially would be expected. The causes assigned for this fact are lower prices, especially for oats and other small grain, higher cost of production, and higher railroad rates, all resulting m a lower return to the farmer. More pronounced liquidation is expected later. A slight decrease in imemployment is recorded during September. Weekly car loadings during the month, except for the week of September lo, in which the Labor Day hoUday occurred, show a continuous increase. For the week of September. 24, a decrease is reported in the loading of grain and grain products, which is in accord with the fact that the crop has been largely moved. Loadings of grain, however, exceeded those for the corresponding weeks in 1919 and 1920. The largest gain recorded for the week was in the loading of mer- chandise and miscellaneous freight; next in order followed coal, Uvestock, forest products, ore, and coke. Bank clearings outside of New York City, which form probably the most reliable index of the volmne of business transactions to be foxmd, are given below. For the covmtry as a whole they show the normal seasonal advance of approximately 3 per cent. New England clearings, however, remained almost the same as for August, while those for the Middle West showed a sHght decline. Clearings for southern cities showed a gain of 14 per cent, reflecting improved conditions in that section. For cities m the remaining districts — Middle, Pacific, and Other West — clearings showed an increase only slightly greater than the normal seasonal advance. Outside Clearings: Actual Figukes {Unit: $1,000,000) Aug. I02Z Sept. zgai Ratio of Sept. to Aug. Middle * 2,909 3,014 103.6 New England 1,259 i>2S8 99-9 Middle West 3,672 3,591 97.8 Pacific 1,41s 1.479 ^°4-5 Other West 1,680 1,742 103.7 Southern 1,869 2,128 113.8 Total * 12,804 13,212 103.2 * Excluding New York City. BANKING Since the first of October, two federal reserve banks have annoimced reduction of their rediscoimt rates. The Philadelphia bank lowered its rate from 55 to 5 per cent on October 5, and the Minneapolis bank decreased the rate on paper not secured by government issues from 6J to 6 per cent on October 6. The rediscoimt rates of the Boston, New York, and Philadelphia Federal Reserve Banks are now 5 per cent, those of Cleveland and San Fran- cisco 5^ per cent, and those of the remaining banks 6 per cent. The federal reserve ratio stood at 69 per cent on October 5, the same figure as for the week before. This was due to an enlargement of the circulation of federal reserve notes, not fully offset by a decrease of deposits and an increase of gold reserves. Circulation and deposits are both lower than in the corresponding week in September. Call loan rates during the first week in October varied between 4J and 5^ per cent, a decrease from the preceding week, but on the eleventh they were 6 per cent. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH Chabies J. BraiocK, Chairman Wabben M. Persons, Editor CAUBsmoE, Massachusetts, U.S.A. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH GENERAL BUSINESS CONDITIONS Advance Letter — November /, ig2l THE INDEX NO pronounced change in business conditions is evidenced by the October movements of curves A and C of our index chart, representing speculation and business, respec- tively. The decline in curve C, reflecting decreasing interest rates, is at approximately the same rapid rate as that maintained since June. At this stage of the business cycle, a funda- mental condition of improvement is a pronounced decrease in money rates, such as has been taking place since April. Even though such a decline has occiurred, it has been sharp only nd *^ ^ ^ s \ ^ ^ 1 1 i 1 C- Banking ^'^^^v^ 1 . 1 -^ ^ f^ k r y^ r ^ < \ r =5 r ^ =^ H n H n N S m r \\ > s r \ \ ^ ■><1 ^ r H / i ^ S&! sss\ s^ \ ^ \ i t r J r s ^ ^ ;^ N^ ^ -3 ssss oec p h ^ ^ ssr y I te p \ J J \ fNYCiiy Bnnk Clearings A 1 Shares Traded on NY St.Excl- [price of IndustriaJ Stocks o rBtxnk Qeari ngs Outside NY Cih ° 1 Bmdstreet's Price Indices 1 \ -B uai qe; s J > ^k. ■> k - — C 1R! on on 60 -90 )S.( Da ;on gc oni' :^^ er kper - ^ «iw a^ f 3 + 5676910 11 12 I SS'IS 7 e 9 10 5 e 7 a B 10 ; since last spring and actual rates are stOl above those which in pre-war years signaled steady advance in business. Nevertheless the decline is the most important evidence, at this writ- ing, of futtire improvement. The decreased rates for the current borrowings of the federal government, noted below, are regarded as favorable to this development in that they make investment of private ftmds in such securities less attractive, and tend to draw them instead into industrial channels. Curve A, representing speculation, has declined slightly during October, reversing the slightly upward movement maintained duruig the preceding two months. Neither of these movements, however, has been pronounced, and less importance is to be attached to the sidewise movement of cmrve A than the decidedly favorable movement of curve C. The extent to which curve A has been affected by the prospect of a raibroad strike is difficult to gauge, but it is probable that, in spite of professed disbelief in its possibihty, specidative activity has been decreased by the threat of a general walk-out. COFYRXGHX, XO^Ii BY HARVARD I7NIVERSITY The decline in curve A, representing speculation, is due largely to the fact that the usual seasonal advance in New York bank clearings has not occurred; these show an increase of 7 per cent, as against the usual advance of 20 per cent. The failure of the October figure to show the usual pre-war increase over September, however, is partly due to a non-business factor — income tax payments — which increased the September base of comparison. The number of shares traded on the New York Stock Exchange in October was about the same as in September. The average price of industrials was 2 per cent higher. The decline in curve C, representing banking, is caused by decreases in the actual rates on both 4-6 months and 60-90 day commercial paper. The first of these rates is usually the same in September and October, the second usually somewhat higher in the later month. The failure of the threatened raUroad strike to materialize has removed a disturbing factor which was affecting appreciably the immediate business outlook. It appears, how- ever, that the imions have received assiu-ance from the Railroad Labor Board that further reduction of wages will not be even considered for many months, and it is probable that this was the real objective of the imion leaders in the whole affair. Nothing, therefore, has been reaUy settled; and the only thing gained has been the clear demonstration that pubUc opinion was overwhelmingly opposed to a strike. Perhaps this fact will lead to federal legis- lation making more adequate provision for handling railroad labor problems. SPECULATION Between the first and seventeenth of October, the Dow- Jones index of the price of 20 railroad stocks showed an almost continuous decline, from $74.58 to $70.00. There has been a recovery since, though not sufficient to bring prices to the level of October i. A less pronounced decrease in the first half of the month occurred in the price of industrial stocks, but this was followed by a substantial increase, which by October 29 had brought the daily average up to $73.93, which was 3 per cent above the figure for October i, and the highest point reached since the end of May. Shares traded on the New York Stock Exchange have numbered about 500,000 daily, though showing higher totals at the middle and end of the month. The index of the price of forty bonds, which had shown a pronounced in- . crease during September, stood between $79.30 and $79.50 on the first ten days of the month, and then declined sharply reaching $78.77 on October 18. Prices have been some- what higher since. The decrease was due largely to the decline in industrial bonds; highest grade rails have, in fact, uicreased slightly during the course of the month. BUSINESS During the last two weeks there seems to have occurred a slackening in even the slow rate of improvement maintained since the middle of the summer. Grain and cotton prices have shown a pronounced falling off from the high point of the season; prices of commodi- ties in general have weakened, and activity in the steel and cotton goods markets seems to have lessened somewhat without as yet affecting production. It is difficult to gauge the extent to which the slowing down observed was the direct result of the threat of the raihoad strike. Many business men did not expect a strike to occur under conditions so unfavorable to the xmions. In at least one case — retail sales of coal for domestic consumption — ex- pectation of the strike had a stimulating effect. Another element, which is reported to have retarded business expansion during recent weeks, is the expectation of lower railroad rates. On October 22, the Interstate Commerce Commission rendered a decision reducing rates on grain, grain products, and hay. These rates are to be put in force on or before November 20. The principles upon which the reductions in the grain cases referred to above were based are such as to lead to the anticipation of further reduction. Rates on hard wood are among those now imder consideration. Reductions in rates have an important effect upon the market for commodities in which freight rates form a large percentage of cost, either to the manufacturer or buyer, because of the bulk of article or of the materials of which it is made. Buyers of these commodities tend to hold off, either tmtil the reduction of rates will directly affect their expenses, or in expectation that price reduction, corresponding to the rate reduction, wiU be made. It is reported that the sales of steel have fallen off in the last two weeks, in the expectation of lower rates on steel and lower prices, following the recent reduction of 28 per cent on iron ore. This reluctance to purchase occurred at a time when the threatened strike might have been expected to stimulate sales. Recently the market for cotton goods has been quiet. During September there was an increase of 4 per cent in cotton consumption, and an increase in employment of 2 per cent in cotton manufacturing and 3 per cent in cotton finishing. There has been a broadening of the woolen goods market with the demand brought on by cooler weather. Demand has favored moderately priced materials, while a feature of the wool market has been the large sales of medixun grade wool. Idle spindles in the wool-manufacturing industry on October I formed 22.4 per cent of the woolen spindles and 8.6 of the worsted spindles. The figure for idle woolen spindles is the same as on September i, and is about equal to the average for the first half of 1914. There has been a pronounced decrease in idle worsted spindles since August, when they formed 13.3 per cent of total worsted spindles. The October i figure is weU below the pre-war normal. More favorable developments are to be found in the lumber and oil industries. The former is active, with prices in some hnes higher. The greatest improvement and advance is to be found on the Pacific coast; a feature of this trade is the increased sale in the eastern markets of western limiber, which has been transported via the Panama Canal. The yeUow pine Ivunber industry of the southwest has likewise shown increasing business. Advances in crude oil prices, after declines since the first of the year, began in the Peimsylvania field on September 27. These have been followed in other fields, and the prices of refining prod- ucts have in some instances moved upward. It is reported that sales of the crude product have slackened since the recent increase. In the steel industry the market is, on one hand, reported to be less active than recently, while, on the other hand, production is reported to be holding up well. On October 24, Judge Gary announced that the U. S. Steel Corporation had cut the price of standard rails $7.00 per ton. The price is now I40, compared with a pre-war price of I30 and a maximum, dviring the war, of I57. On October 11, the Iron Age reports price recessions in wire rods, steel bars, and tin plates, and states that there has been price cutting in line pipe. The copper market has weakened somewhat. The production of bituminous coal has increased steadily since the first of the month. BANKING Average interest rates on commercial paper for October were lower than for September; they are usually slightly higher. Rates have remained throughout the month at about the height reached in the last week of September; on commercial paper the general rate is now si to 5| per cent, with some sales at 6. Call loan rates have ranged from 4 to 6 per cent during the month; they are now 6 per cent. Interest rates on treasury certificates have been reduced f of one per cent from the rate of September i, because of large oversubscrip- tion to the last issues. Seven and one-half and twelve months certificates issued on August I bore rates of si and si per cent, respectively; six and twelve months certificates issued on September 15, 5 and 5i per cent; while the five and ten and one-half month certificates to be isSiied on November bear rates of only 4i and 4! per cent. On October 26 the federal reserve ratio was 70.8 per cent. This high ratio was largely the result of increased gold holdings, which formed about 95 per cent of the total cash re- serve used to obtain the reserve ratio. Gold reserves amoimted to 2,733 miUions of doUars on the fifth, shghtly less on the eleventh, and 2,786 on the twenty-sixth. Of the other ele- ments in the ratio, reserve notes have decHned since October i, while total deposits have varied irregularly, but on October 26 were higher than during the preceding weeks of the month. Important items of the last report of the Federal Reserve Board follow, compared with the same items at the time of greatest expansion: Combined Federal Reserve Banks (In millions of dollars) Maxiuum Amount Date (1920) October 26, xg2X Percentage change Total earning assets , Total bills on hand F. R. notes in actual circulation Reserve ratio Excess reserves 3.422 3,127 3,40s 42.2% 201 Oct. 15 Nov. s Dec. 23 MiNIMlJM May 14 May 14 1,562 1,371 2,409 70.8* 1,365 -54-4 -56.2 -29-3 Reporting Member Banks (In millions of dollars) Maxhtom Amount Date (1920) October 19, 1921 Percentage change Loans and investments f Aggregate deposits Bills payable and bills rediscounted with F. R. banks . Ratio of accommodations at F. R. banks to loans and investments 17,284 14,600 2,278 13-5% Oct. IS Jan. 16 Nov. 5 Nov. 5 11,477 13,350 829 5-6 -33-6 - 8.6 -63.6 Computed on revised basis. t Including bills rediscounted with federal reserve bank. harvard university committee on economic research Chaiu.es J. BtJtLOCK, Chairman Wareen M. Persons, Editor Caubkidoe, Massachusstts, U.S.A. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH GENERAL BUSINESS CONDITIONS Advance Letter — November 75, ig2I THE INDEX THE improvement of business last month has been somewhat less than the usual sea- sonal amount. This is shown by the slight decline in October of curve B of our index chart, representing business. The movement of our curve during the last three months has thus been sidewise. The decUne in October was due to the fact that outside bank clearings did not show the usual seasonal advance. They rose from 13,210 to 14,240 millions of dol- lars, a gain of 8 per cent, against a usual increase of twice as much. During the preceding +& I f^ S k , ^ ^ 1 1 1 1 C- Banking >^ 1 . 1 ■^1 ^ ^ <' y, > ^ s. 1 =:= r "^ 1 H =^ H i r\ n n r \ "V n f" "\ ^ s r ^ \ r-< ^ n k ^ t N Ste ^ s^ ^ t -^ 'ISTK r 1 \ i t r jt r ^ %s^ n W ^ -s oec IS xtio n ) k fife 3^ \ I \ ■ A \ Shnris Tmned on NYSt.Exch \Prv:e of IndustrloJ Stocks iBs-nk Oeari ngs Outside N.Y. City ^ ^ -B JS ie; a \ ^. -N \ - S" -b C ■ R ate Qte on on 4-f 60 3M( -9C Do Con yc n1. cm '^^ en iper \ S» ^ ^ /* month, although increased by the September payments of the income tax, they had shown only the usual seasonal advance. October figures, since they are compared with those of a month of income tax pa3Tnents, natiurally show a less than seasonal increase. To this result, imdoubtedly, another factor, namely, the threat of a railroad strike, contributed. The im- portance of this disturbing factor is vincertain. Probably we must have the November figures before we undertake to decide how far the volume of business this fall has risen above or fallen below the usual seasonal increase. Bradstreet's index of commodity prices on November i stood at $11.35, or 1.4 per cent higher than on October i. This slight increase was not sufficient to offset the small figure for outside bank clearings. Marked changes in commodity prices were increases in provisions and textiles, and a decrease in breadstviffs. Of the 106 commodities quoted 32 advanced, 27 decKned, and 47 remained unchanged. The failure of business to show decided improvement since the trough of the depression last May is not unexpected. A sidewise movement or halting improvement of business was POFVSIGHI, I93I, BV BABVASD UNIVEKSITX indicated early in the year by the course of curve C, representing banking. In January we stated that, "If we are to have a marked revival in business, we must first have a consider- able decUne in rates on commercial paper; and unless low rates continue for a year or more a revival in business can scarcely develop into a boom." It is only since July that curve C, representmg interest rates on commercial paper, has shown a pronoimced decUne, and at present the evidence indicates that the downward movement will continue. The recent issue of treasmy certificates bearing interest at 4! and 4I per cent, rates well below those of previous issues and of other securities on the market, were greatly oversubscribed. During the first week of November, the rediscoimt rates of the federal reserve system were lowered throughout the country. The decreases amounted to i per cent in some districts and | per cent in others. All these factors are evidence of the easier money conditions, which favor continued improvement in business. The reduction of rediscovmt rates by the federal reserve system has followed the building up of cash reserves and general Uquidation of loans, although the latter movement has not been marked in the predominantly coimtry districts. The position of the federal reserve banks in these districts, however, has recently shown some improvement. Under such con- ditions, and with business and agricultural interests lurging the reduction of rediscoimt rates, the action taken seems inevitable. The lowering of rediscount rates was immediately fol- lowed by reduction of commercial paper rates on the New York market. There can be no doubt that lower interest rates will have a stimulating effect on general business conditions; that effect is abeady evident in the present strong and active bond market. SPECULATION Security prices, trading on the New York Stock Exchange, and New York clearings reflected somewhat increased speculative activity in the first two weeks of November. Since the middle of October, when both stock and bond prices slumped, their upward movement has been pronounced. Recently the bond market has displayed undeniable strength. The Dow- Jones average of the prices of 40 bonds rose from $78.77 on October 18 to $81.21 on November 12, the highest point touched in more than two and a half years. The average price for October as a whole, however, was lower than for September. Railroad and in- dustrial stock prices made even greater percentage gains than bond prices, from $70.00 and $69.46, respectively, on October 17 to $73.58 and $75.75 on November 9. On October 28, following the news that railroad strike orders had been withdrawn, the niunber of shcires traded on the New York Stock Exchange exceeded 1,000,000, for the first time in over four months. Since then the daily average has been 600,000 shares. BUSINESS The most important advance during October is shown by the production statistics of iron, steel, and coal. Steel ingot production rose 38 per cent, pig iron 26 per cent, and bitu- minous coal 25 per cent. Forty-three per cent more pig iron was produced in October than the low total reached in July. Ninety-six furnaces were in blast on November i, as against 82 on this date the month previous. Price advance during the month has been slight, and in two important commodities — wheat and cotton — there has been some recession from the highest point of September. The daily price of wheat during October has been well below the average for the preceding month, and lower during the last two weeks than in the first two. A low of 99^ cents per bushel for December futures was reached on November 3 in the Chicago market. Cotton prices have likewise shown a tendency to dechne. The spot price of middling upland on the New York market reached a low for the month of 18.5 cents on the eighteenth; it has not exceeded 20 cents since the eighth. Reports on the liquidation of farm loans are not favorable, although there has been a marked strengthening of the position of the federal reserve banks in the predominantly agricultviral districts. Congestion of wheat storage facilities is reported from many wheat collecting centers, and some grain has been reshipped from Montreal and New Orleans to the Central West. The further decreases which have taken place in the prices of certain farm products mean further curtailment of the farmers' purchasing power, and hence of the market for manufactured articles in the country. The London Grain Reporter's estimate of the world wheat crop is somewhat larger than last year, although European production is reported to be considerably smaller. BANKING The most striking recent development in the field of banking has been the general re- duction of the rediscount rates of the federal reserve system. On November 2, reductions were announced by the federal reserve banks of New York, Philadelphia, St. Louis, Chicago, Kansas City, San Francisco, Richmond, and Atlanta, followed, on November 3, by the federal reserve banks of Boston and Dallas, and on November 4 by those of Cleveland and MinneapoUs. It is significant that during the same week the Bank of England reduced its rate from 5§ to 5 per cent. Present and previous federal reserve rediscoimt rates are as foUows: {Unit: one per cent) District Present rate Previous rate District Present rate Previous rate District Present rate Previous rate Boston 4i Ah 44 5 S S S 54 Richmond Atlanta Chicaffo 54 54 5 5 6 6 6 6* Minneapolis Kansas City Dallas 54 5 54 5 6 New York Philadelphia Cleveland 6 6 St. Louis San Francisco . . . 54 * Bankers' acceptances, s4 per cent. The 4I and 5 per cent rediscount rates, at least, are weU below recent open market rates on short time paper. For October rates at New York averaged 5.94 per cent on 4-6 months commercial paper, and 5.62 per cent on 60-90 day paper. These rates represent a slight decrease from the previous month, and a continuous decline during the past year for 60-90 day paper and during 14 months for 4-6 months paper. During the week ending November 4 rates on the earlier maturities ranged from 5 to 5i per cent, | of one per cent lower than in the preceding week, while rates on the later matmrities stood at 5 J per cent, | to | of one per cent lower. During the same week, the New York call loan rates varied from 4I to 6 per cent. The United States Treasury certificates of indebtedness dated November i were heavily oversubscribed, the aUotmenta amounting, in all, to less than one-third of the total subscrip- tion. In view of the low rate at which the certificates were issued — 4 J and 4I per cent — this result was remarkable. It indicates the extent to which funds are at present seeking investment in securities not involving risk. Subscriptions were principally from banks, though a good demand was reported on the part of corporations. For many months there has been a steady increase in the gold reserve of the federal reserve system, a general reduction of loans by member banks at the central institutions, and an easing of money and credit conditions. The position of the federal reserve system, as a whole, has been steadily growing stronger as a result of these factors. The favorable development, however, has been "spotty," at least up to recent weeks. The position of the reserve banks in the predominantly city districts has shown the most marked and continu- ous improvement, while that of the reserve banks in the predominantly coimtry districts has been much less favorable. Liquidation has been accomplished much less thoroughly in the country districts than in those dependent on manufacturing. Low prices for agricul- tural products, except for cotton, have deterred such development, and the shortness of the cotton crop doubtless has been to some extent an unfavorable element in the south. The reserve ratios of the various banks, therefore, show wide divergences. At present, the New York and Boston reserve banks have the highest adjusted reserve ratio ^ — over 80 per cent in each case. During the past six months the reserve ratio of these banks has increased fairly continuously, and that of New York rapidly, while at the same time the Chicago and San Francisco ratios have likewise shown a steady improve- ment, though one which has not been so rapid. The Philadelphia and Cleveland ratios reached the lowest point in the last half-year on Jime i and 15, respectively, and have been higher since. None of these six banks has shown a reserve ratio of less than 50 per cent since May i. Of the remaining six banks, fovir — Atlanta, Richmond, Minneapolis, and Dallas — show a pronoimced decline from May to August. The last two reached low points for these months on August 24 and 31, respectively. The ratio of the Richmond bank reached its lowest for the last half-year on September 14, and that of the Atlanta bank on September 28. Each of these four banks has shown a recovery since the dates mentioned. That of the Dallas bank has been most pronounced. Its ratio rose from 9.2 per cent to 33.4 per cent on November 9. MinneapoHs has increased from 22 per cent to 47.9 per cent, and Richmond from 26 to 41.4. The increase in the Atlanta ratio has been least. The ratio of the St. Louis bank has increased steadily since July 6, and is now 68 per cent. The Kansas City bank ratio moved up during the early summer, reaching a high point, for the last six months, of 60.8 per cent on August 24. It is the only bank showing a marked decrease since that date. The present ratio is 48.4 per cent. On November 9 the reserve ratio of the combined federal reserve banks was 71.4 per cent, an increase over that of the preceding week. This increase was due to increased gold reserves and a decrease in total deposits, while federal reserve notes in actual circulation increased slightly. ' All the reserve ratios quoted for each of the twelve federal reserve banks are computed from "adjusted" figures. That is, the reserves of lending banks are increased, and of borrowing banks are decreased, respectively, by the amount lent or borrowed. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH CAMBRIDGE, MASS., U.S.A. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH GENERAL BUSINESS CONDITIONS Advance Letter — December i, 1021 THE INDEX CHART THE movements of the speculation and money curves of our index chart during Novem- ber give indications more favorable to business advance than at any time during the past four months. Curve C, representing money, again moved downward, at about the same rapid rate that has been maintained since June. Such a decline usually precedes, and hence forecasts, a marked upward movement of curve A, representing speculation, and then a rise of curve B, representing business. This effect has now clearly shown itself in u f" s k ^ N "^ c ^LnJy ^ ^ ^ V ^ f y, r > ^ ^ r ^ '/ ^ ^ =v^ i f 3 n f % \ V L h > s \ V ,_ ^ ^ r k ^ /. w V «&s saj^. ISl^ ^ \ rrm "?S7h p \ \ i € r J f ^ ^ i^ r ^■«5, ^ t:^ pec -ulc itiO h ^ ^ k g. \ i t«s ^ \ A B [Price of Industral Stocks Ba,nk Clearings Oiiside N YCity Bradstreetfe Trice Indices ^ -E usi ne; s ^ ^ ■> i, C teo teo n4- n6C 6M -9 %i lorn IP om m r per \ ^ *- ^ r « 10 II 12 I B 9 10 II IS 2 3 4 5 Forecast: Since June curve C, representing money, has moved downward sharply, reflecting decreased interest rates. A decline of curve C usually precedes, and hence forecasts, an upward movement of curve A representing speculation, and then of curve B, representing business. Curve A has risen substantially in November, after three months of slow and halting recovery from the low point reached in July. This upward movement of curve A shows, for the first time in a significant fashion, the expected result of easier money conditions, evidenced by the decline in curve C. The movement of these two curves forecasts a significant rise of curve B, representing business, which has been recovering very slowly from the depth of the depression reached last May. curve A. The latter moved upward during the month to a position higher than it has held during the preceding four months. The advance diurmg November has been substantial, after three months of slow and halting improvement from the low point reached in July. Although a rise of curve A for two consecutive months would give greater assurance of the permanence of the improvement shown, the present upward movement is significant, be- caiise certain conditions conducive to speculative activity are now present. Interest rates have decreased, bond prices have risen since the end of Jtme, and the investment market is now strong and active. On the index chart we find that the curve representing speculation has moved sharply upward and has crossed the cturve representing money, COFYKIGHI, igil, BV BARVAKO UOTVESSETy which is sharply falling. Such a movement of the two curves normally precedes an active stock market. Their present relation is just the reverse of that which obtained in January 1920 when curve C, rising abruptly, crossed curve A, which was dechning rapidly from the high level of the previous year; and this changed relation expresses very accurately and strikingly the difference between the business conditions existing at the two dates.^ SPECULATION Security prices have shown a substantial and steady increase since the slump in the middle of October. The Dow- Jones index of the price of 40 bonds rose from $78.77 on October 18 to $84.04 on November 29, an advance of 6.7 per cent. The advance dtiring November has been largest for railroad bonds, but public utiUties and industrials each gained over 4 per cent from the first to the twenty-nmth. Industrial and railroad stock prices have shown an advance sunilar to bond prices. The Dow- Jones index of the price of 20 industrials increased from $69.46 on October 17 to $78.01 on November 28, and the price of 20 raihroad stocks from $70.00 to $76.06. The number of shares traded on the New York Stock Exchange daily in November has been well above the level of the first three weeks of October, and the total for the month, despite three hohdays, is 18 per cent above that of the preceding month. BUSINESS The railroad situation has been considerably clarified by the developments of the month, which gave promise of prompter and more sweeping readjustments than had been generally anticipated. On November 16 the Association of Railway Executives annoimced their plan for a temporary ten per cent reduction in freight rates on agricultural products. They also petitioned the Interstate Commerce Commission for a rehearing on the reduction of 20 per cent in the rates on hay and grain in western and intermountain territory, and asked for a general inquiry as to the further rate reductions which the Commission could require. The Commission has subsequently postponed the date on which the reduced hay and grain rates are to go into force from November 20 to December 27. On November 23, it an- nounced that hearings would be held at its ofi&ces December 14 to determine "whether and to what extent, if any, further reductions in rates, fares and charges can lawftdly be required by order or orders of the Commission .... upon any commodities or descriptions of traflSc." Obviously further readjustment of railroad rates involves consideration of the wage scale of the raihoad employees. The threatened strike was averted apparently on the understanding that the United States Railroad Labor Board would give its attention to the revision of working rules before considering appUcations for wage reductions. The Board has drawn up working rules, which become effective December i, for the six shop crafts. These include the carmen, sheet metal workers, machinists, blacksmiths, boiler- makers, and electricians; approximately 400,000 men are affected at present. The issuance of these rules paves the way for consideration of requests for changes in the wages of the shop crafts. '■ All the series upon which curve A is based advanced during November. New York bank clearings were about s per cent above those for October; they are usually s to 6 per cent less. Shares traded on the New York Stock Exchange in- creased about i8 per cent, and the Dow- Jones index of the price of 20 industrial shares rose almost 7 per cent. The decline in curve C, representing money, was due to a substantial decrease in the rates on commercial paper, at a time when the decrease is usually slight. Indications of business improvement during October continue to appear. Unemploy- ment was generally less than in September. The value of the building permits issued in October exceeded that for any one month except April 1920. The gain over September was 25 per cent, the largest advance between these months since 1908. The increase was due to the large October total for New York City. A strike of the garment workers in New York City has to some extent curtailed pur- chases of wool fabrics during the past two weeks. The percentage of idle spindles on November i amounted to 20 per cent of the woolen spindles and 8 of the worsted, a slight improvement over October i. There has been a substantial recovery since spring in the carpet and rug looms in operation; the percentage of idle loom hours on November i, however, was still almost one-third of the total reported. Production in the iron and steel industry is holding up, although new business is appar- ently less than current output. Leaders in the industry have stated that the directly imfavorable effects of limitation of armaments on the trade would be overbalanced by the ultimate benefits to be derived. Serious readjustments would doubtless be reqiiired in plants specializing in armor plate and ordnance, but the tonnage of steel required for naval construction is only a fraction of one per cent of the yearly output of the industry as a whole. Railroad bujong is a promising factor on the steel market; it includes roUing stock, track equipment, and rails. Prices have felt the effect of competition, arising from the desire of producers to keep up the rate of operation. Those for steel, as shown by the composite figure of the Iron Age, are still declining, while pig iron prices, though slightly below the level reached at the end of September, are weU above the low of August. Fabricated struc- tural steel contracts dviring October were at 54 per cent of rated capacity, the highest monthly percentage shown for the year. Exports of iron and steel were slightly greater than in September, though stUl less than one-third the amovmt shipped in the same month in 1920. Copper prices advanced during November. Lead sales have been less than in October, but the decline is not as great as is usual at this time of year. Both lead and zinc prices have declined slightly at St. Louis, perhaps because a reduction of railroad rates on these metals is expected. The situation in the coimtry districts remains much the same. Buying has been re- stricted by the curtailed piurchasing power of the farmer; even in the cotton area, despite the high level of prices, there seems to be little increase in purchases. Cotton prices de- clined dvuing the first two weeks of November; the spot price of middling upland on the New York market reached 17 cents per poimd on the fourteenth, the lowest since August 31. Exports diuing October were 874,510 bales, 67 per cent larger than in September. Wheat prices have recovered somewhat since the low reached during the first week of the month. Wheat exports fell from 30.8 million bushels in September to 18.4 miUions in October. MONEY Evidences of easier money conditions continue to come to sight. The fact that large fimds are seeking investment is shown by the size of subscriptions to investment secxurities of the best class. The last three issues of United States treasury certificates of indebtedness amoimted to 1,100 million dollars, while the subscriptions were over three times as large. Subscriptions to the recent issue of 6 per cent refunding mortgage bonds of the New York Telephone Company were nearly ten times the amoimt offered — $50,000,000. Although the importance of such oversubscriptions is to be discounted because the bids were padded to gain advantage in the expected allotment, nevertheless they show a real demand of remarkable size. In the case of the telephone bonds, the large niimber of separate sub- scriptions supports this conclusion. The position of the federal reserve system has continued to grow stronger during the month. The reserve ratio of the combined reserve banks rose from 71 per cent on Novem- ber 2 to 72.3 per cent on November 23. Reserves have increased steadily, whUe federal reserve notes in actual circulation and total deposits have decreased. There has been a slight expansion, since August, of the aggregate deposits of reporting member banks. Interest rates on commercial paper have decreased during the month; the current rate is S to 5i per cent. CaU loan rates have ranged between 4 and 6 per cent; at present they are 4§ per cent. SUMMARY The growth of investment funds (despite the lack of business profits), decreased interest rates, and higher prices for investment securities are natiaral at this stage of the business cycle. We are just beginning to emerge from a condition of profound business depression. Prices have fallen abruptly, bank loans have been to a large extent Uquidated, and note circulation has been greatly curtailed. The position of the banks has been greatly strength- ened, not only by these last two factors, but also by the accmnulation of gold in the bank reserves of the United States, because of the large excess of gold imports over gold exports. Business, now at a low ebb as to production and with prices greatly decreased, does not absorb the fimds previously needed for ciurent operation. The surplus fimds of the banks and of business interests have thus expanded, and are now seeking investment. Speculation and investment in industrial stocks have been deterred by the small profits to be reaHzed from industry; hence these funds have found an outlet in government securities and bonds, rather than seeking investment in stocks or being used for specidation. With further accumiilation of funds, however, the narrowness of such outlets, the decreasing return to be gained, and the attractiveness of low stock prices, tend to bring more and more money into the stock market. The increase in stDck prices and activity on the New York Stock Ex- change diuing November is the first tangible evidence that easier money conditions are having their expected result. Two other factors, less immediate in their effects than those just discussed, are favorable to future business development. Recent action of the Interstate Commerce Commission and Railroad Labor Board points toward a not too remote readjustment of raihroad rates and wages. In the second place, the proposal for the limitation of naval armaments made by the United States gives promise of a curtailment of governmental expense. Such curtail- ment in the United States would hft a portion of the burden of governmental expenditure from American industry. For Europe as a whole the lightening of the burden of arma- ment would mean more rapid recovery, which would react favorably upon the economic situation here. Beyond these merely material results, the improvement in international relationships expected from the conference will go far to allay international distrust and restore confidence. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH CAMBRIDGE, MASSACHUSETTS, U.S.A. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH GENERAL BUSINESS CONDITIONS Advance Letter — December 75, ig2l THE INDEX CHART BUSINESS activity and wholesale commodity prices receded slightly in November, as is evidenced by the downward movement of curve B of our index chart, representing business. The decline was due to a decrease in both series upon which the curve is based: Bradstreet's price index fell sHghtly, from $11.35 on November i to I11.31 on December i, while bank clearings outside New York City were 4 per cent less in November than October, as against a usual decline of only 2 per cent. But although the gains of the early fall have +5 -3 - u ^ s ki 1 ^ a;. N ci-Mone 4> ^ ^ f y, "> V ^ ^ ^ ^ ■a ^ ^ =^ V\ ^ ^ r \ \ \i I r > \ k ''^n ¥ ^ n k s <• V/ k H !£a: ^lV| *«* ?i^ ^- t A Y J r ^ 5%. ?^ r l«Si! -ITU), < F pec -ulc itiO P\ 1 1 L ' \ 1& g^ \ 1 \ A B ■N.Y.CituBo/ikCie!\ring5 Shares Triidecl ontheN Y5t Excl Price of Industrial Stocks Biink Clearince Outside WYCity \ i -E jsi ne; « ^ S» ■^ C t% n4- n6( m )-9 %i yc 1P om w. r per S ^ i^ ^ ■* m 10 II 12 Foeecast: Since June curve C, representing money, has moved downward sharply, reflecting decreased interest rates. A decline of curve C usually precedes, and hence forecasts, an upward movement of curve A, representing speculation, and then of curve B, representing business. Curve A has^ risen substantially in November, after three months of slow and halting recovery from the low point reached in July. This upward movement of curve A shows, for the first time in a significant fashion, the expected result of easier money conditions, evidenced by the decline in curve C. The movement of these two curves forecasts a significant rise of curve B, representing business, which is now above the level reached in the depth of the depression last May. not been ftdly maintained, nevertheless the level of the last month was above that of the early summer. The increases which have taken place in prices since the low level are chiefly responsible for maintaining curve B at its present height. Conditions for an immediate upward movement of business are not yet present; but a continuance of the upward move- ment of curve A, representing speculation, which became pronounced during November, and the expected persistence of easier money conditions, represented by curve C, mean a significant improvement in business conditions in the spring of 1922. COPYSIGHI, I92I, BY HAEVAKD DMIVERSnV BUSINESS Now that statistics reflecting business conditions are available for November, it is pos- sible to judge the extent of the economic improvement which we have witnessed this fall. The general indices of business activity show that the increase in business activity during the fall was less than the seasonal amount. Bank clearings outside of New York City, the most reHable of such indices, increased 3 per cent in September, 8 per cent in October, and fell off 4 per cent in November, as against usual increases of 3 and 16 per cent in the first two months and a decline of 2 per cent in the last.^. The statistics for business failures show the same condition, since the increase of such failures in October and November has been somewhat greater than usual. The increase in car loadings from midsmnmer untU the end of October was about that seasonally expected, as has been the decrease since.^ On the other hand, buUding activity and the volume of manufacture for certain industries have shown a more than seasonal improvement. Conspicuous among the latter are the iron and steel and paper industries. The advance in the textile industries through October is about that to be expected at this time. No significant improvement has appeared in the lumber and leather industries, and the production of petroleum has fallen off slightly. Despite apparent progress towards price stabilization, shown by a decreasing number of price changes week by week, one of the most potent factors in delaying business advance has been the expectation of further price reductions. Adjustments in railroad rates and the anticipation of further adjustments have played an important part in the tmcertainty of prices. Perhaps the most disturbing of price fluctuations — unconnected, in this case, with raUroad rates — has been that in the price of cotton. The phenomenal advance in August and September was followed by a somewhat irregular recession, which brought the spot price of middling upland on the New York market to 16.7 cents per pound on November 12. A price of 18.9 cents was reached on the twenty-third, and this has not been exceeded since. One reason for lower prices is found in the governmental ginning figures, which show that 7,641,000 bales of this year's crop have already passed through the gins, whUe the October crop estimate was only 6,537,000 bales. The violent fluctuation in the price of this raw material has had the effect of slowing down the distribution of cotton goods because of the imcertainty it has created as to their future price. Most of the movements of the indices of commodity prices in the United States and European coimtries have been slight and irregular during recent months. The accompa- nying table presents price indices for the United States and important foreign countries since May 1921. The minimum figure of each series is in boldface type. Prior to May there occurred an almost continuous decline from the high points reached in 1920 for all the coxmtries except Germany. For the United States, the decline in prices was steady from the first half of 1920; the low poiut of the decline appears to have been reached in the three months Jime-August. There has been an advance since, but it has been irregular and not large. Until faU the movement of wholesale prices in most foreign countries was similar to that which occurred in the United States. Canada, indeed, had shown an uninterrupted decline ' If the 1921 figures are corrected for income tax payment the September increase practically disappears, while that of October becomes approximately ii instead of 8 per cent. " During the week of July g, a holiday week, car loadings reached a midsummer minimum of 639,698 cars, rising to 776,252 cars the next week. They increased fairly continuously to the week of October 22, when 962,698 cars were loaded, dropped to 952,621 the following week, and numbered 786,671 on the week of November 19. Wholesale Price Index Numbers of Selected Countries (Average for iqij = loo) United States Canada United Kingdom France Italy Ger- many/ Nether- lands Swe- den Nor- way Q Den- marlc A U.S. B.L.S. Brad- street's Dun's lo-com- modity price index Econo- mist Statist Japan IQ2I May June July August . . . September October . . November December a 148 148 152 152 ISO b 117 117 120 120 122 123 123 b 141 140 13S 138 137 136 138 139 b III 113 109 107 III 119 118 112 c 183 179 176 172 169 d 182 179 178 179 183 170 d 191 183 186 183 176 163 d 330 331 332 345 333 d 547 520 542 580 599 d 1428 1387 1467 1723 1777 1993 2687 3283 i 182 182 176 180 180 169 c 218 2X8 211 198 182 175 d 294 294 300 297 287 286 276 b 257 254 253 254 224 202 186 d 191 192 196 199 207 Month in which TngYimiim was reached * May 1920 Feb. 1920 May 1920 June 1920 May 1920 Mar. 1920 Apr. 1920 Apr. 1920 Nov. 1920 Dec. 1921 July 1920 June 1920 Sept. 1920 Nov. 1920 Mar. 1920 a. Average for month. h. First of month. c. Middle of month. d. End of month. «. The decline was continuous from the month in which the highest number was recorded except for Germany, and minor fluctuations in the indices of the Economist (United Kingdom) and of France and Sweden. /. July i9i4 = roo. g. December sr, 1913-June 30, 1914 = 100. h. July i, igia-Jime 30, 1914 = 100. i. Not stated whether average for month, or representing prices for one day. of commodity prices, also had Sweden; but elsewhere there were evidences of approximate price stabilization, with some tendency toward a recovery from the low points reached last spring. In September, however, the indices for Norway and Denmark showed a substantial decline, and in October recessions occurred in the United Kingdom, France, and the Nether- lands. Italy and Japan were not involved in this downward movement, the extent and significance of which it is not now possible to appraise. Obviously the renewal of liquidation in commodity markets reflects and tends to accentuate the slow recovery of Europe from the depression which began last year. Wool consumption continues to gain. In retailing ready-to-wear garments, special sales have appeared earlier than usual, and have in many instances involved substantial price reductions. The strike of the garment workers in New York continues; a similar strike began in Philadelphia on November 28, and in Chicago during the same week. Coin- cident with the willingness of consmners to pay the prices asked for woolen goods there has occurred a steady advance in the price of raw wool. The market has been active, and has even been characterized as "speculative." The price of fine wools has been afifected by the emergency tariff, and a growing scarcity of these wools is reported. At a recent auction of government wool, the entire amount was disposed of at price advances of from 10 to 15 per cent. The London market has shown a price decline of 5 to 15 per cent, due, it is said, to lack of American buying. This has acted to place the American manufacturer of wool at a disadvantage compared with the British. In the iron and steel industry the gain made in production during September and Octo- ber has not only been held, but an additional advance recorded; our adjusted index of physical production, from which increases due to seasonal influences have been removed, stood at 37.0, 44.2, and 53.6,^ for each of the last three months. The number of furnaces in blast rose from 96 on November i to 120 on December i. A merger of seven of the inde- pendent steel companies is being contemplated. Steel prices, according to the index of the Iron Age, are slightly higher than in the third week of November; pig iron prices have fallen oflf somewhat since the middle of that month. There has been a fairly continuous increase in the price of wheat from the low of Novem- ber 3. This has had less effect upon the prosperity of the country districts because this year's crop was largely marketed. Liquidation in the agricultural districts apparently stopped before substantial gains had been made, and at least one more crop year will be needed to lift appreciably the load of indebtedness from farming communities. The War Finance Corporation has been active in meeting the demand of farmers and stock raisers for longer time credit than that ordinarily furnished by the banks. Loans made by this body from January to November exceed $119,000,000. These have been used partly to assist export trade. Loans through the Stock Growers' Finance Corporation to stock raisers are reported to have prevented the sale of immature cattle and the depletion of herds. The Federal Reserve Bank of Chicago states that dairy farming has felt the depres- sion less severely than other branches of agriculture, and that recovery in this line has been cosrespondingly more rapid. SPECULATION Except for industrial stocks, the advance in security prices came to a temporary halt in the first two weeks of this month. The Dow- Jones daily average prices of 20 industrials, though somewhat irregular from day to day, advanced to $80.69 on December 13, the high- est point reached in more than a year. Railroad stock prices, on the contrary, though still well above the October low of $70.00, moved downward from $76.66 on November 29 to $74.21 on December 13. Railroad bonds, like railroad stocks, declined in price, while industrial bonds were firm and public utilities advanced. Thus, the average of 40 issues was maintained between $84.06 and $84.13 during the first week of the month, but has fallen off since. Although security prices as a whole have done no more than hold their groxmd during these two weeks, their level is high in comparison with recent months. The volume of shares traded has decreased since the first week of December, and New York bank clearings have tended so far this month to decline rather than increase. MONEY The easier money conditions evident in November have continued into December. Interest rates on commercial paper have remained imchanged at 5 to 5I per cent. Call loan rates have fluctuated between 4I and 6 per cent; they are now 5 per cent. The reserve ratio of the federal reserve system has again risen, despite an increase of federal reserve notes in actual circxilation during the first week in December. A slight expansion in banking activity, however, is shown by the fact that loans and investments and aggregate deposits of reporting member banks reached a minimum in October and August, respec- tively. ' Provisional. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH CAMBRIDGE, MASSACHUSETTS, U.S.A. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH THE VOLUME OF MANUFACTURE special Letter — September 20, ig2I FLUCTUATIONS in the voliune of production constitute one of the most significant elements of the business cycle. Outwardly the cycle is a matter of markets and demand, of prices and profits. But behind these forces lie such factors as raw material supplies, producers' stocks, employment conditions, manufacturing outpiit. The variations of these INDEX OF THE VOLUME OF MANUFACTURE (iprp MontMy Average = loo) Per Cent les ISO 115 110 105 100 95 90 as ao 75 1 Jan. FebMor Apr? Moy JuneJuly Aug. SepT Oct Nov. Dec Jan. reD. Klai: Apt? May June July Aug. SeptOcT. Nov. Dec. Jan. leti. Mar Apr May June ISIS I9SO 1981 imderl3dng factors reflect critical processes of industrial adjustment. In the development of crisis conditions as well as in the restoration of normal trade, changes in the volume of production are fimdamental. Accurate measurement of these variations is essential to a fuU understanding of indus- trial conditions. The precise character of the fluctuations, their period and amplitude, and the striking contrasts among the several more important industrial groups, must be ascertained. An index of the physical volume of manufacture, for groups of industries and for the groups combined, will throw needed light upon the course of the business cycle. Such an index, for the several manirfacturing groups combined, appears above. ^ CO ^ « I 1o A CJ) o u k u ^ A n\ (U o IV ^ 1 ill \a A w. A k ) Y V N k \ 1 V ^ V J T \ A \. 1 J f \ i f 0^ • — 1 i V J \ \ \ A \ -\ ^ A V "^ S COFXBIGBI, I92I, BV HABVABO mnVEBSITX The index registers definitively the cycle through which manufacturing activity has passed in the United States since the beginning of 1919. The opening months of that year witnessed the industrial sag started by the armistice. Spring brought a recovery, and summer an acceleration, of output. The steel strike, beginning in late September and last- Table a. — Monthly Indices oe the Volume or Manufacture for Eight Groups of Manufacturing Industries and for the Groups Combined (iQig Monthly Average = 100) Month All groups Iron and steel Lumber Paper Petroleum TertUes Leather Food Tobacco 1919 January. . . February . March . . . . April May June July August . . . September October. . , November December . 1920 January . . February . March . . . . April May June July August . . . September October. . , November December . 192 1 January . . February , March . . . , April May .... June July (I) 101.6 86.7 92.6 93-7 9S-7 95-9 101.9 107.2 103.8 104.4 102.2 102. 1 II5-9 104.6 118.0 108.8 111.8 109.6 102.3 104.9 101.4 101.2 88.9 77-9 78.3 75-0 80.6 75-8 79-3 7S-9 (2) 130.1 II3-7 II3-5 94.6 81.2 90.9 103. 1 "3-S 97.6 73-2 94.0 103.4 124.2 120.5 138.3 III.O 121. 2 124.9 118.0 I2S-3 125.0 126.7 111.3 99-S 92.7 73-7 65.2 S0.5 52.4 42.3 34-0 (3) 7S-S 75-8 88.8 97.8 112. 1 I03-S 103-7 123.4 113.8 123.4 102.0 80.1 lOI.I 101.3 117.4 122.3 127.3 irS-4 103. 1 117.8 108.7 108.1 86.2 62.8 S9-S 65-3 78.2 84.6 100.7 93-3 (4) 93 -o 82.8 87.0 88.4 9S-I 98.3 103-3 108.8 106.2 1 1 6.4 108.6 109.8 120.9 105.8 119-S 121. 7 119. 2 122.5 123.0 122.5 119-3 119.7 108.0 94-6 84-3 78.5 83.2 79-3 73-6 7S-6 71.0 (5) 89-7 83.8 92.8 92.4 100.7 96.1 103.7 107.4 108.3 II2.0 107.0 107.6 102.3 97.0 III. 6 109.3 115.0 iiS-9 122.9 13^-2 134.6 I3S-2 131. 2 134.6 131.6 115.0 117.9 124.9 122.8 122.7 (6) 96-S 72.7 79.0 92.9 99.2 100.2 109.4 103.4 105-3 "9-3 104.9 IIO.I 126.6 110.5 121.0 119-3 110.5 104.1 92.8 88.2 83-4 78-7 62.1 S4-5 68.1 76-3 89-3 89-3 96.2 100.8 89-5 (7) 90.8 81.6 94-0 84.7 90.9 95-2 80.7 70.5 73-3 77-8 70.1 72.1 63-S 62.8 72.0 7S-8 83.2 81. 1 (8) 1 14.9 87-3 92.1 92.8 98.5 83.8 85-9 88.4 97-7 III. 2 109.0 108.6 112. 2 85-3 90.8 7S-9 85.1 80.8 80.8 83-3 78.0 81.1 91.0 88.4 84.1 74-4 85.6 82.4 79-9 82.7 82.3 fo) 87-3 83.8 97-7 82.0 88.0 92.9 99-3 lOI.I 108.6 125.4 117.0 113.0 "S-7 98.9 123. 1 108.0 III. I 114.4 98.1 104.8 104.9 106.3 93-1 73-S 88.7 9S-0 106.6 96.6 100.6 108.0 ing into December, and the soft coal strike, from the first of November till the second week in December, seriously reduced the volume of manufacture. The railroad tie-up of mid- winter also impeded production. The peak of output came in March, 1920. There ensued a period of gradual decline, followed by a sharp slump in the fall. December, 1920, found manufacture generally at a low level. It remained there during the first six months of the current year, unmistakable gains in some lines being offset by heavy losses in others.^ The index discloses clearly the extent, as well as the tuning, of the fluctuations of manu- facturing output since January, 19 19. The slump following the armistice reduced produc- tion nearly 15 per cent below the 1919 average. The succeeding boom carried it 18 per cent above this average. The depression of recent months has been a more extreme movement. From January to June, production ran consistently between 20 and 25 per cent below the 1919 average, and in July probably fell over 30 per cent below. These conclusions hold for the more important industrial groups combined, but the individual groups show striking deviations. An examination of these differences is highly instructive. Fortunately the method by which the general index is developed facilitates such an analysis. The index for all manufacture is a composite of indices registering the volume of production in eight groups of manufacturing industries. In three instances the group index rests upon a single series showing production of an important commodity month by month. In other cases, the group indices are based upon as many as four or five series. The several groups, designated by their basic products, together with the supporting series and the sources from which these have been drawn, appear in the following table: Table B. — Monthly Series Indicative of the Volume oe Manufacture Group Series Source Iron and steel. . . : Pig iron produced Steel ingots produced Lumber cut (5 varieties) Newsprint produced Book paper produced Fine paper produced Wrapping paper produced Paper board produced Sole leather produced Crude oil run Cotton consumed Wool consumed Wheat flour produced Cattle slaughtered Hogs slaughtered Cane sugar melted Cigars produced Cigarettes produced Manuf. tobacco produced Iron Age * Lumber 11 (t :]« Bulletin, Fed. Reserve Board * Paper Monthly Report, Fed. Trade Comm.* Leather (( « u U U * U it U it u « u u u u w u u u u u Monthly Survey, Dept. of Comm.* Petroleum Monthly Report, Bureau of Mines * Textiles Bulletin, Dept. of Comm.* Food Market Reporter, Dept. of Agric* Bulletin, Fed. Reserve Board * Tobacco Market Reporter, Dept. of Agric* A/ w r DO IS 19 IS so 192 1 PETROLEUM (^ r r V- ^^ Per Cent eoD TEXTILES ISO 100 50 7 .JvV^ FOOD \, 1 r \ u J V -vT ^^- [919 1980 I9SI TOBACCO igiQ ISSO I9SI / \^ ^f\ Kf 4v ^ r 1919 igSO 1931 1919 1980 1921 S weakness. Only iron and steel and paper, the last two to decline, failed to turn the corner toward recovery during the first six months of the year. Of the many noteworthy contrasts suggested by these differences, perhaps none is so significant as that between the two fundamental groups, iron and steel upon the one hand, and textiles upon the other. Since January, 1919, the movements of textile production have consistently anticipated those of iron and steel. Textile output started upward in March, 1919; iron and steel, in June. The slump in textiles began in May, 1920; that in iron and steel in November. In 1921, textiles showed increased production as early as January; only during the most recent weeks has the iron and steel trade shown improve- ment. This apparent relationship is of great interest. It invites most careful study in later fluctuations of manufacturing output. Distinct as are the differences of variation in output among the several industrial groups, a general movement in manufacture as a whole is discernible. Under the stimulus of rising prices, manufacturers resort to every available means for increasing output. Pro- duction rises to a maximum — as in the second half of 1919. But the effort to enlarge out- put encounters increasing difBculties, and production falls continually short of demand. "Underproduction" is the cry of the hour. Then the crisis breaks. Orders are cancelled, stocks accumulate, even customary output cannot be marketed. Production falls to a minimimi — as in the first half of 192 1. Against industrial management is laid the charge of previous "qverproduction." Only when stocks have been exhausted and prices again stabilized, do the accmniilations of demand restore confidence and lead once more to in- creased manufacturing activity. The cycle of production is undeniable; it has aheady shown its many phases in the United States since January, 1919. Its influence is fimda- mental. There is urgent need for more accurate and complete knowledge of its character and course. TTie Index of the Volume of Manufacture is a means to this better industrial diagnosis. Note on Methods Employed The methods employed in constructing the monthly Index of the Volmne of Manu- facture are essentially those developed in the derivation of the armual unadjusted index of the physical volmne of production.^ That index was set up as a weighted geometric mean of relatives obtained by referring the items of each series to a fixed base. The base' adopted for the monthly index is the monthly average of 1919. The first step in the construction of the index is the reduction of the monthly items of each series to percentages of the series' 1919 average. These percentage relatives have been computed, for the months from January, 191 9, to June, 1921, for all nineteen series upon which the general index is based.^ The data will be given in a later pubUcation. The weighting of the different series of relatives is generally according to the values added by manufacture in the several industries and groups in the latest normal census year before 1 See Review of Economic Statistics, Preliminary Vol. n, pp. 253-255, 309-325, 332-336. " While the 1919 monthly items for leather are not available for publication, it has been possible to include leather in the general index from the start. 6 the war — 1909. Minor alterations of these weights have been made in the case of the tobacco group and the petroleum trade. The latter is manifestly of greater relative impor- tance today than ten years ago. And within the tobacco group the production of cigarettes has shown an increase out of all proportion to the other tobacco products. These facts have been recognized in the adoption of the weights given in Table C. Table C. — Relative Weights Employed in Calculating Indices of Volume of Manufacture Senes Pig iron produced Steel ingots produced Lumber cut , Newsprint produced , Bookpaper produced Fine paper produced Wrapping paper produced Paperboard produced Sole leather produced Crude oil run Cotton consumed Wool consumed Wheat floiu- produced Cattle slaughtered Hogs slaughtered Cane sugar melted Cigars produced Cigarettes produced Manufactured tobacco produced For single industnes 3-45 3-4S For groups of industnes 20.6 3-4 3-4 1-7 3-4 1-7 For eight groups combined 25-7 16.2 13.6 S-4 4.1 24-3 13-5 4.1 A general revision of the above weights is contemplated as soon as the 191 9 census figures are available. There is no reason to beHeve, however, that this revision wiU appreciably affect the index. The index presented above is of the "unadjusted" type — no allowances have been made for seasonal influences and long-time trends. In the index for the several groups com- bined, seasonal disturbances peculiar to the different lines of industry undoubtedly in some measiu'e counterbalance one another. In the individual group indices, upon the other hand, the movements may be in considerable measure the result of such seasonal variations, combined with elements of normal growth. An, analysis of these seasonal factors and trends is highly desirable. It wiU be presented in a forthcoming letter and our monthly Review. At that time more complete details also will be given regarding the materials and methods utilized in constructing the Index of the Volume of Manufacture presented in this letter. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH INDUSTRIAL EMPLOYMENT Special Letter — October 8, ig2l EMPLOYMENT conditions are one of the fundamental elements of the business cycle. Measurement of fluctuation in emplo3Tnent throws useful light on the course of the labor market; it also aids the understanding of industrial conditions. An index of employ- ment is presented below, together with the index of the physical volume of manufacture which appeared in our Special Letter of September 20. According to these indices employ- ment and production in manufacturing industries for July, the most recent month for which comparable figures are available (not shown on the chart), were 20 or 30 per cent below the average of 1919. Chart i. — Indices of Industrial Employment and Production (ipip Monthly Average = 100 per cent) Apr i9ei The index of employment, in addition to showing changes in employment as such, serves two other purposes. First, it indicates approximately the buying power of the wage-earning population from current earnings. An increase or decrease in employment brings with it a corresponding increase or decrease in the aggregate earnings, even if wage rates remain the same. Moreover, the rate of wages itself responds to fluctuations in em- ployment, since wage rates tend to be high when employment is active, and low when employment is dull. Thus full employment at high rates increases the purchasing power of labor; tmemployment decreases that power. Second, the employment index gives us an indirect measure of production, as is shown by Chart I.^ Throughout the thirty months covered, the curves representing employment and the volume of manufacture show change ' The coefficient of correlation is 96 per cent. COPYRIGHT, 192 1, BY HARVARD TJOTVEESITY of direction at nearly the same time. Every critical movement is shown with ahnost equal clearness in both curves: the relapse after the armistice, the recovery in the second, third and fourth quarters of 1919, the culmination of the boom early in 1920, the ebb of pros- perity, at first slow, then precipitate, until the early months of 192 1, and the xmcertain course during the past six or eight months. Two significant differences between the curves appear: (i) The employment curve was much less disturbed than the production curve by the steel and coal strikes of Septem- ber to December 1919, and the railroad congestion in February 1920. Partial explanation for this may be found in the fact that the supply of basic materials, which are the chief constituent of the production index, was perhaps affected much more by these disturbances than was the manufacture of semifinished and finished goods, which maintained itself during the interval, presimiably on a carryover of raw materials. Again, employers doubt- less tend to hold their labor force over a temporary shortage of raw materials incident to labor disputes in key industries; this is of course particularly likely at a time of prosperity. (2) It is clear that the production curves, if the temporary effect of these labor disputes is smoothed out, rose to a greater height in the boom of 1919-20, and recently fell to a greater depth, than did the employment curve. Employment rose only about 8 per cent above its 1 919 average, production increased twice as much; in the present depression, emplo)maent has fallen only about 20 per cent, while production has decreased 25 per cent, below the average for 1919. To what is the difference in the amphtude of the two indices due? It is impossible to get conclusive evidence on this point, but three considerations bear upon it. First of aU, employment data are based on the number of employees on payrolls, and therefore tend to understate the real extent of labor activity in prosperity, since overtime work does not increase the number reported, and to overstate the extent of activity in depression, since part-time work does not decrease the number. In other words, no distinction is made, in the employment figures, between workers who are engaged part time, fuU time and more than full time. To a certain extent this source of inaccuracy is counterbalanced by a second consideration — fluctuations in the efl&ciency of labor. During prosperity labor efficiency is likely to diminish, partly because of industrial fatigue due to speeding and working over- time, partly because of indifference to duty in the face of plentiful opportunities for work in other estabhshments, and partly because of the hiring of mitrained or inferior workers; in a depression the reverse is the case. A third consideration is the fact that the produc- tion index is necessarily derived, in large measure, from data on the production of basic materials, such as sole leather and steel ingots, rather than finished goods, such as shoes and machinery. The reason is that, in more advanced stages of manufacture, units of output are so varied and complex that few reliable data on production at these stages can be obtained. The emplojTnent data, on the other hand, relate to industrial groups at all stages of fabrication. If the production of basic materials responds more decisively to the course of the business cycle than do finished goods, the production curve would naturally swing through a greater range of variation than the employment curve. Changes in labor efficiency would thus lead to greater fluctuations in the employment curve than in the production curve; while the existence of overtime or part-time work and the alleged sensitiveness of the production of basic materials would lead to narrower fluc- tuations in the employment curve, such as Chart i actually shows. THE DATA UTILIZED In Chart 2 are shown the two basic indices upon which the employment index of .Chart i is based. One set of data is collected by the United States Bureau of Labor Statistics from establishments employing 600,000 to 800,000 men, well distributed among the leading industrial states, about three-fourths of the total being in Pennsylvania, Massachusetts, Michigan, New York, Ohio, and Illinois. The other series, collected by the New York Chart 2. — Compaeison of Leading Employment Indices {1919 Monthly Average = 100 per cent) Julu 1919 July ideo State Industrial Commission, relates solely to establishments in that state, and covers 400,000 to 600,000 workers. Neither bureau purports to cover the entire field of industry, but each secures reports from about a dozen large and fairly representative industrial groups. According to an investigation recently completed for the Committee on Economic Research, emplo3anent conditions in New York reflect with great accuracy those in the entire indus- trial section of the United States.. From the New York imemployment data collected prior to 1 916 an index has been constructed which is found to serve, not only as an employment index, but also as an index of the voltmie of business from 1903 to 19 16. Throughout that period the employment figures registered general business conditions as well as did pig- iron production, outside clearings, and other statistical series conmionly used."^ In view of these findings, which will be presented in a forthcoming issue of the Review, there seems to be ground for giving as much weight to the New York series as to that of the Bureau of Labor Statistics; accordingly the two were averaged, forming the middle curve on Chart 2. The seasonal element was found not important in either series. The New York figures are based upon establishments nearly but not wholly different from those covered by the United States Bureau of Labor Statistics. Eighty-eight per cent of the wage earners embodied in the latter series are employed in other states than New York, and possibly some of the reporting 12 per cent employed within the state are not employed in establishments reporting to the New York Commission. Though practi- ' Moreover, the old unemployment series supports the newer employment series, showing nearly the same course of employment conditions in the years 1914-16, when both were reported. 3 cally independent as to source, the two indices follow nearly the same course during most of the period under review.^ There are certain important discrepancies; for example, they disagree as to the precise month in which turning points occurred, upward in early 1919 and downward in early 1920. Equally striking is the fact that during 1921 the New York curve has, in general, continued downward, while the Biureau of Labor curve has moved fairly steadily upward. The contradiction is more apparent than real; comparing the two sets of figures we find that it is due in part to the heavier weight (about 35 per cent) given in the New York figures to metals and metal products, an industry which slimiped very late and has shown Uttle tendency to recover, and in part to the lighter weight (about 12 per cent) given in those figures to the textile industry in which some important branches have recovered very rapidly since the early part of the year. INDUSTRIAL GROUPS In Chart 3 are shown employment indices for certain industrial groups, selected partly from the New York Industrial Commission, and partly from the United States Bureau of Labor Statistics. Examination of Chart 3 will disclose certain striking general facts, such as the depressed condition of late 1920 in the wool industry and that of early 1921 in the automobile industry, the recent tendency to recovery in many Unes and the persistent failure of iron and steel, machinery and car-building to show any marked sign of improve- ment up to July. Chart 3 also enables us to determine (i) the order in which the depression overtook the various industries, (2) the relative severity with which it appears to have affected them, and (3) the order in which recovery appeared in each important industry. A similar review, based on the production data, has appeared in the Special Letter of Septem- ber 20. Judgments based on the employment figures, where both employment and produc- tion data for the same industries are available, are nearly identical with those based on production figures. Apparently the first to feel the depression of 1920 were the food and leather industries, in which decline began in the months from December 1919 to February 1920. Then followed automobiles (April), boots and shoes (April), clothing, etc. (April), woolens (June). After a considerable lapse came cotton manufactures (August), machinery, including electrical machinery (August), wood manufactures (October), iron and steel (November), paper (November), printing and book-making (December), and chemicals (December). From the employment data, the depression appears to have been most severe in the automobile and woolen industries, considerably less so in car-building and iron and steel. Machinery and leather rank next, then come boots and shoes, paper, wood manufactures, clothing; and least affected of aU were cotton manufactures, printing, and food, beverages and tobacco. The recovery from depression seems first to have set in about February i92i,when it ap- peared in many lines — woolen, cotton, clothing, automobiles, wood, food, beverages and tobacco, leather, and boots and shoes. Paper and printing did not show recovery until June or later. For iron and steel the data for August show a turn for the better, as do also those for car-buUding. Most other industries continue to improve, and there seems ground to believe that better times are ahead for industry generally. 1 The coefficient of correlation between the two employment indices is 95 per cent. 4 Chart 3. — Indices of thk Volume of Employment in Leading Groups op Manufacturing Industries (iprp Monthly Average = 100 per cent) IRON, STEEL & MACHINERY AUT0 M0 BILE5 & RAB.WAY CARS CHEMICALS,OILS,FAINTS,ETC. 150 100 50 ^ '■^ <'^^. J5 ■ Iron and Steel "Mjichinery 1919 1920 1921 1919 1920 1921 1919 1920 1921 150 TEXTILES CLOTHtNG,tTC. .^ ^ ^ \ r LEATHER,BOOTS & SHOES 1919 1920 1921 1919 1920 1921 ^ ^^ :>' ' ^ \cC Bather )otsand Shoes .... Be 1919 1920 1921 PAPER & PRINTING FOOD, BEVERAOES & TOBACCO r^ ■^v /V" \^ \ ^ WOOD MANUFACTURES i_J^ ^^ ^ /^ V > ^ 1919 1920 1921 1919 1920 1921 1919 1920 1921 The emplojmient index seems clearly to act as a satisfactory register of the labor market and of production both for all industries and for each leading industry separately. The emplojTiient and production indices render nearly the same verdict as to the timing of the ebb and flow of the industrial cycle. The production of basic materials, embodied in the production index, we also conclude, reflects faithfuUy changes in production aU along the hne, for the production curve agrees with the employment curve, which covers indus- tries at all stages. In ampUtude, or sensitiveness of fluctuation, however, the two curves show noticeable but accountable differences. On the whole there seems good reason for making the indices of employment and production joint members of a composite index of industrial activity. PRESENT UNEMPLOYMENT It is possible to obtain, as a by-product of the employment index, a rough test of the current estimates of unemployment. The estimate of 5,700,000 unemployed in manufactur- ing, mining, transportation, etc., recently announced by the Department of Labor, was based largely on the decrease in numbers employed between early 1920 and the present. Since this figure refers to the shrinkage in the number on the pay roUs, it does not yield an accurate estimate of the volimie of unemployment in the sense of involuntary idleness, because in the boom of 1920 thousands were employed whose activities have since been absorbed by the farm, the home, and the school. Undoubtedly the past year and a half has witnessed an extraordinary shifting of industrial labor; nevertheless the number leav- ing positions in industry is not the mmiber unemployed, but rather a maximum or upper limit of the true figure. The Department of Labor estimate, therefore, is to be regarded as a maximum; that is, the number unemployed is probably less than this figure. The amount of "unemployment" or shifting of labor from manufacturing industries is considerably less, when measured by the employment indices here presented, than accord- ing to the data of the United States Emplo)Tnent Service, upon which the estimate of the Bureau of Labor Statistics was primarily based. According to the latter, the drop in em- ployment from January 1920 to September 1921 was about 42 per cent. According to the New York state index it was 28 per cent, and according to the long-established employment series of the Bureau of Labor Statistics itself, the drop was only 23 per cent. The valuable trade-union data collected by the Massachusetts Department of Labor and Industries shows a decHne of 26 per cent. In short, these three sets of figures agree with each other and aU fail to confirm the figures for the decrease of employment given by the United States Employment Service. This estimate included 3,900,000 unemployed for manufacturing and mechanical industries alone. According to our estimate based on preliminary data of the Census Bureau, there were in January 1920 not more than 12,000,000 persons employed in this group. From the three sets of figures referred to above there would appear to be about 3,000,000 fewer persons employed in the manufacturing and mechanical industries at the present time than twenty months ago. On the basis of these industries, then, the estimate of 5,700,000 seems unnecessarily high even as an upper limit of the number of unemployed. Even without endeavoring to test or revise the figures for transportation, mining, and other industries, the pubUshed figure can probably be reduced with safety by at least a million and possibly to a figure as low as 4,000,000. SOURCES AND METHODS Payroll data of the United States Bureau of Labor Statistics. The data of the Bureau of Labor Statistics have appeared monthly in the Labor Review since October 1915; those utilized consist of month-to-month link relatives, based on identical establishments in the comparison of any two months. These links were combined to get chain relatives with January 1916 as 100 per cent. The base was then shifted to the year 1919 as 100 per cent, and new relatives computed for each of the fourteen industries as follows: (i) iron and steel,^ (2) car-building and repairing,^ (3) automobiles,^ (4) cotton manufactures,^ (5) cot- ton finishing, (6) woolens,^ (7) silk, (8) men's clothing, (9) hosiery, (10) leather,^ (11) boots and shoes,^ (12) paper ,^ (13) cigars, and (14) coal. The series used by the Bureau in rep- resenting the combination of these groups is constructed by combining the chain relatives in a weighted arithmetic mean, using as weights the total numbers employed in those industries at the 1914 Census of Manufactures. In March 1920 the series covered over 800,000 employees, and in July 1921 nearly 600,000. The percentage distribution by states is as follows: March 1920 July 1921 March 1920 July 1921 Pennsylvania 19.7 16.4 Missouri 2.2 2.7 Massachusetts 12.8 14.9 Indiana 2.8 2.4 Michigan 13.0 13.3 Maine 1.7 2.1 New York 12.6 12.2 Rhode Island 1.5 1.8 Ohio 9.0 7.1 Alabama 1.7 1.8 Illinois 6.8 6.9 Connecticut 1.4 1.5 New Hampshire 2.5 3.8 Colorado 1.3 i.i New Jersey 2.7 3.1 All others 8.3 8.9 Payroll data of New York Industrial Commission. Since June 1914, the New York In- dustrial Commission has pubUshed several forms of employment indices. The index used here is the index of employment based on the figures for June 1914 as 100 per cent; for a given month, only those estabhshments are used which also reported in June 1914. Eleven major industrial groups are used, but indices for the component industries can be, and in certain cases were, segregated in this study. The groups are the following: (i) metals, machinery and conveyances, (2) textiles, (3) clothing, nullinery and laundering,^ (4) furs, leather and rubber goods, (5) food, beverages, and tobacco,^ (6) printing and paper goods, (7) paper, (8) chemicals, oils, paints, etc.,^ (9) wood manufactures,^ (10) stone, clay and glass products, and (11) water, fight and power. The series covered over 600,000 employees in March 1920, and over 450,000 employees in June 192 1, or about one-third of the factory workers in the entire state. 1 These groups were selected for separate presentation in Chart 3. 2 These groups were selected for separate presentation in Chart 3. In addition, machinery was talcen from Group i, printing and bookmaking from Group 6. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH Charles J. Bullock, Chairman Warren M. Persons, Editor Cambridge, Massachusetts, U. S. A. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH THE VOLUME OF MANUFACTURE INDEX FOR ALL GROUPS COMBINED ADJUSTED INDEX FOR THE TEXTILE GROUP Special Letter — October 22, ig2l ALL GROUPS INDUSTRIAL output fell in July to the lowest level witnessed since the signing of the armistice. The Index of the Voliune of Manufacture declined to 68.5, more than 6 points under the lowest previous figure. There is every indication that industrial output really touched bottom in July. While the data upon which the Index is based are not yet complete for August, the four group indices which are at hand — those for iron and steel, paper, textiles, and food — point to an index number for August of about 80.0. This Chakt I. — Index of the Volume as Manxtfacxxire P" {igiQ Monthly Average = 100) represents a substantial improvement; and recent reports for September and October in- dicate that the improvement is not ephemeral. An upward movement of industrial output seems definitely under way. The course of the several group indices diu-ing recent months is highly instructive. The most significant movement is that of iron and steel. In October 1920, the index for this industry stood at 126.7; in July 1921, at 34.0. In only one of the intervening months was COPYEEGST, I92I,EY EARVABP UNIVEBSIXY there the slightest pause in the precipitate decline. This rapid fall of output in the iron and steel trade has been the most striking feature of the industrial situation for nearly a year. It has completely offset the substantial gains made in many other Unes. The fact that the movement seems finally to have run its course, suggests the probability of improved business conditions ahead. The textile group in July showed a declnie of about ii points. Undoubtedly this was in part a seasonal movement. Even allowing for seasonal variation, however, a slight recession seems to have occurred. But in the textile group, as in iron and steel, August wit- nessed a recovery of lost ground. The index for the textile group for August is 100.7, equiv- Table a. — Monthly Indices of the Volume or Manufacture for Eight Groups of Manufacturing Industries and for the Groups Combined (igiQ Monthly Average = 100) Month AU groups Iron and steel Lumber Paper Petroleum Teidles Leather Food Tobacco 1919 January ... February . . March .... April May June July August . . . September October . . . November December . 1920 January . . February . March .... April May June July August . . . September October. . . November December . 192 1 January . . February . March .... April May June July August . . . W 101.6 86.7 92.6 93-7 9S-7 95-9 101.9 107.2 103.8 104.4 102.2 102. 1 "5-9 104.6 118.0 108.8 111.8 109.6 102.3 104.9 101.4 101.2 88.9 77-9 78.3 7S-0 80.6 75-8 79-3 7S-9 68.5 W 130.1 II3-7 II3-S 94.6 81.2 90.9 103. 1 II3-S 97.6 73-2 94.0 103.4 124.2 120.5 138.3 III.O 121. 2 124.9 118.0 125-3 125.0 126.7 111.3 99-5 92.7 73-7 65.2 50.5 52.4 42.3 34-0 46.0 (3) 7S-S 75-8 88.8 97.8 112. 1 I03-S 103.7 123.4 113.8 123.4 102.0 80.1 lOI.I 101.3 117.4 122.3 127.3 115-4 103. 1 117.8 108.7 108.1 86.2 62.8 59-5 65-3 78.2 84.6 100.7 93-3 86.8 (4) 93-0 82.8 87.0 88.4 9S-I 98.3 103-3 108.8 106.2 116.4 108.6 109.8 120.9 105.8 "9-5 121. 7 119. 2 122.5 123.0 122.5 119-3 119.7 108.0 94.6 84.3 78.5 83.2 79-3 73-6 7S-6 71.0 84.4 (5). 89.7 83.8 92.8 92.4 100.7 96.1 103.7 107.4 108.3 112.0 107.0 107.6 102.3 97.0 III. 6 109.3 115.0 "5-9 122.9 132.2 134.6 135-2 131.2 134.6 131.6 115.0 117.9 124.9 122.8 122.7 "9-5 (6) 96.5 72.7 79.0 92.9 99-2 100.2 109.4 103.4 105.3 "9-3 104.9 no. I 126.6 110.5 121.0 119-3 110.5 104. 1 92.8 88.2 83-4 78.7 62.1 S4-5 68.1 76.3 89-3 89-3 96.2 100.8 89-S 100.7 (7) 109.4 91.0 92.8 108.7 106.3 108.6 99.8 102.6 102.0 101.4 87.4 89.9 90.8 81.6 94-0 84.7 90.9 9S-2 80.7 70.5 73-3 77.8 70.1 72.1 63-5 62.8 72.0 75-8 83.2 81. 1 76-3 (8) 114.9 87-3 92.1 92.8 98.5 83-8 85-9 88.4 97-7 III. 2 109.0 108.6 112. 2 85-3 90.8 7S-9 85.1 80.8 80.8 83-3 78.0 81. 1 91.0 88.4 (9) 87-3 83.8 97-7 82.0 88.0 92.9 99-3 lOI.I 108.6 125-4 117.0 113-0 "5-7 98.9 123. 1 108.0 III. I 1 14.4 98.1 104.8 104.9 106.3 93-1 73-S 84.1 88.7 74-4 95-0 85.6 106.6 82.4 96.6 79-9 100.6 82.7 108.0 84-S 101.8 96.3 .... alent to the index for June and substantially above the index for any other month since June 1920. As a whole the textile group is now operating on the average level of 1919. Three of the other group indices — those for lumber, petroleum, leather — show slight declines from May to July. The movement is not serious in any case, possibly excepting limiber, and presumably is due largely to seasonal influence. An increase of output in these lines seems reasonably assured during the more active months of the fall. The paper industry has behaved throughout the cycle much like the iron and steel trade. It was late in experiencing the shock of business reaction in 1920, showing no marked decline until November. While the subsequent slump was not as severe as in the iron and steel industry, the paper trade registered the same sudden decline from late 1920 until July of this year. Like iron and steel, it showed a marked improvement of output in August and seems now to have turned the corner toward better conditions. The group which has shown the most consistent gain during recent months is that pro- ducing foodstuffs. The index for this group has risen steadily since May and showed in August the highest figure since January 1920. Undoubtedly some of this movement is due to seasonal influences, but an increase of output is clearly taking place in this fundamental line of production. In this the group is but showing a tendency now typical of manufac- ture in the United States. ADJUSTED INDEX FOR THE TEXTILE GROUP The group index of the volume of manufactiire in the textile trades, given by months from January 1919 to August 1921 in Table A, rests upon series of relatives whoUy uncor- Per cent 150 140 130 Chart 2. — Adjusted (Ordinates a and worsted go Indices of ti of Trend = "I IE Volume oi' MANtrFACxunE formal" = 100) ined textile group Cotton goods ,• ^.. \ ^:::^ ^ ^ ^ \ \ L r' V' \--- Y \. ••J / I ■••• v\ f/ ^S^ V V^l Y/ \ % ^ »*■«; (/^ \ au f i v-'-v.. \ ff^ ^y \/ ^ ^1 'V' \ ./ bu Jixn 19 uly 19 3ct. iQ.n 19 July ( 20 Jet Jan Apr JL^Iy rected for long-time trends and seasonal variations. In brief the relatives are simple per- centages of 1919 monthly averages. Within certain limits the index in this form is entirely 3 satisfactory. For some purposes, however, it is desirable that the index should show the cyclical changes of industrial output undisturbed by other movements. Such an index of the volmne of manufacture for the textile group, adjusted for elements of growth and sea- sonal variation, appears above. The general course of the adjusted index is not unlike that of the unadjusted index charted in the diagram on page 5 of the Special Letter of September 20. Upon the whole, as might be expected, the adjusted index shows a steadier movement. Its rise from the post-armistice slump is not as rapid as is that of the unadjusted index, and its decline after the peak of early 1920 is not so abrupt. Finally, it registers a somewhat more moderate recovery since the extreme of depression. Combining, as it does, the corrected figures for the two principal textile industries, cotton manufacture upon the one hand, and woolen and worsted manufacture upon the other, the adjusted index is a highly significant barom- eter of textUe production since January 1919. Cotton Goods The series imderlying the index of cotton manufacture is the monthly consmnption of raw cotton exclusive of linters. The data are presented graphically in Chart 3. The monthly items of this series are available from September 191 2 to date. Monthly average figures run back to the crop year ending August 31, 1909. By estimate from the series giving cotton consumption inclusive of linters, the average monthly consumption of raw cotton exclusive of linters may be carried a nmnber of years further back. There is thus every opportunity to obtain reliable measiu-ements of the long-time tendencies and seasonal Chart 3. — Monthly Constjmption of Raw Cotton {in thousands of bales; exclusive of linters) /« 6O0 500 1 ill hi k k \ i \ A r- — 7^ h } — — i r ;* 300 --*: ^ ^ ^ s^^ , r factors of cotton manufacture as represented in the consumption of the basic raw material. The long-time trend of the series has been determined by fitting a straight line to the monthly averages of the crop years 1905-14.^ Seasonal factors have been obtained from the monthly Hnk relatives September 1912 to date.^ Finally each item of the monthly ^ The equation is y = 8.8i6a; + 405.9, the unit of consumption being 1,000 bales, the unit of time 12 months, and the origin September i, 1910. ^ Full details wiU be given in a later issue of the Review of Economic Statistics. series of raw cotton consumption has been reduced to the form of a percentage of normal consumption, normal being taken as the trend corrected for seasonal influences. The re- sultant percentage relatives register the cyclical fluctuations of cotton manufacture, undis- turbed by the influence of normal growth or seasonal variation. They appear for the months January 19 19 to August 192 1 in column (b) of Table B and in Charts 2 and 5. Woolen and Worsted Goods The adjusted index for woolen and worsted goods rests upon two varieties of data, one consisting of the single series wool consumption, the other of three series reporting the per- centage of active wool machinery. The three machinery series employed are those showing the percentage active of (i) woolen spindles, (2) worsted spindles, and (3) wide looms. These three series have been combined into a single machinery index which in turn has been averaged with the series of wool consimiption. The analysis of the wool consumption series follows the standard method adopted in handling the records of raw cotton consumption. The data appear in Chart 4. The trend of wool consumption has been ascertained by taking a centered three-year moving average of wool available for consumption for the fiscal years from 1904 to 1914.-^ Seasonal factors have been approximated, partly on the basis of the monthly items which are avail- able from January 1918 to date, partly through consultation with experts in the industry.^ Chart 4. — Monthly Consumption op Wool {in millions of pounds) This measurement of trend and seasonal factors permits of a reduction of the monthly figures of wool consumption to percentages of normal consumption. These cycHcal items for the period January 19 19 to August 192 1 are given in full in column (d) of Table B and are shown graphically in Chart 2. ' The equation is y = .262* -|- 42.43, the unit of consumption being one million pounds, the unit of time 12 months, and the origiir January i, 1909. 2 The details of method here and elsewhere in this study wiU be set forth in subsequent issues of the Review. The utilization of the data on active wool machinery follows a somewhat different course.^ For these series, data are available in quarterly form as far back as 1913- No trend is discernible, nor is any to be expected. The percentage of idle, or of active machin- ery in any industry wUl seldom show any tendency toward marked increase or decrease. The problem is to determine the normal percentage of inactivity. The earUer study of the volimie of manufacture, based upon annual data, showed that production in the woolen and worsted industry was normal during 1914. The percentage of machinery active during 1914 has consequently been taken as a base from which to measure the deviations of the current monthly data on wool machinery. Thus it has been assumed on the basis of the 191 4 records that, in the reports of active machinery, 75 per cent active for wide looms is normal. Hav- ing reduced the original data by this method to percentages of normal, the relatives have been combined into a single machinery index, by first averaging the percentages for woolen and worsted spindles and then in turn averaging these results with the percentages for wide looms. The wool machinery index is given in the fourth column of Table V and in the upper right-hand diagram of Chart 5. The adjusted index of the volume of woolen and worsted manufacture is finally obtained as an imweighted geometric mean of the two series, one based upon wool consxmaption, the other upon active woolen and worsted machinery. This index for the woolen and worsted trade, showing the cycHcal movements of the industry properly corrected for trend and seasonal variation, appears in the index chart for the textile group and is given in full in column (c) of Table B. 2 This examination of cycHcal fluctuations in the textile group would not be complete without reference to several additional series which throw some Hght upon the subject, though they are not in shape for satisfactory incorporation in the index. The most signifi- cant of these series are (i) the percentage of capacity operated among mills turning out finished cotton fabrics, returned regularly to the Federal Reserve Board by the Associa- tion of Finishers of Cotton Fabrics; (2) the per cent of normal production of knit underwear, reported by members of the Knit Goods Manufacturers of America; (3) the consumption of raw silk, estimated by the Silk Association of America on the basis of imports and stocks in storage; and (4) the per cent of carpet looms active, in the wool machinery figures of the Bureau of the Census. These series are given in Table B and shown graphically in the dia- grams of Chart 5. While the series cannot be reduced to a form directly comparable with the indices, the general character of the fluctuations disclosed is significant. The per cent of capacity operated in the production of finished cotton fabrics and the per cent of normal production of knit underwear follow the same general course as the adjusted index for cotton manufacture. The range of fluctuation is different but the timing of movements is almost identical. Raw silk consumption from April 1920 to date follows a course not imlike that shown by the index of woolen and worsted manufacture; a rapid decline features the first six months of 1920; the second six months show consumption upon a low level, and 192 1 witnesses a conspicuous recovery. Only the series for carpets and rugs displays marked ' As a matter of fact the original items are reported as percentages of wool machinery idle. Conversion to percentages active is a matter of simple subtraction from lOO. " The final step in the construction of the adjusted index for the textile group is a simple matter of taking a weighted geometric mean of the adjusted indices of the two major textile industries. The weights employed in obtaining the mean are: cotton manufacture s, woolen and worsted manufacture 3. The index has already been presented on pages 3 and 4. It is given numerically in the first column of Table B. 6 divergences. Apparently this industry kept upon a high level of output much longer than others of the textile group and has experienced, since the extremes of depression, distinctly less improvement. Chart s — Series Indicative of the Volume of Textile Manufacture 150 140 130 120 110 100 90 80 70 60 50 COTTON G00D5 Per cent of nonmol ' consumption of raw cotton W y \.^ * \ , ^^v Vy N \r^ \ v \ k \ \ J — --\,~-~. — V—T'^^ -^^-v.^ 1919 1920 1921 150 140 130 120 110 100 90 80 70 60 50 WOOLEN AND WORSTED GOODS , »^» Per cent of normal consumption of- wool ...H....... » M H »i activtty of m&chinerjj ^ A // VT 1919 ^v V /■ 'm "r: Hi ^[ JiL /! tt ^l 1920 1921 100 90 80 70 60 50 40 30 20 10 FINISHED COTTON FABRICS Per cent of mill ccpacitg operated 50 40 30 20 10 'Cbnsutnf^ion SILK, GOODS of raw silk in thousands of bales jA r^ i / ^ \ r \ \ J \> 1 \ r\ "V V ^ 1919 1920 1921 1919 I9£0 1921 100 90 60 70 60 50 40 30 20 10 KNIT GOODS ,^ Pep cent of "nopmol production of knit underwear — ^ > \A ^ \ ,/ \ / I / \ / \ r 1919 1920 1921 100 90 80 70 60 50 40 30 20 10 CARPETS AND RU65 Per cent of carpet looms active ^,,,_^ ^ A J \i J \— / 1919 1920 1921 While these comments must be looked upon as provisional — for the series will not per- mit of the standard analysis upon which the adjusted indices are based — they suggest that the entire textUe group has been subject to much the same experience since January 1919. 7 The character of that experience is registered with reasonable accuracy in the adjusted index of the volume of textile manufacture. Table B. — Series Indicative of the Volume oe Manufacture * Month 1919 January February- March April . . May . . June . . July .. August September October . . November December 1920 January . February March April . . May .. June . , July .. August September October . . November December 192 1 January . February March April . . May .. June . . July .. August Adjusted Indices Textile group t (a) 94.6 77.2 73-8 87-5 91-3 98.0 107.9 109.0 111. 6 118.1 110.4 II3-9 119. 1 111.4 112. o 111. 7 101.2 100.3 92.6 91.4 89.1 79-4 69.1 61.3 61.9 75-2 79-7 81.6 86.3 96.2 88.8 106.3 § Cotton industry J (b) I08.S 92.1 82.1 93-7 93-7 94.9 103.7 104.9 107. 1 112. 6 105.9 109.8 II3-S 107.3 109.2 1 10.3 102.4 109.3 105.0 100.2 98.6 79-4 72.S 64.8 66.6 81.7 80.0 77.1 80.7 89.1 80.5 95-4 Woolen and worsted industry |[ (c) 75-3 S7-S 61.8 78.1 87.4 103.4 115.1 116.4 119.7 128.2 118.4 121.4 129.3 118.6 116.7 1 14.0 99.2 86.9 75-3 78.2 75-1 79.2 63.8 55-8 54-7 6S-S 79.2 89.7 96.7 109.4 104.7 Woolen and Woested Goods Per cent of norma 1_ consumption o{ wool Cd) 77-5 Si-S 55-1 71.9 80.0 103.4 121.9 123-5 127.4 143.0 121. 1 127.9 I44-S 122. 1 118.9 112.9 91.8 84.7 76.4 87.4 78.1 84.1 58.1 50.8 62.8 69.1 79-7 86.5 90.0 109.9 101.6 127. 1 Per cent of normal activity of spindles and looms Finished Cotton Fabrics W 73-3 64.2 69.4 84.8 9S-S 103.6 108.7 109.6 112.4 115.2 116.0 115.2 115.6 115.0 1 14.6 115.0 107. 1 89.2 74-3 70.0 72.2 74-7 70.0 61.4 47.6 61.9 78.6 93-1 104.0 108.9 107.8 Per cent of miU capacity operated (0 SI 5° 41 33 25 29 33 51 67 66 68 74 62 Knjt Goods Cakpets AND Rugs Per cent of normal production of knit underwear (g) 81.7 80.3 81.7. 82.1 82.2 80.3 73-4 67-3 74.2 504 23.2 II.O 16.4 28.0 50.1 49.6 55-4 65-5 51.2 71.1 Per cent of caipet looms active (h) 34-2 34-4 38.6 42.9 5I-I 55-5 61.4 64-5 62.8 65-9 65-5 66.5 69.8 71.4 72.3 71.8 71-5 70.9 67.9 67.7 64-3 653 61.S 59-9 42.6 36.5 36.1 35-7 38.0 44.1 * For data on consumption of raw cotton, wool, and raw silk, see issues of Current Siaiistics, Section III, Table i. t Weighted geometric average of columns (b) and (c). || Geometric average of columns (d) and (e). { Per cent of normal consumption of raw cotton. § Provisional. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH Charles J. Bullock, Chairman Waeeen M. Persons, Editor Caiuridoe, Massachusetts, U.S.A. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH THE COMMODITY PRICE SITUATION special Letter — November 8, ig2i THE present commodity price situation diflEers strikingly from that in 1913, the last year preceding the price upheaval accompanying the war. Individual prices now are in very different relationships to each other and to the general level. This dislocation seriously curtails the purchasing power of important groups of consmners. Unusual or abnormal as the present situation seems on the surface, similar situations have, nevertheless, been re- current in the past. The price situation of the present period of business depression differs from that of similar periods of the past in degree rather than in kind, while the relatively low price at present of many raw materials and of other articles used in production is a favorable circiunstance for manufacturing activity. Two types of changes are always taking place in the price movement of individual com- modities. The first is the long-time change, which means that a single commodity is be- coming cheaper or dearer during a number of years. The long-time trends of the prices of different commodities differ greatly; some tend to rise rapidly, others slowly, some to de- crease. The result of these differences, over a number of years, is to change radically price relationships; certain commodities, for instance, growing dearer, command in exchange larger and larger amounts of those growing cheaper. The second type of change is the fluctuation accompanjdng the business cycle, that is, low prices for most cormnodities in times of depression, and high prices in times of pros- perity. This, likewise, affects commodities unequally; certain of them fluctuate violently, others only sKghtly, from times of prosperity to times of depression. Like depressions of the past, today's depression has created great disparities in commodity price relationships when compared with those of a period of business expansion. Thus, many raw materials are relatively cheap compared with finished products; whereas during the months of pros- perity preceding the crisis of 1920 the reverse relationship generally obtained. COMPARISON OF PRICES AT PRESENT AND IN 1913 Chart I, on the following pages, compares graphically the present (October) prices of 49 commodities with their prices in 1913. The upright bars show the percentages which the present price of each commodity forms of the 1913 price. Such percentages wiU hereafter be referred to as "relative prices." The relative price of each commodity in 1913 is thus 100, as is also the average relative price of the 49, while the relative price for October is the percentage which the actual price forms of the actual 1913 price. The present average relative price of the 49 commodities considered is 133. Several of these commodities are selling far above the average level, while others are far below it. The effect on purchasing power of such differences may be illustrated as follows: in 1913, a ton of bituminous coal sold for $3.55 and a bushel of wheat for I1.04. In October 1921, a ton of coal sold for $6.75 and a bushel of wheat for $1.34. Both have increased in price, but coal has increased more than wheat. To put the matter in another way: COPYRIGHT, 192 1, BY HARVARD TJNTVERSrTY Chart I. — Relative Wholesale Commodity Prices, October 192 i. (Price in 1913 = 100) Pei^Cent aoo £30 Bjxpley Gate Wheat Cotton Gray goods Silk WorstedyDm Corn Rice Wherftloir Coltonyom Rnishedgoocb V*ol 5erge GRAINS AND FIDUR FIBRES AND TEXTILES Iron ore Steel billeb Tank plates Coppepwire Tin Zinc Rqiron Steel rate Copper Lead Tlnfiote METALS AND METAL PRODUCTS 3.4 bushels of wheat would buy a ton of soft coal in the year 1913, while in October 1921, five bushels would be required. Such differences involve hardships for the individual con- cerns and industries whose products are selling at a comparative disadvantage. Anthracite coal, cotton finished goods, serge goods, brick, window glass, newspaper, and tobacco are all selling at prices twice as high, or more, than their respective 1913 figures. On the other hand, hve sheep, beef, hides, copper, copper wire, tin, coffee, and rubber are selling well below their 1913 prices and very far below the October average price for the 49 commodities. These two groups show the extreme variation; between them he other com- modities, presenting similar, if less pronounced, differences. Bituminous coal, cotton gray goods, worsted yarn, and alcohol, although at relative prices somewhat below those of the first group of commodities enumerated, are nevertheless quite high. Corn, live hogs, lard, and zinc, although not as low comparatively as others, are much below the general level. Some commodities are selling close to the average level, but they are surprisingly few. Wheat,, wheat flour, pig iron, and coke may be mentioned. The actual and relative prices for all commodities are given on page 8. The differences between the changes which have taken place in the prices of individual Relative Prices oe 49 Commodities by Groups Relative price (1913=100) 20- 29 30- 39 40- 49 SO- 59 60- 69 70- 79 80- 89 90- 99 100-109 Number of commodities I O O o I 4 5 3 6 Relative price (1913=100) Nmnber of commodities IIO-II9 5 120-129 3 130-139 2 140-149 3 150-159 2 160-169 3 170-179 I 180-189 2 J90-199 I Relative price (1913=100) Number of commodities 200—209 2 210-219 2 220-229 • ^ 230-239 I 240-249 O 250-259 O 260 269 O 270-279 o 280-289 I Total ^ 49 Chart I. — Relative Wholesale Commodity Prices, October 1921. (Price in 1913 = 100) Percent / AvePixge Price 0cfaberl9gl. '/. /\ ri / / ZL ■ I I I mi U I i 250 200 150 100 50 Beevea Sheep Hams Pork Hides CoalMhc) Cote Brick Unseed oil Atolnol Paper Sugar Hogs Beef Lard Cii,lfskins LetJhep CoB,l(bihjiJ FfeTroleum Glass Steel betms Coffee Rubber Tobaccc LIVE 5T0CK,MEAT PRODUCTS, HIDE3 ,18 LEATHER FUEL BUILDING MATERIALS MISCELLANEOUS commodities is shown in another way in the frequency table on page 2. It gives the nmnber of commodities whose relative price falls in each of 27 price groups. These groups extend, by ten units each, from the group 20-29 to the group 280-289. If ^ commodities had made about the same advance in price over 19 13, the entire 49 would faU in the same group — presumably the group 130-139, since the average relative price, 133, is at present within that group. Instead of this, or some approximation of it, being the case, there is a decided scatter from 20 to 289. Up to this point we have been comparing the commodities individually. The same sort of difference exists, however, between groups of commodities. For purposes of comparison the average relative price for each of seven groups has been computed. For each of four groups, namely, fibers and textiles, fuels, building materials, and miscellaneous, the aver- age relative price is above the general average of 133 ; while in the case of the three other groups — grains and flour; live stock, meat products, hides and leather; and metal and metal products — the group average is below the general average. Two groups — fibers and textiles, and buUding materials — are very high, their average relative levels being 173 and 179, re- spectively. On the other hand, the group consisting of Uve stock, meat products, hides, and leather is 3 per cent below the 1913 level. Figures for all groups may be found in the table on page 8, and the data are presented graphically by the charts on page 4. In each one of these group charts the horizontal Hne designated "A" represents the average price level of all 49 commodities, i. e., 133 on the base of 100; and the line designated "B " repre- sents the average price of the commodities within the particular group. Certain contrasts may be pointed out between individual commodities within some of the separate groups. For example, in the group of grains and flour there is pronoimced disparity between the current levels of corn and wheat. Corn is 13 per cent below its 1913 price; while wheat is 28 per cent above its 1913 price, being somewhat below the average level for the 49 commodities considered, although the group of grains and flour as a whole is far below the average relative price. In the groups of metals and metal products the non- ferrous metals bring down the group average to the point at which it stands; iron and steel 3 prices coincide almost exactly with the general average. The fibers and textiles group is at a price level much above the general average, and the individual commodities within this group are uniformly high, with the striking exception of wool, the current price of which is only i8 per cent above the 1913 figure. Among the fuels, petroleum is comparatively low, Chart II. — Relative Wholesale Commodity Prices by Groups, October 1921 (Price in 1913 = 100) Pec Cent 25 O 200 150 A B 100 Bsrlevf Gate Wient Com Rice Vtel flour GRAINS AND FLOUR Per Cent- aso Ironore Steelbte Torkphte Copperwire Tin Zinc Pig iron MrA Copper LeKl Tinflsle METALS AND METAL PRODUCTS Pep Cent 250 Coil(bnW Coke C(w(Di11iii) P^lem FUEL Percent 250 aoo ISO A 100 B I :i I m. BccTOo oii«p Hams Ports Hides Hogs Beef Lird CtJfsKins f-ealtier LfVE STOCK, MEAT PR0DUCTS,M1DES«LEATHER Percent 230 200 R ^ '7 ,< - ISO A h- 100 50 ConCjfi tiuayGjJi Uiii\ v'ivilfilyani Mmyam rmtied goods Wool Serge FIBRES AND TEXTILES 102 per cent of its pre-war price; on the other hand, coal, both bitiuninous and anthracite, is at an extremely high level. A similar contrast exists between the low relative price of steel beams and the high prices of the other conmiodities in the building materials group. Our use of prices in the year 19 13 as the base of comparison for present prices should not be taken to indicate that we consider that price relationships of commodities in that year are relationships to which we must return and, therefore, constitute a standard for future price revisions. Our object in making such comparison, in this Special Letter and that of August 12, is to set forth the disparities or maladjustments in purchasing power of different commodities resulting from the price upheaval since the war. In the following sections we have an additional object: that is, to show that there are influences normally at work to change the prices of individual commodities and also their adjustment to one another. 4 THE PRE-WAR PRICE TREND OF VARIOUS COMMODITIES The diverse trends of various commodities and groups of commodities during an ex- tended period, say since 1898, show that previous to 1913 the relationships among com- modities had been changing gradually, producing slowly the same sort of "maladjustment" as that existing today. The lines on Chart III show the direction of movement of the prices of seven groups of commodities for which indices are pubUshed by the United States Bureau Chart III. — The Direction of Movement oe the Prices oe Various Groups of Commodities, 1898-1914 of Labor Statistics. Each line passes through the arithmetic average price for the period. Each has a slope depending upon the actual price movement of the commodity group for which the line is determined. It wiU be seen from the chart that farm products, lumber and building materials, and food all increased more rapidly than "aU commodities," while cloths and clothing, fuel and hghting, chemicals and drugs, and metals all increased less rapidly. In other words organic materials increased most rapidly, manufactured articles second, and mineral products least. The course of prices of individual commodities supports the general conclusion stated in the preceding paragraph. For instance, the average annual price increase ^ of various im- portant commodities is as follows: pork, 4.24; corn, 3.80; cotton, 3.47; oats, 3.14; beef, 3.12; hides, 3.05; wheat, 2.42; print cloths, 2.27; calf leather, 2.16; sheetings, 1.46; bitu- minous coal, 1.46; pig zinc, 1.43; lead, 0.15; pig iron, — o.oi; copper, —0.20. It is evident that with the great variety of long-time movements in progress durmg the period 1898-1914 the intercommodity relationships were changing from year to year radically. There is no reason for supposing that a diversity of long-time movements of commodity prices will not develop in the years to come as it has developed in the past. The comparison made in the first section brings out wide disparities between present price relationships and 1 Expressed as percentages of the average price of the respective commodities for 1898-1914. Similar figures for all the groups of the Bureau of Labor Statistics are: farm products, 2.61; lumber and building materials, 2.27; house-furnishing goods, 2.05; food, etc., 193; all commodities, 1.83; miscellaneous, r. 40; cloths and clothing, 1.28; fuel and lighting, 1. 18; chemicals and drugs, 0.4s; metals and metal products, —0.02. Similar figures for 27 agricultural and mineral raw materials, 71 articles manufactured from them, and the 98 commodities combined, are, respectively, 2.62, 1.74, and 2.22 those of 1913. Doubtless these disparities are partly attributable to a fundamental diver- gence of the long-time movements of the prices of individual commodities. COMMODITY PRICE MOVEMENTS DURING BUSINESS CYCLES We have seen that the prices for different commodities and for different groups of com- modities move in different directions over a period of years, such as 1898-1914. Such long-time movements are not the only fluctuations which affect prices. The prices of com- modities in general and of many different groups of commodities alternately rise and faU during periods of business prosperity and depression. Although the prices of commodities in general and many groups of commodities rise and faU together during business cycles the extent of the rise or fall varies greatly for different commodities and groups of commodities. There are certain basic commodities, such as cottonseed oU, coke, pig zinc, pig iron, bar iron, mess pork, hides, print cloths, sheetings, and worsted yams, which always have more violent price movements than do other com- modities, such as steel rails, soda crackers, dress goods, wire, and shoes. In general, the prices of raw materials and of articles used in production respond in larger measure to changes in business conditions than do the prices of finished products. Chart IV shows the Chart IV. — The Price Index for Ten Important Commodities (A), Compared with THE Price Index of the Bureau of Labor Statistics for All Commodities (B), Monthly, 1903-15 Percent / vv I I fe i \s ■s^ ej^ t^A Y a K r \ \ r ,Jy*^ ^ 1904 1906 1906 1907 1908 1909 1910 1911 1912 1914 1915 price index of the ten basic commodities named above compared -vyith the index of the Bureau of Labor Statistics based upon over 300 commodities. The ampUtude of fluctuations of the ten-commodity index, (A) , that is, the distance from peak to trough and trough to peak for the period 1903-14, ranged from 33 to 51 points and averaged 38 points, compared with a range of 4 to 19 points and an average of 9 points for the index of the Bureau, (B). In other words, during the pre-war period the prices of certain commodities and groups of commodi- ties regularly increased in larger measure during business prosperity and declined in larger measure during depression than did other commodities. It appears, therefore, that a "normal" relationship between the prices of various commodities for one phase of the busi- ness cycle is not "normal" for another phase. In periods of depression, like the present, 6 it is usual for materials used in production to show much more drastic price recessions than do finished goods and goods ready for consumption. The relationship which we have found for the pre-war period holds for the post-war period. Our ten-commodity index and the Bureau's index both reached the highest point in May 1920, and both reached bottom in Jidy 192 1. Our index declined 61 per cent, how- ever, compared with 46 per cent for the Bureau's; in September the former index was 115 of the average for 1913, the latter, 152. The extent of the price decline which occurred in 1920-21 was imparaUeled; the irreg- ularity of the decline, however, in which certain basic commodities reached very much lower points than did commodities in general, is "normal" for a period of business recession. Only the degree of the irregularity is imusual. The price situation in which we now find ourselves in 1921, therefore, is not unprecedented. Rather, it is a situation to be expected in the present phase of the business cycle. Price changes are in prospect, as they always are in prospect; it is probable that the prices of certain groups of commodities wiU advance very much more than will the prices of other groups, as always occurs when business is rising from the trough of depression. Our analysis leads us to the conclusion, consequently, that although many considerable price changes have occurred and others are in prospect, the present maladjustment of prices is merely the result of a somewhat greater disturbance than is usual in the present phase of the business cycle. The great decline in the prices of certain commodities, such as farm products, has di- minished the producing power of a considerable portion of our population and has thus acted as a depressing influence upon industry. But there is another side to the story; al- though price disparities have a depressing influence on certain industries they ultimately have a stimulating effect on other industries. In fact, a situation in which many materials for manufacture are relatively cheap compared with finished goods — the reverse of that obtaining before the declines of 1920-21 — is one which should lead to industrial activity. SOURCES AND METHODS Current data. Four of the commodities used in the comparison of present prices with those of 1913 are composites of two or more similar commodities, as foUows — (i) cotton yarn: southern two-ply chain warps, 2/20's, and southern frame cones, id's; (2) cotton gray goods: print clothes, 28-in., 64 X 64, gray goods 39-in., 68 X 72, and brown sheeting 3-yd; (3) cotton-finished goods: standard prints, and standard staple ginghams; (4) wool: Ohio fine delaine, and Ohio quarter blood. The quotations utilized in this study are those prevailing on or near the first of the month designated. They have been taken from the following sources: the American Wool and Cotton Reporter, Bradstreet's, the Commerce Monthly published by the National Bank of Commerce in New York, the Daily Metal Re- porter, Dun's Review, the Iron Age, and the New York Journal of Commerce. Annual data, i8g8-igi4. Lines of secular trend were fitted by the method of least squares to the group indices and to the prices of leading individual commodities which are published in Bulletin No. 269 of the United States Bureau of Labor Statistics. The slope of each line thus found was expressed as a percentage of the average price 1 898-1914 of the group or commodity for which the Une was determined. Monthly data, igoj-ij. The criteria for the selection of the conunodities used in our ten-commodity index, sources, methods, and data for the period 1890-1921 will be pub- lished in The Review of Economic Statistics for November, 1921. 7 Comparison of Wholesale Prices of 49 Commodities in October 192X with Prices FOR THE Year 1913 Commodity Unit Price for the year 1913 Price in October igsi Index No. for October 1921 (Price in 1913=100) Grains and Flour Barley, No. 2, Milwaukee Corn, No. 2, mixed, elevator Oats, No. 3, white, elevator Rice, domestic, good Wheat, No. 2, red winter, elevator Wheat flour, straight winter Live Stock, Meat Products, Hides and Leather Beeves, best native steers, Chicago Hogs prime, Chicago Sheep, prime, Chicago Beef, family Hams, smoked Lard, western steam Pork, new mess Calfskins Hides, native steers, No. i Leather, oak, scoured backs. No. i Metals and Metal Products Iron ore, old range, Bessemer, hematite Pig iron, Bessemer, Pittsburgh Steel billets, Bessemer, Pittsburgh Steel rails, standard Bessemer, Pittsburgh Tank plates, Pittsburgh Copper, electrolytic. New York Copper wire, net, base price, f . o. b. mill Lead, New York Tin, New York Tin plate, Pittsburgh Zinc, New York Fibers and Textiles Cotton, middling upland, spot. New York Cotton yarn. New York * Cotton gray goods. New York * Cotton-finished goods. New York * Silk, shinshiu. No. i. New York Wool, clean basis, Boston ' Worsted yarn, 2-40's half blood, Bradford spun, Boston Serge, i i-oz.. New York Fuels Coal, anthracite, stove sizes, New York Coal bituminous (Pitts.), f. o. b. Chicago Coke (ConnellsviUe), furnace, f. o. b Petroleum, crude, New York Building Materials Brick, Hudson river, hard. New York Glass, window, 10 X 15, New York Linseed oil, New York Steel beams, Pittsburgh Miscellaneous Alcohol, 190 proof, U. S. P., New York Coffee, Rio No. 7, New York Paper, news, roll, spot, New York Rubber, plantation, first latex crepe. New York Sugar, standard granulated, New York Tobacco, medium leaf, Burley, LouisviUe bu. bu. bu. lb. bu. bbl. 100 lbs. 100 lbs. 100 lbs. bbl. lb. lb. bbl. lb. lb. lb. ton ton ton ton lb. lb. lb. lb. lb. box lb. lb.' ' ' lb. yd. yd. lb. lb. lb. yd. ton ton sh. ton bbl. 1000 box gal. lb. gal. lb. lb. lb. lb. lb. •71 .698 .429 •0579 1.0446 4-52 9-23 8.48 5-76 21.58 .164 .1096 22.06 .20 .181S •44 4-35 17.12 25-79 28.00 .015 .1569 .1681 .044 •4433 3-SS .0576 .129 .225 .0567 •OS99 3-65 •SS .96 3-SS 2.46 2.4s 6.S4 2.16 •49 .los .251 .1124 .0225 .82 .0442 .0862 .69 .61 ■47 .0675 1-335 6.00 9-75 7.20 4-25 15-00 -25 .1025 25-00 .19 -145 •45 6.45 21.96 29.00 45.00 .016 .1238 -1363 .047 .267s 5-25 ■0505 .211 ■355 .1008 .128 6.0s •65 1.80 2.425 10.56 6.75 3-25 2.50 iS.op 5-13 ■70 .016 4.70 .0825 ■0475 .1625 ■055 •25 112 97 87 no 116 128 133 97 106 85 74 70 152 94 "3 95 80 102 III 148 128 112 161 107 79 81 107 60 148 88 173 164 158 178 214 166 118 188 200 158 207 190 132 102 179 229 238 143 107 151 187 73 211 20 124 2S9 * Includes two or more commodities. See Sources and Methods, p. 7. harvard university committee ON ECONOMIC RESEARCH CAMBRIDGE, MASS., U.S.A. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH THE VOLUME OF MANUFACTURE INDEX FOR ALL GROUPS COMBINED ADJUSTED INDICES FOR TEXTILES AND IRON AND STEEL Special Letter — November 2g, ig2l ALL GROUPS COMBINED IN our Special Letter of October 22, it was stated that "an upward movement of industrial output seems definitely under way." More recent reports substantiate this view. This is true despite the somewhat disappointing record of September. Upon the whole, during September industry accomplished little more than a consolidation of the gains of August. Certainly no substantial new advance was registered. But the significant fact is that the Chart i. — Index of the Volume oe MANtrEAcruEE {igig Monthly Average = 100) Percent Apr? ~ July 1921 progress made during late summer was firmly held. More recent weeks, from all accounts, have witnessed a real gain in output over the greater part of industry. The Index for Octo- ber win probably show a marked rise, if for no other reason because of material improvement in the iron and steel trade. While the data for September and October are still incomplete, the Index for the former will probably be about 82, for the latter, considerably higher. Be- yond question the underlying tendency at present is toward a larger volume of manufacture. COPXEIGHT, igll, BY HARVABD ITNlyEESITY In part this situation is a natural consequence of seasonal influences. The fall months normally witness an iacrease of manufacturing activity in practically all lines. It will not be surprising if, for similar reasons, there is a moderate decline in output when winter sets in. Certainly an uninterrupted upward movement is not to be expected if the winter proves severe. But it is encouraging to find industrial output responding once more to seasonal factors and no longer completely imder the influence of business depression. This fact promises well for the spring when the forces underlying industrial output should be unusually buoyant. Turning to the indices for single industrial groups, the most significant movement is shown by the index for iron and steel. In Jiily this index fell to the extraordinarily low figure of 34.0. In other words, the output of the iron and steel trade in July was only about one-third of the 1919 monthly average. In August the index reboimded to 46.0; in Septem- ber it remained on approximately the same level; and in October it rose to 63.9. Of course, production is still very seriously below normal, but recent changes nevertheless represent real progress.^ Another line of manufacture which has recently made substantial headway is the paper trade. The index for this group stood at 71.0 in July; it has since risen to 88.2. Part of this change is typical of the trade at this season of the year; part of it is due to the breaking of the exceptional drought which visited the northeastern states dxuing the smnmer and seri- ously interfered with the operation of paper mills depending upon water power. Whatever the causes of the advance, the improvement is unmistakable. "Many paper miUs which were practically shut down during the summer are now nmning close to capacity." A further rise of the paper group index may reasonably be expected. The food group is another which has recently shown somewhat greater output. Manu- facture of foodstuffs — represented by flour, meat, and sugar — was, during September, as in August, fairly close to the 1919 average. Undoubtedly, this was due in part merely to the characteristic fall activity of these trades. Yet it is significant that the manufacture of food has been greater this fall than during the corresponding months of 1920. When it is remembered that 1919 was a year of record-breaking in several Unes of food manufac- ture, the recent level of the food group index is gratifying. The textile group has held firmly the ground gained during the first six or eight months of the year. The group index for September was 105.7, ^ figure not equaled since May 1920. Wool manufacture continues somewhat more active than cotton, but cotton has recently made consistent gains. As to output both industries appear now to have recovered largely from their coUapse of 1920.^ The volume of manufacture in the remaining groups shows no change requiring detailed comment. In the production of liunber and leather no significant movement has occurred since May. In the output of petroleum products a slight downward tendency seems dis- cernible but no radical change has been recorded since March. Tobacco manufacture rose in August to the highest level reported since early 1920, but output in this less important industry has at no time shown the sustained curtailment evident elsewhere. In general these groups have contributed little to the improvement of the past two or three months. '■ A more detailed analysis of the situation in the iron and steel trade is to be found further on in this letter in the section dealing with an adjusted index for the iron and steel group. ^ The course of cotton and wool manufacture is more satisfactorily examined in the following section devoted to the adjusted index for the textile group. Table A. — Monthly Indices of the Volume of Manufacture for Eight Groups of Manufacturing Industries and for the Groups Combined (igiQ Monthly Average = loo) Month AU groups Iron and steel Ljmber W (3) 130.1 75-5 "3-7 75-8 II3-S 88.8 94.6 97.8 81.2 112. 1 90.9 I03-S 103.1 103.7 "3-S 123.4 97.6 113.8 73-2 123.4 94.0 102.0 103.4 80.1 124.2 lOI.I 120.5 101.3 138.3 117.4 III.O 122.3 121. 2 127.3 124.9 iiS-4 118.0 103.1 I2S-3 117.8 125.0 108.7 126.7 108.1 111.3 86.2 99-S 62.8 92.7 S9-S 73-7 65-3 65.2 78.2 So-S 84.6 52.4 100.7 42.3 93-3 34-0 86.8 46.0 99.6 47-4 63-9 Paper Petroleum Textiles Leather Food (7) (8) 109.4 II4.9 91.0 87-3 92.8 92.1 108.7 92.8 106.3 98.S 108.6 83.8 99.8 85-9 102.6 88.4 102.0 97-7 IOI.4 III. 2 87.4 109.0 89.9 108.6 90.8 112. 2 81.6 85-3 94.0 90.8 84.7 75-9 90.9 85.1 95-2 80.8 80.7 80.8 70.S 83-3 73-3 78.0 77.8 81.1 70.1 91.0 72.1 88.4 63.5 84.1 62.8 74-4 72.0 85.6 75-8 82.4 83.2 79-9 81. 1 82.7 76.3 84.5 85-7 96-3 80.3 93.8 Tobacco 1919 January . . . February . March . . . . April May June July August . . . September October. . . November December . 1920 January . . February . March . . . . April May June July August . . . September October. . November December. 1921 January . . February March . . . , April May June .... July August . . . September October . . . (I) 101.6 86.7 92.6 93-7 95-7 95-9 101.9 107.2 103.8 104.4 102.2 102. 1 iiS-9 104.6 118.0 108.8 111.8 109.6 102.3 104.9 101.4 101.2 88.9 77-9 78.3 7S-0 80.6 7S-8 79-3 75-9 68.5 81.2 81.6 (4) 93-0 82.8 87.0 88.4 9S-I 98.3 ro3-3 108.8 106.2 116.4 108.6 109.8 120.9 105.8 II9-S 121. 7 119. 2 122.5 123.0 122.5 "9-3 119.7 108.0 94.6 84-3 78.S 83.2 79-3 73-6 75-6 71.0 84.4 88.2 (s) 89.7 83.8 92.8 92.4 100.7 96.1 103.7 107.4 108.3 II 2.0 107.0 107.6 102.3 97.0 III. 6 109.3 115.0 iiS-9 122.9 132.2 134.6 I3S-2 131.2 134.6 13 1. 6 115.0 117.9 124.9 122.8 122.7 119.7 119.7 118.2 (6) 96.5 72.7 79.0 92.9 99.2 100.2 109.4 103.4 ioS-3 II9-3 104.9 IIO.I 126.6 iio.s 121.0 II9-3 110.5 104. 1 92.8 88.2 834 78.7 62.1 S4-S 68.1 76-3 89.3 89.3 96.2 100.8 89-5 100.7 ioS-7 (9) 87-3 83.8 97-7 82.0 88.0 92.9 99-3 lOI.I 108.6 125.4 117.0 113-0 iiS-7 98.9 123.1 108.0 III. I 1 14.4 98.1 104.8 104.9 106.3 93-1 73-5 88.7 9S-0 106.6 96.6 100.6 108.0 101.8 118.3 113-1 * Subject to slight revision. At the same time it should be noted that no one of the indices has recently registered a significant decHne. The period through which industry is now passing affords an excellent example of the need of an index properly adjusted for seasonal variations and long-time tendencies. We know that industrial output has recently increased, but we do not know how much of the increase is to be charged to customary fall activity. To put the question specifically — to what extent does the rise of the general index represent a net improvement in underlying 3 conditions — a real turn in the cycle? Only an adjusted index can provide the answer. Such an index for the textile group was presented in the Special Letter of October 22. This index is carried into October in the next section. In a final section of this letter a similar adjusted index is developed for the iron and steel trade. THE TEXTILE GROUP The cycles of output since January 191 9 in cotton and wool manufacture appears clearly in the index chart below. The numerical items of the index are given in fuU in Table B.^ It will be recalled that in this adjusted index, allowances have been made for the normal growth of the industries and their characteristic seasonal movements. Fluctuations of the Chart 2. — Adjusted Index for the Textu-e Groxjp {Ordinates of Trend = "Normal" = 100) Percent 130 K i \ 110 100 \ -? A V v V, V\ A f 90 80 70 60 50 4. I ^ V r J V — \ J Jan App 19 July 9 Ocl Jan. (^pn 19 July ( SO 3ct. Jon ^pi? July t9£l index thus reflect the business cycle alone, disturbed, if at aU, by fortuitous factors like strikes or extraordinary weather conditions. The course of the index during recent months gives striking proof of the substantial recovery the textile trades have made since last winter. The decline in July was apparently only a temporary relapse. The volume of textile manufacture is now upon the level of the early fall of 1919, even when allowance is made for normal expansion of the trades. Textiles have been among the first to recover in 192 1, just as they were among the first to coUapse in 1920. Of the two lines, cotton and wool, the latter has moved somewhat earlier throughout the cycle. The precipitate drop in wool manufacture occurred in 1920 four or five months earlier than the corresponding decline in cotton manufacture. More recentiy wool has shown a speedier recovery than cotton. As early as June the woolen and worsted trade was weU back to normal; not until September was cotton manufactiure on a similar level. ' The adjusted index here presented is a weighted arithmetic mean of relatives based upon two series: (i) raw cotton consumed; (2) wool consumed. The more elaborate methods employed in deriving the adjusted textile index in the Special Letter of October 22 give substantially the same results as shown in Chart 2 and Table B. The more refined analysis confirms the simpler but does not give enough better results to justify the greater effort. The data on woolen machinery will still be studied in interpreting the textile index, but the data will not for the present be incorporated in the index. Table B. — Adjusted Indices for the Textile Group and foe Cotton and Wool Manufacture Montli January . . February . March . . . April May June July August . . . September October . . November December Textile Group 1919 96.4 76.6 71.9 85.1 88.1 97-4 IIO.I III. I "3-9 123.6 III.O 116.1 X920 124.6 112. 6 112.4 110.8 98.2 99.6 91.6 9S-I 90.4 80.9 66.8 59-1 64.9 76.5 79.6 80.4 83.8 96.2 87.9 107.0 112. 1 Cotton 1919 107.8 91.6 81.6 93-1 93-0 94.2 103.2 104.3 106.4 111.9 105-3 109.2 1920 112. 8 106.6 108.7 109.6 101.9 108.6 104.4 99-7 98.2 79.0 72.1 64.4 66.1 81.2 79-S 76.8 80.1 88.5 80.0 95-0 101.8 96.2 Wool 19 19 77.2 51-4 54-9 71.7 79.8 103. 1 121. s 123.2 127.0 142.6 120.8 I27-S 144.2 121.8 118.S 112. s 9I-S 84.4 71.2 87.2 77-9 84.0 58.0 S0.7 1921 62.6 68.9 79-4 86.3 89.6 109.6 101.3 126.8 129.4 Examination of the available data would indicate that practically all sections of the textile trades have shared in the improvement since midsummer. True, the reports of active machinery hours show no significant increase in general woolen and worsted manu- facture as compared with May-Jime. But wool consumption exhibits a marked rise, and carpet and rug manufacture seems to have experienced a steady improvement straight through from spring to September. The production of knit underwear likewise has shown a continuous increase and now is in larger voltune relatively than at any time since the reporting service for this trade was inaugurated in early 1920. As a whole, manufactiure in the textile trades was clearly in September upon a higher level than at any time since early 1920. Table C. — Series Indicative of the Volume of Textile Manufacture in 192 i Month Woolen and Worsted Goods Per cent of machinery hours reported active * Sets of cards Woolen spinning spindles Combs Worsted spinning spindles Wide looms Carpets AND Rdgs Per cent of loom hours reported active* Knit Underwear Per cent of actual to normal production Finished Cotton Goods Per cent of factory capacity operated Silk Goods Consumption _ or raw silk in thousands of bales January . February March . . , April May June July August . . . September October . . 33-9 35-7 49.4 64.2 75-0 80.2 82.1 79-4 78.9 79.0 31.6 3S-S 49-5 65-9 77.1 81.4 81.8 80.0 78.8 79.1 37-1 49.0 73-8 88.7 94.6 98.6 95-9 87.4 91.2 97.6 34-8 44-7 62.1 74-3 86.S 89.6 93-4 85-7 88.S 92.2 33-3 40.0 54-7 61.7 73-4 80.1 82.S 79.2 76.7 74.2 42.S 36.S 36.1 35-7 38.0 44.1 47.1 42.1 57-4 6S-S 16.4 28.0 So.i 49.6 55-4 6S-5 SI. 2 71. 1 84.4 33 SI 67 66 68 74 62 71 75 22.2 16.S 25.6 28.9 27.2 33-8 32-3 32.8 31.2 * As of the first of the month. S THE IRON AND STEEL GROUP Examination of the adjusted index for the textile group suggests the development of similar indices for other important divisions of industry. An adjusted index, corrected for long-time tendencies and seasonal variations, has already been obtained for the iron and steel trade. The index is given graphically in Chart 3, numerically in Table D. The chart exhibits strikingly the cycle through which the iron and steel trade has passed since January 1919. The post-armistice reaction in this basic industry was slight, and almost immediately gave way to a strong upward tendency. This movement was broken Chart 3. — Adjusted Index for the Iron and Steel Group {Ordinates of Trend = "Normal" =100) J cent V \pri \ /> k ^J"^ 110 \ .^-^ 1 W^ ^ 1 V ^ i' \ 100 90 80 70 \ /^ \ 1 \ \j ' \ / \ V V ' (- ^ 60 50 -A > t +0 10 \ / pr ' ^ Jon. ftpp 19 July 9 Qd. Jon ~ "^ ( ra July EO 3ct Jan ^p^ 19 July SI Oct from September to December 1919, by a serious steel strike. During the first eight or nine months of 1920 the trade was unusually active. The slump in production was pronounced by November. It has since become one of the most serious setbacks ever experienced by the industry. Output last July was only about one-third of normal. The moderate in- crease of production since summer is little more than would be indicated at this season of the year, but is encoiu-agkig none the less as evidence that the trade is at last headed toward increased activity. The adjusted index for iron and steel rests upon two important series: (o) monthly production of pig iron; (b) monthly production of steel ingots. The former is by far the most valuable single production series available in monthly form for the United States. The two series together furnish an imusually clear and dependable record of the course of production during recent years. Monthly data on pig iron production are to be had from September 1901 to date,^ The characteristic seasonal variation of the industry has been ascertained from the monthly data, September 1901 to December 1920. Upon the other hand, the normal growth of out- ' See "Current Statistics" for this series for recent months. 6 Table D. — Adjusted Indices for the Iron and Steel Group and por Iron and Steel Manutacture Month Iron and Steel Group 1920 Iron 19 19 1920 Steel 19 19 1920 1921 January . . February . March . . . April May June July August . . , September October . . November December 132.0 121.6 111.9 103.6 85.5 91.4 102.6 106.4 94.8* 66.1* 89.8* 97.9* 121. 6 123.9 130.3 115.6 120.S 121. 6 iiS-3 114.8 112. 7 107.6 994 94-5 89.8 77-7 S8.6 56.1 S3-I 42.2 34-0 37-2 37-0 44.2 126.6 118.9 111.8 91.4 75-4 81.0 92.9 103.1 94.8 66.1 89.8 97-9 112.8 117.6 119. 6 98.9 105.8 113. 1 114.1 iiS-4 115.8 116.5 107.5 98.1 88.5 77-4 52.5 40.8 39-8 39-5 32.7 34-4 36.9 40.7 134.7 122.9 111.9 109.7 90.6 96.6 107.S 108.0 t t t t 126.0 127. 1 135.6 124.0 127.8 125.8 115.9 114.5 111. 2 103.2 95-3 92.7 90.4 77.8 61.6 63.8 59-8 43-5 34-6 38.6 37-1 46.0 * Based on pig iron only. t Data not available. put has been determined from the i i-year pre-war period, 1904-14.^ Both original monthly items and the line of trend are plotted in Chart 4. On the basis of the long-time trend and typical seasonal fluctuation of the trade, a normal output has been calculated for each Chart 4. — Monthly Production of Pig Iron {in millions of long tons) -Mc\C[-i\ ;, vvPf 1 f"fr^ ^^ \ ran^ BIO 1915 1920 month. The figures for iron in Table D express actual production as a percentage of the normal output of the month. Like methods have been followed in analyzing the data for steel ingot production. Monthly data in this case have been reported only since July 191 7; but monthly average 1 Full details regarding methods and results will be given in a later issue of the Review or Economic Statistics. 7 figures are available for a number of years before this. The normal growth of output in the industry has been calculated on the basis of the period 1905-14. The original data and the line of growth are shown in Chart 5. The percentages of actual to normal steel ingot pro- duction from January 1 919 to October 192 1 appear in Table D. Chart 5. — Monthly Production oe Steel Ingots {in millions of long tons) r \ M I \ ^"-^ ^ /: 3, \/ / -' .^1- r 7 ^ - ^ \/ f — -^ v/ V \ 1 Comparison of the percentage relatives of iron and steel in Table D shows that the two are nearly every month much the same. The adjusted index for the iron and steel group is a combination of the two constituents/ each already corrected for trend and seasonal varia- tion. The cycle of the iron and steel trade is by this means presented in the simplest and clearest possible form. ' A weighted arithmetic mean, steel being given a weight of 2, iron a weight of i. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH CAMBRIDGE, MASSACHUSETTS, U.S.A. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH THE VOLUME OF FOREIGN TRADE special Letter — December lo, ig2l THE spectacular expansion of our foreign trade induced by the war has been followed by an equally spectacialar decline. Thoughtful observers would probably agree that during both the rise and the fall there has been a tendency in some quarters to overstate our foreign trade situation and its possibiUties. Our views in the matter were stated in July 1919, when the Review said: "We may venture the opinion that many of the claims and predictions regarding our new trading position are overoptimistic." A year later (April 1920) the fact was emphasized that we must soon expect a marked diminution in both terms of the balance; "the value figures wiU diminish with the deflation of prices." In the present year this diminution has occurred, and upon a striking scale; our combined exports and imports in the first ten months of the year have been about one-half of the total trade in the first ten months of 1920. The decline has been accompanied by a lengthy series of failures, particularly among inexperienced traders who had rushed into foreign trade as though it were a new El Dorado. This, too, is not without abundant precedent in business history and required no great sagacity to predict; every boom has left behind a similar trail of wreckage. Beyond the very general conclusion that our new creditor position will event- ually mean a disappearance of our export balance, it is stiU much too early to attempt to assess, as though it were a completed episode, the net effect of the war and of our new commercial and financial relations with Europe and the neutral countries upon the future magnitude of American foreign trade or its relation to our domestic business problems. An analysis of our merchandise trade for the first three quarters of the present year, may be expected, however, to provide some useful fiuther evidence. In the retrospect, if there be occasion for surprise, it is not that the slump has come this year but that it was so long in coming. Probably the most difficult single fact to explain in American foreign trade id recent years was the pronounced expansion that occurred in the autmnn months of 1920, analyzed in detail in the Review's Supplement for June 192 1. In spite of falling prices and the rapid decline of imports which had begun in September, the monthly average of exports, owing chiefly to a great increase in steel products and wheat, rose from $611,000,000 in the third quarter to $716,000,000 in the final quarter of last year. What has happened since that time is shown by the accompanying table. Com- bined exports and imports in the first nine months of the year feU from $10,438,000,000 in 1920 to $5,425,000,000 in 1921, a decrease of 48 per cent. Exports declined from $6,080,000,- 000 to $3,552,000,000, a fall of 42 per cent, and imports fell off from $4,358,000,000 to $1,873,000,000, a decrease of 57 per cent. The monthly averages for the quarterly periods indicate that imports have been diminishing in value since the spring of 1920, the pro- noimced decline beginning with the final quarter of last year, since when it has progressed virtually without interruption to the present time. The monthly average imports of the COFYSIGET; igai, BY HAEVABD UinVESSIIV Monthly Averages of United States Foreign Trade in 1920 and 1921, tor Quarterly Periods {Units of $1,000,000) Total trade Exports Imports Balance 1920 First quarter I217 II79 1082 1023 73.4 729 686 611 716 509 336 3.39 488 493 471 307 225 215 184 + 241 Second " + 194 + 140 + 409 + 284 + 121 Third " Fourth " 1921 First quarter Second " Third " + IS5 Total trade Jan.-Sept. 1920 Total trade Jan.-Sept. 1921 10438 5425 6080 3SS2 4358 1873 4- 1722 + 1679 third quarter of this year were only $184,000,000, or but 37 per cent of the average monthly imports in the spring quarter of last year. The decline in exports occurred between Decem- ber 1920 and April 1921, exports falling from $720,000,000 in the former month to $340,- 000,000 in the latter. Since last April exports have remained about stationary, varying between $325,000,000 (July) and $367,000,000 (August). In October, in spite of the fact that the harvest exports had begun, exports showed no upward tendency, amounting to but $344,000,000, only 41 per cent of the exports of October 1920. Imports also remained weak, being only $188,000,000. CHANGES IN QUANTITY To determine the real significance of this reduction of our foreign trade to about one- half its former value, it is important to recall that the phenomenal growth which preceded it, culminating in exports of $8,228,000,000 and imports of $5,278,000,000 last year, the record figures in our history, was more in the value figures than in the physical volume of trade. It was chiefly the result of rapidly rising prices. In the fiscal year 1919, for instance, when export prices had increased 127 per cent over the 1911-14 average and valiie of domes- tic exports 217 per cent, the quantity of domestic exports had increased only 40 per cent over the pre-war average. Comparing year by year from the beginning of the war, the quantity exported increased steadily to 191 7, when it was 71 per cent in excess of the 1911-14 average, declined in 1918 to but 25 per cent over the pre-war figure and partly recovered in 1919, when it was 40 per cent greater than before the war. These figures, which are for fiscal years, are indices based on 100 leading exports. Like most existent indices of the physical volume of exports, they are open to the criticism that in view of the violent changes in our export trade since 1 914 no system of averages based, as these are, on pre-war conditions would be representative of post-war years. Therefore a new index based on 175 selected commodities, representing three-fourths of our exports in 1919, has been constructed to show the quantity of exports in the calendar year 1920 compared with 1919. This index is based upon the ratios of quantities of the 175 selected commodities exported in 1920 to the quantities exported in 1919. The computation shows that the physical voliune of exports in 1920 was 94.1 per cent of the volume exported in 1919. In other words, there was a faU of 5.9 per cent in the quantity of exports last year compared with the preceding year. An index, similarly constructed, but based on 50 selected com- modities comprising 63 per cent of the exports in 1920,^ shows that the physical volume of exports in the first three quarters of this year was 94.0 per cent of the quantity exported in the same period last year.^ A similar index for physical volume of imports has not yet been computed, since this is a task of somewhat greater difl&culty by reason of the greater nximber and diversity of prod- ucts. The foreign trade index of the Federal Reserve Board, however, places physical quantity of imports in 1919 at 171. i per cent of the 1913 figure, and the quantity imported in 1920 at 171. 7 per cent, indicating that last year there was virtually no change from 1919. For the first nine months of 192 1 its index shows a fall of about 32 per cent in volume of imports compared with the same period of 1920. From this index it appears that the in- crease in physical volume of imports down to last year was even greater than that in ex- ports, though the increase in the value figures was much less than for exports; and that in the present year the import trade has fallen off much more sharply, both in value and in quantity, than has the export trade. In the absence, however, of an index of physical volume of imports constructed by the methods used in our new index of physical volume of exports, this conclusion must be taken as tentative.^ Our indices of physical volume of exports make clear a most interesting fact. In voliune exported, the decline this year has apparently been no greater than it was in 1920. In both 1920 and 1921 it amounted to about 6 per cent. But in 1920 the fall was covered up by a 4 per cent increase in the total value of exports, from $7,920,ooo,ooo-«in 1919 to $8,228,000,000 last year, whereas in the present year the decline in voliune has been accompanied by a fall of 42 per cent in value. In other words, the pronounced decline of our export trade this year has been for the most part but the expression of falling prices. According to the price indices for "exported goods" and "imported goods" prepared by the Federal Reserve Board, though both sets of prices turned downward after April 1920, export prices were 6 per cent and import prices 10 per cent higher in 1920 than the averages for the year 1919. In the first nine months of the present year, on the other hand, the monthly average of ' The 50 commodities were 63.1 per cent of total exports in the first nine months of 1920, and 64.7 per cent of the total in the first nine months of 1921. 2 The formula for the index is as follows: S ^1920 91921 This is the Fisher formula. In the absence of price data, it was necessary to work with values and quantities. The for- mula was therefore thrown into the foUowing form. 91921 i "1920 ~ ■'^i = ~^y~;, = 98.1 I, = ^'"^' = 90.0 91920 2 111921 ;; — 91921 I = \l h h = 94-0 ' The Reserve Board's index of physical volume of imports includes 25 of the most important imports, the value of which in 1913 formed 47.7 per cent of the total import values. Being computed on the year 1913 as base and with the import values of 1913 as weights, it is subject to the cautions mentioned above in discussing quantity indices of exports, as also to the criti- cism that it is not very comprehensive. export prices fell 48 per cent compared with the same period of last year, and the monthly average of import prices fell 49 per cent. In the same period, according to the Reserve Board's index, general prices have fallen 41 per cent below the average for the first nine months of last year. Apparently both on the rise and on the fall the variation in export and import prices has been more pronounced than that in general prices. Export and import prices rose faster and higher, and since their turn in April 1920 have fallen further and more sharply. This fact would tend, of course, to accentuate the decUne in value of exports and imports. CHANGES BY GROUPS AND PRINCIPAL COMMODITIES Both in value and in quantity the changes in the great groups of exports and imports this year have been more striking than the changes in the total trade. In exports the losses have been heaviest in raw materials and in manufactures, the former faUing 51 per cent and the latter 47 per cent below their value in the first nine months of 1920. Total food prod- ucts, on the other hand, have fallen but 24 per cent, and crude foodstuffs have declined only 3 per cent. In imports the decline has been heavy in all three groups, but has been most severe in food products, which fell 6(5 per cent, while raw materials declined 58 per cent and manufactiu-es 46 per cent. The result has been to bring about a marked alteration, both in exports and imports, in the relative importance of the three groups of products Exports by Groups, January-September Inclusive, 1920 and 1921 Absoltjte Figures Percentage of Totai. Groups ig2o I92I 1920 1921 Decrease in 1921 Raw materials millions $1397 1489 3069 millions $684 1136 1637 per cent 23.0 24.5 50.4 percent 19.3 32.0 46.1 millions $713 353 1432 percent 23.7 46.7 Food products Manufactures Total exports * $6080 $3552 100 100 $2528 41.6 Including "miscellaneous" and "foreign" exports. Imports by Groups, January-September Inclusive, 1920 and 1921 Absolute Figures Percentage of Total Groups 1920 1921 1920 igsi Decrease in 1921 Raw materials millions $1515 1487 I331 • millions $630 512 717 per cent 34.8 34.1 30.5 percent 33.6 27.3 38.3 millions $885 975 614 percent 58.4 65.6 46.1 Food products Manufactures Total imnorts * I4358 $1873 100 ICO $2485 S7.0 Including "miscellaneous" imports. 4 which comprise international trade. Iii exports, manufactures and raw materials have declined in importance, and food products, which last year were one-fourth of total exports, are this year one-third of the total. In imports the reaUgnment of the groups has been even more pronovmced, manufactures gaining markedly in relative importance at the expense of food products. Raw materials have continued to constitute about a third of the total. It is interesting to note, moreover, that all of these changes run directly counter to the main tendencies of the pre-war period, which were, for exports, the marked increase (since about 1900) of exports of manufactures and the decline in relative importance of food products, and for imports the increase in food products and raw materials at the expense of manufac- tures. The evidence from the war and early post-war periods was that these tendencies were in no respect fundamentally altered but for the most part intensified; the present year, therefore, represents the first important upset of the preexisting tendencies in the character of our trade. Whether this interruption possesses any lasting significance or is merely a vagary of the present depression period it is too early to venture any opinion. As was stated in the Supplement of last June, however, it may well be that our new creditor posi- tion, entailing as it must an increased inflow of goods from Europe as a means of debt pay- ment, win check for a time the pre-war tendency of manufactured imports to decline in importance. The very marked reduction in food imports is doubtless for the most part the reaction from the heavy imports last year, at greatly inflated prices, of Cuban sugar. In the first nine months of 1920 imports of sugar alone amoimted to $887,000,000, or over one- fifth of the total imports; in the fiorst nine months of this year sugar imports amounted to but $204,000,000, or II per cent of the total imports. While sugar imports have thus fallen 77 per cent in value, their quantity has diminished 32 per cent.^ The group changes in exports are best explained by reference to the detailed table, on the final page of this Letter, of quantity and value changes in the principal commodities exported in the first nine months of this year. Of the thirteen important exports which increased in quantity, compared with the same period last year, six were food products and five, raw materials; only two were manufactured products. The increases in crude food- stuffs were phenomenal, in some cases surpassing all recent records. Corn exports in- creased 813 per cent in quantity and 300 per cent in value, notwithstanding the fact that the price dropped from 96 cents a bushel in January to 60 cents a bushel in September. Exports of wheat increased 82 per cent in quantity and 4 per cent in value, though the price dropped from $2.13 per bushel in January to $1.34 in September. Lard increased in quan- tity 69 per cent, barley 82 per cent, rice 66 per cent, and hams and shoulders 25 percent. These figures indicate sufficiently why, while the other great groups were reduced in value about one-half, total food products, in spite of heavily falling prices, diminished in value but 24 per cent, and crude foodstuffs but 3 per cent. On the other hand, of the twenty lead- ing exports which have decreased in quantity this year all but four were manufactures or raw materials. Especially marked are the declines in steel products, automobiles, oil prod- ucts, coal, and cotton textiles. As pointed out in the last Special Letter, the slump in iron and steel production which became pronounced in the autimm of 1920 has "this year be- come one of the most serious setbacks ever experienced by the industry. Output last July was only about one-third of normal." This decline has been markedly reflected in the ' From 6 970,000,000 lbs. in the first nine months of 1920 to 4,748,000,000 lbs. in the same period of 1921. S export trade, the steel group showing declines in quantity in the first nine months of this year which ranged from i8 per cent for structural steel to 65 per cent for bar iron. THE BALANCE OF TRADE While the value figures of trade have declined so sharply this year it is a significant fact that the trade balance has shown no essential change. For the first nine months of 1920 the excess of exports was $1,722,000,000, and for the same period of this year it was $1,679,- 000,000. Indeed on a relative basis the trade balance has increased strikingly in our favor. In the first nine months of last year the balance was 1 7 per cent of the combined exports and imports; in the same period this year it was 31 per cent of the total trade. There is as yet therefore no apparent tendency for our export balance to disappear as the logical eventual outcome of our new creditor position. It seems clearer than ever that, as stated in previous studies in the Review, there wiU be no radical disturbance of the trade balance imtil Europe begins the payment of the interest charges, now amounting to about $522,000,000 a year, upon our government's credits advanced during the war and the early post-war period. What arrangement of these payments wiU be made by the present administration, and what relation the whole subject will bear to tariff policy, will be watched with interest during the new session of Congress, which is expected to consider both the permanent tariff bill and that on debt refunding. The accompanjdng table shows the changes in exports and imports and the balance of trade by continents in the first nine months of 192 1. In the trade with Europe the relative The Trade Balance by Continents, January-September Inclusive, 1920 and 1921 Countries Exports 1Q2I Imports Balance Europe North America Asia South America. Oceania Africa Totals . . millions $3297 1418 629 432 187 117 millions $1861 919 355 228 130 60 millions $991 I37I 1064 638 152 143 millions I555 593 407 224 67 26 millions +$2306 + 47 - 435 — 206 35 26 millions +$1306 + + + + + 326 52 4 63 34 $6080 $3552 14358 I1873 +$1722 +$1679 Countries Decune of Exports millions per cent Declimie of Imports Ratio of Trade Balance 10 Total Trade millions per cent 1920 ZQ2Z Europe North America Asia South America. Oceania Africa Totals ■. . I1436 499 274 204 57 57 43-5 35-2 43-5 47-2 30.S 48.7 $436 778 657 414 85 117 44.0 56.8 61.7 64.9 56.0 81.8 53.8 1-7 -25.7 -19.2 10.3 -lO.O S4.0 21.6 -6.8 0.9 32.0 39.S $2527 41.6 $2485 S7.0 16.S 30.9 declines in exports and imports compared with the same period of last year have been about equal, being 44 per cent in each case, and the ratio of the trade balance to the total trade has remained unchanged, though in absolute figures the decline has been ahnost one-half, from $2,306,000,000 last year to $1,306,000,000. In the trade with North America imports have diminished much more markedly than exports, owing particularly to the decHne in imports of sugar from Cuba; as a result our excess of exports has increased from $47,000,000 to $326,000,000. With Asia and Latin America also, our imports have decreased much more markedly than our exports, reducing our adverse balance with Asia from $435,000,000 last year to $52,000,000 this year, and giving us for the first year since before the war a favor- able balance in the Latin-American trade. From this brief summary it is evident that the relatively large trade balance this year has been due mainly to the reduction of our imports from the non-European countries; except for price deflation, affecting about equally our European exports and imports, our balance with Europe this year has not exerted much effect upon the total trade balance. RECENT TENDENCIES So far as the value figures of trade are concerned, there is little evidence that we have yet "turned the comer" in foreign trade. There is no evidence from the data that the decline of imports has terminated. The fact that exports have shown no further tendency to decline since last spring may indicate that the downward movement has come definitely to an end, though the failure of exports to show any expansion in September and October, despite the appearance of the new harvest, leaves the question in considerable doubt. The outstanding conclusion from the present analysis, however, is that the value figures of trade are a most imcertain indication of what is reaUy happening, since they are dominated by the change of prices. As already stated, the dechne in physical volmne of exports this year has been but 6 per cent, a decline no greater than that of last year. Relating these changes to those given in a previous section for the war and the early post-war period, the quantity of exports in the first nine months of this year would stiU be about one-fifth greater than the pre-war average (1911-14). The most satisfactorj' indication of recent tendencies in ex- ports wovild be given by a monthly index of physical volume exported. According to the foreign trade index of the Federal Reserve Board, the physical quantity of exports reached its lowest point in April 1921, when it was about 10 per cent below the average for 1913. From April to August there was a strong upward tendency, bringing the index for August to 41 per cent above the 1913 average. This rise was due chiefly to the spectacular increase during the summer months in lard, wheat, corn, and other food products. In September, however, the latest month for which the index has been computed, there was another de- cline, of about 15 per cent. In imports, the tendency has been downward throughout the year. From these indices it appears probable that we have passed the low point of exports, though there can be no marked advance untfl manufactured exports, and particiilarly steel products, follow food products and raw materials upward. In imports there is no evidence, in either the value or the quantity figures, that the low point has yet been reached. Quantities and Values or the Principal Exports, January-September Inclusive, 1920 AND 1921 Exports which Increased in Quantity in ip2i Corn Wheat Lard Cotton, raw Hams and shoulders .... Tobacco leaf Cotton seed oil Pipes and fittings Barley Rice Fuel and gas oUs Crude oil, mineral Cotton cloths, unbleached Values Quantities Ratio of ig2o 1921 Unit 1920 1921 quantities 1921/1920 millions millions millions percent $19 $78 bu. "•5 105.0 913 368 381 bu. 130-5 237.4 182 99 93 lbs. 410.7 695.7 169 862 317 lbs. 2101.8 2226.2 106 41 40 lbs. 150.8 187.9 125 192 168 lbs. 356.2 403.9 113 25 22 lbs. 1 13.0 220.3 195 35 52 ^ lbs. 530.8 785-1 148 18 17 bu. II. 2 20.4 182 32 17 lbs. 294.1 489.4 166 37 39 gals. 587.6 665.7 113 21 17 gals. 250.8 282.7 "3 34 16 yds. 125.3 1 60. 1 128 Exports which Decreased in Quantity in ig2i Bacon Wheat flour Rye Sugar, refined Boards, deals, etc Cotton cloths, colored . . Cotton cloths, bleached Coal, bituminous Copper, ingots Oil, lubricating OU, illuminating Naphthas Locomotives, steam .... Automobiles, commercial Steel, sheets and plates . Iron, other bars Steel raUs Tin plates Structural steel Steel wire, not barbed . . , Ins 183 95 86 76 108 43 197 92 108 95 126 42 35 79 39 25. 30 25 16 95 40 43 32 28 9 107 56 70 73 104 25 9 41 13 17 15 24 lbs. bbls. bu. lbs. Mft. yds. yds. tons lbs. gals. gals. gals. No* No* lbs. lbs. tons lbs. tons lbs. 460.1 16.2 43-9 851-3 1.22 370.5 153.7 23.6 460.S 293.4 621.9 470.4 1.30 22.11 1,923.8 1,028.6 .421 380.2 ■33 283-7 354.7 13.0 25.2 786.0 .82 172.2 61.4 17.5 417.6 195-4 532-8 403.5 .79 5-94 1,031.9 362.9 .282 189.7 .27 130.6 77 80 57 92 67 46 40 74 91 67 86 86 61 27 54 35 67 SO 82 46 Unit 1,000. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH CAMBRIDGE, MASSACHUSETTS, U.S.A. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH THE VOLUME OF MANUFACTURE INDEX FOR ALL GROUPS COMBINED ADJUSTED INDICES FOR IRON AND STEEL, TEXTILES, AND PAPER Special Letter — December 24, ig2l ALL GROUPS COMBINED MANUFACTURE is no exception to the rule that business conditions after depression tend to become spotty. Liquidation and price readjustment are more quickly efifected in some trades than in others and the possibilities of early recovery vary widely. For these rfeasons it is difficult to generalize about such periods as the past four months. That there has been substantial improvement in some industries is obvious. It is just as Chart i. — Index of the Volume of Manufacture (ip/p Monthly Average = 100) 120 115 no 105 100 95 90 85 60 75 7 4- 3C -1 h Id i8 }- \ f] i /> k ^ p i V N I 1 1 f^ %^ -s ^ J V ^ ^ \ /» J f \ 1 f / ] \ V \ 1 \ \ 4 r J V \ si ^ ^ ^ V i\ > J )U «.'5 _- y 1234-56 76 9 10 1 1 12 I 23436789 10 II 12 I 23456789 10 II )2 1919 1980 1981 clear, however, that other lines have merely held their own, or, at most, have shown an increase of activity no more than seasonal. Still others have receded somewhat. The general tendency is not easily ascertained. Upon the whole, however, it would appear that the volume of manufacture has shown a real increase during recent months. COPYMGHT, I921, BY HARVAED UNIVERSITY Table A. — Monthly Indices of the Volume of Manufacture for Eight Groups of Manufacturing Industries and for the Groups Combined {,1919 Monthly Average = 100) Month AH groups Iron and steel Lumber Paper Petroleum Textiles Leather Food Tobacco 1919 January. . . February . March . . . . April May June July August . . . September October . . . November December. 1920 January . . February , March . . . . April May June July August . . . September October . . , November December. 1921 January . . February , March . . . . April May June July August . . . September October . . . November. (I) 101.6 86.7 92.6 93-7 95-7 95-9 101.9 107.2 103.8 104.4 102.2 102. 1 "S-9 104.6 118.0 108.8 111.8 109.6 102.3 104.9 101.4 101.2 88.9 77-9 78-3 7S-0 80.6 75-8 79-3 7S-9 68.5 81.5 81.0 90.6 * W 130.1 II3-7 "3-S 94.6 81.2 90.9 103. 1 "3-5 97.6 73-2 94.0 103.4 124.2 120.5 138.3 iii.o 121.2 124.9 118.0 125.3 125.0 126.7 111.3 99-5 92.7 73-7 65.2 50-5 52.4 42.3 34-0 46.0 47-4 64.0 67.0 (3) 75-5 75-8 88.8 97.8 112.1 103-5 103.7 123.4 113.8 123.4 102.0 80.1 lOI.I 101.3 117.4 122.3 127.3 115.4 103.1 117.8 108.7 108.1 86.2 62.8 59.5 65.3 78.2 84.6 100.7 92.8 86.8 101.5 89.8 (4) 93 -o 82.8 87.0 88.4 9S.I 98.3 103.3 108.8 106.2 116.4 108.6 109.8 120.9 105.8 119.5 121. 7 119.2 122.5 123.0 122.5 1 1 9.3 119.7 108.0 94.6 84.3 78.5 83.2 79-3 73.6 75-6 71.0 84.4 88.2 98.0 (s) 89.7 83.8 92.8 92.4 100.7 96.1 103.7 107.4 108.3 112.0 107.0 107.6 102.3 97.0 III. 6 109.3 115.0 "5.9 122.9 132.2 134.6 135.2 131.2 134.6 131.6 115.0 117.9 124.9 122.8 122.7 119.7 119.7 118.2 (6) 96.5 72.7 79.0 92.9 99.2 100.2 109.4 103.4 105.3 1 19.3 104.9 IIO.l 126.6 110.5 121.0 "9-3 110.5 104.1 92.8 88.2 83.4 78.7 62.1 54-5 68.1 76.3 89.3 89.3 96.2 100.8 89.5 100.7 105.7 no. 6 (7) 109.4 91.0 92.8 108.7 106.3 108.6 99.8 102.6 102.0 101.4 87.4 89.9 90.8 81.6 94.0 84.7 90.9 95.2 80.7 70.5 73.3 77.8 70.1 72.1 63.5 62.8 72.0 75-8 83.2 81. T 763 85-7 80.3 (8) 114.9 87.3 92.1 92.8 98.5 83.8 85.9 88.4 97.7 III.O 109.0 108.6 112.2 85-3 90.8 75-9 85.1 80.8 80.8 83.4 78.0 81.1 91.0 88.4 84.1 74.4 85.6 82.4 79.9 82.7 84.S 96.7 91.6 97.9 (9) 87.3 83.8 97-7 82.0 88.0 92.9 99.3 101. 1 108.6 125.4 117.0 113.0 115.7 98.9 123. 1 107.8 111. 1 114.6 98.1 104.8 105.0 106.3 93-1 73-5 88.7 95.0 106.6 96.6 100.6 108.0 101.8 118.3 113.1 117. 2 * Subject to revision. The general movement of industrial activity is best shown by the Index of the Volume of Manufacture, appearing above. The Index touched bottom (68.5) last July. It rose in August to 81.5, reflecting a decided increase in industrial output. In September there was no appreciable change, but in October a further rise to about 90.6 was recorded. Current reports would indicate no material advance since October. But it is significant that there has been no serious retrogression. Ground once gained has been held. And though there are in some lines factors which may cause hesitation, and possibly moderate relapses, the forces making for increased production appear to be continually gathering strength. At this period of the year a sUght decUne in industrial output wiU not be surprising. But at bottom conditions seem to be improving, and manufacturing enterprise should be in posi- tion to take full advantage of the favorable impulses of the spring. Since, as already stated, the situation during recent months has been spotty, it is de- sirable to examine with care conditions in the several groups of industries, and to consider especially whether any changes observed are more than seasonal movements characteristic of the several trades during the faU months. The iron and steel industry may well be examined first. Activity in this industry is shown in the adjusted index appearing in Chart 2 and Table B. In the construction of the index full allowance has been made for the nor- mal e^ansion and customary seasonal variations of the trade; the cyclical movement is Chast 2. — Adjusted Index for the Iron and Steel Group (Ordinates of Trend = "Normal" = loo) T a 9 10 II 1921 thrown into bold relief. The index shows clearly that the increase in iron and steel produc- tion since July has been more than seasonal. In July the industry was operating at about one-third its normal rate for that month; in October it was operating at a rate about 45 per cent of normal for October. November brought a further advance which raised output to sHghtly more than half of normal. While new orders have been falling off somewhat and many concerns appear to be waiting until after the opening of the new year before making further commitments, there are encouraging evidences of a larger demand for heavy steel for raikoad and construction requirements. Should such a demand develop in the early future, the industry may make a quick recovery, for the demand for light steel is already upon a level which is satisfactory, considering general business conditions. 3 The textiles are another important group for which an adjusted index is at hand. The index is given graphically in Chart 3, and numerically in Table B. The index shows clearly the improvement which has taken place in textile manufacture since last winter. Further- more, it is obvious that the increase in output since mid-summer has been much more than a seasonal movement. The index for October, however, discloses no general improvement over September, an increase in wool manufacture being offset by a decline in cotton manu- facture. In November, a decided increase in the mill consmnption of raw cotton seems to have taken place.^ Figures for November i from practically aU the textile trades except Chart 3. — Adjusted Index for the Textile Group {Ordinates of Trend = "Normal" = 100) MU 1^0 120 110 100 K K ^ J ^ ^ / \ VJ ^ « ^ a i ( > L, / 90 ao 70 \ J / P •V V ^ V ^ Si f ] [ / / \ V ^ J ^ V ^/ r > ^ ( ^ 60 fiO \ \. ) '-•\ V -J I 23456769 10 11 12 123456789 10 II 13 123436789 10 II IS 1919 1920 1931 sUk suggest increased manufacturing activity — most decided in the production of car- pets and rugs. Textile manufacturing as a whole seems to be more than holding its own. Recently it has shown a volume well above the average indicated by pre-war tendencies. Conditions in the paper trade continue to improve. In the final section of this letter an adjusted index of the volume of paper manufactmre is developed. The increase in output since mid-summer has been notable. Part of this increase has been seasonal. But when due allowance has been made for seasonal influences — as is done in constructing the ad- justed index — a decided increase remains. In July, the low month of the depression, the index stood at 63.4. By October, it had risen to 80.6. As a whole, the paper trade appears to expect an early recovery of normal activity. Recent weeks have witnessed some imcertainty about output in the group of industries tunaing out food — wheat flour milling, meat slaughtering, and sugar refining. The unad- justed index for the group (see column 8 of Table A) shows output from August to October distinctly above the level of the earUer months of the year, and not far below the monthly average of 1919. The question is whether this recent increase has been more than seasonal. An adjusted index, provisionally calculated,^ would indicate that output was relatively ' Because of a decline in orders this has resulted in an accumulation of stock at the mills. ' The index will be presented in full next month. much larger in August than in any month since. Production in October, when seasonal variation is taken into account, seems to have been less than in any other month since May. July, August, and September were upon a somewhat higher level than the first six months of the year; October, while showing a volume distinctly above that of October, 1920, fell distinctly below the output of the late summer and early fall. Changes in the other groups are not so noteworthy. While complete data are not yet in hand for October, it is probable that the group index for lumber will show a moderate rise, for petroleum products, a slight fall. Leather production was somewhat greater in Octo- ber than in September though the increase evidences no material change in the con- dition of this trade. Tobacco manufacture in October almost reached the high level of August, and stood nearly one-fifth above the 1919 monthly average. These four groups are similar in that no one of the four showed any increase of output between August and October. Had not other more important groups — previously discussed — showed a de- cided gain during the fall, it would be difi&cult indeed to offer any favorable characterization of recent tendencies. Even as it is, any such characterization must be used with appro- priate recognition of the contrasts which feature the present industrial situation. Table B. — Adjusted Indices for the Iron and Steel, Textile, and Paper GROtrps * (Ordinates of Trend = 100) Month Iron amd Steel Group Tejchle Group Paper Grotip 1920 January . . February . March . . . April May June July August . . , September October . . November December 132.0 121.6 III. 9 103.6 8S-S 91.4 102.6 106.4 94.8 t 66.1 1 89.8 t 97-9 t 121.6 123.9 130-3 115.6 120.S 121. 6 iiS-3 114.8 112. 7 107.6 99-4 94-5 89.8 77-7 S8.6 S6.1 S3-I 42.2 34.0 37-2 37-0 44-3 53-6 96.4 76.6 71.9 85.1 88.1 97-4 IIO.I III. I II3-9 123.6 III.O 116. 1 124.6 112. 6 112.4 1 10.8 98.2 99.6 91.6 9S-I 90.4 80.9 66.8 59-1 64.9 76.S 79.6 80.4 83.8 96.2 87.9 107.0 112. 1 112.9 98.2 96.9 91.9 91.9 9S-S 96.6 100. 1 101.6 103.2 105.6 106.2 116.7 121.9 116. 2 120.1 120.7 iiS-3 116. 2 iiS-3 1 10.6 111.9 104.5 102.4 99-S 84.3 87.1 82.2 78.4 69.2 68.7 63-4 71.7 79-4 80.6 * See Table C for the constituent series. t Based on pig iron only. THE PAPER GROUP The adjusted index — shown in Chart 4 and Table B — exhibits clearly the cycle through which the paper industry has passed since January 1919. Typical seasonal move- ments and the normal growth of the industry have been eliminated. Comparison of the index with the corresponding indices for the iron and steel and textile groups discloses a number of highly significant points. There is striking similarity in the general character of the curves for paper and iron and steel. In both these trades, the post-armistice slump was followed by a rapid upward movement culminating in early 1920; in both there was no S precipitate decline until the final quarter of 1920; in both, the depths of depression were not experienced until last summer. Upon the other hand, paper differs strikingly from iron and steel in the violence of the fluctuations of output. Relatively, paper production rose only about two-thirds as high in 1920 and fell only half as low in 192 1. In this respect, paper has behaved much more like the textiles than like iron and steel. The paper index rests upon adjusted relatives of five series giving the monthly produc- tion of the major varieties of paper: newsprint, book paper, fine paper, wrapping paper, and paper board. The series have been published by the Federal Trade Commission for the months from January 1918 to date. Earlier monthly data for newsprint are to be found in the files of the Paper Trade Journal. These have been used, as weU as data from the Chart 4. — Adjusted Index for the Paper Group (Ordinaies of Trend = "Normal" = 100) 130 120 110 100 A ■% i f V f )w ■% < ** ^ J ^ s. s * N s. ^ /' t^ ^ ^ \ SO 70 60 \ ^ \ ^ \ \, / f y s/ f \.0^ "" 183456769 10 II 12 123456769 10 II 12 I 234567S9I0 1II2 1919 1920 1921 Federal Censuses of Manufactures for 1899, 1904, 1909, and 1914, in determining the differ- ent trends of output.^ The original series and lines of trend appear in the several diagrams of Chart 5.^ Examination of Chart 5 demonstrates beyond question that the five paper series are of the same family. It is clear, however, from the plots for 1918, that the several members of the fanuly were affected differently by the exigencies of the war. And the different lines have reboimded differently from the depression of last summer. Wrapping paper and paper board — the courser papers — have made greater relative gains than book and fine paper. Newsprint, as might be expected, has shown throughout the period less extreme fluctuations than the other members of the group. Despite the differences noted, however, the five series combined afford an unusually reliable record of the 'course of manufacturing output in the paper trade as a whole from January 1919 to date. ' Full details of the methods employed will be presented in a later issue of the Review or Economic Statistics. 2 Wood pulp also is shown though it is not included in the index. Adjusted relatives of the five paper series are given in Table C. Chakt s- — Series Indicative of the Volume of Paper Manufacture 400 360 320 260 240 200 160 120 60 40 WOOD PULP Uoft .- 1000 short tons ; air dry basis A t««^ -P J.. ^ _« . sf u^ J f w f \ / \ /* I9ia 1919 1920 1921 150 135 120 103 90 73 60 45 30 15 NEWSPRINT Unit: looo short tons A^ y^ V' i_ :7^ Ay. w ^ V— W y V T ' I r V 1918 1919 1920 1921 100 90 80 70 60 50 40 30 20 10 BOOK PAPER Unit-.iooo net tons ii^ ■^ ^ V 4 .- «— ^ ^ Ai- r"/^ p.- ,— — : —I y % \J L / V / FINE PAPER Unit: looo net tons 1918 1919 1920 1921 1918 1919 1920 l©21 60 72 64 56 48 40 32 S4 16 8 WRAPPING PAPER Unit: 1000 net tons k* Ajl Kk ifJU T ^ / v\ .^J. r'fH ,1— . -A — — t' T \ J \ ./> / \r W V 1918 1919 I9E0 1921 250 225 200 175 ISO 125 100 7E5 50 25 PAPER BOATTO Unit: 1000 net tons -mm% 1916 1919 1920 1921 Table C. — Adjusted Relatives of Series Indicative of the Volume of Manufactuke * (Ordinates of Trend = loo) Iron and Steel Textiles Papeh Month Pig iron Steel ingots Raw cotton Raw wool Newsprint Book paper Fine paper Wrapping produced Paper board produced produced consumed consumed produced produced produced produced 1919 January . . . 126.6 134.7 107.8 77.2 97-7 93-0 104.I 100.6 98.4 February . . I18.9 122.9 91.6 51.4 99.0 94.8 104.2 93.9 95-9 March III.8 111.9 81.6 54.9 98.7 86.4 94.7 89.7 91.1 April 91.4 109.7 93.1 71.7 98.7 88.9 94.3 86.0 93-7 May 75-4 90.6 93.0 79.8 90.2 99.4 93-7 98.8 93.8 June 81.0 96.6 94.2 103. 1 97.8 95.9 93.2 99.4 92.9 July 92.9 107.5 103.2 12I.5 97-5 100.9 103.5 100.7 99.1 August 103.1 108.0 104.3 123.2 98.1 101.8 103.8 lOO.O 109. 1 September . 94.8 t 106.4 127.0 102.7 102.5 103.6 102.0 107.7 October . . . 66.1 t III. 9 142.6 105.9 106.7 103.6 lOI.I 113.1 November . 89.8 t 105.3 120.8 102.6 107.0 106.5 105.4 113.2 December . 97-9 t 109.2 I27.S 107. 1 121. 1 1 14.0 I2I.O 121. 1 1920 January . . . 112.8 126.0 112.8 144.2 108.1 123.2 116.8 13 1. 133.8 February . . 117.6 127. 1 106.6 I2I.8 107.2 121.2 115.3 I18.5 120.3 March 1 19.6 I3S-6 108.7 I18.5 108.8 124.6 124.8 I2I.3 126.7 April 98.9 124.0 109.6 II2.5 107.7 121.0 127.S 129. 1 122.4 May 105.8 127.8 101.9 91.5 109.2 116.3 III. 2 II9.I 122.3 Jime 113.1 125.8 108.6 84.4 109.7 122.0 111.6 1 1 6.6 121.4 July 114.1 115.9 104.4 71.2 IIO.I 122.6 111.3 113.4 1 19.0 August iiS-4 114.S 99-7 87.2 IIO.I 112.8 97.7 113.7 114.2 September . iiS-8 III. 2 98.2 77-9 109.7 "5.3 105.1 III.O 117.8 October . . . 116.S 103.2 79.0 84.0 104.2 X08.4 96.2 107.S 99-7 November . I07-S 95-3 72.1 58.0 106.8 110.2 96.3 106.2 76.6 December . 98.1 92.7 64.4 50.7 107.6 101.4 95.9 104.4 73.7 192 1 January . . . 88.5 90.4 66.1 62.6 101.7 79.1 77.6 85.3 64.7 February . . 77-4 77.8 81.2 68.9 96.S 81.8 77.6 90.3 82.5 March 52.5 61.6 79-5 79.4 90.0 75.9 71.4 87.4 79-4 April 40.8 63.8 76.8 86.3 95.4 62.9 63.9 86.6 74.1 May 39-8 59-8 80.1 89.6 65.0 62.9 60.3 87.0 63.2 June 39-5 43-5 88.5 109.6 72.1 67.6 52.5 76.0 65.7 July 32-7 34-6 80.0 101.3 78.4 61.2 48.8 63.7 S2.I August 34-4 38.6 9S.O 126.8 86.3 66.6 46.S 79.0 63.5 September . 36.9 37.1 101.8 129.4 89.7 72.9 56.1 88.3 77.0 October . . . 41.0 46.0 96.2 140.6 83.S 78.2 59.1 90.0 83.0 November . 51.0 54.9 108.9 For original items of these series, see "Current Statistics." t Data not available. HARVARD UNIVERSITY COMMITTEE ON ECONOMIC RESEARCH CAMBRIDGE, MASSACHUSETTS, U.S.A.