INTERNATIONAL LABOUR OFFICE STUDIES AND REPORTS Series D (Wages and Hours) No. 15 THE WORKERS STANDARD OF LIFE IN COUNTRIES WITH DEPRECIATED CURRENCY GENEVA 1925 CONTENTS Page PREFACE ix PART I. INTRODUCTION. GERMANY General Economic Conditions 1 CHAPTER I. Real Income of the Working Classes Data Available and their Interpretation Real Wages of Industrial Workers, 1920 to 1924 Estimate of the Aggregate Income of the Industrial Population, 1921 to 1924 Position of Professional Workers, 1920 to 1924 10 10 13 19 27 CHAPTER II. Cost of Living Problems Price Control Rent Control Measurement of the Cost of Living 32 32 33 37 CHAPTER III. Adjustment of Wages by Collective Agreement. . . Limitation of Period of Collective Agreements and Introduction of Greater Elasticity Procedure of Adjustment: Revision of Wage Rates, Cost-of-Living Bonuses, Family Allowances, Local Cost-of-Living Allowances Cost-of-Living Bonuses Family Allowances Local Cost-of-Living Allowances Causes of the Fall in Real Wages ; Technical Difficulties of Adjustment and the Reduction of National Income 44 44 47 48 50 53 54 CHAPTER IV. Attempts at Automatic or Semi-Automatic Adjustment of Wages (Sliding Scales and Stable-Value Wages) Sliding Scales (Automatic Adjustment) Stable-Value Wages (Semi-Automatic Adjustment) 60 60 65 CHAPTER V. 73 The Gold Wage and the Stabilisation of Wages. . . . PART II. AUSTRIA INTRODUCTION. General Economic Conditions CHAPTER I. Real Income of the Working Classes Data Available and Their Interpretation Real Wages of Certain Classes of Workers, 1920 to 1924 Position of Professional Workers . 81 84 84 87 92 VI Page CHAPTER II. Measures to Check the Fall in Real Wages Cost-of-Living Problems Price and Rent Control and Food Subsidies Measurement of the Cost of Living Methods of Adjusting Wages to the Cost of Living The Sliding Scale. .• Cost-of-Living Bonuses and Family Allowances PART III. INTRODUCTION. 95 95 95 97 98 99 105 POLAND General Economic Conditions 109 CHAPTER I. Real Income of the Working Classes Data Available and Their Interpretation Real Wages of Certain Classes of Workers, 1921 to 1924 Effects of Unemployment on Working-Class Income Position of Professional Workers 112 112 114 118 119 CHAPTER II. Measures to Check the Fall in Real Wages Cost-of-Living Problems Price and Rent Control Measurement of the Cost of Living Methods of Adjusting Wages to the Cost of Living The Sliding Scale Cost-of-Living Bonuses and Family Allowances The Gold Wage 121 121 121 122 123 123 127 128 CONCLUSION 131 APPENDICES I. Monetary Inflation and Price Changes on Various Markets in Germany II. Methods of Constructing Index Numbers of Real Wages of Industrial Workers and Real Salaries of Professional Workers in Germany, 1920 to 1924. Nominal Wages and Salaries in Marks of these Workers III. Method of Estimating the Percentage Reduction through Unemployment in the Aggregate Income of the Industrial Population from November 1923 to September 1924 in Germany IV. Note on Data concerning the Period of Validity of Collective Agreements in Germany V. Application of the Sliding Scale at Flensburg (Germany) . . . VI. Methods of Forecasting the Rise in the Cost of Living in Germany 137 143 152 157 159 161 VII Page VII. VIII. IX. X. XI. Suggestions of the German Ministry of Labour as to the Possi• bilities of Stabilising the Purchasing Power of the Earnings of Labour (19 July 1923). Suggestions of the Central Industrial Organisations as to the Fixing of Wages, dated I September 1923 Currency Inflation, Price Changes, and Unemployment in Austria Nominal Weekly Wage Rates of Skilled and Unskilled Workers in Certain Important Industries in the Vienna District and Nominal Monthly Salaries of Various Classes of State Officials at Vienna, 1914 to 1924 Currency Inflation, Price Changes, and Unemployment in Poland Methods of Constructing Index Numbers of Real Wages of Certain Classes of Industrial Workers and of Real Salaries of Public Officials in Poland. Nominal Wages and Salaries of these Workers 163 167 170 173 176 PREFACE The Fourth Session of the International Labour Conference, held in November 1922, adopted a resolution proposed by the Workers' Group instructing the International Labour Office to undertake a documentary investigation into the standard of working-class life, as compared with the pre-war standard, in countries with very depreciated exchanges, more particularly Germany; and also into the measures adopted or contemplated to secure an adequate living wage for the workers. The resolution runs as follows x : The Conference, having regard to the very marked unrest among the workers, due to the recent fall in the standard of life in certain countries, having regard to the fact that such fall in the standard of life proceeds in conformity with the continual depreciation of the exchange in certain countries, as for instance in Germany, having regard to the fact that in such countries there is a danger that the deterioration of the conditions of life may bring the population, and hence the workers, into dire straits and may even imperil the coming generation, having regard to the great number of workers affected by this state of affairs, which is capable of spreading into other countries, having in mind Part XIII of the Treaty of Versailles and in particular the words "the provision of an adequate living wage", instructs the International Labour Office to make a documentary investigation into the standard of working-class life in relation to the pre-war standard in Germany and the other countries with a seriously depreciated exchange, in order to throw light upon the situation and to ascertain the means already adopted or contemplated in such countries for securing to the workers an adequate living wage. The International Labour Office therefore undertook this investigation, the results of which are embodied in the present report. While conditions were disturbed and continually changing owing to inflation, it was impossible to appreciate rightly the facts recorded, or to detect the general tendencies indicated by them. The publication of this report was therefore inevitably delayed until monetary stabilisation was an accomplished fact and it was possible to obtain a general and detached view of the problems raised by 1 International Labour Conierence, Fourth Session (Geneva 1922), Vol. I, pp. 611-612. Geneva, 1922. — x — inflation. Though this inevitable delay may have some drawbacks, from the scientific point of view it has even greater advantages. In conformity with the resolution of the Conference attention was primarily devoted to Germany. Among other countries with heavily depreciated exchanges Austria and Poland were selected, partly because the material was more or less adequate for a complete view of the situation, but also because the experiments and measures adopted in those countries are complementary, as it were, to those to be found in Germany. The material available, however, is inadequate for a complete and detailed study of all aspects of working-class conditions in these three countries. It was therefore thought best to examine mainly the changes in the aggregate income of the working classes since before the war, as the standard of life of a social class is primarily determined by its income. Changes in wages, the cost of living, and unemployment are first examined, and an attempt is then made to measure the present aggregate income of the industrial population in Germany, Austria, and Poland, in comparison with the pre-war period. This estimate and comparison form the central point of the present report. For Germany, the information is fairly complete and exhaustive, but for Austria and even more for Poland the result is more fragmentary. In all three countries the scope of enquiry was limited to manual workers in industry, including mining and transport. Only brief notes could be given for professional workers, chiefly public officials. Agricultural workers have been excluded altogether ; this class, however, suffered less from the effects of inflation. As will be seen from perusal of the report, the chief factor in reducing the income of the working classes was the technical difficulty of adjusting wages rapidly to the rise in the cost of living. An account is therefore given of the methods adopted to effect such adjustment. In so far as they were successful they proved to be the best means adopted or contemplated in Germany, Austria, and Poland to secure an adequate living wage for the workers. No doubt the fall in the national income, which if not proved may be taken as very probable in the three countries in question, also had some share in lowering the workers' standard of life. This, however, is an obscure and controversial question. There is no material available to throw any real light on it, and no attempt was made to consider it in detail. The steps taken to check or reduce unemployment, which became serious on the cessation of monetary inflation, have not been — XI — considered here, as they have already been dealt with in other publications of the Office1. No attempt was made to study the many effects of the reduction in income on the workers' standard of Jife. This is a very wide and difficult problem, any real contribution to which would require systematic investigation on the spot; a purely documentary enquiry could not be made to cover any such attempt. Housing conditions, which are of primary importance in the standard of life of the working population, have been considered in detail in another report published by the Office2, and have therefore been left on one side here. in the present volume each of the three countries considered has been treated separately. This division svas necessitated by the progress of investigations, which were not completed at the same time for all three countries. It must also be remarked that the problems and solutions of each country are not precisely the same, and the geographical division may fairly be considered to make for clearness in presentation. Nevertheless, to avoid too frequent repetition, reference is frequently made in the sections on Austria and Poland to what has already been said on the same subject in the section on Germany. . 1 Remedies for Unemployment; Unemployment Enquiry. Geneva. 1922. " U n e m p l o y m e n t Problems", in International Labour Review, Feb.-Mai - ., J u n e , Sept., Dec. 1923. " T h e Labour Market", in Industrial and Labour Information, 1924, passim, 2 European Housing Problems since the War. Studies and Reports, Series O, No. 1. Geneva 1924. PART I GERMANY INTRODUCTION General Economic Conditions The economic life of Germany has been dominated since the war by two powerful factors whose influence is discernible in every sphere and which have very materially affected working-class conditions; these factors are currency inflation and the shrinkage of the national income. The first of these factors is perfectly plain and unmistakeable and can be traced in detail with the aid of various statistics. In 1923 the volume of Reichsbank notes in circulation amounted to 2,593 million marks. It was gradually increased during the war years, and at the end of 1918 stood at 22,188 million marks. Expansion continued fitfully during the ensuing months; at the end of 1922 the volume of notes in circulation was five hundred times that of 1913. In 1923 the increase became quite inordinate, the note issue mutiplying a hundred or even two hundredfold in the course of one month; in November 1923 the total volume of notes reached 400,267,640,302,000 million marks. Reichsbank notes did not, however, account for anything approaching the total volume of currency. Even during the war, but especially afterwards, quite a number of public and private bodies had engaged in different forms of note issue ; such currencies were usually local. The state railways issued emergency notes for an amount equal to about a quarter of the Reichsbank note issue. In addition to this, currency notes were issued, with government approval, by private banks and certain municipalities. The emergency currencies of all kinds issued by public and private institutions without the approval of the central government also totalled very large sums — exactly how much it is extremely difficult — 2 — to gauge. Official statistics put the amount, at the time when it reached its maximum, at 160,000,000 million paper marks 1. The effect of this huge inflation was to render the paper mark practically useless. It had long since ceased to be the standard of value and from the middle of 1923 onwards its utility as a medium of exchange was greatly impaired. The introduction of a stable currency became an absolute necessity. The earliest attempts to introduce such a currency were made locally. A number of municipalities, those of Berlin and Hamburg, for example, had already issued gold marks convertible into dollars. Again, notes backed by stable-value loans had been placed in circulation 2 ; the certificates themsehres of stable-value loans had been issued in the form of coupons of small denomination which could be employed as emergency currency. The decisive step towards stabilisation was taken on 15 November 1923, when the Rentenbank was founded with a capital of 3,200 million rentenmarks secured by a general mortgage on private property, including land. The Rentenbank, which was granted autonomous status, issued rentenmarks each equivalent to 1,000,000 million paper marks. This new currency, which was issued with great prudence, was inadequate at first to meet the whole of the demand for a stable currency. The issue of other emergency currencies of similar character was therefore continued up to the end of 1923. The issue of paper marks, on the other hand, was suspended until January 1924, at which date the amount in circulation (including other forms of emergency paper money) had dropped to barely one-quarter of the total circulation, calculated at gold values. From February 1924 onwards the various emergency currencies (whether paper or of stable value) were gradually withdrawn and partly replaced by rentenmarks, the issue of which continued. That of paper marks was also resumed, but within much restricted limits. In September rentenmarks accounted for about half the currency in circulation and paper marks for somewhat less than half, the balance being made up principally by metallic token money and some few forms of emergency currency in process of withdrawal. 1 2 Wirtschaft und Statistik, 1924, No. 6. These stable-value loans, issued by public or private bodies, were usually based on the dollar exchange rate or on the price of certain staple products, e.g. coal, wheat, or potash. Obviously, in the latter case their stability was only relative, but this system was none the less an improvement on the papermark. — 3 — From January to September 1924 the total volume of the currency was increased by about 50 per cent. ; but the main object of this increase was to meet the requirements of commerce, as the velocity of circulation had notably decreased since stabilisation had been effected. The enormous inflation of the currency during the last few years, and more particularly in 1923, entailed a correspondingly great depreciation in the purchasing power of the mark. The depreciation, however, was not equal in degree or rapidity on the various markets owing to differences in economic organisation. There was thus a dislocation, as it were, of the purchasing power of the mark between different markets, due to the delay in certain markets in adjusting prices to each new stage of inflation. In the foreign money market the reaction was very marked and almost instantaneous, though it was also influenced by speculation. In the wholesale market the reaction was already somewhat more belated but still very marked; its effects were particularly felt in foreign trade. In the retail market, on the other hand, the reaction was delayed and slighter. In the labour market it may be said to have been wholly inadequate. Obviously in normal times the action of economic forces would quickly have eliminated these discrepancies in the puchasing power of the mark in different markets. But the continued progress of inflation precluded any such readjustment. The dislocation of purchasing power thus persisted practically unabated throughout almost the whole period of inflation, and it was only in 1923 that there were signs of fresh developments. The index number of the dollar exchange was then about twice that of internal wholesale prices, and the general index of wholesale prices (imported and home products) was, in turn, practically double that of the retail prices of the main articles in current consumption. At this moment, however, the increase in prices became so rapid that the paper mark ceased entirely to be the standard of value and was only used as a medium of exchange. In industry and commerce it had become the practice to keep accounts and effect all transactions on the basis of a stable unit — usually the dollar — prices being converted into marks on payment, at the current exchange rate. This practice was first adopted for foreign transactions; subsequently it spread by degrees to the interior of the country, first to the wholesale and afterwards to the retail market. The effect of this development was gradually to eliminate the discrepancies in the purchasing power of the mark. — 4 — The rise in prices gathered such momentum in the autumn of 1923 that wholesale and retail dealers were no longer content to fix their prices in stable units at the current rate of exchange, according to the cost of their goods; they began in some measure to discount a future rise and to fix their prices in marks by what they expected to be the cost of replacement a few days later. The consequence was that in November 1923 retail prices rose above wholesale prices, which themselves had been above the level of the foreign exchanges since October. In November the value of the mark was stabilised by the introduction of the rentenmark, and a drop in prices immediately ensued. The fall, however, was neither very considerable nor of long duration, and in April 1924 a slight rise began again, followed by some fluctuations until the autumn. Throughout this period the reversed relation between wholesale and retail prices and between the dollar exchange and wholesale prices continued 1. In considering the repercussions of these movements on the labour market, it is essential to distinguish two aspects of the problem. The prices ruling in the labour market, i.e. wages, should be viewed from two different angles. From the producers' point of view they are the cost of labour, while from the workers' they are the income of the working classes. Detailed examination of the second aspect of the problem will be deferred to Chapter I; here only the first will be considered. It is extremely difficult to estimate the cost of labour, and wage statistics afford at best very inadequate data. The only information to be gained from such statistics is the cost of labour per hour's work for certain classes of workers. Such data are of value only in so far as it can be assumed, first, that output per worker per hour has remained constant throughout the period under review, and, secondly, that the proportion of skilled to unskilled labour in the various industries has not varied. Neither of these hypotheses can be readily verified, and they have, in fact, given rise to keen controversy. There seems, however, some justification for asserting that average output per hour has decreased by comparison with that of pre-war times, owing partly to the general demoralisation engendered by the critical economic situation, but even more to the physical enfeeblement of the population through underfeeding during and after the war and to the 1 This twofold development in the purchasing power of the mark in different markets emerges plainly from a comparison of the index numbers of prices, which are given in detail in appendix T. falling out through death or disablement of many of thé most robust workers. Similarly, it is probable that the proportion of skilled to unskilled workers has changed to the detriment of the former, owing to a setback in training. The difference between the wages of skilled workers and those of unskilled workers has become less marked x, and this has deterred many young workers from beginning their apprenticeship. Furthermore, the war had a most prejudicial effect on the organisation of training and apprenticeship. Finally, many disabled ex-service men who were formerly skilled workers have been unable to resume their trade. Thus the figures of the increase in the cost of labour given by the statistics of hourly wages tend to underestimate such increase, though subject to this important reservation they are none the less of interest. Table I gives the index numbers of hourly wages of skilled and unskilled workers in all the industries of the country and compares them with another element in the cost of production— the cost of raw materials as indicated by the index of wholesale prices of industrial raw materials. Even allowing for the fact that the index numbers for the cost of labour are undoubtedly too low, the difference between these and the cost of raw material is sufficiently great to secure acceptance for the statement t h a t the cost of labour has lagged an appreciable distance behind the other factors in the cost of production a . Thanks to the stabilisation effected in 1924, the difference was much less marked in that year than during the preceding years, but it still varied from month to month. 1 2 See chapter III, "Cost-of-Living Bonuses". The legitimacy of directly comparing the rate of wages per hour (cost of labour) with the wholesale prices of industrial raw materials for the same dates is open to dispute. As a general rule wages are only paid at the end of the week and it is then that they enter into the cost of production. The raw materials employed by the workers whose hourly wages are considered as factors in the cost of production are bought earlier, the difference in time varying considerably between one industry and another and one establishment and another. If the comparison is made with the index number of wholesale prices of raw materials on the date corresponding with the workers' pay-day, it will be noticed that the difference between the two figures is evengreater than that shown in the table, as the following index numbers furnished by the German Ministry of Labour indicate (Base: 1913 = 1). Date April August September October 1922 1923 1923 1923 Cost of unskilled labour Wholesale prices of industrial r a w materials 40.2 1,128,540.0 28,040,000.0 10,643.000,000.0 68.6 2,220,000.0 56,440,000.0 25,760,000,000.9 — 6— TABLE I. TION: INDEX NUMBERS OF TWO ELEMENTS IN COST OF PRODUC- (1) COST OF LABOUR (HOURLY WAGES) FOR SKILLED AND UNSKILLED WORKERS IN THE PRINCIPAL INDUSTRIES; SALE PRICES OF INDUSTRIAL ( 2 ) WHOLE- MATERIALS 1 RAW (Base: 1913-1914 = 1) Month Cost or skilled labour Wholesale prices of industrial raw materials Cost of unskilled labour Calculated from prices in paper marks 19 2 2 April July October 192 3 J anuary February Marcii April . May June July August September October November December 29 49 163 40 67 223 69 112 569 816 2,058 2,559 2,510 3,301 8,098 29,194 835,366 20,909,000 8,146,000,000 542,460,000,000 943,300,000,000 1,126 2,794 3,492 3,550 4,500 11,015 42,952 1,128,541 28,040,000 10,507,000,000 711,100,000,000 1,211,900,000,000 3.523 7,971 7,790 7,566 11,601 27,882 93,351 1,478,063 38,036,000 10,197,500.000 813,000,000,000 1,541,000,000,000 Calculated from p rices in r e n t e n m a r k i 19 2 4 J anuary February March April May June July August September 0.863 0.857 0.864 0.899 0.997 1.064 1.043 1.024 1.069 1.050 — — — — — — 1.046 1.219 3 1.441 1.450 1.474 1.509 1.522 1.490 1.391 1.382 1.385 1 The index numbers of hourly wages have been supplied by the Federal Statistical Office. They are calculated from the official statistics of wage rates per normal working week, the reduction to hourly wages being effected according to information possessed by the Statistical Office as to the normal hours of work in the various industries at the various dates. The index numbers of wholesale prices of industrial raw materials are taken from the official index, which distinguishes between industrial products and food products: it covers mainly raw or semi-finished products. The two index numbers (wages and prices) refer to the average of each month and are therefore calculated for corresponding periods (in this connection see footnote 2 on the preceding page). s To convert the index number of prices in rentenmarks to one for prices in paper marks the decimal point should be moved twelve places to the right, as one rentenmark equals 1,000,000,000,000 paper marks. — / — The repercussions of currency inflation and the resultant depreciation and dislocation of the purchasing power of the mark have been the determining factors in the economic activity of Germany since the war. In the first place, depreciation alone, which reached the fabulous level described above, entirely wiped off the charges on all debts contracted when the currency was still more or less stable. This was obviously detrimental to all creditors, but it was exceptionally favourable to large borrowers, i.e. the state and private undertakings which had contracted large loans; in industry, for example, the liabilities represented by equipment and plant were reduced to nil. In the second place, the rise in prices in itself was favourable to industry. As the price of an article increased during manufacture, expenditure incurred at the outset of the process became proportionately less. In the third place, the dislocation of purchasing power in different markets again diminished certain costs of production. As noted above, labour costs were materially reduced ; similarly, many other forms of expenditure could be kept at a lower level than that which they reached abroad. In consequence of this, German industry was able to export extensively to other countries, in spite of protectionist measures, and at the same time to cope with foreign competition in the home market under peculiarly favourable conditions. Nevertheless, the relatively easy conditions induced by inflation often led to less economical management, waste of raw material, and cessation of technical improvements. It should further be noted that in certain cases the profits made through the rise in prices were illusory, as the increase in the paper-mark price of the finished goods did not wholly compensate for the depreciation of the currency in terms of the dollar. Transactions which promised to yield very considerable profits on a paper-mark reckoning revealed an actual loss when calculated in gold. In spite of this, and of the economic disorganisation caused by the war, the loss of valuable industrial regions, the heavy burden of taxation, the high rate of interest and the high price of coal *, 1 In 1913 taxation represented 10.9 per cent, of the income of the population: during recent years, before stabilisation took place, the percentage was 27.7 (estimate by the Ministry of Labour). The rate of interest multiplied . more than five-fold and the cost of coal rose sharply owing to the contracts passed with the Micum. German industrial activity continued without interruption throughout the post-war years up to 1923, though a depression of exceptional gravity was developing in all other countries. The unemployment statistics given later * show that wellnigh the whole of the working-class population was employed despite all the difficulties of demobilisation. But in 1923 matters took a fresh turn, and thereafter the position was completely reversed. In the first place, the dislocation of the purchasing power of the mark between various markets began to disappear. Production costs, which were comparatively low and which had so far remained below those of foreign industries, gradually rose until they reached an exceptionally high figure. Money became very scarce and could only be obtained at exorbitant rates 2, which were still further increased by premiums to cover the risk of depreciation. The rise in prices assumed such fabulous 'proportions that the economic life of the country was to some extent paralysed. This state of affairs, added to the effects of a long-continued industrial boom, precipitated the crisis which had been delayed by currency inflation but was now intensified by the stabilisation of the mark and the resultant fall in prices. Yet another factor tended to aggravate it. During the period of inflation an enormous amount of capital had been invested in extensions and improvements of factories; apart from foreign investments this was the only means of avoiding depreciation of profits; but all this new plant stood unused and simply increased interest charges and the cost of upkeep. : From the end of November and more particularly in December 1923 unemployment grew to alarming proportions. An appreciable renewal of activity occurred in the spring of 1924, but this seems to have been due to the opening of large temporary credits by the Reichsbank rather than to a genuine improvement in the situation. When a policy of credit restriction was deliberately embarked on in May of the same year, bankruptcies multiplied; the crisis came to a head in' July. Unemployment again increased during the summer months, though it did not reach the level of December 1923. It will probably be many months before the situation bequeathed by the war is finally cleared up, and inflation has merely delayed the process. 1 See chapter I, p. 21. - 2 During the last months of 1923 the Reichsbank discount rate reached 90 per cent. — 9 — Notwithstanding the industrial activity in Germany from the end of the war till 1923, there has been a certain shrinkage in national income. It has already been stated that this reduction was the second dominant factor in Germany's economic development during that period. Unfortunately, there are no precise statistics from which the extent of the change can be accurately gauged: Furthermore, even with such statistics as might, in default of better, give some general indications, comparisons with the pre-war position are extremely difficult, since both the territory and the population of Germany have been considerably modified by the Treaty of Versailles; post-war Germany is by no means the same economic unit as the Germany of 1914. Although the shrinkage in the national income cannot at present be investigated on statistical lines, the fact of such shrinkage may be accepted as fairly self-evident. A similar phenomenon has been noted in several of the victorious countries where more precise calculations have been possible. What is true of these countries should, a fortiori, be true of Germany. Economic organisation for war needs was carried furthest in Germany, and re-adjustment to peace conditions has involved the most farreaching changes; it is in Germany that the population was weakened most by under-feeding; and it is on Germany that the heaviest reparation charges have been imposed, while areas of great economic importance are no longer under her control. There is, moreover, a very strong presumption that productivity has declined during recent years owing to the weakening of physique and morale. A fall in the national income would therefore seem to be a very probable consequence of the war. At the same time, these conditions do not appear to be final. During recent years the whole machinery of production in Germany has been considerably improved, as was pointed out by the Committee of Experts of the Reparations Commission. The exceptional profits realised during the period of rising prices have been extensively invested in the undertakings themselves with a view to improving and developing the means of production. Though the present depression may be long and seArere, therefore, once it is at an end economic activity can probably be resumed with improved plant and means of transport which may raise Germany's productivity above its pre-war level. CHAPTER I Real Income of the Working Classes DATA AVAILABLE AND T H E I R INTERPRETATION The determination of working-class income has always been a matter of great difficulty, and even an approximately correct estimate can only be achieved after extensive enquiries among working-class families themselves. In addition, to measure the changes which have occurred several enquiries at different periods would be necessary, conducted on such bases and methods as would render the results comparable. In Germany there had been no such investigations, or at any rate none on sufficiently broad lines to be accepted as representative of the whole country. I t has been necessary to use data of another kind. An endeavour has been made to ascertain from wage statistics, first, the income of workers in normal employment, and then b y comparison with the cost-of-living index number the real value of such income on a pre-war basis. It was next considered how far these results must be modified in the light of the general employment situation so as to give an approximate indication of the total income of the working classes; for this correction statistics of total and partial unemployment were used. It was possible to make fairly detailed calculations and estimates for industrial workers, as reasonably complete statistics are available, especially for the period 1922-1924. Owing to the lack of satisfactory data the calculations have had to be more restricted and approximate for professional workers, while for agricultural workers they had to be completely abandoned. Most German wage statistics are based on the wage rates fixed . by collective agreements and reported to the Government by the employers' and workers' associations. Such data, however, do not invariably include certain supplementary allowances granted to the workers, which are sometimes fairly considerable. Thus to »ome extent they underestimate the income of workers in normal employment. It should be noted, moreover, that in some cases the statistics for the pre-war period and for 1920-1922 are calculated on figures — 11 — indicating the workers' average earnings and not nominal wage rates. Although the differences between these two kinds of data are inconsiderable at times of normal activity, yet their significance is not entirely similar, whereby the continuity of the series is somewhat impaired. In view of the very large variations in prices in Germany, it is absolutely necessary to compare nominal wages — which in themselves have no real meaning — with the cost of living, in order to ascertain what the workers were able to purchase with their earnings. In other words, the real value of the wages must be ascertained. To this end the rates of nominal wages for given periods are divided by the index number of the cost of living. In this manner the real value of wages in relation of their pre-war value is obtained for each date. I t should be borne in mind, however, that the official index number of the cost of living which is used in these calculations is based on a theoretical budget so framed as to represent the standard of life of a typical working-class family (father, mother, and two children) after the war, and more especially in 1919-1920, during which period the methods of calculating the index number were devised. The pre-war base of the index number is thus calculated from this theoretical post-war budget, valued at pre-war prices. Consequently, the index number of the cost of living indicate» the increase in expenditure required to maintain at a uniform level as from pre-war days a standard of living corresponding to post-war conditions. The purpose of the present study is to compare the real value of the workers' income after the war with that which it had before. What is needed, therefore, is an index number of the cost of living which would indicate the increase in expenditure required to maintain the pre-war standard. The rise in wages would then be compared with this increase in order to determine the fall in real income resulting from the twofold movement of prices and wages. But in reckoning the official cost-of-living index number it was assumed t h a t the standard of life had fallen; the index number is based upon reduced consumption, excluding wholly or partially goods whose price had undergone the greatest proportionate increase. Part of the increase in expenditure is thus not indicated. I t follows t h a t the index numbers of the real value of wages in terms of their pre-war value, if calculated from the official cost-of-living index number, are apt to be unduly high. This important reserA-ation must be made in respect of all figures quoted in this chapter. — 12 — It should also be pointed out that for a correct estimate of the real incomes of workers in normal employment the real value of the wages during the period in which they are spent should be known. The wage statistics refer to the periods during which the money was earned. It may be assumed that wages earned during one week are largely spent during the next ; consequently wages earned in the course of one month are spent during a period falling one week later than the month during which they were earned. The cost-of-living index for the later period has therefore been employed in calculating real wages. This distinction is indispensable for the years 1922 and 1923, during which the rise in the cost of living was particularly rapid; it was observed in the calculation of the index numbers of real wages published by the Federal Statistical Office, and reproduced below. For 1920-1921, a period of relatively slow increase, it was not thought necessary to make the distinction. Calculations were based on the cost-of-living indexes for the corresponding months, as the increase in cost of living was much smaller, and the available data were in any case not sufficiently precise to call for such correction. The tables given in the following pages show the principal official statistics published in Germany for the years 1920-1924 in respect of the wages of industrial workers and the salaries of professional workers. In the first case, for purely technical reasons, two periods are distinguished: from 1920 to the first months of 1922, and from then to the first months of 1924. Statistical information for the first period is meagre, and the data are not directly comparable with those of the latter period. Moreover, the sources and statistical methods employed are not uniform for all industries, and in some cases vary from one month to the next in the same industry. Too great reliance should not be placed on the information given ; only an approximate and very general idea can be gained from it. From April 1922 onwards, on the other hand, the information is far more satisfactory. It has been systematically published by the Federal Statistical Office; the methods employed are sufficiently uniform and the general lines sufficiently broad to permit of comparisons between industries and give an accurate picture of the situation. It should be emphasised, however, that the figures for the period July-November 1923 are so enormous in comparison with the base figures that here again the results are not to be regarded as absolutely accurate, since the slightest error in the base figure would lead to enormous differences. — 13 — AU the tables in this chapter refer exclusively to index numbers of real wages calculated in the above-described manner. A series of tables corresponding to those here given, and referring to nominal wages expressed in marks, will be found in an appendix 1, together with further particulars as to the methods employed in computing them. REAL WAGES OF INDUSTRIAL WORKERS, 1920 TO 1924 Tables I and II show that at the beginning of 1920 the real wages of workers in various industries had already depreciated considerably. The w7ages of skilled workers ranged from 55 to 70 per cent. of their pre-war A'alue, those of unskilled workers from 70 to 80 TABLE I. INDEX NUMBERS OF REAL WAGES OF SKILLED WORKERS IN CERTAIN IMPORTANT INDUSTRIES, 1 9 2 0 TO 1922 (Base: 1913 = 100) Textiles s Metals Date 1920 Feb Apr. May June Julv Nov. Dec. 1921 Jan. Feb. May, July Aug. Sept. Oct. Nov. Dec. 1922 Jan. Feb. Apr. Mines 67 1 Building2 i 56 | 57 3 62 Wood Average of several Berlin* towns 3 69 56 s 48 57 78 67 60 89' 61 94 94 94 Men 72 State Print- undering: takings Women Berlin* (higher5 grade) 89 55 6 59 56 63 69 71 64 61 71 72 68 66 73 73 85 75 54 69 65 68 74 67 81 57 73 61 69* 94 78 80 2 85 1 74 70 2 72 74 76 67 74 70 90 78 72 69 59 62 Quarterly average of earnings per shift in the principal mining districts. ?3 Weekly rates of wages fixed by collective agreements in several towns. Average hourly earnings ascertained by a census in some twenty towns, multiplied by the normal hours of work as fixed by collective agreement. *3 Weekly rates of wages fixed by collective agreements for Berlin. Weekly rates of wages in towns in the highest cost-of-living class (class A). * Figure refers to January 1920. ' Figure refers to the third quarter. 1 Appendix II. — 14 — per cent. The wages of women workers in the textile industries alone maintained their level more consistently; they have, as a matter of fact, always been so low that they quickly reached an irreducible minimum. TABLE II. INDEX NUMBERS OF REAL WAGES OF UNSKILLED WORKERS IN CERTAIN IMPORTANT INDUSTRIES, 1 9 2 0 TO 1 9 2 2 (Base: 1913 = 100) Metals Date Mines ! Building 2 Wood 3 Several towns Berlin * State undertakings 5 1920 April May June July November December 85 • 94 6 101 84 102 111 89 104 93 '6 95 107 90 3 1 1. 1 . 1922 January February April Í 70 ? 63 77 1 11 1 1 11 1 1. 1921 January February April July August September October November December 67 S 111111 February 80 3 66 79 88 88 88 94 95 91 88 93 95 100 2 85 112 96 99 101 90 100 95 121 104 96 100 94 79 85 96 116 2 102 - 1 2 Quarterly average of earnings per shift in the principal mining districts. Weekly rates of wages fixed by collective agreements in several towns. 3 Average hourly earnings ascertained by a census in some twenty towns, multiplied by the normal hours of work fixed by collective agreements. *6 Weekly rates of wages fixed by collective agreements. Weekly rates of wages in towns in the highest cost-of-living class (class A). 6 Figure refers to the third quarter. This general depreciation in wages grew progressively less during 1920 and 1921. The available figures do not indicate precisely the date at which the index numbers reached their maximum in each industry. It may be said, however, that the highest real wages were recorded in 1921, especially during the second half. At this period the wages of skilled workers represented from 65 to 95 per cent, of their pre-war value, and those of unskilled — 15 — workers from 85 to 110-115 per cent. It would therefore appear that some classes of unskilled workers were in a better position than before the war. Tables III and IV show that during 1922 and up to November 1923 the situation grew steadily worse. Depreciation was not uniform, however ; it progressed with a number of fluctuations, each waAre of depreciation being followed by a slight reaction which brought about temporary increases. These fluctuations did not entirely coincide in all industries, but it may be said that in the majority of cases abrupt rises in wages occurred during March and August 1923, while January and October of the same year were conspicuous for particularly low real wages. The workers in all industries suffered, though some more than others. Among the industries most seriously affected may be mentioned the printing trades and state undertakings. The mining, chemical and textile industries suffered least; in the last-named the women workers were better able to maintain the level of their wages than the male workers for the reasons given in connection with the fluctuations in 1920 and 1921. The fluctuations of real wages- were much alike for skilled and unskilled workers in the same industry, though the latter were relatively less affected. From December 1923 onwards, prices having been stabilised, a general reaction was experienced. Real wages definitely increased in all industries, and continued to rise gradually — albeit with some few fluctuations — during the whole of 1924. These fluctuations, which were in any case slight, were attributable merely to variations in the cost of living, nominal wages having increased almost uninterruptedly in most industries. In September 1924 the wages of skilled workers stood barely 10 per cent, below their pre-war value ; those of unskilled workers approximated still more closely to the pre-war standard, and in some industries they appear to have risen even above it. This general tendency emerges very clearly from table V, which gives average real wages calculated for each year from the figures in tables I to IV. The figures for 1920 and 1921 are obviously of less value than those for 1922 to 1924, as they have been calculated from tables I and II. In certain cases they are even based on a single annual return only, and should be accepted with reserve; for 1922 three monthly returns only have been employed, and for 1923 the average of the first eleven months has been taken, December being included with the first three quarters of 1924. This method of procedure is permissible, considering the character of Weighted average tor ail these industries — 16 — r>.io KÍ«Ot>.r>tu«U>ÍO^ií5Wr* r-^c^r-.c©CQCsC7>CTiCO W5 C M O <£> » Í C i n O CC O « ODl>t£! i O l x D M O t M i î I M N t D O l ^ IM>COCOÛ000Q0»00 XilOO M X t N C 5 r ( O O O < i l t N r i < Ä f l 0 O ^ ií5rNOOitNtMflOOl>CO<íi^ CO I N [ M » O r i r i ^ O [>• OO CO OC Oí C l O l 3 i O l I r>y3 ^ o ^iorM>tO(ûv?ioioco»oi> M » o * f f ) < í ^ f O r i r-.c-oor-*r^cococTici I or^T-f cor>or-»TH (N(NMOI>Wi^(0«H00D (ÛCOCO O O C O < f 0 0 ( N CO CO •*# r^r-io riCOa>CO(N^OOlCOOlí5 LirNOO»[^tMi:oorM»u:rN W tO ^ O COM O rH Ol oocooocoooQCiaioo State undertakings (mainly transport) n I ! c q Uî COW(M^ff)t>iO'00 ^ s Coal mines Building Men H. 1 en en en CO oor*rfCiœcûQa> O O Î X O W i û U Î O ^ r^cocoHO^ocíCTícn 5 CS CO O co d r-* Or>>a)KÎCÛrl(NOCO(NiÛiO torvOOcoxtOCico^Or» l > » í t u CO O) CO O O O» r s co co O co c i co O^ co 1 Or^.í£> as as r-» OlOOifOOSCTiXlCCO'HtOCO [>cooodMocorM.orNCíi C5 *# l í j r . O CO CN CO ift cociascoasaiasasai r-> <£> IO Gioir» O rH Oí * í IO O) (N.** O O r ( l > iûr>cnffir^r>(û!DcowrNCû ÖCOC-«. co ai ai o a> Cl O C-* CiCitO tOriWOW^sJfCO^OOlO tOOrnOCiCHOOOlOMÍ r i n n TH -^H [NOOtûtûOCOCicÛ © T H O O S O ^ H O O O r l r i ^ ^H r-l *r< *-< ^ c. O Textiles (T>(JÍ t > Wood -rH Building T* Coal mines Chemicals ^ -r* -î-i 5Í t ¡ Ctf_ ^ £ O CO CO J i Ol Q Cl O Oï Oí ^ tfi *"* tí *- ^ +* fi (M-50(Ni-5faS-s < 5!iX!0!z;QS^CiHg 2 — 18 — the averages for the preceding years, and has the further advantage t h a t it throws into relief the very striking changes entailed by stabilisation of the currency. The general rise of real wages in the various industries from 1920 to 1921, their gradual decrease during 1922 and 1923, and t h e renewed increase as from December 1923, can be clearly seen from the table. In connection with these various estimates attention must again be drawn to the reservations respecting the value of index numbers of real wages. They probably do not fully indicate the depreciation, owing to the methods of constructing the cost-of-living index numbers employed in calculating them. TABLE V. AVERAGE INDEX NUMBERS OF REAL WAGES OF SKILLED AND UNSKILLED WORKERS IN THE PRINCIPAL INDUSTRIES, 1920 TO 1924 (Base: pre-war = 100) Industry 1920 1921 Average of several months Average of several months 1922 1923 Average Average of April, of period Jan.to July, October Nov. 1923a 1924 Average of period Dec. 1923 to Sept. 1924 Skilled workers Mines Building Wood Metals Textiles: men women Chemicals Printing State undertakings, mainly transport Weighted general average 82 62 69 ! 56 ! 72 * 89 ! 92 75 69 i 84 1 — — •— 66 75 68 67 73 62 80 91 75 60 72 56 56 58 72 81 72 52 86 76 84 81 85 92 83 86 62 66 50 62 76 83 87 82 89 91 86 93 86 69 92 68 70 82 79 82 81 63 105 84 100 92 88 91 91 100 85 86 69 79 Unskilled workers Mines Building Wood Metals Textiles: men women Chemicals Printing State undertakings, mainly transport Weighted general average 87 75 62 ! 80 ! 108 92 100 ' — — — — — 88 100 • — 5 -a ? $ ¿f 6 ~ ¿ •-S, 86 "' 94 i Figure based on a single return during the year; the month referred to can be ascertained from tables I and II. — 19 — ESTIMATE OF THE AGGREGATE INCOME OF THE INDUSTRIAL POPULATION, 1921 TO 1924 The information on real wages afforded by the preceding tables indicates the income of workers in normal employment only. To form an idea of the income of the whole industrial population, it is also necessary to consider the income of workers employed overtime, totally or partially unemployed workers, sick and disabled workers, and those involved in industrial disputes, and finally any supplementary income received by wage earners. Direct information on this head being unobtainable, another method of enquiry must be adopted. The index numbers of real wages, which show fluctuations in the wages of a worker in normal employment, might serve at the same time as index numbers of the aggregate income of the entire working population if all workers without exception were in normal employment. The index numbers of such aggregate real income could then be obtained by increasing the indexes of real wages to allow for supplementary earnings, and by decreasing them in proportion to the losses caused by unemployment, sickness, disablement, industrial disputes, etc. Of all the agents enumerated it will be sufficient, however, to deal with those which provoke comparatively sudden and marked fluctuations. When exceptional losses or gains stand in a more or less constant ratio to the aggregate income of the industrial population, they in no way modify the index number. The latter being merely a ratio between two quantities (real wages at a given date and those at the basic date), no modification is entailed by increasing or decreasing both quantities by the same percentage. As regards losses caused by sickness or disablement, it may be said that the number of workers affected remains practically constant over a period of several years. It is true that, owing to privation and to the high death rate among the younger and more robust men due to the war, the standard of health in Germany is probably below its pre-war level, but it is impossible to determine the effect of such conditions on the income of the whole working population. The losses entailed by industrial disputes also show considerable fluctuations. It has been thought unnecessary, however, to take these into account, as the resultant proportionate decreases in the income of the working classes is always very slight; even — 20 — in years when it reaches its highest figure it is not more than one per cent. 1 It is also impossible to determine the value of supplementary earnings which go to swell the income of the industrial population ; the variations to which they are liable can scarcely be such, however, as to modify appreciably the proportion of working-class income for which they account. The most variable factor, then, is the extent of employment among the population — that is to say, the opportunities available for earning normal wages and special overtime pay. The index number of real wages (or of the income which the population would receive if all workers were in normal employment) must therefore be modified in accordance with employment conditions. For the index number of the income of the entire industrial population, the income received during the last years before the war at current wage rates, allowing for the degree of employment then prevailing, is taken as base, equal to 100. The index number of real wages should therefore be reduced or increased according as the degree of employment was greater or less during the years under review^ than in pre-war years. This gives an index number of the income, the regular income at least, of the entire industrial population. The degree of employment is indicated with reasonable accuracy by unemployment statistics. Prepared from information furnished by the trade unions, they cover a large portion of the population and include the principal industries 2. They are, moreover, substantially confirmed by the official statistics of unemployed in receipt of relief 3 . 1 In 1919, when this reduction reached its maximum, the number of worker-days lost in consequence of disputes was 32 million, as against approximately 8 millions in 1912-1914. The difference (24.5 million) is equivalent to about 0.8 per cent, of the aggregate worker-days available yearly in German industry. 2 For 1924 the percentage of wholly unemployed workers is based on reports from trade unions with a membership of over 4 millions, and the percentage o£ partially unemployed from those of organisations with a membership of over 3% millions. For the pre-war period the figures of total unemployment cover a smaller number of persons — from 1 y2 to 2 millions. The data for this period, however, are not of such outstanding importance in this estimate as to make this difference in scope a matter of serious concern. No information has been published regarding partial unemployment during the pre-war period. 8 According to an estimate by the Minister of Labour, Dr. Brauns (Wirtschaftskrisis und Sozialpolitik ; Volksvereinsverlag G.m.b.H., München-Gladbach, 1924), the number of unemployed in receipt of relief (including those in the occupied territories) at the time of the unemployment crisis towards the end of 1923 amounted to 3 million totally unemployed and 2.3 million partially — 21 — TABLE VI. UNEMPLOYMENT AMONG TRADE UNIONISTS, 1912 TO 1924 Date 1912: yearly average 1913: yearly average 1914: first half 1920: yearly average 1921: yearly average 1922: yearly average 1923: yearly average 1922: J a n . Apr July Oct. Percentage of wholly unemployed Percentage of partially unemployed 2.0 — 2.9 — 3.2 — 3.8 — 2.8 5.4 1.5 2.6 10.2 27.8 3.3 0.9 0.6 1.4 1.6 0.6 0.8 4.7 Date Percentage of wholly unemployed Percentage of partially unemployed 1923: J a n . Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. 4.2 5.2 5.6 7.0 6.2 4.1 3.5 6.3 9.9 19.1 23.4 28.2 12.6 14.9 23.6 28.5 21.7 15.3 14.5 26.0 39.7 47.3 47.3 42.0 1924: J a n . Feb. Mar. Apr. May June July Aug. Sept, 26.5 25.1 16.6 10.4 8.6 10.5 12.5 12.4 10.5 23.4 17.1 9.9 5.8 8.2 19.4 28.2 27.5 17.5 It will be seen that during the period 1920-1922 the intensity of unemployment was not widely different from that which existed before the war. It is noticeable, however, that in 1920 the percentage shown by the statistics is very slightly higher than that of the two years immediately preceding the war, but, as the relief granted to unemployed during t h a t same year was considerably higher than before the war, the effects of unemployment on the unemployed. These figures, however, cannot be directly compared with those of the trade unions unless unemployed not in receipt of relief are included and agricultural and professional workers excluded. As regards total unemployment, the former group may be estimated at 300,000 or 400,000, the latter at 200,000. This gives a total of 3,100,000 totally unemployed, which represents 26 per cent, of the 12 million industrial workers in Germany, a figure closely approximating to that given in the trade union statistics for December 1923 (28 per cent.). As regards partial unemployment, relief was granted only to such workers as had lost more than a third of their normal hours of work. Assuming that from 200,000 to 300,000 partially unemployed fulfilling these conditions were not in receipt of relief, a total is obtained of 2,600,000 partially unemployed who had lost more than one-third of their normal hours of work. No deduction is necessary to eliminate unemployed agricultural and professional workers, as practically none of these are included in this category. This gives a percentage of 22. The percentage thus arrived at is 28.3, which is fairly close to the first figure given. — 22 — aggregate income of the industrial population must have been much the same in these two periods. In 1921 the percentage of unemployment was practically the same as that recorded before the war. For the above-mentioned reasons the reduction in the income of the entire working class must have been somewhat slighter than in 1912-1914; but the difference must be so very trifling that there appears to be little object in attempting an estimate, especially considering the very approximate nature of the data respecting real wages available for 1921. For 1922 the difference must be distinctly greater. Unemployment was less prevalent than during the preceding year; but the difference is still small. On the other hand, the increase in workingclass income due to overtime must have been substantially greater during this year than before the war; for not only was economic activity more intense, but the increase in pay for overtime over normal rates was greater. The aggregate income of the industrial population must thus have been greater (though not to any considerable extent) than that shown by the index numbers of real wages of workers in normal employment. In 1923 unemployment assumed much greater proportions. During the first half of the year the percentage of totally unemployed among trade unionists was not very high as yet, but for several months those partially unemployed represented about onequarter of the membership. It was in the autumn that the crisis developed with full intensity, a quarter of the members being totally unemployed, and more than a third and sometimes even one-half on short time. In the first two months of 1924 there was a slight improvement in the situation, and from March onwards unemployment was distinctly on the wane. In June it increased again, though it did not reach the pitch of November and December 1923. If working-class conditions are to be accurately gauged, it is essential therefore to attempt a more precise estimation of the decrease caused by unemployment in the income of the industrial population. For this estimate it has been assumed that the trade union statistics give a fair indication of the extent of unemployment among the entire industrial population. The first step was to determine the proportionate increase in unemployment since 1912-1914, so as to find the changes in the degree of employment since before the war. Were no relief of any kind paid to the unemployed, the resulting percentage would be that by which the wage index number should be reduced to obtain the index of the income of the entire industrial population. Allowance must, however, be made for the relief paid — 23 — to the unemployed, as this, though small, is not negligible. The chief basis for this adjustment was the official rate of relief, though some allowance was made for other relief which the unemployed might obtain from their trade unions and public or private institutions, together with any possible supplementary earnings. By this means there was obtained the percentage decrease in income through unemployment per unemployed person. If this is applied to the percentage of unemployed the result gives the percentage reduction in total working-class income due to the increase in unemployment since before the war. These figures were calculated separately for total unemployment and partial unemployment and then totalled. The result of these estimates will be found in table VII. Particulars of the methods employed and the calculations made are given in an appendix 1 . No figures are given for dates prior to November 1923, as during the period of inflation the rates of relief were so frequently changed TABLE VII. APPROXIMATE ESTIMATE OF THE REDUCTION IN INCOME, THROUGH UNEMPLOYMENT, OF THE WHOLE INDUSTRIAL POPULATION, TO SEPTEMBER NOVEMBER 1923 1924 All unemployment Percentage Percentage reduction Percentage reduction in workreduction in workin working-class Excess Excess ing-class ing-class income over income income Percentage 1 9 1over Percentage 1912-1914 due to 2-1914 due to due to total and percentage percentage partial total partial unemunemunemployment s ployment ployment l Total unemployment Date 1923 November December 1924 January February March April May June July August September 23.4 28.2 20.4 25.2 14-15 17-19 Partial unemployment 47.3 42.0 15-17 13-15 29-32 30-34 18.4 22-26 6-8 4-5 19-22 12.1 1-3 10-13 4.9 0-2 5-8 0.8 1-2 5-6 3.2 9-12 4-6 14.4 13-16 7-9 23.2 13-16 7-9 22.5 | 9-12 4-6 12.5 1 Two figures are given, one calculated from the wages of unskilled workers (lower figure), the other from those of skilled workers (higher figure). 2 Two figures are given. Up to May 1924 the lower figure is based upon the percentage of partial unemployment over and above that of 1912-1924 and the higher figure on the absolute percentage of partial unemployment for the given period. From May 1924 onwards, from which date official relief was continued in part only by the local authorities, the lower figure is based on the percentage of partial unemployment over and above that of 19121914, assuming unemployed relief to have been maintained in full; the higher figure is based on the full percentage of unemployment assuming relief to have been completely withdrawn. 1 26.5 25.1 16.6 10.4 8.6 10.5 12.5 12.4 10.5 See Appendix III 23.5 22.1 13.6 7.4 5.6 7.5 9.5 9.4 7.5 16-18 15-17 9-10 5-6 4 5-6 6-7 6-7 5-6 23.4 17.1 9.9 5.8 8.2 19.4 28.2 27.5 17.5 42.3 37.0 — 24 — and its real value is so difficult to ascertain t h a t no estimate approaching accuracy seemed possible. A rough estimate will be given later. The figures given in table VII are merely approximate estimates. At the same time they suffice to show t h a t the reduction caused by unemployment in the income of the entire industrial population was considerable; it amounted to something like one-third in November and December 1923 and varied between one-fifth and one-quarter during the first two months of 1924. In March a sudden improvement occurred, and in April and May the reduction in the income of the whole of the industrial population was less than 10 per cent., while it slightly exceeded this figure during the summer months. For the first ten months of 1923, although no very accurate calculation of the percentage reduction can be made, a rough estimate based directly on the unemployment percentage will give some indication of its extent. Assuming unemployment relief to have been nil and the loss of work due to partial unemployment to have been one-third of the normal day, the diminution of income due to unemployment over and above that of pre-war days cannot have exceeded 8 or 9 per cent, during the period January to August 1923* ; from September onwards the reduction grew steadily greater until it reached its climax in.December. The various statistics and estimates presented in this chapter now provide a basis for a final estimate of the changes in the income of the industrial population. In table VIII these are shown by approximate index numbers, obtained by applying to the index numbers of real wages of industrial workers in normal employment 2 the percentage reduction in the income of the industrial population caused by unemployment. These figures are, of course, to be regarded simply as approximate indications of the development of an exceedingly complex phenomenon, seeing that many reservations have had to be made at 1 During this period the average aggregate figure amounts to 5 per cent* for total unemployment and 19 per cent, for partial unemployment. Subtracting 3 per cent, from the percentage of total unemployment so as to obtain the excess of unemployment over that of pre-war days, and assuming that the partially unemployed lose, on an average, one-third of their ordinary wages, the result is 2 per cent. + 6.5 per cent. = 8.5 per cent. 2 To apply the same percentage reduction to skilled and unskilled workers presupposes that unemployment affected both classes equally. It is probable, however, that unskilled workers have been more severely affected than skilled. The higher figure for the aggregate income of the industrial population (calculated from the index number of real wages of unskilled workers) should therefore be reduced on the whole, and the lower figure (calculated from the index number of the real wages of skilled workers) somewhat increased. It seems, therefore, that the exact figure falls within the limits of the estimate. — 25 — every turn. It should be remembered, moreover, that up till November 1923 the percentage reduction through unemployment is obtained by a simple estimate and that up to 1922 the data for TABLE VIII. ESTIMATE OF THE AGGREGATE INCOME OF THE INDUSTRIAL POPULATION, 1920 TO 1924 (Base: 1913-1914 = 100) Date Approximate index numbers 1920: Yearly average 1921: » » 1922: » » 1923: Average for the period January to August 1923: November December 1924: January February March April May June July August September 60 - 85 i 75 -105 l 70-90 1 58-76» 36 - 47 46 - 60 55 - .69 61 - 75 69 - 81 75 - 85 81 - 90 80 - 91 76 - 85 76 - 87 78 - 88 i These figures are the results of a very rough approximation. real wages themselves are of very uncertain value. For 1920 and 1921 in particular the statistics by no means represent all industries and are compiled from very heterogeneous sources. From November 1924 onwards the data upon which the estimates are based are more satisfactory; they are none the less liable to error in various ways and the arbitrary element in their composition must not be overlooked. The greatest caution should be exercised in drawing any conclusions from these figures. They nevertheless show fairly clearly: (a) the gradual improvement in the position of industrial workers up to 1921; (b) the continuous downward trend of conditions in 1922 and during the first months of 1923, owing to inflation ; (c) the extreme distress which developed in the autumn of the same year (culminating in November) under the twofold influence of a depreciating currency and growing unemployment; (d) the immediate relief brought by stabilisation in December 1923; (e) the gradual improvement up to May 1924, thanks to the rise in real wages and the decline in unemployment; and (/) finally, from June onwards, the pause in development at a level still from 10 to 25 per cent, below that of the pre-war period. It was inevitable that so great a reduction in the income of an important portion of the population should have far-reaching — 26 — repercussions on general living conditions. Unfortunately it is extremely difficult to measure by precise statistical methods the changes which occurred. The few figures provided by the Ministry of Labour and given below merely illustrate the importance of these changes. TABLE IX. CONSUMPTION OF CERTAIN FOODSTUFFS PER INHABITANT IN THE WHOLE OF GERMANY BEFORE AND AFTER THE WAR Article i Meat Cocoa Coffee Tea Beer Milk 1913 43.15 0.78 2.50 0.06 102.10 106.90 1923 21.80 2 0.80 0.61 0.04 50.00 62.90 1 Quantities are given in kilogrammes for the first four articles and in litres for the last two. 2 Figure for 1922. It will be seen that there has, on the whole, been a notable reduction in the consumption of these foodstuffs — except in the case of cocoa, which shows a slight increase. It should be noted, however, that the articles considered — and considered merely because they happen to be the only ones for which statistics are available — only account for a relatively small part of total consumption. Some idea of the general state of health can be gained from the death rate from tuberculosis. It has gradually increased from 1.5 per thousand in 1921 to 1.6 in 1922 and 1.7 in 1923. These figures, which refer to the 46 towns with more than 10,000 inhabitants and have been furnished by the Ministry of Labour, indicate a progressive decline in the physical condition of the population. The birth rate also has fallen greatly. The general figure for the whole country fell from 28 per thousand in 1913 to 26 in 1920 and 1921, to 24 in 1922, and to 22 in 1923. At the same time it is true that a gradual fall was recorded during the thirty years preceding the war 1. It is thus difficult to say how far the diminution now 1 The b i r t h r a t e s were as follows : Date 1871-1880 1881-1890 1891-1900 1901-1910 1911-1920 1903 1904 1905 Birth rate per 1000 40.7 38.2 27.3 33.9 22.7 34.9 35.1 34.0 Date 1913 1914 1915 1916 1917 1918 1919 1920. 1921 1922 1923 Birth rate per 1000 28.3 27.6 21.0 15.7 14.4 14.7 20.7 26.7 26.1 23.6 21.6 S O U R C E S : Statistisches Jahrbuch für das Deutsche Reich, 1923; a n d report of he Ministry of Labour to t h e International L a b o u r Office. — 27 — recorded continues the former general tendency and how far it is the consequence of exceptional economic conditions. POSITION OF PROFESSIONAL WORKERS, 1920 TO 1924 Information of practical value concerning salaries of professional workers is extremely meagre. It is confined to particulars of the salaries paid to state officials and bank employees from 1920 to 1924 (these particulars are given in table X) and a few data respecting the salaries of commercial employees attached to various industries for the period December 1923 to July 1924 (table XII). These categories cannot be regarded as truly representative of all professional workers. They are therefore given here as examples only. TABLE X. INDEX NUMBERS OF REAL SALARIES FOR CERTAIN CLASSES OF PROFESSIONAL WORKERS, 1 9 2 0 TO 1 9 2 4 Bank employees i State officials Date Unskilled 1920 1 February 1 April 1 August 1921 1 January 1 July 1 August 1 October 1922 1 January 1 April 1 July 1 October 1923 January April May June JulyAugust September October November December 1924 J anuary February March April May June 1 Skilled Unskilled Skilled Higher grades — — — — — 77 51 •— — 54 — 88 115 58 73 80 50 36 — — — 82 76' 82 97 48 51 65 35 37 54 — •— — — — — 77 62 75 69 50 37 45 46 41 30 36 37 75 84 105 114 179 113 117 49 53 70 76 120 76 96 — — — — — — 61 63 66 77 73 79 78 73 53 62 40 42 44 51 51 56 55 51 38 49 33 34 35 41 42 46 46 43 32' 44 68 70 68 81 80 95 53 55 54 64 63 82 48 49 48 57 56 79 83 147 104 51 89 64 99 81 — — — — — ~ • — — •— ~ Up to April 1922 the figures refer to the towns in the highest cost-of-living class; from July 1922 onwards they apply to Berlin alone. Conditions in the former class of town correspond; however, to those in the capital. — 28 — The position of these workers shows marked differences in comparison with that ascertained in respect of industrial workers. In the beginning of 1920 the salaries of bank employees had already depreciated further than the wages of manual workers. At that time the salaries of skilled salaried employees were only 50 per cent. of their real pre-war value, and those of unskilled employees somewhat less than 70 per cent., whereas the real wages of manual workers fluctuated between 55 and 70 per cent, for skilled workers and 70 and 80 per cent, for unskilled workers. The position of public officials was somewhat better, the unskilled grades being particularly favoured ; the real value of their earnings was practically the same as before the war, while those of skilled officials were still about 80 per cent, of their pre-war value. In the upper grades, on the other hand (there are no data for the corresponding classes of bank employees), the real value of salaries fell considerably, and was barely 45 per cent, of the pre-war value. In the following years, however, the position was completely reversed; the real value of bank employees' salaries rose progressively up to the end of the year 1923, whereas in the case of officials it steadily decreased. These two opposite tendencies are no doubt obscured by fluctuations which were of monthly occurrence and particularly affected bank employees, but they emerge quite clearly from the yearly averages shown in table XI. TABLE XI. AVERAGE INDEX NUMBERS OF REAL SALARIES OF CERTAIN CLASSES OF PROFESSIONAL WORKERS 1 9 2 0 TO 1 9 2 4 {Base: Pre-war = 100) 1920 Class of workers Average for several months 1921 1922 1923 1923-1924 Average for several months Average for April, July, October Average for period Jan. to Nov. Average for period Dec. 1923 to Sept. 1924 Skilled workers Bank employees State officials (lower grades) State officials (higher grades) 53 66 68 77 48 » 55 45 48 60 35 ! 42 36 39 54 102 111 112 86 71 69 ' Unskilled workers Bank employees State officials (lower grades) 80 76» 75 » Figure drawn from a single return during the vear. For the month referred to, see table X. — 29 — From December 1923 onwards a rise was experienced; it persisted throughout the first half of the following year, so that in June 1924 the real value of officials' salaries was only from 5 to 10 per cent, below pre-war rates. The difference was greater in the case of higher officials, but in June 1924 their salaries were none the less appreciably higher than in 1923. For bank employees other statistics must be used, as those given in table X do not go beyond January 1924 1 . To this class of worker also the year 1924 brought a measure of improvement; in July the real value of their salaries must have been about at its pre-war level. This tendency to rise during the first six months of 1924 is shown, though somewhat irregularly, in the salaries of the other classes of commercial employees for which information is available. It should be noted, however, that in general the same cost-ofliving index number has been used to calculate the real value of professional workers' salaries and for manual workers' wages. As this index number is based upon the working-class standard of life, it considerably underestimates the increase in the cost of living for the professional classes. The actual real [value of salaries, therefore, probably falls short of that shown in the tables. To gain a somewhat more accurate idea of conditions among professional workers it would be further necessary to take into account the extent of their employment, as was done for manual workers. There are, however, no statistics indicating even approximately the percentage of unemployed at each period. The figures given by official employment exchanges nevertheless show that, though relatively small in 1922, the number of professional workers in quest of employment grew appreciably during the following year. Proportionately the increase is slighter than among manual workers ; it may consequently be assumed that unemployment was roughly the same in both classes. It is, moreover, probable that the lack of employment was more severely felt among professional workers. In the first place, unemployment relief, being paid at the same rates as to manual workers, represented a yet smaller percentage of their salary ; in the second place, their trade union organisations were in most cases unable to afford them even such slight assistance as was granted by the workers' organisations, during part at least 1 Among the various categories of commercial employees whose salaries are given in table XII, the only one which corresponds approximately to that of skilled bank employees shown for the preceding period is that of skilled accountants. — 30 — of the period of depression. In the first months of 1924 there appears to have been a slight decline in unemployment; but in May the situation once more became critical for this class of workers, though it was not so grave as in December. On the whole it may be said that, while the real salaries of bank employees for the period 1920-1922 can be taken as indicating roughly the aggregate income of this class, their standard of life during 1923 must have been lowered by unemployment despite the real increase in salaries among those who were still in employment. TABLE XII. INDEX NUMBERS OF REAL SALARIES OF CERTAIN CLASSES OF COMMERCIAL EMPLOYEES IN BERLIN, MARCH AND JULY 1924 (Base: December 1913 = 100) Date Chemi- TexMetal cal tile indus- indus- industry try try Wholesale Retail textile trade trade Insurance Federal Bank- government ing oflices Clerks x 1924 March July 119 111 958 103 121 112 110 102 Shorthand-typists 118 121 122 121 147 149 152 122 117 110 141 103 105 148 152 — 103 105 154 158 122 118 114 116 110 2 1924 March July 112 102 106 112 120 112 110 102 959 980 Skilled forwarding clerks 3 1924 March July 135 ¿123 101 108 120 112 110 102 — Skilled accountants 4 1924 March July 127 121 102 109 120 112 110 102 — i Clerks aged 26, entrusted with simple office work. a Experienced shorthand-typists, aged 24, filling a position not entailing responsibility and8 not requiring knowledge of foreign languages. Forwarding clerks aged 26 filling responsible positions. * Accountants aged 30 responsible for keeping the books. — 31 — The position was probably similar among other classes of professional workers in private undertakings. As for public officials, they may be said to have practically escaped the crisis and suffered only a reduction in salary. At the same time, the financial reforms introduced at the end of 1923 entailed a considerable reduction in the number of officials; those who were dismissed were reduced to living on their pensions until they could find other work — hence a further and considerable reduction in their income. CHAPTER II Cost of Living Problems 1 Before discussing the methods which were adopted for adjusting wages to the cost of living, those taken to influence the cost of living itself, or simply to measure it, will be considered. The Government devoted itself more particularly to this side of the question, leaving it to private organisations to devise the methods of actual adjustment of wages. From the outbreak of war the Government endeavoured to counteract the rise in the cost of living, knowing t h a t this was the most certain way of forestalling wage disputes. Moreover, by artificially maintaining the purchasing power of the mark in spite of repeated issues of paper money, the Government kept the purchasing power of its own resources, derived from the issue of paper money, much higher than it would otherwise have been. Its systematic action chiefly took the line of fixing the prices of articles of prime necessity and rents, measures which were combined with the rationing of foodstuffs and the policy adopted to meet the housing shortage. PRICE CONTROL - Control of the prices of the more important foodstuffs was introduced during the war on a very large scale. At that time all articles of current consumption were rationed and their prices fixed officially. After the revolution this system gradually disappeared, either because official regulations were no longer observed and quantities of foodstuffs were sold in illicit trade, or because the Government deliberately withdrew its control. The prices of a certain number of the most important foodstuffs, such as bread and sugar, were, however, still controlled for several years, and the system was not given up till 1922 and 1923. It is very difficult to determine exactly how effective these measures proved. During the war they were certainly quite effective and helped to keep the level of prices below what it should 1 In the compilation of this and the following chapters considerable use has been made of an article by Dr. Fritz SITZLER: "The Adaptation of Wages to the Depreciation of the Currency in Germany", in the International Labour Review, Vol. IX, No. 5, May 1924, pp. 643-666. — 33 — have been in view of the food shortage and the inflation of the currency. After the war the effects were much more doubtful, for by that time illicit trade had become very widespread and illicit prices were much higher than the official ones. It is equally uncertain whether these measures of price control proved effective in maintaining the real value of wages. It is obvious that the adjustment of wages was made easier by this holding back of the rise in the cost of living. As will be shown later, part of the depreciation of real wages is due to the fact that wages can be adjusted only slowly and at considerable intervals. As long as the rise in the cost of living is not very rapid the depreciation in real wages due to the lag in adjustment is not too marked. Moreover, as wage increases are not very large in absolute figures, they are generally obtained more easily, since employers can easily balance the increased cost of production by raising the prices of their products to correspond. On the other hand, however, it should be observed that the conditions created by the system of price control make any measurement of the cost of living extremely difficult. The method adopted by the Federal Statistical Office in constructing its index number (that of using official prices for the government ration and illicit prices for the additional quantities in the theoretical budget used for the index number) is arbitrary, since it was frequently impossible to obtain the rationed amounts on the official market, and they consequently had to be bought in the illicit market at much higher prices. The cost of living was therefore in reality higher than the current estimate based on official prices or even on the cost-of-living index number. It is difficult to say how far these two tendencies balanced each other; the answer depends very largely upon local and often temporary conditions. When control was withdrawn the rise in prices was much accelerated for some months and steadily tended towards the natural level corresponding to economic and financial conditions. The adjustment of wages at this time was therefore much more difficult, but on the other hand it became much easier to measure the cost of living. RENT CONTROL Government measures for controlling rents were more effective and perhaps more important than those for controlling the prices of foodstuffs. As soon as the housing shortage became pronounced during the war, the Government began to take steps to protect 3 — 34 — tenants against too rapid an increase in rents. This policy was maintained and even intensified after the war, in the hope that under peace conditions it would be possible to settle the housing problem by new building. But owing to economic disorganisation this programme could not be carried into effect, and rent control had to be very strictly maintained in order that there should not be a sudden and substantial rise in the cost of living. It is not necessary to consider in detail here the complicated legislation adopted to prevent a rise in rents 1 ; the main point is that it proved remarkably effective. Until the end of 1923 rent increaseswere extraordinarily slight as compared with the enormous rises in other prices. Towards the end of 1923, however, further measures were adopted to enable landlords to meet the cost of upkeep and also to place a tax on tenants. Expenditure on housing (rent and taxes) thus increased rapidly, and continued to do so throughout 1924. The following table illustrates the importance of these two movements and their influence on the cost of living. It gives the rent index number derived from the official cost-of-living index number and the index number for all items other than rent (food, clothing, heating and lighting) included in the general cost-ofliving index. Finally, the proportion of expenditure on rent to the total budget is given; the budget referred to is that used for constructing the cost-of-living index number, increased by 15 per cent, to cover various items of expenditure not included in it 2. During the first years after the war there was practically no rise in rents. By July 1921 (the first date for which rent figures relating to the whole of Germany were available) rents were at most double those paid before the war, while the prices of other articles of current consumption were already 16 times above the pre-war 1 For a detailed examination of this question, cf. INTERNATIONAL LABOUR OFFICE : European Housing Problems since the War. Studies and Reports, Series G (Housing and Welfare), No. 1, Geneva, 1924. XII + 484 pp. 6s.; $ 1.50. 2 If pre-war prices are applied to the theoretical budget used for the construction of the index number, which is based on post-war conditions, the expenditure on rent is 23.7 per cent, of the total expenditure. If this total is increased by 15 per cent, to cover the various items not included in the theoretical budget, the proportion spent on rent is reduced to 20.4 per cent. In order to obtain the proportion at the different dates this percentage of 20.4 has been multiplied by the ratio between the rent index number and the cost-of-living index number at the date under consideration. Thus for July 20 4 X 2 1921 the percentage is —^-=— = 3.3. — 35 — TABLE I. RENT AND RETAIL PRICES OF OTHER ARTICLES OF PRIME NECESSITY, 1913-1914 TO 1924 Official index numbers (base: 1913-1914 = 1) Date Rent Food, light and beat, clothing Expenditure on housing per cent. of total expenditure of typical workingclass family Calculated from p a p e r - m a r k prices 1913-1914 Average 1 1 20.4 July 2 15.80 3.3 1922 January July 2.4 3.4 26.00 69.58 2.4 1.3 1921 1923 January July August September October November December 38 714 4,932 300,000 54,000,000 22,000,000,000 218,000,000,000 1,453 48,986 764,389 19,500,000 4,763,000,000 852,000,000,000 1,562,000,000,000 Calculated from r e n t e n m a r k prices 1924 January February March June September 0.290 0.340 0.380 0.570 0.700 1.350 1.260 1.28Ò 1.290 1.300 0.7 0.4 0.2 0.4 0.3 0.7 3.6 1 5.4 6.7 7.2 10.4 12.3 1 To convert the index number calculated from rentenmark prices into one for papermark prices, the decimal point should be moved twelve places to the right, as one rentenmark equals 1,000,000,000,000 paper marks. level. During 1922 and until about the autumn of 1923 the rise in rents was relatively rapid, but less so than that in the prices of the other articles considered. In September 1923, however, rent increases assumed enormous proportions, and the difference between the two classes of expenditure tended to diminish. During 1924 prices other than rent, taken as a whole, underwent a number of fluctuations which kept them at a practically stationary level, while rents rose almost uninterruptedly. The effects of the policy of rent control on the cost of living are clearly indicated by the figures given in the last column of the — 36 — table, showing the proportion between expenditure on rent and the total working-class budget. Thus in July 1921 this proportion was only 3.3 per cent., although before the war it was taken as 20.4 per cent. It continued to fall, with a few slight checks, until August 1923, when the minimum of 0.16 per cent, was reached. In general it mayì)e said that from the middle of 1922 until November 1923 rent was practically negligible in the total cost of living. From December 1923, however, the proportion rose again. In September 1924 rent represented 12.3 per cent, of the working-class budget, i.e. more than half the pre-war proportion 1 . Nevertheless, though government rent control was quite effective in itself, it was not directly advantageous to the workers ; as the cost of housing was gradually reduced, the part of wages intended to cover this expenditure similarly disappeared. It is true that the gradual disappearance of the burden of rent kept back the rise in the cost of living during a certain period, and the workers may therefore have been able to obtain a more complete adjustment of their wages to cover the other items of expenditure. If rents had been left free to rise, the workers would have had to obtain even larger increases in wages, and in view of the difficulties which they would have met they would no doubt have been compelled under the circumstances to be satisfied with an even more imperfect adjustment of wages than that which they actually accepted. This advantage disappeared as soon as rent represented only a negligible part of the cost of living, namely, from the middle of 1922 till November 1923. During this period the rise in the prices 1 It is necessary to make some reservations as to the significance of the figures on the importance of expenditure on rent as compared with total expenditure. They are only valid if it is accepted, first, that expenditure on rent represented before the war about 20 per cent, of the total expenditure of a working-class family, and secondly that the workers have been able to maintain their pre-war standard of life unchanged. This figure of 20.4 per cent, may at first sight seem rather high, but it has been adopted by the Ministry of Labour for calculations similar to those carried out here (Wirtschaft und Statistik, No. 5, 1924), and it is not very far from the average proportion of 18 per cent, obtained as the result of an enquiry into working-class budgets undertaken by the Ministry of Labour in 1907 (Erhebung von Wirtschaftsrechnungen minderbemittelter Familien im Deutschen Reiche, 1907. 2. Sonderheft zum Reichsarbeitsblatt, 1909). It would therefore seem admissible to use it as an average proportion, perhaps a trifle high. As regards the standard of life of the workers, it is clear that it has been considerably lowered, as has already been shown in chapter I. If it is assumed that housing conditions have remained unchanged, the fall in the standard of life has taken place solely in the other items of expenditure, which would be equivalent to a modification (sometimes fairly marked) in the proportionate expenditure on rent. — 37 — of articles of prime necessity was no longer balanced by the progressive reduction in the burden of rent, for this burden had already been practically eliminated. The rise in the cost of living (excluding rent) therefore exerted its full influence. The workers had practically no rent to pay, but they no longer received the part of wages intended for the payment of rent. The remaining part of wages had therefore to be proportioned directly to the rise in the prices of articles of prime necessity. Conversely, from September 1923, and especially from December 1923 onwards, when the payment of rent was gradually imposed again on the workers, they were compelled to obtain corresponding rises in wages, if their standard of life was not to be lowered. From the point of view of production, on the other hand, considerable advantages were derived from rent control. Since wages nò longer had to cover rent the cost of labour was correspondingly reduced. This advantage reached its maximum when rent practically disappeared as an item of the working-class budget. At that date the cost of labour was reduced by 15 to 20 per cent., which obviously proved favourable to production, while at the same time the workers benefited indirectly, for without the reduction they would have been unemployed, or at least unemployed earlier. This was probably the most important advantage accruing to the workers from rent control. The cost of maintenance, renewal, and extension of housing, which under normal economic conditions is met out of the national income, was only partly covered in this way, the balance being charged to capital invested in housing, which depreciated and was not developed. The revenue which under normal conditions landlords similarly drew from the national income either totally disappeared or was transferred to other parties. These very serious defects it was which induced a return to a sounder housing economy. But this increase in rents did not go in its entirety to the landlords, as part of it had to be paid out in taxes. Furthermore, from the point of view of production the cost of labour was increased though the workers' standard of living was not raised. MEASUREMENT OF THE COST OF LIVING Besides taking steps to counteract the rise in the cost of living^ the Government also undertook to publish an index number to give as exact and impartial a measure of the cost of living as possible, and thus to facilitate the adjustment of wages. Not that — 38 — such index numbers were absolutely wanting in Germany; on the contrary, this step was necessitated by their number. Many private statisticians, as well as local authorities or joint committees appointed by industrial organisations, published monthly index numbers of the cost of living calculated on the most varying methods, some being satisfactory, others quite inadequate. The results of these calculations naturally varied considerably, and this could not fail to discredit index numbers in public opinion as a measurement of the cost of living. An index established uniformly for the whole country offering full guarantees of impartiality was therefore necessary. The Federal Statistical Office, assisted by the municipal statistical offices, undertook to cover the deficiency. It would be interesting to study in detail the various cost-ofliving index numbers, both official and private, which are published in Germany, and their degree of accuracy 1 . Systematic examination from this point of view of the principal private index numbers and those drawn up by the local authorities, however, would take us too far, besides which the data for the necessary comparisons are very often entirely lacking. Attention is therefore here confined to the official index number, which is of particular importance. It played a fundamental part in the adjustment of wages, although it would be very difficult to state exactly how far the national index number and the local indexes, private and official, have been used in wage negotiations. Assuming that most national and regional 1 Among the best-known private indexes may be mentioned that of Calwer, covering 16 foodstuffs only, weighted according to the pre-war ration of a German marine, and calculated for 200 large towns. Similar calculations have also been made with a reduced scale of rations more in keeping with post-war conditions. Dr. Elsas extended the Calwer index by adding the prices of certain articles of clothing in Berlin and Frankfurt, and subsequently increasing the aggregate expenditure by a certain percentage to cover the cost of rent, heating, and sundries. Dr. Kuczynski, Director of the Berlin-Schöneberg Statistical Office, undertook to determine monthly the minimum cost of living for a working-class family in that town ; his calculations cover foodstuffs (the various items being combined in such a way as to obtain a given nutritive value, for a minimum price), articles of clothing, light and heat, and rent, the total budget being subsequently increased by 25 or 30 per cent, to allow for sundries. Among the indexes regularly drawn up by local authorities are those of Breslau, Cologne, Erfurt, Hanover, Kiel, Ludwigshaven, Magdeburg, Nuremberg, Flensburg, Halle, Hamburg, and Stuttgart. In constructing these indexes the data considered are usually those employed for the Federal Statistical Office's local indexes, which are secured by the municipal authorities. The data, however, are supplemented or modified to suit local conditions or include articles excluded from the official index, more particularly articles of clothing and certain other items such as: taxes and subscriptions, upkeep and renewal of furniture and household utensils, cleaning, travelling, education, amusements, etc. — 39 — collective agreements have been based on the national index number — in so far as an index number has been used at all — and most local agreements on a local index, it may be stated, according to official statistics of collective agreements 1 , that the national index number has been used for more than 80 per cent. of the workers and local indexes for less than 20 per cent. This pi'oportion is obviously a little more unfavourable to the latter than the actual figure, for local indexes are used in many regional agreements. On the other hand, they are very numerous, and the importance of each is therefore comparatively limited. The publication of the official index number of the cost of living was begun in February 1920. At first the prices of 13 articles of food and two items of heating and lighting, and the rent of two rooms and a kitchen were systematically collected through the local statistical offices for over 600 towns with a population of over 10,000. At that time state control of prices was still fairly general. The official prices were therefore recorded for controlled articles, but where the official ration was less than that used for the construction of the index number the extra amount was calculated at prices on the open market or even in illicit trade. The latter operation was clearly difficult, the accuracy of the result depending largely on the technical qualifications of the individual municipal office. Further, certain foodstuffs of a seasonal nature, which were not to be found on the market at certain periods, were replaced by other so-called substitutes of corresponding nutritive value. Similar measures had to be taken when, owing to the disorganisation of transport, even essential foodstuffs were not available for some time on certain local markets. The weighting of the index number was based on a theoretical budget representing the four-weeks consumption of a family of five (father, mother, and three children aged 12, 7, and 1% respectively) under post-war conditions. The general index for the whole country was obtained by first averaging the prices of each article in 39 towns for which data were rapidly available, the average being weighted by the population of each town, and then taking the average of these average prices, weighting them according to the budget. The average was limited to this small number of towns, so that the national index could be published more speedily; but indices were calculated for 1 Reichsarbeitsblatt, Nos. 3 and 5 1923; Mitteilungen aus dem Tarijwesen; eie Tarifverträge im Deutschen Reiche am Ende des Jahres 1921. See also Chapter III. note to p. 45. — 40 — the 600 towns, taking for each of them the average of the recorded prices of the different articles, weighted uniformly according to the theoretical budget. These index numbers constituted an advance, but they were still imperfect and were soon criticised. In January 1921 they were slightly modified and the national index was extended to 47 towns. A more far-reaching change was made in March 1922 as the result of a conference between the officials of the municipal statistical offices. The number of towns covered by the national index was increased to 71, care being taken to select towns of varying sizes in different parts of the country, and in both industrial and agricultural districts. Experience has shown that this selection gives results perfectly representing the fluctuations recorded in the 600 towns. In these 71 towns prices were recorded twice a month instead of once. Illicit prices were no longer recorded, and the introduction of substitutes for seasonal foodstuffs was reduced to a minimum and made more systematic. The method of recording rents was improved. In April 1922 the index was completed by the inclusion of the "clothing" group, which had hitherto been omitted from the calculation owing to the special technical difficulties involved. The prices of some 15 articles were recorded regularly in the 71 towns, and the annual expenditure of the standard family was estimated on the basis of the current prices of the month, such expenditure being then divided by 13 to obtain the figure for the four weeks under consideration. The index for the clothing group was published only for the 71 towns as a whole and was not included in the local indexes. In view of the extremely rapid rise in prices it was soon found necessary to record them four times a month instead of only twice, and from September 1923 the pressure of labour organisations resulted in a reorganisation, so that a weekly index could be published on Wednesday evening, based on the prices recorded on the preceding Monday 1 . The monthly index was still published, but the method of compilation was again revised. An index was calculated by interpolation for each day of the month on the basis of the weekly indexes for that month and for the week preceding and following the month, on the assumption that prices increased daily in geome1 Weekly indexes had been published since April 1923, but until September there was delay in publication. — 41 — trical progression during the interval between two weekly records. An average was then taken of these 30 or 31 daily indices to obtain the monthly index. It may be useful to examine the effects of these revisions on the resulting calculations, so as to determine how far the original data suffered from error. This is easy, for on each occasion the index number was entirely recalculated as from 1920. The figures are given in the table on page 42, and in order to facilitate comparison between the fluctuations of the indexes, the percentage monthly increases in the principal series are also given. The revision of January 1921 was of no practical importance because it affected little but the method of calculating the national index number, which was extended to include the 47 towns. Since the indexes of the various towns remained unaltered the difference was not marked. The change effected in the spring of 1922 was much more important. The first step was to improve the collection of prices and extend the national index to 71 towns. This led to a considerable increase in the national index, due not to the inclusion of some twenty more towns but rather to the change in the method of calculation, for most of the local indexes were found to have risen, as will appear on detailed examination. During 1920 and 1921 the old index is about 10 per cent, below the new one, and the difference was slightly accentuated during the first months of 1922. On the other hand, an examination of the monthly fluctuations will show them to have been fairly similar as a whole. It is more interesting to compare the old index with the new index number in which the price of clothing is included. The difference between the two indexes is decidedly more marked, the old figure being about 25 to 35 per cent, below the new one. This difference tended to diminish, for in a general way the rise of the old index number during the period under consideration was more rapid than that of the new one. If it is agreed that the new index number including clothing constitutes the most exact available measure of the cost of living, it will be clear that the use of the old index number for adjusting wages definitely underestimated the increase in the cost of living as compared with the pre-war period. On the other hand there was a tendency slightly to overestimate the monthly rise after the war. Doubtless this overestimate would eventually have counterbalanced the underestimate, but the old index was not continued long enough for this result to be achieved. — 42 — TABLE II. OFFICIAL INDEX AT DIFFERENT NUMBER OF THE COST OF LIVING AS PUBLISHED DATES AND COMPARED WITH THE REVISED INDEX {Base: 1914 = 1) Percentage monthly increase ( + ) or decrease (—) of of the index the index of number number the index revised revised number in March in April 1922 revised 1922 (exclud- (includin ing ing 1921 clothclothing) ing) Index numbers Date Original index number (39 towns) 1920 Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. 1921 Jan. Feb. Mar. Apr. May June July Aug. Sept. . Oct. Nov. Dec. 1922 Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. 1923 Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Index number revised in January 1921 (47 towns) Index number revised in March 1922 (71 towns) Index number revised in April 1922 to include clothing 6.24 7.42 8.37 8.77 8.44 8.44 7.97 7.78 8.28 8.73 8.28 6.23 7.41 8.36 8.76 8.42 8.42 7.95 7.77 8.27 8.72 9.16 7.00 8.17 9.16 9.75 9.53 9.35 8.87 8.81 9.45 9.99 10.47 8.47 9.56 10.42 11.02 10.83 10.65 10.23 10.15 10.71 11.18 11.58 — — — — — — — — — — — — 9.24 9.01 9.01 8.94 8.80 8.96 9.63 10.45 10.62 11.46 13.97 15.50 10.70 10.33 10.28 10.22 10.14 10.48 11.24 11.92 12.12 13.08 15.94 17.46 11.79 11.47 11.38 11.27 11.20 11.67 12.50 13.33 13.74 15.04 17.75 19.28 — — — — — — — — — — — — 16.40 19.89 23.02 28.04 30.48 — — — — — — — • 18.25 22.09 26.39 31.75 34.62 37.79 49.90 70.29 113.76 195.04 400.47 611.56 20.41 24.49 28.97 34.BOSS.03 41.47 53.92 77.65 113.19 220.66 446.10 685.06 — — — — — — •— — — — — — •— — — — — — — — — • — —. — ' 1.034 1,120 2,408 2,643 2,627 2,854 2.764 2,954 3,521 3,816 6,979 7,650 33,300 37,651 508,631 586,045 13,200,000 15,000,000 3,265,000,000 3,657,000,000 633,000,000,000 657,000,000,000 1,182,000,000,000 1,247,000,000,000 — + 19 + 13 + 5 — 4 0 — 6 — 2 + 6 + 5 — 6 + + + — — — — + + + 17 12 6 2 2 5 1 7 6 5 — + 13 + 9 + 6 — 2 — 2 — 4 — 1 + 6 + 4 + . 4 + — 1 2 0 — 1 — 2 + 2 + 7 + 9 + 2 + 8 + 22 +. 11 + — — — — + + + + + + + 2 3 1 1 1 3 7 6 3 8 22 10 + — — — — + + + + + + + 2 3 1 1 1 4 7 7 3 9 18 9 + + + + + 6 21 16 22 9 — — — — — — — + + + + + + + + + + + + 5 21 19 20 9 9 32 41 62 71 105 53 + + + + + + + + + + + + 6 20 18 19 11 9 30 44 72 66 102 54 — — — — — — — — 63 + 69 + + 133 + 136 8 + 9 + + 5 + 3 + 27 + 29 + 98 + 100 + 377 + 392 + 1427 + 1456 + 2495 + 2459 + 24635 + 24280 + 19287 + 17865 + 87 + 90 • — — — — .—• — 43 —• It should be pointed out that the index of the clothing group does not appear in the local indexes published by the Federal Statistical Office, since it is calculated separately for the whole country and subsequently added to the national index. These local indexes, however, are of subsidiary importance only, and to ascertain how accurately they measure the cost of living it is enough to consider them as a whole, i.e. simply to compare the national index number including the clothing group with the national index number excluding clothing. The latter remained during 1922 and 1923 at a level about 8 to 13 per cent, below the former, thus underestimating the increase in the cost of living as compared with before the war. The rise month by month of the index number without clothing is sometimes faster, sometimes slower than that of the index number including clothing, the difference displaying no particular tendency. There was thus no really satisfactory measure of the cost of living in Germany until the spring of 1922, and even then it applied only to the country as a whole. Before that date the national index number underestimated the change in the cost of living as compared with the pre-war level, but overestimated the monthly rise. The local indexes had the same defects as the national index, and after the latter had been revised they still underestimated the increase in the cost of living as compared with before the wrar. Attention should again be drawn here to the reservation made elsewhere as to the significance of this comparison. Since the index number is weighted according to a budget based on post-war consumption, the rise indicated corresponds to an increase in the expenditure needed to maintain a post-war standard of living since before the war. To maintain the pre-war standard of living would probably have meant a much more marked increase. The most recent reform in the calculation of the cost-of-living index number was to make it weekly so as to allow of more rapid adjustment of wages. This question had been important all along but it became absolutely burning from the middle of 1923 owing to the incredible rise in prices, and the reform was at last carried out. But the last column of the table shows that the reform was barely sufficient; during the last months of 1923 the weekly rise in the cost of living was greater than the monthly rise in preceding months. CHAPTER III Adjustment of Wages by Collective Agreements The measures taken by the Government either to arrest or to obtain an accurate measure of the increase in prices have been reviewed in the foregoing chapter. In the following chapters a description is given of the efforts made to adjust wages to the cost of living, the various methods adopted for the purpose, and the results secured. The first of these measures aimed at adapting the existing machinery for fixing wages to the new requirements arising out of the increasingly rapid rise in prices. The various methods, adopted provisionally at first in the hope that a drop in prices would shortly follow, were gradually systematised and improved ; from being the exception they soon became the rule. Thus the period of validity of collective agreements was by degrees curtailed and the annulment clauses, more particularly those concerning wages, became far more elastic. Similarly cost-of-living bonuses and family allowances were very generally adopted. These opportunist methods, born of the necessity for solving an immediate problem, had serious defects. It was not long before controversy arose around the devising of more systematic methods, but the practical results achieved were of small importance. Sliding scales were discussed at length, but abandoned owing to the numerous difficulties which they entailed. The "stable-value wage" and "gold wage" systems followed. They seemed to promise a solution of the problem, but the Government's financial reforms and the stabilisation of the mark made them superfluous even before they were generally introduced. LIMITATION OF PERIOD OF COLLECTIVE AGREEMENTS AND INTRODUCTION OF GREATER ELASTICITY It is well known how extensively the system of fixing salaries and wages by collective agreement has developed in Germany since the war; in 1921 more than fourteen million persons employed in industry, commerce, and agriculture (i.e. the large majority of the working classes) were working under conditions thus regulated, whereas the corresponding number before the war was only — 45 — one and a half million and in 1918 six million. These figures alone do not, moreover, show the full extent of the development. Agreements have been largely systematised and made more comprehensive; all conditions of work are now regulated by them. They are no longer mainly local agreements, as before the war; they cover large areas or even the whole country. Usually concluded between workers' and employers' organisations and backed by their authority, and liable to be declared compulsory by the Ministry of Labour, they form collectively a real code of labour 1 . It was this huge system which had to be adapted by degrees to the new conditions created by rapid inflation. For detailed study and explanation of the very clearly marked development which took place, the chief sources used were the statistics of collective agreements during 1920 and 1921 2 published by the Ministry of Labour and a wide selection of agreements, whose principal clauses are published by the Ministry and have been carefully analysed 3 . Needless to say, this transformation went forward by stages and did not take place in a uniform manner. The methods adopted vary with the traditions and habits of each industrial group, the nature of the relations between employers and workers, and the economic position of the groups concerned. Two main tendencies are, however, to be discerned. In the first place, the clauses dealing exclusively with wages were gradually detached from the remainder of the agreement. This process was carried out in two stages, as it were. To begin with the annulling or revision of these clauses, independently of the remainder of the agreement in which they were embodied, was facilitated. The most uncompromising method of doing this 1 This development is clearly shown by figures taken from statistics of the Ministry of Labour: ("Mitteilungen aus dem Tarifwesen; die Tarifverträge im Deutschen Reiche am Ende des Jahres 1921", in Reichsarbeitsblatt Nos. 3 and 5, 1923; and Die Tarifverträge im Deutschen Reiche am Ende des Jahres 1922; 31. Sonderheft zum Reichsarbeitsblatt. This enormous extension of collective agreements meant no appreciable increase in the number of agreements; 10,885 were in force at the end of 1913, 11,488 at the end of 1921, and 10,768 at the end of 1922. The average number of persons covered by each agreement, on the other hand, considerably increased; it grew from 147 in 1912 to 1,121 in 1921 and 1,324 in 1922. In 1922 nearly 84 per cent, of the workers whose conditions of work were thus fixed were bound by agreements concluded between their organisations and those of their employers (65 per cent, in 1912). In 1922 agreements covering large areas affected 72 per cent, of the workers (50 per cent, in 1912) and those covering the whole country 14 per cent. (5 per cent, in 1912). 2 See sources quoted in preceding footnote. 3 Reichsarbeitsblatt, 1921, 1922, 1923, 1924: "Mitteilungen aus dem Tarifwesen. Neue Tarifverträge." For further details see appendix IV. — 46 — was to restrict the validity of the wage clauses to a shorter period than that stipulated for the remainder of the agreement. A less rigid method was to allow these clauses to be annulled before the agreement itself; such annulment might take place as from a date specified in the agreement, orbe conditional on any substantial change in the cost of living or general economic conditions, or again be subject to both conditions. By the middle of 1922, however, these intermediate forms had almost entirely disappeared. As a general rule wage questions formed the subject of special negotiations and agreements. A second tendency is the progressive limitation of the period of validity of agreements or at any rate of the wage clauses. So long as the wage clauses were an inseparable part of the agreement, the limitation applied to the whole agreement; subsequently it was restricted to the wage clauses, which had become independent of the remainder of the agreement. It thus became possible to revise wages at fairly frequent intervals according to the increase in the cost of living recorded since the last negotiations. This tendency was fairly general in all agreements, whatever the methods adopted to fix their period of validity. These varied with each agreement, but fall into three main types : (a) The period of validity was expressly stated in the agreement, with the proviso that this might be determined before the stipulated date of expiry in certain cases. (¿>) The period of validity was not expressly stipulated, but the agreement could be annulled in certain specified conditions. (c) Neither the period of validity nor the conditions for annulment were specified, but it was tacitly understood (or expressly stated) that the agreement would continue in force so long as economic conditions remained unchanged. It may be said on the whole that the chief method adopted for adjusting wages to the cost of living was simply to limit the maximum period of validity of agreements. This type of procedure became more general every year, the period of validity being at the same time progressively curtailed. Before the war agreements concluded for a year or more were very common. At the beginning of 1921 the average duration was from three to six months. By the beginning of 1922 it was only from one to three months, and from a fortnight to six weeks towards the end of that year. It was still further curtailed in the middle of 1923 owing to the fabulous increase in prices; agreements were seldom concluded for more than a week or a fortnight. The number of agreements — 47 — without any specified period of validity decreased progressively; some of them, however, provided for a minimum period of validity, which showed a steady tendency to decrease. The growing frequency with which these agreements had to be renewed inevitably brought about a simplification of the negotiation and conciliation procedure. The systems of conciliation through administrative channels, which were introduced as a temporary measure during the period of demobilisation, were maintained throughout that of inflation and grew by degrees into a regular wage fixing system. A similar process of simplification occurred in the procedure for making binding on a whole industry agreements ' which were concluded for part of it only. The Act of 23 January 1923 laid down that the previous publication of agreements which it was proposed to render compulsory should be optional whenever the adjustment of wages to monetary depreciation was the sole matter concerned ; this measure gave rise to no difficulties although it was applied in many cases. The compulsory application of certain collective agreements contributed greatly to ensure a more or less methodical increase in wages. In many cases, though not declared compulsory, the rates specified in agreements governing important industries or large areas were largely instrumental in determining the rates of wages in minor agreements in the same industry or region. Miners' wages, which from 1920 onwards were usually fixed with the assistance of the Ministry of Labour, and those of state employees, which were fixed in consultation with the Ministry of Finance, had a regulative effect on many other industries. The numerous codes of general principles drawn up by the large employers' and workers' associations also helped to secure fairly uniform adjustment. PROCEDURE OF A D J U S T M E N T : REVISION OF W A G E RATES, COST- OF-LIVING BONUSES, FAMILY ALLOWANCES, LOCAL COST-OF-LIVING ALLOWANCES Two chief methods may be distinguished in the adjustment, properly so called, of wages to the cost of living on the frequent occasions when collective agreements were revised: (a) the wage scales were completely revised and the rates increased as thought necessary; (b) the existing scale was maintained as a basic wage and costof-living bonuses were granted in one form or another to balance the increase in the cost of living. — 48 — Cost-of-Living Bonuses Cost-of-living bonuses were widely used during the war and the first years which followed it. At that time it was hoped that prices would shortly be stabilised. Bonuses were therefore looked upon as a temporary measure which could more readily be withdrawn when the cost of living went down than could an actual increase in wage rates. It was also possible by this means to avoid the lengthy and delicate negotiations involved in establishing new wage scales, some of which were very complicated ; time and labour were saved and possibilities of disputes avoided. When all hope of a speedy stabilisation of prices faded, the costof-living bonuses (which had been added one after the other and often formed complicated systems in themselves) were merged by degrees in the wage rates, and thereafter the latter were directly increased. The bonuses were nevertheless retained in some measure, as they afforded an easier solution of certain problems which had acquired added importance from the economic situation. The first and most discussed of these problems was this: were wages to be adjusted in full to the increase in the cost of living or not ? On the assumption that the national income had diminished and that the real value of wages should therefore be reduced correspondingly, the conclusion was that wages should be adjusted in part only to the rise in cost of living. Granting that secondary needs should be sacrificed before fundamental needs it followed that the adjustment must be so made as to ensure in all cases a minimum subsistence level, anything above this being sacrificed by degrees. The simplest and most rigid system for attaining this end was to grant cost-of-living bonuses at flat rates without distinguishing between skilled and unskilled workers and irrespective of the wages they could earn. Another slightly more elastic system was to give skilled workers a somewhat higher cost-of-living bonus than the unskilled, though the proportional difference was less than that between their respective basic wages. Similar results could of course be obtained, though at the cost of some complications, by an appropriate modification of basic wages. Even during the war both methods had been widely practised, and they continued to be applied during the years which followed. The consequence was a marked levelling up of wages, which was attended by the grave danger that skilled workers would lose any incentive to produce their utmost and young workers any encouragement to train for skilled trades. — 49 — It should further be noted that this tendency received a fresh impulse from another quarter. Before the war the wages of unskilled workers had been kept very low by the constant influx into the towns of rural workers accustomed to small earnings. After the war, however, this exodus of the rural population towards the industrial centres had greatly slackened, and in certain areas had even come to a standstill. Unskilled workers, therefore, found themselves in a stronger position than before the war to maintain their standard of living. It was soon realised that these unduly rigid methods were attended by serious dangers, and they were gradually abandoned. In 1923 the difference between the wages of skilled and unskilled workers, which had become very small, grew steadily more marked. The introduction of the gold wage at the end of 1923 stimulated progress along these lines, and in September 1924 the difference was not much less than that of pre-war days. A similar and even more marked development occurred among non-manual workers. Before the war the salaries of state officials — as also those of employees in private banks — varied even more according to grade and category than the wages of industrial workers. These differences were at first greatly reduced, but a reaction then took place, with a return to the old distinctions. At the end of 1923 or the beginning of 1924 the pre-war position had not yet been reached but it was not far distant. TABLE I. RATIO BETWEEN AND UNSKILLED WORKERS IN BETWEEN SALARIES NOMINAL THE WAGES PRINCIPAL OF HIGHER AND OF SKILLED INDUSTRIES1 LOWER GRADE AND EMPLOYEES IN GOVERNMENT OFFICES AND BERLIN BANKS, 1 9 1 3 TO 1 9 2 4 Date 1913 Average 1922 April October 1923 April October December 1924 January February March June September Wages Salaries of of skilled higher officials workers as percentage as percentage of salaries of wages of of lower unskilled workers officials Salaries of middle grade officials as percentage of salaries of lower officials Salaries of higher bank employees as percentage of salaries of subordinates 144 298 218 175 112 112 177 197 133 147 107 107 111 116 119 196 216 270 147 157 178 113 119 122 126 131 131 270 270 270 323 183 183 183 203 145 t The industries included in these statistics are: mining, building, wood, metals, textiles, chemicals, printing, and state workers (mainly transport); from January 1924 onwards the paper industry is also included. 4 — 50 — This two-fold movement is illustrated by the preceding table showing the ratio between the wages of skilled and unskilled workers in the principal industries and that between upper and lower grade employees in government offices and the Berlin banks. In each case and for each month the wage or salary of workers in the higher grade is expressed as a percentage of the wage or salary of workers in the corresponding lower grade. Family Allowances Another problem of a similar kind was that raised by differences in family responsibilities. The idea that married workers with children should be given a higher wage than single workers is not new, but it always came into collision with the principle of ''equal pay for equal work". It asserted itself with compelling force, however, when real wages had to be reduced owing to the shrinkage in national income and it was desired to effect this reduction in such a way that the most essential needs would still be provided for. The system most generally employed was that of granting family allowances, i.e. cost-of-living bonuses proportioned to the number of dependants. To start with they were paid at flat rates, without distinction of skilled or unskilled workers or of those earning high or low wages. Disadvantages similar to those already referred to in connection with the flat rate cost-of-living bonuses were experienced, and so, by degrees, family allowances likewise came to be graduated roughly according to wages. In certain cases the basic wage itself was fixed separately for single and married workers. This system did not find much favour among the workers; their point of view was expounded by Mr. Leipart, the President of the General Federation of Trade Unions 1 . He admitted that workers with large families suffered severely from the rise in prices, but at the same time thought it desirable that single workers should be paid wages enabling them to keep their parents or other persons having a claim upon them. He also pointed out that the expenses of single workers, especially those who do not live at home, are comparatively high; to pay them at lower rates than married workers while their output might even be greater caused grave discontent. He suggested that family allowances embodied in wages should be abolished, while the financial burdens of workers with families should be lightened by a reduction in taxes or by granting their children certain privileges (extension of educational 1 Cf. Korrespondenzblau, No. 34, 1921. — 51 — facilities, free industrial training, free meals at school, and even free clothing), the cost being defrayed by the community. The system of family allowances was not very widely adopted in Germany. It is certainly found in the majority of industries, but it was never universally adopted and the scale of allowances is usually low. As a means of providing for family expenses it was in any case quite inadequate; this can be seen from the following table, showing the ratio between the wages of married workers with two children and those of single workers in several branches of the economic system. TABLE II. AND RATIO BETWEEN THE WAGES AND SALARIES OF MARRIED SINGLE WORKERS IN CERTAIN INDUSTRIES, GOVERNMENT OFFICES BANKS, AND a (Wages or salaries of single workers = 100) April 1922 October 1922 April 1923 October 1923 Coal Mines Hewers and trolley-men Other underground workers Surface workers 111 113 112 107 108 108 111 112 113 109 110 111 Metals Skilled workers Unskilled workers 110 111 107 107 107 108 107 108 Chemicals Skilled workers Unskilled workers 106 106 103 103 105 106 104 105 Printing Hand compositors Assistants 104 104 104 104 104 104 Slate Workers Skilled workers Unskilled workers 124 126 117 119 122 123 129 134 State Officials Higher officials Subordinate officials 117 134 115 135 116 136 116 141 Bank Employees Higher grades Assistants 125 127 111 112 — Industry — i These figures indicate only the ratio between the wages of workers drawing family allowances (for wife and two children) and those of workers not drawing them. Thus, in April 1922 a skilled worker in the metal industry received on an average 772 marks per week if single and 853 marks if he had a wife and two children. The wage of a married worker with two children was thus 110 per cent, of that of a single worker in the same category. • — 52 — ' Family allowances were of most importance in state undertakings. They were introduced by the Act of 30 April 1920 for all officials of the central Government and remained in force, with some few modifications, from that time onwards. Monthly allowances were paid, irrespective of grade or wage, in respect of each official's wife and each of his children under 21 years of age, the allowances for children increasing with their age. A system on similar lines was applied to manual workers in state services, including the railways, but allowances were restricted to the wife and children under 14 years of age. No distinction was made between the various classes of workers. These allowances represented an increase in the salaries or wages of married officials and workers with two children over those of single men of an amount varying from 15 to 40 per cent. In private undertakings family allowances were of most importance for non-manual workers, although even so they were less extensive than in state undertakings. In industry proper, the system was adopted almost universally in the mining, engineering, chemical, textile, and paper industries, but the allowances were hardly more than 10 per cent, of wages. The system was fairly widespread in the stone, china, glass, printing, transport, and food industries. On the other hand, it was rarely adopted in the leather, wood, and building industries. In agriculture the distinction between single and married workers was often made in actual wages, particularly in payments in kind, which for permanent workers were proportionate to the number of dependants. Similarly a dwelling and a small plot of ground were often furnished to married workers gratis or at a rent much below their ¿value. Even in those industries where the system of family allowances was most widely adopted, equalisation funds were seldom established 1. This system, which avoids giving the unmarried worker an advantage over the married worker on the labour market, and has been so widely adopted in France and Belgium 2, was not considered necessary in Germany. This is doubtless attributable to the fact that up to 1923 labour was scarce, so that the married worker's chances of finding work were not seriously affected. Moreover, 1 An equalisation fund is a fund common to several establishments in the same industry or region, such establishments paying subscriptions in proportion to the number of workers they employ, whether married or single. Allowances to 2married workers with children are then paid from the fund. Cf. Family Allowances; the Remuneration of Labour according to Need. Studies and Reports, Series D, No. 13. Geneva, 1924. — 53 — family allowances were- small; they did not entail sufficiently heavy liabilities to justify the establishment of equalisation funds, the administration of which alwavs involves a mass of detail. Local Cost-oi-Living Allowances A further difficulty of an entirely different kind also helped to maintain the custom of paying cost-of-living bonuses. It arose out of the fact that the actual cost of living varied markedly in different localities. These differences, which are always present, had greatly increased with the enormous inflation of recent years 1 ; they invested the question with a new importance, especially as collective agreements fixing wages and salaries over large areas were growing yearly more numerous. Following the example of the State administrations the practice rapidly spread of classifying the various localities according to the cost of living and of granting local cost-of-living allowances or else family allowances graded according to these classes. In certain cases the distinction was applied to wage rates. 1 The increased variation in the cost of living between different localities is clearly shown by the cost-of-living budgets for the different towns published by the Ministry of Labour. These are not index numbers proper, but are used in constructing them. They represent the absolute cost in marks of a number of items of food, heating, lighting, and rent, the quantities being those of the theoretical budget on which the cost-of-living index number is based. The table below gives, for certain dates from 1921 to 1923, the quartiles and median of the cost-of-living budgets for the 71 towns covered by the national cost-of-living index number. In addition a measure of dispersion is given by the difference between the upper and lower quartiles expressed as a percentage of the median. Date 1 Lower quartile Median Upper quartile marks marks marks Dispersion 1921 September 997 1,037 1,081 8.16 1922 March September 2,141 9,814 2,257 10,298 2,387 10,798 10.90 9.56 1923 March June September December 203,026 562,328 881,378 1 88.933« 218,681 600,084 1,003,722 1 95.275 « 239,966 668,201 1,186,597 1 103,298« 16.88 17.64 30.40 15.07 Million marks. » Thousand million marks. — 54 — To sum up, it may be said that cost-of-living bonuses were granted at first as a temporary means of adjusting the wages fixed by collective agreements to the cost of living. As the rise in prices continued, these systems were improved and generalised; by their means certain needs arising out of the economic situation could be satisfied in some measure and certain problems, which had always existed but had become particularly acute owing to the increase in prices, could be met. Cost-of-living bonuses at flat rates and family allowances were a means of placing the burden due to the reduction in national income on those workers whose economic position was strongest, i.e. those in receipt of high wages and those having little or no family responsibilities. Similarly, the local cost-of-living allowances made it possible to even up the increasing differences between different localities. The disadvantages of the levelling up of wages through these various measures having speedily been realised, a reaction took place. On the other hand, distinctions according to the number of dependants or the cost of living in different localities were gradually introduced into the actual rates of wages. CAUSES OF THE FALL IN REAL WAGES: TECHNICAL DIFFICULTIES OF ADJUSTMENT AND THE REDUCTION OF NATIONAL INCOME In this and the foregoing chapter, the principal measures adopted and technical methods employed up to the end of 1923 for adjusting wages to the cost of living have been reviewed ; these are government control of prices and rents to check the rise in cost of living; regular publication of an official index number to measure fluctuations in the cost of living; reduction of the period of validity of collective agreements and the introduction of more elastic annulment clauses to facilitate frequent revision of wage rates; introduction of cost-of-living bonuses in various forms, first as a temporary measure and subsequently in order to cope with certain special problems arising out of the economic situation and the extremely rapid increase in prices. In the first chapter it was also noted, howeArer, that the adjustment of wages had been inadequate and that the real value of workers' incomes had fallen by comparison with pre-war days. The reason was that the methods adopted were makeshifts, expedients improvised on the spur of the moment which did not work well. These difficulties were further added to and enhanced by others of a purely technical nature; not only did they continually — 55 — impede efforts to secure adjustment but they also made it difficult to discern the exact position. The depreciation in wages must not, however, be ascribed wholly to these difficulties. There was another factor already mentioned which may be recalled at this stage: the shrinkage in national income owing to the economic position brought about by the war and the Peace Treaty. It is no easy matter to determine the respective importance of these groups of causes as active factors in the situation. It has already been seen that the principle of reducing wages to allow for the shrinkage of national income had greatly affected decisions as to the scale of cost-of-living bonuses and family allowances. It was upheld especially by employers. The main arguments invoked were the following: (1) Owing to the economic disorganisation engendered by the war and the liabilities imposed by the Peace Treaty, the productivity of the country was materially reduced; national income had inevitably fallen and the reduction in real wages was merely the consequence of this. (2) A rise in nominal wages would lead to a temporary increase in the purchasing power of the working population but without increasing production and the quantity of goods available for consumption. The effect would be to accelerate the rise in prices and thus nullify the temporary advantage gained by the workers; the final effect would merely be to increase price difficulties. (3) The rise in nominal wages and the consequent rise in prices would necessitate further inflation of the currency to provide for the transactions at the higher prices, hence again an acceleration in the rise of prices. (4) In view of the economic position of the country the only means whereby industry could cope with foreign competition, especially on foreign markets, was to produce at substantially lower cost than other countries; since certain factors in the cost of production (interest on capital, contributions to social insurance, etc.) had increased, the cost of labour had to be reduced, and the workers must accept a reduction in their real wages as an unavoidable necessity in post-war Germany. The workers on their side demanded that their real wages, which were already diminished, should be safeguarded from further reduction caused by fresh monetary inflation. The depreciation in wages so caused was very great already, owing particularly to the many technical difficulties of adjustment to the cost of living by the rudimentary methods then employed. — 56 — One of these difficulties was due to the fact that the cost-of-living index number was itself not always an accurate measure of the situation. It was shown in chapter II that up to the spring of 1922 the index number underestimated the increase in the cost of living as compared with pre-war days by 25 or 35 per cent., while on the other hand it tended to overestimate the monthly increase. Thus, even if wages were fully adjusted to the increase shown by the index, their real value would still be materially decreased. Another difficulty which was pointed out at the beginning of this chapter arose out of the fact that wages remained at the same level during the whole run of agreements, whereas the cost of living continued to rise. The duration of agreements was progressively curtailed, it is true, but the rise in prices grew so rapid that the depreciation of wages was even greater over a short period than it had previously been over long periods. Again, the various operations required to adjust wages inevitably entailed a succession of delays during which the increase of prices continued. This was due to a number of causes. In the first place there was £ome delay in the publication of the index number. While it was issued once a month only (i.e. up to September 1923) it represented approximately the level of prices about the middle of the month, whether prices were recorded on the 15th, as in the beginning, or whether two or four returns scattered over the month were averaged, as was done later. The index number was not published, however, until the beginning of the following month and consequently represented what the cost of living had been over a fortnight earlier. A glance at table I of chapter II giving the monthly fluctuations of the index will show that, up to the end of 1921, the discrepancy thus involved was not very great; with a few exceptions it barely exceeded 5 per cent. From 1922 onwards, however, the rise became more rapid. In a fortnight prices increased at a bound by anything from 10 to 50 per cent., except for a few months when they remained more or less stationery. In September 1923 the index was published weekly for a date not more than two days earlier. But at this moment prices were soaring so madly that the change (in those two days) was greater than it had been in a fortnight during the preceding months. It averaged 65 per cent, in September, from 100 to 200 per cent, in October and from 50 to 100 per cent, in November. To this must be added the time occupied in negotiating collective agreements and notifying them to the parties concerned. An — 57 — agreement could not come into force less than a week after the beginning of discussions. If these were at all protracted or were not begun immediately on the publication of the last index number, the intervening period might easily exceed a month. From September 1923 onwards, it is true, thanks to the weekly publication of the index and to the simplification of procedure, the delay was rarely more than a week. Finally, there was a further delay in that wages earned in any given week could only be spent in the following week. No doubt it gradually became customary, during the period of very rapid increases, to spend the greater part of the wage on pay day, making all purchases required for the week. But this shopping in bulk, concentrated into a single day, led as often as not to a further substantial rise in prices. At the same time, a number of perishable foodstuffs could not be purchased in advance, and there was no avoiding the depreciation of that part of the wage reserved for these purchases. Matters were yet worse in the case of articles involving rather large expenditure (e.g. clothing, furniture), for which workers were in the habit of putting aside part of their weekly wage for some length of time. It should further be noted that the depreciation which occurred between the date of paying wages and that at which they were spent was borne entirely by the worker, the employer's production costs being in no way reduced. As the Ministry of Labour justly points out x, the employer paid more in wages than the worker really received. There were thus numerous technical difficulties; most of them were due to delays, which the working classes constantly strove to reduce to a minimum by curtailing the duration of agreements, multiplying pay days, and hastening the calculation and publication of the cost-of-living index numbers. Most of the improvements so effected, however, were of little use, as they were inArariably outstripped by the constant acceleration of the inflation process for which, moreover, they were partly responsible. Again, as such improvements were always achieved spasmodically and the increase in the cost of living was no less irregular, the depreciation due to these technical difticulties varied from month to month and could not be forecast from one month to the next, or even from one week to the next. In spite of all their efforts, 1 Report by the Ministry of Labour on working-class conditions in Germany, communicated to the International Labour Office. — 58 — then, the workers' standard of life was considerably lowered, while they were in a state of complete uncertainty as to the real value of the wages which they were earning or negotiating. The methods of adjustment just reviewed had the further disadvantage of militating strongly against industrial peace. Too frequent a recurrence of negotiations multiplied opportunities for dispute, and' the discontent caused by the depreciation occurring during the run of agreements induced a less conciliatory frame of mind. This was admitted, and the statistics of strikes and lockouts published by the Ministry of Labour clearly indicate this, as can be seen from the following table. TABLE III. STRIKES AND LOCKOUTS IN GERMANY, 1 8 9 9 TO 1923 Year Number of worker-days lost through disputes Number of disputes over wages Number of disputes arising from causes other than wages Number of disputes over wages per cent. of total disputes Official index number of wholesale prices (annual average) Base: 1013 = 100 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 3,381,319 3,711,994 2,426,952 1,950,847 4,158,519 5,285,125 18,984,553 11,567,618 9,016,856 3,665,607 4,151,766 17,848,440 11,466,727 10,723,782 11,761,754 2,843,895 45,511 245,404 1,862,302 1,452,834 33,082,774 16,755,614 28,874,452 27,733,833 12,343,830 1,148 1,444 878 817 1,278 1,826 2,481 3,467 2,389 1,271 1,433 2,748 2,693 2,995 2,453 1,188 117 215 438 482 3,739 3,799 4,139 4,356 1,708 997 1,349 873 831 1,362 2,132 2,867 3,809 2,099 1,001 1,110 3,139 2,503 3,021 2,330 976 55 121 416 359 2,147 1,421 1,453 1,569 483 54 52 50 50 48 46 46 48 53 56 56 47 52 50 51 55 68 64 51 57 64 73 74 74 78 83 88 82 82 81 84 87 89 100 94 93 91 96 108 100 106 142 153 179 217 415 1,486 1,911 34,182 16,619,000 > million Apart from the very material increase in the total number of disputes during recent years, and in the number of workerdays lost through such disputes, it will be observed that the pro- — 59 — portion of lockouts or strikes caused by wage disputes is much larger than that for earlier years, even in periods of full economic activity and rising prices. It must not be forgotten, moreover, that during these same years the arbitration system designed to prevent industrial disputes was greatly developed and that a considerable number of disputes were thereby avoided. Without the assistance of this institution the adjustment of wages by the means which had been adopted would have led to still greater difficulties. CHAPTER IV Attempts at Automatic or Semi-Automatic Adjustment of Wages (Sliding Scales and Stable- Value Wages) The more or less makeshift systems adopted to adjust wages to the cost of living thus teemed with defects and drawbacks, and their results were on the whole unsatisfactory. Both the public in general and the principal parties concerned soon realised these defects, and controversy arose as the improved methods to be adopted. Discussion was widespread, but its practical results were very limited. The present chapter will therefore be a discussion of principles rather than a survey of concrete facts. There were two well-defined stages in the controversy. In the first stage the central idea was the automatic adjustment of wages to the cost of living, so as to avoid all the drawbacks of the expedients previously adopted and to settle the question once and for all. This idea was embodied in the sliding scale; but the system proved in practice infinitely more complicated than had been expected and was gradually abandoned. Towards the end of 1922, when the mark began to lose its quality as a standard of value, discussion turned to the question of stablevalue wages. This was in fact the same thing as adjusting wages t o the cost of living, b u t in a more flexible and limited manner than under the sliding-scale system. An agreement was being reached in the matter, and some practical steps had been taken, when financial reform, more particularly the introduction of the rentenmark, rendered them unnecessary. SLIDING SCALES (A ulomatic A djustment) At first sight the simplest means of maintaining the workers' standard of life, and thus eliminating the recurring causes of friction and dispute due to their continual effort to maintain this standard, would seem to be the automatic adjustment of wages to fluctuations in the cost of living by a system of sliding scales. This system liad already been in use before the war, but as a rule for the purpose of adjusting wages to factors other than cost of living (e.g. the selling — 61 — price of the articles manufactured, the cost price of raw materials, or individual or collective output), either to give the workers some share in the prosperity of the industry or as a method of payment giving a greater incentive to work. The idea of adjustment to the cost of living had nevertheless been put forward by Mr. Braun, a member of the Reichstag, who in 1914 advocated a general sliding scale applying to all workers. The idea was taken up again in 1919 on the trade union side by Mr. Erkelenz. He proposed that statistical offices should be set up in the chief districts to note changes in the prices of the commodities needed for the minimum of subsistence, and that all wages, salaries, and income should be adjusted to such changes. The question was debated at a conference on wage rates held at the Ministry of Labour on 12 January 1920; this inaugurated a vigorous discussion on the theory of the matter. Opinion among the workers was not unanimous. Some considered the adoption of a sliding scale as the best protection against the continued debasement of their standard of life. Others remained very mistrustful of the system, the very fact of its being automatic causing them misgiving. They were afraid of losing the advantages of trade union action and of sacrificing one of their fundamental rights — that of negotiating about their own conditions of work. Moreover, they had not sufficient faith in any index number of the cost of living to make their wages permanently dependent on it. Organisations of salaried employees and officials, on the contrary, were distinctly in favour of the system. Among employers opinion was almost unanimously unfavourable on the grounds that so rigid a system involved serious risks for industry, since export trade might be completely lost when the cost of production could no longer be adapted to conditions on foreign markets. The dangers of an accelerated inflation of prices were also strongly urged. The most precise schemes were put forward by independent persons. In June 1921 Mr. Staude, a Bremen technical councillor, proposed a system under which a certain percentage of wages, reduced monthly by 5 per cent., would be adjusted to changes in the retail prices of the chief articles of food. In October 1921 the principal exponent of the sliding-scale system in Germany, Mr. Zeiler, Judge of the Federal Court, advocated a fairly detailed scheme for officials. The authorities showed much greater reserve. Dr. Fritz Foth of the Ministry of Labour, however, expressed fairly optimistic views — 62 — in a semi-official article 1. On the basis of certain statistics he argued that wage rates were as a matter of fact adjusted spontaneously every month to changes in the cost of living in the previous month. No sliding-scale system would effect much more, since for technical reasons the adaptation of wages to the cost of living by this means would similarly apply to the fluctuations of the previous month. This view was very much criticised. It must be admitted that many of the critics misinterpreted Dr. Foth's idea, but in view of the official wage and cost-of-living statistics hitherto used and the revised index number of the cost of living, which was not yet published at the date of Dr. Foch's article, his theory is seen to be not entirely in accordance with the facts *. In February 1922 the Minister of Labour stated his views. He feared that automatic adjustment would give a further impetus to price inflation and kill the export trade by raising the cost of labour. On the other hand, he fully recognised the need of effective measures and recommended the adoption of a more flexible non-automatic system. These views were developed in further detail in a memorandum addressed to the Reichstag and published at the end of September, which in some measure closed the discussion. The automatic adjustment of wages by means of sliding scales was definitely abandoned. As a matter of fact there had been very little practical result for this press campaign. The most important and best-known experiment was that of Flensburg, details of which are given in an appendix 3. It was initiated at the beginning of 1920 before the discussion reached its height, but given up at the end of 1922. Reference may also be made to the experiments of Breslau, Frankfurt-onMain, Kiel, Cologne, and the Barckhauser firms. The complete failure of the idea of a sliding scale was very largely due to the fact that the problem proved on examination much more complicated than had at first appeared and that it was not possible to reach agreement on the method of settling the various questions raised. One of the most delicate problems was that of the basic wage. On the sliding-scale principle the basic wage must retain its real value, if not permanently, at least for a considerable period, and thus its determination became of particular importance. The most diverse opinions were expressed. Some took existing or pre1 2 8 Reichsarbeitsbla.lt, No. 4, 1922, p. 131. See Chapter I, tables I and IV. See appendix V. — 63 — war wages, thus choosing a specific period of no particular intrinsic value. Others maintained a one-sided view and demanded that the basic wage should provide for certain ideal conditions of life, while yet others claimed that wages should depend entirely on the economic situation of the country and the particular industry. Suggestions as to the choice of the index number by which wages were to be adjusted automatically were also very varied. Reference has already been made x to the small degree of confidence placed in the many cost-of-living index numbers published in Germany, precisely on account of the variety of their results. At the time when the question of sliding scales was most under discussion the official cost-of-living index number had not yet been revised and was therefore not of a nature to inspire general confidence. In the few cases in which sliding scales actually were adopted. use was made of cost-of-living index numbers published by local authorities or compiled by joint committees. Other indexes were also proposed as some sort of substitute for a cost-of-living index number, but they all possessed serious disadvantages. Thus the rate of exchange of the dollar (which was suggested later in another form in connection with the question of gold .wages) represented fluctuations in the purchasing power of the mark not at home, but abroad, and was further far too much subject to the influence of speculation. Similar objections applied to the index number of wholesale prices, although the speculative element was of rather less importance. An index number of the prices of certain fundamental commodities such as wood or coal was still less satisfactory, such prices varying with the conditions of the market and being open to speculative manipulation by either of the contracting parties with the deliberate intention of influencing wage rates. Fairly soon, however, it was realised that allowance should be made, not solely for the cost of living, but also for the economic position of the country and of the industry. Several proposals were put forward to this end. Mr. Zeiler suggested that two calculations should be made on different lines; the first was to be based on an index number of economic conditions and the second on the index number of the cost of living. The higher of the two results would then be used 2. For industrial workers the index of economic conditions was to be separately calculated according to conditions in each industry ; for 1 2 Cf. Chapter II, p. 38. Soziale Praxis und Archiv für Volkswirtschaft, 1922, XXXI, p. 97. — 64 — oflìcials, it was to be based on the general prosperity of the country, in other words, on aggregate wealth as evidenced by income tax data y Mr. Zeiler's scheme was severely criticised by Dr. Vollbrecht 2. He stated that in calculating the mean income of the population Zeiler allowed equal shares to capitalists, employers, and workers. He thus ignored the important influence of the economic power of the various groups on the sharing-out. Moreover, when general prosperity declines — i.e. when the quantity of goods available for consumption grows less — prices rise, and the rise in the cost-ofliving index number, based upon fixed quantities of consumption goods, balances or even outweighs the fall in the index of general prosperity. The results at which Mr. Zeiler aimed would thus not be attained. Another method of taking some account of factors other than fluctuations in the cost of living is to adjust wages only partially to these fluctuations on the ground that these other factors may justify, if not demand, a reduction of real wages. This system was generally advocated by those who were anxious not to accelerate the rise of prices too much, or considered that, as the national income had diminished since before the war, real wages should be correspondingly reduced. These principles are the same as those already discussed in connection with adjustment by collective agreement 3 . If the object is to prevent too rapid an inflation of prices, the partial adjustment operates uniformly without other distinction, i.e. a proportion of each wage is adjusted to the cost of living, the rest remaining unaltered. This proportion may be fixed once and for all or be increased according as the rise in prices is accentuated or prolonged. Other considerations arose when it was desired to reduce real wages to correspond with the diminution of the national income. As a rule an attempt was made to maintain the absolute minimum of subsistence and to reduce the excess over this by bearing more hardly on high wages and workers with little or no family responsibilities. In this case the proportion of wages to be adjusted to the 1 Under this system, wages would be adjusted to the cost of living monthly or quarterly, and to national prosperity or to the position in a given industry annually. In this manner wages would derive from the first adjustment a stable value throughout the year; their exact amount would be fixed yearly, according to the prosperity of the community as a whole or of the particular industry, so that each class would receive its proper share of the national income. 1 3 Soziale Praxis und Archiv für Volkswirtschaft, 1922, XXXI, p. 241. Cf. Chapter III, pp. 48-55. — 65 — cost of living may be fixed once and for all, a higher proportion being allowed in respect of low wages or workers with families. It is also possible to adjust only what is considered the minimum of subsistance and grant the same additions to everyone ; allowance may or may not be made for family responsibilities in this case. The last difficulty was due to the marked reluctance of many workers to accept the consequences of the sliding-scale system in the event of a fall in prices. They considered that their real wages were already sufficiently reduced to give them the right to enjoy the full advantage of a reduction in the cost of living without any corresponding reduction in wages until the pre-war level had been restored. They similarly objected to the adoption of sliding scales on the ground of a possible temporary fall followed by a further rise. Under all the systems proposed the adjustment of wages would be based on the index number for the previous month. It might therefore quite well happen that a fall in prices in the previous month would lead to a reduction of wages just during the month when the cost of living again began to rise. This brief survey will sufficiently bring out the very great difficulties and imperfections of any automatic system of adjusting wages to the cost of living, difficulties and imperfections which were still further increased by the fact that the rise in prices in Germany was already enormous. STABLE-VALUE WAGES (Semi-Automatic Adjustment) Sliding scales having rather failed to win public approval, discussion turned towards some less rigid method of adjustment. In the meantime, the depreciation of the currency was so rapid that the paper mark lost its function as a standard of value and became simply a medium of payment, the rate of which was estimated daily. The new methods of adjusting wages to the cost of living, which were intended to protect them from the depreciation in the value of the mark, were baptised "stable-value wages". Until the middle of 1923 the opposition of employers to the efforts of labour organisations to obtain an adjustment of wages was based on the arguments already advanced in the discussion of sliding scales, namely, the danger of accelerating inflation, an unrestricted rise in prices, and the ruin of industry, which would be unable to meet foreign competition. Stress was laid on the Austrian example. — 66 — Opinion among trade unionists was still not unanimous. Some proposed a gold wage, others supported semi-automatic adjustment based on the cost-of-living index number, the index number of wholesale prices or, as suggested by the Afa1, on the two indexes combined, or, again, on the dollar rate of exchange. Yet others approved the suggestion of Dr. Potthoff, a well-known writer on social questions, to pay wages in kind. - One of the chief obstacles to using the cost-of-living index number was that it was still imperfect, especially in speed and frequency of publication. Part of the discussion and demands, therefore, turned on revision of the method of calculating the index. In addition, the issue was often confused by differences in the interpretation of the terms "gold wages" or "stable-value wages". The Government, while recognising the need for some kind of action to improve the conditions of the workers, refused to take any steps which might compromise its efforts to stabilise the mark. Besides, the Cabinet was not unanimous. The Minister of Finance, in a statement in the Reichstag in March 1923, went so far as to maintain that all attempt at adjusting wages should be given up so as not to interfere with the efforts of the Government to stabilise prices. The Minister of Labour, on the other hand, in an interview with trade union representatives, recognised that the adjustment of wages was still necessary, in spite of the temporary fall in the dollar rate, and that in any case it could not be dropped suddenly. The statement made by the Minister of Finance created much indignation among trade unionists. They failed to secure its withdrawal, but the question was placed before the whole Cabinet, and the Chancellor summarised the matter in a press circular in which he endeavoured to reconcile the two points of view. By degrees opinion in the trade unions became more definite. By the end of June the organ of the General Confederation of Trade Unions, the Korrespondenzblatt, was definitely in favour of adjusting wages to an improved cost-of-living index number, published weekly, basic wages being revised once a month in accordance with the economic situation. The delicate question of the selection of the basic wage, which had so long obscured the problem of the adjustment properly so called of wages, was therefore temporarily left on one side; the basic wage had to be fixed in each case by negotiation between 'employers' and workers' organisations. 1 Allgemeiner freier Angestelltenband unions). (Federation of salaried employees" — 67 — . The views of the other trade unions were more or less the same, and the workers' representatives on the general Joint Industrial Association (Arbeitsgemeinschaft) * were therefore able at the end of the month to agree on a general programme. They met with strong opposition from the employers, but the latter ultimately accepted the idea of adjusting wages to the dollar rate, which seemed to them less dangerous for industry. Their point of view was that the direct adjustment of wages to the cost of living would lead to a vicious circle : the rise in nominal wages would temporarily raise prices ; the purchasing power of the population would increase without any corresponding increase in the commodities available for consumption ; the rise in prices would instantly lead to a fresh rise in wages, and there would be no direct advantages to the workers. The effect would be to depreciate the purchasing power of the mark at home more than abroad, with the result that the cost of labour would be substantially increased. The premium on production, hitherto enjoyed by Germany as a result of her relatively cheap labour, would disappear entirely, giving way to very unfavourable conditions. If, on the other hand, wages were adjusted to the dollar rate, the first effect would no doubt be to raise prices, but this rise would not affect wages except in so far as the increase in wages and prices would necessitate fresh issues of paper money and thus depreciate the rate of the mark in New York. Since, however, this increase in the cost of labour in marks would be accompanied by a depreciation of the purchasing power of the mark abroad, the cost of labour from the point of view of foreign competition would remain the same. These proposals were declared unacceptable by the workers' representatives on the Joint Industrial Association, for they did not wish to make the level of real wages dependent on foreign speculation. Negotiations were therefore broken off. The Reichstag soon had an opportunity of expressing its views on wage policy, for two resolutions on this subject were laid before it. That of the Social Democrats favoured the immediate introduction of a system of adjusting wages and salaries of officials and of workers in private industry to the. cost of living. The second resolution, tabled by the Centre Party, simply recommended 1 This is an institution founded since the war for bringing together representatives of employers' and workers' organisations to discuss and study economic questions affecting them and if need be to conclude general agreements. — 68 — the Government to take the steps necessitated by the importance of the question. The Reichstag passed the second resolution. The Federal Economic Council, too, expressed itself in favour of weekly adjustment of wages, but considered this only a temporary measure pending the possibility of adopting gold wages. It further left the parties concerned to decide whether the adjustment should be based on the general index for the whole country or the particular index for each town. Under the pressure of the distress caused by the steadily accelerated rise in prices, the demands of the workers became more and more urgent, and serious strikes had already broken out in Berlin. An agreement was at- last reached and both parties accepted as a starting point the guiding principles proposed by the Minister of Labour on 19 July 1923. These principles definitely rejected the system of a purely automatic sliding scale, a combined system being recommended in its place. Basic wages were to be fixed at regular intervals of one or two months by the usual method of negotiation, taking into account not only the cost of living, but also economic conditions in the country as a whole and in the particular industry; the relative strength of the two contracting parties would also exert an influence. During this period basic wages would be adjusted once a week to the cost of living on lines laid down in the agreement. It was considered advisable, however, that small joint committees should be appointed for this purpose, if necessary with an impartial chairman, which would undertake the adjustment of wages. The official cost-of-living index number and if need be the index numbers of wholesale prices or the dollar rate of exchange were recommended as indicating fluctuations in the cost of living 1 . In this way both parties were more or less satisfied; basic wages could be revised in accordance with changes in the situation by means of the periodical negotiations, while the semi-automatic weekly adjustment guaranteed the workers that the real value of wages would on the whole be maintained during their period of validity. The expectations raised by this agreement were not fulfilled. The rise in prices was particularly violent during August, and there were enormous differences in wages between different localities and different trades. Certain undertakings had to close down, being unable to obtain a sufficient quantity of paper money to 1 Further details will be found in appendix VII, where these guiding principles are given in full. — 69 — pay their wages. In other branches of industry the real wages of the workers were terribly reduced. On 1 September employers' and workers' organisations entered into a general agreement supplementing and defining the principles laid down by the Ministry of Labour. The principle of this agreement was the same. The period of validity of the basic wage was strictly limited : from four to eight weeks. The use of official index numbers was definitely recommended, the index for the whole country for collective agreements covering the whole country or large areas (in which case regional differences were allowed for in the basic wage) and the official or private local indexes for local agreements. Stress was also laid on the need for making the adjustment as far as possible in respect of depreciation in the week during which the wages were spent. With this end in view the use of the index number of wholesale prices and the dollar rate was recommended as s means of estimating the future rise in the cost-of-living index number, but no indication was given of the method to be employed. Any mistake made in the estimate would have to be corrected the following week, when the cost-of-living index number was published, and the excess subtracted from or the deficit added to the wages being fixed. The system of making pay day more frequent than once a week was to be reserved for quite exceptional circumstances (e.g. the first week of August 1923). The practical results of all these long discussions and attempts at conciliation were not very important. Even before the Ministry of Labour and the organisations of employers and workers had devised their systems, a number of collective agreements had already set up a system of stabilisation. In the ready-made clothing trades, for instance, a national agreement of 24 June fixing wages for the whole country provided that they were to vary weekly in proportion to the percentage rise in the cost of living as grven by a special index number calculated by a joint committee. In the clothing industry a similar system had been in force since 1 September 1922. Every three weeks tailoring employers and workers in 200 towns systematically recorded the prices of articles of prime necessity, independently of any official body. These statistics were used for fixing wages. At a meeting held on 26 June 1923 at Eisenach, the central committees of the employers' and workers"' organisations in the tailoring trade improved their system by arranging for the collection of price data with a view to fixing wages once a week. — 70 — In July 1923 the wages of shop assistants in Berlin were adjusted to the depreciation of the currency by the weekly addition to the basic wage of a certain percentage calculated on the gold rate of exchange. In the Berlin metal industry the agreement of 10 July, which put an end to the strike, laid down that wages were to vary weekly by the same percentage as the cost of living as shown by a special index number calculated every Friday and used for fixing wages on the following Saturday. A general agreement for the transport industry concluded in July embodied the principle of fixing basic wages by collective agreement in the usual manner and of paying a supplement varying weekly with the cost of living. The practical application of the principle was left to the special groups concerned. In the coal mining industry a general agreement was concluded on 8 August containing similar provisions. The basic wage was to be fixed once a month by collective agreement. If a rise in the cost of living was anticipated, the agreement was to fix a cost-of-living bonus at the same time. This bonus would be paid weekly as long as the cost of living rose at the same rate. As soon as there was any change in the rate, a new bonus would be fixed by agreement between the parties in accordance with the official cost-of-living index numbers. A special index was to be used for the occupied areas. When the general rules made by the employers' and workers' organisations were published, they were fairly rapidly adopted in the large towns and industrial districts. They distinctly relieved a situation which was growing very difficult owing to the incredible rise in prices. At the same time the application of these rules demanded of the contracting parties and the conciliation authorities some knowledge of the subject and of the economic situation^ which knowledge they did not always possess; unsound methods were applied in some cases and the effect of their failure was to strengthen the opposition in certain quarters" to the adoption of these rules. However, even though these principles might not be formally accepted they were at least agreed to in principle. For instance, the employment of the cost-of-living index number was very widespread, though expressly stipulated in only a small number of agreements. At all events the proposed system marked an advance on previous methods. It was merely a matter of time that it did not gain — 71 — acceptance in various quarters and that there was no opportunity of effecting improvements in the light of the first experiments ; the system had hardly been introduced before it was rendered superfluous by the stabilisation of prices consequent on monetary reform. It is none the less interesting to examine the various problems which arose and the various means adopted or contemplated to solve them. The fundamental question then, as before, was the choice of the index number. As has been shown, agreement was not reached on this point until the cost-of-living index had been revised, extended, and calculated weekly, until representatives of the workers had been admitted to membership of the commission in Berlin responsible for compiling the index for the whole country (as they were to the committees responsible for compiling the indexes for the different towns), and until the parties concerned were left free to choose between the index for the whole country, the local indexes, and private indexes calculated by local authorities or private joint committees. The very method of using the index number, once it had been selected, might vary considerably. The rules of the Ministry of Labour on this point were intentionally vague. The agreement between employers' and workers' organisations was rather more precise; it suggested the multiplication of the basic wage by a coefficient which was simply the cost-of-living index number. It will be recalled that when sliding scales were under discussion the great difficulty was that wages could not be fixed solely by the cost of living, but that various other factors for which no index was available had also to be taken into consideration. Here the problem was solved by adjusting wages to the cost of living alone at short intervals. All the other factors could be taken into consideration in the periodical negotiations for fixing basic wages. Such negotiations no doubt also took into account the cost-ofliving index number, but other factors were fully considered. There was therefore no need to apply the procedure of partial adjustment to the automatic part of the system, for the same result could be obtained by modifying the basic wage. In order that the privation due to the reduction of national income might be borne by those most able to bear it, i.e. persons with high wages or few family responsibilities, it was enough to fix basic wages and family allowances in the desired proportions. No decision was taken in this respect either, the settlement of the question being left in each particular case to the contracting parties. — 72 — The problem has been complicated by another difficulty: the positively fantastic inflation of prices from the middle of 1923. The device of making pay-day more frequent than once a week necessarily increased the work of the wage offices, especially in view of the calculations for adjusting wages to the cost of living. A half-way measure was to make advances on wages between the regular pay-days, fixed more or less arbitrarily in the form of a flat rate or a percentage of the previous week's wages, and only to calculate the adjusted wages for paying the difference on the ordinary pay-day. Either of these measures was unproductive expenditure which it was desirable to avoid; the agreement of 1 September between employers' and workers' organisations stipulated that they should be used only in exceptional cases. It therefore became necessary to estimate in advance the rise in prices for the following week, and various methods were proposed. One method was to extend the cost-of-living index number, as it were, estimating that its future increase would be the same as the average rise during the previous weeks. This is a rather rudimentary method, for it assumes that the rate of increase of the index number is more or less uniform, while in reality it was irregular and definitely accelerated (see chapter II, table II). The use of the indications to be derived from changes in the index number of wholesale prices and the dollar rate was generally recommended, for they were known to be closely related to changes in the cost-of-living index number. Both the instructions of the Ministry of Labour and the agreement between employers' and workers' organisations recommended their use without exactly defining the method of calculation to be used. This was in fact a difficult question, for any solution must be empirical, and since the relations between wholesale and retail prices changed rapidly a solution which was satisfactory at a specified moment might lose all value some months later 1 . In spite of their imperfections, the proposed solutions seemed fairly practicable. As was stated above, the fact that they were not adopted on a large scale is to be attributed to the lack of time for their introduction. 1 For further information on methods of estimating the rise in the cost of living, see appendix VI. CHAPTER V The Gold Wage and the Stabilisation of Wages The results of the discussion on sliding scales were of no practical importance. The system of semi-automatic adjustment known as "stable-value wages" had not yet been widely adopted when the financial reforms of the Government more or less restored the stability of the currency and arrested the rise in prices. Parallel with this general stabilisation the so-called "gold wage" system was introduced. Of all the new expressions employed in discussions on wage policy the term "gold wages" has probably led to most confusion. Some writers used it for systems of stabilisation based on index numbers of the cost of living or wholesale prices, probably associating the idea of the stability of real value with the value of gold, which was fairly stable before the war. For others gold wages were quite simply pre-war wages. In order to prevent confusion these conceptions have been ignored here. The idea of a gold wage had already been considered in the discussion of sliding scales. It was then suggested that wages should be adjusted according to the dollar rate, but in such cases the gold index was intended to be used simply for an approximate adjustment to the cost of living. If a cost-of-living index number was not used, it was because the employment of the index of the proper dollar rate presented certain advantages. It was established daily and published immediately; its use in adjusting wages - would not tend to depreciate the purchasing power of the mark at home more than abroad and thus make competition impossible. But obA'iously it could give only a rough measure of fluctuations in the cost of living and was often influenced by factors which did not affect retail prices on the home market. The principle of these systems was to fix basic wages in paper marks in accordance with the conditions prevailing when the agreements were concluded and subsequently to vary them in proportion to the percentage variation in the gold index. The problem was thus approached from the point of view of adjusting wages to the cost of living, but there was no suggestion — 74 — of dissociating the two functions of money as a standard of value and as a medium of exchange. The gold wage system advocated at the end of 1923 consisted in making an agreement fixing the actual wage in terms of some stable monetary unit (with a gold basis) and paying it either in paper marks, at the rate of exchange of the stable unit, or partly in paper marks and partly in the stable unit. If the latter were not used to pay wages, it might be wholly fictitious. This idea involved the acceptance in principle of the dissociation of the two functions of money, as a standard of value and as a medium of exchange. Owing to its remarkably rapid depreciation, the paper mark was incapable of fulfilling the first-named function, which was therefore conferred on another monetary unit. The paper mark was retained as a medium of exchange, no other being legal tender. Moreover, this system was merely an extension to the labour market of customs which had already become common elsewhere. This system of accounting in one unit and paying in another had been adopted in industry and commerce at an early date. It had rapidly become the custom in foreign business to keep all accounts in some fairly stable foreign currency, such as the dollar, the Dutch gulden, or the Swiss franc, and when the transactions had to be settled in paper marks to calculate the sum in paper marks at the current rate of exchange of this foreign currency. This procedure had definite advantages of simplicity and security which secured its wide adoption. The Government, on the contrary, did not regard it very favourably, for it justly considered the custom as the death warrant of the paper mark, which would gradually be reduced to a simple token currency. For a long time the Government therefore refused to adopt the system of gold accounts for re-organising its taxes, and it hesitated long before placing its customs duties on this basis. Insurance companies, however, obtained legal power to keep all their accounts on a gold basis, but retail traders were strictly prohibited by law from quoting their prices in anything but paper marks, though this could not prevent them from using a stable unit, generally the dollar, for their accounts and converting their prices to paper marks at the current rate. The result was to establish a much closer relation between the fluctuations in retail prices and those of the dollar rate. Under such conditions, and with such a widespread example, it was quite natural that proposals would be put forward for adopting the same system for fixing wages. — 75 — Messrs. Tarnow and Wagner-Roemich were the first advocates of the new idea of the gold wage 1 . As early as the autumn of 1922 they endeavoured to demonstrate the evolution of the process of splitting up the currency into a unit of value used for accounts and a unit of payment. They considered it inevitable that the procedure would be extended to all spheres of economic activity, and therefore thought it obvious that it should be quite generally adopted as soon as possible for settling all transactions and particularly in employment contracts. But they did not as yet contemplate the replacement of the paper mark as a medium of exchange. They merely advocated the general adoption of a stable standard of value in terms of which all transactions would be effected, settlement being made entirely in paper marks at the rate of exchange of the selected standard. For this, Wagner Roemich proposed the gold gramme, which could be adopted internationally. With more practical insight Tarnow suggested that the use of the dollar, which was already very widespread in the country, should be sanctioned by law and applied to the labour market as to all others. Both writers were severely criticised. To most minds the term "gold wage" suggested the idea of pre-war wages, or of existing wages in other countries, or at any rate very high wages. The arguments as to the danger of an increase in the cost of labour and inflation were therefore taken up again with renewed fervour; it was further claimed that the available paper marks would be insufficient to pay gold wages. The opposition of employers was very evident from the outset. The trade unions were far from being agreed, and a vigorous discussion was carried on in the pages of the Korrespondenzblatl, the organ of the Social Democratic unions. The Deutsche "Techniker-Zeitung, the organ of technical employees, was opposed to the idea, but the Christian unions seemed to be in favour of the reform. Professor Hirsch, Secretary of State in the Ministry of National Economy, in a statement to the Socialisation Commission in the summer of 1923, expressed himself in favour of gold wages. He considered the enormous margin between wages paid in paper and sale prices calculated in gold to be one of the fundamental reasons of the lack of productive capacity in industry-, of the postponement 1 Cf. T A R N O W in Holzarbeiter-Zeitung, 23. Sept. and 21 Oct. 1922, a n d Korrespondenzblatt. 21. Oct. 1922. Also W A G N E R - R O E M I C H , in Soziale Praxis, 5 Oct, 1922. — 76 — of technical improvements, and the unhealthy development of a class of profiteering middlemen. The remedy, according to him, was to place the whole economic system on a single accounting basis, that of gold. The Government expressed no official views. Public attention was diverted from the question of gold wages by the discussions on the adjustment of wages to the cost of living which took place in the summer of 1923, and were described in the previous chapter. The question of gold wages assumed fresh importance with the first issue of gold mark currency. From the autumn of 1923 onwards several loans on a gold basis had been floated by various public bodies, and the bonds of such loans were used to cover the issue of gold marks by the Federal States and the municipalities. Berlin also put notes valued in dollars into circulation. Finally, the rentenmark was introduced. The existence of stable-value currencies thus strengthened the idea of the gold wage. It was no longer necessary to resort to a fictitious standard of value or to a foreign currency. The gold mark itself, for such was the name given to these various currencies, could be used, and payments could be made in part at least in the currency chosen as a standard. The practice of accounting in gold marks or dollars had, moreover, become general in retail trade as elsewhere, and consumers whose income continued to be fixed in paper marks found themselves in a difficult position. The demands of workers' organisations for the adoption of a gold wage system grew thus more numerous and insistent. As before, the Ministry of Labour attempted to secure agreement among the central industrial organisations, but it met with uncompromising opposition from the employers' organisations, which refused to accept any measure which might entail an increase in wages. Once again the Minister published his views in a press notice. While recognising that the demands for gold wages were legitimate he considered that the moment for introducing such a reform should be chosen with the greatest circumspection; the economic position in each industry in particular should be taken into account. It was in the large towns, and more especially in Hamburg, where stable-value currencies were most current, that the system was first adopted. Thence it spread by degrees to the whole country. From the third week in November 1923 miners' wages were fixed in gold marks. On 1 December the Government introduced the same method of calculation for its workers and officials. Prior to — 77 — this the bank employees had after a long dispute secured an arbitration award to the effect that their salaries should be fixed on the same basis. The first agreements mentioning gold wages fixed rates in gold marks, but gradually, with the success of the rentenrnark, they were fixed in that currency. The employers were, moreover, not slow to adhere to the new system and the Federation of German Employers' Associations published instructions with a view to its introduction. The employers insisted that the system should be introduced very gradually, but evidently expected it to lead to the reorganisation of the entire economic system on a gold basis. One clause of the agreement governing the introduction of gold wages was criticised by the workers; it stated that the consent of both parties was required and abolished reference to joint arbitration committees. The workers looked on this as a.skilful move on the part of the employers to retain full freedom to introduce gold wages or not as they pleased. The trade unions were equally unwilling to accept the employers' reservations regarding the level of the gold wage. On the whole, however, public opinion was in favour of complete adoption of the gold wage and the abandonment of other methods of adjustment. The gold wage was merely a temporary measure to facilitate the transition from an economic system based on the constantly depreciating paper mark to one based on the rentenrnark or the stabilised mark. It was advocated before the introduction of a stable-value currency, and meant recognition of the complete dissociation of the two functions of money, the standard of value being exclusively a fictitious or foreign monetary unit. It was not applied in practice, however, until the appearance of the first German stable-value currencies, which were a concrete representation of the theoretical concept of a standard of value and obviated complete dissociation, the gold mark acting both as a standard of value and, to some extent at'least, as a medium of exchange. Before the introduction of the rentenrnark, the proportion of wages which could be paid in stable-value currency was usually very small. When the rentenrnark was put into circulation the problem was to decide what proportion of wages could reasonably be paid in the new currency. The Government strove to keep this proportion very small at first, and to increase it only very gradually in order to avoid the danger of further inflation. It also endeavoured to divide the available quantity of rentenmarks equitably among the various branches of production, to ensure more or less even — 78 — distribution among the various economic groups. But the public showed in general such a strong preference for the new currency that this scheme had soon to be abandoned. The employers' organisations realised that too rapid an increase in the circulation of the rentenmark involved the risk of fresh inflation; in their regulations they expressly stipulated that the proportion of wages paid in rentenmarks should not exceed ten per cent, at first, and should only be increased slowly in subsequent payments. But the predominance which the rentenmark speedily achieved solved the problem of itself. Stabilisation further made it possible considerably to simplify methods of fixing wages, and this was a definite advantage for workers and employers alike. Cost-of-living bonuses were rendered. unnecessary by the relative stability in the cost of living and tended to disappear. Similarly, the graduation of wages according to the local cost of living (either by local allowances or by a gradation of wage rates) grew less and less necessary, for with stabilisation the enormous discrepancies in the cost of living between different districts were considerably reduced. Family allowances, on the other hand, were still maintained, as wages were still far below their real pre-war value. As pointed out in chapter III, advantage was taken of the stabilisation of prices partially to restore the differences between the wages of skilled and unskilled workers and between the salaries of higher-grade and subordinate professional workers; this was a notable step forward, in view of the disadvantages which are apt to accompany excessive levelling up. But the fundamental problem, which subsisted when events themselves had set aside the questions just examined, was that of the actual level of wages. This problem had in a sense fallen into the background, or had at any rate been obscured by a host of other problems engendered by the general economic upheaval. During 1923 especially, inflation had grown to such proportions that all other considerations had faded before the urgent necessity for arresting the growing depreciation of the wages of labour. Once the currency was stabilised, all the problems of adjustment were removed at a stroke, and the same was true of the problems peculiar to the brief transitional period. The question of the actual level of wages once more became of prime importance. The employers hastened to emphasise the fact that gold wage did not mean pre-war wage. On the contrary, they repeated that as the national income had diminished owing to the war and its — 79 — consequences real wages should be correspondingly reduced. Furthermore, the relative importance of the various trades and industries in the economic system had greatly changed since pre-war days, and their share of the national income should also be modified. The workers, on their side, wished to increase their real wages. They pointed out that, as the purchasing power of gold was lower than before the war, their standard of life would not be the same even if nominal wages were fixed at pre-war rates. They further declared that their wages had been reduced to an exceptionally low level by the depreciating action of inflation, and it was high time for the injustice which they had suffered to be remedied. The Federation of Salaried Employees pointed out, again, that the stabilisation of prices and the introduction of a gold wage afforded the employers real advantages. These reforms greatly reduced unproductive expenditure on wage accountancy and gave employers greater security in business transactions; premiums to cover the risks incidental to inflation were no longer necessary. Such gains as these should enable the employers to grant real increases in wages. It seems that the workers' point of view prevailed, for throughout this first period of stabilisation nominal wages increased steadily and appreciably month by month. There were, it is true, a few slight fluctuations in real wages, for the cost of living varied somewhat during the months under review, but the general tendency was definitely towards a rise. The following table illustrates the two movements. TABLE I. WEEKLY NOMINAL WAGES AND INDEX NUMBERS OF REAL WEEKLY WAGES PAID TO SKILLED AND UNSKILLED WORKERS IN THE PRINCIPAL INDUSTRIES, DECEMBER 1 9 2 3 TO JUNE 1 9 2 4 Date 1923 December 1924 January February March April MayJune July August September Skilled workers Index of Nominal real wages weekly wages (Base: 1913 = 100) (rentenmarks) 70 28.81 27.31 28.12 29.13 31.54 33.75 35.52 35.63 35.76 35.83 74 78 79 81 86 91 90 91 89 Unskilled workers Index of Nominal real wages weekly wages (Base: 1913 — 100) (rentenmarks) 24.27 85 22.87 23.08 23.21 24.43 26.16 27.16 27.24 27.35 27.38 88 92 90 90 95 100 98 100 97 — 80 — It should be recalled, however, that the position of the workers did not improve during the stabilisation period as much as wage statistics would appear to show. Unemployment was growing, as already mentioned, concurrently with the stabilisation of the financial system. It must not be forgotten that the total income of the working classes was considerably reduced by unemployment, as was shown in chapter I. From the first months of 1924, unemployment was doubtless less serious, but even in the autumn of that year its disappearance could not be predicted. From the point of view of production it should be noted that, though the purchasing power of the working classes was enhanced by the increase of wages, the cost of labour increased to a yet greater extent. It will be remembered that during 1922 and 1923 expenditure on rent was practically eliminated from the workingclass budget, and hence from the cost of labour. The new policy again made rent an appreciable charge, and expenditure on rent gradually resumed its pre-war importance in working-class budgets, the cost of labour being correspondingly increased. To some extent, it is true, this is compensated for by the increase in hours of work which was resorted to in spite of the social evils which it may entail. It may thus be said that the stabilisation of wages led quite. naturally to the disappearance of the difficulties and disadvantages connected with inflation, but the fundamental problem of the level of real wages, i.e. the distribution of national income, remains. Now, however, it stands by itself, and efforts to solve it are no longer hampered by the innumerable and ever-varying obstacles of currency inflation. PART 11 AUSTRIA INTRODUCTION General Economic Conditions The development of economic conditions in Austria falls into two well marked stages. Up till the autumn of 1922, i.e. before the League of Nations intervened, the country was in a state of complete disorganisation and was fast going to ruin; this was the first phase. The second phase is that of reconstruction, undertaken with the assistance of the League and energetically carried through by the Government. During the first phase economic conditions were much the same as those described in Germany, but the disorganisation which the war had left in its wake was still greater. When the AustroHungarian Empire was dismembered, Austria was cut off from the principal industrial and agricultural districts which had until then been her main source of wealth. One-third of the six million inhabitants of this small country were concentrated in a capital which had formerly been the intellectual, banking, commercial, and administrative centre of a nation of 29 million people. In the circumstances, and although the fact is as little open to statistical proof as in Germany, it must be concluded that the national income diminished very materially during the post-war years. Weighed down by the burden of enormous administrative services which it could not abolish in a day, and of subsidies for feeding the population, the Austrian Government adopted a policy of monetary inflation which, once begun, had perforce to be continued. The issue of bank notes increased progressively. At the beginning of 1920 the total issue already exceeded 30 thousand million; during that year it had to be multiplied fourfold. In 1922 the increase became particularly rapid, though it did not reach the pitch recorded in Germany at the end of 1923. When 6 — 82 — the policy of stabilisation was initiated in December, the note issue of the Austro-Hungarian Bank totalled about four billion kronen. The effects of inflation were closely akin to those found in Germany. Prices rose in all markets, but not to the same extent or with the same rapidity. The dislocation of the purchasing power of the krone cannot, with the statistical data available1, be traced with the same degree of accuracy as was possible for Germany ; in particular, the lack of any satisfactory index number of wholesale prices before 1923 makes it impossible to show the various stages of dislocation. If, however, the index number of the dollar rate is compared directly with that of the retail prices of the principal foodstuffs, clothing, light and heat 2, it will be found that the differences in the purchasing power of the krone increased up till July 1922, at which date the index number of the dollar rate was about double that of retail prices. Soon afterwards the dollar exchange grew more stable and for some months retail prices rose fairly rapidly until the index number was very near that of the dollar exchange; they remained at this level during the whole of 1923. It should be noted, however, that in the circumstances retail prices still remained below world market prices, as the purchasing power of the dollar had itself depreciated and was no longer at its pre-war level. The cost of living taken as a whole (i.e. allowing also for rent, which had undergone only a trifling increase owing to the stringent control exercised by the Government) remained about 25 per cent. below the level of retail prices for articles in current consumption. Wholesale prices may be assumed to have been somewhere between the dollar exchange and the retail price level during the inflation period. At the beginning of 1923 (when an index number of wholesale prices was first published) they were about 25 per cent. above the dollar index; in the autumn of 1923, having steadily increased in the meanwhile, they were as much as 35 or 40 per cent. above it. Even at the last-named date the difference was not quite equal to the depreciation in the purchasing power of the dollar itself, which may be estimated at from 40 to 50 per cent. It may thus be said that wholesale prices in Austria were still below world prices. 1 2 The relevant data available are given in appendix VIII. Cost-of-living index number of the Paritätische Kommission, excluding rent. — 83 — Nevertheless, in this country there was no such reversal of the relations between the index numbers of retail and wholesale prices as was found in Germany. The reason is that the rise in prices was not so enormous as in Germany; there was not the same need for retail dealers to forestall an increase in wholesale prices. Only a very rough idea of the cost of labour can be gained from wage statistics, but the fact that real earnings were generally reduced, as is shown in chapter I, affords sufficient grounds for assuming that it increased much less than the other factors in costs of production. It can therefore be said that during the whole period of inflation Austrian industry reaped the full benefit of the rise in • prices and the dislocation of the purchasing power of the krone. The depreciation of the currency led, however, to great scarcity of capital, which very soon impeded economic activity. The rise in the level of wholesale prices above the dollar rate also had a depressing effect from 1922 onwards. The instability consequent upon extremely rapid inflation during the first six months of this year added further to the confusion. A crisis was imminent when the League of Nations intervened. Thanks to the energetic action of the Government and to the guarantee afforded by League control, inflation was arrested and a new bank of issue was founded; the reforms required to balance the budget were introduced ; first a short-term and then a long-term loan were successfully floated in the principal foreign countries; the krone exchange was immediately stabilised and the rise in prices became very gradual. The abrupt withdrawal of all the advantages of inflation placed industry in a position the great difficulty of which was further enhanced by the fact that the scarcity of capital had not yet been remedied; unemployment, which had appeared before the financial reforms, increased considerably. The task of restoring the economic system to a sound footing was continued during 1923 and 1924 and industry remained in an unsatisfactory position during the whole of this period. Unemployment remained serious and varied only under the influence of seasonal factors. CHAPTER I Real Income of the Working Classes DATA AVAILABLE AND THEIR INTERPRETATION The data from which conditions among Austrian workers can be determined are still more inadequate than in Germany. There is an equal absence of investigations into post-war working-class budgets suitable for the present purpose, but to make matters worse the available statistics of wages and unemployment are less complete than the corresponding information for Germany. The principal data in respect of wages refer to Vienna alone and are actually published for a certain number of industries only. But the capital city is of fundamental importance in the economic life of the country, so that these statistics may be regarded as affording some general indication of the movement of wages in the country as a whole. Before these statistics can be applied to the purposes of the present enquiry, however, certain reservations of material importance must be made as to the value of the figures reproduced in tables I and II. The index numbers of real wages given therein have been calculated by the International Labour Office from figures compiled by the Central Statistical Commission x and later by the Federal Statistical Office 2 on information furnished by the workers' organisations. In addition to this certain figures for the pre-war period have been taken from an article by Dr. Benedikt Kautsky, Secretary of the Vienna Chamber of Labour 3 . These data are, however, not always strictly comparable as between one period and another. From 1914 to 1922 most of the figures refer to minimum wages, whereas those for 1923 refer mainly to average wages or those most frequently paid. The difference between these and minimum wages, however, is slight, to judge by a comparison for the various industries which is possible with the December 1922 figures, and in many cases there is 1 2 Mitteilungen der statistischen Zentralkommission, Statistische Nachrichten. * "Löhne und Gehälter in Deutsch-Oesterreich", Die Wirtschaftskurve der Frankfurter Zeitung, 1923, No. 4. — 85 — no difference. It is thus probable that minimum wages were almost universally adopted in practice at the time. While there are good grounds for assuming that the same conditions prevailed in 1923 and 1924 and during the years immediately preceding 1922, it is extremely probable, on the other hand, that in 1914 the wages most frequently paid were slightly higher than the minimum wage. ¡ As the figures upon which the index numbers are based are mainly minimum wages, they therefore tend to underestimate the level of the most usual wages during that period, whereas, for similar reasons, the 1923 indexes overestimate it. It was impossible TABLE I. INDEX NUMBERS OF REAL WAGES OF SKILLED WORKERS IN CERTAIN IMPORTANT INDUSTRIES, VIENNA DISTRICT, ! 1920 TO 1924 {Base: 1913 = 100) Building Date Brick'ayers Carpenters Wood Metals Textiles Skilled Skilled Lace work- work- makers Weavers ers ers Clothing Chemicals Skilled workMen's Ladies' ers Tailors Printing Printers Bookbinders ' 1920 Dec. — 1921 Mar. June Oct. Dec. 75 88 81 86 1922 Mar. June Sept. Dec. 105 85 110 111 1923 Tan. Feb. Mar. Apr. May June July Aup. Sept. Oct. Nov. Dec. 1924 Jan. Feb. Mar. Apr. May June July Aug. Sept. — 58 69 40 72 84 54 85 83 108 64 67 62 64 74 88 100 82 106 107 79 67 54 90 69 65 81 83 111 108 104 105 10P 115 116 110 111 112 112 109 107 104 101 102 105 111 112 107 107 108 108 106 89 85 80 83 82 84 90 89 96 95 98 95 111 112 124 122 122 129 126 131 130 107 108 119 118 118 124 122 127 126 93 98 98 99 97 102 98 96 100 .— — — 50 44 — — 158 — — — — 73 72 85 62 64 67 75 55 — — — — 48 54 66 72 — 69 ,97 — 94 — — 81 — — 66 79 82 50 49 74 — — 76 94 98 67 57 85 86 77 •98 72 88 103 77 74 70 69 73 83 86" 86 84 83 82 105 88 85 84 85 84 92 96 95 93 92 102 99 74 72 71 72 70 75 79 78 76 75 78 76 83 80 80 81 8'. 88 91 87 110 HI HI 109 73 70 70 71 73 76 80 75 96 96 98 94 82 79 78 78 80 83 85 84 82 81 80 84 76 73 68 68 70 72 81 77 78 82 83 81 94 90 91 90 93 96 100 105 102 101 HO 107 103 102 102 100 98 98 94 93 110 96 100 100 99 97 97 93 98 97 74 77 77 76 74 74 71 75 74 HO 111 HI 115 116 117 116 117 117 96 97 97 100 100 102 98 102 101 82 85 85 84 82 82 82 81 84 87 88 88 87 92 93 92 95 94 109 112 112 110 117 116 112 119 118 — 66 62 — — 86 — TABLE II. INDEX NUMBERS OF REAL WAGES OF UNSKILLED WORKERS IN CERTAIN IMPORTANT INDUSTRIES, VIENNA DISTRICT, 1920 TO 1924 (Base: 1913 = 100) Metate Clothing — 86 — — 60 66 60 — 79 82 74 77 — — — — — — — — 121 94 121 122 82 55 56 93 84 80 102 104 107 73 62 92 — — 124 143 1923 January February March April May June July August September October November December 122 119 115 116 119 126 127 121 122 123 123 120. 92 88 83 86 84 87 93 92 99 98 102 99 98 94 88 87 92 104 109 108 106 104 103 116 81 87 88 89 92 96 100 95 116 116 116 114 137 131 129 130 134 138 141 139 136 134 133 139 1924 January February March April May June July August September 122 120 136 134 134 142 139 144 143 96 101 101 103 101 105 102 99 104 114 113 113 111 109 109 105 102 122 116 117 117 121 121 123 119 123 122 136 141 141 140 137 136 137 134 141 Date Building 1920 December 49 1921 March June October December 93 108 97 106 1922 March June September December Wood Chemicals to obtain more satisfactory information for the pre-war period, and this method therefore had to be adopted. Another point to be noted is that the index number of the cost of living used to calculate, real wages has the same defects as those already pointed out in the case of Germany. Based as it is on a budget theoretically framed to represent a post-war standard of living, it indicates the increase in expenditure required to main- — 87 — tain post-war conditions at a uniform level as from pre-war days. Had it been possible to construct an index number indicating the increase in expenditure required to maintain a pre-war standard of life, the increase would have been slightly greater than that shown by the index actually employed, and the real wages so calculated would have been somewhat lower than those given in the table. In order to determine real wages it was thought necessary to calculate their value (as was done for Germany) at the time when they were spent and not when they were earned. These indexes refer to the end of the month, and it was thus necessary to calculate an index number indicating the cost of living during the first week following the month considered for wages; this was obtained by interpolation from the indexes of the Joint Committee x. REAL WAGES OF CERTAIN CLASSES OF WORKERS, 1920 TO 1924 Notwithstanding the reservations made above, the index numbers of real wages shown in tables I and II give a general idea of the position. It will be seen that in December 1920 real wages in the various industries had already depreciated very greatly. At this time skilled workers received only from 40 to 70 per cent, of the real value of their pre-war wages. Unskilled workers were little better off, so far as can be judged from the few figures available. During the first half of 1921 real wages in the various trades rose continuously until in June the wages of skilled workers varied between 55 and 90 per cent, of real pre-war wages and those of unskilled workers between 70 and 110 per cent. The explanation is that prices increased much more slowly during that period, so that the workers were able by degrees to improve their position. From September 1921 onwards the rise in prices again gathered speed, and on the whole real wages depreciated during the autumn. From the end of 1921 up till September 1922 it was found possible to increase the real value of wages by adopting more or less systematic methods of adjustment to the cost of living in a large 1 The cost-of-living index numbers refer to the 15th of each month. To find the cost of living during the first week of any month an index number was calculated for the third or fourth day of the month. To find the index for 3-4 April 1921, for example, the formula used was as follows: (1 X March index) -f (2 x April index) 3 — 88 — number of industries. In the wood, printing, and clothing industries, however, after an increase at the beginning of 1922, wages failed to adjust themselves to the very sharp rise in the second quarter of the year, and again depreciated. Thus, at the time when the stabilisation of the krone was beginning to check the rise in the cost of living, real wages, expressed as a percentage of their pre-war value, exhibited great differences between various industries; they varied from 50 to 110 per cent, for skilled workers and from 60 to 125 per cent, for unskilled workers. From the autumn of 1922 onwards the cost of living was more or less stabilised. Real wages, which had fallen heavily owing to the rise in prices during the preceding period, increased again for some months; they then remained at approximately the same level during the year 1923; at the end of this year, and more particularly during the following year, they again showed a slight tendency to rise. In September 1924 the wages of skilled workers had, on an average, reached the pre-war level (with variations from 95 to 131 per cent, according to industry), while those of unskilled workers tended to exceed it (104 to 144 per cent.). Allowance should be made, in interpreting there estimates, for tendency to overestimate the level of real wages, since the figures used for calculating the pre-war series are somewhat low and refer in most cases to minimum wages and also because of the method of constructing the cost-of-living index after the war. The general movement of real wages emerges clearly from a calculation of yearly average index numbers, such as are given in tables III and IV. It will be seen that they increased steadily in all industries; those of skilled workers increased from 57-83 in 1921 to 76-124 in 1924 (first nine months) and those of unskilled workers from 62-101 in 1921 to 110-140 in 1924. TABLE I I I . YEARLY AVERAGES OF INDEX NUMBERS OF REAL WAGES OF SKILLED WORKERS IN CERTAIN IMPORTANT INDUSTRIES, VIENNA DISTRICT, 1921 TO 1924 (Base: 1913 = 100) Building Date 1921 1922 1923 1924 (first 9 months) Wood Metals Carpenters Skilled workers Skilled workers Lace makers Weavers 3 83 103 110 78 99 107 82 72 89 64 74 81 76 83 91 75 123 119 98 100 97 75 Brick- • layers Textiles — 89 — Printing Date 1921 1922 1923 1924 (first 9 months) TABLE IV. Printers Bookbinders Clothing Chemicals Tailors Skilled workers Men's Ladies' Average (unweighted) of these 7 industries 57 72 76 78 90 98 73 77 93 75 64 81 81 81 73 81 89 91 114 114 99 83 101 YEARLY AVERAGES OF INDEX NUMBERS OF REAL WAGES OF UNSKILLED WORKERS IN CERTAIN IMPORTANT INDUSTRIES, VIENNA DISTRICT, 1 9 2 1 TO 1 9 2 4 (Base: 1913 = 100) Date Building 1921 1922 1923 1924 (first 9 101 114 121 months) 135 Metals Clothing i Chemicals Average (unweighted) of these industries 62 71 92 78 92 101 84 100 133 135 80 99 110 110 114 120 138 123 Wood 1 These figures refer to semi-skilled tailors, who are of a somewhat higher class than the workers in the other industries given. As was pointed out for Germany, the indexes of real wages merely indicate the fluctuations in the income of normally employed workers. To obtain some indication of the aggregate income of the industrial population several other factors must be taken into account, the most important being the extent of unemployment. The problem of measuring the reduction in aggregate workingclass income through unemployment is yet more difficult in Austria than it was in Germany, owing to the nature of the statistics available for measuring the extent of unemployment. The chief of these are supplied by the unemployment insurance service, which establishes each month the number of unemployed in receipt of relief, but the way in which they are published makes it hardly possible to calculate an accurate percentage and they give no definite information on the subject of partial unemployment. They can, however, be supplemented by the results of an enquiry among the various workers' organisations undertaken by the Trade Union Committee to ascertain the percentage of totally and — 90 — partially unemployed among their members in December 1922. This enquiry, which was carried out on sufficiently broad lines 1 to render the results fairly representative of the whole country, gave the proportion as 33.24 per cent, totally unemployed and 22.3 per cent, partially unemployed. With these two kinds of data, it is possible to make an approximate estimate of the volume of total unemployment at various dates. From thé unemployment insurance statistics the number • of unemployed registered at any given date can be expressed as a percentage of those registered in December 1922. If these figures are applied to the percentage of total unemployment in December 1922, as ascertained by the trade union enquiry, an approximate estimate of the percentage of total unemployment at various dates can be made 2. Subtracting from these figures the average amount of unemployment before the war, i.e. 4 per cent. 3 , gives a result representing the increase in unemployment since the war. 1 The aggregate membership of the organisations consulted is over 600,000, and covers most of the important industries. 3 The results of these calculations are as follows : Date Percentage of totally unemployed (trade union enquiry) (1) (2) N u m b e r of unemployed as percentage of figure for Dec. 1922 (unemployment insurance statistics) (3) Estimated percentage of totally unemployed, from columns (2) and (3) (4) 1922 April July December 33.24 — — 37.6 26.3 100.0 12.5 8.7 33.2 1923 February October — — 142.0 64.3 47.2 21.4 1924 February June September — — 106.6 53.9 66.1 35.4 17.9 22.0 • 3 The only statistics giving an idea of the percentage of unemployment before the war are those published by the trade unions, which refer to the former territory of Austria. They indicate unemployment amounting to 5 or 6 per cent, during the first six months of 1914. This was, however, an exceptional period, and just as wages for 1914 are only taken as typical of pre-war figures, so a typical percentage of unemployment is used. The employment exchange statistics show that unemployment was on the increase from 1910 to 1914; minimum unemployment is generally estimated at 2 per cent., and it may therefore be taken that the average percentage during the few years immediately preceding the war was about 4 per cent. — 91 — Finally, some idea is required of the income of the unemployed represented by the various forms of relief. This again is a very difficult matter, especially for the period since the autumn of 1922, when the accelerated rise in prices necessitated a constant increase in the rate of allowances, though their real value depreciated rapidly. • It may nevertheless be estimated that from that date onwards the relief given to totally unemployed workers was equal to about one-quarter of average wages for skilled workers and to one-third for unskilled workers. The percentage reduction in aggregate working-class income through total unemployment may thus be estimated for skilled workers at three-quarters of the unemployment percentage calculated, and to two-thirds of this percentage for unskilled workers. This gives the following figures for the dates at which unemployment reached its maxima and minima since 1922. TABLE V. ESTIMATED PERCENTAGE REDUCTION IN AGGREGATE WORKING-CLASS INCOME THROUGH TOTAL UNEMPLOYMENT, 1922 TO 1924 Estimated percentage of totally unemployed Estimated excess of total unemployment over pre-war figure Estimated reduction in income through total unemployment Per cent. Per cent. Per cent. 1922 April July December 12.5 8.7 33.2 8.5 4.7 29.2 6 3 19-22 1923 February October 47.2 21.4 43.2 17.4 29-32 11-13 1924 February June September 35.4 17.9 22.0 31.5 13.9 18.0 21-24 9-10 12-13 Date Although the reduction in the aggregate income of the industrial population was relatively slight up to the autumn of 1922, it became very great from the end of that year onwards and especially during the beginning of 1923. The reduction remained very great up till the spring of 1924, although there was some improvement during the summer of 1923. Throughout this period the losses due to total unemployment must have largely counterbalanced the — 92 — gains which those workers who remained in employment derived from the rise in real wages; by the end of 1922 and the beginning of 1923 they must even have exceeded these gains. From the summer of 1924 onwards, however, the workers' position seemed to be definitely improving as unemployment showed a tendency to decline (apart from seasonal fluctuations) and real wages a fairly general tendency to rise. No precise information being available, it has not been possible, up to this point, to take into account the reduction in income caused by partial unemployment. Unemployment insurance statistics give no information on the subject. The trade union enquiry shows, however, that in December 1922 the proportion of partially unemployed among trade unionists was about 22 per cent. Assuming that the development of partial unemployment was on the whole parallel to that of total unemployment, one may conclude that the percentage was somewhat less during 1922, and materially greater at the beginning of 1923, this level being again practically reached at the beginning of 1924 after a temporary fall during the second half of 1923. The relief paid to partially unemployed were much less than that payable to totally unemployed, no payments being made from unemployment insurance funds; the aggregate income of the industrial population therefore suffered a further reduction through partial unemployment. The position in Austria is thus similar to that described in Germany. Real wages were at first depreciated by inflation ; later, with the advent of stabilisation, the income of part of the working population increased again thanks to a rise in real wages, but the income of other sections underwent a considerable reduction through unemployment. POSITION OF PROFESSIONAL WORKERS Professional workers were in no better position than manual workers. Although the available statistical data are inadequate for a thorough investigation of the question, the figures given below in respect of real salaries 1 paid to unmarried officials afford a fair indication of the position. 1 The real value of salaries has been calculated according to the index of the cost of living for the same month as that in which the salaries were paid. At the period under review the cost of living did not fluctuate very greatly from one month to the next, and it is unnecessary to distinguish between the real value of salaries when they were paid and when they were spent. — 93 — It should be noted, however, that the real value of these salaries was calculated according to the cost-of-living index number of the Joint Committee, which is based on working-class conditions. It is highly probable that the rise in cost of living was still greater for professional workers than for manual workers, notwithstanding the relatively greater benefit which the former derived from rent control. Consequently, the figures given in the appended table do not fully indicate the reduction in the income of this class of workers. Again, the figures given refer only to unmarried officials. Family allowances have therefore not been taken into consideration ; in any case they only represent a trifling increase in basic salaries. TABLE VI. INDEX NUMBERS OF REAL SALARIES OF CERTAIN CLASSES OF STATE OFFICIALS (iN THEIR TENTH YEAR OF SERVICE) IN VIENNA, JUNE 1923 AND JULY 1924 (Base: 1914= 100) Class June 1923 July 1924 Upper-grade officials (with university training) Middle-grade officials Lower-grade officials Subordinate officials 49 56 90 82 65 68 96 82 In June 1923 the depreciation in salaries was still considerable for the upper and middle grades of state officials ; their real value was barely half the pre-war figure. Those in the lower grades and subordinate posts were less severely hit, the index numbers of their salaries being roughly the same as those of the real wages of skilled workers at the same dates. In some of the very low grades (there is no information for them comparable with that given in the table) real salaries were then equal to those paid before the war x. It must not be forgotten, as has already been pointed out, that in view of the special conditions of professional workers the real values indicated by the foregoing calculations are an even greater overestimate than those given for manual workers. In general, therefore, most professional workers were in a worse position than manual workers. 1 Cf. LEAGUE OF NATIONS: Twelfth Report by the High Commissioner of the League of Nations for Austria. Doc. C. 13, 1924, II. — 94 — It is difficult to establish a continuous series of figures showing the salaries paid to officials during all the post-war years, as the methods of fixing them were very frequently changed and often render comparison extremely difficult. However, as in June 1923 the cost of living had been more or less stable for some six months one is led to conclude that during previous years, when prices rose very steeply, the depreciation in salaries was even greater than that shown in the table. Salaries were, it is true, automatically adjusted as from June 1922, but only to the cost of living in the preceding month; their depreciation must consequently have been greater during the second half of 1922 than in June 1923. Table VI also shows that real salaries increased somewhat in June 1924, but the increase was not so great as that in manual workers' wages. The greatest increase occurred in the two higher grades, but even their position remained somewhat precarious. The real salaries of the officials of lower and subordinate grades shown in the table were by that time little below pre-war figures. In interpreting these figures allowance must be made for the fact that, unlike manual workers, these officials did not suffer from unemployment; on the other hand, a large number have been dismissed during recent years, and those who were unable to find a position in private industry have been reduced to living on their pension. No statistics are available for a comparison of the position of professional workers in private employment before and after the war. As a general rule their post-war salaries were slightly higher than those of officials and their position might be regarded as somewhat better, but for the fact that they suffered from unemployment, which did not affect the state officials who remained in the service. CHAPTER II Measures to Check the Fall in Real Wages l Among the measures taken to remedy the decrease in real wages the first to be examined, as in the case of Germany, will be those intended to check the rise in the cost of living or to measure it as accurately as possible. Subsequently an account will be given of methods of adjustment, properly so called, of wages to the cost of living. COST-OF-LIVING PROBLEMS Price and Rent Control and Food Subsidies During the war and the ensuing years the Government undertook to control the prices of foodstuffs and to grant subsidies for feeding the population. The prices of a large number of foodstuffs in current consumption were fixed and the commodities themselves were methodically rationed. In the case of certain staple foods (bread, flour, fat) the authorities went even further and entirely controlled supplies; they bought enormous stocks of such foodstuffs and re-sold them to retailers, generally at a lower price, the difference being covered by the state. The amount of these indirect subsidies to the whole population varied according to the prices at which the Government was able to buy and also according to the prevalence and intensity of distress in the country. The effect of these various measures was to keep the prices of staple foodstuffs well below what they would have been had economic forces been allowed free play. But this period also witnessed a widespread development of illicit trade, which afforded facilities for obtaining more than the official ration at prices considerably above those fixed by the Government. As in Germany, this state of affairs made it very difficult to measure the cost of living accurately. At the end of December 1921 the policy of control 1 In the compilation of this chapter, more particularly the section on methods of adjusting wages to the cost of living, considerable use has been made of an article by Dr. Carl FORCHHEIMER: "Sliding Wage Scales in Austria", in the International Labour Review, Vol. X, No. 1, July 1924, pp. 30-47. — 96 — was finally abandoned. Prices thereupon rose considerably above the official rates, but on the other hand the exorbitant prices quoted on the illicit market practically disappeared. Rent control was more effective. A serious shortage of housing had developed since the war and, had the Government not intervened, enormous increases in rent would undoubtedly have followed. The protection afforded to tenants was very complete, and was still further increased by the fact that the Government was compelled to organise systematic distribution of available accommodation in order to secure that all sections of the population were housed. Rents were thus kept at a level very much lower than that of the retail prices of articles in current consumption 1 . The difference emerges clearly from the following table showing the index numbers of rent set against those of the cost of living and of the retail prices of foodstuffs, clothing, lighting and heating. TABLE I . INDEX NUMBERS OF THE COST OF LIVING, RETAIL OF FOODSTUFFS, CLOTHING, LIGHTING AND HEATING, PRICES AND RENT IN VIENNA, 1921 TO 1924 (Base: 1914 = 1) Date 1921 December Food Clothing Lighting and heating Cost of living Rent i 579 1,287 518 4 527 1922 March June September December 904 2,421 13,531 10,519 1,428 2,712 19,159 15,829 860 1,670 12,638 13,503 14 21 33 166 790 1,866 11,271 9,375 1923 March June September December 11,637 14,132 12,509 12,860 15,152 15,323 15,981 18,673 15,068 14,364 15,646 14,525 346 496 524 724 10,151 11,513 11,841 11,249 1924 March June September 13,930 13,457 15,623 19,521 20,634 21,088 14,889 14,673 14,760 751 1,002 1,024 11,996 12,442 13,162 ; • The increase in rent being so trifling as compared with that in the prices of articles of current consumption, the item of expen1 Cf. INTERNATIONAL LABOUR OFFICE: European Housing Problems since the War. Studies and Reports, Series G, No. 1. Geneva, 1924. — 97 — diture on rent in working-class budgets was reduced to practically nothing; in July 1923 it was barely one per cent, of the total expenditure of a working-class family, whereas before the war it might be estimated at about 20 per cent. This change, however, directly benefited certain classes of workers only, those who, having obtained wages approximately equivalent in gold value to pre-war rates, had practically no rent to pay. But this applied only to unskilled wage earners in a few industries. Most of the workers were in the same position as those of Germany; they had practically no liabilities under the heading of rent, but the corresponding amount was not included in their wages. The actual gain was thus nil. Industry, on the other hand, benefited, as in Germany, by the reduction in the cost of most labour by the full amount which Tent represented in wages before the war. From the autumn of 1923 onwards the situation was altered somewhat, as rents rose slightly more rapidly than the prices of articles of current consumption. Measurement of the Cost of Living The problems of the measurement of the cost of living did not, during the first years after the war, arouse the same interest in Austria as in Germany. When the rise in price became rapid there were no means available for measuring it. This was a particularly difficult task owing to the existence of two markets for the majority of commodities: one on which official prices were quoted and the other, the illicit market, on which prices reached extraordinary figures. Neither could offer anything hut an entirely false indication of the position. Again, if it were proposed to combine the prices of both markets, what relative value was to be attributed to each series ? The fact of using illicit prices in the preparation of an official index number would, moreover, have involved official recognition of illicit trade, to which theGovernment objected. Up till 1920, therefore, the parties to wage negotiations had to resort to rough estimates of the rise in the cost of living arrived at by very rudimentary expedients. Only at the beginning of 1920 did the Federation of Salaried Employees in Industry begin to record the cost of living, but the index number constructed by a joint committee was restricted in the main to commodities then controlled by the state. The workers in the metal industry undertook to prepare another index, based •on a more complete list of articles, and used illicit prices for the 7 — 98 — excess over official rations in the quantities taken for the standard budget adopted. The results obtained by the two organisations were naturally substantially different. The metal workers' index showed a greater increase over pre-war prices than that of the salaried employees. The reverse occurred when government control was abandoned, and the prices of articles formerly controlled rose considerably; the employees' index, which was based strictly upon these prices, showed a much greater increase than the other. At the beginning of 1921 the Federal Statistical Office began the publication of an official index number. This was based on a theoretical budget representing post-war conditions of life; the food item was revised every month so as to secure a certain nutritive value for a minimum price, allowing for market conditions. Thi& index number was especially criticised in regard to the method of determining the food item, which, if it was not to be unduly arbitrary, required precise information as to market conditions. which was not readily obtainable. The scientific principles on which the budget was compiled were often little understood b y the working classes and aroused distrust. As a result the index number was little used for adjusting wages. In many industries the cost of living was still estimated by the rough methods previously employed. At the end of 1921, when the food subsidies were abolished and cost-of-living bonuses and family allowances were introduced tobalance the consequent rise in prices, the Government appointed a Joint Cmmittee to construct a cost-of-living index number. This committee comprised employers and workers' representativeswith a number of experts who were to ascertain the prevailing; prices of various commodities on the 15th of each month. This index number, like that of the Federal Statistical Office, covered four groups of commodities : food, clothing, rent, and lighting and heating; it was based upon a fixed standard budget representing a post-war standard of living. It was adopted almost without exception in all negotiations. METHODS OF ADJUSTING W A G E S TO THE COST OF LIVING As regards the adjustment of wages three main periods can bedistinguished. During the first of these periods, which ended with 1921, themethods employed varied greatly from one industry to another.. Some were very crude, others more complex, but they were all. — 99 — empirical. To begin with there was no satisfactory index number and the contracting parties were compelled to estimate the rise in prices as best they could. A great variety of principles was applied, but the prevalent aim was to preserve only the minimum subsistence from the effects of monetary depreciation. During the second period, which lasted from the beginning of 1922 to the autumn of 1923, the index number prepared by the Joint Committee was available and universally accepted. Nearly all wages and salaries were more or less automatically adjusted according to this index number, although the employers soon began to oppose the principle of the sliding scale. The beginning of the third period may be placed in the second half of 1923. The practice of automatic adjustment gradually disappeared; the clauses in collective agreements providing for this procedure were abolished or replaced by more elastic provisions which left it to free negotiation to determine how far the rise in the cost of living was to be reflected in wage rates. Whereas during the first and second periods adjustment to the cost of living was in the forefront of the negotiations, subsequently economic conditions in the country and in the industry concerned came once more to be regarded as important. The Sliding Scale First Period. Towards the end of 1919 the Government convened an industrial conference. Representatives of the Government, employers, and workers were summoned to discuss the question of wage adjustment and if possible devise a method to be adopted once and for all in negotiations between employers and workers. There were two problems to solve: the establishment of a costof-living index number and the adaptation of wages in accordance with it. Agreement was reached on the first item of the programme, but the parties remained at variance as to the application of the index number to the calculation of wages. It was generally agreed, however, that the whole wage should not be adjusted to changes in the index; as the quantity of goods available for consumption remained unchanged, such a procedure would merely have accelerated the rise in prices and produced a vicious circle. It was only over the detailed application of the principle that opinions differed. The Chancellor, Dr. Renner, proposed that wages should be divided into two portions, one fixed and the other variable. The latter, being required to cover the cost of a minimum — 100 — subsistence, should be adjusted in full to the rise in prices. The former, which represented the difference in standard of living between social classes, was to remain constant. Not only did such a method raise very grave difficulties in fixing the relative size of the two parts of the wage, but it had the further defect of gradually obliterating the distinction between skilled and unskilled workers' wages. Other suggestions were advanced. The salaried employees in private undertakings proposed that the variable portion should not be a fixed amount for all classes of workers but à uniform percentage of their wages. In the course of these discussions the workers' representatives were faced with a very awkward dilemma. During periods of rising prices it was to their advantage that the variable portion of the wage should be as large as possible. On the other hand, in the event of a fall in the cost of living, it was preferable that the stable fraction should be larger so that the gains secured by previous negotiations should not be lost. In the end the question remained unsettled. The parties to collective agreements were left to specify the details of the method to be employed. The conference had nevertheless secured general acceptance for the principle of employing a cost-of-living index number in adjusting wages. In 1920 estimates of the rise in prices were made in a large number of industries, and methods were also devised for adjusting wages thereto in some measure. These methods differed from one industry to the next. In the case of salaried employees in private undertakings it wras at first agreed in principle that the rise in the cost of living should be reduced by one-fifth at most and one-tenth at least before it was applied to wages ; later on a somewhat complex mathematical formula was employed in calculating cost-of-living bonuses. ' In the metal industry, which employs a very large proportion of the Austrian working-class population, cost-of-living bonuses were calculated not on the percentage increase in the index number, but on the absolute rise in price of given quantities of certain commodities. The disadvantage in this system being that it more or less levelled out wages, the following proportional increase system was adopted: the absolute increase in prices was compared with the wage of a typical worker, and the percentage increase thus obtained was applied to the different wage classes. This method was introduced in a large number of other industries. — 101 — The cost-of-living bonus for state officials was directly calculated from the cost of a number of staple commodities in quantities corresponding to the official rations ; the bonus was further multiplied by the number of dependants. Second Period. The publication of the index number prepared by the Joint Committee instituted* in December 1921 under the Act abolishing food control marked the opening of the second period. This index number soon enjoyed the general confidence, as extensive publicity was immediately given to the methods employed in its calculation and for the further reason that workers' and employers' representatives assisted the official experts. Thus the practice spread of adjusting wages to this index. The principle previously applied in most industries, of adjusting to the cost of living only that part of the wage which corresponded to the minimum subsistence, had had to be abandoned in October 1921 owing to the enormous rise in prices. The rise since the preceding month amounted to 80 per cent, according to the metal workers' index and 90 per cent, according to that of the salaried employees. This being so, adjustment restricted to such portion of wages as represented the minimum subsistence would not only have eliminated any difference between high and low wages in a few months but would also have materially decreased the real value of the total wage. When the Joint Committee's index was published, the whole wage and not only part of it was adjusted. At this period, as also in the beginning of 1922, the system of completely automatic adjustment (known as the sliding-scale system) was generally applied in most industries. Moreover, in June 1922 it received official sanction, which encouraged its extension to private undertakings. The salaries of public officials had not hitherto been adjusted very methodically to the cost of living. The Act of 13 July 1921 laid down that such salaries were to comprise a basic salary, graduated according to rank and seniority, and a local household allowance and cost-ofliving bonus varying with the cost of living in the place of residence. Owing to the very rapid increase in prices immediately after this Act had come into force, constant adjustment became necessary; a further Act, dated 28 June 1922, therefore introduced the system of automatic adjustment to fluctuations in the cost of living. Basic salaries (inclusive of the household allowance in force in March 1922) were multiplied by a coefficient calculated from the — 102 — Joint Committee's index; the cost-of-living bonus instituted by the Act of 13 July 1921 was retained in its old form, unaffected by adjustment. At about this time, however, the employers began to oppose the sliding-scale system, on the ground that allowance should also be made in fixing wages for the economic position of industry, which was becoming difficult and was already reflected in fairly widespread unemployment. In February 1922 the employers invited the trade union committees to examine with them the question of adapting the slidingscale system to the industrial situation. The results of these negotiations were negligible, but in the following May the employers intensified their campaign against the system. During the spring and summer prices rose very rapidly. During the first months of this period wages continued to be adjusted in full to the cost of living, but in September negotiations grew very difficult and the presidents of the Rational Council were compelled to intervene ; the workers' earnings were then increased by only two-thirds of the rise in the cost of living. Third Period. At the end of 1922 came the stabilisation of the krone. One effect of this was to aggravate the economic depression which had begun during the year. At the same time the rise in the cost of living became very slight and there was even a fall in the index numbers in some months. The sliding-scale system therefore began to be abandoned by degrees, though this aroused keen irritation among the workers. Disputes became frequent, and in the autumn there even appeared to be some danger of a general strike. The position grew so serious that Mgr. Seipel, the Chancellor, felt bound to intervene. In November 1923 the sliding scale was finally abandoned in the collective agreement for the metal industry; shortly afterwards, following the negotiations which took place in December, the system was also discarded for state officials, after having been applied since June 1922. These two examples were followed in certain other industries. For workers in the chemical industry, for instance, bonuses varying with the cost of living had been abolished in August 1923; the agreement of December 1923 contained no reference to a sliding scale. In the textile industry the collective agreement which came into force on 1 January 1924 did not entirely abolish the use of a — 103 .— cost-of-living index number, but greatly restricted it. The costof-living bonus was fixed by a joint committee, which was required to allow as far as possible both for the cost of living and for the conditions prevailing in the industry. In some cases, however, automatic adjustment was continued. In the building industry, for instance, the sliding scale, though temporarily abandoned in May 1922, was still used up to the end of 1923, and the principle was retained in the collective agreement concluded at that time. Similarly, bank employees' salaries were still calculated according to the cost-of-living index, and its use was not challenged during the strike of February 1924. The system was similarly to be found in agreements for other subsidiary branches of industry; but as time went on the principle was less and less strictly applied, and employers prevailed in their claim that when wages were fixed the economic position of the industry should be considered first, or at any rate concurrently with other factors. They strove especially to escape the obligation to adjust wages automatically to the cost of living, as this prevented any forecast of the cost of labour. It is difficult to form an idea of the results of the various methods of adjustment described, but an examination of the wages actually paid in the various industries, according to the data given in the preceding chapter, affords some basis for estimating their respective merits. In the first place, tables III and IV show yearly average index numbers of real wages, which give a fairly clear indication of the general development. In 1921 adjustment was still somewhat rudimentary, as it was in most cases not automatic and affected only a part of wages. The consequence was that throughout this year the real wages of skilled and unskilled workers depreciated; at the same time they show an improvement on the figures for December 1920, when the adjustment methods were still less satisfactory. In the autumn of 1924, when the increase in prices became very rapid, the unsatisfactory character of these methods became patent. Tables I and II show that at this time real wages again depreciated. Towards the end of 1921, this state of affairs led to the adoption of the principle of adjustment in full to the cost of living, the result being that real wages rose in December 1921. During 1922, when the system of automatic adjustment reached its greatest extension, the average level of wages was distinctly higher than during the — 104 — preceding year. This tendency is clearly apparent, although in some industries inadequate methods of adjustment led to a temporary fall at the moment when the rise in prices became most rapid. During 1923 and 1924 the average level of real wages continued to rise, although at this time, and especially from the middle of 1923 onwards, the sliding scale was gradually being abandoned. The reason is that the rise in prices had by then practically ceased and the variations in the cost of living were small. Thus in spite of the progressive abandonment of the system of automatic adjustment the workers gradually succeeded, by repeated negotiations, in increasing their real earnings; at the beginning of 1924 the real wages of skilled workers were little below their pre-war value, while those of unskilled workers were even above it. The comparatively favourable position of the latter is to be attributed, it would seem, to the practical disappearance of expenditure on rent. To sum up, when prices had remained relatively stable for a year, the depreciation in real wages, which was largely due to the delay in adjusting wages (despite the employment of automatic systems), had completely disappeared. Workers were able to secure real wages equivalent to, if not higher than, their pre-war earnings. In spite of these favourable conditions, however, the aggregate income of the working-class population was still, owing to unemployment, less than it had been before the war. It is interesting to note that during the period when the sliding scale system was most widely used, i.e. in 1922 and the beginning of 1923, real wages continued to exhibit more or less irregular fluctuations which differed little in extent from those recorded during the years when this system was much less widely used, i.e. in 1921, at the end of 1923, and in 1924. The reason is that, on the one hand, automatic adjustment was always effected according to a cost-of-living index number referring to an expired period and, on the other hand, as such adjustment was only carried out at monthly or even two-monthly intervals, some considerable time elapsed between the time when prices were recorded for the purposes of such adjustment and the time when wages were spent. During this intervening period the cost of living continued to rise and the real value of wages was consequently more or less reduced. Thus the very sharp rise in June 1922 entailed a material decrease in real wages, although the latter were generally adjusted automatically to market condi- — 105 — tions during the preceding month. The fall in prices at the end of 1922, following upon a rise, brought, on the contrary, an increase in real wages. At the beginning of 1923 a renewed rise in prices led to a fresh reduction in real wages. In May of the same year the rise became less rapid and prices fell in July and August, real wages being thereby once more increased. Cost-of-Living Bonuses and Family Allowances The preceding section described the various attempts to solve the question whether wages were or were not to be adjusted automatically to the cost of living. A number of subsidiary questions, resembling those to which attention has already been drawn in the case of Germany, also came up for discussion. During 1920 and 1921, as stated above, it was generally decided that, to obviate too rapid an increase in prices, only a part of wages should be adjusted to fluctuations in the cost of living. The practice then arose of distinguishing between a fixed basic wage and a cost-of-living bonus, which was usually a flat rate for all workers but varied with prices. It will be recalled that the result of this system was a gradual levelling up of the wages of skilled and unskilled workers. Towards the end of this period the extremely rapid rise in the cost of living made it necessary to adjust the whole of wages. At the same time the distinction between the fixed basic wage and the variable additions was very often retained; the latter, however, were not uniform for all classes of workers in a given industry, but were reckoned proportionately to the basic wage, so that the total earnings of the various classes of workers in the same industry were increased in the same proportion. This system had the further advantage that it did not accentuate the levelling process, the drawbacks of which were beginning to be realised. The differentiation in wages was not, however, restored to its pre-war scale; the tendency to revert to the pre-war position did not appear until later. This development can be followed in certain industries for which figures of skilled and unskilled workers' wages are available. Table II gives the wages of skilled workers as a percentage of those of unskilled at the same dates. It will be seen that the fluctuations in the different industries were not uniform. In the metal industry skilled workers' wages, which in 1914 were 33 per cent, above those of unskilled workers, were only — 106 — TABLE II. PERCENTAGE RELATION OF NOMINAL WAGES OF SKILLED WORKERS TO THOSE OF UNSKILLED WORKERS IN CERTAIN INDUSTRIES, VIENNA DISTRICT, 1 9 1 4 TO 1 9 2 4 Jletals Building Clothing Chemicals 1914 133 136 148 203 1921 March June October December 109 110 111 111 110 111 113 111 — — — — — — — — 1922 March June September December 110 108 106 106 118 124 124 124 136 136 136 136 — — 122 120 1923 March June September December 106 105 105 120 124 124 124 124 136 136 141 141 120 120 121 121 1924 March June September 120 120 120 124 124 124 141 141 141 120 120 119 Date 10 per cent, above them in 1921, and 5 or 6 per cent, in 1923. Only in December 1923 did a sharp reaction restore the difference to 20 per cent. In the building industry the difference in 1921 became as slight as in the metal industry (about 10 per cent.), but the reaction took place earlier. By June 1922 the wages of skilled workers were 25 per cent, above those of the unskilled, and this level was roughly maintained during the following years. For the clothing industry no information is available prior to 1922 ; by this time the difference was little less than before the war, and by the end of 1923 it was even nearer the pre-war figure. In the chemical industry the wages paid to the classes of workers used for comparison showed a very substantial difference in 1914, the wages of skilled workers being more than double those of the others; in 1922 and 1923 the difference remained at something like 20 per cent. When prices were rising very sharply the difference in cost of living as between different localities grew steadily more marked • — 107 — (as was noted in Germany). It therefore became necessary, in collective agreements covering large areas, to grade bonuses according to the local cost of living. This system was very widely adopted and was applied in particular to metal workers and state officials. The very fact that the cost of living was in the forefront of wage negotiations called attention to differences in family responsibilities, and the practice of granting allowances proportionate to such responsibilities was resorted to in many cases. Family allowances were granted to state officials from February 1916 onwards. After many modifications during the ensuing years, the system was confirmed by the Act of June 1922, which introduced automatic adjustment of salaries, though family allowances were not necessarily subject to this. At the end of 1921, and for some time afterwards, the family allowance system spread widely in private industry. The reason was that, when the Government decontrolled prices and withdrew the subsidies previously granted in respect of certain staple foods, it compelled employers to pay all their staff supplementary bonuses in the form of family allowances in order to compensate for the anticipated rise in prices. It also established an equalisation fund, to which employers contributed in proportion to the number of workers employed by them, and from which they drew the sums required to pay allowances in respect of children. The supplementary allowances for the worker himself and his wife were not paid out of this fund, and it was not long before they were gradually merged in wages. Family allowances alone were retained, and at first were frequently adjusted to the rise in prices. In June 1922 the amount per child was fixed at 1,155 kronen (i. e. slightly more than 3 per cent, of the average wage) ; but after that date the allowances were no longer increased, and as wages continued to follow the general upward movement of prices they gradually lost their value, until eventually they were barely equivalent to 0.3 per cent. of the average wage. Thus the system established on a national basis at the beginning of 1922, and applicable to all workers, gradually lost its utility. Side by side with this national system there were many examples of collective agreements providing for family allowances. Among the industries where the allowances were most generally adopted may be cited the mining and chemical industries. The Vienna Union of Salaried Employees also adopted them. Generally, however, they represented only a trifling addition to wages and the establishment of equalisation funds was not necessary. P A R I ' III POLAND INTRODUCTION General Economic Conditions The general economic conditions which prevailed in Poland from the Armistice onwards closely resemble those already described in Germany and Austria. Poland, like those two countries, passed through a period of intense monetary inflation which dominated all factors in economic life; subsequently she entered upon a period of financial stabilisation, which was accompanied by a severe economic depression. The fact that the new state was constituted from territories formerly belonging to Russia, Germany, and Austria was a source of further very complex problems ; the economic system had to be completely reorganised and made to pivot upon fresh centres. Although monetary inflation did not reach such ah incredible pitch as in Germany, it was none the less enormous. The process began towards the end of 1919 after a period of relative stability, and with a few brief intervals continued up till the beginning of 1924. The note issue rose to 1,098 million Polish marks at the beginning of 1919; two years later it was something like 50,000 millions. By the beginning of 1923 it had multiplied more than tenfold and reached 793,000 millions. During this year, and particularly during the latter months, inflation increased at a stupendous rate ; in February 1924 the total note issue was 313,659,000 million Polish marks. At the beginning of 1924, however, the Government was given special powers with a view to restoring the finances to a sound footing. Drastic action was immediately taken. The budget was balanced within a very short time, the conversion of state revenue to gold values being systematised, the levying of taxes being expedited — especially the tax on capital — the customs tariff revised, and the railways reorganised so as to do away with the deficits which were a burden on the Treasury. Paper marks were issued with great prudence and only in the amounts necessary to meet the needs of commerce and purchase foreign securities. The Government's promise to make no further use of the printing — 110 — press to cover expenses having thus been made good, the Polish mark was finally stabilised 1 . At the end of April a new bank of issue was founded in the form of a private company with a capital of 100 million gold francs. It was commissioned to issue a stable official currency, the zloty or Polish florin 2, secured by a gold reserve of not less than 30 per cent, and by foreign securities. The new currency quickly supplanted the paper mark, which ceased to be legal tender as from 1 July 1924. As in other countries, however, monetary inflation had resulted in practically uninterrupted depreciation of the purchasing power of money, coupled with a dislocation of purchasing power as between various markets. Both aspects of this twofold development can be traced by means of the available statistics of prices quoted in four principal markets; the foreign money market (the dollar rate of exchange), the wholesale market, the retail market, and the labour market 3 .' During the greater part of the inflation period the fall in purchasing power generally reacted more quickly and powerfully on the foreign market — especially the foreign money market — than on the home market. Only in the autumn of 1923 was the position reversed; from September 1923 onwards the index number of . wholesale prices was permanently above that of the dollar. The repercussions of inflation were less immediately felt in the retail market than in the foreign security or wholesale markets. The adjustment of retail to wholesale prices was still several months in arrear in 1922; this delay was gradually shortened, and in the beginning of 1924 the index number of the cost of living rose above that of wholesale prices. It should be noted, however, that this index number was kept at a lower level than that of the prices of most articles in current consumption in consequence of government control of rents, which did not rise to any great extent. If the index numbers of wholesale prices and of retail prices of foodstuffs only be compared, there is a much smaller difference than between the indexes of wholesale prices and the cost of living. The foregoing sketch of conditions particularly applies to the former Russian territories; in those parts of the country which were previously German or Austrian, the situation was further 1 This stabilisation was effected at first at the rate of 10 million Polish marks to the dollar. The rate was subsequently reduced to 9,300,000 millions. 2 The ratio between the two currencies was fixed at 1,800,000 Polish marks to 3 the zloty. These statistics will be found in appendix X. — Ill — complicated by the transition from the old to the new unit of currency. In Poland, as in the other countries, it is extremely difficult to ascertain exactly the changes in the cost of labour, owing to the profound changes which the country underwent after the termination of hostilities and their repercussions on the conditions of work. Assuming, however, that nominal wages rose more slowly than the cost of living (as would appear from the statistics available) and that, on the other hand, weekly hours of work were generally reduced, it may be concluded that the cost of labour must have been subject to fluctuations not greatly different from those of the cost-of-living index number. Up till the autumn of 1923 inflation thus created conditions particularly favourable to production. The incessant and extremely rapid rise in prices placed industry in a privileged position throughout and counterbalanced most of the difficulties which the economic disorganisation consequent upon the war was likely to cause. The purchasing power of the mark being lower on the foreign market than at home, Polish industry possessed an assured advantage over foreign competitors. At the same time the cost of labour had not increased in the same proportion as other costs of production, and was therefore a smaller charge on the employer. At the end of 1923, however, the position was gradually reversed. The purchasing power of the mark in the home market fell below that in the foreign market, so that now industry was in an unfavourable position. It can be taken that at the beginning of 1924, when the cost of living had risen above the level of wholesale prices, labour accounted for a far greater proportion of production costs. In addition to this, capital had been growing steadily scarcer and could only be obtained at exorbitant rates, to which a premium to cover risks of depreciation was added. The policy of monetary reform which was initiated at the beginning of 1924 and the resulting stabilisation of prices further aggravated the industrial situation. The possibility of an imminent crisis had to be faced. Matters did not come to a head at once, but towards the end of the spring of 1924 the position became serious and unemployment grew alarmingly. Inflation having ceased to stimulate industry, the latter had to be reorganised on sound economic lines. To reduce costs of production and middlemen's expenses and extend credit became the great problems of the hour. CHAPTER I Real Income of the Working Classes ' DATA AVAILABLE AND T H E I R INTERPRETATION I t is even more difficult to form an idea of the living conditions of the workers from their incomes in Poland than in Germany and Austria. The only data available are, first, certain wage statistics for a few classes of workers in a few important towns, and, secondly, a cost-of-living index number for Warsaw alone 2 . No satisfactory information being procurable, it is impossible to say exactly how real wages compared with pre-war wages during the period prior to 1921. A few figures covering the period 19141921 were, it is true, compiled b y the Warsaw Institute of Social Economics, but no index number was constructed to show the general movement of the cost of living during those years. The available index number covers five foodstuffs only; from comparison of this index with the above-mentioned wage statistics, coupled with general information, it appears that real wages reached lowwater mark during the first half of 1918, and t h a t each of the three ensuing years witnessed some slight improvement. From 1921 onwards statistics of wages and of the cost of living were published by the Central Statistical Office at Warsaw; this gives a somewhat clearer view of the position. But the essential data for calculating index numbers of real wages in comparison with pre-war figures are practically available only for Warsaw. The statistics given in this chapter are therefore restricted to t h a t city. There are no wage statistics for the textile industry at Warsaw, as this industry is hardly represented there; the wages paid in Lodz have therefore been used, in combination with the Warsaw cost-of-living index number, on the assumption that the difference in this respect between the two towns was not very great. 1 In the compilation of this chapter, and more particularly of the following one, considerable use has been made of an article by Mr. J. SZTURM DE SZTREM : "Wage Problems in Poland during and after the War", in the International Labour Review, Vol. X, No. 3, Sept. 1924. 2 There are other cost-of-living index numbers for certain other important towns in Poland, but they have not been prepared in such a way as to permit of comparison with pre-war conditions. — 113 — It is impossible, however, to extend the cost-of-living index for Warsaw to the old German and Austrian territories as, from information received, prices in these areas appear to have been widely different from those recorded in what was formerly Russian Poland. All that can be said, judging from the available information, is that pre-war wages appear to have been lower in the formerly Russian districts than in the German and Austrian provinces. Since 1914, on the other hand, the increase appears to have been proportionately greater. Certain reservations must be made even in respect of the Warsaw figures. In the first place, the wage statistics for the different periods have been drawn from different sources, and in spite of all efforts to secure homogeneous series it is quite possible that the classes of workers covered are not quite the same in the different sources. In the second place, pre-war wages were expressed in roubles; it was therefore necessary to convert them first into paper marks for comparison with the figures for the period 1922 to February 1924, and then again into gold francs for comparison with the figures for subsequent months. The index numbers of nominal wages have thus been prepared on a paper-mark basis for the period up to February 1924, and on a zloty basis for the following period. The cost-of-living index numbers based upon prices in paper marks and zloty published by the Central Statistical Office have been applied for the corresponding periods. This gives a continuous series of index numbers of real wages convering both periods. As in the case of Germany and Austria, the real wages shown in the tables are calculated for the time when they were spent and not when they were earned. The method employed is similar to that adopted for the other two countries. Owing to the fact that in Poland both the cost-of-living index number and the wage statistics were repeatedly modified and that the period covered varied from one month to the next, the interpolation formula had to be slightly changed, but the principle remains the same 1 . It should also be noted that the index numbers employed for the calculations are based upon a theoretical budget representing a post-war standard of living. As was pointed« out in the case of Germany and Austria, had this index been prepared on the basis of pre-war conditions — as it should be for the present purpose — 1 Cf. appendix XI for particulars of the sources and methods employed. 8 — 114 — it would indicate a still greater increase, and real wages would consequently be lower than those shown in the tables. REAL WAGES OF CERTAIN CLASSES OF WORKERS, 1921 TO 1924 Table I gives the index numbers of real wages paid to skilled workers in certain industries in Warsaw and Lodz from January 1921 to February 1924. Table II gives similar data for semiskilled and unskilled workers. These statistics, it should be noted, cover only certain groups of workers in these two towns; tbey cannot therefore be regarded as representative of the general position in the whole country, least of all in the former German and Austrian provinces; they have by no means the same general value as the Vienna figures have for Austria. Not only do these figures show substantial differences in the level of real wages between different industries, but in each separate industry they, also fluctuate irregularly. The general development is thus more clearly to be seen from tables III and IV, in which the average yearly indexes have been calculated from the available monthly indexes. So far as can be judged from these statistics, it seems that except for a few industries the highest level of real wages recorded during the period of inflation was reached in 1921, more particularly during the first half of that year. In most industries, however, the increases in skilled workers' wages during the three years following the war was not sufficient to restore them to their pre-war level. Exceptions are to be found in the case of women spinners (whose wages were comparatively low before the war), printers and workers in hides and skins industry. The wages of semiskilled and unskilled workers appear to have about re ached the pre-war level in some cases, and slightly exceeded it in others, in the middle of May 1921. From the summer of 1921 onwards the movements of real wages show fairly wide divergencies. Skilled workers' wages tended generally to decrease up to the beginning of 1924 in the engineering, textile, and hide and skin industries. The textile workers in Lodz seem to have been particularly affected, but at that time this industry was suffering from a severe depression. In other industries the decrease did not become appreciable until the last months of 1923. The only workers who drew higher real wages from 1921 to 1923 than before the war were the printers; but their wages also depreciated greatly i 15 — TABLE I. INDEX NUMBERS OF REAL WAGES OF SKILLED WORKERS IN CERTAIN INDUSTRIES AT WARSAW AND IN THE LODZ TEXTILE INDUSTRY X (Base: January 1914 == 100) Build- Engineering ing Date 1921 January February March April May June July August September October November December 1922 January July August September October November December 1923 January February March April * May June July August September October November December 1924 January February March April May June July Brick- Skilled Weavlayers workers ers 77 76 89 96 90 78 71 72 68 76 95 95 68 74 87 88 84 72 69 56 64 72 74 75 — 75 76 76 74 71 64 61 67 68 87 105 96 82 78 76 69 49 50 44 53 50 58 70 61 51 48 57 77 78 96 94 82 78 79 106 107 107 108 109 105 • — 43 36 — a 81 97 83 83 84 85 82 Printing Textiles 72 82 80 81 77 66 53 56 61 53 54 55 • Brew- Bakereries ies SpinSkilled Com- Skilled Skilled ners work- work(wo- positors workers ers ers men) 102 117 114 117 110 94 76 80 78 67 68 70 82 92 113 140 136 117 95 112 104 117 120 132 66 64 91 100 100 91 78 89 82 92 115 117 65 75 88 95 124 106 90 94 87 98 122 125 102 110 129 132 124 106 119 98 111 96 98 100 102 101 119 114 102 97 • — • 53 64 53 54 40 38 61 72 59 59 44 42 123 122 151 146 130 123 — 117 99 98 107 103 91 87 124 101 101 97 94 84 79 45 59 47 48 46 35 38 49 64 51 53 50 38 41 107 107 138 166 151 129 126 106 86 87 111 134 122 104 99 69 88 135 — — • — • 38 44 70 58 — 64 66 78 — — — Hides and skins 50 72 66 .— — — — — — — 138 188 191 191 89 92 111 111 85 75 57 49 — — — — — — .— 65 57 102 138 120 117 — 84 84 108 130 119 101 96 95 86 — 65 54 108 131 133 130 » Up till September 1923 the indexes are calculated from the average wage rate for the whole month; from September 1923 onwards they are calculated from the rates in force during the second half of the month only, and from March 1924 onwards from the rates in force during the whole month or, for certain classes of workers, on the flrst day of the month following that considered. » Wages current during the period 16-21 January only. — 116 — TABLE II. INDEX NUMBERS OF REAL WAGES OF UNSKILLED WORKERS IN CERTAIN INDUSTRIES AT WARSAW AND IN THE LODZ INDUSTRY1 TEXTILE (Base: January 1914 = 100) Engineering Date 1921 January February March April May June July August September October November December 1922 January July August September October November December 1923 January February March April May June July August September October November December 1924 January February March April May June July Textiles Breweries Hides Bakeries and skins Assistants Labourers Labourers Labourers Assistants Labourers 92 100 117 120 113 97 94 76 87 98 100 102 108 117 137 140 132 113 110 89 101 114 117 119 111 126 123 126 118 101 81 85 82 71 72 74 75 74 106 116 116 106 90 103 95 108 134 137 66 80 95 102 122 105 110 112 111 120 149 152 86 93 110 112 105 90 101 83 94 82 83 84 97 95 93 90 81 78 118 111 107 107 103 93 89 73 63 73 62 61 46 43 135 77 76 150 101 101 97 94 84 79 118 106 105 102 98 88 83 67 62 72 87 76 63 60 70 77 71 82 99 86 72 68 80 51 66 52 54 51 38 43 60 60 77 93 85 72 69 68 94 71 72 92 112 102 87 82 53 54 57 62 — — • — 90 2 93 97 96 — — 97 101 101 100 101 102 99 • — — • — — 53 60 • — 45 74 59 .— — 78 78 79 79 — — — — — — — — — . — — • — — — — — — — — — — — — — — — — — •— — — — . — i Up till September 1923 the indexes are calculated from the average wage rate for the whole month; from September 1923 onwards they are calculated from the rates in force during the second half of the month only, and from March 1924 onwards from the rates in force during the whole month or, for certain classes of workers, on the first day of the month following that considered. 3 Wages current during the period 16-21 January only. — in — TABLE I I I . WORKERS AVERAGE I N D E X NUMBERS OF REAL WAGES OF SKILLED IN CERTAIN INDUSTRIES TEXTILE AT W A R S A W A N D IN T H E LODZ INDUSTRY (Base: January 1914 = 100) Building Engineering Bricklayers Skilled Weavworkers ers Date 1921: annual average 1922: a v e r a g e for.6-7 m o n t h s 1923: annual average 1924:average of first 7 months TABLE IV. WORKERS Breweries Printing Textiles Spinners (women) Compositors Bakeries Hides and skins Skilled Skilled Skilled workwork- workers ers ers 82 74 66 91 113 90 97 110 84 71 50 56 133 100 97 106 73 53 48 53 119 94 83 93 103 85 64 72 177 101 119 126 AVERAGE I N D E X N U M B E R S OF REAL W A G E S OF U N S K I L L E D IN CERTAIN INDUSTRIES TEXTILE AT W A R S A W A N D IN T H E L O D Z INDUSTRY (Base: January 1914 = 100) Engineering Hides Textiles Breweries Bakeries and skins Date 1921: a n n u a l average 1922: average for 6-7 m o n t h s 1923: a n n u a l average 1924: average of the first 7 m o n t h s Assistants Labourers 100 116 89 Labourers Labourers Assistants Labourers 98 105 110 94 104 60 80 101 100 60 75 54 70 81 88 94 100 79 during the last two months of 1923. At the end of 1923, the indexes of real wages paid to skilled workers were o a the whole little more than half the pre-war figures. Semi-skilled and unskilled workers' wages also tended generally to fall between 1921 and 1923, but to a less degree than those of skilled workers. In 1924 the whole position changed; the real wages of every class of worker for which figures are available showed a marked increase during the first months of t h e year, and were subsequently stabilised at various levels. In several industries real wages have — 118 — reached their pre-war level, but this phenomenon is not quite general as yet. E F F E C T OF UNEMPLOYMENT ON THE AGGREGATE INCOME OF THE INDUSTRIAL POPULATION In the sections on Germany and Austria it was possible to supplement the data in respect of real wages by an estimate of the losses caused by unemployment and so to give a general idea of the aggregate income of the industrial population. Such an estimate is impossible for Poland. The few statistics given below are inadequate to measure unemployment even approximately; they give at best a general idea. of the changes in the volume of partial and total unemployment. The figures for total unemployment are taken from employment exchange returns, the others from a regular enquiry conducted by the Central Statistical Office in three or four thousand undertakings in the principal industries. The increase in unemployment at the beginning of 1924 must be ascribed primarily to the reduction of work in the winter, this reduction being considerable in Poland. From April onwards, however, when the seasonal depression should have come to an end, there was a further increase due to the economic depression, which continued throughout the summer. Taking the industrial population as a whole, it is clear t h a t from the spring of 1924 onwards the improvement in conditions due to increased real wages was largely if not wholly neutralised by the losses due to unemployment. TABLE v. TOTAL AND PARTIAL EMPLOYMENT, 1923 AND 1924 Total unemployment Partial unemployment Date 1923 March June September December I 1924 January February March April May June July August Average number of days' work Number of totally unemployed registered per worker per week in 3.000-4,000 at employment industrial exchanges undertakings 114,600 76,400 52,400 61,600 110,500 111,700 112,600 121,000 128,000 136,000 163,000 4.92 4.88 5.15 5.32 5.08 4.92 4.98 5.23 — 119 — POSITION OF PROFESSIONAL WORKERS To judge by the position of state officials, professional workers have been even worse off than manual workers since the war. The following table gives index numbers of the real salaries (including family allowances) of certain classes of officials; they have been calculated in the same manner as those given in tables I and II *. The nominal monthly salaries of the classes here considered in January 1914 were 1,080 marks, 270 marks and 140 marks respectively. TABLE VI. INDEX NUMBERS OF REAL SALARIES OF CERTAIN CLASSES OF STATE OFFICIALS IN WARSAW (Base: January 1924 = 100) Higher-grade officials Middle-grade officials Lower-grade officials 1921 July 14 34 47 1922 July October 20 18 41 43 45 65 1923 January April July November December 22 25 17 11 9 40 59 45 15 13 —. 1924 January February March April 11 27 28 26 16 39 40 38 Date 89 -70 — • 57 54 11 will be seen that during the post-war period real salaries were considerably below their pre-war level in each case, those of the higher-grade officials being extraordinarily low. There was a slight improvement, it is true, during 1922 and 1923, but the last months of the period of inflation caused even greater depreciation than in 1921. The financial stabilisation of 1924 doubtless produced an improvement, but in the lower grades at any rate salaries still failed to reach the real values of the beginning of 1923, which themselves represented a considerable reduction. 1 See appendix XI for particulars of the sources and methods employed. — 120 — Again, it is true that special bonuses, which have had to be disregarded here, were paid in some cases, but their amount was inconsiderable. Another point to be noted is that the tendency towards equalisation, already alluded to in connection with skilled and unskilled workers' wages, was very marked in the salaries of higher and lower officials. In conclusion it should be noted that during the period of financial reform state officials were in a peculiar position, for retrenchment in state expenditure was more than ever imperative at the very time when industrial workers were securing increases in wages. For this reason state officials cannot be taken as representative of professional workers in private employment during that period; but they were free from the unemployment which affected every other class of workers. CHAPTER II Measures Adopted to Check the Fall in Real Wages As in Germany and Austria, the measures taken in Poland to check the decrease in real wages fall into two categories, first, those concerning the cost of living itself and, secondly, the procedure adopted for adjusting wages to changes in the cost of living. COST OF LIVING PROBLEMS Price and Rent Control After the Armistice the Polish Government strove to reduce the cost of living by continuing the policy of price control which had been applied during the war in the territories embodied in the new state. The control of a certain number of staple foodstuffs was abandoned by degrees during 1920 and 1921. On the other hand, rent control, which was far more stringent, Avas continued most effectively up to the middle of 1923. Expenditure on housing was kept well below the price of articles in current consumption, and the increase in the cost of living was thus mitigated, as will be seen from the subsequent table. It has already been shown that similar measures were adopted in Germany and Austria, but that the working classes did not derive much direct advantage therefrom, as the amount previously corresponding to rent was not added to their real wages. They benefited indirectly, however, from the fact that the cost of labour liad risen less than prices in general and that economic activities could be continued on a scale which the economic disorganisation of the country would otherwise have precluded. — 122 — TABLE I. INDEX NUMBERS OF THE COST OF LIVING AND OF RETAIL PRICES OF FOOD, CLOTHING, HEATING AND LIGHTING, AND RENT IN WARSAW, 1921 TO 1924 (Base: 1914 = 1) 1 Food Date Clothing Lighting and heating Total cost of livinc Rent 1921 Marcii September 329 607 241 910 119 254 12 21 180 398 1922 March September 813 1,574 967 1,834 394 849 74 314 526 1,077 1923 March June September December 13,809 11,330 16,366 31,724 64,100 110,022 1,526,274 2,556,305 6,092 10,855 42,680 866,582 7,618 767 12,780 1,262 51,756 9,172 101,629 1,196,567 1924 March 3,761,494 4,778,527 2,707,644 597,957 2,872,968 Measurement of the Cost of Living. Satisfactory statistics of the cost of living are available in Poland only from 1921 onwards. Prior to this the retail prices of certain foodstuffs were recorded by a number of municipalities, but no index number was constructed from these prices in any systematic manner to cover a sufficient number of commodities. This registration of prices was, however, reorganised on government instructions^ «nd indexes of the retail prices of foodstuffs and lighting and heating were calculated on more or less uniform lines in the various towns. The most accurate measure of the cost of living was afforded by the cost-of-living index for Warsaw alone constructed by a committee appointed under the Decree of 27 May 1920. This committee, which was attached to the Central Statistical Office of Warsaw,. comprised representatives of various Ministries and of workers' and employers' organisations. The prices of the various articles were reported to it by the municipal authorities, which were responsible for recording them, though the committee reserved the right to correct these figures if they were deemed too high or too low. It then combined them so as to arrive at an index number based on a theoretical budget representing the average consumption of a working-class family of four persons. The groups of com- — 123 — modities included in the list covered food, clothing, lighting and heating, rent, and sundry expenses (education, amusements, travelling, and laundry). The method of compiling this index afforded a guarantee of impartiality ; it was soon relied on by both employers and workers, and was very widely used for adjusting wages to the cost of living, Some criticisms, however, came from the workers' side; they contended t h a t the index number should be higher, and t h a t the theoretical budget on which it was based should be revised so t h a t it covered all the essential expenditure of a working-class family. In December 1922 a difference of opinion arose in the committee itself; the employers protested against the committee's right to increase the prices reported by the municipal authorities and demanded t h a t the committee should be abolished and its functions entrusted to the Central Statistical Office. Except fer these difficulties, the index was generally accepted as a fairly adequate measure of changes in the cost of living; it was only revised in a few details during the whole period of its publication. However, owing to the extremely rapid rise in prices during June and July 1923, it became necessary to curtail the period elapsing between the collection of price data and the adjustment of wages; the index was therefore prepared for the last week of each month. As this change failed to give satisfaction, the index was calculated every fortnight from the month of August onwards. The possibility of calculating the index weekly was under consideration when prices were stabilised and further modification became superfluous. Subsequently the index was prepared once a month only and based on prices in zloty. In addition to the Warsaw index number others were prepared in a certain number of towns by special local committees, as mentioned previously. The methods employed varied in different localities, so that direct comparison is not possible; in any case these index numbers do not seem to have played as important a part in the adjustment of wages as the Warsaw figure. METHODS OF ADJUSTING W A G E S TO THE COST OF LIVING The Sliding Scale. The essential character of the wage adjustment problem in Poland was unquestionably thrown into relief by the adoption of a slidingscale system. — 124 — In 1919 and 1920 no systematic method appears to have been followed. Wages were generally increased by the ordinary means of individual and collective negotiations; but agreements had frequently to be revised, and this led to a considerable waste of time in discussions and to a large number of disputes. The introduction of automatic adjustment was proposed at an early date with a view to avoiding these difficulties, but the lack of any satisfactory index number greatly hampered the application. of such a system; during wage negotiations frequent differences of opinion arose between employers and workers as to the extent of the increase in the cost of living during a given period. It was to supply this deficiency and to provide a sound basis for automatic adjustment that the Government appointed the joint committee attached to the Central Statistical Office to calculate the changes in the cost of living every month. The regular publication of official index numbers led to the general adoption of the slidingscale system. Even in 1920 a number of collective agreements provided for automatic adjustment; in 1921 and 1922 this method became quite general 1 . The most usual method at that time was to adjust wages once a month according to the change in the cost of living as shown by the index number for the past month. Certain agreements laid down that wages were not to be increased unless the rise in prices exceeded a certain figure — 5 per cent, for instance. It should be noted that with this method wages were adjusted one month in arrear, so that when the rise in prices grew more rapid real wages were materially decreased. To compensate for this in part at least, special cost-of-living bonuses were often paid to cover the difference. In some cases these bonuses were exceptional payments and the amount was fixed by special negotiations. Thus in the autumn of 1923 the state officials demanded a bonus equivalent to two months' salary, and although the Government refused to pay the full amount claimed it none the less granted a special allowance. In other cases these bonuses were fixed more methodically; certain agreements stipulated that wages should be regularly adjusted to the cost-of-living index for the preceding month, but that at the end of each month a bonus should be paid, calculated 1 It is interesting to note that the sliding-scale system was not applied merely to industry and commerce but also to agriculture, though in a somewhat different form. In certain districts the collective agreements for agricultural workers stipulated that the part of wages paid in cash should vary with the average price of certain agricultural produce. — 125 — by applying to wages the percentage increase in the index for that month. In this mariner the losses resulting from the delay inherent in the ordinary sliding-scale system were greatly reduced, but not entirely obviated; as prices were rising almost continuously, the supplementary bonus paid at the end of the month had less purchasing power when spent than it would have had if paid simultaneously with the wage which it was intended to supplement. During most of 1921 and 1922 the monthly increase in prices was comparatively slight and the sliding scale appears to have worked smoothly wherever it was used; the official index number was usually adopted without question as a true indication of the increase in the cost of living 1 . In 1923, however, the rise in prices became so rapid that the methods employed in 1921 and 1922 were no longer adequate to prevent a reduction in real wages. Various measures to improve these were then proposed in working-class quarters; the object of the proposals was either to reduce delay in adjustment, to make the introduction of the sliding scale compulsory in all industries, or to secure the adoption of a gold wage. To ensure closer adjustment it was urged that the index number should refer to the last week in each month instead of to the average of the month; the delay would thus be reduced. As mentioned at the beginning of this chapter, the committee attached to the Central Statistical Office acceded to this request, and in June and July 1923 the index numbers were published for the last week of the month. From August 1923 onwards the index was published every fortnight and adjustment was effected for a similar period in the majority of industries. Mention has already been made of the conditions which induced the committee to consider publication of the index weekly. At this time, i.e. in the autumn of 1923, the extremely rapid rise in prices was continually adding to the difficulties of the position. Faced with a large number of industrial disputes which then broke out, the Government decided to introduce measures making the automatic adjustment of wages compulsory, as this appeared to have proved effective in 'many industries. On 7 December 1923 a Bill for the purpose was introduced in the Diet by the Minister of Labour and Social Welfare. The system was to apply to all establishments covered by the Eight-Hour Day Act of 1 Strikes over wage questions were fairly frequent, it is true ; but the point at issue in these cases was mostly an increase in the basic wage and not strictly speaking its adjustment to the cost of living. — 126 — 18 December 1919 and to any other concerns employing more than four workers. The index number to be used for the adjustment of wages for a given period was that for the preceding period, and the length of such periods was to be fixed by the Council of Ministers. The Bill further laid down that automatic adjustment would cease to be compulsory as soon as the gold value of wages rose above the level of June 1914 for the same class of workers. Sanction was given for the adoption of any system which was more advantageous to the workers than that prescribed by the Act ; but individual and collective agreements with less favourable provisions were to be null and void. The Minister of Labour and Social Welfare was, however, to be empowered in cases of absolute necessity to grant either temporary or permanent exemption to certain undertakings. The Act was passed with a few amendments 1. At the beginning of February 1924, however, the currency was stabilised. The relative positions of the cost of living and the goldfranc exchange were reversed, the former rising above the latter. The Act thus lost much of its utility as, under one of its clauses, it could be suspended on account of stabilisation; moreover, adjustment ceased to be compulsory as soon as the gold value of wages rose above the 1914 level. The changes consequent upon stabilisation influenced the attitude of employers and workers towards the system to be adopted for adjusting wages. Up till 1923 wages fixed at a gold value equal to the 1914 rate would have been higher than 1914 wages simply adjusted to the cost of living. The consequence was that in 1923 the workers' demand that their wages should be fixed at an amount equivalent in gold value to the pre-war wage encountered the strongest opposition on the part of the employers, as has been shown. In January 1924, however, stabilisation was accompanied by a certain lise in the cost of living. If, therefore, from pre-war days up to that moment the workers had received increases proportionate to those in the cost of living their wages would by that time have exceeded the gold value of the pre-war wage. In the circumstances they came more and more to favour adjustment to tbe cost of living, whereas the employers inclined towards the suggestion previously made by the workers that wages should be fixed at the 1 The Act was extended to cover all industrial and commercial undertakings without distinction. It could be temporarily suspended by the Council of Ministers if the Polish exchange in Swiss francs remained stable for at least two weeks. — 127 — 1914 gold value and that the sliding scale should be discarded. They generally refused to grant increases fully equivalent to the lise in the index, pointing out that in Poland prices had risen above world maiket rates, that the economic position of the country was growing extremely difficult, and that the workers must make some sacrifice. The principle of the sliding scale was, however, still included in a large number of agreements, and in several cases the workers demanded the adjustment of their wages to the cost of living, vvhiclt was still rising. But once their real earnings exceeded the pie-war figure they lost the support of the law, and as they feared an increase in unemployment they were usually obliged to accept a compromise. The depression which began in the spring of 1924 and the resulting need for cutting down production costs led by degrees to the final abandonment of the sliding-scale principle. It may, however, be said that during the whole period of inflation this was the fundamental method employed to adjust wages to the cost of living, although the proposals to make it universal and compulsory could not be carried into effect. Cosi-oj-LU'ing Bonuses and Family Allowances Certain problems associated with the question of adjustment ana akin to those already considered in the chapters on Germany and Austria had similarly to be dealt with in Poland. As in the other two countries, a variety of cost-of-living bonus systems were adopted, either as temporary measures, when the increase in the cost oí living was regarded as temporary, or to compensate for the differences between the cost of living in various localities, or again to provide allowances proportionate to the worker's family responsibilities. The last-named of these allowances are of the greatestinterest. During the war family allowances had already been granted in the form of grants in kind distributed to the workers in proportion to their family responsibilities. When these grants were stopped certain of the workers induced their employers to continue the allowances either in cash or in kind. Pro\ision for family allowances is, however, found in a fewcollective agreements only. Of 500 agreements examined by the Central Statistical Office, and referring to various dates, only 7% per cent, provided for such allowances. They are found mainly in Upper Silesia, where the system was introduced under — 128 — German administration. The industries where the system was most prevalent were the mining, metal, and petroleum industries. The Government also paid family allowances to its officials and workers. The Gold Wage The sliding scale was not the only method of preventing the depreciation of real wages which was considered in Poland. In 1923, when prices were rising alarmingly and adjustment was therefore growing very difficult, proposals of a different kind began to engage the attention of certain groups of workers and of the Socialist Party. These proposals were no longer restricted to the adjustment of wages to the cost of living in terms of the ordinary currency, but aimed at fixing them in terms of a stable monetary unit, generally known as the zloty or "Polish florin" (which was taken as equivalent in value to the gold franc), and at paying wages in paper marks at the rate, of exchange of the gold franc quoted on pay-day. This idea was in fact the same as that which was being discussed in Germany at about the same time under the name of the "gold wage", and was the product of similar circumstances. Owing to the extremely rapid rise in prices the Polish mark, like the German mark, had practically ceased to be a standard of value. In most industrial and commercial transactions it had become the practice to fix prices in zloty; even the Government had adopted this method. The demands of the working class that their wages should be fixed in zloty grew particularly insistent after the spring of 1923; in June the Socialist Party introduced a Bill into the Diet for the adoption of this system. This Bill laid down that the wages of all manual and non-manual workers, the salaries of officials, and all other financial liabilities incurred under labour agreements should be fixed in that theoretical monetary unit, which was based on gold, and paid in paper marks at the rate of exchange fixed for gold treasury bonds by the Minister of Finance, that rate being itself based on the gold franc exchange. The level of wages provided for under this Act was not the 1914 one, but was to be fixed by converting into " Polish florins" the paper-mark rates in force during the first week of April 1923. These proposals were supported by various workers' organisations and favourable resolutions were passed by several congresses during the second half of 1923. The employers, on the other hand, — 129 — opposed them and argued that this system would decrease Poland's ability to meet foreign-competition and that it might still further depreciate the mark. The Bill was not passed. It was mentioned earlier that when the Government was compelled to intervene it favoured the extension of the sliding-scale system to all industries. In actual fact the gold wage system did not come into general use until after stabilisation had been effected, i.e. in the beginning of 1924. Although at this period the zloty was not yet in circulation, the - Government's financial policy had had the effect of stabilising the paper mark, which was converted into gold at a fixed rate. The practice of keeping accounts in zloty thus spread rapidly in the labour market as elsewhere, and when the new currency was issued in April 1924 it was universally adopted for fixing wages. There was, however, only one advantage to be gained from this system, namely, the disappearance of the huge figures required for accountancy in marks. Notwithstanding the belief which was held at first, it was very inadequate to ensure stability in real wages. The cost of living continued to rise to an extent which, though not comparable with that of the inflation period, was none the less considerable. Adjustment was thus still necessary, and this explains the workers' reluctance to discard the sliding scale. In Poland as in Germany the gold wage system was but a temporary measure and was only applied for a very short time, ae in July 1924 the paper mark was practically withdrawn from circulation, having been replaced by the zloty. 9 CONCLUSION There is a certain similarity in the general economic conditions of the three countries studied here and the problems to which these conditions gave rise during the years following the war. There was first a period of excessive monetary inflation and then one of stabilisation. Inflation reached the most incredible pitch in Germany ; in Austria it was less than, but still very greatly in excess of the average level of inflation in other countries. In Austria, too, the stabilisation period began earliest — at the end of 1922, thanks to the intervention of the League of Nations. In Germany the financial reforms which led to the stabilisation of the mark date only from November 1923, and in Poland from February 1924. In Germany and Poland a beginning was made with economic reconstruction between the end of the war and 1921, in which year perceptible progress was made. But monetary inflation soon put a stop to this; 1922, and even more 1923, saw a marked retrogression. In Austria, on the other hand, the disorganisation left by the war was too great and inflation began too soon for any economic reorganisation to be attempted. Conditions became steadily worse until the League of Nations intervened. In all three countries the economic situation brought about by excessive inflation was becoming critical when steps were taken to restore the finances of the country. The sudden stabilisation of prices resulting from this precipitated the crisis, which appears to have been most serious in Germany. These difficult economic conditions naturally had a deleterious effect on the position of the workers. In Germany and Poland the income of the working classes, measured by their real wages, was higher in 1921 than in the preceding years, and even in some cases approached the pre-war level. From 1921 it steadily decreased until about the end of 1923. The stabilisation of the currency was accompanied by a rise in real wages, but also by an outbreak of unemployment. Taking the industrial population as a whole, the individual benefit enjoyed by workers still in employment through the increase in wage rates was counterbalanced by the losses suffered by the — 132 — imemployed. In Austria the workers appear to have been in a worse position in 1921 than before the war. In 1922, although the methods of adjusting wages had become sufficiently effective to increase the real value of wages in spite of the great rise in prices, the total working-class income decreased owing to the growth of unemployment. In 1923 and 1924 unemployment assumed such proportions that, although workers still in employment received relatively high wages, the total income of the working classes was definitely less than before the war. It may be noted that on the whole during the period under consideration skilled workers suffered more in proportion than unskilled; the difference between their respective wages was very greatly decreased. It began to increase again to some extent, however, when it was realised that this levelling up of wages formed a check to the recruiting of skilled labour. On the whole the decrease in the total income of the working classes in the three countries considered may be contributed to three factors. (1) Economic disorganisation and the extensive changes of frontiers due to the war brought upon a decrease in the national income for some years at least, quite apart from the distribution of this income. (2) Monetary inflation caused a steady rise in prices, to which wages could not adjusted precisely or rapidly. It thus led to a continual fall in the real income of the workers. (3) Monetary stabilisation increased the unemployment which had already appeared and deprived a large proportion of the working population of employment and therefore of income. The first two of these three factors exerted their influence concurrently for a long time, and it is not always possible to distinguish their respective effects. Nevertheless the influence of inflation on wages is the most apparent and doubtless the most characteristic of the period. Inflation, together with the defensive measures to which it gave rise, is thus the basis on which the present report has been built up. The first defensive measure adopted in the three countries was the indirect one of price control. The adjustment of wages was facilitated if the rise in the cost of living was checked or retarded. Control was applied chiefly to food prices and, even more effectively, to rent. The effect was to reduce the cost of labour, since wages were adjusted to a reduced cost of living, and to obviate unemployment to some extent, but real wages were little if at all increased by these means. — 133 — The direct methods of adjusting wages in the three countries are similar, although the procedure adopted varies slightly. In Germany wages were chiefly adjusted by the ordinary method of negotiations between employers and workers. The period of validity of agreements had, however, to be curtailed progressively, and the wage clauses had to be made more or less independent of the rest of the agreement. As the rise in cost of living grew more and more rapid in 1923, negotiations had to be renewed every two or three weeks and procedure naturally had to be simplified. Purely automatic adjustment of wages, though discussed at length, was never generally adopted in Germany. In Austria, on the other hand, it was at first extensively used, but owing to vigorous objections from employers was gradually abandoned, while stabilisation made it much less necessary. In Poland, also, the system was in great favour at one time and it was even proposed to make it compulsory. The problem of adjustment of wages to the cost of living raised further allied problems. Allowance had to be made for differences in the cost of living between different localities and cost-of-living bonuses graduated accordingly. This procedure is to be found in all three countries, but most of all in Germany, where it was particularly necessary as many collective agreements cover very large areas. The decrease in wages bore most heavily on workers with families, and attempts were made to meet this difficulty by paying allowances graded according to the number of the worker's dependants. These allowances were never of any great amount in any of the three countries, and seem in no case to have compensated fully for the greater expenses of workers with families. The system was fairly general in Germany, where many collective agreements provided for family allowances, but these are generally very small. Many Austrian collective agreements also include such provisions. In this country the most interesting measure in this connection is the establishment by law of a system of family allowances (or more precisely children's allowances) on a national basis. These allowances, however, were not regularly adjusted to the cost of living, and rapidly lost most of their value. In Poland the system }vas less frequently adopted, except in a few industries and for public officials, who, as in other countries, were among the first to receive family allowances l. 1 The International Labour Office has published a detailed report on this question: Family Allowances; The Remuneration of Labour according to Need; Studies and Reports, Series D, No. 13; Geneva, 1924. — 134 — As the currency was stabilised, what was known as the gold wage appeared in Germany and Poland. Even before stabilisation was effected the system had been discussed and even put into practice in certain limited spheres. At this period the gold wage was taken to represent a system of fixing wages in a stable-value foreign currency (e.g. the dollar or the Swiss franc) and paying them in the depreciated national currency at the current rate of exchange. In effect this was only an application to the labour market of a procedure already adopted on other markets when the national currency ceased to be a real standard of value owing to constant depreciation. Later, as the result of financial reform and ef the introduction of the rentenmark in Germany and the zloty in Poland, this system became general, but in a slightly different form. Wages were fixed in the new national currency and paid partly in this currency, partly in the old, the latter being calculated at a rate fixed by the government. This was a purely transitional method which has already disappeared in Poland, where since the middle of 1923 the zloty has entirely replaced paper money. In conclusion, it should be pointed out that the subjects dealt with in this report are already problems of the past. At the present moment the three countries considered seem definitely to have embarked on the policy of stabilisation and new problems are arising. The chief problem now is not to prevent a fall in real wages through an incessant rise in prices, but to find means of increasing the national income as a whole and of securing a fair distribution of it. These problems, it is true, have existed all the time, but during the period of inflation they were overshadowed and remained in the background. They are now re-appearing with all their old importance, but it is too soon as yet to forecast the solutions which may be adopted. APPENDICES APPENDIX I Monetary Inflation and Price Changes on the Various Markets in Germany In chapter I of this report a general outline was given of the course of monetary7 inflation in Germany and the price fluctuations which it caused in the various markets. This appendix gives a number of tables specially prepared to display the differences in the purchasing power of the mark in these markets, thus supplementing chapter I. Table I shows the amount of the Reichsbank note issue from 1920 to 1924, table II the amount of various currencies in circulation at the end of 1923 and in 1924, i.e. when stabilisation occurred. They clearly indicate the gradual development towards a stable and orderly currency system. Table III gives the index numbers of the dollar rate in Berlin and the index numbers of wholesale prices of home and imported goods. The index of the dollar rate indicates the purchasing power of the mark over gold on the foreign money market, while the index of wholesale prices for home products show the purchasing power of the mark over certain raw and semi-finished materials on the home market. The index of wholesale prices for imported goods indicates changes resulting from a combination of two phenomena, namely,-the depreciation of the purchasing power of the mark in terms of foreign currency and the depreciation of the purchasing power of such foreign currencies in terms of the goods considered. This index is of interest here only as one of the factors determining the general index of wholesale prices, itself expressing the depreciation in the purchasing power of the mark on the wholesale market in general, which depreciation in turn affected the retail market. In table IV, therefore, the general index of wholesale prices is compared with retail prices. As rents were strictly controlled and lagged far behind the general increase, the table gives first the index of retail prices for food, clothing, lighting and heating, and, second, the index of the total cost of living, i.e. including rent. Finally, by comparison of the weekly indexes of wholesale prices and the cost of living given in table V, the increasing speed with which retail prices were adjusted to wholesale prices can be followed. In actual fact the improvement was even more rapid than the table would appear to show, as the index here used, in default of other data, is that of the cost of living, the rise in which was checked by the influence of rent. — 138 — TABLE I. REICHSBANK Date (end of the month) 1913 1914 1915 1916 1917 1918 • 1919 1920 December December December December December December December December 1921 March June September December : 1922 March June SOURCE. NOTES Total circulation in million marks 2,593 5,046 6,918 8,055 11,468 22,188 35,698 68,805 69,417 • 75,321 86,384 113,639 130,671 169,212 IN CIRCULATION, Date (end of the month) 1 9 1 3 TO 1 9 2 3 Total circulation in million marks 1922 September December 316,870 1,280,095 1923 January February March April May June July August September October November December 1,984.496 3,512,788 5,517,920 6,545,984 8,563,749 17,291,061 43,594,738 663,200,000 28,228,815,000 2,496,822,909,000 400,267,640,302,000 496,507,424.772.000 L E A G U E O F N A T I O N S : Monthly Bulletin of Statistics. — 139 — TABLE I I . LEGAL CURRENCY IN CIRCULATION IN GERMANY, END OF 1 9 2 3 AND B E G I N N I N G OF 1 9 2 4 (In million gold marks) 1923 Currency Paper Currencies * Reichsbank notes Private banknotes R a i l w a y currencies Yai'ious currencies sanctioned b y t h e Government 1924 31 Oct. 30 Nov. 144.6 400.3 496.5 483.6 587.9 689.9 1,097.3 1,520.5 0.4 0.1 0.1 0.1 0.0 0.0 0.0 0.0 25 7 114.8 109.0 64.6 51.1 26.6 5.5 3.4 2.3 1.9 1.4 ;i.o 176.2 518.5 607.9 550.2 640.4 717.5 1,097.3 1,520.5 — — — 1.0 5.8 25.9 194.7 326.8 1,771.6 1,853.1 31 Dec. 31 Jan. 29 Feb. 31 Mar. 30 June 30 Sept, I. Total I I . Small Change I I I . Stable Currencies. Rentenmarks Gold loan certificates (small ; denûmination) R a i l w a y currencies Currencies secured b y gold loans 78.0 50) .3 1,049.1 1,196.3 1,563.8 1,760.3 216.4 240.0 240.0 207.9 146.6 35.4 141.9 144.2 144.2 123.6 55.5 46.1 216.2 234.7 146.0 71.4 50.4 9.5 Total 124.1 969.3 1,665.7 1,726.5 1,987.3 2,080.9 1,836.6 1,860.3 T o t a l legal currency in circulation 300.3 1,487.8 2,273.6 2,277.7 2,633.5 2,824.3. 3,128.6 3,707.6 [ SOURCE: Wirtschaft und Statistik, 1923 and 1924. i Converted into gold marks at the dollar rate of exchange on the date of issue. 7.2 — 140 — TABLE III. INDEX NUMBERS OF THE DOLLAR RATE AND OF WHOLESALE PRICES OF HOME AND IMPORTED PRODUCTS, 1920 TO 1924 (Base: 1913 = 1) Dollar rate in Berlin Month Wholesale prices of home products Wholesale prices of imported products Based on p a p e r - m a r k prices 1920 January July 15.4 9.4 9.6 12.6 27.3 19.0 1921 January July 15.5 18.3 13.6 13.7 18.2 17.2 1922 January July 45.7 117.5 33.8 93.0 50.8 138.5 4,281 84,186 1,100,632 23,549,000 6,017,200,000 522,285,700,000 1,000,000,000,000 2,390 69,695 867,771 22,250,000 6,662,300,000 706,900,000,000 1,193,000,000,000 4,758 100,244 1.325,388 32,410,000 9.257,300,000 819,600,000,000 1,603,000,000,000 1923 January July August September October November December " 1924 January February March April May June July August September j Based on re u t e n m a r k prices ' 1 1 1 1 1 1 1 1 1 1.083 1.041 1.082 1.117 1.122 1.061 1.064 1.130 1.208 1.625 1.763 1.828 1.856 1.738 1.650 1.583 1.574 1.577 i ' ì S O U R C E : Wirtschaft und Statistik, 1921 to 1924. To convert index numbers based on rentenmark prices to a paper-mark basis, (h-decimal point should be moved twelve, places to the riifht, as oae rentenmark = 1,000,000,000,000 paper marks. 1 — 141 — TABLE IV. INDEX NUMBERS OF WHOLESALE PRICES, RETAIL PRICES., AND COST OF LIVING, 1920 TO 1924 (Base: 1913-1914 = .1) ! Month Wholesale prices Retail prices of food, clothing, lighting and heating (cost of living excluding rent) Cost of living (including rent) Based on p a p e r - m a r k prices 1920 January July 12.6 13.7 — — 8.47' 10.65 1921 January July 14.4 14.3 15.7 — 11.79 12.5 1922 January July 36.7 100.6 25.4 69.6 20.4 53.9 2,785 74,787 944,041 23,948,000 7,095,000,000 725,700,000,000 1,262,000,000,000 - 1,120 37,651 586,045 15,000,000 3,657,000,000 657,000,000,000 1,247,000,000,000 i 1923 January July August September October November December 1,453 48,986 764,389 19,500,000 4,763,000,000 852,000,000,000 1,562,000,000,000 Based on r e n t e n m a r k prices -' 1924 January February March April May June July August September 1.173 1.162 1.207 1.241 1.225 1.159 1.150 1.204 1.269 . 1.350 1.260 1.280 1.310 1.330 1.290 1.310 1.270 1.300 1.100 1.040 1.070 1.120 1.150 1.120 1.160 1.140 1.160 S O U R C E : Wirtschaft und Statistik, 1921 to 1924. This figure refers to February 1920, no flgure being available for January 1920. To convert index numbers based on rentenmarfc prices to a paper-marfc basis, th8 (tecimal point should be moved twelve places to the right, as one rentenmarfc =» Í,000,000,000 paper maris. 1 2 — 142 — TABLE V. WEEKLY INDEX N U M B E R S OF WHOLESALE P R I C E S AND COST OF LIVING, APRIL 1923 TO JANUARY 1924 (Base: 1913-1914 — 1) Date 1923 5 April 14 & 25 » 5 15 25 May » » 5 15 25 June » » 3 July 10 17 » 24 » 31 » 7 August » 14 » 21 » 28 4 September 11 18 » 25 » 2 October 9 » 16 » 29 » 30 » 6 November 13 » 20 » ¡> 27 4 December 11 » 18 » 25 » 1924 2 January 8 15 22 29 » » » » S O U R C E : Wirtschaft Wholesale prices Cost of living 4,844 4,923 5,738 2.793 2,846 2,886 3,061 3,254 6,239 3,463 7,105 3,737 9,034 4,168 4,843 12,393 6,002 17,496 6.950 24,618 9.272 11,785 33,328 16,180 48,644 21,511 57,478 28,892 79,468 1 39.336 183,510 [1 71.476 483,461 149,531 436,935 663,880 753,733 1,246,598| 1,695,109 i ,183,434 2,981,532 | 1,845,261 5,051,046 11,513,231 36,009,183 1 14,244,900 36,200.000 28,000,000 84,700.000 | 40,400,000 307,400,000 | 109,100.000 1,093,000,000 691,900,000 14,600,000,000 3,045,000,000 18,700.000,000 13,671,000,000 129,000,000,000 98,500,000,000 265,600,000,000 218,500,000,000 1,413,400,000,000 831,000.000,000 1,422,900,000,000 1,535,000,000,000 1,337,400,000,000 Ll,515,000,000,000 1,274,500,000,000 1,269,000,000,000 1,244,600,000,000 1,163,000,000,000 1,200,000,000,000 1,150,000.000,000 1.224,000,000,000 1,147,000,000,000 1,197,000,000,000 1,198.000,000,000 1,157.000,000,000 1,148,000,000,000 1,130,000.000,000 1,110,000,000,000 1,080,000,000,000 1,060,000,000,000 und Statistik, 1923 to 1924. Date 1923 4 April 11 » 18 25 2 May 9 » 16 23 30 6 June 13 20 >> 27 4 July 11 » » 16 23 30 6 August 13 » 20 » 27 » 3 September 10 » 17 ¡> 24 1 October 8 » 15 » •» 22 29 » 5 November 12 » 19 » 26 » 3 December 10 » 17 » 22 » 29 » 1924 7 Janua ry 14 » 21 » 28 » APPENDIX II Method of Constructing Index Numbers of Real Wages of Industrial Workers and Real Salaries of Professional Workers in Germany, 1920 to 1924. Nominal Wages and Salaries in Marks of these Workers. INDUSTRIAL W O R K E R S ' W A G E S The data for the period 1920-1922 (Part I, chapter I, tables I and II) are drawn from, various sources. Only a few or the series can really be compared with those for the period 1922-1924 (Part I, chapter I, tables III and IV) ; these are the figures for mining, building, state undertakings, and, to a lesser degree, the metal industry. Nominal wages in marks, from which the index numbers of real wages have been calculated, are shown in tables I to IV of this appendix, which correspond to tables I to IV of chapter I, Part I. The methods of calculation are described below. Data for the Period 1920-1922 The wages of skilled and unskilled workers in the mining industry were arrived at by calculating first the weighted average (weighted in accordance with the number of workers employed) of the average earnings per shift in five districts: Upper Silesia, Lower Silesia, Dortmund, Aix-la-Chapelle, and Saxony, in order to obtain the average earnings per shift in all mining districts (Source: Statistisches Jaìirbuch für das deutsche Reich, 1921-1922). These figures, like the original district figures, are quarterly averages; to convert them into index numbers of real wages, a quarterly cost-of-living index was obtained by calculating the arithmetic mean of the three official monthly indexes. From the method of calculation, these wage data correspond roughly to the wage index numbers for the years 1922-1924 given in tables III and IV. Certain changes have, however, been made in the delimitation of the districts, in the case of Upper Silesia especially, which slightly impair the homogeneity of this series. In the building industry the weekly wage rates in twenty important towns were used (Source: Statistisches Jahrbuch für das deutsche Reich, 1921-1922). This series corresponds exactly to that given in tables III and IV. In addition to this, for purposes of further comparison figures calculated from data in the Wage Census of February 1920 are given (Lohn- und Gehältererhebung vom Februar 1920; Statistik des deutschen Reichs, Vol. 293). These figures, like those for 1920 in respect of the wood, metal, and textile industries, which are drawn from the same source, were calculated as follows: The figures given by the 1920 Wage Census represent average hourly wages. These figures were multiplied by the normal working hours provided for in collective agreements to give weekly wages. — 144 — For an index number of real wages data for pre-war years are also required, and those published in the Census for purposes of comparison were selected. These generally represent hourly rates as laid down in collective agreements in force in the towns for which wages were recorded in January 1920. These hourly rates were then converted into weekly rates, being multiplied by the normal hours of work prescribed in collective agreements at that date 1. The base thus obtained does not entirely correspond to that employed in tables III and IV for the indexes of real wages during the period 1922-1924; but it is more closely in relation with the data of the February 1920 Census. The figures for the wood industry were also obtained from the Wage Census and treated in the same manner as those for the building industry. The Census further gives the average earnings of workers during four weeks as compared with nominal wage rates for a similar period before the war. The index numbers of nominal wages thus obtained correspond very closely to those obtained by multiplying average hourly wages by the normal hours of work. For the metal industry two series are given, one based on figures obtained from two different sources but referring to a number of towns, the other drawn from a single source but limited to Berlin. The composite series covers, (a) for February 1920, figures taken from the Wage Census, and referring to some twenty towns; from these an index number of real wages has been calculated by the method employed for the wood industry, taking the pre-war data for the twenty towns as base; (b) for October 1921 and January and April 1922, figures of weekly wage rates in eleven important towns (Source: Statistisches Jahrbuch für das deutsche Reich), from which the indexes have been calculated, taking pre-war data for the same towns as base. The Berlin series is based upon the weekly wage rates specified in collective agreements for that town (Source: Wirtschaft und Statistik, 1921 to 1924). The cost-of-living index number used for the conversion into real wages was the official national index. It will be noted that the two series are more or less parallel, although the Berlin series is usually somewhat below the other. Neither of the two series is absolutely comparable with those given in tables III and IV, but that of the average for several towns seems more suitable for comparison, although it is somewhat heterogeneous. The figures for the textile industry were again drawn from the February 1920 Wage Census and treated in the same way as those for other industries similarly obtained. As in the case of the wood industry, it may be observed that these indexes correspond closely to those obtained from figures of average earnings during a period of four weeks. It should be noted, however, that the figures for male wage earners represent the average earnings of weavers and spinners, while those for women refer to the earnings of weavers only; in tables III and IV they also include the earnings of spinners. For the printing industry the indexes in table I constitute a continuous 1 The normal weekly hours of work specified in collective agreements and used for these calculations was as follows (Wirtschaft und Statistik, No. 6, 1924): Pre-war Post-war Building 53 47 Wood 52 46 Metal 59 48 Textile 57 46 — 145 — series, but it is limited to the city of Berlin, and is therefore not directly comparable with the series given in tables III and IV (Source: Statistisches Jahrbuch für das deutsche Reich, 1921 to 1922). The figures for state undertakings refer to the weekly wages (including family allowances) of workers employed in towns graded in class A, i.e., with the highest cost of living (Source: information supplied by the German Ministry of Labour). They can be compared directly with those given in tables III and IV for the"period 1922 to 1924. Data for the Period 1922-1924 (Tables 111 and IV) All these figures have been obtained from official sources (Statistisches Jahrbuch für das deutsche Reich, 1923; Wirtschaft und Statistik, 1924; information supplied by the German Ministry of Labour to the International Labour Office). The number of towns to which they refer varies slightly in different industries, but in most cases some twenty important centres are included. These figures are based upon the weekly wage rates fixed in collective agreements, except those for the mining industry, which are reckoned from the average earnings per shift. The full rate for adult workers was used in all cases. When several agreements were in force during one and the same month, they were each allowed for in proportion to the period of operation. The rates thus determined for each town were then averaged, the average being weighted according to the number of workers of the classes concerned in each town. Method of Calculating the Real Value of Wages when Spent In chapter I it was seen that before 1922 no distinction was made between the period in which wages were earned and that in which they were spent. Not until 1922 was the available information sufficiently accurate and the rise in prices sufficiently rapid to necessitate such a distinction. The preparation of a cost-of-living index number for monthly periods one week later than the calendar month employed for wage statistics was an easy matter for the Federal Statistical Office; it had weekly price returns at its disposal, and had simply to average the returns for the weeks composing this later monthly period. From September 1923 onwards the cost-of-living index was even calculated daily by interpolation between the weekly indexes, and an index number for any monthly period could therefore be obtained by simple averaging. SALARIES OF PROFESSIONAL WORKERS Two tables of professional workers' salaries are given. Table V shows the salaries paid to roughly corresponding classes of skilled and unskilled bank employees, and unskilled, skilled, and higher-grade officials in the Federal civil service (Classes III, VIII, and XI). Table VI shows the salaries paid to certain classes of commercial employees in various industries and administrations. From these figures the index numbers of real salaries of professional workers, shown in Tables X and XII of chapter I, Part I, were calculated. The method of calculation is similar to that employed for the index numbers of real wages paid to industrial workers. 10 — 146 — TABLE I. NOMINAL WEEKLY WAGES (PER SHIFT IN MINING), IN MARKS, OF SKILLED WORKERS IN CERTAIN IMPORTANT INDUSTRIES, 1920-1922 Metals Date Miningi marks 1913-1914 6.27 1920 February 37.63 April May June July November December 1921 January February May July August September October November December 1922 January February April 1 2 3 Buildings Woods marks marks I 37.69 Í 37.52 1 180 1 175 245 3 Textiles s State Printing under6 (Berlin)* takings Women Average of several towns Berlin* Men marks marks marks marks marks marks 42.44 23.94 15.96 34.38 34.56 174 147 120 161 3 31.43 I 35.91 Í35.152 183 169 3 254 52.76 276 275 57.77 65.96 291 68.22 67.31 211 236 251 271 286 342 69.99 388 365 102.50 2 521 571 2 851 2 131.63 953 407 563 296 321 385 410 535 635 613 353 6 172 230 230 264 264 264 264 293 293 293 293 341 341 485 485 485 521 521 754 Quarterly average oí earnings per shift in the principal mining districts. Weekly wage rates fixed by collective agreement in a number of towns. Average hourly earnings as determined by census in some twenty towns, multiplied by the normal hours of work prescribed in collective agreements. 4 Weekly wage rates fixed by collective agreements in Berlin. 5 Weekly wage rates in towns graded in the highest cost-of-living class (class A). 6 Figure for January 1920. — 147 — TABLE II. NOMINAL WEEKLY WAGES (PER SHIFT IN MINING), IN MARKS, OF UNSKILLED WORKERS IN CERTAIN IMPORTANT INDUSTRIES, 1920 TO 1922 Metals Date Mining ! Buildings Wood » marks 1913-1914 4.14 Average of several Berlin * towns marks j 29.48 Í29.263 marks 156 150 — — — — — — — — — — — — — — — — — — — — — — 321 — — 367 523 29.68 marks I 22.23s /22.29 s State under- 6 takings marks marks 30.62 23.70 136 158 222 216 216 240 240 240 240 1920 February 25.07 April May June July November December — I 174 I 156 8 238 37.67 •— 1921 January February May July August September October November December 1922 January February April .— 267 40.17 46.67 — — 48.63 52.14 — —. — 329 — .—. 375 56.55 83.59 — — 109.19 — — 498 — — — 914 • — • — — s 334 2 529 a — 794* — 245 — 259 — — — — — — 595 269 269 269 269 317 317 446 446 446 482 482 708 i2 Quarterly average of earnings per shift in the principal mining districts. Weekly wage rates fixed by collective agreement in a number of towns. » Mean hourly earnings as determined by census in some twenty towns, multiplied by the normal hours of work prescribed in collective agreements. * Weekly wage rates fixed by collective agreements in Berlin. s8 Weekly wage rates in towns graded in the highest cost-of-living class (class A). Figure for January 1920. — 148 — es o »o co INI L O T H CS O0 L O corico co v i d LO Ti I O ^i co - Ì ^ OO CO O O »í r H ( N ( M co <¿* M cs * i o © en co c i co i o ^H CO 0 0 OO » O CO I O CO CS OO O l «H - - Ì © oo co i n co o es m co CO CO CO O l I O CO ^ O i ^ l O C O CO O T-t CO d IO r% r» i o co co CM CO 0 0 GC r^ - H CO CO ODQOIN I O O CO co • H t N O C i T-t C-O W O 0 0 T H CO T Í CM cscscscocococococo OsiocOcOCOCOOOOO cn oo r i CO oooor>.r^oooo T Í Ci O CO O Cl O CS O •¡HrsiOO CS CO CO L O T Í Ci v i T Í CO LO O O CO CO CS O CS T Í IO © CO v i CO c o CS 0 0 CSviOOOOviOOCOOOOO CiCOCOOCivivivivi oo T H O -H 0 0 T-H Ci O o r*. TH C i LO TH CO I O I O CO ^ CO L O I O 0 O C M > C i 00 Pc o CO CO I O L O v i I>.00CÌrH(r0LOLOlOLO T Í CO c o o . oo LO co r^ T Í rHCSCO-viLOCSCOCOCO cor-tT-iior--LOcor->.oo CO o L O O CO < t - H r H ( Û T H l > LO CO I O L O C-» C i CO T H O CO t> O CS 00 Ci O CS C i T Í I O o TH o o o o es v i co r^ oo O l co vi O vi Vi CS L O GS CD OO OO OO M ^ i o ViO r > CO O T H < f OOOO O CS CS ( M CO v i CO O vi r^. O O CS Ci CO O - Ì CS C-N 0 0 O CN O r M > 0 T-t TTÌ < T CO cs r^ -vi CO vi OO CS CO O ^ W r H I O nrM>o T Í es cs ai ci < «H d TÍ co -.+ CO T - I CS OO CO 0 0 0 0 »4* Q> CO LO L O O O0 C i O CS L O CS TH T Í CS CS CS co TÍ CS O T Í CO v i CN LO O l > CO ci v i © v i T Í CO PN ^ l > c D O - H C O O CO v i cO O CO C i O O O r- LO co v i T i LO CO cs cs cs CO v i CO- © v i CS c O O CO ( M 0 0 O CO ( M CS TH TÍ IO (£> IT^ CO 0 0 O CO CO v i rO C O C O ^ c i v i i n LO c= r^ CN -vi v i oo £ fi o LO CS d v i CO cs T Ì co r* CO O C i r^ in r^ a i LO v i T Í CS v i v i OO CO 0> r» fi GJ d cs " CO LO O C CO CO CS C i -vf < T i T-< o co co T-H LO cO O O CS CO CO O I O CO O CO cs ai co o cs - i cs LO r-> oo c i cn co co cs cs cs cs cs cs cs Ci cO O O Ci OO Ci O CS O v i O r» ai o cscscocococococoeo CS T-< LO v i CO c i O T-H cocor^cìviincor^oo cscscscscococococo cs es o co v i LO T-t C i O >i-~" Î--Q 5 ^ to a. tobe veni ^^^ ^ co « cj ¿2 9-¿« o o 0£ CO CO CO v i v i v i CS CO co co co co O C O C N O C O [ N < f COCO C i u 3 vi cs cs cs co ^ CS v i C i LO LO LO r v co O vi r~v c s v i CS < i LO C O C O C S C O T Ì C S O T Ì T Ì T H CS CO C i C i T H TÍ K t - 22 Aprii July Octob H cs ci •> F C i CS O C i C i O v i I O T Ì OtOCOOOCiCSCOCOO O TH v i uO c o r^ r-* r^ r-« r~» ^ m T Í CO T Í C i L O v i v i O T i C i C i O LO v i T i v i t O c o c o c o C i v i LO r-* co cn C i C i C i C i TH T Í T i TH TH TH TH T Í T Í SS -H ^ en uo to o uo es - H co r - es io en oo en o ( M UO O ¡M .S (M 00 en ^- o C S CO CO r-& C-J r- io to to ^ > 0 ( N o o to OO r- to en o r- CS O CO C S *-i i CS ^ H CO r^ TH i - - r - ' CO <» tO ir- C fS « t/1 to GO l O «H OO v ^ co o "-Ï 1 li TH w o H O) GO CO T-l GS ce o LO «O I O H C O ^ en ^ en •H *# CS -G Ü IO IO ** v*# L o m i > *' tO O ^ Î C O T H co C l t O CO r M > i N t O T H CS *?H ^ TH C S CO co to to r LO •H o co v # I O I O o to en o es U 0 O t O t O OO cC So m O í N ^ »o r- r - en ^ tO C S UO tO O o *# o co io to oO TH I O r ^ v í CO T * OO TS O O 00 ir- £ CS CS te oo • * # s CT) CS °S CJ ÏS p N W O OO 0 0 t O r* es <* TÍ »# co en v*« co co oo CTilOrH -H IO 00 O O O r - TH i o GO -vf CO ••S* TH M 15 .S o a cscoco*#tOr-»r^r-Nr» csesescscscscscses •# (M co en o en co [ - N r - v# uO -H co CS *# <* IO to o en o 00 l > *H tO O r- -& O o to ¿4 CS CS iO 00 O ctf S T5 n rrt CS ^ CS CS ir* en o en o o TH TH es co c IO r N Q O O r í Q O •>H OO KO O V * en io~ coT ^ o en io co io io TH ^ O O r . co r- res C O co r-» *TH C S es CS I O r- es uo to en r- oo to oo o en r>> rO CS co VA' o o C S C ^ OO o o r- v* co r o o I| 00 O to 00 es r-HH T H C S o co o rH en o es en o o TH *0 CS CJ >, >-. rrt s :=3 sg "S« C E >i -#©I>.CStOtOtOtOtO ooentOT-i-HT-iTr-iTH cocs'vooecnericnencn cscsescsescscscscs virxenvíiencocscoto cocOv^v#totorv,rxOO víi o i o es i o es 00 TH co CS o TH CO t O C S C S • H Q C K Í CS t o TH cn oo co oo 0 0 o co ( M CS < p V f I O UO » O CS M CS CS C^ CS CS CS IO CS o co ^ to co O t D T H C S CS 0 0 CS t-v Ï> en TH co «^ C S o T-t co oo co en en cn CS C S CS C S CS C S CS CS CS cococoiocnooTHCS cscscscsescocococo to co 00 v# O IO r H C N O es en v# o C~v CO es o TH _ -2 £;.£>.£ •rHv^vyiorvOOOO cscscscsescocococo co CS CS 0 0 0 0 CS 0 0 «TH e n r H C S t O C S -«ri to io r» *-i -^ to to oo ^coo v # CO v # O M O v í O •+-J n O O O C S V ^ - T H ^ G S C O C O COTHtCtOOví^víi io co rH co v# .coese (N r i CO 00 IO IO IO IO Gocnococntotototo THT-i -£ s - ^ 5 3 == o '-s'-s<;cQ Oetober November d v-H (M i N l O Q O ^ i o GO CO < * 8 18,9 393,0 •Í «ÏH CO 24.27 C! Í10US lä §§=11 « > "^3 211 14,231 149 11-5feS .¡H IO Ci o f co" <*"*-" ( M * * I O CD CM CM 00 Ul o I "^ I- -»ri CM O ^ § •^ =•= V^ > o r-* r» r-» CM oo oo v-< O~O~VO~CTÍ co 00 Ci CM CM W Ü O O O -S" C O C O CMCMfMCOCOCOCOCOCOCOCOCO o o o o o o o o o o o o oooo*# rN r> M N (N r i CO C M O O r * C 0 O C 1 0 0 C O I N I N O - Í O O O O O O r n t > c O * * t O O O O O v * ^ ^ - < c O r ^ O O O r-i C Í D Í O O r H ^ Í N O O CO I O < f CS CO O O X N O W O Q r l C O t N I N N - r i ' Ì ' O O O U O a O I M M C i O O COCOCOCOCOCOCOCO^CT)^r!COCMl>.CM TH 1,248 kronen O IOCDCDCOCDCOOOOOO COCOCOCOCOCOCOCO-^ kOOOClCN CNCNO0W CM O O C i CO O CO (M CO CM OO CO CO ( M ( M v - < í * v # v # v # O »o o s O oo o o *# O Ê O £M O * c T O * v o " ^ CM Í M ri Ì ^ T H CM O Í O CM CO ! -H" ^-T (Nr\CON co CM o1 r* O O r- iO 1,462 kronen «s ~ rt o " o " o~ o " oT oí" o C O C O C O C O C O C O C O - ^ - * CM CD ^ o c o O v # ^ ^ < M C O O O O 5 o í« £ S CO"CC"T-* r-*^w«ïH—* o '3i o o C O O O O O O O L O L O O O O 00 öl • « rrtr-iOr^OOOOOOlOVOVOCíCTi O cD 0 0 O v - í C7i CM ( M < * < ^ I O C7ì r i C N t û C ! CM -O« IO CM r< CO O O O O O O O O O O O W N N O O O O O O C O C O O C M t D O O O CMCOCOO-rHCMCOOOOO ( N C O W t D C N C O O O H H CMiO-^T-^-^OOOOOOO COlNincCCOrirlri'HH^lC O O O O O O O O O ( M O t û O O t û O O O O - H O C O C O O O O O O O O c D r " - 0 0 O O O O O O O - r * T ^ H r ! ^ ^ r H ( N í N C N C N C N C N C N O O O O O O - H C D c O •^-(CMtMCMCMCMCMCOCO (MCMCMCMCMCMCMCMCM O O i O O O v ^ ^ O O O O O ^ - ^ l O T - I * # CD t O I O I O I O I O »O CD CD OiCOVOCOCOaiCTiCTiCDCTieMCM - O O O O O O C D O O O O O C M O O O O O » O O O T - ' C T 1 C T Î I O » O L O I O » 0 0 0 O c D c O c D O c O c O O O CMCOCOCOCOCOCOtOcO COCOCOCOCOCOCOCOCO [Mmowißwooo O O O O O O O O O CM CM COCOIOCOCDCOCOOOOOOOCMCM CMCMCMCMCMCMCMCMíMíMCOCO OO CD v * CD CU O CM T C O C N O O O t O c D c D ^ - ^ O C T S C v ï O c r î C r î O CMCMCMOr-fOOOOOOOOCOOOOO CD CO - H 0 0 - r " * 3 * • * *-< CM CM r ^ r - r ^ c o ^ c D t O t D t O c O O O O O O r i « l O i O W l O l 0 1 0 ^ CMCMCMCMCMCMCMCMCMCMCMCO OOOOOO-^-r-iOO » 0 0 0 0 0 CO CO CO CO CO CO CO CO CO O O O O O O O M ' Î CM * # O < f CO O r H r H ^ ! 5 0 0 C 0 0 0 r N C 0 ( ? a i O C O C 0 C 0 C 0 C O C 0 C 0 C 0 ^ - * ^ i O CN CO Í O O 1 ^ 0 0 OO r " IO IO IO IO l O IO tO O -r« CM~00~C0~ cor-N-^r^ • ^ CO O O O O O O O O O O O O 1 O O C D C O C M ^ C M * r * ï - O O C M < r C N O ' Î O Î ^ i n c O C M C î O C O O < O O O O O O O O O O O O O O O O O O O O O O O O Q O O O O O O O O O O O O O O O O O c o c o c o c o ^ o o o o 00 -s^ 00 - ^ T I r^'ví»cD<í r^oocTJCM-^uo-^CMCooio»oiooco o co o o r-. C M O O CM_ C ^ CM^ O ^ CM CM r » OC a 1* C O O O O O i C M O C O O r " Ì O O O O O O O O O O O O O v r v r c O 0 0 - ^ < r c D c D C O C M c D c D O C N C M c D O O - ^ i O r > . r ~ » C ^ C M i O O O O O O O O O O sì C O O • i !û O O O CO C \ l O LO CD C O ^ O ^ O ^ O ^ ooíoo"io oTó"r>.co"co'coco**-r-co"T- , o"o" CO O O L O CO CO L O O r ^ O C M C O r r i - O T - C M C M C O CO CO CO «J* *•?• «"S* ••5' »-T * 5 " «C **S" "^3* r » " o LO CN cN co »? [N r> >?iooor-