Ac 
 
Cornell University Library 
 HD5119.S5N3 
 
 Hours of work as related to output and h 
 
 3 1924 013 870 781 
 
Hours of Work as Related 
 
 to Output and Health 
 
 of Workers 
 
 SILK MANUFACTURING 
 
 Research Report Number 16 
 March, 1919 
 
 National Industrial 
 Conference Board 
 
^m ^axk 
 
 HnU O^allege of Agriculture 
 
 Kt <S,atnell 3ltntnecaitg 
 
 3Itbt£ii;g 
 
HOURS OF WORK AS RELATED 
 
 TO OUTPUT AND HEALTH 
 
 OF WORKERS 
 
 Silk Manufacturing 
 
 Research Report Number i6 
 March, 1919 
 
 Copyright 1919 
 
 National Industrial Conference Board 
 
 IS Beacon Street 
 Boston, Mass. 
 
Cornell University 
 Library 
 
 The original of tiiis book is in 
 tine Cornell University Library. 
 
 There are no known copyright restrictions in 
 the United States on the use of the text. 
 
 http://www.archive.org/details/cu31924013870781 
 
CONTENTS 
 
 PAGE 
 
 Foreword vii 
 
 Introduction 1 
 
 Statistical Basis of this Report 2 
 
 General Features of Silk Manufacturing 3 
 
 Source and Preparation of Silk 6 
 
 Principal Processes of Silk Manufacturing 7 
 
 Thrown Silks 7 
 
 Spun Silks 10 
 
 Fatigue Factors 11 
 
 Present Hours of Work in the Silk Industry .... 11 
 
 Saturday Half Holiday 12 
 
 Nominal and Actual Hours 12 
 
 Hours and Output 15 
 
 The 50-Hour Group 16 
 
 The 621^-Hour Group 16 
 
 The 54-Hour Group 17 
 
 The 55-Hour Group 18 
 
 Factors Bearing on Efficiency 18 
 
 Machine Efficiency 22 
 
 Relative Efficiency of Piecework and Daywork 23 
 
 Statistical Data as to Output 26 
 
 Difficulties of Securing Comparable Data 26 
 
 Comparisons of Output of Individual Establishments 28-35 
 
 Conclusions as to Output 37 
 
 Health of Silk-Mill Workers 39 
 
 General Health Conditions in the Industry .... 40 
 
 Accident Hazard 44 
 
 Conclusions as to Health 44 
 
 Appendix 47 
 
LIST OF TABLES 
 
 PAGE 
 
 Table 1 : Geographical Distribution of Establishments and Em- 
 ployees Covered by the Investigation, 1917 2 
 
 Table 2: Geographical Distribution of Silk Manufacturing 
 
 Establishments; United States Census, 1914 2 
 
 Table 3: Establishments and Employees, Grouped According 
 
 to Nominal Hours of Work per Week 12 
 
 Table 4: Percentage of Nominal Time Actually Worked During 
 55 and 50-Hour Weeks in Mills A and B, and 54 and 51-Hour 
 Weeks in Mine 13 
 
 Table 5: Summary of Changes in Output Accompanying Re- 
 ductions in Hours of Work, by Hour-Groups 15 
 
 Table 6: Comparison of Weekly Output of a Broad-Silk Mill 
 for 10 Weeks under a 55-Hour Schedule with that for an 
 11-Weeks Period under a 50-Hour Schedule 28 
 
 Table 7: Comparison of Weekly Output of a Large Velvet Mill 
 for Two 10-Week Periods under a 55-Hour and a 50-Hour 
 Schedule Respectively 30 
 
 Table 8: Comparison of Weekly Output of a Broad-Silk Mill 
 for a 6- Weeks Period under a 55-Hour Schedule with that for 
 a 6-Weeks Period under a 50-Hour Schedule 31 
 
 Table 9: Comparison of Weekly Output of a Broad-Silk Mill 
 for Three Months- under a 55-Hour Schedule with that for 
 Three Months under a 50-Hour Schedule 32 
 
 Table 10: Comparison of Weekly Output of a Broad-Silk Mill 
 for Two Months under a 55-Hour Schedule and for Three 
 Months under a 50-Hour Schedule 33 
 
 V 
 
PAGE 
 
 Table 11: Comparison of Weekly Output of a Throwing Mill 
 for Six Months under a 54-Hour Schedule with that for a 
 Similar Period under a 50-Hour Schedule 34 
 
 Table 12: Summary of Output Data Covering Reductions in 
 
 Ten Establishments from a 55 to a 50-Hour Week 36 
 
 Table 13: Percentage of Deaths of Males and Females Due to 
 Tuberculosis, of the Total Mortality for the Silk, Wool, and 
 Cotton Industries and for all Manufacturing and Mechanical 
 Pursuits, hy Age Groups, U. S. Census, 1909 41 
 
 Table 14: Percentages of Deaths of Male Silk- Workers from 
 Certain Causes, 1897 to 1900, Compared with Similar Per- 
 centages for 79 Selected Occupations, as shown by the Indus- 
 trial Experience of the Prudential Insurance Company ... 42 
 
Foreword 
 
 THIS report is the fourth of a series which the National 
 Industrial Conference Board is preparing, based on 
 an iijvestigation of the hours-of-work problem in major 
 industries. 
 
 The purpose of the investigation is to establish the 
 facts of this controversial subject and to present such con- 
 clusions as are clearly warranted by a scientific analysis 
 of these established facts. The investigation was not 
 undertaken with the intention of attacking or defending a 
 workday of any specific length, but, instead, to assemble 
 the available results of actual experience. 
 
 The discussion of hours of work should be sharply dis- 
 tinguished from that of a basic workday. The latter is 
 primarily a wages problem, but is often confused with 
 the problem of determining what number of hours of work 
 per day or per week will be most productive without im- 
 pairing the health of workers. 
 
 Conclusions as to the proper length of the workday, 
 moreover, must take into account not only the eflFect on 
 output and on health, but also the need of the worker for 
 a reasonable period of leisure for recreation, home life, 
 and self-development. 
 
 The problem varies so widely in different industries — 
 because of differences in type of worker, in character of 
 product, in management, and particularly in the extent 
 to which automatic machine processes influence results, — 
 that the experience of any one industry must not be taken 
 as representative of industry in general. 
 
Hours of Work as Related 16 Oiitput 
 and Health of Workers 
 
 SILK MANUFACTURING 
 
 INTRODUCTION 
 
 The Board's investigation of the hours-of-work problem 
 in the silk manufacturing industry was conducted along 
 the same general lines as the earlier investigations in the 
 cotton' and wool" manufacturing industries. The results 
 of the silk investigation, however, present a striking con- 
 trast with those for the two other textile industries studied. 
 
 In the northern cotton industry, reductions to less than 
 56 hours per week involved a general loss in ouput in more 
 than 90% of the establishments' and in a majority of cases 
 this loss was approximately proportional to the reduction 
 in time. Again, in the wool industry, a reduction to a 
 54-hour schedule resulted in a general loss in output, 
 though the loss was somewhat less pronounced than in the 
 cotton industry. 
 
 In the silk industry, however, a considerable proportion 
 of the mills reported output maintained after substantial 
 shortening of schedules to less than 54 hours per week, in 
 a number of cases, indeed, to 50 hours per week. 
 
 In the wool and northern cotton industries, moreover, 
 the predominating schedule was 54 hours per week, 
 whereas in the silk industry a large majority of the mills 
 reporting were on a 50-hour schedule. It is, therefore, on 
 this group that interest centers, and one important feature 
 of the investigation was to ascertain the reasons for the 
 difference in results. 
 
 ' Research Report No. 4 — Hours of Work as Related to Output and Health 
 of Workers, Cotton Manufacturing. 
 
 2 Research Report No. 12 — Hours of Work as Related to Output and Health 
 of Workers, Wool Manufacturing. 
 
 'Throughout this report the word "establishment" is used to designate a 
 corporation, partnership, or individual submitting a schedule. When a single 
 management operating several plants returned a consolidated schedule, this 
 was necessarily treated as one establishment. 
 
 1 
 
STATISTICAL BASIS OF THIS REPORT 
 Schedules of inquiry were addressed to members of the 
 Silk Association of America. Replies, stating the results 
 of reductions in hours, were received from 71 establish- 
 ments employing 30,095 workers.' This information was 
 supplemented by reports of field agents who visited a 
 large number of mills; 13 of these establishments employ- 
 ing 6,230 operatives did not file schedules. In all, the 
 report covers 84 establishments, employing 32.7% of the 
 total silk-mill workers in the United States as shown in 
 the 1914 Census of Manufactures. A classification, by 
 states, of establishments and of employees follows : 
 
 TABLE I : GEOGRAPHICAL DISTRIBUTION OF ESTABLISH- 
 MENTS AND EMPLOYEES COVERED BY THE INVESTIGATION, 
 
 I917 
 
 (National Industrial Conference Board) 
 
 
 
 Establishments 
 
 Employees 
 
 States 
 
 Nximber 
 
 Percent 
 
 Number 
 
 Percent 
 
 Tota 
 Pennsylvania . 
 New Jersey . 
 Connecticut 
 New York . . 
 Massachusetts 
 Maine . . . 
 
 
 84 
 
 19 
 39 
 12 
 9 
 3 
 1 
 1 
 
 100.0 
 
 22.6 
 
 46.4 
 
 14.3 
 
 10.7 
 
 3.6 
 
 1.2 
 
 1.2 
 
 35,325 
 
 12,740 
 
 10,439 
 
 8,872 
 
 1,912 
 
 1,074 
 
 139 
 
 149 
 
 100.0 
 
 36.1 
 
 29.6 
 
 25.1 
 
 5.4 
 
 3.0 
 
 .4 
 
 Maryland . , 
 
 
 .4 
 
 
 
 The industry is largely centered in New Jersey, Penn- 
 sylvania, New York, and Connecticut.' The distribution 
 
 'This total includes all factory employees, but does not include office, sales, 
 or engineering forces. 
 
 • The distribution of the silk industry in 1914 as shown by the U. S. Census 
 of Manufactures is given in the following table: 
 
 TABLE 2: GEOGRAPHICAL DISTRIBUTION OF SILK MANUFACTURING ESTAB- 
 LISHMENTS IN THE UNITED STATES, I914 
 (Compiled from data given in the U. S. Census of MantJactures, 1914) 
 
 
 Establishments 
 
 Employees 
 
 
 Number 
 
 Percent 
 
 Number 
 
 Percent 
 
 Total 
 Pennsylvania 
 
 902 
 284 
 368 
 
 44 
 143 
 
 19 
 5 
 
 39 
 
 100.0 
 
 31.5 
 
 40.8 
 
 4.9 
 
 15.8 
 
 2.1 
 
 .6 
 
 4.3 
 
 108,170 
 
 44,755 
 
 28,263 
 
 10,668 
 
 11,659 
 
 4,495 
 
 874 
 
 7,456 
 
 100.0 
 41.4 
 26 1 
 
 Connecticut 
 
 9 9 
 
 New York 
 
 10 8 
 
 Massachusetts 
 
 Maryland 
 
 4.1 
 g 
 
 All other states a 
 
 6.9 
 
 a Includes California 1; Delaware, 1; Georgia, 1; Illinois, 3; Maine, 1; Michigan, 2; 
 Ney Hampshire, 2j North Carolina, 4; Ohio, 1; Rhode Island, 12; Virginia, 9; West 
 Virginia, 1; Wisconsin, 1. 
 
of establishments covered by the investigation is there- 
 fore representative. 
 
 The 84 establishments upon which this report is based 
 furnished definite information reflecting the judgment of 
 the respective manufacturers as based, upon experience. 
 The data were supplemented by statistical comparisons of 
 output supplied by a considerable number of establish- 
 ments. The information assembled, therefore, affords 
 ample basis for valid conclusions. 
 
 The length of the work-week is, of course, only one of 
 many factors that affect output. In making output com- 
 parisons to show the effects of shortened schedules, it is 
 usually desirable to compare periods shortly before and 
 shortly after reductions in hours, thus avoiding or mini- 
 mizing the effects of many of the collateral factors such 
 as changes in personnel, season, or materials. The expe- 
 rience should, however, cover a sufficient period to make 
 the results fairly representative. 
 
 Output obviously is not the sole test of the justification 
 of any given hours-of-work schedule; the social needs of 
 the worker must likewise be carefully considered. The 
 broad social aspects of the hours-of-work problem are of 
 great importance. They may, however, be more intelli- 
 gently dealt with if the facts as to output and health are 
 first clearly determined. It is the purpose of this report 
 to assist in establishing the facts on these points, leaving 
 the social factors, which are less susceptible to statistical 
 treatment, for later discussion. 
 
 GENERAL FEATURES OF SILK 
 MANUFACTURING 
 
 Hand time feature of the work. Silk manufacturing on 
 the whole involves more handwork and consequently 
 production is more dependent upon the skill and 
 general efficiency of the worker than in cotton and wool 
 manufacturing. In some processes, however, the machine 
 dominates. In a few processes, indeed, the machine may 
 be said to control output. 
 
 As in cotton manufacturing it is sometimes possible to 
 run certain machines unattended for a time, but silk 
 yarn as a rule is so delicate that breakages necessitate a 
 
 3 
 
relatively large amount of handwork, particularly in 
 weaving. It is said that, on the average, a silk weaver 
 spends one-third of his time repairing broken threads or 
 remedying other disturbances In the warp. The per- 
 centage of the theoretical maximum efficiency' generally 
 attainable in silk weaving is, therefore, rather low as 
 compared with cotton weaving. With good warp an 
 automatic cotton loom may run practically all day without 
 stopping, attaining an efficiency of perhaps 98%; indeed, 
 by running unattended for a time at noon or after nominal 
 closing time, efficiency for a nominal day may rise con- 
 siderably above 100%. No such efficiencies are reported 
 in silk weaving. 
 
 The amount of hand time varies with the individual and 
 with the material being worked. Some processes are 
 almost exclusively handwork. Speaking in general, the 
 machine almost controls output in spinning. In weaving, 
 the machine also frequently dominates, but the operative 
 is a much more important factor than in spinning. In such 
 processes as sorting, trimming, inspecting, cutting, and 
 folding, handwork clearly controls output. 
 
 Size of silk establishments. A striking feature of the silk 
 industry is the wide range in size, equipment, and general, 
 arrangement of the plants. In a large percentage of cases, 
 weaving and throwing' are done by different establish- 
 ments. Many weaving mills purchase their yarn or have 
 it thrown on commission. Not infrequently a number of 
 very small establishments are located in a single building. 
 The capital and equipment essential to the operation of 
 such mills, especially if engaged in a restricted or special 
 line of work, may be small as compared with the amount 
 usually required in either the cotton or wool industries. 
 This is a consequence of the highly specialized nature of 
 the industry, the great value of the product, and the 
 opportunity for personality and individual skill. On the 
 other hand, a mill doing general silk manufacturing is 
 necessarily elaborate and costly. Particularly noteworthy 
 is the intricacy and massiveness of some of the machinery 
 required in such a mill. The raw silk is, of course, very 
 
 'The percentage efficiency of a loom for any particular kind of product dur- 
 ing a given period may be approximately determined by comparing the actual 
 number of picks (filling, or woof yarns; woven with the maximum possible 
 number that could have been woven had the looms worked at their known speed 
 without any stoppages whatever. 
 
 'The process of making yarn from reeled raw silk; from the Saxon word 
 "thrawan," meaning to twist. 
 
much more costly than the raw material used in any 
 other textile industry and, on the whole, the overhead 
 manufacturing charges are heavy/ 
 
 High percentage of women and minors. The number of 
 women and minors employed in the industry is larger 
 than for either cotton or wool manufacturing. The Census 
 of Manufactures in 1914 showed that approximately 54% 
 of the total work-force were women, and over 7% minors 
 under sixteen years of age. In those establishments 
 making a specialty of throwing, characteristic of certain 
 regions of Pennsylvania, for example, the percentage of 
 women and children is considerably higher than the 
 average. So high a percentage of women and minors 
 clearly introduces a special consideration into a study 
 of the hours-of-labor problem. 
 
 Compilations based on data for the 84 establishments 
 reporting show close agreement with the figures of the 
 United States Census of Manufactures with respect to the 
 proportion of men, women, and minors employed. 
 
 (National Industrial Conference Board) 
 
 
 N. I. C 
 
 . B. 19l'7 
 
 Census 
 
 ,1914 
 
 
 Number 
 
 Percent 
 
 Number 
 
 Percent 
 
 Total . . . 
 
 Men . . . 
 
 Women 
 
 Boys 
 
 Girls 
 
 33,775a 
 
 13,564 
 
 18,280 
 
 657 
 
 1,274 
 
 100.0 
 
 40.2 
 
 54.1 
 
 1.9 
 
 3.8 
 
 108,170 
 
 41,599 
 
 58,763 
 
 2,499 
 
 5,309 
 
 100.0 
 
 38.5 
 
 54.3 
 
 2.3 
 
 4.9 
 
 a Four establishments employing 1,550 persons failed to make returns concerning the 
 division of their employees into men, women, boys, and girls. 
 
 Wages. The average wages of silk-mill workers before 
 the war were higher than for cotton-mill workers but 
 lower than for wool-mill operatives. They also ranked 
 low among other leading industries. Compilations made 
 from the United States Census of Manufactures, 1914, 
 show that the average annual earnings of silk-mill workers 
 were approximately 27% below the average for 20 leading 
 industries." It should be pointed out, however, that a 
 
 'According to the data given in the U. S. Census of Manufactures, 1914, the 
 total capital invested in the silk industry was $210,072,000, as against 
 $867,044,000 for cotton manufacturing and $389,653,000 for woolens and 
 worsteds. The value of the products for the same year were $254,011,000, 
 $676,569,000 and $379,484,000 respectively. The proportion of these values 
 added by manufacture was as follows: silk, 75.9%; cotton, 56.8%; wool, 54.0%. 
 
 ' The industries were those standing first in value of products. 
 
 5 
 
portion of the workers were unskilled and fitted only for 
 comparatively simple tasks where the physical effort is Ugh t. 
 During the war substantial increases of wages were 
 made in the silk industry as in practically all other 
 industries. Changes in rates of wages accompanying 
 reductions in hours are noted in the Appendix. 
 
 Source and Preparation of Silk 
 
 All of the natural silk' used in the United States is im- 
 ported." Broadly speaking, it falls into two classes: (a) 
 Reeled-raw" silk, which goes to throwing mills to be made 
 into yarn, and (i) silk waste, which includes all unreelable 
 cocoons and parts of cocoons, and manufacturing waste 
 from which spun-silk yarns are made by processes similar 
 to other textile spinning. 
 
 ' Silk is chiefly produced by the Bombyx mori under domestication. The 
 worm envelopes itself in a cocoon by exuding, from its under-lip, two fine 
 thread-like streams of fluid which unite to form a single strand. The cocoon is 
 completed in about three days. The worm then enters the chrysalis stage and 
 changes within a fortnight into a moth which, unless killed, breaks its way out 
 of the cocoon, injuring the silk fibers and rendering them unreelable. A cocoon 
 from which the moth has thus escaped is called a "pierced cocoon" and is 
 classified as silk waste. In order to prevent this piercing of the cocoons about 
 98% of the worms are killed by heat, thus preserving the continuity of the 
 fiber. Normally a cocoon has, in addition to much waste fiber, a continuous 
 filament of reelable silk from 300 to 700 yards in length. So fine are the silk 
 filaments that, if placed end to end, a single pound might measure 1,000 to 1,500 
 miles, or even more. In order to render the fibers commercially useful, several 
 are in the reeling process simultaneously unwound from their respective cocoons 
 and, by the aid of the natural gum with which they are coated, combined 
 into one thread. 
 
 Artificial or imitation silk is fiber of finer strand than natural silk, obtained 
 from a cellulose base by chemical process. The yarn does not stand the strain of 
 warping as well as natural silk yarn and its use in the various processes of 
 manufacture should not be long delayed. Prior to the beginning of the war the 
 greater part of the artificial silk used in the United States was imported. 
 Recently, however, imports have been reduced and the manufacture of artificial 
 silk has increased. According to the Semi-Annual Report of The Silk Associa- 
 tion of America, 1918, the importations of yarns fell from 2,780,063 pounds in 
 the fiscal year 1914-1915 to 293,421 pounds in 1917-1918. Reports from other 
 sources indicate that the domestic manufacture increased from 320,000 pounds 
 in 1914 to 6,500,000 pounds in 1917. Returns from establishments making 
 artificial silk are not included in the statistical compilations of this report. 
 
 ^Compilations made from Information Series No. 3 of the U. S. Tariff 
 Commission show that silk imports for the fiscal year ending June 30, 1917, 
 totaled 40,351,423 pounds, valued at $160,571,808, distributed as follows: 
 Reeled-raw silk for thrown yarn 33,868,885 pounds; silk cocoons and waste 
 for spun yarn 6,482,538 pounds. 
 
 'Reeling, which requires much handwork, is not commercially possible in 
 this country. The low wages prevailing in China, Japan, and Italy, the main 
 sources of the raw silk used in America, effectively discourage any attempt to 
 establish this tedious but important process of manufacture in the United 
 States. Spun-silk manufacturing is, however, an important part of the silk 
 industry in America. This is possible for the reason that in spun-yarn manu- 
 facture, as in throwing and weaving, machinery rather than handwork plays 
 the leading part. 
 
 6 
 
PRINCIPAL PROCESSES OF SILK 
 MANUFACTURING 
 
 THROWN SILKS 
 
 The term throwing as used in the silk industry includes 
 the series of operations by which reeled or raw silk is 
 made into yarn. It is essentially a spinning or twisting 
 process including various doubling processes, by which 
 several raw silk threads .are combined. In other textile 
 spinning the process involves a drawing or attenuating of 
 the roving and then twisting it into yarns. In reeled raw 
 silk, however, the threads or fibers have already been 
 formed by the silkworm and no drawing is possible. 
 
 The chief products of throwing mills are organzine and 
 tram.^ Organzine is silk especially twisted for use as warp. 
 Two or more raw silk threads which have already been 
 twisted in the single are doubled and then twisted the 
 reverse way. Tram is often made of poorer qualities of 
 silk and is used for filling. Single untwisted threads are 
 loosely twisted together and unlike organzine require only 
 one spinning. 
 
 The principal processes of silk manufacturing with 
 reference to fatigue and health hazards of operatives may 
 be briefly described as follows :* 
 
 Winding: After soaking the raw silk to soften the gum 
 which adheres to it, and drying it — processes performed by 
 unskilled labor and almost ezclusively by men — the skeins are 
 placed on hexagonal power reels and wound on spools or bobbins 
 attached to rapidly revolving spindles.* The work of winding 
 consists of putting the skeins on swifts, as they are called, 
 joining the ends of broken fibers, removing the bobbins when 
 filled and replacing them with empty ones. Ordinarily one 
 winder will tend from SO to 90 reels or swifts. The usual num- 
 ber is 72. The work, which is usually performed by women 
 
 'An important branch of the silk industry is the manufacture of sewing 
 and embroidery silks. The several kinds are the results of as many methods 
 of handling raw silk. Though differing materially among themselves from 
 a technical standpoint, they involve no new principles, and are essentially 
 matters of doubling, winding, and twisting. Some spun-silk thread is also 
 made in this country. 
 
 ^ The machinery and procedure necessarily vary in the several processes of 
 silk manufacture. Moreover, there are some processes which while important, 
 perhaps even essential, are, nevertheless, relatively minor operations. In this 
 report it is obviously impracticable to attempt to discuss the industry in detail. 
 Accordingly some operations are not included in the description of processes. 
 
 ' Silk is also put through a similar process, often called " soft silk winding, " at a 
 later stage in the manufacture of silk fabrics. 
 
 7 
 
and girls, is simple and light, but it ordinarily requires close 
 attention on the part of the operative, with almost constant 
 activity, and some stooping. 
 
 Spinning: In first-time spinning, the silk as wound on 
 bobbins by the winding frames is rewound on other bobbins 
 and at the same time given a certain number of twists per inch. 
 A spiijner tends from a few hundred to approximately a thou- 
 sand spindles, the number varying with the quality of its fiber 
 and size of the yarn. The work consists chiefly in piecing broken 
 thread and in changing bobbins. In case a thread becomes 
 broken the bobbin affected is automatically stopped. This 
 also happens when the bobbin has been wound full. Thus the 
 threads can be pieced together for a particular bobbin, or a 
 bobbin can be removed without interfering with the operation 
 of any other spindles on the frame. Because of the large 
 number of spindles attended and the high rate of speed, some- 
 times more than 10,000 revolutions per minute, the operative 
 must constantly walk up and down the aisle, and in remedying 
 trouble much stooping is required. While in some localities 
 many men are engaged in spinning, women are chiefly employed. 
 
 Second-time spinning is substantially similar to first-time 
 spinning, but with the twist in the opposite direction, doubling 
 two or more strands together. 
 
 Doubling is the uniting of two or more strands into a loose 
 thread. It is very similar to winding and is light, semi-skilled 
 work largely performed by women and girls. 
 
 Reeling: The process of reeling consists merely in unwinding 
 the thread from the^ bobbin and rewinding on a hexagonal 
 frame or reel so that it may be dyed. The work of the operative 
 consists in putting the bobbins in the proper position on the 
 winding frame, in straightening the threads in case they become 
 snarled, and in tying ends of broken threads. 
 
 Considerable manual effort is required in lifting the reels 
 from the frames and carrying them to benches for the sub- 
 sequent process of lacing.' Where girls or young women are 
 employed in reeling it is customary to have men handle the 
 reels, since they weigh twenty pounds or more. These workers 
 are called "reel carriers" and usually do nothing else. Where 
 men are employed for the reeling proper, however, they also 
 handle the reels. 
 
 Quilling is essentially another form of winding and consists 
 in transferring the silk threads from bobbins to small spools or 
 
 ^Lacing, which is light work usually performed by children and new em- 
 ployees, consists in dividing the skeins into a number of parts by passing short 
 strings between the strands of thread. Bundling consists in twisting the skeins 
 into tight rolls and packing them into bundles preparatory to dyeing. This 
 work is light but calls for considerable judgment and dexterity. An important 
 feature is the detection of imperfections in the thread. 
 
 8 
 
quills which are placed in the shuttles of the looms. An empty- 
 quill is placed on the bobbin and when wound full with thread 
 is removed and replaced with another empty one. Women and 
 girls are almost exclusively engaged in this work and there is 
 opportunity to sit perhaps one-fourth of the time. Close 
 inspection and the removal of all dust and dirt from the machin- 
 ery constitute an important part of the operative's duties. 
 
 Warping: Silk warping is the process of unwinding warp 
 threads from the spools on which they were wound in the 
 doubling and spinning process and placing them methodically 
 on loom beams. The work consists mainly in removing rough 
 threads and knots, and in repairing broken threads. It involves 
 constant standing and considerable bending over, though there 
 arc intervals of inactivity. 
 
 Drawing-in and Twisting-in: The process of drawing-in 
 consists in passing the ends of warp threads through the eyes 
 of the harness and the reeds of the loom. It is skilled work. 
 Hand twisting-in is the process of uniting, by twisting, the 
 thousands of new threads with the corresponding ends of the 
 piece of fabric just completed. The work, which is highly skilled 
 labor, is done while sitting and requires much reaching. Both 
 men and women are employed. 
 
 Weaving: Plain silk weaving is very similar to plain cotton 
 and wool weaving. On account of the adaptability of the 
 Jacquard loom to weaving a great variety of patterns, that 
 form of loom is extensively employed. Although in Jacquard 
 weaving the threads are automatically controlled by perforated 
 cards, there is, nevertheless, great opportunity for skill on the 
 part of the operative, particularly in watchfulness to prevent 
 unnecessary breakage and in promptly repairing threads which 
 become broken. In addition the operative must, of course, 
 keep the shuttles filled with quills. 
 
 Partly because of the peculiar properties of silk and partly 
 because of the higher value of the product which makes spoiled 
 work more costly, silk weavers do not tend as many looms as 
 weavers in cotton mills. In the Paterson, N. J., district ordina- 
 rily not more than two looms are tended by a single operative. 
 This, however, is apparently a result of arbitrary restrictions 
 by labor organizations in that district, as four and even five 
 looms are frequently tended by a single operative in other 
 localities.* To catch defects in warp and filling, and generally 
 supervise the operation of the loom, requires constant attention,- 
 though, of course, the amount of actual physical effort varies 
 with the material, style of fabrics, number of looms, and, by 
 no means least, with the skill and alertness of the weaver him- 
 self. An efficient weaver can obviate a large amount of un- 
 necessary labor. 
 
 'Where, however, the two-loom system prevails the fabric is, as a rule, fancy 
 and elaborate. 
 
Both men and women are extensively employed in weaving. 
 
 Dyeing: In general, silk is dyed either in the yarn or in the 
 piece.', Dyeing is sometimes done by hand manipulation of the 
 fabrics in dye vats or by running the silk through dye by 
 power-driven reels. Silk dyeing lacks mechanical precision 
 and as a rule is fatiguing work, so that very few women are 
 employed; daywork wages largely prevail. 
 
 Printing: Printing is done by means of engraved designs on 
 copper cylinders, a separate cylinder being used for each color. 
 The cloth is then passed through in a manner somewhat re- 
 sembling the feeding of paper into a cylinder printing press. 
 The colors are set by steam. The work is performed by men. 
 
 Cloth Picking: The cloth as it comes off the loom contains 
 rough ends of threads which have to be picked off. In addition, 
 grease stains and dirt spots have to be removed. Picking is, on 
 the whole, light work, requiring skill; it is performed by women 
 and girls. After this, the goods are put through what is called 
 the finishing process by which glue is deposited^ upon the cloth 
 by means of a steam atomizer; many other finishing processes 
 are also employed. The fabric is then pressed by hot rolls to 
 make it smooth and at the same time give it the luster or gloss 
 which characterizes it in its commercial form. Finishing is 
 skilled work generally performed by men. 
 
 Beyond this are several comparatively simple processes of 
 preparing the silks for shipment. This work is generally 
 performed by women. 
 
 SPUN SILKS 
 
 The manufacture of spun silk involves more work, both hand 
 and machine, than does the manufacture of thrown silk. More- 
 over, the highly developed state of the raw silk liber from which 
 thrown silk yarns are made renders unnecessary many of the 
 involved processes required in the manufacture of spun silk and 
 other yarns such as cotton and worsteds. 
 
 As already pointed out, those mills which use silk waste 
 instead of reeled fiber have to prepare the waste and spin it. 
 The process called dressing is in many respects similar to the 
 preparation of cotton or wool, but differs in some cases, 
 particularly in that the waste has to be boiled, usually in 
 a solution of soapsuds in order to remove the natural silk gum 
 and to cleanse it from dirt and impurities. After drying, the 
 mass is run through revolving cylinders which pull the fibers 
 apart, straighten them, and leave them in the form of sheets 
 or laps. These sheets are later cut into short lengths, automati- 
 
 • Weighting of silk is usually done in connection with the dyeing, by the 
 addition of metallic salts or by use of vegetable or organic matter deposited in 
 the fiber, sometimes to the extent of half the weight of the silk. Silk is usually 
 weighted in the skein; piece-dyed goods are but little weighted. 
 
 10 
 
cally hung over small rods, in which position they are combed 
 and cleaned and later run into laps of several feet each. In 
 this form the laps are taken to the spinning mill and drawn into 
 a loose strand called sliver. The mass is put through roving 
 frames and is wound on bobbins. It is then run through the 
 spinning frame. This process is similar to those described in 
 previous reports of the Board on cotton and wool manufac 
 turing. Among the subsequent processes is gassing, in which 
 fuzz is removed from the thread by gas flame. The yarn is 
 also passed through a machine for removing knots, after which 
 it may be reeled and woven. 
 
 Fatigue Factors 
 
 Most of the heavy work in the silk industry is per- 
 formed by men. As was pointed out by various manu- 
 facturers, however, the physical condition of a worker, 
 the degree of fitness to perform a certain task, often is 
 much more of a fatigue factor than is the mere physical 
 effort normally required for that operation. In every mill 
 there are operatives who appear to suffer fatigue while 
 engaged in comparatively light work, while others will 
 finish a day of much heavier work with little or no notice- 
 able fatigue. Again, as in other factory industries, noise, 
 monotony, poor ventilation, standing, reaching, stooping, 
 eye-strain, etc., alone or In combination, may cause serious 
 fatigue. Sometimes the operatives most exposed to these 
 objectionable conditions are women and children, who 
 suffer In consequence considerable fatigue, although the 
 required physical effort may be light. 
 
 As noted elsewhere, good lighting and cleanliness are 
 essential to the efficient conduct of the Industry. On the 
 whole, the silk manufacturing Industry compares very 
 favorably with manufacturing In general; many mills, 
 indeed, are excellently constructed and arranged. There 
 Is some unguarded machinery, but probably not more 
 than In other industries where machinery plays a leading 
 part. 
 
 PRESENT HOURS OF WORK IN THE 
 SILK INDUSTRY 
 Of the 84 establishments covered by this report 64, or 
 76.2%, were on schedules of from 50 to 523^ hours per 
 week. In 55 establishments the hours were 50 and in 
 three others they differed from this schedule by only a 
 fraction of an hour.' In other words, a 50-hour week was 
 
 1 These 58 establishments are hereinafter treated as composing the 50-hour 
 group. 
 
 11 
 
in force in practically two-thirds of the establishments 
 included in the investigation. 
 
 Only about 3.4% of the workers were in establishments 
 where the weekly schedule was more than 64 hours. 
 
 TABLE 3 : ESTABLISHMENTS AND EMPLOYEES, GROUPED 
 
 ACCORDING TO NOMINAL HOURS OF WORK PER WEEK 
 
 (National Industrial Conference Board) 
 
 
 Establishments 
 
 Employees 
 
 Nominal Weekly 
 
 
 
 
 
 Hours of Work a 
 
 Number 
 
 Percent 
 
 Number 
 
 Percent 
 
 Total .... 
 
 84 
 
 100.0 
 
 35,325 
 
 100.0 
 
 60* 
 
 58 
 
 69.0 
 
 22,717 
 
 64.3 
 
 51 
 
 1 
 
 1.2 
 
 263 
 
 .7 
 
 try. ■.:■.■.■. 
 
 5 
 
 6.0 
 
 3,641 
 
 10.3 
 
 11 
 
 13.1 
 
 7,513 
 
 21.3 
 
 55 
 
 9 
 
 10.7 
 
 1,191 
 
 3.4 
 
 a A number of establishments designate regular periods, usually a half-hour per 
 week, for cleaning machinery, etc., the actual operating tiine thus becoming 
 correspondingly less than the nominal hours. 
 
 h In 1 of these nulls the hours were 50^', 
 
 c In 2 of these mills the hours were 52 ^. 
 
 d In 1 of these mills the hours were 535^. 
 
 The high percentage of women and minors, the exacting 
 character of the work, and the specific demands of labor 
 have all contributed to the rather general observance of 
 the 50-hour schedule. Only about one-fifth of the mills 
 covered by this investigation are running on schedules as 
 high as the legal limitation of working hours for women 
 in their respective states, and in these mills the schedules 
 for men and women are the same. In no estabhshment 
 were the nominal hours for men higher than for women. 
 
 Saturday Half Holiday 
 In the silk industry as in the northern cotton and wool 
 industries the Saturday half holiday, is almost universally 
 observed. Without exception, the 84 establishments 
 covered by this report make provision for closing 
 at about noon on Saturday. Furthermore, with the 
 exception of five plants, all of the establishments had 
 observed the Saturday half holiday prior to the latest 
 reduction in hours. 
 
 Nominal and Actual Hours 
 
 The average actual hours of workers in any silk milt 
 
 ^how some variation from week to week. Normally, of 
 
 course, the actual hours are somewhat less than the 
 
 nominal hours. This margin in the silk industry is not 
 
 12 
 
due to general absence among the work-force at any given 
 time, but reflects the absence of a varying proportion of 
 the force in any given week, averaged over the entire 
 number. The great majority of workers put in practically 
 the full nominal schedule. Many employers, recognizing 
 that more or less absence is inevitable, endeavor to keep 
 substitute hands available for the more important tasks. 
 
 The margin between actual and nominal hours appears 
 to be fairly constant, irrespective of the length of the 
 nominal factory schedule. A comparison for three estab- 
 lishments representing three distinct geographic sections 
 of the industry before and after reductions in nominal 
 hours, is given in Table 4. It will be noted that there was 
 no marked difference in the percentage of lost time when 
 hours were shortened. 
 
 TABLE 4: PERCENTAGE OF NOMINAL TIME ACTUALLY 
 WORKED DURING 55 AND 5O-HOUR WEEKS IN MILLS A 
 AND B, AND 54 AND 5I-HOUR WEEKS IN MILL C 
 
 (National Industrial Conference Board) 
 
 
 Before reducing 
 nominal hours 
 
 After reducing 
 nominal hours 
 
 
 Number of 
 operatives 
 
 Percent full 
 time worked 
 
 Number of 
 operatives 
 
 Percent full 
 time worked 
 
 MILL A 
 
 Whole Mill . . . . 
 
 Men 
 
 Women ... 
 
 269 
 
 209 
 
 60 
 
 a 
 
 94.7 
 95.1 
 94.0 
 
 238 
 
 170 
 
 68 
 
 b 
 
 94.6 
 94.6 
 94.8 
 
 MILL B 
 
 Whole mill 
 
 Men 
 
 Women 
 
 184 
 
 36 
 
 148 
 
 c 
 
 88^0 
 89.4 
 87.6 
 
 224 
 37 
 
 187 
 
 d_ 
 
 90.8 
 85.6 
 
 MILL C 
 
 Whole mill . . . . 
 
 Men 
 
 Women 
 
 228 
 41 
 
 187 
 
 e 
 
 9a9 
 95.2 
 90.0 
 
 163 
 
 23 
 
 140 
 
 / 
 
 92.7 
 97.2 
 92.2 
 
 Mill A. u. Velvets, Connecticut. Full or nominal time, September, 1916, 55 hours 
 per week. 
 6. Full or nominal time, September, 1917, 50 hours per week. 
 Mill B. c. Silk ribbons, Paterson, N. J. Full or nominal time, September, 1915, 
 65 hours per week. 
 d. Full or nommal time, September, 1916, 50 hours per week. 
 Mill C. «. Broad silks. Pennsylvania. Full or nominal time, September, 1917, 
 54 hours per week. 
 /. Full or nommal time, September, 1918, 51 hours per week. 
 The foregoing percentages are not givpn as being necessarily representative. 
 
 13 
 
Time spent in a silk mill is, of course, not necessarily 
 all spent in physical effort. While silk operatives have on 
 the whole less opportunity for rest during working 
 hours than do cotton or even wool workers, there are, 
 nevertheless, intervals in which relaxation is possible. In 
 some processes these intervals are considerable. Undue 
 weight should not, however, be attached to mere freedom 
 from actual physical effort, for often the chief fatigue 
 factor, even under excellent working conditions, is not so 
 much actual physical effort as it is the tension of continued 
 watching, of being constantly on the alert. These con- 
 siderations obviously have a bearing on the determination 
 of the proper length of the workday. 
 
 The silk industry as a whole is subject to considerable 
 variations in the number of workers employed, but these 
 variations cannot be said to be seasonal; rather they are 
 the reflection of market prices and conditions in general. 
 
 14 
 
HOURS AND OUTPUT 
 
 Of 84 establishments which reported the effect on out- 
 put accompanying reductions in hours of work, two re- 
 ported that production was increased, 17 that it was 
 maintained, and 65 that it was reduced. A summary 
 table in the Appendix epitomizes the results as reported 
 by the managements of the several establishments. The 
 reductions in weekly schedules covered by the report, 
 with two exceptions, occurred during the period from 1909 
 to 1917; the greater number were made in 1916. 
 
 Of the 84 establishments covered by the report, 58, em- 
 ploying 69% of the total number of workers, reduced to 
 60 hours, one reduced to 51 hours, five to 52}^ hours, 
 eleven to 54 hours, and nine to 55 hours. 
 
 Table 5 shows the various hour-groups and compares 
 
 TABLE 5 : SUMMARY OF CHANGES IN OUTPUT ACCOMPANY- 
 ING REDUCTIONS IN HOURS OF WORK, BY HOUR-GROUPS 
 
 (National Industrial Conference Board) 
 
 Production Maintained 
 
 
 Production Decreased 
 
 
 Previous 
 hours 
 
 Number 
 establish- 
 ments 
 
 Number 
 employees 
 
 Previous 
 hours 
 
 Number 
 establish- 
 ments 
 
 Number 
 employees 
 
 Reduced to 50 Hours 
 
 55 
 55 
 54 
 
 54 
 
 55 
 54 
 
 56 
 55 
 
 60 
 
 58 
 
 8 
 
 ir 
 
 1 
 
 ,1/ 
 
 I}' 
 
 S69 
 
 3,139 
 
 268 
 
 263 
 
 1,255 
 1,409 
 
 842 
 1,244 
 
 168 
 173 
 
 55 
 
 54 
 
 55 
 
 60 
 58 
 56 
 
 60 
 
 58 
 
 57yi 
 
 2 
 
 2 
 
 6 8 
 ij 
 
 4) 
 1 6 
 ij 
 
 14,563 
 
 Reduced to 61 Hours 
 
 4,388 
 
 Reduced to BS\ Hours 
 
 977 
 
 Reduced to BU Hours 
 
 570 
 4,172 
 
 Reduced to 66 Hours 
 
 685 
 
 630 
 
 149 
 
 71 
 
 Establishments . . 
 
 Employees .... 
 
 Percent of establish- 
 ments 
 
 Percent of em- 
 ployees 
 
 
 19 
 22.6 
 
 9,120 
 25.8 
 
 
 65 
 77.4 
 
 26,205 
 74.2 
 
 * Increase shown in italics. 
 
 15 
 
the number of employees and the previous hours of the 
 establishments maintaining output, when hours were 
 reduced, with similar data from the establishments 
 reporting that production was decreased. 
 
 This table shows that for approximately one-quarter 
 of the establishments reporting, output was maintained 
 (or in two cases increased), while in the remaining three- 
 quarters a decrease was reported. The proportions of 
 employees in the two groups were approximately similar. 
 In order to make an intelligent analysis of these figures, 
 however, it is necessary to take up the discussion by hour- 
 groups. 
 
 The 50-Hour Group 
 
 As stated at the beginning of this report, it is in the 
 50-hour group of establishments that interest centers. It 
 is the largest group ; its mills cover a wide range of size and 
 represent every branch as well as every important manu- 
 facturing district of the industry.' In nine of the 58 mills 
 in this group output was maintained, two of the nine, in 
 fact, showed an increase, while in 23 others the decrease 
 in output was relatively less than the reduction in hours, 
 although in several of the 23 cases the loss was substantial. 
 These 32 establishments employed approximately 68% 
 of the employees in the 50-hour group or about 44% of 
 the number in all groups covered by the investigation. 
 The data indicate that reduction to a 50-hour work-week 
 in the silk manufacturing industry has in a great majority 
 of cases resulted in a decrease in output. Taking the 
 industry as a whole, however, there has undoubtedly been 
 some gain in the efficiency of the workers and, in conse- 
 quence, the loss in output was less than proportional to 
 the reduction in hours. 
 
 The 523^-Hour Group 
 
 In the 523^-hour group, three establishments reported 
 output maintained, while two reported that it was de- 
 creased; all were large plants. One mill of the three 
 maintaining production attributed a part of its gain in 
 efficiency to a bonus system and to a system of rest periods ; 
 in^another the actual reduction in time was only one 
 
 'The geographical distribution of the mills in the 50-hour group is as follows: 
 New Jersey, 36; New York, 7; Pennsylvania, 8; Connecticut, 5 ; Massa- 
 chusetts, 2. 
 
 16 
 
hour and fifteen minutes, effected by extending the lunch 
 period. The weight of the evidence for this group, though 
 obviously far too limited to be accepted as conclusive, is 
 in favor of the shorter schedule. 
 
 Only one mill reduced to 51 hours; consequently no 
 deductions concerning this schedule can be made, but the 
 fact that it maintained output gains some significance 
 when considered in connection with the 50 and 523^- 
 hour groups. 
 
 The 54-Hour Group 
 
 In the 54:-hour group, eight estabHshments reported 
 that output was decreased and three that.it was main- 
 tained. Of these three mills two made a reduction 
 of only one hour each by closing earlier on Saturday; 
 in one of the two there was considerable change in the 
 working force which appeared to increase the average 
 efficiency. In four of the eight mills in this group reporting 
 a decrease in output, the loss was less, proportionally, 
 than the reduction in time. The eight mills reporting 
 output as decreased, with one exception, reduced from 
 either a 58-hour or a 60-hour schedule. The average 
 reduction was, therefore, over four hours per week as 
 compared with a reduction of two hours or less for the 
 three mills in this group which maintained production. 
 
 The 54-hour group included 7,500 workers, or about 
 one-fifth of the total number covered by the report; 
 approximately 2,100 were in mills reporting output main- 
 tained, against 5,400 in establishments reporting output 
 decreased. The three establishments maintaining output, 
 as well as the eight reporting a decrease, include throwing 
 mills in the processes of which machinery dominates pro- 
 duction. It should be pointed out, however, that in those 
 establishments which maintained output the throwing 
 was done in combination with other processes where hand- 
 work plays a more important part. For mills doing throw- 
 ing exclusively, there is evidence that a 54-hour week will 
 not maintain production as against a somewhat longer 
 schedule. The evidence for the whole group indicates 
 that a 54-hour schedule did not maintain maximum out- 
 put, but the basis of experience is so limited and the 
 qualifications and collateral conditions so varied that this 
 cannot be stated as a positive conclusion. 
 
 17 
 
The 55-Hour Group 
 
 In the 55-hour group, three mills maintained output 
 and six reported a decrease; all were small establishments, 
 together employing less than 4% of the total number of 
 workers covered by the table. In one establishment it was 
 stated that maintenance of output was due to the inaugu- 
 ration of a bonus system and not to the reduction in time. 
 In another case it was stated that a throwing mill which 
 maintained output had reduced hours some fifteen years 
 ago, since which time important changes have taken 
 place in the throwing industry. The weight of the evi- 
 dence, though again limited, indicates that maximum out- 
 put was not maintained with a 55-hour week. 
 
 FACTORS BEARING ON EFFICIENCY 
 
 The preceding discussion shows that results of reduc- 
 tions in hours-of-work schedules in the silk industry 
 varied widely in different establishments. An important 
 fact is that such differences are found in the case of each 
 different weekly schedule. This means, of course, that 
 changes in the length of the workday are not the only 
 factor affecting output. It becomes necessary, therefore, 
 to examine some of the other factors in the problem such 
 as the character of the product, the size of the establish- 
 ment, changes in wages, efficiency of management and the 
 relative importance of machine and handwork. 
 
 The character of the product apparently did not have 
 a controlling influence on output with the possible excep- 
 tion of those processes which, like throwing, are largely 
 dominated by machinery. In respect to finished goods, 
 it appears that both the group of mills maintaining pro- 
 duction and the group reporting output as decreased 
 covered a considerable variety of product. 
 
 Similarly, ability to maintain production when hours 
 were reduced does not appear to depend on the location 
 or the size of the establishment. Some large plants 
 reported decreases while others reported that output was 
 maintained. Very small plants and those of moderate 
 size likewise are found in each group. 
 
 Again, changes in output are not attributable to 
 changes in rates of wages. In a great majority of estab- 
 lishments reporting on this point, piece rates or hourly 
 
 18 
 
rates of wages were substantially increased when hours 
 were reduced, but results on the whole were not signifi- 
 cantly different than for those establishments in which 
 rates remained unchanged. 
 
 In some cases the payment of a bonus tended to 
 increase the efficiency of workers but there is not a 
 sufficient basis of experience to warrant conclusions on 
 this point. 
 
 Differences in output results among different establish- 
 ments, when hoursof workwere shortened, are undoubtedly 
 explained in part by differences in the efficiency of manage- 
 ment. In this respect, there is, of course, a wide range. 
 Under any given schedule of hours it will be found that 
 some establishments will fail to secure the results obtained 
 by others. It does not necessarily follow that the estab- 
 lishments which do not maintain production are less 
 efficient than those maintaining output, since the char- 
 acter of the process, the class of employees, the raw 
 material, or numerous other factors may have had an 
 important bearing. Comparison of the efficiency of dif- 
 ferent mills is every difficult. But that variations in the 
 efficiency of management are bound to have an important 
 influence on the ability of different establishments to 
 maintain production under any given schedule of hours is ' 
 indisputable. 
 
 It is, however, of secondary interest to explain the 
 differences in results in different establishments. The 
 important question to determine, especially in view of the 
 general failure of cotton and wool manufacturing estab- 
 lishments to maintain output on even a 54-hour schedule, 
 is how some silk establishments are able to accomplish 
 this with a still shorter week. 
 
 The principal reason apparently is to be found in the 
 greater importance of hand-time work in silk manufactur- 
 ing with the consequently greater opportunity for increas- 
 ing the efficiency of the individual worker. Silk machinery 
 is, on the whole, less automatic than wool-mill equipment, 
 and very much less automatic than cotton machinery. 
 
 In this connection an official of a large silk establish- 
 ment said: 
 
 The cotton loom running on good work practically runs 
 all day long without any warp breakage, and all that is neces- 
 sary is to keep changing the shuttles, so that it is practically 
 
 19 
 
all machine work and very little hand-time work for the 
 weaver. On silk looms, warp breakage is very much greater 
 and, of course, varies according to the work that is being done, 
 but on a general average a weaver loses one-third of his time 
 repairing broken threads or other disturbances in the warp. 
 This so-called hand-time work, of course, may vary among 
 the various weavers, depending on their ability and dexterity. 
 
 If thirty-three per cent (33%) of a weaver's time is spent on 
 handwork, in a shorter hours-of-work period the energy may 
 be sufficiently increased and the fatigue sufficiently decreased 
 to permit a weaver to produce as much in nine hours as he 
 formerly did in ten.^ 
 
 The energy of a weaver and power of work may vary 
 according to the hours of work, the longer the hours, the 
 more the loss of energy and power to work towards the end 
 of the day, or in other words, the loss caused by fatigue. 
 
 This is also the case throughout most of our operations 
 requiring a lot of handwork, and we find that people can 
 practically deliver in nine hours the same amount of work as 
 done before in ten, whereas, where the operation is more or 
 less automatic, the loss in production is proportionate to the 
 loss in time. 
 
 Another large manufacturer said, in substance: 
 
 The ability of a worker to maintain output on a 50-hour 
 schedule as against a 55-hour schedule varies with the amount 
 of handwork. This is not an ironclad rule but I can safely 
 say that, throughout our mills, in those departments where 
 machine work controls to a high degree, output has fallen 
 about proportionally to the reduction in time. On the other 
 hand our experience has been that in processes where hand- 
 work controls there has been a gain in efficiency in some cases 
 su^icient to offset the loss of time at the worker's disposal. 
 Between these limits there are various sets of conditions, but 
 the fact seems to be that the ability to increase hourly effi- 
 ciency grows less as the part a machine plays in any task 
 increases. In spinning, the machine almost controls output. 
 Machines are less the master in dyeing and weaving but even 
 there the machines largely dominate. In contrast with these 
 processes we have the opposite extremes where handwork 
 controls, such, for example, as sorting materials, trimming, 
 inspection, cutting and folding. 
 
 In line with these opinions is the statement of another 
 silk manufacturer submitted to a committee of Congress 
 in connection with tariff discussion.* He said in part: 
 
 ' This statement refers to a reduction from a 55-hour to a 50-hour week. 
 ' U. S. Tariff Commission — Tariff loformation Series No. 3. Schedule L. 
 Statement of Mr. Horace B. Cheney. 
 
 20 
 
The hands of the silk worker are one of his most important 
 assets. Take a man from the fields of Siberia, from the plains 
 of Austria, from Southern Italy, or from a farm in the United 
 States or Sweden — and that is about the conglomeration we 
 get in this country — and it is a different matter to make a 
 silk worker of that man, a very different matter, from taking 
 men who have been brought up in countries where silk is 
 produced, where the very habits and occupations have de- 
 veloped the technique and the kind of hand that makes the 
 silk worker. They work to better advantage with it. Ma- 
 chinery does not do away with the use of hands in silk manu- 
 facture. The hands still remain, and will always remain, in 
 my opinion, a very important factor in the operation. A man 
 with clumsy, awkward hands handling silk warp is a very 
 different factor from the man whose grandfather before him 
 handled the silk fabric. 
 
 From the statements just given, it is clear that many 
 operations in silk manufacture involve much hand-time 
 work in which the quality and quantity of the product 
 turned out depend to a considerable extent on the skill of 
 the individual worker. , There can be no doubt that this 
 fact explains in part the ability of some silk establishments 
 to maintain production on a schedule of hours much 
 shorter than is practicable in either the cotton or wool 
 manufacturing industries. 
 
 If machinery always ran at a uniform speed without 
 stoppage during each spell of the workday, a maximum 
 of efficiency would be attained. That is, actual production 
 would coincide with the theoretical maximum of 100%. 
 If the machinery were stopped for 25% of the time, 
 there would be a 25% loss in production. Similarly, the 
 effect of a 25% reduction in work-hours, unless offset by 
 an increase in the efficiency of the operative, would be to 
 bring a corresponding loss in output. Where the speed 
 or type of machinery cannot be improved under such 
 reductions in hours, a loss in output is unavoidable in 
 processes where the machine dominates, unless it is pos- 
 sible to increase the efficiency of the operative. Con- 
 versely, there is more opportunity to offset a reduction in 
 the work schedule where handwork is an important factor. 
 For instance, in cotton spinning, handwork represents only 
 a small part of the process, perhaps not more than 5%. If 
 machine time is reduced 20% by a reduction in the hours 
 of work, and there is no opportunity to increase the speed 
 of the machinery, it becomes almost impossible for the 
 operative to increase his own efficiency to an extent which 
 
 21 
 
will fully offset the loss in machine time. The same 
 principle holds generally true in the case of silk throwing 
 although the element of handwork here is rather greater 
 than in the case of cotton spinning. For this reason it is 
 exceedingly difficult for silk mills exclusively engaged in 
 throwing processes to maintain production when hours of 
 work are shortened below 54 or 56 per week. 
 
 In the case of weaving, however, and in some other 
 processes in silk manufacture, the operative has much 
 greater opportunity to increase personal efficiency. Thus 
 weavers must be on the alert for many troubles. Among 
 these may be mentioned the following: twisting and 
 sticking of yarn at the cross-rods; unequal running of the 
 loom and variations in the harness due to extremes of 
 heat and moisture in the weave room; imperfections in 
 the fabric due to damaged reeds and worn blades; spotting 
 of goods due to improper oiling; spotting or streaking 
 due to loose particles from dirty harness or machinery; 
 streaks due to badly drawn ends in the harness or crossed 
 in the reed; imperfections caused by too many knots 
 (smashes) ; imperfections in the fabric due to looms , 
 working irregularly or out of time; broken picks in filling 
 resulting from improper functioning of the stop motion 
 device. 
 
 The extent to which the machine dominates cannot be 
 determined from cursory observation. Some processes 
 which seem almost automatic really involve a consider- 
 able amount of hand-time work. 
 
 Machine Efficiency 
 
 Some mills reporting for this investigation stated that 
 they were able to increase the hourly efficiency of their 
 workers by installing improved machinery and by in- 
 creasing machine speeds. These changes in a few cases 
 appear to have been sufficient to offset a substantial 
 reduction in working time. In many establishments, 
 however, a high state of machine efficiency had been 
 attained prior to the latest reduction in hours. It is clear 
 that neither the installation of new machinery nor the 
 increase of speeds is a factor the beneficial effects of 
 which should be attributed to the shortening of the work- 
 week, except in cases where the new machinery or the 
 increased speed involved such extra strain on the part of 
 the workers that a longer work schedule would not have 
 
 22 
 
been practicable. If, however, a reduction in the hours of 
 work permits an increase in the speed of machinery beyond 
 a point which would be practicable under a longer schedule 
 of hours, this gain is fairly to be used as an argument in 
 favor of the shorter work-week. But if speed of machinery 
 could have been increased under the longer week without 
 increasing the operatives' fatigue, a reduction of hours 
 means a relative loss in production even though actual 
 production may be as great as before. 
 
 By the use of electric power and especially by means of 
 the individual motor drive, some establishments have 
 been able to speed up production. There always are, 
 however, speed-limiting features, such as the quality of 
 the material used. Thus both the quantity and quality 
 of yarn production is affected by the quality of raw silk 
 and silk waste, while loom production is in turn affected 
 by the quality of the yarn. 
 
 On the whole, it appears that improvements in machin- 
 ery or increases in speed of machinery were not an im- 
 portant factor in the output comparisons presented in 
 this report. 
 
 Relative Efficiency of Piecework and 
 Daywork 
 
 As in the case of the other industries thus far investi- 
 gated, considerable evidence was obtained to the effect 
 that the piece-rate system was more conducive to the 
 maintenance of output with shorter hours than was the 
 day-rate system. Thus, the appendix table occasionally 
 contains such statements as the following by manufactur- 
 ers reducing to a 50-hour schedule from a longer week: 
 
 Same output for daywork; 10% increase in output for 
 piecework. 
 
 Pieceworkers speeded up and there was no loss of pro- 
 duction; dayworkers did not speed up, with consequent loss 
 of 8 to 10% in production. 
 
 Total output decreased 3 to 6% for piecework and 10% 
 for daywork. 
 
 These statements are by no means typical, since in 
 many cases the output of pieceworkers was reduced 
 practically as much as that of dayworkers. Nevertheless, 
 
 23 
 
these statements indicate rather definitely that it is more 
 practicable to maintain production under the piece-rate 
 system than under the day-rate system, when hours are 
 shortened. While the statements of manufacturers indicate 
 that in a few cases time-workers made some improvement 
 in hourly efficiency, this improvement, as a rule, was too 
 slight to offset the reduction in hours, and in some cases 
 there was a decrease in even the hourly efficiency, which 
 would mean that the reduction in total output was more 
 than proportional to the reduction in hours. 
 
 The result was that in a great majority of cases the 
 output of the mill as a whole showed a loss. 
 
 In connection with the relative efficiency of piece and 
 time work, the following statement by an executive of a 
 large mill is of interest: 
 
 When we . . . run on short time ... all the piecework 
 operatives endeavor to make the same wages in less time, 
 with the result that our hourly production is materially 
 increased. This has led to the belief that if we operated 
 short hours but paid full wages for the shorter hours, we 
 would not get increased production per hour to amount to 
 enough to make up for the increased wages. 
 
 Our operatives wish to make about a certain amount a week. 
 If we pay that amount for slack work, they have no in- 
 centive to do good work. 
 
 In estimating the influence of the piece-rate system it 
 must be remembered, however, that differences in personal 
 efficiency may result from other factors, such as age, 
 temperament, and general physical and mental fitness.' 
 
 'The Board's report on the wool industry mentioned certain reasons why 
 one weaver may make a better output showing than another. These apply 
 with equal if not greater force to silk weaving and are as follows in substance: 
 
 1. Differences in the warp. 
 
 2. One weaver may tie ends or change shuttles more quickly than another. 
 
 3. When the material is not woven properly the weaver must pick out and 
 pull the beam back, and get the proper tension — otherwise there will be 
 thick or thin places. One weaver may perform these operations more 
 quickly and more skillfully than another. 
 
 4. Variations in the complexity of the weave may be more easily adopted by 
 one weaver than another; one will seize the idea of the new weave 
 boldly, while another will lack confidence. 
 
 5. The watchful weaver catches the shuttle when there are a few more 
 winds of yarn on the bobbin, and in this way avoids the work of finding 
 the piece of thread and closing the gap; another may be too late for this. 
 
 6. Correcting defects in complex weaves requires judgment and skill. 
 Weavers show considerable difference in this respect. 
 
 24 
 
From the general evidence here reviewed it appears, 
 therefore, that in a majority of establishments reporting, 
 output was not maintained with a 50-hour weekly sched- 
 ule. Nevertheless, maintenance of output or an improve- 
 ment in hourly efficiency was reported in a sufficient 
 number of cases to fairly raise the question whether, 
 with efficient management and good quality of material, 
 such a schedule or a schedule not much in excess of 50 
 hours might not be a reasonably practicable one for the 
 industry. Before reaching final conclusions, it is desirable 
 to analyze certain statistical evidence of a more detailed 
 character obtained from several establishments during the 
 course of the investigation. 
 
 25 
 
STATISTICAL DATA AS TO OUTPUT 
 
 DIFFICULTIES OF SECURING COMPARABLE 
 
 DATA 
 
 The factors influencing efficiency of operation are 
 so numerous that the effect of any single factor, such as a 
 change in the length of the workday, can be measured in 
 only an approximate way. By its very nature the problem 
 is one which does not lend itself to refined calculations. 
 The essential thing to determine, however, so far as 
 output is concerned, is whether a given change in the 
 length of the workday resulted in a change in the total 
 production, and in case of a decrease, whether that de- 
 crease was important or negligible. For this purpose the 
 collective experience of a substantial number of estab- 
 lishments permits of conclusions which, though broad, 
 are nevertheless valid. 
 
 Some of the difficulties involved in a statistical com- 
 parison of output under different hours-of-work schedules 
 have already been set forth in previous reports on experi- 
 ence in the cotton and wool manufacturing industries and 
 in the boot and shoe industry.' Thus differences in the 
 weight, quality, and texture of the fabric all have an 
 important bearing. Again, the great variety of weaves 
 or patterns for which the silk industry is noted introduces 
 difficulties, since complicated patterns generally require 
 more attention and skill on the part of the weaver, or 
 possibly a slower speed of the loom, than do plain fabrics 
 or fabrics of comparatively simple pattern. Still again, 
 many silk establishments produce a wide variety of fabrics 
 ranging all the way from ribbons to dress goods. Unless 
 output comparisons under different schedules of hours are 
 confined to goods of substantially similar patterns, or to a 
 variety of fabrics the proportions of which are fairly con- 
 stant, the result may be misleading. 
 
 The delicacy of the silk fiber introduces a still further 
 difficulty into statistical measurements. Because of its 
 extraordinary fineness, relatively small differences in 
 
 ' National Industrial Conference Board: Research Reports Nos. 4, 7, and 12. 
 
 26 
 
sizes of yarn may have an important eflFect on the total 
 quantity of the finished fabric. The quality of the fiber 
 is equally important. Fiber of poor quality is a source of 
 frequent breakage of yarn, with consequent loss of time 
 of the operative in repairing broken ends. 
 
 Owing to differences in climatic conditions while the 
 worms are being reared, to differences in the care with 
 which the raw silk is prepared, and to various other 
 causes, the quality of silk fiber varies greatly in different 
 years. 
 
 During the war some inferior silk fiber which formerly 
 went to other countries was brought into the United 
 States. On the other hand, the quality of the fiber of the 
 1917 crop was unusually good. On the whole, American 
 silk establishments have in recent years endeavored to 
 secure a better grade of raw silk than that formerly used. 
 Over any considerable period of time these variations in 
 the quality of the raw material might be sufficient to in- 
 validate output comparisons, although they are not an 
 important factor in the comparisons presented in this 
 report, which are for periods shortly before and shortly 
 after a reduction in hours. 
 
 The difficulties noted, which by no means exhaust the 
 list of complicating factors, necessarily qualify con- 
 clusions as to the statistical evidence presented in this 
 chapter, and indeed forbid as definite conclusions as were 
 reached in the corresponding studies of output in the 
 cotton and wool manufacturing industries. Nevertheless, 
 the material here assembled gives a somewhat broader 
 basis for conclusions than that obtained from the data 
 already given, and, if used with due recognition of such 
 collateral factors as those just mentioned, is of decided 
 value. 
 
 Figures of production based on book records were 
 obtained from 10 establishments representing various 
 major establishments of the industry. 
 
 It will be noted that, in several cases, the production 
 per loom per week is given as the "indicated" production. 
 In such cases, this figure has been arrived at by first de- 
 termining the actual production per loom per hour, based 
 on the actual number of hours worked and the actual 
 number of looms in operation, and then applying this 
 average per loom per hour to the nominal number of 
 
 27 
 
schedule hours per week. This basis of comparison gives 
 an accurate idea of the efficiency of the two different 
 hours-of-work schedules, where the actual hours worked 
 show only a small variation from the nominal factory 
 hours. 
 
 Comparison of Output of a Large 
 Broad-Silk Mill 
 
 A modern mill operating about 1,000 broad looms and 
 employing approximately 600 workers furnished the 
 comparison given below. The mill is located in a large 
 industrial center. 
 
 TABLE 6: COMPARISON OFWEEKLY OUTPUT OF ABROAD-SILK 
 MILL FOR lO WEEKS UNDER A 55-HOUR SCHEDULE 
 WITH THAT FOR AN I I-WEEKS PERIOD UNDER A 5O-HOUR 
 SCHEDULE 
 
 Period 
 
 Hours per 
 week 
 
 Production: 
 per week per 
 loom: yards 
 
 1916 
 January— April, 10 weeks 
 
 55 
 50 
 
 94 9 
 
 April-June, 11 weeks 
 
 86.3 
 
 9.1% reduction in hours per week. 
 9.1% reduction in yards per loom 
 
 per week. 
 
 For the 50-hour week the loss in production as compared 
 with the 55-hour week was therefore proportional to the 
 reduction in time. The wage rates of weavers were ad- 
 vanced 20% at the time of the change in hours. There 
 were certain interruptions in the work in both periods, for 
 which, however, adjustment was made. There seems 
 little reason to doubt that the reduction in time was 
 largely responsible for the decrease in output. 
 
 The weekly production per loom at this plant for the 
 two periods was as follows: 
 
 28 
 
FIRST 
 
 PERIOD 
 
 
 SECOND PERIOD 
 
 Week Ending 
 
 Looms 
 
 Product 
 
 per Loom 
 
 (yards) 
 
 Week Ending 
 
 Looms 
 
 Product 
 
 per Loom 
 
 (yards) 
 
 Jan. 17th 
 
 962 
 
 87.7 
 
 April 10th 
 
 983 
 
 92.2 
 
 24th 
 
 964 
 
 85.3 
 
 17th 
 
 972 
 
 85.0 
 
 31st 
 
 978 
 
 98.3 
 
 24th 
 
 974 
 
 87.5 
 
 Feb. 7th 
 
 979 
 
 98.1 
 
 May 1st 
 
 976 
 
 82.1 
 
 14th 
 
 975 (a) 
 
 93.3 
 
 8th 
 
 977 
 
 92.6 
 
 21st 
 
 974 
 
 93.6 
 
 15th 
 
 958 
 
 91.0 
 
 28th 
 
 978 (i) 
 
 89.8 
 
 22d 
 
 978 
 
 84.7 
 
 
 
 
 29th 
 
 977 
 
 86.6 
 
 March 6th 
 
 969 
 
 95.7 
 
 
 
 
 13th 
 
 978 
 
 96.6 
 
 June 5th 
 
 977 (c) 
 
 69,3 
 
 20th 
 
 977 
 
 99.3 
 
 12th 
 
 976 
 
 81.6 
 
 27th 
 
 978 
 
 95.4 
 
 19th 
 
 989 
 
 87.6 
 
 
 
 
 26th 
 
 977 
 
 78.9 
 
 April 3d 
 
 979 
 
 98.5 
 
 
 
 
 Average yards 
 
 per week per 
 
 Average yards 
 
 per week per 
 
 
 loom 
 
 
 . 94.9 
 
 loom 
 
 
 86.3 
 
 (a) 5 hours lost. (b) 10 hours lost. (c) 10 hours lost. 
 
 These broken weeks are not included in the averages. 
 
 This table brings out very clearly a fact which should 
 always be kept in mind in comparing production under 
 different hours-of-work schedules, namely, that even 
 under the same weekly hours of work there will be con- 
 siderable fluctuation in the output from week to week, 
 although perhaps not so marked as shown in this instance, 
 where the mill was making a variety of fabrics. For this 
 reason it is unsafe to draw conclusions from the experi- 
 ence of a single week, although collectively even such 
 limited experience may be of value. In averaging results 
 over a period of several weeks, however, such inconsist- 
 encies tend to counterbalance each other and give a fair 
 indication of the real effect of the change in hours on 
 output. 
 
 It will be noted that there were two broken weeks in 
 the first period and one in the second period. In the 
 first period under the longer schedule this broken time 
 had much less effect on the weekly output than in the 
 second period. These broken weeks are not included in the 
 averages. 
 
 Comparison of Output of a Velvet Mill 
 
 Table 7 presents a comparison based on data furnished 
 by a large velvet mill which reduced hours from 55 to 50 
 per week in 1916. 
 
 29 
 
TABLE 7: COMPARISON OF WEEKLY OUTPUT OF A LARGE 
 VELVET MILL FOR TWO lO-WEEK PERIODS UNDER A 
 55-HOUR AND A 50-HOUR SCHEDULE RESPECTIVELY 
 
 Period 
 
 Hours per 
 week 
 
 Average weekly 
 production per 
 weaver: yards 
 
 1916 
 
 10 weeks previous to June 16 
 
 10 weeks after June 16 
 
 55 
 
 50 
 
 306.0 
 280.7 
 
 
 
 In this case a reduction of 9.1% in time was accom- 
 panied by a loss of 8% in output per weaver. A com- 
 parison for the same number of looms showed a decrease 
 in the weekly production per loom of 5.6%. 
 
 The management stated that the mill was operated to 
 full capacity in each period on a product of uniform 
 character and that conditions were unusually favorable 
 for a fair comparison. There is, therefore, little doubt 
 that the reduction in hours was chiefly responsible for 
 the decrease in output. 
 
 In the manufacture of velvets quality is a prime con- 
 sideration and speed of machinery cannot easily be in- 
 creased. For the same reason it is less practicable to 
 increase the speed of handwork. Therefore, the element 
 of machine time is the dominant factor, with the result 
 that the loss in output was very nearly proportional to 
 the reduction in hours. 
 
 In a prior reduction from 60 to 65 hours per week in 
 this plant in 1915, by the adoption of a Saturday half 
 holiday, a loss of 16% in production was shown. This 
 loss is so marked as to indicate that some other factor 
 than the shortening of hours was also involved. 
 
 In the reduction from 60 to 55 hours per week as well 
 as that from 55 to 50 hours, there was a slightly greater 
 output per weaver per hour during the first five weeks 
 than during the second five weeks after the change in 
 hours. This indicates a temporary increase in efiiciency 
 due to speeding up directly after the reduction in hours 
 and a subsequent relapse into a slower settled speed. 
 Similar experience, it may be noted, was reported by a 
 number of establishments in other industries inves- 
 tigated as well as in the silk industry. 
 
 30 
 
Comparison of Output of a Broad-Silk Mill 
 
 A broad-silk mill which reduced hours in 1916 from 55 
 to 50 per week submitted the comparison given in Table 8 
 for a period of 6 weeks under a 55-hour schedule and 
 for a 6-week period under a 50-hour schedule. The 
 management stated that conditions were on the whole 
 such as to permit of a fairly accurate comparison. 
 
 The mill is located in an important silk manufacturing 
 center and operates both ordinary broad-silk looms and 
 Jacquards. 
 
 TABLE 8: comparison of weekly output of a broad- 
 silk MILL FOR A 6-WEEK PERIOD UNDER A 55-HOUR 
 SCHEDULE WITH THAT FOR A 6-WEEK PERIOD UNDER A 
 50-HOUR SCHEDULE 
 
 Period 
 
 1916 
 
 Two weeks ending February 23 . . 
 
 " March 8 . . . 
 
 " March 22 . . 
 
 Average 
 
 April 19 
 May 3 . 
 May 17 
 
 Average 
 
 Hours 
 
 Number 
 
 per week 
 
 of looms 
 
 55 
 
 106 
 
 55 
 
 105 
 
 55 
 
 107 
 
 50 
 
 103 
 
 50 
 
 102 
 
 50 
 
 112 
 
 Actual 
 
 yards per 
 
 hour per 
 
 loom 
 
 1.015 
 1.091 
 1.088 
 1.065 
 
 1.169 
 1.100 
 1.161 
 1.143 
 
 9.1% reduction in hours per week. Loom hours, first period, 33,390. 
 2.4% indicated loss in yards per loom per week. Loom hours, second period, 
 31.700. 
 
 This shows that reduction of 9.1% in working time was 
 accompanied by an increase of 7.3% in hourly output. 
 While this was not sufficient to offset the reduction in 
 hours, the indicated loss in total output was only 2.4%. 
 
 Ordinarily interruptions resulting from colds and other 
 sickness are more common in February and March than 
 in April and May. It is possible that the gain in hourly 
 efficiency shown for the 60-hour period may have been in 
 part due to the more favorable weather. As stated, how- 
 ever, it is the opinion of the management that the com- 
 parison gives a fair representation of the output effects of 
 shortening the work-week. 
 
 31 
 
Comparison of Output of a Broad-Silk Mill 
 
 The comparison given in Table 9 is for a large broad- 
 silk mill located in a silk manufacturing center. The 
 product is mostly fancy tie silks varying from 76 to 180 
 picks per inch; there was also considerable variation in the 
 width of the goods produced. The output of the periods 
 given below were regarded by the management as being, 
 on the whole, fairly comparable. 
 
 table 9: COMPARISON OF WEEKLY OUTPUT OF A BROAD- 
 SILK MILL FOR THREE MONTHS UNDER A 55-HOUR 
 SCHEDULE WITH THAT FOR THREE MONTHS UNDER A 50- 
 HOUR SCHEDULE 
 
 Period 
 
 Hours per 
 week 
 
 Number of 
 looms 
 
 Indicated 
 
 production 
 
 per week 
 
 per loom: yards 
 
 1916 
 January 
 February 
 March . . 
 
 55 
 55 
 55 
 
 SO 
 50 
 50 
 
 180 
 180 
 196 
 
 220 
 224 
 224 
 
 71.88 
 76.12 
 73.48 
 
 Average 
 
 April . . 
 
 May . .... 
 
 June 
 
 73.83 
 
 63.50 
 71.90 
 68.35 
 
 Average 
 
 67.92 
 
 9.1% indicated reduction in time 
 8 . 0% reduction in output 
 
 The loss in production in this case was nearly propor- 
 tional to the reduction in time. There were several 
 collateral factors which qualify the statistical data. 
 Thus good weavers were scarce at the time hours were 
 reduced and efficiency suffered somewhat during April, 
 May, and June; moreover, the attendance of the weavers 
 was not quite as regular during those months. Again, 
 rates of wages were advanced 10%, and this may have 
 influenced production somewhat. In the opinion of the 
 management, these collateral factors had only a minor 
 bearing and the greater part of the decrease in production 
 was attributed to the shortening of the workday. 
 
 32 
 
Comparison of Output of a Broad-Silk Mill 
 
 _ The comparison given in Table 10 is for a large broad- 
 silk mill which reduced hours from 55 to 50 per week in 
 1916. It is located in a silk center and produces chiefly 
 dress and tie silks. 
 
 TABLE ID : COMPARISON OF WEEKLY OUTPUT OF A BROAD- 
 SILK MILL FOR TWO MONTHS UNDER A 55-HOUR SCHEDULE 
 AND FOR THREE MONTHS UNDER A 5O-HOUR SCHEDULE 
 
 Period 
 
 Hours per 
 week 
 
 Looms 
 
 Indicated 
 
 production 
 
 per week per 
 
 loom: yards 
 
 1916 
 February . . 
 March . . . 
 
 Average 
 
 May . . . 
 
 June 
 
 July 
 
 Average 
 
 55 
 55 
 
 50 
 60 
 50 
 
 265 
 270 
 
 278 
 277 
 283 
 
 61.05 
 59.40 
 
 60.23 
 
 60.90 
 52.90 
 60.20 
 
 58.00 
 
 The 9.1% reduction in working time at this mill was 
 accompanied by a 3.7% loss in output. It will be noted 
 that there was a marked increase in hourly efficiency, 
 although not sufficient to offset the shortening of hours. 
 However, omitting the figures for June, which obviously 
 are not fully comparable, the weekly production showed 
 a slight increase under the shorter schedule. 
 
 The preceding comparison is for the output of the 
 entire mill. For a portion of the looms running on the 
 same quality of goods the comparison was distinctly in 
 favor of the 50-hour week, as the following table shows: 
 
 Production: Same Quality of Goods 
 
 Period 
 
 Hours per 
 week 
 
 Looms 
 
 Actual 
 
 production 
 
 per loom 
 
 hour: 
 
 yards 
 
 Indicated 
 production 
 per loom 
 per week: 
 yards 
 
 1916 
 
 March 
 
 April .... 
 
 June .... ... 
 
 55 
 50 
 50 
 
 27 
 34 
 47 
 
 .861 
 
 .956 
 
 1.020 
 
 47.36 
 47.80 
 51.00 
 
 33 
 
For these goods, therefore, the June production per loom 
 under a 50-hour schedule showed a gain of 7.7% over 
 that in March with a 55-hour schedule. Allowing for the 
 fact that efficiency of operation is usually greater in 
 periods of full demand than when work is running light 
 the comparison still suggests a decided gain in efficiency 
 under the shorter schedule. The increase in hourly effi- 
 ciency (18M%) was, however, so great as to leave no 
 doubt that other factors than the shortening of the 
 workday contributed to the result. 
 
 Comparison of Output of a Throwing Mill 
 
 The throwing department of a mill which reduced its 
 working time in 1917 from 54 to 50 hours per week fur- 
 nished the following comparison: 
 
 TABLE II : comparison OF WEEKLY OUTPUT OF A THROW- 
 ING MILL FOR SIX MONTHS UNDER A 54-HOUR SCHEDULE 
 WITH THAT FOR A SIMILAR PERIOD UNDER A 5O-HOUR 
 SCHEDULE 
 
 Period 
 
 Hours per 
 week 
 
 Average 
 
 monthly 
 
 amount of 
 
 silk thrown: 
 
 pounds 
 
 Average 
 
 per throwing 
 
 employee: 
 
 pounds 
 
 1917 
 Six months previous to May . . . 
 Six months after May 
 
 54 
 50 
 
 4,800 
 4,965 
 
 120.0 
 110.3 
 
 In this case the reduction of 7.4% in work-hours was 
 accompanied by a loss of 8.1% in output. As already 
 explained, the throwing process is one in which machinery 
 dominates to an unusual degree, making it correspondingly 
 difficult for the operative to offset a reduction in work- 
 hours. 
 
 This mill is located in an industrial center, although little 
 silk manufacturing is done in its vicinity. An executive 
 stated that its schedule of hours was reduced to 50 per 
 week in order to prevent the establishment's best opera- 
 tives from going into other lines of work where hours 
 were below 54 per week. In the opinion of the manage- 
 ment the comparison given above fairly shows the results 
 of the reduction in hours in this department. 
 
 34 
 
A marked increase in average weekly production in this 
 establishment which occurred between 1912 and 1917 was 
 reported as having been largely due to a change from day- 
 work to piecework and to the installation of improved 
 machinery. 
 
 Other Comparisons 
 
 Several other statistical comparisons were secured in 
 the course of the investigation. 
 
 A comparison for a weaving mill making broad silks, 
 located near an industrial center, which changed its work- 
 week from 55 hours to 50 hours in April, 1916, is given 
 below; all looms were run at the same speed, each weaver 
 operating three looms. 
 
 Period 
 
 Hours per 
 week 
 
 Number of 
 looms 
 
 Production 
 per week per 
 loom: yards 
 
 1916 
 March 14r-28 
 
 55 
 50 
 
 110 
 122 
 
 162.05 
 
 April 11-25 
 
 156.31 
 
 This shows a 3.5% decrease in output per loom per 
 week accompanying a reduction of 9.1% in time. There 
 was, therefore, a marked increase in hourly efficiency. 
 
 A comparison of output of a broad-silk mill for one week 
 in March, 1916, under a 55-hour schedule with that for a 
 week in April of the same year under a 50-hour schedule 
 showed a loss of only 1.2% in production per loom; the 
 reduction in time was 9.1%. A week is too short a period 
 to warrant final conclusions. In this case, however, it 
 was specifically stated by the management that the weeks 
 were representative and the opinion was furthermore 
 expressed that under the shorter schedule there was a 
 noticeable increase in personal efficiency of the workers. 
 It may be noted that it was the opinion of the nianagement 
 that, in this general section, arbitrary restrictions of out- 
 put were commonly practised. 
 
 Another mill making broad silks and ribbons in April, 
 1916, reduced hours from 55 to 50 per week with an im- 
 mediate output loss of 5%. A later comparison showed 
 
 35 
 
a 73^% loss. The management stated that the reduction 
 in working time was unquestionably the cause of the 
 marked decrease in output, though there was some in- 
 crease in hourly efficiency. 
 
 The statistical evidence thus gathered is summarized 
 in the following table. The reductions in hours were made 
 in 1916, with one exception: 
 
 TABLE 12: SUMMARY OF OUTPUT DATA COVERING REDUC- 
 TIONS IN TEN ESTABLISHMENTS FROM A 55 TO A 5O-HOUR 
 WEEK 
 
 Kind o£ Mill 
 
 Percent 
 
 decrease in 
 
 weekly output 
 
 per loom 
 
 Change in 
 
 hourly 
 efficiency 
 
 Weaving: Broad .... . . 
 
 Weaving: Broad and Jacquards 
 
 Weaving: Broad 
 
 Weaving: Broad, all looms . . 
 
 Looms on one class of goods 
 
 Weaving: Broad and narrow 
 
 Weaving: Broad . . 
 Weaving: Velvet (a) . . 
 Weaving: Broad 
 Throwing (i) . 
 
 1.2% 
 2.4% 
 3.5% 
 3.7% 
 *7.7% 
 
 8.0% 
 8.0% 
 9.1% 
 8.1% 
 
 Increase 
 
 Increase 
 
 Increase 
 
 Increase 
 
 Marked increase 
 
 Increase 
 
 Increase 
 
 Increase 
 
 Unchanged 
 
 Decrease 
 
 Where hourly efficiency is merely maintained the total weekly output is 
 decreased about in proportion to the reduction in hours. 
 
 ♦Increase in italics. 
 
 (fl) Per weaver. 
 
 (&) Per throwing operative. In this mill previous hours were 54 per week, making the 
 percentage of lost time 7.4 instead of 9.1, as in the other mills of this group. The reduc- 
 tion was made in 1Q17. 
 
 These various comparisons are on their face conflicting. 
 Nevertheless, two facts are definitely established. First, 
 there was in a majority of cases some loss in output 
 and, in processes where machinery dominated to an un- 
 usual degree, a practically proportional loss. Second, in a 
 majority of instances there was some increase in hourly 
 efficiency and in certain cases a decided increase. 
 
 The results of this statistical analysis are, therefore, in 
 substantial harmony with those presented in the preceding 
 chapter, based on evidence of a more general character. 
 
 36 
 
CONCLUSIONS AS TO OUTPUT 
 
 The experience with a 50-hour week in the silk industry 
 presented in the preceding discussion shows clearly that 
 such a schedule has, in a great majority of instances, been 
 accompanied by a loss in output. For 58 establishments 
 reporting results secured with such a work-week, only 9 
 maintained the production previously secured with longer 
 schedules; in 2 of these 9 establishments there was an 
 increase. On the other hand, 23 establishments, employ- 
 ing more than one-half of the total workers included in 
 the 50-hour group, reported a loss in output somewhat 
 less than proportional to the reduction in hours. Only 
 one establishment reported a loss more than proportional 
 to the reduction in time. In 19 establishments, or about 
 one-third the total number, the loss was reported as about 
 proportional. These facts are presented in the following 
 summary for the 50-hour group : 
 
 SO-Hour Group 
 
 Establishments reporting output increased . . . 
 Establishments reporting output maintained . . . 
 Establishments reporting a decrease in output less 
 
 than proportional to the reduction in hours . . 23 52.0 
 
 Establishments reporting a decrease in output 
 
 about proportional to the reduction in hours . . 19 21.0 
 
 Establishment reporting a decrease in output 
 
 greater than proportional to the reduction in 
 
 hours 1 1.8 
 
 Establishments reporting a decrease in output but 
 
 not stating its amount 6 8.6 
 
 Establish- 
 ments 
 
 Percent of Total 
 Employees 
 in Group 
 
 2 
 
 1.6 
 
 7 
 
 15.0 
 
 58 100.0 
 
 This statistical evidence, in connection with the state- 
 ments of employers given in the text and in the appendix, 
 indicates that the point of maximum output in the silk 
 manufacturing industry as a whole lies somewhere between 
 50 and 54 hours per week. Because of the limited experi- 
 ence available with intermediate work schedules it is im- 
 possible to fix that point definitely. The point of maxi- 
 mum production, moreover, is not the same for all classes 
 of mills. In processes where machinery dominates 
 to an exceptional degree as, for instance, in throwing or in 
 certain kinds of weaving, reduction of work-hours to 
 even 54 per week has usually involved some loss in output. 
 
 37 
 
But, taking the industry as a whole, it may fairly be con- 
 cluded that the point of maximum production is somewhat 
 nearer 50 hours than 54 hours per week. 
 
 This conclusion is in sharp contrast with that reached 
 in a similar study of the cotton manufacturing and wool 
 manufacturing industries. In the latter, output was not 
 maintained on a schedule of 54 hours per week; while in 
 the cotton industry, reductions to even a 56-hour week 
 almost invariably involved a loss in production. The 
 reason for this difference in results is largely to be found 
 in the much greater importance of skilled handwork in 
 the manufacture of silk goods, which affords a correspond- 
 ingly greater opportunity for the worker by his own effort 
 to influence efficiency of production. 
 
 38 
 
HEALTH OF SILK-MILL WORKERS 
 
 The same difficulties that were experienced in attempts 
 to determine the relation between hours of work and 
 health of workers in the cotton and wool manufacturing 
 industries were also encountered in the study of the silk 
 industry. Practically no records covering a sufficient 
 period to permit of satisfactory study were available, 
 while the absence of any definitely recognized standard 
 for the measurement of health conditions, and above all 
 the great number of factors entering into the problem, 
 made it difficult to draw conclusions from such general 
 statements as were secured.' It is, for instance, obvious 
 that such influences as nationality, inheritance, personal 
 hygiene and habits have a very important bearing on 
 health, so that even though a change in hours of work may 
 have a significant effect, this may be overshadowed by the 
 collective effect of other factors. 
 
 Six establishments reported an appreciable improve- 
 ment in health of workers with a 50-hour schedule, as 
 compared with conditions obtaining under a longer week. 
 In one of these a marked lessening of fatigue was reported. 
 A large majority of the establishments covered by the 
 investigation, however, reported that they observed no 
 noticeable change in health conditions when hours were 
 shortened to 50, 52, or 54 hours per week from longer 
 schedules. 
 
 A representative of a large silk manufacturing estab- 
 lishment conducting practically all branches of the in- 
 dustry stated: 
 
 The results of our Benefit Society for the periods covered 
 by days of different working hours appear to show an improve- 
 ment in health with shorter hours. We doubt very much 
 whether this can be so simply disposed of, as there are many 
 other factors that enter into it. The most important is that 
 in the beginning the Benefit Society took into membership 
 anybody regardless of his physical condition, but the fifty- 
 hour period has been upon the basis of a physical examination 
 and all undesirable risks have been rejected; also during both 
 the fifty-five and fifty-hour period, there has been medical 
 
 ' Some of these factors are noted in Research Report No. 4, Hours of Work 
 as Related to Output and Health of Workers — Cotton Manufacturing. 
 
 39 
 
supervision, free medical advice and treatment for all people 
 who are at work. Therefore, it probably will be unsafe to 
 draw any conclusions from these figures as to the effect of 
 working hours upon the health. 
 
 General factory conditions in the silk industry compare 
 favorably with those in other leading branches of the 
 textile industries. The value and delicacy of the product 
 make cleanliness almost imperative. Good lighting, like- 
 wise, is essential. There is, furthermore, comparative 
 freedom from dust, which is a source of considerable 
 trouble in both the cotton and wool industries. Thus, the 
 report of the United States Bureau of Labor Statistics on 
 Women and Children Wage-Earners in the United States, 
 said:' 
 
 A silk mill is, as a rule, clean. There is no dust or lint or 
 fumes in the air, so the need for special means of ventilation 
 is not so imperative as in other textile mills. . . . The 
 manufacture of silk requires that it should be conducted in 
 well-lighted and clean rooms. Silk thread must be watched 
 carefully as it is being wound or spun or woven in order to 
 detect broken or knotted threads. The best lighting facilities 
 are therefore necessary and are in a great majority of cases 
 provided. The mills are constructed with the plan of provid- 
 ing the greatest possible amount of natural light. 
 
 On the other hand the report just cited contained con- 
 siderable criticism of improper ventilation. This was 
 attributed, not to lack of sufficient air space, but to failure 
 to keep a sufficient number of windows open. This 
 criticism applied to conditions as they existed several 
 years ago. Faulty ventilation was also observed in a few 
 of the mills visited by representatives of the Conference 
 Board in the course of its investigation. In the larger 
 establishments, however, ventilation was not only good 
 but often almost ideal. In this connection it should be 
 pointed out that many small silk establishments con- 
 ducting a few specialized processes rent space in a 
 building where power may be purchased. In such cases, 
 there is more likeHhood of overcrowding than in larger 
 establishments conducting more processes of manufacture. 
 
 General Health Conditions in the Industry 
 
 With the exception of tuberculosis, the silk industry 
 makes a favorable comparison with other leading textile 
 industries from the standpoint of health. Statements of 
 
 'U.S. Senate. 61st Congress, 2d Session, Document 645. 
 
 40 
 
manufacturers furnished In this investigation indicated 
 generally favorable conditions of health among their 
 employees. Frequently these conditions were reported as 
 excellent and almost never as distinctly unfavorable. 
 
 The silk industry is, however, characterized by a dis- 
 tinctly high death rate from tuberculosis. This is clearly 
 brought out in Table 13 compiled from data collected in 
 the Census of 1909. 
 
 TABLE 1 3 : PERCENTAGE OF DEATHS OF MALES AND FEMALES 
 DUE TO TUBERCULOSIS, OF THE TOTAL MORTALITY FOR 
 THE SILK, WOOL, AND COTTON INDUSTRIES AND FOR ALL 
 MANUFACTURING AND MECHANICAL PURSUITS, BY AGE 
 
 GROUPS 
 
 (U. S. Census Mortality Statistics, 1909) 
 
 
 Males 
 
 Females 
 
 Age Group 
 
 SiUc 
 
 Wool 
 
 Cotton 
 
 AUMtg. 
 
 and 
 Mecb. 
 Pureulttf 
 
 Silk 
 
 Wool 
 
 Cotton 
 
 AUMJJ. 
 
 and 
 
 Mecb. 
 
 Piiraulti 
 
 All ages 10 and over 
 
 19.8 
 
 27.7 
 50.0 
 29.8 
 28.5 
 13.3 
 7.0 
 
 22.3 
 
 30.8 
 44.8 
 41.7 
 35.3 
 16.7 
 13.6 
 
 21.1 
 
 19.6 
 48.8 
 37.8 
 27.1 
 19.4 
 9.4 
 
 15.5 
 
 22.1 
 32.5 
 30.8 
 25.0 
 15.6 
 8.2 
 
 37.7 
 
 38.8 
 53.4 
 46.7 
 30.0 
 22.2 
 20.0 
 
 29.3 
 
 36.8 
 51.6 
 42.3 
 20.0 
 11.8 
 5.9 
 
 29.7 
 
 38.2 
 50.9 
 35.6 
 32.2 
 14.9 
 9.1 
 
 27.4 
 
 15 to 19 . . . 
 
 20 to 24 
 
 25 to 34 
 
 35to44 . . . . 
 45 to 54 . . 
 55 to 64 . . 
 
 37.9 
 44.4 
 40.8 
 26.3 
 12.3 
 9.3 
 
 As compared with the average mortality for all manu- 
 facturing and mechanical pursuits, this table shows a 
 distinctly high death rate from tuberculosis for silk-mill 
 operatives. Furthermore, it is higher for female workers, 
 who constitute considerably more than half the total 
 work-force of the average silk mill, than for males.' 
 
 A similarly high death rate is shown by more limited 
 data based on the industrial experience of the Prudential 
 Insurance Company during the period 1897 to 1900 
 inclusive.* In these four years there were 142 deaths of 
 silk workers among males included in the industrial 
 experience of the above named company. The following 
 
 ' It should be borne in mind that the census data give only the occupations 
 followed at the time of death. It does not necessarily follow that the occupation 
 reported for the worker at the time of death was followed by him during the 
 greater part of his working life, or that the disease causing death was contracted 
 while he was engaged in his last occupation. 
 
 ^This Prudential Insurance Company material was furnished through the 
 courtesy of Dr. Frederick L. Hoffman, Third Vice President and Statistician. 
 
 41 
 
table shows the percentage distribution of these deaths hy 
 principal causes, and also gives corresponding percentages 
 for the same causes for 79 selected occupations. 
 
 TABLE 14: PERCENTAGES OF DEATHS OF MALE SILK-WORK- 
 ERS FROM CERTAIN CAUSES, 1 897 TO I9OO, COMPARED 
 WITH SIMILAR PERCENTAGES FOR 79 SELECTED OCCUPA- 
 TIONS, AS SHOWN BY THE INDUSTRIAL EXPERIENCE OF 
 THE PRUDENTIAL INSURANCE COMPANY 
 
 
 Silk Workers 
 
 79. 
 Occupations 
 
 Causes 
 
 Number of 
 Deaths 
 
 Percent 
 of Total 
 
 Percent 
 of Total 
 
 Total 
 
 Typhoid 
 
 Alcoholism 
 
 Tubercular Diseases 
 
 Nervous Diseases 
 
 Circulatory Diseases 
 
 Respiratory Diseases 
 
 Liver Diseases 
 
 Other Digestive Diseases 
 
 Urinary Diseases 
 
 Accidents 
 
 Suicide 
 
 142 
 
 11 
 1 
 
 55 
 7 
 
 10 
 
 17 
 4 
 4 
 
 12 
 6 
 2 
 
 13 
 
 100.0 
 
 7.8 
 0.7 
 
 38.8 
 4.9 
 7.0 
 
 12.0 
 2.8 
 2.8 
 8.5 
 4.2 
 1.4 
 9.1 
 
 100.0 
 2.7 
 0.9 
 
 23.0 
 
 11.4 
 9.4 
 
 14.9 
 2.7 
 4.9 
 
 10.1 
 8.0 
 1.9 
 
 10.1 
 
 
 
 While for several of the causes listed in the above table 
 the number of deaths is so small that the percentages 
 must be used with great caution, the comparison as 
 a whole brings out strikingly the decided preponderance 
 of tubercular and respiratory diseases as causes of death 
 among this group of male silk-workers. Thus, nearly 
 39% of the total was attributed to tubercular diseases 
 and an additional 12% to respiratory diseases. The per- 
 centage of deaths from tubercular diseases, it will be 
 noted, was distinctly higher than the average for the 
 79 selected occupations. 
 
 With the exception of tubercular diseases and typhoid 
 fever, for which the high death rate among these silk 
 workers apparently reflected local conditions in Paterson, 
 the percentages of deaths due to the specific causes men- 
 tioned were as a rule lower for silk workers, and frequently 
 decidedly lower, than the corresponding averages for the 
 79 selected occupations. 
 
 42 
 
It may be considered as established that the death rate 
 from tuberculosis among silk-mill operatives is distinctly 
 high as compared with the average rate for factory indus- 
 tries in general. There is a strong presumption, moreover, 
 that conditions in the industry itself are partly responsible, 
 although these conditions have not been identified. What- 
 ever the effect of factors directly involved in manufac- 
 turing processes, it is certain that various collateral 
 factors, such as nationality, personal habits, and living 
 conditions, have an important influence. 
 
 Whatever the cause of this high mortality from tuber- 
 culosis among silk workers may be, it calls for most serious 
 study in order to locate the responsibility and, so far as 
 possible, remove or minimize the cause. 
 
 In view of the rather general disposition among writers 
 on industrial hygiene to attribute a high occurrence of 
 tuberculosis to factory dust, it is worth emphasizing the 
 fact, already pointed out, that the silk manufacturing 
 industry is exceptionally free from dust hazards, except 
 perhaps in the plush and velvet mills. Yet, on the whole, 
 the tuberculosis death rate among silk-mill operatives is 
 rather higher than that for workers in cotton and wool 
 mills, where dust is characteristic of certain processes. 
 
 Special studies of health conditions in the silk industry 
 in Italy, under the direction of the Department of Industry 
 and Labor, show a similar high mortality from tubercu- 
 losis, particularly for the younger age groups. This has 
 been attributed by one writer' to the fact that children go 
 to work at so early an age in the silk industry in Italy that 
 they are unable to stand the strain of the work and the 
 unsatisfactory ventilation in the factories. This, the 
 report stated, tends to bring on an anaemic condition 
 which later results in tuberculosis. 
 
 In the United States, the laws of most of the states in 
 which the silk industry is conducted aim to protect chil- 
 dren from working while too young, and also prescribe 
 their hours of work. It should be remembered, however, 
 that the proportion of young persons working in the silk 
 industry is higher than in most factory industries, and 
 higher than in the other textile industries. It may be 
 that this has a bearing on the high incidence of tuberculosis 
 among silk workers later in life.' 
 
 *Dott. Luigi Carozzi. " Contribute alio Studio delle Condizione Igienico- 
 Sanitarie nella Trattura della Seta." Bulletin of the Department of Industry 
 and Labor. Anno IV.N. 7-8 Rome. 1913. 
 
 43 
 
Accident Ha2ard 
 
 The accident rate in the silk industry is distinctly low. 
 Thus, in the Pennsylvania Rate Manual,' the silk manu- 
 facturing industry takes a rate of 25 cents per $100 of 
 payroll as against 55 cents for cotton and wool spinning 
 and weaving, and $1.05 for wool combing, and much higher 
 rates in many other industries. 
 
 Conclusions as to Health 
 
 The evidence necessary for satisfactory conclusions as 
 to the health of silk-mill workers under different hours-of- 
 work schedules is not in existence. As already noted, 
 there was the same absence of records as in the cotton and 
 wool industries. No morbidity statistics are gathered by 
 the Government, and even the Census mortality statistics 
 by occupation are limited. Special effort was made to 
 see whether information was contained in literature on 
 the subject which would permit of valid conclusions. 
 Here, again, the same paucity of information was de- 
 veloped.^ 
 
 There is no doubt that the mortality from tuberculosis 
 among silk-mill workers is high; it is somewhat higher 
 
 ' The following table shows the rates for various industries as quoted in 
 the Pennsylvania Rate Manual. 
 
 SILK MANUFACTURING 
 
 $ .25 
 
 
 .40 
 
 
 .45 
 
 Linotype and hand composition 
 
 .50 
 
 Printing 
 
 Cotton thread .... 
 
 Cotton spinning and weaving 
 
 Wool spinning and weaving , . 
 
 Shoe findings manufacturing 
 
 Electric apparatus mantifacturing 
 
 .50 
 .55 
 .55 
 .55 
 .60 
 .75 
 
 Shoe stock manufacturing 
 
 Textile dyeing, finlahln)! and printing 
 
 Rubber tire 
 
 .80 
 .85 
 .90 
 
 Bleacheries 
 
 Wool pulling 
 
 1.00 
 1.05 
 
 Wool scouring 
 
 Wool combing . 
 
 1.05 
 1.05 
 
 Machine shop, without foundry ... 
 
 1.05 
 
 Blacksmithing and shoeing 
 
 1.30 
 
 Paper manufacturing except writing, bond and ledger 
 
 Can manufacturing 
 
 Iron foundry 
 
 Steel castings 
 
 Rolling mills 
 
 1.35 
 1.50 
 1.60 
 1.60 
 1.75 
 
 Boiler making 
 
 2.05 
 
 Pulp manufacturing, sulphur process, no saw or barking mills 
 Wool separating (chemical) 
 
 2.75 
 3.00 
 
 
 3.50 
 
 Saw and barking mills 
 
 3.75 
 
 'Thus, in Diseases of Occupation and Vocational Hygiene, by Kober and 
 Hanson, covering nearly 900 pages, the only reference to the silk industry 
 noted in the index is a brief reference to the use of chemicals in the manufacture 
 of artificial silk. 
 
 44 
 
than the rate for the wool and cotton manufacturing in- 
 dustries which, in turn, show a higher rate than the 
 general average for all factory industries combined. 
 
 With the exception of tuberculosis, however, health 
 conditions among silk-mill workers, so far as these can be 
 judged by mortality statistics, compare favorably with 
 those in other factory industries. 
 
 The National Industrial Conference Board, in the sum- 
 mer of 1918, secured the co-operation of a substantial 
 number of establishments for the keeping of careful 
 records of absenteeism in industry over a 24-day period 
 which would show not only the total amount of time lost 
 but also the cause. It is hoped to present a report on the 
 results of this study in the near future. In the meantime 
 a similar record is being kept for another 24-day period 
 which will reflect conditions at the end of the winter. It 
 should be understood that these records will throw little 
 light on the effect of hours of work on health. Probably 
 in most cases there has been no change in work-hours 
 between the periods for which these absence records were 
 kept. In any case it would be necessary to have records 
 covering a much longer time before any safe conclusions 
 could be drawn as to the effect of a single factor such as 
 a change in hours of work. 
 
 It is iioped, however, that the keeping of these records 
 for specific periods may lead to keeping permanent 
 records by these establishments, and may encourage 
 other employers to adopt a similar policy. Until such 
 careful records of the amount and causes of absenteeism 
 are kept by a large number of industrial establishments, 
 it will be practically impossible to present valid con- 
 clusions as to the effect of hours of work or any other 
 single factor on the health of industrial workers. 
 
 45 
 
Appendix 
 
 The following table epitomizes the important data 
 submitted by those establishments which reported the 
 eiFects on output accompanying reduction in work-hours. 
 For convenience, the information is assembled by hour- 
 groups and the establishments classified according to the 
 effect on production. 
 
 In examining the data given in the column "Change 
 in Rates of Wages," the distinction between rates of 
 wages and total wages must be kept clearly in mind. 
 The statement that there was no change in piece or hourly 
 rates when hours were shortened means that weekly 
 earnings were reduced for dayworkers, and for piece- 
 workers also unless these increased their output. Likewise 
 an increase in efficiency as shown by an increased hourly 
 output does not necessarily mean that total output was 
 maintained. 
 
 47 
 
REDUCED TO 50 HOURS 
 Production Maintained 
 
 Establishments 9 
 
 Employees 3,766 
 
 Previous 
 H<nirs 
 
 55 
 
 55 
 
 55 
 
 Number 
 Employees 
 
 209 
 
 150 
 
 114 
 
 55 356 
 
 55 268 
 
 55a 1,713 
 
 55 106 
 
 55 582 
 
 54 268 
 
 Product Change in Rates 
 
 of Wages 
 
 TWN Piece rates increased 
 
 10%; increase of 
 11>^% in time 
 rates. 
 
 M Piece rates increased 
 
 proportionally. 
 No change in time 
 wages. 
 
 W Piece rates increased 
 
 10 to 15%; 10% 
 increase in hourly 
 rates. 
 
 W Piece and day wages 
 
 increased 10%. 
 
 W N Piece and hourly 
 
 rates increased 
 10%. 
 
 W V* Rates increased 
 
 proportionally. 
 
 W T P Piece rates increased 
 
 5%; hourly rates 
 increased propor- 
 tionally. 
 
 W Piece and day rates 
 
 increased 10%. 
 
 W Piece rates increased 
 
 10%. Same rate 
 for time-workers. 
 
 Statement of Establishment 
 
 Total output per employee in- 
 creased for both piece and 
 daywork. 
 
 Same output for daywork; 
 10% increase in output for 
 piecework. 
 
 At start 2% decrease in output 
 — later on none. 
 
 Practically no 
 total output. 
 
 reduction in 
 
 Total output per employee re- 
 mained the same as under 
 the 55-hour schedule. 
 
 No effect on output directly 
 traceable to reduction in 
 hours. No figures available. 
 No marked change. 
 
 No loss in production or in 
 earnings of employees. 
 
 Practically no effect on output. 
 
 No change in total output per 
 employee or in cost. Better 
 attitude. 
 
 55 
 
 Production Decreased 
 
 Establishments 49 
 
 Employees 18,951 
 
 Previous 
 
 Number 
 
 Product 
 
 Change in Rates 
 
 Statement of Establishment 
 
 Hours 
 
 Employees 
 
 
 of Wages 
 
 
 55 
 
 271 
 
 WN 
 
 Piece rates increased 
 10%; same weekly 
 wages for day- 
 workers. 
 
 Output for piece andj|time 
 workers reduced 5%. Tried 
 to speed up and succeeded in 
 some departments. 
 
 600 
 
 Piece rates increased 
 15 to 25%. Hourly 
 rates unchanged. 
 
 Decrease of 10% in total out- 
 put. 
 
 W designates broad weaving. 
 N '* narrow weaving. 
 
 WV " velvets. 
 T " throwing. 
 
 * " dyemg. 
 
 G designates general mill. 
 
 P " _ printing. 
 
 M This mill does winding, warping, and quilling. 
 
 a This mill reduced to 50}j hours. 
 
 48 
 
Previous Number Product 
 Hours Employees 
 
 55 170 W 
 
 55 155 
 
 55 61 
 
 55 37 
 
 55 500 
 
 55 150 
 
 W 
 
 55 4,582 G 
 55 175 W T 
 
 W 
 
 W 
 
 W 
 
 55 400 N 
 
 W 
 
 55 494 W 
 
 65 919 
 
 55 308 WV* 
 
 Change in Rates 
 of Wages 
 
 Increase of 10% in 
 piece and hourly 
 rates. 
 
 Piece rates in- 
 creased 10%. Same 
 weekly wage for 
 daywork. 
 
 Rates increased 
 proportionally. 
 
 Pieceworkers in- 
 creasedlO%. Same 
 weekly wage for 
 time-workers. 
 
 Increase of 10 to 
 15% in all depart- 
 ments. Bonus to 
 weavers. 
 
 Piece rates in- 
 creased 5 to 10%. 
 Same weekly wage 
 for time-workers. 
 
 55 
 
 186 
 
 W 
 
 Piecework and day 
 wages raised 10%. 
 
 55 
 
 54 
 
 W 
 
 Increase of 10% in 
 piece and hourly 
 rates. 
 
 55 
 
 276 
 
 W* 
 
 Wages increased. 
 
 55 
 
 721 
 
 W 
 
 Increase of about 
 10%. 
 
 Weavers' rates in- 
 creased about20%. 
 
 Piece rates increased 
 15-25%; hourly 
 rates unchanged. 
 
 Piece and hourly 
 rates increased 
 10%. 
 
 Statement of Establishment 
 Decrease of 10% in production. 
 
 Loss of 5% in total output. 
 
 Slight loss in production; gain 
 in efficiency. 
 
 A reduction of 10% in produc- 
 tion. 
 
 Production decreased 10%. 
 
 Less production. Employees 
 more contented but do not 
 work any harder. 
 
 Production curtailed 8%. 
 Production decreased 8%. 
 
 Loss in output 3.7%. 
 Undoubtedly lost production. 
 
 Decrease in production but 
 efficiency of operatives in- 
 creased. 
 
 Loss in output greater than 
 loss in time. 
 
 Some loss in output but effi- 
 ciency of operatives in- 
 creased. 
 
 Slight increase in efficienc)^ of 
 operatives. Production de- 
 creased. 
 
 Decrease of 10% in total out- 
 put. 
 
 Total output per employee de- 
 creased 3 to 6% for piece- 
 work and 10% for daywork. 
 
 W designates broad weaving 
 N " narrow weaving. 
 
 WV " velvets. 
 
 T designates throwing. 
 
 * " dyeing. 
 
 G " general mill. 
 
 49 
 
Previous Number Product 
 Hours Employees 
 
 55 188 W 
 
 55 60 
 
 55 
 
 55 359 
 
 55 802 
 
 55 163 
 
 N 
 
 174 W 
 
 55 122 W 
 
 65 78 N 
 
 55 623 W N 
 
 N* 
 
 55 559 W 
 
 55 122 W 
 
 55 162 WV 
 
 55 320 N 
 
 55 113 W 
 
 W* 
 
 W 
 
 W designates broad weaving. 
 N " narrow weaving. 
 
 Change in Rates 
 of Wages 
 
 Piece rates increased 
 10%; time-work- 
 ers received same 
 pay as before. 
 
 Rates increased 
 proportionally. 
 
 Increase of 10% all 
 around. 
 
 No change in piece 
 rates; hourly 
 
 rates advanced. 
 
 Piece and hourly 
 rates increased 
 10%. 
 
 No change in piece 
 rates; 11% increase 
 in hourly rates. 
 
 Piece and time rates 
 increased 10%. 
 
 Piece and time rates 
 increased 10%. 
 
 Piece and time rates 
 increased 10%. 
 
 Piece 
 
 rates 
 10%. 
 
 and hourly 
 increased 
 
 Piece and time rates 
 increased 10%. 
 
 Advanced 10%. 
 
 Piecework rates in- 
 creased 10%. 
 
 Statement of Establishment 
 Production decreased 10%. 
 
 Loom records show 8 to 10% 
 loss, but we feel sure that 
 under normal working con- 
 ditions the reduction in out- 
 put owing to the reduc- 
 tion of working hours would 
 be very slight. 
 
 Pieceworkers speeded up and 
 there was no loss of produc- 
 tion. Dayworkers did not 
 speed up, with consequent 
 loss of 8 to 10%. 
 
 No gain in hourly output. 
 
 No increase in output per em- 
 ployee. 
 
 Loss in output 7.5%. 
 
 Output for daywork decreased 
 10%. 
 
 Reduced production 5%. 
 
 Decrease of 10% in total out- 
 put. 
 
 Total output decreased 3 to 6% 
 for piecework and 10% for 
 daywork. 
 
 Output per employee un- 
 changed for piecework; 10% 
 decrease for daywork. 
 
 Help better satisfied with 
 shorter hours, and they work 
 harder. Reduction of 10% 
 in time has given a reduction 
 running from up to 5% in 
 production. 
 
 Brought about a loss in pro- 
 duction running from up 
 to 5%. Workers better sat- 
 isfied. 
 
 Reduction in output practi- 
 cally in proportion to the 
 hours worked. 
 
 50 
 
 WV designates velvets. 
 * " dyeing. 
 
revious 
 Hours 
 
 55 
 
 Number 
 Employees 
 
 40 
 
 Product 
 W 
 
 Change in Rates 
 of Wages 
 
 Piece rates increased 
 10%; same weekly 
 rates for time work. 
 
 55 
 
 146 
 
 N 
 
 Piece and hourly 
 rates increased 
 10%. 
 
 55 
 
 483 
 
 WT 
 
 Proportional in- 
 crease. 
 
 54 
 
 287 
 
 T 
 
 Proportional increase 
 for piece and day 
 work. 
 
 54 
 
 813 
 
 WT 
 
 Proportional increase 
 for piece and day 
 work. 
 
 54 
 
 164 
 
 N 
 
 Piece rates increased 
 10%. 
 
 54 
 
 187 
 
 W 
 
 Proportional increase 
 for piece and day 
 work. 
 
 54 
 
 242 
 
 WT 
 
 Piece rates in- 
 creased; time- 
 workers paid same 
 weekly wage. 
 
 54 165 W Piece rates increased 
 
 }4%; time-work- 
 ers same per week. 
 
 54 62 W Piece rates increased 
 
 15%; same weekly 
 wage for daywork. 
 
 Statement of Establishment 
 
 Output decreased 5% forpiece- 
 work and 10% for daywork. 
 
 Decrease of 10% in output in 
 both piece and time workers. 
 
 Production reduced about 10%. 
 
 Curtailment in production. 
 Later, new machinery in- 
 stalled and output was main- 
 tained. 
 
 No important increase in 
 hourly output. 
 
 Decrease in output for piece 
 and day work. 
 
 Slight decrease in production. 
 
 Output decreased; percent un- 
 known. 
 
 Immediate reduction of 8% in 
 production. 
 
 About 8% loss in production. 
 
 54 
 
 96 
 
 W 
 
 Piece and time rates 
 increased 10%. 
 
 Loss of production about 5%. 
 
 54 
 
 142 
 
 W 
 
 Proportional 
 
 Decrease in production. 
 
 64 
 
 356 
 
 T " 
 
 increase. 
 
 Reduced production. 
 
 54 
 
 67 
 
 W 
 
 Proportional 
 increase. 
 
 Production decreased about 
 7% 
 
 54 
 
 337 
 
 WV* 
 
 
 Employers lost in output; em- 
 3 oyees lost in wages. Went 
 jack to 54 hours aftef a 
 month's trial. 
 
 64 1,425 
 
 54 
 
 45 
 
 WTS 
 
 TSp. 
 
 W designates broad weaving. 
 N *' narrow weaving. 
 
 WV " velvets. 
 T " throwing. 
 
 61 
 
 Loss in production, but effi- 
 ciency of operatives in- 
 creased. 
 
 Proportional loss in production. 
 
 * designates dyeing. 
 
 Sp. " spooling, twisting. 
 
 S " spinning. 
 
REDUCED TO 51 HOURS 
 Production Maintained 
 
 Establishments 
 Employees . . 
 
 Previous Number Product 
 Hours Employees 
 
 54 263 W 
 
 Change in Rates 
 of Wages 
 
 Proportional increase 
 for time-workers. 
 
 1 
 263 
 
 Statement of Establishment 
 
 Records show no decided 
 change in production. 
 
 Previous 
 Hours 
 
 55 
 
 55 
 
 Sib 
 
 REDUCED TO 52}^ HOURS 
 Production Maintained 
 
 Establishments 
 Employees . . 
 
 Number Product 
 Employees 
 
 1,003 G 
 
 252 
 
 W 
 
 1,409 WT 
 
 Change in Rates 
 of Wages 
 
 No change in piece 
 rates; bonus sys- 
 tem given. 
 
 Increase of 10% In 
 wages. 
 
 Piece wages not 
 changed; day 
 
 wages adjusted ac- 
 cordingly. 
 
 3 
 
 2,664 
 
 Statement of Establishment 
 
 Output reduced very little, if 
 any, on account of change of 
 hours. 
 
 Output for piecework re- 
 mained unchanged; 2% de- 
 crease in output for day- 
 workers — 13% of total force. 
 
 More satisfaction and happi- 
 ness all around. Production 
 per week stays the same. 
 
 Previous 
 Hours 
 
 55K* 
 55 
 
 Number 
 Employees 
 
 737 
 240 
 
 Production Decreased 
 
 Establishments . . , 
 
 Employees ... 
 
 Product 
 
 W 
 W 
 
 Change in Rates 
 of Wages 
 
 Piece and hourly- 
 rates increased 
 10%. 
 
 2 
 977 
 
 Statement of Establishment 
 
 Proportional loss of production. 
 
 Pieceworkers maintain output 
 and show greater efficiency; 
 10% decrease in output of 
 dayworkers. 
 
 Previous 
 Hours 
 
 56 
 55 
 
 54Kc 
 
 REDUCED TO 54 HOURS 
 Production Maintained 
 
 Establishments 
 Employees . . 
 
 Number Product 
 Employees 
 
 842 ST* 
 
 1,014 W T 
 
 230 T* 
 
 W designates broad weaving 
 T " throwing. 
 
 * " dyeing. 
 
 G " general mill. 
 
 3 
 
 2,086 
 
 Change in Rates 
 of Wages 
 
 Statement of Establishment 
 
 Production was not lessened. 
 
 No change in total output. 
 
 Production practically remains 
 the same. 
 
 S designates spinning 
 
 h This mill was reduced to 525i hours. 
 
 c This mill was reduced to 53^ hours. 
 
 52 
 
Production Decreased 
 
 Establishments 8 
 
 Employees 5,427 
 
 Previous 
 Hours 
 
 Number 
 Employees 
 
 Product 
 
 Change in Rates 
 of Wages 
 
 60 
 
 70 
 
 T 
 
 
 60 
 
 500 
 
 T 
 
 Proportional 
 increase. 
 
 58 
 
 139 
 
 W* 
 
 
 58 
 
 2,489 
 
 G 
 
 7% increase for time 
 work. 
 
 58 
 
 425 
 
 NTS 
 
 No change in rates 
 for piece or time 
 workers. 
 
 Statement of Bstablishment 
 
 Decrease of 12% in output of 
 dayworkers. 
 
 Output lowered in proportion 
 to time. 
 
 Decreased production. 
 
 Loss in production. Gain in 
 proficiency. 
 
 Total output per employee de- 
 creased 6%. 
 
 58 
 
 228 
 
 N 
 
 Loss in output has not been 
 as marked as decrease in 
 hours. 
 
 58 891 W Bonus for piece- 
 
 workers. Hourly 
 rates increased to 
 make up same 
 weekly wages. 
 
 Slight decrease in production. 
 
 56 
 
 685 
 
 WNT 
 
 Proportional reduction. 
 
 REDUCED TO 55 HOURS 
 Production Maintained 
 
 Previous Number 
 Hours Employees 
 
 60 
 
 58 
 
 58 
 
 168 
 
 52 
 
 121 , 
 
 Establishments 
 
 Employees 
 
 . . 3 
 
 . . 341 
 
 Product 
 
 Change in Rates 
 of Wages 
 
 Statement of Establishment 
 
 Sp.* 
 
 No increase in wages. 
 Bonus system. 
 
 Output maintained by chang- 
 ing workers from day rates to 
 piece rates. 
 
 T 
 
 No change in piece 
 or hourly rates. 
 
 No perceptible change in out- 
 put. Output maintained by 
 speeding up machinery. 
 
 W 
 
 Bonus given. 
 
 No change in output notice- 
 
 W designates broad weaving. 
 N " narrow weaving. 
 
 T " throwing. 
 
 S '* spinning. 
 
 able. 
 
 * designates dyeing. 
 
 Sp. " spooling, twisting. 
 
 G " general mill. 
 
 53 
 
Production Decreased 
 
 
 
 Establishmei 
 Employees . 
 
 Its 
 
 . . 6 
 
 
 
 850 
 
 Previous 
 Hours 
 
 Number 
 Employees 
 
 Product 
 
 Change in Rates 
 of Wages 
 
 Statement of Establishment 
 
 60 
 
 123 
 
 T 
 
 Same wages for 55 
 as for 60 hours. 
 
 Decrease of 10% in output for 
 daywork. 
 
 60 
 
 102 
 
 Sp.* 
 
 
 Proportional decrease. 
 
 60 
 
 300 
 
 Sp.* 
 
 
 Proportional decrease. 
 
 60 
 
 105 
 
 W 
 
 
 Loss in production about in 
 proportion. 
 
 58}id 
 
 149 
 
 TN 
 
 No change in piece or 
 hourly rates. 
 
 Slight decrease in production. 
 
 57K 
 
 71 
 
 T 
 
 Same pay for 55 
 
 Proportional decrease. 
 
 hours as for 57^. 
 
 W designates weaving. 
 
 N " narrow weaving. 
 
 T " throwing. 
 
 * designates dyeing. 
 
 Sp. " spooling, twisting. 
 
 d This mill reduced to 55}i hours. 
 
 54 
 
Publications 
 
 of the . 
 
 National Industrial Conference Board 
 
 15 Beacon Street, Boston, Mass. 
 
 Research Report No. 1. Workmen's Compensation Acts in 
 THE United States — The Legal Phase. 62 pages. 
 April, 1917. $tm. 
 
 Summary of Research Report No. 1. 8 pages. 
 arch Report No. 2. Analysis of British Wartime Re- 
 ports on Hours OF Work as Related to Output and 
 Fatigue. 58 pages. November, 1917. 31-00. 
 Research Report No. 3. Strikes in American Industry in 
 
 Wartime. 20 pages. Marchrl918. 50 cents. 
 Research Report No. 4. HourS of Work as Related to Out- 
 put and HeaEth of Workers — Cotton Manufac- 
 turing. 64 pages. March, 191 8kJj^.QP:_- .^ 
 Research Report No. 5. -The CxNAm^i^^^tfmjaM^WTES 
 Investigation Act. 28 pages^ April, 1918. SOcente. 
 Research Report No. 6. Sickness iNSURSfliCE ^r .^icknesS 
 - Prevention? i4 pages. M^,1^!W^ 59^^ ] 
 , Research Report No. 7. Hours of ^W!5|^!fe^s Rel^^d -p^ 
 _, Output and Health of Wpiix^^l^^^^^^^^^ikGE 
 Industry.^ 76 pages. Jiiae, 1918."';'"~^ja2b2ii£:i-'- ' 
 - Research Report No. 8. W4RT1ME Employment of Women 
 m the Metal Trades. 80 pages. July, 1918. 31.00. 
 Research Report No. 9. Wartime Changes in the Cost 
 of Living/^ July, 1914, to June, 1918. 82 pages. 
 August, 1918. 3I.QO: 
 Reifarch Report No. 10. Arbitration and Wage-Fixing in 
 
 Australia. 52 pages. , October, 1918. 31.00. 
 Research Report No. \\. T^^ Eight-Hour Day Defined.: 
 
 12 pages. December, 1918. 50 cents. 
 Research Report No. 12. HovTi.s of Work as Related to 
 Output AND Health of Workers — Wool Manu- 
 facturing. 69 pages.- December, 1918. 31-00. 
 Research RejpprtN 0.13. Rest Periods for Industrial 
 ""* WoRKEks, 55 pages. January, 1919. 3100. 
 
 fearch Repqri No. 14. Wartime Changes in the Cost 
 ■ Living— July, 1914, to November, 1918. 33 pages. 
 
 ,00. ■ ■ ^ ; " 
 
 leportNo. 15. Problems of Ji^dustriai. ReAdjust- 
 
 MfeKT in the United States. 5,8 pages. 31-00. . 
 
 Res^i^^M^port No. 16. Hours of Work as Related to 
 
 v£^.|;j ii^!UT And Health of Workers — ^Silk Manu- 
 
 ^^"^ facturing. 54 pages, ^^arch, 1919. 31.00. 
 
 INDUSTRIAL NEWS SURVEY 
 
 Important industrial news in concise form. 
 Weekly. 32.00 per year