ſae;** ✉ğ ✉ §§ ¿ §####- ș4 ș* ≡ §§§ ::::: §§ §§§§§§§ ¿ ¿ ĒĢĒ §§ |-§§ * §§§§ 、。、、。、、。 ::::：::::：:********* 、。 §§§§§ ;&3&&3&& §§§§ §§ 、、。 ！******* -¿&& -¿??¿ ****** ;，&&\;&y&+&y&#łºśw{.***g*.***** 、、、。、、、。、、。§5€$£§§§§§§）; 、。gjë…«******** ae ț¢&#șaeșiķae 、、、。 ×× ȚUȚIȚIȚĂȚIȚIE[[№ſ &ººººººººººººººººººººººººººš!！！！！！！！!！！!！！！！！！!！！！！！！！！！！！！ №-№ºººººººººº！！！！！！！！！！！!！！！!！！！！！！！！！！！}Sſae \\}\\\\ ~~~~）;- T_- ：°′SÑ ºutritiºn [] · ：Z | | | | } 8 H | } # { | { # | }; } [] 8 # } Q , , , w ** .…” "...g. f º- • *::: , , , , . . . . . . . COMMITTEE ON LABOR TURNOVER THE NATIONAL ASSOCIATION t { ..., , EIGHTH ANNUAL CONVENTION NEW YORK, N. Y. . May 31st, June 1st, 2nd, 3rd, 4th, 1920 NOT RELEASED FOR PUBLICATION UNTIL JUNE 4th, 1920 & - COPYRIGHT, 1920 THE NATIONAL ASSOCIATION OF CORPORATION SCHOOLS OF CORPORATION SCHOOLS) . . . . . ; COMMITTEE ON LABOR TURNOVER A MR. L. L. PARK, Chairman AMERICAN LOCOMOTIVE COMPANY Schenectady, N. Y. MR. J. C. VANDER PYL AMERICAN HARD RUBBER COMPANY College Point, N. Y. MR. E. B. MERRIAN GENERAL ELECTRIC COMPANY Schenectady, N. Y. MR. H. W. DAHLBERG THE GREAT WESTERN SUGAR COMPANY Denver, Colorado MR. E. R. NEAR THE NATIONAL CITY BANK OF NEW YORK New York City MR. G. A. GLYER MARSHALL-WELLS COMPANY Duluth, Minn. DR. F. S. CRUM THE PRUDENTIAL INSURANCE COMPANY OF AMERICA Newark, N. J. MR. E. O. BERG BRIDGEPORT BRASS COMPANY Bridgeport, Conn. sº --- é. 7 4. cº REPORT OF COMMITTEE ON LABOR TURNOVER The work assigned the Sub-Committee on Labor Turnover this year included (a) a study of the causes of and means for reducing labor turnover and (b) an analysis of the disturbing elements in industrial employment and means for neutralizing them. - It has been easier to study the more common causes of labor turnover than to analyze the disturbing elements in employment, since the latter ran well below the surface, and a close investi- gation of actual conditions is necessary to an intelligent discus- sion of these elements. - The committee has, therefore, confined its report this year to the presentation of certain factors in labor turnover and the man- ner in which they affect different classes of employes, together with the remedies usually suggested for overcoming the trouble represented. Difficulties in Such a Study As might be expected, the Committee experienced difficulty in attempting to standardize the conditions prevailing in various industrial communities. One Association member in replying to our request for an opinion concerning the importance of certain factors in labor turnover stated: “Our city is all by itself and no one knows the answer.” It is obvious that conditions in the larger labor centers, as well as in the individual industries, them- selves, vary so greatly and have such marked peculiarities that the rules applying to one have no connection whatever with others. * - There are, however, certain factors which appear common to most industrial and commercial organizations, and the committee has endeavored to discuss the importance of these and the means for overcoming them. - «w” * Study of Turnover by Labor Groups In approaching the study of the factors in turnover it be- comes apparent that these factors must vary in importance in different groups of employes, and the committee has based its study upon the more common divisions: unskilled, semi-skilled, skilled and office forces. The term “unskilled” is here assumed to cover those groups of employes which require little training to enable them to perform their duties; “semi-skilled,” those whose training can be accomplished in a few weeks, while “skilled” applies to those whose preparation has been a matter of years. “Office force” applies to the clerical organization below the rank of “supervision.” Any effort to study turnover except by some such grouping as the foregoing cannot but be too general to accomplish definite results. Few causes relate equally to all classes of employes, and some sub-division is necessary to intelligent study. We have chosen the above classification because the terms are most readily understood and most commonly applied. Factors in Turnover To review the causes most commonly found in the problem of turnover and the remedies which suggest themselves as apply- ing to them, the following list is given. These are by no means new, and they are here given for the purpose of applying them to the grouping already outlined. EXTERNAL FACTORS CAUSES 1. Competition. In the larger labor centers there is a tendency for some firms to draw employes from others by the use of unfair methods. There is also a tendency for employes to drift from one firm to another when work is plen- tiful and change of employment is easy but without permanent good to themselves. REMEDIES Labor Agreements, whereby neigh- boring firms protect each other against unfair practice, tend to re- duce the changes due to this cause. While not an absolute safeguard it discourages a large amount of drift- 1Ing. ... 6 Cooperative Employment Agencies which handle transfers within a com- munity also tend to reduce inter-plant changes and insure better placement. 5 CAUSES 2. Housing. In most localities the housing problem is serious and affects nearly all classes of labor. This heading includes the question of homes for employes with fami- lies, rental of houses for would-be employes, boarding places for day and night workers and bunk houses for groups of laborers. 3. Transportation. The problem of securing adequate transportation in large cities or at plants located far from center of residence of employes is often a serious one. This usually relates to traction service, though frequently the steam railways are involved. 4. Climatic Conditions. Expo- sure to weather or extremes of heat and cold affect some classes of help. In lattitudes where winters are severe or summers are very hot the industries are apt to be affected in certain seasons. REMEDIES Housing Projects. The plans found effective include: (a) Housing plans “en bloc,” such as the building of model cities or other groups of homes for employes. (b) Building houses for employes in -any part of community on easy terms of payment. * (c) Free real estate agencies for helping employes build their own homes or secure houses for rental. (d) Operating boarding or bunk houses or helping employes find suit- able places to board. (e) Consulting service in building or rental, advising as to wise or un- wise investments, plants, estimates, etc. - Improved Traction Service can be often secured if pressure is exerted by employing firms. Bus Service is maintained by some concerns as a supplement to car serv- ICe. • “Jitney” Service has at times been inaugurated as a further solution of the question. Prizrate Auto Service in which groups of employes arrange for transportation by private autos owned by fellow-employes. Protection against extreme condi- tions can sometimes be made. Sheds, enclosures, screens and special means for heating or cooling have been found possible at times. INTERNAL FACTORS 5. Poor Placements is responsible for some degree of turnover. Lit- tle effort to determine the fitness of the applicant for the work of the company or to place him where will be of greatest value to his employer and to himself, can re- sult only in much dissatisfaction and lack of interest. Job Analysis as a means of deter- mining the requirements of the task and Tests of Applicants to discover their qualifications are acknowledged essentials in intelligent placement. Interviews by experienced employ– ment clerks are valuable but they should be supplemented by detailed information regarding applicants and jobs. CAUSEs 6, Working Conditions as re- lated to the sanitation, safety and comfort of employment have much to do with the attitude of employes toward their work. Where condi- tions encourage a “grouch” there cannot grow the best spirit of co- operation. . 7. Faulty Management is re- sponsible for misunderstandings between management and men, lack of cooperation and many un- necessary grievances. The fore- man is the medium for expressing the company's policy to the men and he often misrepresents the management through personal feeling and lack of knowledge. 8. Monotonous Work. The ele- ment of repetition with a limited number of operations tends to weary some operatives of their tasks, although others prefer work of that character. The mere fact of repetition does not constitute an objection to a job, but it does affect adversely a certain percent- age of employes. 9. Business. Fluctuations. Sea- Sonal work or work in which there is much extreme of short time or overtime tends to discourage permanency of employment. Men are apt to seek places which offer less hazard of loss of time or less, pressure due to hurried overtime jobs. 10. Remuneration. Wages lower than prevailing standards will tend to cause turnover, unless there be other conditions which offer great- er attraction. The basis of wage payments is always ground for Some dissatisfaction and change of employment. REMEDIES Modern Shops with sanitary equip- ment, safeguarded machinery, neces- sary aids to comfortable work, proper lighting and heating and good house- keeping are aids in developing con- tentment and plant morale. A Definite Labor Policy is essential to friendly relations between employ- er and men. - The Training of Foremen is recog- nized as an important means of pre- paring them to properly represent the management and use diplomacy in handling matters of discipline. Employe Representation in Man- agement is being considered as an im- portant means of preventing misun- derstandings and adjusting griev- a11CeS. Plan for Transfer of operatives who cannot stand work of too con- fining a nature is advisable. It is sometimes possible to provide regu- lar rotation of work to give variety of occupation to those who react un- fºrably to the requirements of the taSk. - Making Monotonous Work More Interesting by labor-saving devices, modified conditions and modified processes is a better program than that of constantly changing opera- tives. Eartending the Business Season in some industries can be accomplished by careful planning, spreading out the year's work more uniformly. Filling in the Vacant Seasons with work other than the usual line, when possible, aids a concern to hold its organization. Fair Compensation will help to steady employment, if other condi- tions are favorable. Wage understandings between em- ployers and men tend to maintain proper wage relations. CAUSES 11. Outside Trouble, such as dif- ficulties at home, sickness, legal en- tanglements, etc., tend to cause transfers of employment to other communities. 12. Lack of Skill. Many em- ployes because they cannot readily earn fair wages within a reasona- ble time after starting work, be- come discouraged and quit. Others also find themselves in a narrow groove with little chance for ad- vancement and leave to widen their experience. 13. Lack of Ambition is charac- teristic of a large number of the habitual applicants found in the employment offices, who go from place to place because they cannot make good anywhere. 14. Roving Tendency. The ten- dency of many capable and efficient employes to rove periodically from city to city or from shop to shop is particularly difficult to overcome as their habit is not based on dissat- isfaction with local conditions but on a desire to roam. REMEDIES Free Health Clinic by which em- ployes receive free medical advice will tend toward better health and less pessimism. Health Insurance is intended to re- lieve anxiety as to loss of time and insure security of position in time of illness. Free Legal Advice will assist in keeping employes out of some legal difficulties which might otherwise lead to their leaving town. Family Contacts, such as commu- nity houses, playgrounds, visiting nurse, milk stations, company stores, etc., tend to eliminate many home troubles and prevent exploitation of the employes. Training Courses, either on the job or in vestibule schools, give better chance for the earning of a compe- tency in a short time. They also give uniformity to the training and are likely to avoid the difficulties which later develop when training has been indifferent. Educational Courses which give supplementary training or schooling in subjects allied to the trade but not immediately part of it, help to produce efficiency and supplement manual skill. A System for Promotion will offer means of advancement to capable men who are anxious to become more skilled. Personal Contact with Welfare Workers who endeavor to stimulate interest in self-betterment and cre- ate a desire for improvement, should help overcome this trouble. Social Activities which develop ifi- terest in local affairs and help to tie up employes to the general organiza- tion assist in holding them. Athletic Contests which give oppor- tunity for personal exploit serve to increase the local attractions. Stock Ownership by Employes with the added income involved aids in lo- calizing interest. Bonuses for long term services and Pensions for old employes help to secure permanency of employment. Factor Chart To enable the committee to present a comparison of the im- portance of the factors just cited, a chart was prepared and sent to a number of typical industries and commercial houses, with a request that they rate the importance of each factor as viewed from experience in their particular fields. The average of the ratings returned from these firms is given in the following chart, in which the horizontal lines indicate by their length the relative importance of each item. - * *~ While it is true that the rating of no two firms will be the same, owing to the great variation of conditions prevailing, the results obtained from the chart are suggestive of certain domi- nating circumstances to which some attention may well be given. CHART SHOWING RELATIVE IMPORTANCE OF FACTORS IN LABOR TURNOVER Unskilled Semi-skilled Skilled Office. FACTORs Labor Labor Labor Help Earternal Competition . . . . . . . . . . Housing . . . . . . . . . . .* . . — *-mºmºmºsºm-º. Transportation ... . . . . . *m-º-º: * Climatic Conditions... — *-*- Internal Poor Placement . . . . . Working Condition . . Faulty Management... — *s- Monotonous Work.... — * Business Fluctuations. – * Remuneration .... . . . . = Personal Outside Trouble. . . . . . *E*E* Lack of Skill......... — Lack of Ambition. . . . Roving Tendency . . . . — *-*-* = Approximate Per Cent - * * Annual Turnover... 167% 126% 98% . 32% Notes on Factor Chart . In comparing the marks on the chart for the various groups, it is evident that: External factors affect all classes nearly equally; Personal factors show some difference in the various groups; Internal factors show the greatest change in the different groups. It is well that we should note that in averaging the marks of the different firms, many of the extreme marks are neutralized by other extremes, making the average more moderate than any individual case. We may, therefore, consider the differences found as being more exaggerated in any individual case than indicated by the average shown. In summarizing the conclusions which we feel justified in drawing from the chart, we are giving below some of the phases of labor turnover which we believe are worthy of consideration. 1. Competition for Labor is generally recognized as a seri- ous problem in all groups of employes. While there can be no criticism of such transfer of employment as adds materially to the progress or advantage of the individual, the endless chain of industrial connections brought by the preying of one concern on the organization of another can only add to the difficulties natu- rally attendant on the maintaining of a stable working force. Cooperation between neighboring concerns cannot be too cordial in employment relations, and it is to be hoped that in the larger business centers there may be less destructive effort and more constructive progress made in promoting industrial stability. 2. Insufficient Housing is becoming a constantly pressing source of difficulty among all groups. Many firms hesitate to launch housing movements, but it is well to recognize the possi- bility of doing something toward housing without a concern entering the landlord class. Means for rendering service to would-be builders, advice as to proper value of real estate and cost of building plans, free service in locating vacant houses, and other means may be used to advantage even if no building projects are contemplated by the employer. Every firm could help toward meeting the housing need to some extent, even though its plan might be a modest one. - - - 10 3. Poor Placement appears particularly strong in connec- tion with unskilled and semi-skilled groups. As might be ex- pected, we are making better progress in placing skilled help and office help than in locating those whose capacity is limited and whose talent is less pronounced. The chart emphasizes the need of greater study in the placement of unskilled labor. Too fre- quently the employment office has assumed that a laborer can work equally well anywhere and under any conditions. It is apparent that distinctions will be found in this field as well as in the others. - 4. Working Conditions among unskilled and semi-skilled workers apparently have an important influence. While this condition is probably based upon the nature of the work which these groups perform, it is worthy of consideration that condi- tions may be made to contribute more to industrial stability if the same attention is given to the surroundings of the less-skilled workers as is given to the places where skilled labor and clerical forces operate. 5. Faulty Management is prominent in the three labor groups, somewhat emphasized in relation to semi-skilled labor. A more careful study of the needs of this field and more prac- tical training of foremen for dealing with its problems will un- doubtedly help reduce the difficulties encountered. 6. Monotonous Work is naturally more pronounced in the unskilled and semi-skilled groups. While the shifting of em- ployes to provide variety is commendable, it is better, where possible, to change the methods to relieve monotony to a greater extent. Motion study has aided in simplifying the movements required for any operation, but it has tended to increase monot- ony. Would it not be desirable if several different groups of movements could be devised for any task, enabling the operator to vary his efforts with equal skill? It is well for us to remem- ber that all people do not react to monotony in the same manner, and that some operators prefer work which is highly objection- able to others. 7. Remuneration is prominent in all groups and probably will always be one of the important factors in employment. This item emphasizes the desirability of proper grading of rates to encourage the attainment of higher degrees of skill. Where there is no incentive to trade training, the tendency is toward 11 the easiest path to competency, usually at the expense of mental development. 8. Lack of Skill is paradoxically most prominent among the “unskilled.” One would hardly expect to find complaint of . lack of skill at work in which no skill was required, but the defini- tion of the term is undoubtedly intended to include “lack of intelligent effort” which is compatible with laboring work. The great difficulty in securing men who are willing to undertake unskilled work has probably been responsible for the presence of men not suitably fitted to the task and, therefore, performing their duties in an unsatisfactory manner. The prominence of this item in the clerical field suggests the need of better schooling for office forces. In all groups there is indication of the im- portance of practical training to prepare for the requirements of the work demanded. 9. Lack of Ambition appears charged against all groups to an unusual extent, though most prominent among the unskilled. Means for awakening interest in self-education and a desire to do better work are possible in any organization, but they demand a larger degree of personal contact than is found in most estab- lishments. 10. Roving Tendency appears to vary in inverse ratio to the education required for the work. Where the drudgery is greatest, the tendency to drift seems strongest. Unless recrea- tional or other means can be provided to neutralize the effect of the work, it is difficult to check the desire for a change of lo- cation. 11. Per Cent of Turnover. As would naturally be expected, the per cent of turnover is least in the clerical forces, somewhat greater among skilled men, still greater with the semi-skilled, and greatest among the unskilled. The nature of the work and the lack of advancement in the less skilled positions tend toward this condition. Our educational, social and betterment plans are apt to benefit most those in the groups which already have the greatest stability, and special study is needed of means for help- ing those in the less fortunate classes to find proper opportunity and incentive to remain longer in their present fields of endeavor. The results of the above study may seem to lack direct value to those contributing to the chart or reading the report. It is hoped that it may suggest ways of studying the problem of labor * 12 turnover to those who have not thoroughly covered their field. The nature of some of the replies received indicate that even in well organized personnel departments there is lack of certainty as to the best methods of meeting certain phases of the problem, and it is the hope of the committee that the discussion at the convention may bring out further suggestions of value to our members. \ Method of Figuring Labor Turnover In addition to its study, of the factors in labor turnover, the committee is pleased to print as part of its report the following paper on “How to Figure Labor Turnover,” by Dr. Frederick S. Crum, Assistant Statistician of the Prudential Insurance Com- pany, and a member of this committee. The paper is reprinted from the Quarterly Publication of the American Statistical Asso- ciation of June, 1919. It has received such favorable comment by men interested in employment statistics that the committee believes its publication here will be of practical value to the members of the Association. 13 [Reprinted from Quarterly Publications of the AMERICAN STATISTICAL Association, June, 1919.] HOW TO FIGURE LABOR TURNOVER By FREDERICK S. CRUM The term labor turnover has been variously defined and as its calculation is dependent upon its definition, attention should first be given to just what the term means. So far as I am able to find out from reading an already somewhat extensive literature on the subject labor turnover was first used and is still most widely used to express the phenomenon of hirings necessary to replace separations in any given industry, establishment or de- partment of an establishment. Before the subject was analyzed too closely, perhaps, by writers of the academic type, labor turn- over or the flow of labor in any given industry or establishment, was measured in terms of the average or normal payroll, and it was expressed in the form of percentage of such payroll dur- ing the period under observation—usually a week, a month or a year. If the average payroll of a given establishment were 1,000 in a business which required that such payroll be kept at that figure and during a year 500 additions to the payroll were neces- sary to replace separations therefrom, the yearly labor turnover would be expressed as 50 per cent, or *%000. On an increasing payroll in a growing business the separations from the payroll would still represent the numerator and the average payroll the denominator. On a decreasing payroll in an establishment gradu- ally going out of business, perhaps, or declining from any cause whatsoever, the numerator would be the number hired and the denominator the average payroll, as before. In other words, average payroll by this most generally employed method is al- ways the denominator and the number of employes replaced during the period under observation is the numerator in the calculation of labor turnover. As this is the older, so, I believe, 14 it is also the best method of measuring labor turnover. What- ever method of calculation is ultimately agreed upon, the statisti- cal problem of labor turnover can still be truly stated as Richard B. Gregg has stated it, namely, as the “measurement of the move- ment of industrial workers in and out of their employment, and the analysis of its causes and results.” At the National Conference of Employment Managers held in Rochester, N. Y., May 9–11, 1918, a committee was appointed to draft a report which would define labor turnover and state how it should be calculated. The report made by this committee reads as follows: “Labor turnover for any period consists of the number of separations from service during that period. Separations in- clude all quits, discharges and lay-offs for any reason whatsoever. “The percentage of labor turnover for any period considered is the ratio of the total number of separations during the period to the average number of employes on the force report during that period. The force report gives the number of men actually working each day as shown by attendance records. “To compute the percentage of labor turnover for any period, find the total separations for the period considered and divide by the average of the number actually working each day through- out the period.” In brief, this committee advocated the use of attendance force rather than payroll in the calculation of turnover percent- ages. This seems to have been the main principle advocated in this report, but it is somewhat ambiguous and misleading in some of the statements. For illustration, labor turnover for any period does not necessarily “consist of the number of separations from service during that period.” In the current and recent history of shipyard plant construction we find many illustrations of the falsity of this statement, When a shipyard nears completion many of the men engaged for its construction must be laid off from day to day, and these separations do not all represent labor turnover, according to the best and most usual definitions of that term. On a decreasing payroll only the number hired can truly represent replacements or turnover. The statement that “to compute the percentage of labor turn- over for any period, find the total separations for the period con- & 15 sidered and divide by the average of the number actually working each day throughout the period” is to me ambiguous. Probably all that is meant is that the average daily attendance is recom- mended as the divisor. If so, it does not mean that only those are to be included in the denominator of the fraction who have worked or been in attendance “each day throughout the period.” It is conceivable that the turnover in a given year would be rep- resented by infinity in an establishment where no employe could show a record of 100 per cent attendance. The point in this committee report, however, which challenges discussion is the recommendation that the working force, as shown by attendance records, be made the basis upon which labor turnover is calcu- lated rather than payroll force. - Anyone who has had wide experience in studying compara- tive labor turnover figures in such an extensive industry as ship- building will at once question the validity of this new method apparently advocated by at least some of the employment man- agers at the National Conference in Rochester, last May (1918). One of the principal objections to it is that attendance is a more fluctuating factor than payroll. On an extremely hot day, 1,000 or more men out of a force of 5,000 might lay off for the one reason “too hot to work.” This is not an isolated instance; there are plenty such illustrations in the daily attendance reports of the 160 or more shipyards which it was my privilege to review for a period of eleven months this past year. Heat, cold, blizzard, rain and snow are meteorological conditions which cause wide fluctuations in daily attendance in shipyard and other outdoor employments. Again, an epidemic period such as that we have been experiencing in the shipyards since September causes most violent fluctuations, daily, weekly and monthly, in the attendance records—much wider variations, quite naturally, than are re- flected in the payrolls. In the first week of October, 1918, it was reported that approximately 6,000 out of 19,000 employes of three large shipbuilding plants at Gloucester, N. J., were ill with influenza. Obviously, the resulting temporary low attend- ance had little or no real relation to labor turnover. Why, then, should this attendance factor be used in calculating turnover percentage? Another objection to the attendance records as a basis for the calculation of labor turnover percentages is that very fre- 16 tº quently the separations or quits when recorded are of employes who may not have been in attendance for a day, two days or longer. Theoretically, the payroll is, therefore, a better denomi- nator for the labor-turnover fraction, as it always contains the whole labor force, while the attendance force at any given time may range from 100 per cent of the payroll down to 50 per cent or even less in severe weather, during an epidemic, obstructed traffic, etc. These latter phenomena, while affecting attendance seriously, may not and often do not affect the payroll or the separations so far as these enter into the labor-force records. One of the best statements that I have seen advocating pay- roll in preference to attendance as representing average force for labor turnover calculations is that by Mr. E. Goldberger, efficiency engineer, Packard Motor Car Company, published in the November, 1918, issue of Industrial Management. The state- ment follows: “The total number of men that have been hired and been paid off, i.e., actually replaced, represents a complete turnover, and if there be an excess of men hired over paid off, they actu- ally are to be accounted to increase in force, while if there be an excess of men paid off, they represent a decrease in force. “There are more reasons in favor of counting the men on our books rather than those really attending, as some men leave without notice, passing through a transitory stage of not attend- ing and still being on the books, so that, when they leave and are counted in the turnover, they do not represent the turnover of the force working, but of that on our books. “It is much easier to count the men on the books than those attending, as the first information can be made available quickly in the employment office, while the last one has to be made avail- able by many clerks in different departments or by the time office. “It is very likely that high wages, where they are unwar- ranted, will force us to face the problem of absentees, and to study a standard coefficient of attendance. This is, in my opin- ion, a good enough reason to keep the same denominator for all three coefficients — labor turnover, labor fluctuation and labor attendance.” The following quotation* also is in conformity with my own * From “Standardizing Labor Turnover” by Harry W. Kimball, Indus- trial Management, December, 1918. 17 point of view as regards the method of calculating the labor turnOver : “What most persons mean by labor turnover is labor replace- ment. The term labor turnover arose in connection with discus- sions of the cost of hiring and firing workers. It was recognized that men drifted in and drifted out of industries. Competent men often did not stay, and incompetent men had to be fired after expensive trial. Effective labor service finds the right man for the right job and then gets him to stick. If he does not stick, then someone else has to be hired to take his place, or if a mistake in judgment is made and an unfit person employed, then that person has to be fired and another worker found for his job. This is labor turnover, but a better and more accurate phrase is to think of it as necessary labor replacement. “The National Employment Managers’ conference by decid- ing that the number of employes actually working each day rather than the number appearing upon the payroll should be the proper base for computing labor turnover introduced another factor into personal fluctuation. This factor, which may be termed labor variation, is something different from either plant change or labor replacement. “The number of the working force actually working on any one day is dependent upon sickness and accident, upon explain- able absences, and upon habitual loafing usually after pay day. Increasingly, employment offices are keeping records of these daily absences, seeking for the causes of these constant leaks and endeavoring to stop them. But such absences do not mean usu- ally that a man is fired or even that someone is hired to take his place. This labor variation is not labor turnover as the term is ordinarily used. Yet if the daily working force rather than the payroll sheets are used as the base for figuring labor turnover, then this element of temporary absences affects to some degree the percentages. “A report, then, of personnel fluctuation for any period of time should show in percentages the plant change, the labor re- placement and the labor variation. These are three distinct ele- ments in the personnel fluctuation. They are due to different causes, require different remedies and should never be confused. “One other subdivision may be desirable. Persons whose places must be refilled leave for avoidable and unavoidable rea- 18 sons. Among the unavoidable reasons are marriage, death, ill health, retirement, family leaving city, and military service. It is worth while to divide the leavings under labor replacement into avoidable and unavoidable, making separate percentages. The decrease of the avoidable leavings from week to week, and from year to year, is the true measure of the efficiency of the labor service. “While a record should be kept of all transfers from one department to another within the plant, these transfers should not enter into the calculation of any personnel fluctuation, as they do not increase or decrease the working force. “The simplest way of calculating these three elements of per- sonnel fluctuations is by the week, and the simplest base to use is the number on the payroll.” Daily attendance records are extremely important and hours- of-work records, also very important for several purposes, should be available to the manager of a plant or establishment, but I seriously doubt their utility in the calculation of labor turnover. Payroll measured against daily attendance gives a reasonable notion of the percentage and total amount of absenteeism. Such ratios could be computed weekly, monthly, quarterly and yearly. In other words, daily attendance should preferably be consid- ered as a coefficient of payroll. Among shipbuilders, for the country as a whole, approximately 85 per cent of the average payroll was in daily attendance. There were wide variations, however, from this figure when individual yards, different sec- tions of the country and the different seasons of the year were considered. For a proper measurement of daily attendance good payroll records must be freshly available and, if available, there would appear to be no good reason why labor turnover should not be measured against payroll rather than against attendance—a more widely fluctuating, and, at least as regards labor flow, a more artificial factor than payroll. Perhaps one objection to the use of payroll records is that they sometimes are not cleared fre- quently enough. This is not a valid objection in labor turnover calculations, for the net gain or loss in the payroll is readily obtained when the number hired and the number of separations are known for a given period. An effort should be made to bring 19 about standard and uniform practice in the keeping of all labor or work-time records, and payroll records unquestionably should be more uniform and accurate than at present. Probably the most generally feasible plan would be to have the payrolls cleared at least twice a month, preferably, perhaps, on the first and fif- teenth of each month. Such payroll figures would, I believe, give an excellent basis from which to calculate labor turnover or re- placement percentages, labor fluctuations, and daily attendance. If the payroll were cleared weekly the accuracy of the calcula- tions would, of course, be somewhat greater. Unless some sound standard as to the factors which shall be used in calculating labor turnover is adopted and applied gen- erally at an early date, there will be created a still greater statis- tical chaos than already exists. At present not only is there no universally or, even generally recognized method of calculating labor turnover, but there is also an “utter lack of any standards in estimating cause, extent and cost of labor turnover.” An efficient employment manager is probably the first and most important step in any effort to bring about stability in the labor force of an industrial plant. The employment manager can perhaps best demonstrate his efficiency by pointing to the reduction of the labor turnover under scientific as compared with the previous and presumably unscientific “hire and fire” method of foremen. To the employment manager, therefore, it is impor- tant that his methods of calculating labor turnover shall be uni- form and on a basis that can be defended by sound common sense and logic. Labor turnover should be measured for the principal depart- ments of any given plant as well as for the plant as a whole. Transfers from one department to another would enter into the labor turnover of the departments affected, but such transfers would not count in the labor turnover of the whole plant. In the United States there is a large amount of interstate migration, but these migrants are not added to the statistics of immigrants and emigrants for the United States. The analogy illustrates how transfers should be handled. The labor turnover of the * “An Initial Survey of the Problem of Labor Turnover,” Confidential Reports Series, Report No. 1, The National Association of Corporation Schools. * 20 departments of any given plant if summed would show too large a labor turnover for the plant as a whole, provided there had been interdepartmental transfers. Turnover figures are valuable only as they are used and under- stood by the employment manager and by the plant management as an index to the labor conditions. When the labor turnover figures are reduced to simple percentages which always have definite significance, the management will get light upon many problems, the solution of which is necessary to effect a reduction in turnover. As the necessary remedies are applied the labor turnover percentage furnishes an index of the result of the remedies. A few of the many partial solutions of the big prob- lem of turnover are profit-sharing, bonus payments, group in- surance, other insurance provision for employes such as disability allowances, pensions, etc., good restaurant facilities, quick pro- motions for the most capable, careful fitting of the employe to the job, etc. Mere expression of labor turnover in percentages will never solve the problem, but a simple mathematical statement of its size is likely to persuade the manager that it is important to seek for the causes why men leave in such large numbers. Next, he is apt to study the cost of turnover and, finally, he will be in a position to justify to himself or to the owner of the business that it is better to spend money for preventive measures than for the wasteful and reckless hiring and firing methods so usual in American business. The mathematical formula of labor turnover will thus be transformed into a better understanding of the human element, with the result that the average length of serv- ice of the employes will be considerably extended and the labor turnover correspondingly reduced, to the great and mutual ad- vantage of employer and employes. ILLUSTRATIONS OF LABOR TURNOVER CALCULATIONS For an increasing payroll the following example will illus- trate the method when payrolls and changes in same are avail- able in weekly reports. Some months will embrace five weeks and others four. Any given week should be included in that month which embraces four or more of its days. When five weeks are included in a given month the turnover percentage 21 for the month must be multiplied by 10.4 to reduce it to a yearly basis, and when only four weeks are included the factor for multiplication is 13. INCREASING PAYROLL. Labor Turnover & No. No. No. 2–Percentage—s Week Ending Payroll Hired Lost Replaced Weekly Yearly October 5, 1918. . . . . 2,593 37 15 15 .58 30.1 October 12, 1918. . . . . 2,615 42 17 17 .65 33.8 October 19, 1918. . . . . 2,640 26 19 19 .72 37.4 October 26, 1918. . . . . 2,647 17 20 17 .64 33.4 November 2, 1918. . . . 2,644 22 18 18 .68 35.4 Total . . . . . . . . . . . . 13,139 142 89 89 3.39 35.2 Average Payroll . . . . 2,628 r In this example the fundamental data are assumed to be available, week by week, and the establishment is assumed to be a growing one with weekly additions to the force generally larger than the subtractions from all causes whatsoever. In this example the weekly percentages of turnover varied only slightly, or from .58 to 72. Multiplying the weekly turnover percentages by 52 reduces them to a yearly basis. For the month ending November 2 the turnover percentage was 3.39. As thes period was really five weeks, or 10.4 part of a year, the turnover per- centage reduced to a yearly basis is 3.39 multiplied by 10.4, or 35.2. In the next example the week ending October 26 is purposely shown as indicating a greater loss than gain in the payroll. The replacements for that week represent the number hired rather than the number lost, or 17 rather than 20. For the five weeks as a whole, however, the number hired are in excess of the sepa- rations, so in the totals the 20 is included as replacements. When large numbers of industrial establishments are being followed week by week, as is done in the Emergency Fleet Corporation, the weekly turnover percentage is a valuable index of the general trend of this phase of the labor problem. It should also serve a useful purpose in any industrial establishment. The payroll, however, should be cleared weekly, if feasible, and at worst, not less often than every two weeks if the labor turnover is to be calculated with a close approximation to exactness. 22 DECREASING PAYROLL. Labor Turnover No. No. No. a-Percentage—s Week Ending Payroll Hired Lost Replaced Weekly Yearly October 5, 1918. . . . . 6,754 70 95 70 1.04 53.9 October 12, 1918. . . . . 6,729 47 83 47 .70 36.3 October 19, 1918. . . . . 6,693 27 78 27 .40 21.0 October 26, 1918. . . . . 6,642 50 45 45 .68 35:2 November 2, 1918.... 6,647 39 60 39 .59 30.5 Total . . . . . . . . . .... 33,465 233 361 233 3.48 36.2 Average Payroll .... 6,693 In this example the payroll is fairly continuously diminish- ing. In such instances the number hired rather than the number of separations represent the turnover, for only such losses are replaced as are represented by number hired during the period under observation. During the five-weeks period, in the exam- ple, obviously the turnover percentage is represented by 233, the number hired, divided by 6,693, the average payroll, or 3.48. The factor 3.48 multiplied by %, or 10.4, gives 36.2 as the labor turnover percentage, yearly basis. During the week ending October 26 fifty men were hired as against forty-five lost from the payroll. In the calculation of the weekly turnover percentage for that week, the number of replacements must be considered as forty-five, therefore, and not fifty, although the fifty should be used when calculating the turnover for the five-weeks period. If labor turnover percentages are to be calculated only on a monthly basis, the payroll should preferably be given for the first and fifteenth of the month so as to get as true an average or mean payroll as possible. If this is not feasible, the next best thing is the payroll at the beginning and end of each month from which approximate means can be derived for each month by taking the sum of the two numbers and dividing by two. The simplest formula for labor turnover and one which will serve every practical purpose is as follows: T = R P; +P.+ P. --— P. N * In this formula: T = Turnover. R = Replacements, or number hired to replace losses. P = Payroll. N = Number of Payrolls used in striking an average for the period under observation. 23 A PARTIAL BIBLIOGRAPHY OF RECENT ARTIC ON \ LABOR TURNOVER. The Problem of Labor Turnover—By Paul H. Douglas. The American Economic Review, pp. 306–316, June, 1918. Standardizing Labor Turnover—By Harry W. Kimball. Employment and Labor Maintenance, p. 506, December, 1918. Labor Turnover and Safety—By H. W. Forster. Proceedings of the National Safety Council. Sixth Annual Safety Congress, Part I, pp. 153–155, September, 1917. Also in same report, pp. 200–202, Report of Committee on Standard Forms relating to the Conservation of the Health of Workers. By J. W. Schereschewsky, Chairman, Surgeon, U. S. Public Health Service. Labor Turnover in Cleveland and Detroit–By Boris Emmet. Monthly Labor Review, pp. 11–30, January, 1919. U. S. Bureau of Labór Statistics, Washington, D. C. Labor Turnover—By George J. Eberle. The American Economic Review, pp. 79–82, March, 1919. Methods of Reducing the Labor Turnover—By Boyd Fisher. Proceedings of Employment Managers' Conference, Minneapolis, January 19 and 20, 1916. Published in Bulletin of the U. S. Bureau of Labor Statistics, whole No. 196, May, 1916. The New Apprenticeship as a Factor in Reducing Labor Turnover—By Charles A. Prosser. Proceedings of Employment Managers' Confer- ence, Minneapolis, January 19 and 20, 1916. Published in Bulletin of the U. S. Bureau of Labor Statistics, whole No. 196, May, 1916. Employment Managers Solving a Business Problem—By Dr. Ernest Fox Nichols. The Nation's Business, p. 14, 15 and 17, March, 1916. Methods of Reducing the Labor Turnover—By Henry S. Dennison, of the Dennison Manufacturing Co. Bulletin of the United States Bureau of Labor Statistics, whole No. 202, pp. 56, 57, 58, 59, September, 1916. The Cost of Labor Turnover—By Magnus W. Alexander, of the General Electric Co., West Lynn, Mass. Bulletin of the United States Bureau of Labor Statistics, whole No. 227, pp. 13–28, October, 1917. Also, same Bulletin, How to Reduce Labor Turnover. By Boyd Fisher, Vice-President, Detroit Executives’ Club, pp. 29–47. Also, same Bul- letin, The Tabulating of Labor Turnover. By E. H. Fish, Employ- ment Manager, Norton, Co., Worcester, Mass.; Chairman Committee on Labor Turnover of the Employment Managers' Association of Boston, pp. 50–55. Also, same Bulletin, Determining Cost of Turnover of Labor. By Boyd Fisher, Vice-President, Detroit Executives’ Club, pp. 60–66. Also, same Bulletin, An Actual Account of What We Have Dome to Reduce Our Labor Turnover. By J. M. Williams, Secretary, Fayette R. Plumb (inc.); President, Philadelphia Association for the Discussion of Employment Problems, pp. 173–190. Handbook on Employment Management in the Shipyard. Special Bulletin, Labor Loss, Employment Management Branch, Industrial Relations Division, United States Shipping Board Emergency Fleet Corporation, Philadelphia, 1918. Hiring and Firing—By Lee K. Frankel, Ph.D., Supervisor, Welfare Divi- sion, with the cooperation of Laura S. Seymour. Suggestions for Employers, Industrial Service Bureau Bulletin No. 1. Metropolitan Life Insurance Company, New York, 1918. 24 Absenteeism—Reports from plants in different localities. United States . ºn: of Labor, Information and Education Service, Washing- lº ton, & ... Reducing Labor Turnover in Our Shops—By Fred H. Colvin. American Machinist, pp. 27, 28, 29, January 3, 1918. ..i The Employment Manager and the Reduction of Labor Turnover—By : Thomas T. Read, E.M., Ph.D., New York, N. Y.; Technical Depart- ment, New Jersey Zinc Co., Bulletin No. 130 (October, 1917), 1833. . Transactions of the American Institute of, Mining Engineers (subject to revision) (New York Meeting, February, 1918). Standardigation of the Causes of Leaving Jobs—By J. D. Hackett. Indus- ? trial Management, pp. 233–234, March, 1918. Labor Factors in Our Shipping Program—How to Procure and Maintain : Shipyard Workers and Hasten the Building of our Emergency Fleet. # Roy Willmarth Kelly, Industrial Management, pp. 210–217, March, 918. Methods of Arriving at Labor Turnover—By J. M. Van Harlingen and T. J. Dwyer. Employment and Labor Maintenance, pp. 319, 320, . April, 1918. How to Reduce the Turnover of Labor—Methods Successfully Applied at an Ohio Steel Works—Get the Right Man for the Job and Advance # *; Fit Men. By Ernest C. Gould. The Iron Age, pp. 874–5, April 4, § 1918. Confidential Report Series, Report No. 1–An Initial Survey of the Prob-; lem of Labor Turnover, Issued by the National Association of Cor- poration Schools, compiled in the Office of the Executive Secretary ºf with the Cooperation of the Committee on Employment and Dr. Paul, Kreuzpointer, Investigator. Available to Class “A” Members only, . May, 1918. º The Problem of Labor Turnover—By M. C. Hobart. American Machinist, Vol. 48, No. 20, pp. 821–822, May 16, 1918. º Keeping Track of Labor Turnover—By E. H. Fish. In Plants Where Semi-Skilled Men Must Be Trained to Meet Labor Needs, a Careful Compilation and Analysis of Turnover Records is Worth Many Times: the Expense Involved. Employment and Labor Maintenance, pp. 445, 446, September 12, 1918. t Computing Labor Turnover—A Questionnaire. Industrial Management." Pp. 239–246, September, 1918. . Diagnosing Labor Ills—By William Lamkie. Employment and Labor: Maintenance, p. 502, June, 1918. Why Men Leave Their Jobs—By Henry P. Dutton. Employment and; Labor Maintenance, pp. 147, 148, August, 1918. º: Absentee Record Forms—U. S. Department of Labor, Working Condiº tions Service, Grant Hamilton, Director-General, Division of Labor: Administration. Circular No. 4, March 17, 1919. Proper Systems Reduce Labor Turnover—Wide Variation in Turnoved Percentages Due to Factory Methods—Many Plants Have Over 500 Per Cent. Yearly. Automotive Industries, pp. 619–20–21 and 667; March 20, 1919. º: Labor Turnove, and Industrial Training—Training Bulletin No. 6, U. S. Department of Labor, United States Training Service, Washington. D. C., 1919. . Labor Turnover in Cincinnati—By Emil Frankel. Monthly Labor Review: March, 1919, U. S. Department of Labor, Bureau of Labor Statistics; Washington, D. C. Some Reasons for Labor Turnover—By Harry W. Kimball. Industria; Management, pp. 324–6, April, 1919. -, * ſ º t 3 9015 07324.7974_ | 2 Ll– O Õ OC LLI > Z TD Makers Syracuse. N Y. PAT , 1908 Gaylord Bros. . JAM. 21 § § tº: : f§ : º | º # : § : § #: :§#: ; ; : # : # .* # § | -# i º : §: *** * : A's. ¥º: 3.3% , º, . ::::::::::::::: *:::::::::::::: §: $::::::::::: i : | ;3 * 2, ºft # *:::::::::: rºš - § §§ sº § §§ § à º ;: ; i º § i § ..º º ; # ; i. ; {: § g 3. | #: # ;# % # # tº: