MASTER NEGATIVE NO. 94-821 52 COPYRIGHT STATEMENT The copyright law of the United States (Title 17, United States Code) governs the making of photocopies or other reproductions of copyrighted materials including foreign worlds under certain conditions. In addition, the United States extends protection to foreign works by means of various international conventions, bilateral agreements, and proclamations. Under certain conditions specified in the law, libraries and archives are authorized to furnish a photocopy or other reproduction. One of these specified conditions is that the photocopy or reproduction is not to be "used for any purpose other than private study, scholarship, or research." If a user makes a request for, or later uses, a photocopy or reproduction for purposes in excess of "fair use," that user may be liable for copyright infringement. The Columbia University Libraries reserve the right to refuse to accept a copying order If, in its judgement, fulfillment of the order would involve violation of the copyright law. Author: Church, Alexander Hamilton Title: The proper distribution of expense burden Place: New York Date: 1913 ^M-szisz-n MASTER NEGATIVE # COLUMBIA UNIVERSITY LIBRARIES PRESERVATION DIVISION BIBLIOGRAPHIC MICROFORM TARGET ORiGINAL MATERIAL AS FILMED - EXISTING BIBLIOGRAPHIC RECORD jf(aes» t34732 Ohurolu A^lexander^ Hamilton, 1866« I • •• Th« proper distribution of expense burden^. •• New Tork» The Engineering ASgasine oo«» 1913 •"' 144 p« 19 om« (industritl nanageyent library 'The oontents of this book appeared originally in the Engineering nagasine as a series of arti- oles«* Pref. RESTRICTIONS ON USE: TECHNICAL MICROFORM DATA FILM SIZE: 35 IV? t»^ REDUCTION RATIO: . 12: 1 IMAGE PLACEMENT: lA (ua) IB MB DATE FILMED: (plZokf INITIALS: TRACKING « : //ISM 0/S3I FILMED BY PRESERVATION RESOURCES. BETHLEHEM. PA. > CO .^: .^/ a^ en 3 3 CT O > ^a Is 13 7— en cnx ^-< OOM O ^ '^S'.. CJl 3 3 > o m CD O ^ O O CO X < N o: w: .v'V ■a a''' ^'V. 'V:. %7^ ^ e^ ^ o o 3 3 Ul o 3 3 s 3 3 O do O^ 00 S K5 I ro 1.0 mm 1.5 mm 2.0 mm ABCDffGMIJKLMNOPO»STUVWXV; abcdemnopgrstu67890 ABCDEFGHIJKLMNOPQRSTUVWXYZ abcdefghijklmnopqrstuvwxyz 1234567890 ABCDEFGHIJKLMNOPQRSTUVWXYZ abcdefghijklmnopqrstuvwxyz 1234567890 ABCDEFGHIJKLMNOPQRSTUVWXYZ ^ ^ abcdefghijklmnopqrstuvwxyz 2.5 mm 1234567890 .^^ *> V ^o ^fcP ^^ ¥a^ V .** ■^ ^^ ^^ ¥rf. fp ^'Sr m o ■o m -o > C c*> I TJ ^ m 30 o m -A. '' '*«■. ^1%. '^1^ ■4^ V << 5^ l§ °i M CO roP c»rsi S fo ^f^ 1— • IS> CJI o E = 0> cr o >> I? 3 J— •■*■ Z *< ::d »^ C/) •——I N>C oi< o>x ^-< CX>M VO o '■f^ ^^ i '4 ^ 1 1 f^ f • 1 j ^n< li m ♦It .;-rj;**i ■^■.*;-.jy't;i*'ilrte;UH.>'^i*-v; i t jl " -ntirtiitfi,iiiiiiiiimij|i W"-=M\^i4-iin^^^i>uiiutni4itUtHiiuiinm:Hii ;:'e:; :^«»flw*fiiiiiiiiiiip' -'■s-;yitiufm-''-ii4M irt,.«,H;B^iiM»**»'^**''**^^ «*-.—— Columbia ^niberjtitp cc LIBRARY School of Business u INDUSTRIAL MANAGEMENT LIBRARY THE PROPER DISTRIBUTION OP EXPENSE BURDEN A. HAMILTON CHURCH NEW YORK THE ENGINEERING MAGAZINE CO. 1916 Na O ij^ .i/**! Copyright, 1006 By JOHN R. DUNLAP Copyright, 1912 By THE ENGINEERING MAGAZINE CO. WAVERLY PRESS Baltwou. U. & A. • J- PREFACE. The contents of this book appeared originally in The Engineering Magazine as a series of articles. These at once took rank as a standard reference work on one of the most difficult ques- tions of cost-finding; and a steady and continued demand for the numbers of the Magazine (now out of print) in which they were contained has led to the republishing of the entire group, here revised and adapted for presentation in volume form. The accurate distribution of general expense is admittedly one of the most perplexing, but yet one of the most important, problems with which the manufacturer must deal. The simple but thorough analysis conducted in this volume, and the clear, conmion-sense demonstration pre- sented, will furnish a reUable guide to the solu- tion of highly complex conditions in factory accounting. Much of the published literature in this field has been purely descriptive, and has gone little farther than to present specialized adaptations employed in certain individual shops, and per- PREFACE haps not well suited to any but the one estab- lishment for which each was designed. Mr. Church's material is of far greater value. He is not concerned with the size, ruling, or print- ing of forms and cards — matters which should be designed by the accountant to fulfill his special piupose. He goes to the root ideas of cost-finding, and lays down broad principles by which safe and reliable figures may be obtained for machine, piece, and job costs. These principles will properly distribute all expenses of manufacture, marketing, and man- agement, so that the truth may be known as to the profit or loss of any line of product, and changes in manufacturing cost from time to time may be instantly detected and the cause discovered. With increasing competition and increasing complexity of manufactured output this knowledge is indispensable. The Editor. CONTENTS. Chapter I. Interlocking General Charqes with Piece Costs The Growth of Cost Accounting — Competition Makes Accurate Knowledge Urgent — The Expense of Accurate Cost-Finding Justified — Its Complexity Varies with the Business — Its Advantages Demon- strated — Direct and Indirect Wages — Job Costs the Starting Point — The Elements Composing Job Cost — ^Prime Cost, Shop Cost, Inclusive Cost, and Selling Price — The Machine Rate — Expense Apportionment often the Most Important Component — Its Impor- tance Demonstrated 9 Chapter II. Distributing Expense to Individual Jobs What Goes into Indirect Expense — Methods of Apportioning Indirect Expense to Work — The Per- centage-on- Wages Method — Its Fallacy Pointed Out — Conditions under which it may be Safely Used — The Hourly-Burden Plan — Its Principle Illustrated — Its Results Differ from the Preceding — Examples — Important Features and Weaknesses Pointed Out — ^The Machine-Rate Method — Its Antiquity — Gen- erally Unsuited to Modem Conditions — Reasons for Its Unsuitability Indicated — General R(Ssum6 — The Elements of an Ideal System — The Shop Considered as an Assemblage of "Production Centres" — How this Clears Up the Problem 31 6 CONTENTS CONTENTS Chapter III. The Scientific Machine Rate and the Supplementary Rate Essential Differences from the Old Machine-Rate Plan — Examples Worked Out — The Distribution of Charges to Production Centres Explained — What Constitutes the "Supplementary Rate" — The Effect of Idle Machinery Exhibited — Full Outline of the Plan Revealed — Its Working Shown in Typical Cases — The Principles Restated 55 Chapter IV. Classification and Dissection of Shop Charges The Fallacy of General Averaging Pointed Out — Discriminating Apportionment of Expense Among Production Centres is Necessary — It is not Difficult — The Various Items Discussed — Floor Space, Light- ing, Power, Machinery, Supervision — The Cost of Tools — Wages of General Foremen — Depreciation — Repairs — Periodical Readjustment by Latest Rec- ords — The Workings of the Supplementary Rate Demonstrated — Their Correctness Proved 77 Chapter VI. Appobtionicbnt of Office and Sbllinq Expense Items Entering into Expense Burden after Pro- duction is Complete — ^The Difficulties of Appor- tioning Them to the Product — General Averages Misleading— Possibilities of Approximately Correct Distribution of Office and Selling Cost— Basis for Distribution Commonly Used — Examples of Clas- sification of Product for Distribution of General Expense— Tabulation of Items and Their Apportion- ment to Various Classes of Product — Value of the Results Secured— The Features and Effects of the Proposed System Reviewed 123 Chapter V. Mass Production and the New Machine Rate The Method Applied to Automatic, Semi-Auto- matic, and Repetitive Work — The Special Problems Which Appear Here — How Idle Time is Adjusted — Apportionment of Idle Time Charged between Indi- vidual Jobs and the Supplementary Rate — Adapta- tion to Mixed Shops — Differences between the Old and the New Methods Summarized — Correct Figures and Examples from Actual Shop Work — The Ac- counts Presented, Analyzed, and Explained 101 II Ill II! Ill INTERLOCKING GENERAL CHARGES WITH PIECE COSTS. THE PROPER DISTRIBUTION OF EXPENSE BURDEN. Chapter I. INTERLOCKING GENERAL CHARGES WITH PIECE COSTS. IXTRITING no less than sixty-eight years ^ ^ ago, Charles Babbage said :* " The great competition introduced by machinery, and the application of the principle of the sub-divi- sion of labor, render it continually necessary for each producer to be on the watch, to dis- cover improved methods by which the cost of each article can be reduced, and with this view, it is of great importance to know the precise expense of every process, as well as of the wear and tear of machinery which is due to it." And in a subsequent chapter dealing with the "causes and consequences of large factories," he points out that in reference to any particular manufacture there will be a • Economy of Manufacture, by C. Babbage, M.A., Lon- don, Knight, 1832. 11 12 THE DISTRIBUTION OF EXPENSE BURDEN GENERAL CHARGES AND PIECE COSTS 13 • certain minimum proportion of indirect expen- diture for supervision, lighting, clerical work, passage of materials from process to process, repairing, etc., which cannot be exceeded in any similar factory without a less efficient pro- duction resulting. This is probably the earliest reference made by any pubhc writer either to costs or estab- lishment charges. Yet notwithstanding that the importance of such matters was realized by Babbage so early in the development of manufacturing industries, the lesson for vari- ous reasons fell upon deaf ears, and it is only today, when the principle of competition has attained its full growth and has become a matter of life and death, that the commercial organization of manufactories is felt to have become a matter of prime urgency. In the present book an attempt will be made to thresh out the question of establishment expenditure and its interconnection with costs, not merely of the output as a whole, but of the piece and process in detail. To many persons who have not given sufficient attention to the matter, such a problem will present itself, in the first place as hopeless, and in the second as superfluous. Show a person of this type how he can reduce his pay roll by $1,000 a year. and he will listen with respect. Hint to him that knowledge is power, and that absence of knowledge is weakness, and he will be alarmed, not at his own ignorance, but at the prospect of what he terms "unreproductive expenditure." A firm that will spend its money freely on plate glass and mahogany in its offices, often shrinks from the expense of getting to know what is going on in its shops. Why is this? The true answer probably is that one is a familiar and conventional item of expenditure, and the other is "something new." One is an obvious and tangible asset, the other an easy subject of criticism for the irate share- holder who remarks "they have nothing of this kind of thing at so and so's." To spend money today for the sake of saving a much larger amount at some future period, even where the certainty is admitted, demands some moral courage; but to expend money for the purpose of raising efficiency, detecting waste, and preventing loss, demands strong judgment and confidence as weU. Hence innovations in th|s department are made much more slowly than changes in the shops. The difficulty of deaUng adequately with establishment charges in their relation to piece and process is usually in prof>ortion to the heter- 14 THE DISTRIBUTION Of EXPENSE BURDEN GENERAL CHARGES AND PIECE COSTS 15 ogeneity of the business carried on. In highly speciahzed businesses working on modern lines, it probably costs no more to maintain a close inter- connection of charges and costs than to take out prime cost in any form. But in older places, where all sorts of things from steam engines to twist drills are included in one output, the diffi- culty is much greater, but on the other hand the need for and utility of such a connection is greater in equal degree. At what stage in the growth of a business from very small beginnings a fully developed cost and expenditure system becomes a neces- sity is difficult to say. But that it should be introduced as early as possible is desirable for reasons imconnected with its immediate use. In a small works under the personal super- vision of the proprietor, who is able to give so much of his time to the affairs of the shop that he is fully conversant with the progress of work through it day by day, a cost sjrstem of any kind has but little present value. But as it is in the nature of new businesses to expand into large ones, and that sometimes with consider- able rapidity, with a consequent loss of grip of detail on the part of the proprietor, it becomes very important that records shall be available representing what was accomplished in the shop at the most vigorous period of its life, namely, when the need for expansion was becoming felt. It is not uncommon experience that expan- sions do not always produce the satisfactory effects anticipated. During changes of this kind something has evaporated, it is not known what, but the absence of which is recognized keenly enough in its practical effects. The real element varies, of course. It may be a less intense watch on the part of the proprietor, or it may be, with still greater probability, a dislocation of the previously existing relation- ship between work and the incidence of expen- diture, which again may be due to a variety of causes that must needs be recognized before they are remedied. We have not, however, to consider the case of the private proprietor alone. Probably to- day the larger number of new businesses are joint-stock concerns from the very commence- ment cf their career. Innumerable companies are formed for the exploitation of some par- ticular patent article, not infrequently imder the supervision of the patentee, who presently demonstrates his unfitness for practical busi- ness management. How many vicissitudes do concerns of this type undergo, due to changes of management! Every capitalist with a taste for ill < II ^1 16 THE DISTRIBUTION OF EXPENSE BURDEN financing inventions can call to mind exam- ples of this class. Every one can remember with what hopes and fears new blood has been introduced, and with what anxiety the next half-yearly or yearly balance was looked for, to lay bare the result of the change. It is not yet recognized how immensely the task of a competent directorate, to say nothing of the advantages to a stranger coming fresh to a business to take over and reorganize its work- mg, would be faciUtated by the adoption, from the very first, of a thoroughly comprehensive method of recording shop work, including the connection of expenditure of all classes with the items of output on which they are incident. No board would think of carrying on opera- tions without a set of conmaercial books ar- ranged by an accountant in whom they had confidence; yet the equally important technical records are nearly always treated in an amateur spu-it, with the result that no two imdertakings collect this information on similar Unes. It is commonly left to the taste and fancy of the manager to say what records, if any, shall be kept, and how far they shall represent anything real and useful. So far is this attitude carried that many otherwise experienced business men really and conscientiously do not beUeve that shop accounts are of any use at all. GENERAL CHARGES AND PIECE COSTS 17 Yet the same men would undoubtedly give short shrift to the cashier who shirked the detail- ing of his petty-cash account on the groimd that to keep track of small detailed expenses was not possible, or at any rate not worth doing. Nevertheless, the possible waste represented by an undetailed cash accoimt is as nothing to the possible waste represented by an undetailed wages roll. And still more important is it to remember that, as the sum of cash expended is made up of hundreds of small items, which are and should be analyzed and classified so as to give an intelligible idea of the wisdom of the expenditure, so the productive activities in the shops are made up of innumerable small jobs, which, if properly marshalled, wiU tell precisely the same tale as to the wisdom of the expenditure on them as the money payments in the cash analysis. It is true that the broad results of a half- year's work can be read in unmistakable figures in the balance sheet. But the mischief is not only done by that time, but in the absence of proper shop accounts, it cannot be ascertained where is the element at fault. To introduce reform one must first know where reform is necessary. It is no answer to this to say that practical experience suppUes the deficiency. A •w Ml 18 THE DISTRIBUTION OF EXPENSE BURDEN man with a file and a true plane surface can supply a duplicate of that plane surface if he has sufficient skill and works long enough, but he will produce the same result more quickly and surely if he has a machine tool of the high- est class to aid him. And in proportion to the accuracy of the machine tool so will be the ease and speed of the perfonnance. A modem sys-r tem of organization is a high-class machine tool. It can be done without, but not economically That is all there is to it. The wise man will make his own choice. ' Thus far the argument has been directed to the desirability of a modem method of shop accounting, even though only of the class which merely records, and does not work up the results into new forms. Attention will now be directed to the particular subject of this discussion — the connection of work with the expenditure properly incident upon it. It will, of course, be readily imderstood that this further develop- ment implies the existence of a highly organized cost system. It cannot be worked at all in those cases where wages are merely analyzed each week, usually by a more or less forced agreement, into the order numbers on which they have professedly been expended. Such ai> rangement, though admirable for the purposes GENERAL CHARGES AND PIECE COSTS 19 of the commercial accountant, is useless for the technical accountant. And while the results required by the former are readily furnished by the operations of the latter, the converse of the proposition is a clear impossibility. The clearest demarcation exists between direct (viz., allocable,) and indirect wages — be- tween the wages of fitters and planers, and those of foremen, timekeepers, and messengers. But between direct and indirect expense, the distinction is not so obvious. Workmen's time is not without exception and at all times being employed on orders for customers or stock. Whilst it is evident that some wages never become direct expenditure on orders, all classes of wages are liable to become indirect expendi- ture. The same thing may be said of material. For this reason the unit selected from which all the subsequent erection is built up, is neither the wages as such nor the material as such, but the job, A definition must be given of what will throughout these chapters be meant by the term "job." In this respect the distinction drawn by Mr. Slater Lewis* between the job and the works order is adhered to. The works order may, and usually does, consist of a con- 'The Commercial Organisation of Factories." • II' 20 THE DISTMBDTION OF EXPENSE BT7BDEN ii ^\ Ml siderable number of distinct jobs. Practically the job may be defined as the amount of time spent by any particular workman on any par- ticular piece or similar set of pieces. Thus, a works order for a lathe will include such jobs as planing bed, cutting leading screw, milling slide rest, etc. If the works order were for six lathes of similar pattern, the jobs would be extended similarly — as, for instance, plan- ing six beds, cutting six leading screws, etc. In the case of work done otherwise than on customer's stock orders, the job would be for items such as new screw for No. 45 lathe, alter- ing position of band saw No. 67, etc. It will be evident that some jobs will be charged with material and some not. Mere process work is of the latter class. Generally speaking, while the works order sets in motion the activities of all classes of labor, the job is individual to each man, his day's work being made up, it may be of one job, or it may be of several jobs. The whole output of a shop, for any given period, consists of a number of jobs, and of nothing else. When we know all there is to be known about these jobs, we can form a very shrewd idea of the way in which things are going in that shop. Now the elements which enter into the cost a ' II GENERAL CHARGES AND PIECE COSTS 21 of a job are many. To ascertain merely the actual wages and material spent upon it is undoubtedly something, but to consider these factors alone and unqualified as a basis for the comparison of the relative profitableness of work may easily lead to serious misapprehen- sions. In many cases the desirability of introducing a third element into the costs becomes at an early stage obvious to the practical mind. The difference between a small cheap tool and some high-class expensive appliance seems to demand recognition in the costs of the work done on either. In seeking to remedy this, an attempt is frequently made to introduce what is termed a "machine rate" into the accounts — ^a charge per hour for each tool being made and debited to the job, in addition to wages and materials. This is not always an advance, however. While it is easy to see that a charge is desirable, it does not follow that the fixing of such charges is either simple or easy. Usually some arbi- trary proportionate rate is made, supposed to represent the interest on capital sunk in the tool, and this reduced to a rate per hour is charged against the jobs done on that tool. The first objection to this plan is that tools are not in constant employment. But the interest, I ' iil I n 22 THE DISTRIBUTION OF EXPENSE BURDEN on the contrary, does not cease to run because the machine is idle. Hence this rough-and- ready settlement of the problem is deceptive and may even lead to false conclusions. This illustration has been introduced here merely to show, by a simple and familiar case, the general tendency of all attempts to deal with shop establishment charges. It is true that by means of the machine rate no more is usually attempted than to burden particular jobs with some of the interest on the machines concerned. But this, although , as we have said , ineffectual even for the special purpose intended, gives a very definite and clear idea of the prin- ciple involved, viz., the spreading of shop ex- penses over jobs, job by job, upon some recog- nized basis closely connected with the time occupied and the way of doing the work. Even on this rudimentary plan, it is easy to see that the prime cost and the shop cost have become different things. If two jobs take three hours each, the payment for labor being the same in each case, say 20 cents per hour, we may have the respective costs of the process $1.05 and $1.95 respectively, if we assume ma- chine rates of 15 cents and 45 cents in each case. N GENERAL CHARGES AND PIECE COSTS 23 But interest is by no means the only item of indirect expense worthy of being identified with jobs. It is, in fact, one of the smaller of many items. Thus it is that prime cost is in extreme cases not even the larger half of shop cost, and in mass production may even cease to have any tangible value whatever. Before going on to speak of shop expenses in detail and technically, something must be said of their relation to the administrative side of the business. Shop estabUshment charges are, without doubt, in many cases the "lost factor," absence of which makes the difference between success and failure in a commercial sense. There is a wide difference between technical excellence and commercial efficiency. It is not making an\ article that is the object to be attained, but! making it at a price. And where various classes ' of things are being simultaneously made, some- thing more accurate than practical instinct is desirable as a guide when it is required to decide which line it will pay to exploit commer- cially to the greatest degree. In a very small undertaking this practical instinct may be suffi- cient, but we are not considering the case of very small undertakings. The moment the work • t t \ ] f ' d * 24 THE DISTRIBUTION OF EXPENSE BURDEN gets too large in volume for all its minutest de- tails to be carried in one head, the necessity for a close analysis and interiocking of expenses with jobs becomes marked. Nor must it be supposed that the "depart- mentalization" of imdertakings is more than one step, though certainly a very necessary step, in the solution of this problem. To separate the expenses incurred in or in connection with the work or organization of each shop does not usually help us to any knowledge of the interior economy of that shop, if more than one class of articles are passing through it. In the rare case that a department contains machines of one size and pattern, performing exactly the same operations on a single variety of work, it is true that very little would remain to be done, but this condition of affairs rarely obtains. Generally speaking, it remains true that departmentalization is quite a separate transaction from the interlocking of charges with costs. An important fimction discharged by shop cost as distinguished from prime cost is the information afforded to the draughtsman when considering the question of design. The de- signer has to arrange to give full play to the actual conditions of the shop much as an archi- GENERAL CHARGES AND PIECE COSTS 25 tect has to consider the circumstances of the site, by which he is often fettered. It is not a question of producmg a purely theoretical plan free from all limitations of practice that he has to study, but how to arrange to make a given article with a given set of plans, and a given organization, so that the capabiUties of these fixed elements are uti- lized to the utmost. It will be evident that the best chance of domg this exists where the actual shop condi- tions are fully known, and mter-related with the work abeady done. Great discrimination is of course necessary in considering these fac- tors, but it is the natural course of things that as the industry increases in complexity, the intelligence required to conduct it successfully requires more and more careful training and selection. This is a tendency against which it is useless to protest. A not unimportant matter on which the shop- charges question has a very definite bearing is that of valuation for stock taking. When it is considered that the truth of a balance sheet must obtain all its vitality from the valuation of finished work and of work in progress through the shops, the immense importance of a correct basis for such valuation will appear. Unfor- A ''vr. :1 m 26 THE DI8TBIBUTI0N OF EXPENSE BURDEN m tunately, there appears to be no sort of uni- fonnity of practice with regard to this matter. In some cases prime cost is taken as the basis of reckoning; in other cases arbitrary percent- ages are added, and even in these latter there is a divergence of practice. Sometimes they are added to time only, sometimes to prime cost, but in no case, within my knowledge,* is any attempt made to discriminate between various classes of manufactures. Now, a mo- ment's consideration will suflBce to show that any system of general percentage must be most unfair. So must the method of basing valu- ation on simple prime cost. The reason is obvious. All valuation is an attempt to repre- sent certain facts. The facts are undubitably these: the charges incident on a variety of articles as truly represent part of the cost of such articles as the actual direct wages paid on them. And these charges are rarely, it would be safe to say never, identical in theh* incidence on different classes of articles nor are they constant from period to period. Therefore an attempt to represent their value either by ignor- ing this factor of production or by applying an arbitrary increment or percentage eqiially on * Exception must be made, of course, in the case of work« organized on the writer's system. GENERAL CHARGES AND PIECE COSTS 27 all, will produce not any approach to facts, but merely a fancy figure, which will be not even constant in its error. It is, in fact, a guess, and not the less so because based on figures. Arrangements of this kind probably originated the unkind saying that figures will prove any- thing, "except facts." The cost of production, as dealt with in this discussion, will be considered uniformly as divided into three great divisions. First, the bare cost of wages and materials, which will be called Prime or No. 1 cost. Secondly, the prime cost plus the expense of production in- curred in passing through the shops; this I shall refer to as Works or No. 2 cost. Thirdly, the works cost plus the expenses of the commer- cial management and selling organization, which is termed Inclusive or No. 3 cost. This classi- fication will be remembered better by examina- tion of the diagram (Figure 1) which represents the total sale price of an article dissected into its constituent factors. In this diagram it will be seen that the cost of materials and of wages, taken together, make up Prime or No. 1 cost. Material, wages, and shop establishment charges make up No. 2 or Works cost. And material, wages, shop charges, and general establishment charges make up Inclu- I i 11 28 THE DISTRIBUTION OP EXPENSE BURDEN sive or No. 3 cost. This last, subtracted from the sale price, gives the profit on that order. The organization of no works can be con- sidered complete mitil it is able not merely to connect its costs o/ all classes with its jobs, but also to check its financial position by aggregat- ing its profits on sales item by item. Ctf course -Selling Price. (400- r ■ Inelu8iv« or No.3 Coat, |350- -Works or No^ Cost, (250 ' Prime or No.1 Coat. $150 llateria $50 Waffea 1100 Shop Charge HOO General Eitablishment Charge 1100 Profit 150 FIG. I. ANALYSIS OF THE SALE PRICE OF A MANUFACTUBSD ARTICLE the latter process is merely a corollary or deduc- tion from the cost factor. Having ascertained the latter in the form shown by the diagram, the net profit on any sale becomes a trifling matter of arithmetic. Some attention should be given to a clear understanding at this stage of the very definite line drawn between works cost and inclusive GENERAL CHARGES AND PIECE COSTS 29 cost. Works cost, as its name implies, repre- sents the expenditure of all sorts upon the work, up to its delivery into warehottse. It is the cost of production, and of nothing but production. Inclusive cost is works cost plus what the com- mercial arm has spent on it to effect a sale. The distinction will become clearer if we sup- pose the case of a business having its works in the country, and its conamercial office in the metropolis. The expenses of the former would fall into works cost, and the expenditure of the latter would be classed as general establish- ment charges. It is, of course, understood that all the correspondence, packing, etc., was per- formed by the metropolitan office, and that the works has nothing to do but to manufacture. There are, of course, certain expenses which may be doubtful and their inclusion in one division or another is subject to the actual cir- cumstances. Of this kind are draughtsmen. In some cases, the cost of drawings is certainly a works expenditure and chargeable to the job. The same applies to patterns. But, generally speaking, the basis of division is this: if a draw- ing or pattern is chargeable against a given order, it should be treated as part of prime cost, but in all other cases as a general establishment charge — never in any case as a shop establish- ment charge. fi r I 30 THE DISTRIBUTION OF EXPENSE BURDEN The necessity for and the justice of this distinction between shop charges and general charges (works cost and inclusive cost) hardly needs insistence. It has been ably and exhaus- tively argued in the work of Mr. Slater Lewis, already referred to, but the argument may be briefly recapitulated here. Nothing is naturally more distinct than thev operations of making and seUing. They require different instincts and widely separate talents. An imdertaking may be most efficiently organ- ized and managed on one of these sides, and yet be imsuccessful, because what is gained in one] set of efforts^is lost in another. What is more useful, therefore, than to make this natural division of work to be reflected by a similar division in the system of accounting? It is the easiest and simplest of all the modifications with which we shall meet, because it is the most obvious and fimdamental. In the chapters which follow, works cost will occupy nearly the whole of the field. This is because as yet general establishment charges are not capable of any great amount of detailed analysis. But by keeping them separate and distinct, and thus excluding them from affecting the operations of the works, we are enabled to follow the latter in very great detail, and to feel sure of our result. DISTRIBUTING EXPENSE TO INDIVIDUAL JOBS. Chapter II. DISTRIBUTING EXPENSE TO INDIVIDUAL JOBS. A SSUMPTION is made in what follows that -^^ a system of taking out prime costs is ah-eady in use, which summarizes the expendi- ture in each department into: (1), wages and' materials expended on stock or customers'^ orders; and (2), other expenditure, such as repairs, supervision, and similar indirect items of expense. Having obtained the total of indirect expen- diture for any shop, there are several recognized ways of establishing a ratio or proportion be- tween the value of work done in a given period, such as a month, and this indirect expense. This ratio may be established as a percentage on wages, or expenses may be divided over the number of hours worked, or they may be dis- tributed in some approach to a differential method of treatment by means of machine rates varying in value. Each of these methods must be considered in detail. p II T 34 THE DISTRIBUTION OF EXPENSE BURDEN The Percentage-on-Wages Method. Where any attempt is made to distribute shop charges over work at aU, this method is the one most usually employed, at any rate in Great Britain. The probable reason of this is that, though capable of considerable refine- ments, it is as usually employed a ''handy," or in other words a rough-and-ready, method. In order to distribute shop charges on this plan, all that is necessary is to obtain first, the total of charges; the second, total of wages on actual orders during, say, a month in each case, and — ^find the percentage of one to the other. Thus, if in a given month we have spent: Items. Direct Expenditure on Orders. Shop Establishment Charges : Repairs and Maintenance . . Spoiled Work Supervision General Laboring Share of Storekeeping, Time- \ keeping, Etc j Proportion of Rents, Rates' Power, Light Heat, Etc. for this Department Total of Shop Charges 29 Wages. Material. Total. 100 15 2 5 7 56 4 1 156 10 3 $ 7 8 17 89 Total Wages and Material in Shop } 129 61 )* !•*, DISTRIBUTING ^EXPEKSB TO INDIVIDUAL JOBS ^S We find that as against $100 direct wages on order, we have an indirect expenditure of $59, or, in other terms, our shop establishment charges are 59 per cent of direct wages in that shop for the period in question. This is, of course, very simple. It is also as usually worked very inexact. It is true that as regards the out- put of the shop as a whole a fair idea is obtained of the general cost of the work— that is, of its works (or No. 2) cost. And in the case of a shop with machines aU of a size and kind, performing practicaUy identical operations by means of a fairly average wages rate, it is not alarmmgly incorrect. ^ If, however, we apply this method to a shop m which large and smaU machines, highly paid and cheap labor, heavy castings and small parts, are aU in operation together, then the result, unless measures are taken to supplement it, is no longer trustworthy. The reason why, under these conditions, it is no longer to be regarded as a scientific method IS not far to seek. The case of the difference in the value of work done on different machines as regards one item alone, viz., the interest factor, has ah-eady been aUuded to. But this IS not the only point of differentiation. The space occupied by machines varies, the power f 1 36 THE DISTRIBUTION OF EXPENSE BURDEN taken by them varies. And in the case of men at high and low rates of wages, the cheap man takes up as much room as the expensive man, and probably takes much more supervision; large parts, again, use powerful cranes and occupy the time of laborers in their progress about the shops, to say nothing of the room required for their evolutions. Now on the simple percentage plan, it is not only that these factors of difference are ignored, but the basis actually used for the distribution of charges, viz., the cost of the job as meas- ured in dollars or shillings, though apparently a "natural" one, is really quite arbitrary and without any real relation to the actual problem. That this is so will be allowed when it is considered that the effect in the costs of lowering the wages rate of a man, is also to reduce the amount of establishment charges which gets allocated to the job by the system of percent- ages. If this represented anything like the facts it would not be so bad; but unfortunately, the highest probability exists that by substitut- ing a poorly paid man for a good man on any particular job, the true proportion of charges would be actually increased. He would take longer, for one thing, yet as his longer time occupied is counterbalanced by a smaller total DISTRIBUTING EXPENSE TO INDIVIDUAL JOBS 37 of wages, the percentage would not be increased. Again, he would probably absorb more of the foremen's attention, but this would not show in the accounts. In short, the result of the change would really be to increase the cost of the works—the works cost— but the apparent effect would be to reduce it, because the prime cost might be reduced, and the burden of the charges which were before allocated to that job would fall on other jobs with which it has nothing to do. In certam kmds of business it is possible to apply a compensating arrangement. Where the class of work is fairly steady, and the bulk of the work done in any one shop is on different parts of the same kind of machine, certain carefully considered but still to a large extent arbitrary variations of the incidence may be arranged. The work may be divided, for exam- ple, into three or more classes Skccordmg to size of castings, such as heavy, medium, and light, the percentages bemg so adjusted that the heavy class takes x per cent more than the medium class, and the medium class y per cent more than the light class. By this means the whole of the shop charges will be distributed as before over the wages expended on orders, but not uniformly. The heavier work gets burdened f \t 38 THE DISTRIBUTION OP EXPENSE BURDEN more heavily than the light work, and therefore absorbs more, while the light work absorbs less than it would have done without this com- pensation. The method requires, however, very elabor- ate and careful arrangements for determining: (1), the basis on which to divide the ratio of percentages; and (2), an equally close consider- ation of every article manufactured to decide in which class it should be included. Where this is done the method gives satisfaction; but it should not be forgotten that anything like hasty or careless dealing with the varying fac- tors vitiates the value of the whole method. But with proper precautions it is a very con- venient and elastic system, if confined to shops which fulfil the conditions above stated. The Hourly-Bukden Plan. The broad principle of this method, in all its varieties, is based on the consideration that the shop as a whole may be considered as a factor of production, and can be charged for at an hourly rate against the work done on orders. The standing charges, the cost of maintenance, and the shop organization, are summarized as for the percentage plan, but the resulting figure is dealt with quite differently. Instead of spread- J Wages. Material. Total. 29 5 59 DISTRIBUTING EXPENSE TO INDIVIDUAL JOBS 39 ing the shop charges over the cost of individual jobs, as measured by the direct wages spent, these charges are spread over them according to the number of hours during which work has been done on them. Item. Total of Shop Charges (as before.) Total Hours Worked on Orders During Month— 3,000 Hours, — - « $0.0197 per Hour Worked. oUUU Thus, if we have $59 as before, representmg the whole of the shop charges for one month, we require to ascertain the time actually worked reproductively, that is on orders, and reduce this total of $59 to a rate per hour on such re- productive time. There are variations of this naethod, but the principle is the same in all, viz., making an hourly charge for the shop organization and expenses. A httle consideration will show that this method may produce quite a different works (or No. 2) cost from the same set of figures arranged on the percentage plan. It is claimed by the advocates of the hourly-burden method that it is far superior in that a nearer approach to the real conditions of work, i.e., to facts, is realized by making the hour rather than the dollar or shilling the basis of distribution. |i 40 THE DISTRIBUTION OP EXPENSE BURDEN I li « !r / Some of the obvious faults of the percentage system are undoubtedly removed by it. The elimination of wages as the basis of distribution certainly removes the great reproach of the former system, viz., that it makes the works cost on any particular job slavishly follow the prime-cost figures, and therefore tends to make cheap labor bungling over a job for an unneces- sarily long time appear profitable. It brings into full prominence the essential fact that to have work hanging about is a costly proceed- ing, however low-priced the actual labor may be. Perhaps this is the most important fea- ture of the method — it contributes to a clearer view, to more exact picturing, of the immense importance of the time factor in production. But when we consider the system as fulfilling the r61e of a scientific method of apportioning estabUshment charges, it is easily seen to fall short of perfection in a serious sense. Just as the percentage method stakes all upon the single factor of wages, so the hourly-burden plan stakes everything upon the single factor of time. The mere statement of this peculiarity is sufficient to show that we have here a method which sacrifices a great deal for the sake of simplicity. To take one item alone, no dis- crimination between large and costly and small DISTRIBUTING EXPENSE TO INDIVIDUAL JOBS 41 and cheap machines is brought into play. Sim- plicity is a good thing, but as an end in itself is worth nothing. What is required is an exact method of apportioning estabUshment charges, a method which will allow us to obtain figures which are closely approximate to facts, because they follow the actual and known conditions obtaining in the shops, as far as we know how to make them. If any method fails to do this, when it can be done in greater degree by an- other method, its simplicity is nothing in its favor. At the same time, it may be allowed that for particular conditions the hourly-bur- den plan will give satisfactory results. In shops, for example, where there is uniformity in the class of machines employed and in the kinds of work passing through, the simple hourly burden gives substantial accuracy. These, it / will be observed, are also the conditions under ( which the percentage plan has been stated to be apphcable. The reason of this is not the similarity of the two methods, but because in such cases, the conditions are at a maximum of simplicity. In other words, where we have a simple set of facts to represent, their representa- tion is an equally simple matter. We have here an unportant key to the dis- cussion of the suitability of any system. Ahnost 42 THE DISTRIBUTION OP EXPENSE BURDEN I any method will represent simple conditions. But will any given method break down when the conditions become complex? This is the touchstone. As work ia not arranged to suit methods of record, but mce-versa, it is evident that we must seek a method capable of record- ing with approximate accuracy imder the most complex and difficult conditions. We may then feel quite secure that it will also record simple conditions in a simple manner. It is putting the cart before the horse to consider the simple matters first. The Machine-Rate Method. This is perhaps the oldest and most wide- spread of all. It is not a modem method. By this is meant that it belongs to a former period, in which the close accuracy, the monthly bal- ances, and other refinements of accounting had not been heard of, and that therefore it cannot be considered as among the scientific systems of distributing shop establishment charges. In fact, it does not usually pretend to deal with shop charges, in the general sense, at all. The usual function of the machine rate is to burden the work with a charge proportionate to the interest on and the wear and tear of the machine. As usually arranged it is an hourly charge, DISTRIBUTING EXPENSE TO INDIVmUAL JOBS 43 based on the probable life of the machine under full work. In this sense it is intelligible. But cases do exist in which it has been attempted to make it represent more than this, by increasing the amount arbitrarily, so as reaUy to charge part of the shop expenses through it as weU. This is exceedingly bad practice. Why it is bad will be seen presently. The machine rate may be looked upon as belonging to the Silurian epoch of shop accounting. It is theachthyo- saurus of expense systems. It attempts to do m a heavy, blundering' way what is hardly worth doing at all. Considered in its legiti- mate aspect— as a means of distributing the burden of interest and depreciation on any machine over the work done on that machine— ^it does this successfuUy under one set of con- ditions only, viz., if such machine is never for a moment idle. This being its condition of maxi- mum perfection, it foUows that the greater the proportion of idle hours, the less accurate will be the results obtained from the method. A httle consideration wiU lead to the conclusion that, m a slack time, the work done in the shops will be receiving just the same burden as m a busy time, with every machine running fuU. The balance of expenditure not aUocated to jobs IS just lost sight of. !J.-JJ a kaaaaitflB ^ :^ 44 THE DISTRIBUTION OF EXPENSE BURDEN This is a serious enough matter if we consider the machine rate as a charge for interest and depreciation only. But if we look on it, as is sometimes done, as a medium for distributing some (unknown) fraction of the shop charges, the absurdity of the method will be obvious. It will not compare for this purpose with either the percentage plan or the hourly-burden plan. Yet with all its impossibiUty as a modem system, the machine rate has certain practical advantages, shared by no other method. It does take into account just what is missed by the former systems, viz,, the variation in the cost of work done on different typea of machines. This feature, which practical instinct easily recognizes, has secured its siurival. Though it cannot be considered as a good method, it has had its value as a compromise between bare prime cost and a regular apportionment of charges. It tells something definite — some- thing which is valuable, but which is neither as definite nor as valuable as it might be. The fact that it estabhshes a permanent rela- tion between the work and the machine is a valuable feature. By a permanent relation is meant a relation which does not change or fluctuate with conditions of work in the shop. The charge for the machine being always x, DISTRIBUTING EXPENSE TO INDIVIDUAL JOBS 45 whether the shop is slack or busy, brings a very steady factor into account, which is available as a datum of comparison between work done at different periods as no other factor is available. The prime cost, for instance, is no absolute guide for comparison of efficiencies, since rates of wages vary not always in direct proportion to personal output. The shop charges on the two former systems also fluctuate violently, but the machine-rate value remams constant! This is because it is a pure function of tune. It is true that for all practical purposes the time itself is the very factor which it is desirable to know. And remembering this, it is not diffi- cult to see that we can obtain this information without necessarily being bound to a non-fluc- tuating charge. ^ General R6sum6 of the Foregoing Methods. There have now been passed in review the three fundamental methods of allocating shop charges over work so as to produce works (or No. 2) cost. Of these only the first and second give a real No. 2 cost in the sense that all the charges for a given period are spread over pro- ductive work. It will have been seen that there are two di&- tinct ways only of doing this. The charges ,1 It 46 THE DISTKIBUnON OP EXPENSE BURDEN must be distributed either on "wages value" as a basis, or on "time occupied*' as a basis. For ordinary manufacturing operations there is no other alternative. In certain special in- dustries a distribution on imits of weight or measure might be necessary, but we need not pause to consider these. Generally speaking, we have to choose between wages value or time as a unit on which to distribute. , The weak point of the percentage and the hourly-burden system is that it is an averaged result at which we ultimately arrive, and which we try to regard as the true works or No. 2 cost. Now, it is just as wrong to average such charges as it would be to throw all wages paid into one sum, and average this over the number of hours worked, taking the result as a true wages cost. In the case of wages this is seen to be an absurdity, and no question of simplic- ity or convenience would be allowed to stand in the way of collecting the actual wage cost of each job in an engineering works. An average- wages charge is easily seen to be misleading and dangerous. V^Yet in certain cases where only one class and size of article is turned out in one shop, this averaged-wages cost, although several variations of rates may be included in the total sum before DISTRIBUTING EXPENSE TO INDIVmUAL JOBS 47 averaging, is not only useful but is actually used. It works successfully in such cases be- cause of the extreme sunpUcity of the condi- tions. Only one set of factors being concerned VIZ., a smgle process and a single variety and size of article, the average charge for wages sops up mequaUties of rates, which inequalities are not due to exigencies of manufacture but to exigencies of employment. Here we have an exceUent example of a method of accounting, correct within narrowlimits, but obviously break- ing down the moment those limits are exceeded. The impropriety of applying such a system of averagmg wages to a shop in which several processes, different grades of workers, and sepa- rate sizes and classes of articles are concerned needs no argument. It can be realized at a glance. But this case of wages is exactly on all fom^ with the case of shop establishment charges. Under very simple and uniform con- ditions a method of distributing estabHsh- ment charges which averages such charges per doUar of wages or per hour of time is correct, but m proportion as we depart from these sim- pie conditions we also depart from accuracy. This is not a mere theoretical and pedantic distmc ion. These shop charges frequently amount to 100 per cent, 125 per cent, and even «. 48 THE DISTRIBUTION OF EXPENSE BURDEN IC {: i L ■■I much more of the du'ect wages. It is therefore often actually more important that they should be correct than that the actual wages cost should be correct. If we have to put a dime and also a quarter in a certain collecting bag, it is certainly more important that the quarter should not go astray than that the dime should be looked after. It is true that by no possible method can we avoid averaging shop charges to a certain ex- tent. But we can reduce this extent to a mini- mum, thereby attaining a high average of accu- racy. In the end we have, of course, to choose between one or other of the alternatives already mentioned, viz., wages or hours, for the distribu- tion. Yet by avoiding the attempt to force simplicity of results into the representations of conditions which are by no means simple, and by letting our method follow the natural lines of the actual complexity, we can arrive at results which are no longer averaged to any- thing like the same extent, but are on the con- trary highly differentiated. The Elements of an Ideal System. A little preliminary consideration may be given to the features which might be looked for in an ideally perfect method of interconnecting S 1 DISTRIBUTINa EXPENSE TO INDIVIDUAL JOBS 49 establishment charges and work. Having en- tirely cleared one's mind of any traditions of what IS usual or conventional under the averag- ing regime, it becomes visible that several of the Items of shop charge are naturaUy con- nected with the use and employment of property or plant, and are in the nature of a rent paid for these. In this categoiy are:— Rent, taxes, and msurance on buildings, interest, depreciation on machines, on cranes, shafts, motors, etc. Otheritems are connected with other factors of production. Power, with the use of same by machmes; cost of lighting or heating, with axea. of floor space usuaUy lighted or heated-in short, not to go into detail at this stage, it is readily seen that a large number of shop charges are by no means general in their real nature but can be narrowed down to definite points of mcidence. In an ideal system, it would therefore be expected that this narrowing down should be earned aa far as it was practically profitable to do so, and that only such expenses aa were whoUy general and could not by any reasonable analysis be connected with definite points of incidence, should be treated as general shop charges, and therefore left to be averaged on the former basis. This looks complex. And 1. 50 THE DISTRIBUTION OP EXPENSE BURDEN *1 i m 51 undoubtedly under the most favorable circum- stances the preparations are complex; but, as will be seen, a working method (which after all is the only thmg in which simphcity is important, or at any rate essential) can be de- vised which is not more troublesome than other methods of distributing shop charges. Before gomg on in the ensuing chapters to explain the prehminary steps and the methods of working, an idea will be given of the new point of view on which the method advocated is based. We have seen that it is a very sunple matter to represent the conditions of a shop in which the work is uniform and all the conditions con- stant. All the charge factors, whatever their real point of incidence, press equally on all por- tions of the work in such a shop, and there is therefore no differentiation possible between different portions of the work. In a mixed shop, on the contrary, these factors do not press equally— power, floor space occupied, with its burden of rent, interest, insurance, lighting, etc., interest on capital outlay for machines — all vary as between jobs. There is no justifica- tion for treating such charges as an average if it can possibly be avoided. Hence it is evident that in the distribution of such charges the shop can no longer be considered as a unit, but some- DISTRIBUTING EXPENSE TO INDIVIDUAL JOBS 51 thing smaUer must be taken. Of course, to subdivide the shop into smaller areas would be no solution of this problem. The mixture of the factors of incidence would be as great in half a shop as in a whole shop. A new point of view is necessary. It becomes desirable to look at what the expression ''shop" reaUy signifies. If we regard the shop as an organic whole, (which it no doubt is, or should be), we get no new light. We may perceive clearly enough the cross-fire of factors of incidence bearing unequally in different parts of It, but are no nearer their disentanglement. But if we regard the shop as a collection of small "production centres," each differing from the other, with certain common con- necting bonds which are alone the average or general factors of incidence, then the problem suddenly becomes clear. If, instead of tum- mg our energies to cover up the natural differ- ences between these ''production centres" and make an average of them, we devote our atten- tion to giving the fullest play to this difference, we may claim to have subdivided the shop in a new sense, and to a new end. A production centre is, of course, either a machine or a bench at which a hand craftsman works. Each of these is in the position of a ht- I I 1 ■• , 1 ! c, 52 THE DISTRIBUTION OF EXPENSE BUBDSN tie shop carrying on one little special industry, paying rent for the floor space occupied, inter- est for the capital involved, depreciation for the wear and tear, and so on, quite independently of what may be paid by other production centres in the same shop. Then, in addition to this, there will be a separate debit representing those items of incidence which can only be treated as an average all-round charge. The cost of each production centre is, of course, laden onto the work by a rate per hour. Here we seem to have come back to a machine-rate method, subject to all the defects of that method, par- ticularly as regard the idle hours. If the new process ended here it would be of little use, but it does not end here. It will be seen, when this portion of the sub- ject is reached, that whilst the rate for the pro- duction centre is kept constant, the waste of resources due to idle production centres is not lost sight of, but finds its due place in the works cost simply enough, and in such a way that it has a special significance of its own. One prominent advantage of the plan of re- ducing each shop to its constituent production centres as a basis for distributing estabUshment charges should be noticed. The most heteroge- neous processes may be carried on side by side. DISTRIBUTINQ EXPENSE TO INDIVIDUAL JOBS 53 Having narrowed down the greater portion of the shop charges to their fundamental points of incidence, we need no longer be afraid of any conditions, however complex. Unhke the aver- aging methods, almost any actual workmg con- ditions can be faithfully represented on this system. This is because each production cen- tre is virtually independent of any other, and therefore complexity is indifferent to it. It will be seen, for instance, that it is applic- able to heavy engineering work, to mass pro- duction, and even to that most complex class of work, in which ahnost-automatic machine work side by side with ordinary machine tools. I L'* ''if ' THE SCIENTIFIC MACHINE RATE AND THE SUPPLEMENTARY RATE. Chapter III. 1 (I I ^ r THE SCIENTIFIC MACHINE RATE AND THE SUPPLEMENTARY RATE. TN the preceding chapter, an outline of a new •■• basis of determining the incidence of shop establishment charges was given. Before pass- ing to consideration of the preliminary measm-es necessary to its introduction, the effect and results of the method will be detailed. The apparent similarity of the system to the old device of the machine rate has been alluded to; the points of difference will now be brought into prominence. It will be remembered that the principle on which the proposed method was stated to be based, was that of dividing each shop up into its constituent ''production cen- tres" and giving full play to the natural differ- ences between these as far as practicable, instead (as on the averaging plan) of throwing them into one conamon receptacle or lump sum of shop charges. It was also pointed out that this of itself would be subject to all the defects of the old machine-rate method, inasmuch as the 67 58 THE DISTRIBUTION OP EXPENSE BURDEN problem of idle hours is still unsolved. There is also the problem of those actually general charges, which cannot by any reasonable allo- cation be connected with particular production centres. Leaving these latter out of account for the moment, it will be seen that, admitting all machines or production centres to be running full time, it would be possible so to adjust the rents or machme rates payable for the use of these centres during one hour, that at the end of any period, say a month, the whole of the shop charges would find themselves attached to particular jobs, according to the time taken by each job. TOTAL OF SHOP CHARGES FOR 1 MONTH « $200. Working Machine. Hours per Month. A 200 B 200 C 200 D 200 Rate per Hour on New Plan. 40 Cents 30 Cents 20 Cents 10 Cents Monthly Earnings on New Plan. 180.00 60.00 40.00 20.00 rv r\ L ,'^T. lui^^nts 20.00 $^ ^"^ ^ ^^ ^°^ machines, 800. Hourly burden =-gQQ = 25 cents. ctZ'i^imoO^''*^'* ^^ bourly-burden plan, 800 hrs. @ 25 In the above schedule, there has been given the case of a certain shop, m which the monthly MACHINE RATES AND SUPPLEMENTARY RATE 59 total of shop charges was $200. The tune made by the constituent production centres, viz., ma- chines A, B, C, and D, was 200 hours each, or 800 in all. Now on the hourly-burden plan this would give an average hourly rate of 25 cents for all jobs, on whichever machines they happened to have been done. But it is assumed that the application of the new method and thereby the separation instead of lumping together of the incidence of charges has resulted in the rates of machine A being fixed at 40 cents, of B at 30 cents, of C at 20 cents, and of D at 10 cents. Again assuming the full number of hours to be worked, it is easy to see that we get considerable differences in the resulting incidence of shop charges on the two plans. The charge, for instance, on a job which occupied 5 hours of machine D and 1 hour of machine A would be: On hourly-burden plan— 6 hours at 25 cents = $1.50. On new plan— 5 hours on D at 10 cents = 50 cents; 1 hour on A at 40 cents = 40 cents: total, 50 + 40 = 90 cents. We may now consider what would be the incidence under the percentage-of-wages method of distribution. Let it be assumed that the wages of operators on A and B machines are I f 00 THE DISTRIBUTION OP EXPENSE BURDEN 20 cents per hour, and of operators on C and D machines 15 cents per hour. Reckoning, as before, 200 hours per month, for each machine, we find: — Total wages for one month, |140; and as shop charges are $200 this gives an average of 142 per cent on wages-value for all work done, by whichever operator or whichever machine concerned. Hence we get the charges incident on the job we have been considering. — 6 hours of operator on D at 15 cents - 76 cents, on which charges at 142.9 per cent = J1.07. 1 hour of operator on A at 20 » 20 cents, on which charges at 142.9 per cent = 28 cents. Total charges for job $1.07 + $0.28 = $1.35. We have now worked out this job on three methods, and find different results in each case, viz: — On hoiu-ly-burden plan $1 . 50 On new plan 0.90 On percentage method 1 .35 Which of these is most correct? The nearness of results obtained by the hourly-burden and the percentage plan makes these two methods apparently support each other. But this in reahty is due to the fact that liACHINS RATES AND SUPPLEMENTARY RATE 61 we have assumed a very shght difference (only 5 cents) between the highest and lowest wages rate. This contributes to make the real basis of distribution as nearly uniform in this case as the hourly burden. By this is meant that the value of an hour having been made nearly uniform, it is much the same in effect whether we distribute on hours or value of hours. It will be recognized that in proportion as wages rates differ so will the discrepancy between the distribution on the hour and on the hour- value increase. Between these two methods there is already a difference of 15 cents, which, as just shown, might easily have been much greater if the difference in the assumed wages rates had been made larger. To begin with then, let us ask which of these two may be con- sidered the more accurate. The discrepancy being clearly due to the introduction of the wages factor as a measure of incidence, we need only inquire whether this lowering of the charge under the percentage system is a nearer ap- proach to truth than the other. If we assume that the $200 of shop charge is whoUy made up of standing expenses, rent, interest, depreciation, and such-like, we must decide in favor of the nearer correctness of the hourly-burden plan in principle. It at least is I II lit ^ THE DISTRIBUTION OF EXPENSE BURDEN based on an invariable basis, which, though not representing the real state of affairs, is constant in its error. But we can easily see that the other result has more of the element of chance in it. It will vary according to just what man has happened to be on the particular job. And if the worker did not receive any wages at all, the charges would be reduced to nil. This objection still holds, though in the par- ticular job cited the percentage plan has given us a figure nearer to that provided by the new plan than the hourly-burden figure, which we have notwithstanding argued to be the more correct in principle. This paradox is explained by the consideration that had by any chance the wages not been scaled proportionately to the machine rates, the discrepancy would be far greater between the new and the percentage, than between the new and the hourly plan. In the case of a low-waged operator working a highly rated machine, and vice versa, the dis- crepancies would plainly pass from positive to negative on the percentage plan, whilst on the other two plans they would not vary. The fault of the percentage plan is, then, that it fluctuates according to a factor which cannot be considered as really representing a true and \^ corresponding variation in the incidence of MACHINE RATES AND SUPPLEMENTARY RATE 63 shop charges. The fault of the hourly-burden plan is that it does not vary at all. As between these two evils, our preference must be for that one which produces the most constant error. It has been assumed in the above argument that the new plan does produce a result not merely relatively more correct than either of the two others, but that it gives a normal re- sult closely connected with actual conditions of work. Remembering that we are still discuss- ing the case of a shop in which all machines are making full tune, it is difficult to see how we are to avoid coming to this conclusion. It is evident that the standing charges, at any rate, comprising such items as interest and deprecia- tion on particular machines, the proportion of rent, taxes, insurance, heating, etc. (which may be considered as reduced to a certain uniform charge for each square yard of floor space, and thus debited according to the working space occupied by any machine) and all charges aUied to these, can be as properly assigned to particular production centres as in the case of rents in large buildings which are let out in subdivided areas to tenants. No sophistry is needed to assume that these charges are in the nature of such rents, for it might easily happen that in a certain building y/ :^i\ I 1 64 THE DISTRIBUTION OF EXPENSE BURDEN a number of separate little shops were estab- lished, each containing one machine, all making some particular part or working on some partic- ular operation of the same class of goods, but each shop occupied, not by a wage earner, but by an independent mechanic, who rented his space, power, and machinery, and sold the finished product to the lessor. Now, in such case, what would be the shop charges of these mechanics? Clearly they would comprise as their chief if not their only item just the rent paid. And this rent would be made up of : (1) interest; (2), depreciation; (3) insurance; (4) profit on the capital involved in the building, machine, and power-transmitting and generat- ing plant. There would also most probably be a separate charge for power according to the quantity consumed. Exclude the item of profit, which is not in- cluded in the case of a shop charge, and we find that we have approached most closely to the new plan of reducing any shop into its con- stituent production centres. No one would pretend that there was any insuperable difficulty involved in fixing a just rent for little shops let out on this plan. Nor can there be any more diflficulty in fixing the proper charges for each production centre in a large shop. The little MACHINE RATES AND SUPPLEMENTARY RATE 65 shops are, in fact, a large shop when we take away the partitions which divide them. The' presence of these partitions certainly seems to make the problem more definite and concrete, but this is a mere illusion. There is a new pomt of view involved, which, once grasped, enables us to see separately things which otherwise have a look of conglomeration. This illustration of the Httle independent shops is also useful in helping to discuss the question of idle hours. In each of these little shops, whether idle or busy, the rent is gomg on all the time. Only the charge for power might conceivably be stopped whilst the shop was shut. Now if the mechanic in one of these shops is only working half his time, it is pretty clear that he must double his rate of distribution on such work as he does, if he is to distribute his month's rent over his month's work. But, instead of doubling his rate, he might continue to charge at the normal rate, and then at the month end find how much he was yet in arrear, and distribute this undistributed amount only as a supplementary rate. In this way he would benefit in two ways. First, his normal rate being constant, he can compare costs of jobs worked on at different periods; and secondly, the amount of the supplementary rate, (its II ■I;*- i 06 THE DISTRIBUTION OF EXPENSE BURDEN .1 n ratio to the normal rate,) will serve as a very accurate barometer of the conditions under which he has been working at any period. If the normal rate has been so fixed, for instance, that when fully employed the supplementary rate was nil per cent of the normal, then if he observes it steadily rising 5 per cent, 10 per cent, 15 per cent, etc., he may be fairly sure that things are not going well with him. Of course, in the case of a little shop this would be perfectly clear to him without the aid of the rates to tell him so; but if we, now, throw down all the partitions and make the hundred sepa- rate Uttle shops into one large general shop, this information may be very important indeed to those in authority. The meaning of the term supplementary rate will now be understood. It is this which is the invaluable complement to the machine rate and which makes the great distinction between f the new method and the old machine-rate / method. The supplementary rate is the undis- I tributed balance of shop charges due to idleness \of productive centres, ^ ^ In the illustration just given of the little shops, it will be seen that the imdistributed balance due to slackness of one machine is thrown onto whatever work is done at all on MACHINE RATES AND SUPPLEBfENTART RATE 67 that machine at some time or other. This is, of course, necessary in the case of independent httle shops; but when the partitions are down, and we consider the case of a bona fide large shop with a number of production centres, this is no longer always possible, or even desirable from some points of view. The idleness of a machine may or may not be considered as the fault of that machine. If, for instance, a machine was found to be idle nine- teen-twentieths of its time, this might be due to one of two causes. Either the process was rare but essential, or the machine itself was largely superfluous. In the first case it would be emi- nently fair that the charge should be made very high when it was put in use, since the shop charges due to its presence and up-keep are indubitably incurred for the sake of this occasional use. In the second case it might be rather a matter of accommodation that the machine was retained at all, in which case the shop as a whole should bear the burden, and not the unlucky piece of work that should happen to be put on such machine at any time. Here we have one of those matters of principle which are incapable of a general solution. The procedure proper would vary according to actual circumstances. Nevertheless, the general prac- I* ' Lf .1 •I? t f 'ii : 68 THE DISTRIBUTION OF BXPKNSS BURDEN tice will probably be that idle machines shall get rid of the burden undistributed during their idle periods, not by charging it especially in their working periods, but by transferring it to the general shop-charges account. If we adopt this view of the usual destination of idle hour's burden, then we introduce at once a striking difference between the case of the little independent shops and the same ma- chines considered as production centres of one large shop. In the former, idle time accumu- lates against its own machine; in the latter, all imdistributed charges fall, not into the ^up>- plementary rate of particular machines, but into a general supplementary rate for the whole shop. It is also into this supplementary rate that the really general expenses of the shop fall. The full outline of the new plan may now be seen. First, we consider each machine as an in- dependent production centre, allocating to such centres all the expenses and charges which can, on reasonable analysis, be considered charge- able as a composite rent dji\ machine rate for all the factors of production therein concerned. Second, we charge to a monthly shop-charges account all charges whatever incurred by that shop, including all the items specifically repre- sented in fractional detail by the machine rates, MACHINE RATES AND SUPPLEMENTAKT RATE t)9 and also including, of course, such general Items as cannot be represented in the machine rates, of which the most obvious item is the supervision of a head, or foreman. Then, as each machine is occupied on jobs the latter are debited with so much per hour as machine rate, and at the end of the month the total amount so earned by the machines, is deducted from the total shop expenses, leaving a balance which is distributed over the same job a« a supplementary rate. The ratio of the sup- plementary rate to the amount distributed by the machine rates forms a varying barometer, whose fluctuation is an index to the current efl[iciency of the sho^^^ , It wiU, of course, be obvious, from what has already been said, that when the machines are ftU running full time the supplementary rate wiU consist of the general charges alone, such as the foreman's wages, which have not any individual connection with particular machines. This will be the condition of maximum effi- ciency in the shop. In proportion as aU ma- chmes are not kept fuU of work aU the time, this ratio of the supplementary rate to the amount distributed by the machine rates will begin to rise. The same effect will occur if any general kind of expenditure is increased. I :^ |l ^. h 1; 'I ■ * • I: I •J.' il 70 THE DISTMBUnON OF EXPENSE BURDEN While we have now seen the principle of the supplementary rate in detail, there yet remains the question on what basis the additional dis- tribution shall take place. It may, for instance, be made into an ordinary hourly burden, or, which is simpler, may be reduced to a percent- age of increase on the amount already distrib- uted by the machine rate. In the case of the shop whose charges amount to $200, cited above, we may suppose: — Machine A worked 120 hours @ 40 cents = $48.00 for month. Machine B worked 134 hours @ 30 cents = $40.20 for month. Machine C worked 169 hours @ 20 cents = $33.80 for month. Machine D worked 200 hours @ 10 cents = $20.00 for month. Total 623 hours. Total burden distributed, $142. Now deducting the total burden distributed through these machine rates, viz., $142, from the full month's total of shop charges, viz., $200, this leaves $58 as a residual or undis- tributed amount. This forms the amount to be made the subject of the supplementary rate, but this itself may take two forms. It may be MACHINE RATES AND 8UPPLEMENTART RATE 71 treated as an hourly burden, thus: — $58 spread over 623 hours worked = 9J cents per hour, which amount would have to be added to the amount already allocated by means of the machine rate; or, it might be that the residual amoimt is reduced to a percentage of the amount already distributed, thus: — total burden dis- tributed by machine rates, $142, residual| amount, $58; supplementary rate = per cent of increase on machine-rate amount, = 40.9 per cent. On this latter plan the calculation of the correct incidence of shop charges is reduced to a very simple matter. First we get a certain amount charged against the job by the medium of the machine rate. This represents strictly the normal charge for that machine when under full work. Then at the month end we find the actual ratio of undistributed to distributed charges, and increase the debit against individual jobs accordingly. Thus: — 5 hours of machine B at 30 cents = $1 . 50 Supplementary rate at 40.9 per cent = $0.61 Total charges = 2.11 This represents: (1), a proportion of the really general charges of the shop; and (2), an averaged 72 THE DISTRIBUTION OP EXPENSE BURDEN distribution of the idle hours of any or all the machines in the shop. As by the new machine-rate plan we have only subtracted from the total charges just those charges which can be connected with different production centres, it follows that, as regards the remainder, we are again face-to-face with the problem of finding a suitable basis on which to make the distribution of incidence. The discussion of which is the better and more just way of distributing this residual amount, or supplementary rate, is worthy of some attention. Theoretically, as regards the idle hours, we are averaging these over the work actually done, and this on the basis of their being considered as the fault of the shop as a whole, and not of any particular machine; but this may be as well attained by distributing the supplementary amount directly as an hourly burden, or a pro rata increase of amounts already charged through the machine rates. The general, or, as we might more conveni- ently term them, the floating shop charges, plus the unallocated charges due to idle ma- chines, are obviously not connected in any defi- nite way with the work. They must be aver- aged on one basis or another, and the basis should be as constant — that is, subject to aa MACHINE RATES AND SUPPLEMENTARY RATE 73 few variations— as possible. This definition is better fulfilled by the hour than the hour-value. It is to be preferred, therefore, that the supple- mentary rate should be an hourly-burden rate. This, though not quite as simple in routine as the percentage form of supplementary rate just described, is the more accurate, because the incidence of the imallocated charges can- not be considered as varying within the value of different machines, but rather as pressing equally on all work done in the shop hour by hour, irrespective of whatever machine it may be done on. Nevertheless, the percentage sup- plementary rate is not without its arguments, since it means virtually a pro rata increase of machine rates in slack times, and this is very much like what might be conceived to happen in the case of the separate little shops which have been used all an illustration, supposing that from one cause or another it became neces- sary for the landlord or owner to get the same return out of a smaller number of little shops. He would raise the rents. But this is exactly equivalent to a distribution of the supplemen- tary rate as a percentage on the amount already distributed by machine rates. This is m any case quite a different matter from a percentage distribution of shop charges ■>l . .5 w I 74 THE DISTRIBUTION OF EXPENSE BURDEN on wages, and is not vitiated by the arguments already directed against that method. At this stage the matter may be left an open one, with preference in favor of the hourly-burden method of distributing the supplementary rate. A summary of the complete scheme of ma- chine rates and supplementary rates is now possible. The principle followed is to connect expendi- ture with production centres, (which for the most part mean machines,) wherever this is possible and as far as it is possible. These charges are debited to jobs by machine rents or rates so carefully arranged as to include all of them. All these items of expenditure, and others which we term floating or general items (which do not enter into the composition of the machine rates at all) are collected into a monthly shop-charges account. This account is relieved by the total of charges which have been debited to jobs during the month, by means of machine rates, leaving in the accoimt a residual sum, which has yet to be got rid of and allocated to jobs. The supplementary rate affects this, either by reducing the residual amount to an hourly burden, distributed over the jobs in the usual way, or, more simply but somewhat less accurately, by finding the proportion or ratio 'i MACmNE RATES AND SUPPLEMENTARY RATE 75 of the residual sum to the amount already dis- tributed by machine rates, and then increasing the amounts already allocated to each job by so much per cent. It will be obvious that by this means all the charges will be distributed, leaving nothing in the charges account, and that if all the machines have made full time in the month, the supple- mentary rate will only represent the floating shop charges. In proportion as machines are idle the supplementary rate will rise, because only working hours are credited to the shop account as per the total of the machine rates found at the end of the month. Thus we secure that each job gets its own expenses only attached to it, plus an average of floating charge and of what surplus may be due to slack times or inefficiency in the shop. i -I % r V 'Hit CLASSIFICATION AND DISSECTION OF SHOP CHARGES. Chapter IV. CLASSIFICATION AND DISSECTION OF SHOP CHARGES. TN the preceding chapters the practical effects "■• of the new method have been developed, special stress being laid on the desirabihty of discriminating in regard to the incidence of charges, rather than seeking, as has hitherto been the practice, to throw all classes of charge into one common collection of shop expenditure, and then to average down the whole per unit of time or wages. It might be supposed that the steps neces- sary to do this were very difficult and cumbrous. This is not the case, unless indeed at the outset. Once roughed out and the various factors deter- mined, the subsequent steps present no special difficulty. It will be useful to bear always in mind the illustration already given of the "little shops," forming part of a large mill and rented out to weekly or monthly tenants. If we fol- low the lines that would of necessity be followed by a landlord of such a building in order to the 79 # ■. 1 ii \\l 80 THE DISTRIBUTION OP EXPENSE BURDEN fair proportionment of rents, we shall not be far out of our course. •^ The first item to be dealt with would probably be the' building. For the sake of simplicity, we may assume that this is of one story only. There is, of course, no difference in principle whether the building be one or several stories high. As a matter of fact this method was first worked out for and applied to a factory which consisted of two parts, one being an old build- ing of five stories, and the other a modem shop of three. The process is the same in any case. All the factors of capital and revenue incident on such buildings are carefully mapped out. la-The capital invested in land has first place. To this is added the 6ost of the building. Partic- ular care is taken to consider whether any ine- qualities in cost in the various parts of the latter are significant or accidental — that is, whether they have any special reference to the uses ol that part. Having exhausted the capital items they are reduced to floor areas. That is to say, every square foot of floor space not sub- ject to special conditions is considered a? repre- senting so much capital outlay. Having deter- mined this, the charges incident on the floor spac^ due to this capital outlay are ascertained, then interest, jjound rent, if any, taxes, insur- CLASSIFYINQ AND DISSECTING SHOP CHARGES 81 ; ance, depreciation of buildings — all are reduced to figures, and therefore to so many dollars or pounds per square foot of floor area. The ' ' first cost and working expenses of heating and ventilating appliances are treated on a similar footing. When this is finished, with any factors pecu- liar to the local circumstances taken to account, we have obtained a pretty close idea of what rent charge is due to a production centre occupjring say 13 square feet of working space. Whether we are determining the figures for the purposes of subletting our building or for determining the correct incidence of such items of expense in a large shop, makes no difference whatever to the process or its result. So many square feet of floor, so many dollars per annum out of pocket. If machine A occupies twice the space of machine B, it is costing us just twice as much to house it. There seems to be no escape from or alternative to this position. Where shops are electrically lighted by over- 5 head Ughts, the cost of this also is reduced to a floor-area basis. The items include: — charges ^-^^ due to the capital sunk in leads, switches, lamps, and in some cases where the fight is generated on the premises, a due proportion of the cost of the generating plant; interest, insurance, i 1^ - .i i t \]i ti 1 ' I m 82 THE DISTRIBUTION OP EXPENSE BURDEN and depreciation is taken out on these sums, and to the result is added the actual cost of power of ^ current, carbons, cleaning, and so forth. Here again we get a rent representing the cost of lightmg either one of our little shops or a given area of a large shop. Two items are thus complete — first, a rent for the empty shop or area, second, the rent of the same lighted up. The shop is, in fact, ready for occupation and use. 4 Probably the cost of power will be the next item dealt with. This is a more complex sub- ject, still not insuperably so. The capital out^^ lay on engines, boilers, motors and generators, main shafting, and pulleys being ascertamed, and the resultmg revenue charge therefrom, to this is added the running expenses— fuel, ^v stoking, repairs, and so forth, reduced to a value per horse-power hour. There is not much difficulty in the broad working out of this figure, but the position of shops and various local details make the problem a httle trouble- some. In some cases the cost of power might not be the same in all parts of a large mill or scattered group of shops. This is a matter of which the practical aspect will vary, and no definite rulings are possible. It is evident that in many cases the charge for power should CLASSIFYING AND DISSECTING SHOP CHARGES 83 not be averaged. The cost of transmitting power to a distant shop might sometimes give rise later to the adoption of more efficient means of transmission, and this should of course show in the charges of that particular shop alone, and not reduce the average cost of power in other places which have nothing to do with the improvement. In most cases, however, the aver- age charge per horse-power hour would meet the case fully. The shops are now ready for the introduction of machines. ,, I. The charge due to interest, depreciation, and insurance on the value of machines is of course the item which the new method has in common with the ordinary machine-rate plan. These charges are calculated per annum and then reduced to a rate per hour, based on the prob- able number of hours the machine will be in work under normal conditions. Thus, if we take the annual work of any machine to be 2,500 hours, and the annual charge for interest, depreciation, and insurance to be $150, this gives an hourly rate for these items of 6 cents. The ordinary machine rate being extensively understood at the present day, this particular item need not detain us farther. It should be noted, however, that the items of annual or AN •■!«• \>^ 84 THE DISTRIBUTION OF EXPENSE BURDEN hourly charge for interest, depreciation, and insurance do not form the whole machine rate as in the old method, but are one of many fac- tors, of which several have already been de- scribed. The new method includes all that was signified by the old machine rate, and several other items as well. It should be mentioned that along with each machine are included countersnafts, fencing, belting, or motors, which are specially used to operate that machine alone. Some care is nec- essary to discriminate between what may belong to individual machines and what pertains more properly to the main system of power transmission. In addition to the empty shop lighted and heated, with its necessary main system of power, we have now considered the machines installed and ready for working. It will be evident that up to this point the charge incident on the work done on any machine will be made up of: — (1), a rent charge for the space occupied; (2), a charge for the capital simk in the machine it- self; and (3), a charge for power calculated according to the average amount used by the machine. None of these items can be imagined as being other than properly connected with the particular work done on any given machine, or as CLASSIFYING AND DISSECTING SHOP CHARGES 85 having on the other hand any connection with work done on other machines, whether they happen to be fixed under the same roof or not. We have, in fact, not so much differentiated as kept separate these factors of production cost, which on the ordinary methods of averaging shop charges are jumbled up in one common account. There remain, however, other items perhaps not so definite, but yet capable of being worked out into reasonable approximations to the truth. 5 Of these are the cost of overlooking and super- vision, which are frequently special to a limited group of machines, as, for example, in the case of a subforeman in charge of a half-a-dozen screw-maJdng machines or turret lathes. The approximate cost of such supervision for a year can be taken, divided by the number of ma- chines, and reduced to an hourly rate per ma- chine. The importance of confining a special charge of this kind to just that class of work that incurs it, instead of letting it be part of a general shop charge and thus become spread over all the work of the shop, will be clear. The same result might, of course, be attained by a very elaborate system of allocating such overlookers' time on a cost sheet, but this would be much more troublesome and not appreciably more I II fi 86 THE DISTRIBUTION OF EXPENSE BURDEN accurate than the plan proposed. It would be, in fact, less accurate, as will be seen when the question of partially-idle groups of machines is considered in the chapter on mass production. > The cost of tools may in some special cases be similarly treated, as for example the wheels on emery machines, cutters on milling machines, etc. This requires very cautious treatment. The same remark applies to diarges for oil,y or other material such as rouge ^^used in partic- ular processes. Such materials frequently form a very serious item in the true cpst of a process, and the justice of including them in the working expenses of a production centre will be felt if the case of the separate little shops is consid- ered. The expenses of such a httle shop doing polishing work by means of a buff wheel will be readily allowed to be made up of the standing charges already detailed plus the value of actual emery flour, rouge, etc., consumed on the work. And that substantial accuracy is attained by reducing such item of expense to a charge per working hour will be admitted if we assume, as we are entitled to, that the same machine em- ployed on the same class of work uses approxi- mately in proportion to the amount of work got through. If from the evident circumstances of the case this were not so, then other arrange- CLASSIFYING AND DISSECTING SHOP CHARGES 87 ments would be necessary. The rate might, for example, be higher when engaged on one class of work than for another. It has already been stated that an item of this class requires very careful consideration. /The principal items of shop expenditure are now accoimted for. Those remaining comprise the wages of general foremen, and the cost of repairs done to the buildings, to power, trans- mitting, lighting, and heating plant, and to the machines themselves. This latter question, be- ing mixed up with that of depreciation, needs some discussion. The usual view taken of a depreciation rate is that it is a fixed percentage, say 5 per cent, on a gradually lessening value. The effect of this is to transfer the cost of a machine from the plant account at the outset of its career to the depreciation fund at the end of its life. The value of the machine, if plotted graphically, would show a series of intermediate values year by year, each rather less than that of the previous year. From a financial point of view this is correct. As the machine wears out its value as an asset decreases, and the provision for its renewal in the depreciation account increases. It is usually understood that the cost of minor repairs during working life is being / 88 THE DISTRIBUTION OF EXPENSE BURDEN met out of revenue. Therefore such repairs are properly a shop charge. Repairs or alterations, on the other hand, which practically give the machine a new lease of life, should be charged to plant account, wholly or partially, according to circumstances. The question which interests us here is: — How should the cost of repairs which are met out of revenue be dealt with under the new method? It is evident that they can be either kept isolated from all others and in connection with the particular machine or item of plant that has incurred them, or else be treated as a really general shop charge, such as foreman's wages. The principle of the new method being to keep all expenditure individualized, wherever it can be done, it follows that arrangements must be made to give the cost of repairs expression in the various rate factors which go to make up the total new machine rate. As the charging of repairs is always an ex post facto transaction — that is, the work done by reason of which the repairs have become necessary cannot possibly be charged with them — ^it will serve the purpose if such repairs be averaged over half-yearly periods, though still kept confined to the actual machines f CLASSIFYING AND DISSECTING SHOP CHARGES 89 involved. Thus, if the power plant has been overhauled, the rent charge for power may be correspondingly increased at the next half- yearly balance, until such extra and abnormal expenditure has been wiped out. Generally, therefore, it will be seen that an estimated figure for cost of repairs is included in the vari- ous factors of the new machine rate, and this estimated charge is subject to periodical revi- sion, if on examination it is found to be too high or too low. It may be objected here that we are here breaking away from actual figures, and entering the nebulous region of estimated charges, which it is the purpose of the modem methods of accounting to avoid. The answer is that the breaking away is only temporary, and that the cost of repairs is more properly averaged over a more or less extended period than borne by any particular month in which they may happen to have been executed. And when it is remem- bered that the alternative to this method of averaging is the throwing of the cost of such repairs into the general fund of shop repairs, and spreading it over the work done in a given month, with which it has manifestly no proper connections, the propriety of the plan advocated will not be in doubt. By this means, the cost 90 THE DISTRIBUTION OF EXPENSE BURDEN of repairs to a turret lathe, for example, is kept incident on the work of that turret lathe and no other; and as no particular item of the cur- rent work can by any possibility be imagined to bear properly the burden of such repairs, no substantial injustice is done by a temporary want of coincidence between the actual rate of expenditure in repairs and the estimated rate. On the other hand, the individualizing principle is kept intact, and thia is the more important object to keep in view. Necessarily, the estimate for repairs included in the new machine rates must be no hastily determined or fancy figiu-e. Throughout this method the preliminary determinations must be most carefully made. If /this is done, then their subsequent variation WiU be an easy and almost automatic matter. If we recur to the case of the little shops, the justice of the fore- going contentions will be more clearly seen. It is evident that the cost of repairs to any one* machine is part of the expenses of the littlej shop it is in. But, ag^tin, it would not he\ correct to charge with the extra cost the work which happened to be done in a week or month in which such repairs were carried out. More substantial justice would be done by estimating in advance the probable expenditure, and includ- CLASSIFYING AND DISSECTING SHOP CHARGES 91 ing that in the hourly rate charged on the work. Then, if at the end of a six-months' period it was found that this amount had been under- estimated, the small amount representing the difference between what was actually spent and what was charged to jobs as an estimated figure would be adjusted by raising the rate slightly. The question of repairs done to the power plant or to the buildings of the little shops seems to present more difficulties. It seems clear that the landlord could not charge more for power or for rent because he had just spent a certain sum in repairs, merely to keep up the efficiency. On reflection this will be seen to strengthen the case for the averaging method. For the reason why no more rent could be charged is that, in fixing the rent in the first instance, the probability of a certain expendi- ture for repairs was foreseen and allowed for. It was, in fact, averaged in advance. And any discrepancy, any failure to estimate correctly the probable cost of repairs would, in the land- lord's case, come out of his profit. It would be a dead loss to him. In the case of a large shop, however, there being no item correspond- ing to the landlord's profit to fall back on, the recovery of this unforeseen expenditure from 1 V- J , 1 . r, 1^: 92 THE DISTRIBUTION OP EXPENSE BURDEN the shop is the only course open. It must be got rid of by becoming an expense chargeable to work, and this is done simply enough by raising the rent charge per floor area, in the case of repah^ to buildings, or per horse-power hour, in the case of the power plant. The treatment of all these items is thus iden- tical, whether we are considering the little shops or a number of productive centres in a large shop. It will be necessary now to consider the case of those charges which have no parallel in the Uttle shops, such as foremen's wages, which press equally on a number of productive centres. As there is no natural line of cleavage herein, no reasonable presumption in favor of such items being connected with one productive cen- tre more than with another, we are compelled to treat them as really belonging to the cate- gory of general or non-individual shop charges, and thus to distribute them over the work as a whole, on the hourly-burden plan for prefer- ence. The principle of this method has already been explained, and there is nothing special in the treatment of these general items which differentiates them from the ordinary distribu- tion of shop charges by hourly burden. When all the machines or production centres are run- CLASSIFYING AND DISSECTING SHOP CHARGES 93 ning full time, there is nothing in the shop- charges account to distribute but non-individual items. All the rest will have been already allo- cated to particular jobs by means of the new machine rates. It is when this is the case that the differences between the present method and any previous one are seen at the full. Out of all the many items of expenditure going to make up the ordinary shop-charges account, which are usu- ally lumped together and spread like a thin paste equally over all jobs in strict proportion to time occupied only, we find but one or two items, themselves of relatively small amount, remaining to be treated in this way. All the main items of expenditure, by following the natural lines of action, have been kept immedi- ately connected with the machines and there- fore with the jobs done on such machines, on which they are properly incident. The large and expensive machine bears its own burden; so does the small and cheap machine. Their vicinity in the same shop does not have the effect of altering each other's shop-charge inci- dence. The introduction of a special over- looker on a small group of machines does not send up the hourly burden, and thus affect the work of the whole shop; it is confined in its 94 THE DISTRIBUTION OP EXPENSE BURDEN effect to the machines and their work on which the increased expenditure was actually incurred. Throughout the items of expenditure have been not so much individualized as kept individual- ized, and thus preserved from mutual inter- ference—just as much as if they were actually as well as hypothetically incurred in separate little shops. It would be hard to argue that this is not a decided gain or that it is not worth the trouble involved, whatever that may prove to be. If it is worth while attempting to con- nect estabhshment expenditure with work at all, it is certainly worth doing so in a way which closely represents actual facts. Having thus considered in some detail the mfethod of organizing the shop charges on the new method, with the result that it is demon- strated that when all the productive centres are making full time the great bulk of shop expenditure finds its way to the actual jobs on which it has been incurred, it remains to show what happens when there is a large proportion of idle time in the shops. In such case, while the work actually going through will be charged as before with its due expense, there will obvi- ously remain at the end of the month a certain amount of undistributed charges, exactly pro- portional to the idle time of the various produc- tive centres. CLASSIFYING AND DISSECTING SHOP CHARGES 95 This balance, together with the really general (or non-individual) shop charges, remains to be got rid of. This is effected by means of an hourly-burden rate. There are thus, under the new system, two rates which enter into the cost of each job: 1, the new machine rate, which remains the same whether the shop be busy or slack; and 2, the hourly-bm-den rate, which it is proposed to call the "supplementary rate," since by its aid we spread away all the residual charges, which would otherwise escape allot- ment. This rate necessarily rises when there is a large proportion of idle machines, and falls when idleness is at a minimum. The supple-,, mentary rate thus becomes a significant index to the commercial efficiency of the shop. It should be the aim of the authorities to keep the rate as low as possible. Its rise when work is plentiful will be a sign of the most automatic nature that something is going wrong. It may occur to the reader that the more appropriate way of dealing with idle-machine time would be to distribute the unallotted ma- chine-rate amounts, not as part of the general shop expense, but as a kind of supplementary machine rate, thus confining the effect of idle- ness to the particular machine in fault. This would certainly be the ideal method, but it is er cent. The figures in the expense columns of the above table are obtained by carefully consider- » 136 THE DISTRIBUTION OP EXPENSE BURDEN ing the items with reference to the output. Thus, on examining the advertising account, it was found that $3,000 was spent in advertising special cranes, leaving the remainder of that expense to fall upon the standard lathes. Cata- logues, again, not containing more than a men- tion of the cranes, were adjudged to be borne almost entirely by the lathes. On the other hand, when the work and expenditure of the correspondence department was reviewed, it seemed just to apportion the expense much more equally between the three classes. The result of these several apportionments having been summarized for each class and the ratio estab- lished between them and the volume of business done, it is foimd that while as might have been expected the standard lathes which form the great bulk of the output absorb very nearly the average amount of incidence, viz., 13 J per cent instead of 14J per cent, the other classes, viz., cranes and repairs, differ considerably from the average, the former taking nearly 23 per cent and the latter 10 per cent. In distributing general charges each month, effect is given to these percentages. The total expenditure being found, it is not averaged indiscriminately over the whole output for that month, but in such a manner that when all APPORTIONING OFFICE AND SELLING EXPENSE 137 is distributed the proportion between the vari- ous classes is maintained. This is done by a very simple calculation, not demanding in its practical working anything more than common arithmetic. The result obtained is decidedly instructive. Assuming that the classification has been carefully made, the relative profitable- ness of any class of work can be ascertained. And if at any time it is found necessary to revise the percentage for any class, this can be done without disturbing other classes. For, as will be seen from an examination of the table above, if the percentage of any class is sUghtly reduced, it fails to absorb so large a proportion of the expenses, which consequently fall more heavily on all the other classes. It must not be overlooked that a fundamen- tal principle of the distribution is that the debit for general expenses shall be exactly cleared by the total of the allocation to classes. The percentages here express merely the relations between the classes, and not any definite percent- age on values. If one class takes 10 per cent, another 15 per cent, and a third 20 per cent, this merely means that the general expenses, whatever they are, are distributed on the basis of that relative difference between the shares taken by each class. It does not mean that 'I I 1 138 THE DISTRIBUTION OP EXPENSE BURDEN Class 1 gets 10 per cent only on its value, but that if Class 1 gets a certain burden, Class 2 has one-and-a-half times and Class 3 twice as heavy a burden in proportion to hours or value. That is, if Class 1 has a burden of 4 cents an hour, Class 2 will have 6 cents and Class 3 will have 8 cents. The actual burden will, of course, depend on the amount to be distributed. If then we find that any particular class is absorbing expenses to such a tune that there is no profit left when prime cost, shop charges, and general charges have all been debited, a prima-facie case is made out for supposing that this class is not remunerative at the prices ob- tained. Before we can say that this is actually so, the basis of classification must be carefully examined to see if it has been assessed with imdue severity on that article. This will per- haps involve some trouble, but it is eminently a case in which it will pay to spend trouble. Between the shop charges and the general charges there is, then, this difference: while the first, if properly arranged in the first place, are subject to no appeal, as they are real figures and do show exactly what has taken place, the general charges may be looked on as rather in the nature of a danger signal which gives warn- ing of a probable pitfall, but which may, if APPORTIONING OFFICE AND SELLING EXPENSE 139 careful enquiry approves, be put on one side and its readings modified. It would be better, of course, if the figures pertaining to general charges were as real and rehable as those of the shop charges. But there seems to be no possi- ble hope of their being made so. There is no visible and tangible result connected with con- crete things in the case of general charges. Nothing is produced. Expenditiu-e may, in fact, lead to no result at all — nay, does often lead to pure loss of money and time. It is this vagueness of the general charges that forbids our regarding them as bed-rocks on which we can base deductions without further enquiry. But there is no excuse for not making them as useful as possible, reading them as indications if not as facts. Enough has, perhaps, been said about the character of these general charges to empha- size the remarks which have been made several times in the course of these articles as to the infinite importance of separating them from the shop charges. Mr. Slater Lewis's theory on this point is very clear and distinct, and should be studied by every manufacturer who cares for acciu*acy and clearness in place of confusion and mixed results in his accounts. It is not too much to say that any system of accounts V 1.1 ■ I' I "'1- !l 'f . - »u tr 140 THE DISTRIBXJnON OF EXPENSE BURDEN which lumps both classes of charge together and averages them all round is entirely worth- less. Far from being an improvement on a simple system of prime cost, it is probable that it may easily, by inducing a false security, be positively dangerous and worse than no sys- tem at all. The conclusion of this demonstration has now been reached. An attempt has been made to develop in an orderly manner the method of deahng in an improved way with the very dif- ficult subject of establishment charges. En- deavor has been made to show that present methods are not satisfactory in that they seek to mingle elements which have nothing in com- mon, and which are, on the contrary, naturally distinct. In particular it has been siiown that the old idea of the machine rate was based upon a theory essentially correct, but incomplete. It fulfilled functions which were seen to be desirable, and which have been altogether neglected in the more modem methods that have succeeded. The fault of the machine rate, or perhaps we may say its misfortune, was that it was born and came to maturity in the archaic ages of cost accounting, and that it therefore sinned against the most serious canons of modem theory and practice, viz., that whatever else is APPORTIONING OFFICE AND SELLING EXPENSE 141 done, every dollar of charges must he burdened onto some item of work. The machine rate failed because in the first place it dealt with only one or two items of expense, usually inter- est and depreciation alone, and in the second because in the case of idle time it simply lost sight of the charges which were nevertheless still there. The main feature of the system proposed, as will have been gathered by those who have followed these chapters, is that not only do the new machine rates deal with nearly all items of shop charges on a natiu*al basis, but by the device of the supplementary rate all the lost time is picked up and distributed separately, thereby leaving the significance of the machine- rate figures unalloyed by the accidental condi- tions of the shop at any time. Further it has been shown that the supplementary rate forms in itself a useful barometer of conditions. Although necessities of space have prevented a full discussion of the complex subject of gen- eral charges, enough has perhaps been said to show that the principle of averaging is not the best possible. It has been demonstrated that a certain amount of connection can be devel- oped, not between general charges and indi- vidual articles, but between such charges and I M 142 THE DISTRIBUTION OF EXPENSE BURDEN classes of products. And while the results are not as precise and indisputable as in the case of shop charges, there is still a decided gam by ef- fecting this connection as far as it can be done, the results being looked on as in the nature of a guide or warning rather than as positive and definite information. The whole tendency of what has been said is to the effect that by suitable arrangements on the lines proposed, manufacturers should be in a position to get much further into touch with the details of their work than they have any chance of doing on present methods. A few final observations may be offered on the subject of works accounting generally. Modern methods have taken their rise in the growing complexity of modern industry. They, like the industry itself, tend to become more complicated in proportion as the numbers of their factors increase. The snare of the ' ' simple system" must therefore be avoided. One can- not calculate the weight of the earth or the distance of Uranus by means of common arith- metic. Nor can anyone represent the thousand- and-one interlocked factors of a modern factory by means of a double-entry ledger and an oflice boy. And just as the whole science of navigation hinges on higher mathematics, so Ni APPORTIONING OFFICE AND SELLING EXPENSE 143 the management of a considerable business turns upon intricate principles which are the horror of the rule-of-thumb man and the sheet anchor of the progressive man of business. Now one can traverse the Atlantic in an open boat and without the Nautical Almanac, but it is done more quickly and more surely in the Mauretania. No system, however good, will give immedi- ate results. It must be established, and must have run smoothly for months and even years before the full advantages are realized. For all progress is made by comparing what was done today with what was done some time ago, and this can be done to any purpose only when the basis of comparison is similar. From the nature of the case, no spasmodic efforts at the introduction of new methods are of the smallest use. To be of final and real commercial and technical value, records must be continuous and extend over various fluctuating periods of trade. It will seem superfluous to many who read these pages to insist on the necessity of connect- ing the establishment charges as closely with costs as it can possibly be done, yet the truth is that comparatively few manufacturers are awake to this necessity. They do not realize 11 i ■ •I r \ .1 't v\ 144 THE DISTRIBUTION OP EXPENSE BURDEN that in all industries it is the narrowing of mar- gins of profit, too often to the vanishing point, that has brought this question to the front. Fifty years ago, though there was a need (as evidenced by the tentative intjpoduction of machine rates), there was not a pressing and imperative necessity for the manufacturer to look so closely and minutely into what he was doing. And only the "coming men" feel that necessity now. The older concerns will prob- ably never come to that point, for they will have become exhausted in the effort to find out what was wrong with them. •}■ -I !f ■■ I 1 h iir \i . Date Due Hbc^ -MaJL ^^'Xi j v^ ^ j — . 1 9 XD 4-. Jhe proper distribution of expense^ burden* thSH o\gs\ NEH BAY 1619M &r COLUMBIA UNIVERSITY LIBRARIES 0041415493 j^'^.'^'ljfT^^'f^^ *^^^«^?*r*^''^ m^f''\ " ■*J HK ill^:WH!i'?r'tiynr ^m ^fs* lf**K-'.„J" ^^^ iuttahik*^ jMiihWB ;4^a&»sW*>*-^sj*y««*^H wVitilv? END OF TITLE