Factors That Influence Costs
( Originally Published 1939 )
Having reviewed the comparative costs of various types of intermediaries in different lines of trade it may be of interest to consider some of the principal factors that influence costs. As in retailing these include the size of the establishment, population density, size of orders, lines of merchandise carried, inventory turnover and services to customers. Since the cost of wholesaling is influenced by all of these factors, however, it is difficult to isolate one of them such as turnover, and draw sweeping conclusions as to its effect. The facts seem to indicate considerable variation in the relationship of any one factor to total operating costs in each trade.
In considering the influences affecting wholesale costs, average figures are necessarily used. Since these averages are made up of figures from individual establishments reporting widely varying experiences even in the same kind and size of business, they can indicate only general relationships and broad tendencies. As an example of how widely costs can vary the Ohio State University report on the operating results of thirty-two wholesale grocers for the year 1934 showed that with an average expense for the group of 10 per cent of sales, individual company costs varied from as low as 6.6 per cent to as high as 17.8 per cent.
Even among firms of approximately the same size one firm may have a cost ratio twice as high as another. This is illustrated in the following tabulation based on a confidential analysis of the costs of a group of dry-goods wholesalers for 1936, which shows for each size-group the average, and the range, of expense ratios, respectively.
As we have seen in the case of grocery wholesalers," the size of establishment influences costs-often to a marked degree. The tabulation above illustrates a condition apparently prevailing in a number of trades. The smallest businesses have relatively high operating costs, which drop sharply for the medium-sized companies and then tend to flatten out, sometimes rising slightly for the largest companies. A survey of automotive and equipment wholesalers, shows that relatively small firms (with sales under $100,000) had operating costs of over 29 per cent in 1936. Costs dropped to less than 26 per cent for the next larger group followed by smaller declines for the next size-groups down to about 21 per cent for the largest firms.
Statistics of the 1935 Census for wholesalers in groceries, clothing, dry-goods, drugs, electrical goods, furniture, and hardware, as shown in Figure 18, reveal somewhat the same situation. Cost ratios decline as the size of the business increases-at least up to a certain point. In drugs, dry-goods, furniture and hardware there was a steady decline up to the $2 million size business. In groceries and foods, on the other hand, the lowest operating ratio was eached in the $300,000 to $500,000 size-groups. It is notable that after a certain point is reached costs apparently cannot be reduced further, and in fact often tend to rise again. In all but two lines of business--clothing and electrical goods-wholesalers doing over $2 million worth of business had slightly higher operating expense ratios than those in the next lower size-group.
(2) Population Density
Wholesale operating expenses may also be materially influenced by the population density of the sales territory or by the physical extent of the trade area in which the wholesaler operates. The 1937 Dun & Bradstreet report on the confectionery trade points out that wholesalers located in the relatively populous New England states had an expense ratio of about 9 per cent in contrast with 14 per cent for companies in the Southern states.
The same survey also showed that wholesalers with customers concentrated within a hundred-mile radius ran their organizations at an average expense of from 12 to 13 per cent of sales while those with a more extended territory had costs above 17 per cent.
(3) Size of Account
The size of account is another important factor in costs-too often overlooked. Many merchants in their zeal for volume forget that small, unprofitable accounts drastically increase their operating expense. A case study of one relatively efficient drug wholesale merchant showed that a third of his customers bought less than $10 worth of merchandise a year, but that the total volume ac-counted for by this group represented only 0.2 per cent of his business. When operating expenses were carefully allocated by customers it was found that the actual cost of this class of business was twice the amount of the gross income accruing from it.
(4) Lines of Merchandise Carried
The same group of confectionery merchants mentioned in the section on population density showed marked variations in operating expense between those with most of their volume in strictly candy lines and those doing up to a third of their business in candy and the remainder in tobacco products. The average expense of running the candy business was 15 per cent of sales and of the candy-tobacco combination, 6 per cent.
Further evidence of cost variation within the same general category is provided by an intensive study made by the Department of Commerce of a Louisville wholesale grocery establishment and the costs of handling various types of items carried. This analysis showed that with total operating expenses of somewhat more than 5 per cent of sales the cost of handling one group of commodities ran as high as 11 per cent, in contrast to less than 4 per cent for another group. The commodities in the second group were characterized by high value with little weight or bulk, low sales resistance, non-perishability, rapid turnover, convenient packaging and a limited range of sizes, brands and types.
The effect of emphasis on private label goods is demonstrated by a Dun & Bradstreet study of the wholesale grocery trade covering 1936 operations 23 Wholesalers' private brands required more selling and promotion than nationally advertised goods, and selling costs therefore comprise a substantial part of the larger expense of featuring such goods. Selling costs were 4.6 per cent of net sales for distributors selling three-fifths or more private label merchandise, compared to 2.9 per cent for those selling less than one-fifth private label goods. The same study also showed that firms handling a relatively large volume of perishable merchandise had costs of operation 2.2 per cent higher than other houses.
(5) Inventory Turnover
Carrying a large variety of stock is almost sure to involve higher expenses because of slower turnover, greater warehousing expense, less quantity discounts because of smaller unit purchases, and a larger mark-down on distress lots. A Department of Commerce study shows that a typical wholesale grocer in his eagerness to over-come sales resistance, stocked two and a half times the number of items carried in a typical grocery chain store warehouse.
That this policy tends to raise costs is evident from Census figures on the relation of sales-stock ratios to operating cost ratios, which show a marked tendency for costs to be higher with a de-crease in stock turnover. The results of this analysis, covering wholesale merchants, manufacturers' sales branches and chain store warehouses.
The sales-stock ratios of the wholesale merchants varied from 6.5 to 8.7 in 1929-1935 and their cost ratios varied from 12.4 per cent of sales to 15.8 per cent. At the opposite extreme were the chain store establishments with sales-stock ratios of 19.5 to 21.6, and cost ratios of only 4.1 to 4.5 per cent. Manufacturers' sales branches were between these extremes, with sales-stock ratios of 9.4 to 12.5 and cost ratios of 11.8 and 14.9 per cent.
With a ratio of sales to stocks only half as large, it is clear that independent wholesalers carry much heavier inventories than the chains. In spite of this they probably receive smaller discounts on purchases for they are not likely to buy in as large volume in any single transaction as the chains. A study made by the Federal Trade Commission in 1931 shows that wholesalers in groceries, drugs and tobacco were not able to get as high average discounts as were granted to independent department stores and to both corporate and cooperative chains. Table 30 shows some typical discounts given by manufacturers to wholesale merchants and chains. In almost every instance the wholesaler received the smallest discount -sometimes only half as large as was granted to the corporate chain. Not only larger individual orders, of course, but larger total annual purchases are a factor in getting large quantity discounts.
Low operating costs do not always accompany rapid turnover, however. A. confidential study of dry-goods wholesaling shows that although the larger establishments as a rule had greater turnover than the smaller houses, within any one size-group the individual firms with rapid stock-turns did not always have lower operating expense ratios. High turnover was sometimes accompanied by high expense, but, on the other hand, some firms with less than average turnover had relatively low costs. Dun & Bradstreet's 1937 survey of wholesale grocers tends to confirm the conclusion that there is no direct and simple relationship between turnover and costs.
(6) Credit Operations
Since credit operation depends so much on individual and personal factors it is difficult to arrive at even an approximation of its cost in general. However, the 1933 Wholesale Census shows that the 30 per cent of all intermediary concerns which did not grant credit had smaller total operating expenses than the 70 per cent that did. Aggregate operating expenses of the credit-granting group were about 13 per cent of net sales, while the comparable figure for establishments not reporting a credit business was about 7 per cent. Cost ratios in relation to the amount of business done on credit for various types of operation are shown.
Out of the fourteen types of intermediaries shown in the table only two-manufacturers' sales branches without stocks and bulk-tank stations-show higher costs for concerns doing an exclusively cash business. In these exceptional cases the difference is small and in almost all other cases the cash firms operate at much lower costs. The difference in operating expenses of the two groups is not due entirely to the credit factor, however, since cash establishments do not perform as many services and stock as wide a range of goods as the credit-granting establishments.
As a further example of this general tendency, the expenses of service wholesalers in the confectionery trade covered by the Dun & Bradstreet survey in 1936 were found to increase as the pro-portion of credit business expanded. Firms selling for cash had expenses of less than 12 per cent of sales and as the proportion of credit business became greater there was a consistent increase in expenses up to nearly 15 per cent of net sales for concerns doing 90 per cent or more of their business on credit. However, a similar study for paint and varnish houses for the same year shows just the opposite relationship. Firms doing practically all of their business on credit reported appreciably lower costs than those doing a considerable cash business. This condition was no doubt due to better control of credit since the bad-debt losses of the full-credit firms were only half as large as those of firms with a substantial amount of cash business.
2 MANUFACTURERS' DISTRIBUTION COSTS
In spite of their large and growing importance manufacturers' marketing costs have been subjected to less study and analysis than almost any other aspect of distribution. Manufacturers are regarded chiefly as producers. They themselves often fail to appreciate their significance as distributors and have no accurate knowledge of how much it costs them to distribute their goods. Even when provision is made in cost systems for distribution, the entire administrative expense of a manufacturing company is often charged against production instead of being allocated in part to distribution. There are no standard generally accepted systems of expense allocation for the distributive operations of manufacturers such as are used, for example, in department stores.
Yet the manufacturer is steadily taking more and more responsibility for distribution of his products. The integration backward toward sources of supply on the part of such mass distributors as chain stores and mail-order houses has been paralleled by the development of extensive distribution and sales organization by large manufacturers. This trend is particularly evident in the growing importance of branded articles which must necessarily be advertised and distributed by the manufacturer on a national scale.
Even in 1929, as will be seen from a glance at the Flow Chart, manufacturers sold a large part of their output through other than the usual wholesale channels. More than half of the $69.6 billion total sales of manufacturers was sold directly to other manufacturers, to retailers, and to consumers and consuming institutions. Furthermore, a considerable portion of the $31.8 billion shown in the Chart as sales to intermediary trade was actually distributed through wholesale branches owned by the manufacturers. Between 1929 and 1935, moreover, the share of manufactured goods distributed through their own wholesale branches increased-from 18 to 20.6 per cent of total sales.
The Total Cost Bill
The total costs incurred by manufacturers in distributing their products can only be guessed at, but for 1929 they may have been larger than the estimated $9 billion given in Chapter 5. This figure, which amounts to about 13 per cent of manufacturers' sales in 1929, includes only expenses involved in direct selling, administration of advertising and promotion, warehousing and storage, credits and collections, financial expenses and a pro rata share of general administrative expense. It does not include such distribution expenses as are incurred in the maintenance of separate sales branches, transportation charges on out-going shipments paid by the manufacturer, and expenditures for national advertising. The first of these items has already been considered in the preceding section of this chapter as part of the costs of intermediary distribution. The second and third items will be discussed separately as consolidated estimates for all phases of distribution since it is difficult to determine how much of the total costs of transportation and advertising are met by manufacturers.
But the estimate of $9 billion for manufacturers' selling costs (excluding items treated separately) may well be questioned. Because of lack of adequate and comprehensive data this estimate was based on a study of the experience of 312 manufacturing firms representing less than 5 per cent of total manufacturers' sales in 1931. Since the completion of this study the Census Bureau has published the results of a survey of the distribution costs of manufacturers in 1935. According to this survey, the selling costs of manufacturers averaged 9.4 per cent of sales, instead of the 13 per cent indicated by the earlier survey. Under ordinary circumstances it would be possible to accept the Census figure, which was based on a much larger sample than the other survey. Comparison of the Census figures with the results of the earlier survey and of another confidential study, however, indicates that the distribution cost ratios reported by the Census for a number of important industries were much too low. Until much more comprehensive and reliable data become available it is clearly impossible to make an accurate estimate of manufacturers' distribution costs. About all that can be said is that they probably amount to somewhere between 10 and 13 per cent of sales; or, on the basis of nearly $70 billion sales in 1929, to between $7 billion and $9 billion, exclusive of cost items mentioned above which are considered separately. It may be of interest, however, to consider in greater detail the admittedly inadequate data on the subject which are now available.
a. DISTRIBUTION COSTS OF 312 MANUFACTURERS
This study was made by the Association of National Advertisers and the National Association of Cost Accountants and published in 1933.29 It covered the distribution costs of 312 manufacturers in 1931. The sample included twenty-nine distinct kinds of products ranging from drugs and groceries to chemicals and machinery, with both large and small firms represented, varying from less than $500,000 sales to more than $5,000,000. Reporting firms employed a variety of distributive channels, some selling direct to large retailers, some maintaining their own sales branches and others using regular wholesale channels. Although all of the 312 companies re-ported some advertising expenditures a considerable number were not national advertisers. In spite of the wide range of conditions represented, however, the sample was so small that the organizations making the survey make no claim that the results are adequately representative of all manufacturing industries. On the whole it seems probable that the sample is less representative of the small than of large companies in each industry.
Wide variations exist for different products not only in the total cost ratio but in the various expense items. Distribution costs for consumer products-ranging from 16.5 per cent of sales for radio equipment to 38.8 per cent for drugs and toilet articles-were considerably higher than for industrial goods, which ranged from 9.2 per cent for textiles to 25.8 per cent for machinery and tools. This latter ratio, which was the highest among industrial products, was exceeded by more than half the consumer products. The cost advantage of industrial products, which reflects the fact that a large proportion of these goods are sold on specification direct to the users, was especially noticeable in lower advertising and sales promotion expenditures.
Advertising costs amounting to 18.4 per cent of sales-more than twice as much as for any other consumer product-were chiefly responsible for the high expense ratio for drugs and toilet articles. On the other hand, tobacco products with a higher than average advertising cost-8.2 per cent of sales-reported almost the lowest total expense ratio. Relatively high advertising costs for paints and varnishes and heating equipment, however, were accompanied by high total expenses. Direct selling expenses (chiefly salesmen's compensation and traveling expenses and sales office expense) for these products-amounting to 17 per cent and 16 per cent of sales, respectively, were exceeded by only one other group, office equipment and supplies, which obviously can be sold most effectively by demonstration and personal sales effort. With direct selling expenses amounting to 21.3 per cent of sales-higher than for any other group--advertising expenses of 3.2 per cent were among the lowest.
b. MANUFACTURERS' DISTRIBUTION COSTS: 1935 CENSUS
In the 1935 Census of Business an attempt was made for the first time to get reports on manufacturers' distribution costs. A single reporting form was used for all industries (irrespective of substantial differences in their distribution setup) and each manufacturer was asked to report on: (1) total salaries and wages, bonuses and commissions paid to full-time and part-time officers and employees who devoted all or a major portion of their time to distribution activities such as selling, advertising, sales promotion, credit and the invoicing, installing and servicing of goods sold; and (2) distribution expenses other than salaries and wages, including traveling expenses of salesmen, advertising, credit and collection expenses, losses from bad debts, and rent, interest and general administrative expenses allocated to distribution. The latter amount was exclusive of expenses of manufacturers' sales offices which were covered separately in the Wholesale Census.
Census Data Incomplete
Both the method employed in getting this information and the wording of the question undoubtedly permitted a considerable opportunity for error in reporting, with a minimum chance of detection. Many, even of the large firms, have no clear conception of what items should be included in distribution costs and a considerable proportion have probably never attempted to segregate distribution costs from production costs.
As indicated previously, the results of this Census survey cannot be accepted as providing reliable measures of distribution cost ratios in various industries. The Census Bureau itself recognized the weakness of the basic data in stating:
Many manufacturers do not have accounting systems showing these costs separately . . . there is a decided lack of uniformity in bookkeeping methods used by manufacturers . . . one plant may consider a certain item as an expense while another plant may either ignore it or classify it as some-thing other than an expense. In some plants few or no records of distribution expenses are kept. In such instances figures reported were necessarily estimates. Because of the lack of uniformity among manufacturers in classifying and recording expense data, the accuracy of such data cannot be guaranteed, and any conclusions drawn from analysis of expenses must be made with this fact in mind.
The detailed figures for individual industries in the Census report compared with data from other sources, as discussed with representatives of various industries and trade associations, show that the ratios reported to the Census in many if not most cases give an understatement of the actual level of distribution costs. Several industries out of the 315 covered by the Census show wide differences in costs of distribution for manufacturers as reported by the Census and by other sources, which indicates that the former generally are too low. This opinion is confirmed in an address before the Boston Conference on Distribution on September 21, 1937, by N. H. Engle, Assistant Director of the United States Bureau of Foreign and Domestic Commerce, who said, "manufacturers' distribution costs appear to be consistently low in the Census reports as compared with other studies in this field."
Payroll expense as a whole amounted to 4.1 per cent of sales or to about 44 per cent of total reported expenses, reflecting the fact that labor cost is important here, as it is in other stages of distribution. The high total expense ratio for the chemicals group probably results from the fact that it includes drugs and cosmetics, with high costs, particularly advertising which is included above in "other" expenses.
C. COMPARISON OF DATA FROM VARIOUS SOURCES
By deducting certain expense items included in the survey of the Association of National Advertisers, the results of that study can be put on a nearly comparable basis with the ratios reported for specific industries by the Census. These comparisons are shown in Table 33, which also includes expense ratios reported in a confidential study for the year 1931.32 Even with these adjustments the results of the three studies are not exactly comparable because of differences in classification of industrial firms. However the comparison shows the wide variations in the available data and probably serves to establish the reasonable limits within which the average manufacturer's selling costs, exclusive of branch office and transportation expenses, fall.
In nearly every instance the Census ratios are lower-and usually substantially so-than those from the other two sources. The relative position of various industries, however, is generally much the same in each of the surveys. High expense ratios, for example, are shown for drugs and cosmetics, beverages, furniture, jewelry, machinery, and other similar non-standardized products requiring considerable sales effort and servicing.
Comparatively low expense ratios, on the other hand, were re-ported for certain kinds of standardized industrial goods such as paper and paper boxes, forgings, foundry products and textiles. Mass-distributed standard consumers' goods like meat and flour also involved small distribution expense, while individualized goods like clothing and shoes had higher expense ratios.
3. PRIMARY PRODUCERS' COSTS
No estimates of the costs of distribution of primary producers are available but it is known that their selling activities are comparatively small since sales are usually made in bulk to a limited number of buyers. Because primary products are bulky and are usually produced at a considerable distance from the point of use, transportation is the largest element in the cost of distributing them. Transportation charges are often paid by the buyer rather than the producer, however; in any event this cost is considered separately in Chapter 8.
Many primary commodities are produced under the control or ownership of the processors and users, and little or no distribution expense (except transportation) is involved in the disposition of the products of such captive sources. The mining of iron ore, for example, is controlled to a very large extent by the steel companies with little ore sold in the open market. Other metallic ores such as copper, lead and zinc move directly to smelters located near the mines, which are largely owned or operated by the smelting companies.
The oil industry presents a more complicated picture, since a substantial amount of petroleum production is in the hands of in-dependent operators. Even here, however, distribution expense, aside from transportation, is relatively low. The number of outlets for crude oil is limited because of the necessity for economical transportation by pipe lines.
Coal mining companies incur substantial selling expense in addition to heavy transportation charges. Often they have sizable sales organizations or employ the services of brokers who in turn sell to wholesalers and large retailers. Yet 25 per cent of the total production it is estimated involves no selling cost because it is con-trolled by large consumers such as railroads and steel companies.
The costs of distributing forest products are relatively low. Lumber and paper manufacturers and naval store producers control the largest portion of the production. Distribution expenses are involved, but they are chargeable as purchasing costs to the manufacturing industry and not to primary production.
Farmers are extensively engaged in distribution. The $1.2 billion of agricultural products sold direct to consumers in 1929 was distributed as well as produced by farmers. Since raw materials from the farm are generally grown at a distance from the railroad the farmer incurs a distribution expense in hauling his product to the shipping point. Grain must be hauled from the farm to elevators, livestock to concentration shipping points, tobacco and cotton to warehouses, and milk to bulk plants. It has been estimated that if wheat producers were to have their grain hauled by commercial truckers the cost would average about 10 per cent of the price received for the grain at the elevator.
Whether the farmer sells his produce to local buyers or ships it by rail or parcel post the procedure consumes time and energy and involves a selling cost which should be added to the farmer's hauling and delivery expense. Selling farm produce from house to house or in farmer's markets or roadside stands involves considerable expense and helps to explain why consumers can often buy as cheaply from dealers as from the farmers direct. Selling by mail also involves time and transportation costs in addition to the parcel post charges. No estimate of the cost of distributing these goods (except for transportation) is possible. However, although distribution costs incurred by farmers and other primary producers may be considerable in some instances, the aggregate of these costs is not large when compared to those of manufacturers and wholesale and retail dealers.