Distribution Of Milk
( Originally Published 1939 )
Because of the local nature of the milk business, national figures on costs would be rather meaningless averages. Sources of supply, local regulations, distribution methods and cost accounting practices vary so widely from market to market that it is a question whether any set of figures can even be considered as typical. However, various studies of costs by localities throw light on some parts of a very complex business and illustrate something of the nature of the milk distribution problem.
A recent study of the Federal Trade Commission in the Chicago sales area shows a rising trend in prices after 1917. During the depression years after 1929 retail prices dropped considerably, as did the prices paid to the farmers; but in the Chicago area this was greatly aggravated by a local competitive situation. The data presented in Figure 6 should therefore be interpreted with caution. They show that while the dealer's spread in cents per quart de-creased substantially in the depression, his share of the milk dollar at first increased and then fell again during the early recovery years.
This general tendency toward rising distribution costs and declining prices paid to farmers, is borne out by a report of costs in the milk business in Milwaukee .8 This revealed that beginning with 1923, when the farmer received 58.1 cents of the consumer's milk dollar, his share shrank consistently down to 1934, when he received 46.8 cents. During this same period the retail price of milk declined from 10.5 cents to 9.5 cents but the distributor's gross margin increased from 4.4 cents to 5 cents.
What usually happens in this complex interrelation of producer, distributor and consumer when retail milk prices rise or fall? The Federal Trade Commission came to this conclusion:
A drop in prices charged consumers has usually been accompanied by a reduction in prices paid by distributors to producers ; similarly, an increase in the price paid to producers has been followed immediately, in almost every instance, by an increase in the prices charged to consumers, and, in many instances, the latter increase has been greater than that allowed producers. . . . Generally speaking, from the facts ascertained during this investigation in a limited number of milksheds, while the full extent of the decreases in prices paid to producers has not always been passed on to the consumers, the full amount of increases in prices paid to producers has usually been added to the prices charged consumers.
The handling of milk by large city distributors not only involves distributive expenses such as teaming and hauling costs, freight charges, delivery expense, and advertising and selling costs, but also rather extensive processing. Delivery expense alone accounts for a large part of what the consumer has to pay for a quart of milk, as can be seen from Table 3. The table also shows how much various expense items increased during a fifteen-year period ending in 1931.
An increase of a little more than five cents a quart, or 58 per cent, occurred in the retail selling price of milk over the period shown in the table. Almost two cents of this increase is accounted for by higher costs of raw milk to the distributor. The distributor's spread, or the total of his operating costs and profits, accounted for a little more than three cents of the entire increase but the profit per quart actually decreased during this period. Percentage profits were only half as large in 1931 as they were fifteen years before.
Delivery expenses, which doubled during the period, accounted for most of the increase in the distributor's costs and for more than half of the total increase in the price of milk to the consumer. In 1931 the consumer paid five cents a quart or more than a third of the total price to have milk delivered at his door. Whether this increase in delivery expense reflects wasteful methods, or results from such factors as the increased congestion of city life, smaller units of daily purchase, greater requirements in delivery service or higher wages to employees, cannot be known without further study.
Delivery and Selling Complexities
Delivery and selling expenses vary widely. The Federal Trade Commission found in their study of four large cities that out of an average gross margin of 42 per cent, delivery costs comprised about 26 per cent of net sales in Baltimore and Boston, 31 per cent in Cincinnati, and 34 per cent in St. Louis. This item also varies to a marked degree from distributor to distributor. The Commission's study showed that average costs of seven milk distributors in these same areas for delivering a quart bottle varied from as low as 2.64 cents to as high as 4.78 cents. A large part of this variation was found to be due to different methods of paying the route men, which in turn grew out of varying regulations of the route men's unions.
An interesting comparison was made by the Commission between the costs of delivery of milk and cream in bulk, in quart bottles, and in pint and half-pint sizes. It was found that delivery in small bottles was relatively very expensive, since small pack-ages involve as much time, effort, and expense as larger units. Ex-pressed in terms of quarts, the cost of delivering milk products in bulk was about 1.61 cents, in quart bottles 4.69 cents, in pint bottles 7.12 cents, and in half-pint bottles 13.66 cents per quart."
Consumers must bear part at least of the responsibility for the high cost of delivery, since they demand doorstep delivery and other special services. In a Milwaukee study, it was found that 78 per cent of the families had their milk delivered regularly while only 12 per cent invariably bought it at the neighborhood store. Delivery through retail stores may save one cent a quart, or possibly more, in handling expenses as compared with home delivery, but in many cities consumers who buy milk at stores and deliver it themselves are still taxed for the cost of a delivery service they do not get.
An even cheaper method of distribution has been tried out in New York City to meet the needs of the city's poorer population. Delivery was made direct to consumers from a dealer's wholesale truck within a two-hour period each morning. Sales were for cash and bottles had to be returned or a deposit forfeited. While the actual expense of this method has not been reported, dealers were eager for a chance to sell this milk for eight cents a quart when regular retail delivered price was thirteen cents and the store price, twelve cents.
Home delivery is further complicated and made more expensive by the duplication of territory by various milk companies in every city. In a study of 1,020 city blocks in Milwaukee in 1934 it was found that in every block but one, at least two companies made deliveries. In 800 of these blocks, five milk companies went in and out during the day; in 147 blocks, seven companies competed with each other; in one block seventeen companies made deliveries; while two apartment houses were found in which nine different companies delivered milk daily. This duplication also extends back to the hauling process, where competition causes similar wastes in collecting milk from the farmers.
One of the large manufacturers of candy bars has furnished confidential information on costs and price spreads for his product in 1936. These products are usually sold by the producers through wholesalers to independent retailers or directly to chain stores and other large retail distributors.
The direct factory cost (exclusive of the manufacturer's administrative and selling expense) of producing the standard unit of twenty-four five-cent bars was reported as approximately 49 cents. This unit is quoted to the wholesaler at a nominal price of 64 cents, with discounts and freight allowances which reduce the average price to about 61 cents. Chain stores are quoted at the same price but are often successful in getting extra allowances which bring the net price down to 60 cents.
Wholesalers have no standard or usual resale price. Some of them sell these products to retailers in western markets for as low as 64 cents-the quoted manufacturer's selling price-making their own profit on other items. From this level, the wholesaler's price to the retailer ranges upward to as high as 78 cents, depending upon competitive conditions in various territories. Independent retailers sell these products to the consumer at five cents, or $1.20 for the unit of twenty-four bars. In this case the price spread, covering the selling and administrative expense and profits of the manufacturer and wholesaling and retailing costs, amounts to 71 cents, or 59 per cent of the price paid by the consumer and 145 per cent of the factory cost.
Many chain store organizations, however, having bought at even less than the wholesaler, sell the standard five-cent bars at three for a dime, or 80 cents for the twenty-four bar unit. The price spread between the factory cost of 49 cents and the retail price in this case is only 31 cents, which means that only 39 per cent of the consumer's dollar goes for marketing.
Clearly conditions vary so widely in this industry that no definite conclusions can be drawn. Competition is so acute and the manufacturers and wholesalers know so little about their costs that total costs of production and distribution often may not be covered entirely by prices charged.
Although the clothing industries embrace a multitude of different products varying widely in price and quality, certain pervasive characteristics help to explain pricing policies, price spreads, and marketing methods. Most of these articles are style goods of a seasonal character, which necessitates flexibility in the industry and often involves substantial price spreads. Much, but not all, apparel manufacture is carried on by relatively small enterprises whose activities are largely limited to production and do not include the elaborate marketing methods involved in national advertising of trademarked goods. Many products such as shoes and hats, however, are commonly sold under national brands.
Clothing is generally sold by the producer directly to many kinds of retail outlets most of which, like general merchandise and department stores, also handle a wide variety of other products. In other words the business is marked by a specialization and limitation of function at the producing end and by exactly opposite conditions at the retail end.
Examples of wide variations in cost-price spreads in the distribution of a typical article of clothing may be seen in the prices of men's and women's hats. In comparison with some other kinds of clothing the price spreads of hats are relatively moderate in spite of the fact that rapid style changes, particularly in millinery, are accompanied by losses from obsolescence and mark-downs.
By far the largest volume of both men's and women's hats moves directly from the manufacturer to the retailer. Table 4 is based upon the 1936 experience of manufacturers who distribute in this manner. It shows that one brand of men's hat with a factory production cost of $1.70 (including factory overhead but excluding general administration and selling expense) was sold for $2.13 to the retailer, who in turn sold it to the consumer at $3.50. The total spread between production cost and retail selling price, therefore, was $1.80. In other words, about 51 per cent of the price paid by the consumer went for distribution. In contrast with this low-priced hat, a hat selling to the consumer at $10 cost the manufacturer about $3.74 to produce and was sold to the retailer for $5.75. In this case $6.26 of the consumer's $10, or nearly 63 per cent, went for distribution; and $4.25, or 43 per cent, was paid for the retailer's expenses and profits.
The retail selling prices represent regularly maintained prices for the two brands of hats and are comparable. Why the low-priced hat should carry a total distribution cost of 51 per cent, while the higher-priced hat bears a total charge of 63 per cent of final cost to the consumer, needs explanation. It may be that the consumer pays too much for the quality product and too little for the cheaper article. ntially by representative hat manufacturers.
The manufacturer of the $10 hat advertises extensively and has built up a reputation over many years for making a consistently dependable high-quality product. Thus, in addition to the guarantee of obtaining good materials and workmanship, the purchaser of this hat has to pay for the psychological satisfaction which wearing a universally recognized high-quality product is supposed to carry with it. The manufacturer must pay to establish and maintain this universal recognition, and of course the cost is passed on to the consumer.
Higher Mark-up for Quality Goods
The higher percentage mark-up on the quality product is due to the fact that the turnover on this class of merchandise is smaller, and as a consequence, the selling expense is greater. There are many more buyers for a $3.50 hat than for the $10 one. Expenses chargeable to service and returned goods are also greater for higher-priced products than for lower-priced goods.
If the manufacturer's production costs furnish an accurate measure, the high-priced hat is inherently worth only about $2.00 more than the low-priced hat; yet the consumer must pay $6.50 more for it. This raises a serious question as to whether conventional pricing policies for similar products of different quality are always sound, either from the standpoint of the businessman or the consumer. It is possible that they result in a vicious circle in which high mark-ups and high prices are responsible for a small volume of sales and slow turnover on quality products, which in turn make for higher unit costs of distribution and thereby necessitate higher prices.
If it costs only $1.80 to market a hat costing $1.70 to produce, there is little reason to believe that it should cost $6.26, or nearly three and a half times as much to distribute a hat costing $3.74 to produce. If actual distribution costs could be ascertained and properly allocated between the two, it might be argued that the mark-up and retail price should be higher on the low-priced product, and lower on the high-priced one. It must be remembered that it is the cost of potential distribution rather than current distribution which must be considered. If it is possible so to reduce prices as to double sales, the cost of production and distribution per item may be so reduced that the lower prices may result in more profits to producer and distributor as well as in savings to the consumer.
Even within the same price class, surprising differences in price spreads sometimes occur. Three manufacturers of women's hats retailing at a price of $5 reported production costs ranging from $1.71 to $2.25. Thus 66 per cent of the consumer's dollar, in the first instance, and only 55 per cent, in the second instance, was paid for distribution. Such cost differentials, which are not uncommon in style merchandise, may reflect either actual differences in production or distribution costs, differences in quality, or differences in profit margins.
b. WOMEN'S DRESSES
Most manufacturers of women's dresses are small concerns which do not produce nationally advertised merchandise under their own name, but sell direct to the retail trade, in many cases under the retailer's label. The manufacturer's selling expenses are therefore not a large part of the final price. Although cost accounting is not well developed in this industry, the typical dress manufacturer, following a rule-of-thumb method, breaks down his sales dollar into three equal parts. One-third covers the cost of materials and trimming, another third, labor costs, and the remainder provides factory overhead, administration, and selling expenses and profits.
This distribution of costs must be considered as an average, however. The manufacturer's price to the retailer on a specific article of merchandise, and the latter's price to the consumer, often do not accurately reflect differences in costs because of the practice of adhering to a relatively few standardized retail price levels. One dress may cost more to make than another and yet sell at the same retail price. This is especially true of highly-styled merchandise but also holds to a degree even in the mass market.
In general the higher the retail price of the dress, the larger is the margin of the retailer, both in percentage of cost and in actual price spread. The larger mark-up on more expensive dresses is due to their smaller turnover and particularly to heavy end-of-theseason mark-downs because of style obsolescence. No detailed information on actual production costs for specific products was submitted by manufacturers for this study; but there was substantial agreement that, irrespective of grades and qualities, materials and labor usually account for two-thirds of what the manufacturer receives for his merchandise.
C. LEATHER SHOES
The manufacture of standard popular-priced leather shoes presents sharp contrasts with other kinds of apparel. These products, especially men's shoes, are not as subject to seasonal influences and abrupt style changes as are most other clothes. Their production is carried on in large and highly mechanized establishments. Nearly all of the standard machinery used in the industry is owned by the United Shoe Machinery Corporation and leased rather than sold to the manufacturers. Consequently, methods of manufacture are highly standardized and there is comparatively little variation in production costs from one factory to another.
Most of the industry's production is put out under the manufacturer's brand, more than half of the total production going direct to retailers without the intervention of any intermediary. Specialty stores dominate the retail shoe business and in some cases manufacturers operate their own retail outlets. Confidential figures for 1936, obtained from certain representative manufacturers and shown in Table 6, furnish an accurate measure of price spreads for the popular-priced grades of men's and women's shoes. Spreads -particularly the margins of the manufacturer and wholesaler-are surprisingly small in comparison with other articles of apparel. This may be due to the highly competitive nature of the industry; in part at least to the fact that competitors all use much the same machinery on much the same terms and are left with a relatively narrow field in which to demonstrate their competitive excellence. Larger price spreads for women's shoes seem to be due to the importance of novelties and frequency of style changes and to the fact that larger inventories have to be carried.
The manufacture and distribution of rubber footwear is of course an entirely different industry from the leather-shoe business. These products are made chiefly by companies which produce a great variety of rubber goods. They are distributed through numerous channels, sold in many types of retail outlets and show little uniformity in prices and price spreads. A confidential report from one of the leading manufacturers on costs in 1938 shows the diversity of conditions in the industry and the difficulty of presenting an accurate and representative picture.
One of the typical items sold by this company is a men's short rubber work-boot produced at a factory cost of $1.25 a pair. The manufacturer's selling price to the jobber is $1.45. Jobbers resell this item to the majority of their retailer customers at $1.70 a pair who sell it to the consumer for $2.29. Thus the spread between the factory cost and the retail price of these boots amounts to $1.04 or 45 per cent of what the consumer pays.
4. OTHER PRODUCTS
Distribution costs and profits and price spreads vary so widely among products of different nature, quality and value that it is dangerous to draw general conclusions from specific examples. Most of the varied products discussed below, however, are generally typical of a considerable group of commodities of similar nature.