of small manufacturers who make a single product or small variety of products. The selling agent effects his principal economies by handling the outputs of several mills, and thereby reducing the selling cost per unit of goods sold.

Most manufacturers employ no intermediary in reaching jobbers, but go to them direct. This requires the manufacturer to maintain a sales organization, to employ salesmen, to assume merchandising risks, etc. In the textile trades, the American Woolen Company and some of the largest cotton mills "sell direct," which means that they sell direct to jobbers and the "cutting-up trade," rather than direct to retailers. Selling direct to jobbers usually requires relatively few salesmen, for goods are sold in large units, and the manufacturer has relatively few customers, all of whom are "good pay." It enables the manufacturer to manufacture largely "on order" rather than "for stock," hence reducing the storage function.

Other manufacturers assume still more of the marketing functions and "do their own jobbing," thereby selling direct to the retail trade. Such an undertaking demands an extensive and expensive selling organization, a very large number of salesmen (to perform the selling function), frequently the establishment of branch offices dealing with and giving credit to thousands of retailers who are "slow pay" (financing and risk functions), the manufacture of goods for stock (storage function), and sometimes the delivery of goods to retailers (as is done by the National Biscuit Company). Well may a manufacturer hesitate before he undertakes to develop the organization and assume the expense of performing these functions. Many a manufacturer has thought it over, and decided that the jobber can do these things for him more cheaply than he can do them himself.

Although the writer has no intention of discussing in this place the reasons for the tendency of manufacturers in some lines to assume these marketing functions, it should perhaps be noted that other considerations than mere cheapness or economy in marketing are often the controlling motives-such as better pushing of advertised goods, control of resale prices, etc. In fact, it is a question whether goods marketed direct to retailers by manufacturers are done so any more cheaply than if they were sent through the hands of jobbers; probably the contrary is true in many instances. At least, when goods are marketed in this

way, they usually do not reach retailers at any lower price than when they go through jobbers—except in the case of very large retailers who buy in "jobbing quantities."

It is needless to carry this analysis further and to show how the manufacturer assumes even more of the marketing functions when he sells to consumers through his own retail stores. What has already been said is perhaps rather obvious, but it illustrates the value of thinking of these questions of commercial organization in terms of the marketing functions. It may not be amiss to call attention to the fact that while the manufacturer has tended to assume more of these functions, the consumer, on the other hand, has tended to perform fewer of them. He (or rather she) goes to market for foodstuffs less frequently and dislikes to carry goods home with her. Furthermore, she either dislikes or is unable to perform the storing function, and hence buys supplies from day to day. When she goes shopping for clothing and household goods she is performing the "assembling" function, but the department store has helped her out in this particular by "assembling" all manner of goods under one roof.

As the writer began his studies of marketing in the field of agricultural products and has since extended his investigations to include manufactured products, he has been particularly interested in comparing the methods of marketing these two general classes of products. There are many noticeable differences, including the following: There is a greater diversity of methods in the marketing of manufactured products; exchanges play a more important part in the farm products trade; farm products are more commonly sold at auction than manufactured goods; and the wholesale trade in manufactured goods is concentrated in fewer houses, which do a greater average volume of business and which cover wider territories.

But the most interesting difference lies in the fact that there is greater functional specialization in marketing farm products than in marketing manufactured products: in other words, that farm products pass through the hands of a greater number of successive middlemen. Furthermore, there is a distinct tendency in the selling of manufactured products toward the "elimination of middlemen,” or rather toward the taking over of the marketing functions by producers—a tendency that is not apparent in the handling of farm products. In the farm products trade, for example, there is usu

ally a country shipper—a middleman that has no counterpart in the manufactured goods trade. Also the wholesale distribution of farm products is commonly divided between two agents-the wholesale receiver or commission merchant, who assembles large quantities from various and far-away localities; and the jobber, who sells in small quantities to hundreds of retailers. In the wholesale trade in manufactured products there is no such important division between two functional specialists, although the commission house, manufacturer's agent, and broker, described above, sometimes appear between manufacturer and wholesaler.

What are the reasons for the greater functional specialization in marketing farm products? In answering this question one is helped by referring again to the marketing functions. In general, it may be said that these functions are harder to perform in the case of farm products, and the more difficult they are to perform, the greater need of specialization. The risk in farm products is greater because they are perishable and because their prices are more fluctuating; they are produced seasonally, and hence the storage function becomes more important; they are not so well standardized, and hence the sorting, grading, and packing (or rearrangement function) is more difficult to perform; the producing unit (the individual farm) is much smaller than for manufactured products, which means that the assembling function has to be performed by a country shipper, or sometimes by a coöperative shipping association. The fact that farm products mature first in one section of the country and then in another, also complicates the assembling function for wholesale dealers. Farm products are sold through a larger number of retail outlets than are most manufactured goods, and hence the selling function is more difficult. Without going into detail, the division of the wholesale trade between two successive middlemen can be explained only in terms of functions performed; in this case it will be found that the wholesale receiver or commission merchant specializes in the assembling, storing, and financing functions, and that the jobber specializes in sorting, selling, and delivery. The risk function is divided between the two, except in the case of commission merchants, who do not take title to the goods, and who shift the risks largely to country shippers.

In closing, it may be mentioned that one other feature of this classification of marketing functions is its value to the scientific student of marketing subjects in performing his research work or

trade investigations. With these functions in mind, he can approach practically any kind of dealer or trader and ask a fairly intelligent and comprehensive set of questions without knowing much of anything about the trade. For example, suppose one wishes to study flour brokers. Begin with the assembling function, and ask: For what class of mills do you sell flour? Where are they located? For how many mills do you sell? Under what arrangement do you handle flour for your principals? etc. Then take up the storage function, and ask: Do you keep flour on hand for your principals? If so, is it kept in a public warehouse? Who pays storage fees? Why and how long is it kept on hand? etc. By going through the list in this way, and by having a certain amount of merchandising knowledge which makes one answer suggest another question, an investigator can find out practically everything he needs to know of the functions of any class of traders.

Sheffield Scientific School.

L. D. H. WELD.

REAL WAGES IN RECENT YEARS1

Those who are inclined to argue that the real wages index numbers of the Bureau of Labor report2 of July, 1908, furnish grounds for optimism on the part of the laborer (and there were and possibly still are many such) have a worthy opponent in Dr. I. M. Rubinow. Eliminating the fluctuations by use of the moving average, this writer, in the AMERICAN ECONOMIC REVIEW for December, 1914, shows that in the trend of the series there is little encouragement to be found for the laborer. Dr. Rubinow is not willing to stop here, however. Taking advantage of certain later data, he has computed a series of index numbers purporting to show the trend of real wages from 1890 to 1912. In attempting this task he had certain difficulties to face. In the first place, the Bureau of Labor in 1908 ceased publishing information concerning either retail prices or wages. And it was not until 1912 (Bulletin No. 105) that a continuation was made of the study of these matters. For certain reasons when this work was continued the retail prices of only 15 articles of food from some 39 of the more important cities (instead of 30 articles from 68 localities as had been the case with the old series) were studied. An index, however, representing the price of these 15 articles running over the period 1890-1912 was constructed. Changes were also made in the form of presentation

1A critique of Dr. I. M. Rubinow's article on "The Recent Trend of Real Wages," published in the AMERICAN ECONOMIC REVIEW, Vol. IV (December, 1914), p. 791.

2 Bulletin No. 77. The series is as follows:

Series known as the moving average is as follows: