The tendency for sales per employee to decline as the number of employees is increased, is shown in different form in table 8. This table shows only the relationship between the proportion of total paid employees and total sales by size of store. When sales account for relatively less of the total than does employment, the average sales per employee must be less than the average for all stores. The reverse is true where the proportion of sales exceeds the proportion of employment. It is interesting, therefore, to see that stores with 100 or more paid employees accounted for 17.6 percent of the total number of employees, but only 12.7 percent of the total volume of sales. Similarly, stores employing between 20 and 99 workers had 24.1 percent of the employees, but only 22.3 percent of the sales. For these groups of stores, the volume of sales per employee was lower than for smaller stores for which the proportion of sales was about equal to or greater than their proportion of total employment.

TABLE 8.-Retail stores, sales, and paid employees, by employee size, 1948

NOTE. In this tabulation, data have been omitted for number of stores and sales of those stores which have no paid employees.

Source: U. S. Department of Commerce, Bureau of the Census, United States Census of Business, 1948, vol. I, Retail Trade-General Statistics, pt. I, Washington, 1952, p. 25.

Sales provide the ultimate source of income out of which all costs must be paid. Of course, the proportion of the total sales dollar available to meet wage and other costs, after paying the cost of goods sold, varies among retail lines and among stores within a given line. Nevertheless, these data suggest that there is no major difference between stores with large and small employment in the abiilty to assume significantly higher wage costs. This appears to be true, despite the differences in number of employees, in the number of stores in a firm, and in volume.

S. Diversity of retailing: the chainstore has not been increasing in relative importance

Retailing involves a number of different types of stores and selling operations. Retail stores may emphasize price appeal, service, or some combination of both. Self-service stores usually feature price appeal. Department stores, on the other hand, usually offer a wide variety of services including credit, delivery, liberal policy on returns, and so on, as well as some price appeal. Some stores, such as the variety stores, sell a large number of popular priced items. Some types of food stores combine price appeal with a limited amount of service. Stores also differ in the size of the average sale. The lowest average sale is usually found in the variety store. At the other extreme, we find the relatively high average sale in furniture stores. Thus, while each of these types of stores is local in nature, the impact of increases in costs will vary widely depending upon which combination of price appeal and service is present in a particular store or retail line.

Table 9 shows the relative importance of single unit and multiunit stores in 1948. There were 1,606,885 single unit stores, or 90.8 percent of the total number of stores. These stores accounted for 70.4 percent of the total sales. The table also shows the relative importance of several types of stores. Variety stores accounted for 1.9 percent of total sales in 1948, department stores for 8.2 percent, and mail order catalog sales for 1.1 percent.

TABLE 9.-Number of stores and percent of total retail sales, single units, multiunits, and by selected type of institution, 1948

2 The type of stores contained in this portion of the table include both multiunit and single-unit stores. Source Howard H. Maynard and Theodore N. Beckman, Principles of Marketing. The Ronal! Press New York, 1952, p. 157.

These various kinds of stores usually are in competition with each other. Competition often takes forms other than price, although, with some exceptions, that is usually a factor. Thus, the neighborhood independent retailer may offer convenience and personal attention as an offset to the lower price, credit, liberal returns policy, and delivery service of the downtown store. In periods of shortage, the independent store may offer a more certain source of supply because "they know their customers." However, the consumer may be attracted to the chainstore or department store, where he may obtain a price advantage to offset the inconvenience of traveling or parking his car. Or he may prefer to buy in the department store or specialty shop because of the larger selections or because of a wider range or better styles or for any number of reasons. Here we are not concerned with the relative merits of each type of retailing. Experience has demonstrated that there is room for all types of retailing in our economy.

Nathan E. Handler has described the flexibility of retailing as follows: "The retail business is fluid. Tobacco chains sell haberdashery; clothing chains sell shoes, hats, and furnishings, and lately have been branching out into the feminine-apparel field generally. The automobile-service store goes in for sporting goods; the drugstore stocks all lines and will soon be selling wearing apparel. The lunchroom may be expected to combine its normal service with the selling of anything else that will make the cash register tinkle. Possibly we may buy overcoats or umbrellas or galoshes with our coffee and cake. "It is this adaptability, this capacity to readjust to changing opportunities that suggests that the triumph of bigness in the retail business is not as certain as it seems.' [Italics added.]

In the department of labor materials, considerable emphasis was given to so-called interstate firms. (See E, p. 3, etc.) Some data were presented to show the relative importance of "interstate" firms in terms of employment by kind of business in 1953 (E, p. 5); for selected large chains, the number of stores and employees were shown (E, pp. 6-7). These scattered data covered only 1953 or 1954. Except for one comparison covering larger stores, rather than chainstores, data were not presented to show changes over time. As a result of this comparison, the department of labor materials concluded: "During the 1939-48 period the larger stores, those with 20 or more employees each enlarged their proportion of industry employment from 36 to 42 percent. Since 1948, the big business trend in retailing appears to have continued, perhaps accelerated. Major evidences of this include (1) the supermarket boom, (2) the growth of multistore enterprises, (3) the drift toward self service, (4) the trend toward one stop and one-half stop outlets by most of our major retailers” (F, p. 4). This statement apparently refers to the entire retail trade and specifically to the larger stores, not to multiunit firms. That such tendencies have developed is probably true. The full picture of the trend since 1948 is not clear. But

2 Nathan E. Handler, How Big Is Retailing? And Is It Good? Journal of Retailing, vol. 25, No. 1, spring, 1949, pp. 16-17.

between 1948 and 1951, the proportion of retail employment accounted for by firms employing 20 or more workers remained about unchanged.23

At the outset, it should be noted that there are no census data available for interstate firms. Data are reported in terms of the number of stores owned by a firm, not in terms of the States in which those units are located. There are no data to show whether all units are located within a State or in two or more States. The term "multiunits" is used by the Bureau of the Census to describe the firms with more than one store. The Department of Labor apparently has used the data for multiunit firms, described as four or more units, to show the importance of interstate firms. Such a procedure must overstate the relative importance of interstate firms since some firms which have four or more stores have all of their units within a single State. Unfortunately, data are not available to indicate the relative importance of these intrastate muliunit firms.

The multiunit firms have not followed the pattern of growth shown for large stores. The United States Department of Commerce reports sales data for all retail stores and for multiunit firms. Up through 1951, the sales data were available for firms with 4 or more units; since 1951, data have been available only for firms with 11 or more units.

Table 10 shows the relative importance of multiunit firms in 1938, when the Fair Labor Standards Act was passed, and in 1951, when reports for the four unit category were abandoned. To picture the most recent trends, the 11-unit series is shown for 1951 and 1954. Except for grocery, department stores, and general merchandise stores, the relative importance of firms with four or more units to total sales, declined from 1938 to 1951; the changes since 1951 have been very small.

TABLE 10.-Relationship of monthly average sales in all retail stores and in multiunit firms with 4 or more and 11 or more stores, 1938-51 and 1951-54

Sources: U. S. Department of Commerce, Office of Business Economics, Business Statistics, 1953 biennial edition, pp. 44-45, 49; Survey of Current Business, February 1955, pp. S. 9-10.

Table 11 shows the trend annually from 1935 to 1954 for all retail stores, excluding eating and drinking places. In 1951, the last year for which data were available for the 4 or more units series, such firms accounted for 23.5 percent of total retail trade, as compared with 24.6 percent in 1938, when the Fair Labor Standards Act was passed. Only in 1950 (22.9 percent) did such firms account for a smaller proportion of total sales. The data clearly reflect a small decline in the relative importance in 1949-51 as compared with most earlier years. From 1951 to 1954, the proportion of total sales accounted for by the firms with 11 or more units has risen fractionally. If it is assumed that a similar trend developed for the firms with 4 to 10 stores, then in 1954 it appears that the proportion of volume accounted for by the firms with 4 or more units was about in line with the experience of the past 20 years. These data lend no support to

23 Survey of Current Business, May 1954, pp. 22–23.

the inference that chains have been increasing in relative importance in total retailing.

TABLE 11.-Monthly average sales in retail stores, excluding eating and drinking places, 1935-54

Table 12 shows the annual data for variety stores. The proportion of sales accounted for by firms with 4 or more units declined from 88.2 percent in 1938 to less than 83 percent in 1951. Since 1951 a further small decline has taken place.

TABLE 12.-Monthly average sales of variety stores, 1935-54

Table 13 shows the annual data for general merchandise stores. In this category the proportion of total sales accounted for by multiunit firms rose from 39.8 percent in 1938 to 54.8 percent in 1951; there has been little change since 1951.

1

TABLE 13.-Monthly average sales of general merchandise stores, 1935-54

Table 14 shows the annual data for apparel stores. In 1951 the proportion of sales accounted for by multiunit firms was moderately lower than in the late thirties; little change has taken place since 1951.

TABLE 14.-Monthly average sales of apparel stores, 1935-54

Department stores are among the largest retail stores. It is instructive to note, therefore, that monthly average sales by department stores have increased from $139 million in 1939 to $407 million in 1954, or by 192.8 percent. During the same period, however, total retail sales rose by 306 percent. Thus, while the relative proportion of department store sales accounted for by chains has increased, the relative importance of department stores had declined.

The data discussed above do not support the conclusion that the chainstores are growing in relative importance. Students of retailing have been fully aware of this situation. Thus, in 1951, Arthur Tremain, of Montgomery Ward & Co., pointed out:

** in the past two decades, despite all the economic changes brought about by the depression, the wars, and the effect of the national growth, the relative positions of the chains and the independents have undergone practically no change.