F.— ECONOMIC SCIENCE AND STATISTICS 139 



Commercial establishments and shops in particular must by their very 

 purpose be dispersed among the population and cannot be localised like 

 manufactures. They must therefore be small. The dispersion among 

 the population may, however, be too thorough in the sense that in any one 

 neighbourhood there are more outlets for any one product than is neces- 

 sary for the convenience of the neighbours, and that where deliveries are 

 made errand boys continually criss-cross one another in their rounds. In 

 England one shop to every 60 or 70 head of population has been the usual 

 estimate, but the recently published trial census of distribution in six towns 

 each of about 30,000 inhabitants, gives the observed frequency of 41-4 

 persons per selling-point or about one selling-point to every ten families. 



Even if shops selling only the same wares are taken, this census shows 

 that the population of 183,000 of the combined six towns was provided 

 with 1,973 outlets for cigars, cigarettes and tobacco only, that is one outlet 

 for every 93 persons, or say, twenty-two families ; and with about 1,200 

 outlets for canned goods, for chocolate and sugar confectionery, for tea 

 and coffee and for biscuits, that is one outlet for every 152 persons or, say, 

 thirty-sLx families. 



Is this typically very small size and small clientele of shops the result 

 of a survival of the fittest ? The evidence from costs and profits, to be 

 taken in the next section in connection with distributive firms, shows that 

 there is certainly no greater efficiency in the smaller shops. Rather the 

 contrary. And the tale of bankruptcies, which is particularly sordid and 

 tragic in the retailing trades, would go to suggest that the multitude of 

 small shops is due to a continual trial and error process. Men without 

 any experience or much capital are continually trying to set up as shop- 

 keepers and almost as continually making an error. And the error involves 

 the consumer as well as himself. If for lack of experience and capital a 

 shopkeeper can only sell, say, £450 worth of goods he must charge a 

 margin of 33^ per cent, for his own gross profit in order to have a bare 

 income of £150 a year. Thus, with a small total of sales a huge margin 

 of profit (on sales) is quite compatible with a low (total) profit and forms 

 a vicious circle, for the high margin keeps up high prices and checks sales 

 further. With more sales per shop, however, a fair (total) profit is com- 

 patible with a low margin of profit (on sales) ; and a virtuous circle is 

 engendered. The low margin of profit permits lower prices, these in 

 turn stimulate more sales, and these in turn enable lower margins to be 

 charged which in turn permit lower prices . . . ad infinitum. 



§ 4. The Size of Firms. 

 The British and American censuses limit themselves to the size of 

 plants, whereas State and private industrial policy is equally, if not more, 

 concerned with the size of the firm or unit of control. Marketing and 

 central selling schemes, for instance, whether in agriculture or mining, 

 form an attempt to enlarge the business unit controlling several plants or 

 mines rather than to enlarge any one plant or mine. The Government's 

 Economic Advisory Council reported in 1930 that in the cotton industry 

 ' the first essential step in the path of recovery ' was the formation of 

 amalgamations large enough to secure economies of merchanting and 

 finance rather than building larger mills or sheds. The policy of the 



