F— ECONOMIC SCIENCE AND STATISTICS 133 



the stage of the exact yet summary measurements which may soon test 

 such an hypothesis of delocahsation. But this is something achieved ! 



A statistical series, such as the location factors and the localisation 

 coefficients of all the particular industries distinguished in the Census 

 of Occupations or the Census of Production for different years, enables 

 the economist to see exactly what is the situation or trend in situation that 

 he is trying to explain. He cannot just use a few well-known locations 

 and localisations as illustrations when it suits his argument, but must 

 frame a theory which, as a working hypothesis, will explain all or at least 

 the bulk of the measured observation in the series before him. It is to 

 be hoped that this working out of standard measures of localisation will 

 now permit a realistic understanding of the forces at work upon which 

 to base effective siting policies. 



So far the problem of site has been confined to the localisation of some 

 particular industry with reference to the population as a whole. But the 

 concentration of industry generally into a relatively confined geographical 

 area is also exercising the State government, as the terms of reference of 

 this year's Royal Commission, mentioned earlier, clearly show. 



It is the problem of urbanisation or, rather, suburbanisation. Just 

 as the localisation of a particular industry presents (external) economies 

 of supply of labour and auxiliary services specialised in that industry, so 

 general localisation presents (external) economies of an accessible supply 

 of general labour and services which, with the wider use of semi-skilled 

 occupations, are not much less important. The scarcity and high rents 

 of urban sites and new transit facilities, however, urge labour and factories 

 to (not necessarily the same) suburbs. Hence the economic inefficiencies 

 of ribbon development and of hours and expenses per man getting to 

 and from work, and the strategical problems connected with population 

 congestion in time of war. What is usually lumped by economists under 

 the simple title of localisation thus appears to have four distinct meanings, 

 all of them of separate importance in forming policy. 



1 . The location of a particular industry in a particular area-. Measur- 

 able by the location factor of that area. 



2. The concentration or localisation of a particular industry anjrwhere 

 in the country. The degree of localisation, measurable by the localisa- 

 tion coefficient. 



3. The location of industry in general in a particular * urban ' area. 

 Measurable by density of occupied population in that area.^ 



4. The concentration or localisation of industry in general anywhere 

 in the country. The degree of urbanisation measurable, e.g. by the pro- 

 portion of the total occupied population in towns or districts over a certain 

 population density. 



Research into the location and localisation of industry in general has 

 recently been undertaken in the United States by economists of the 

 Wharton School of the University of Pennsylvania headed by Prof. 



* Contour lines linking places of equal density of population may be drawn 

 over the map of any country ; and cross-sections cut showing typical successions 

 or cycles of city, suburban and rural areas. See Florence, Statistical Method in 

 Economics and Political Science, p. 116. 



