No CrossRef data available.
We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
Published online by Cambridge University Press: 07 November 2014
The analysis of trends in industrial location is one tool with which regional economic development may be examined. Changes in the pattern of location are indicative of the continuing alteration of cost relationships within the profit-maximizing firm. From the broader standpoint, changes in the locational pattern of manufacturing activity may have a considerable impact on the distribution of population. Finally, industrial dispersion poses problems of resource management and allocation among agriculture, housing, industry, and conservation.
This study deals exclusively with the location of manufacturing industry. Its main purpose is to identify various trends towards dispersion or concentration of manufacturing among the cities and counties of the province of Ontario. In Section II input and output measures within the context of the production function are briefly discussed. Following the earlier work of Slater, manufacturing employment is adopted as an input measure. Value added is selected as an output measure with which to assign appropriate weights to individual manufacturing units so that they may be summed to an aggregate measure of activity. Descriptive analyses of changes in the locational pattern as revealed by percentage shares and Lorenz curves are reported in Section III. In preference to either of these, however, the coefficient of variation is stressed as a method to describe the distribution of manufacturing industry. Yearly estimates of the coefficient of variation for value added and employment distributions are developed in Section IV. Time trends are fitted to these annual estimates, and the meaning of the trends discussed. The results of the study are summarized in Section V.
This paper was written while the author was associated with Brown University. He is indebted to John W. L. Winder and Murray A. MacGregor for their advice during preparation of an earlier version of the paper. Philip A. Neher has contributed several useful suggestions. While thanking these gentlemen, the author wishes in no way to implicate them in any errors which may still remain.
1 This point has been emphasized recently by D. W. Slater. See his “Decentralization of Urban Peoples and Manufacturing Activity in Canada,” this Journal, 02 1961, 71-8Google Scholar, hereafter cited as “Decentralization.”
2 The aggregate data include seventeen groups of manufacturing industries: Foods and beverages, tobacco and tobacco products, rubber products, textiles, knitting mills, clothing, wood products, paper products, printing, publishing and allied products, iron and steel products, transportation equipment, non-ferrous metal products, electrical apparatus and supplies, non-metallic mineral products, products of petroleum and coal, chemicals and allied products, and miscellaneous industries.
3 Slater, , “Decentralization,” and “Trends in Industrial Location in Canada,” Resources for Tomorrow (Ottawa, 1961), I, 409.Google Scholar
4 These shortcomings have been recognized by several writers who placed the onus of measuring industrial activity entirely upon employment data. See, e.g., Goodrich, C. et al., Migration and Economic Opportunity (Philadelphia, 1936)Google Scholar; Creamer, D. B., Is Industry De-centralizing? A Statistical Analysis of Locational Change in Manufacturing Employment, 1899–1923 (Philadelphia, 1935)Google Scholar; Kitagawa, M. and Bogue, D. J., Suburbanization of Manufacturing Activity within Standard Metropolitan Areas, University of Chicago Studies in Population Distribution, no. 9 (Oxford, Ohio, 1955)Google Scholar; D. W. Slater, “Decentralization,” and “Trends in Industrial Location in Canada.”
5 Data on capital employed in Canadian manufacturing industry have not been collected since 1943, presumably because of difficulties in maintaining accuracy in this type of data.
6 Ontario data for manufacturing employment and value added in production are collected for cities, metropolitan areas, and counties. Unfortunately, this system has not been extended to townships, nor are data for all cities, towns, and rural areas collected individually. This failure is in part due to the application of the disclosure rule, which forbids publication of data for areas with less than three manufacturing units.
7 This method is used by Creamer, Is Industry De-centralizing? and Slater, “Decentralization.”
8 Apart from the inclusion of Algoma and York, the selected counties are identical to those singled out by Bell and Stevenson in their recent discussion of economic health. See Bell, W. H. and Stevenson, D. W., “An Index of Economic Health for Ontario,” Ontario Economic Review, 09 1964.Google Scholar
9 Fourteen counties were excluded on the basis of lack of manufacturing activity or because they fall into a region generally recognized as Northern Ontario. They were: Algoma, Bruce, Cochrane, Haliburton, Huron, Kenora, Manitoulin, Muskoka, Nipissing, Parry Sound, Rainy River, Sudbury, Thunder Bay, and Timiskaming.
10 Stigler, G. J., The Theory of Price, rev. ed. (New York, 1952), 265.Google Scholar
11 Using the same forty counties, certain highly industrialized counties are subdivided into city and the county area outside the city. Three metropolitan areas outside the city are included, together with county areas outside the metropolitan areas: (i) The remainders of counties: Brant, Essex, Halton and Wentworth, Lincoln, Middlesex, Waterloo, and York. (Note the loss of one county unit due to the necessity of treating Wentworth and Halton as one county outside Metro Hamilton); (ii) The cities of Brantford, Windsor, Hamilton, St. Catharines, London, Kitchener, and Toronto; (iii) The remainders of metropolitan areas: Metro Windsor, Metro Hamilton, and Metro Toronto.
For purposes of convenient comparison through time, the areas of all cities and metro areas are held constant at their 1945 size. This was achieved by using a correction formula to relate measures to the 1945 base size. The underlying assumption of this formula is that most cities have annexed land for the purpose of acquiring rate revenue of residential subdivisions (present or future), and only in a few instances are industrial sites involved in an annexation. Thus, we argue that annexed land may be expected to be industrialized to the level that exists in the residue of the respective county without the city or urban area.
The correction formula employed is as follows:
For year t,
and
Where Z may be termed the “corrected” measure per square mile of urban area “core,” for year t. The sources of city, metro area, and county size data are given in the Appendix.
12 Dates of business cycles are taken from Chambers, E. J., “Canadian Business Cycles since 1919: A Progress Report,” this Journal, 05 1958, 166–89Google Scholar, and dates selected by the Department of Trade and Commerce, Ottawa.
13 If we assume a very general two-input aggregate production function for manufacturing industry, of the form:
V = A Kk Ll
where V is value added in production, K is the flow of services from capital inputs, L is similar flow from labour inputs, and A is the “residual,” all for a given time period, then it is clear that growing dispersion of value added relative to employment may be due to changing capital intensity or increased values of A, the “residual.” This residual combines the effects of levels of technology, quality of labour, and returns to scale, amongst others. For a fuller discussion of the components of the “residual” in a growth context see sec. v of P. A. Neher and K. A. J. Hay, “Misallocation, Education and Economic Growth,” forthcoming.
14 Specifically those cities and metro areas which show strong positive absolute per annum rates of change in the value-added density measure are Hamilton, Kitchener, London, Metro Toronto, and Metro Hamilton. Toronto, Windsor, Brantford, St. Catharines, and Metro Windsor did not demonstrate any rate of change significantly different from zero (at the 1 per cent level of significance) during the period under discussion.
15 Wildgen, F. X., “Small Business and Restrictive Monetary Policy,” Staff paper, Royal Commission on Banking and Finance, 1962.Google Scholar For a concise discussion of this problem with respect to the United States, see Carson, D., The Effect of Tight Money on Small Business Financing (Small Business Administration, Washington, 1963).Google Scholar