The Spatial Components of Intra-Urban Manufacturing Change: Suburbanization in Clydeside, 1958-1968

In an investigation of the suburbanization of manufacturing industry in Clydeside between 1958 and 1968 the effects of three components of industrial change-plant births, plant deaths, and intra-urban transfers-are assessed. It is shown that centrally located industries in 1958 tended not to suburbanize over the period, and that for those industries that did move away from the city centre, it was the relatively peripheral locations of plant births that appeared to exert the principal influence upon the change, rather than the effects of intra-urban transfers. THIS paper reports the results of a study of the suburbanization of manufacturing industry within a large urban area. The term 'suburbanization' has been employed to denote the phenomenon under scrutiny, rather than the more commonly used expression 'decentralization', because this latter term has become somewhat ambiguous, having been used to describe spatial change at both the inter-regional (Keeble, 1968; Parry, 1963; Zelinski, 1962) and the intra-urban (Firn and Hughes, 1974; Keeble, 1968; Zelinski, 1962) scales. Only intra-urban change will be considered in this essay. The novel aspect of the paper will be to identify within an inter-industry comparison the relative importance of three components of spatial change-plant deaths, plant births, and intra-urban transfers-in order to demonstrate the complexity of the suburbanization process taking place within the study area: Clydeside between 1958 and 1968. CONCEPTS AND MEASUREMENTS As already noted, the relative shift of manufacturing establishments to more peripheral locations within the conurbation of Clydeside will be termed suburbanization. Suburbanization will be assessed by calculating the difference between the median distances of plants from the centre of the Clydeside conurbation at two points in time, 1958 and 1968. A shift in the median location of plants of any distance away from the centre of the conurbation will be accepted as indicating a suburbanizing trend. The term is thus not intended to be confined to change having its destination in the suburbs-if indeed such an area could be precisely defined (Johnson, 1974). Of the simple descriptive statistics-the mean, mode and median-it was decided to employ the median to describe the distance of plants from the centre of Clydeside because graphs of establishment frequencies against distance from city centres are typically highly positively skewed. Hence, the mean would be inflated too readily by only a few large distance values and therefore overestimate the general trend, the mode would be too inflexible since only the distance value with the greatest frequency would be taken into account, and the median would be the most appropriate indication of distance since it would be heavily influenced by the distance categories with the highest frequencies but at the same time not be dominated by them. The centre of Clydeside from which the median distances were calculated, was located at