FUNCTIONAL TRANSPORT REGIONS IN SOUTH AFRICA: AN EXAMINATION OF NATIONAL COMMUTER DATA

Transport authorities have been enacted by the recent National Land Transport Bill (B51 – 2008) with the objective to improve transport service delivery by grouping transport functions into a single, well-managed and focussed institutional structure. These authorities operate at the municipal level of government and may consist of single or multiple municipalities (or part of municipalities) and may even extend across provincial boundaries. In considering the area of transport authorities, the focus is placed on dominant passenger movements (to, from and traversing municipalities) and the economic interdependency between municipalities. This paper illustrates the use of national commuter data, also termed journey-to-work data, to derive functional transport areas based on dominant passenger movements. A clustering algorithm, developed in 1975 by Masser and Brown and termed INTRAMAX, is used to cluster municipalities. The procedure identifies municipalities between which significant numbers of people commute and aggregates these areas into functional units. These aggregated regions have stronger transport connections with each other than with outside areas and can be considered functional transport areas. The functional areas are refined by allocating the commuter flows to the transport network which allows the identification of heavily travelled routes that fall outside the area and which can be considered for inclusion in the transport area. The paper demonstrates that national commuter data are suited to identify the dominant passenger movements and the resulting demarcated functional transport areas maximize intrazonal commuting and minimize cross border commuting. The demarcated areas are a reflection of economic interdependency and indicative of an inclusive labour market area. While the results proved intuitively appealing, several improvements can be recommended, including using more disaggregate commuting data and considering alternative spatial aggregation methodologies such as principle component analysis.