In this study an attempt has been made to evaluate the validity of the boundaries of the officially approved ‘development regions’ of South Africa. A variety of statistical methods were used for the evaluation of the empirical validity of the boundaries of these regions. Firstly, two sets of homogeneous regions based on different sets of parameters were delimited. Secondly, two sets of regions indicating areas of different levels of economic development were determined. With the aid of this information, some changes in the boundaries of the existing regions are suggested. A historical account of attempts at regionalisation in South Africa for development management purposes is also given. A semantic assessment of definitions points to the need for a change in the use of the term ‘development region’ — for this specific purpose — to the term ‘development management region’.
Notes
Senior lecturer, Department of Town and Regional Planning, University of Potchefstroom.
Head, Statistical Consultation Service, University of Potchefstroom. Prof Steyn was responsible for the execution of the cluster and the discriminant analyses as well as the TEPI map and gave his assistance in the interpretation of results.
Senior Statistical Consultant, Statistical Consultation Service, University of Potchefstroom. Mrs Uys was responsible for the computerising of basic data.
Head, User's Support, Computer Service Department, University of Potchefstroom. Mr van der Walt was responsible for the reproduction of computerised maps.