This resource is hosted by the Nelson Mandela Foundation, but was compiled and authored by Padraig O’Malley. It is the product of almost two decades of research and includes analyses, chronologies, historical documents, and interviews from the apartheid and post-apartheid eras.
5.1 Regional structure and performance
The range of characteristics involved in location decisions is demonstrated clearly by the varied experiences of Regions E and F. Region F, as noted earlier, is relatively land and resource rich, while Region E's advantages include access to ports, a large urban agglomeration (Durban-Pinetown) and generally good rainfall. Region E is comprised of Natal, all of KwaZulu and part of Transkei, making it a very populous and mostly black region compared to Region F.
Table 12 locates Regions E and F relative to other regions and South Africa as a whole for a number of socio-economic indicators. The last two columns of the table show that both regions have maintained almost constant population shares and density between 1970 and 1990, but that Region E is both much larger and more dense in population. In fact, Region E in 1990 was the most dense if one should exclude the heavily urban Region H, while Region F was below the national average density. Income disparities are also wide.
Personal income per capita in Region F in 1985 was 91 per cent of the national average and the GGP per capita in 1989 was 181 per cent of the average. For Region E personal income per capita was 67 per cent of the average and the GGP per capita 63 per cent for the same years. Since the gap between GGP and personal income per capita is a rough indicator of net tax and transfer incidence, these figures suggest that Region E did indeed benefit while Region F lost from redistribution. A caveat, however, is that ownership of Region F resources is disproportionately absentee. Also, profits tend to flow out of the region to a larger extent than from other regions, so that private rather than state redistribution may explain the gap between GGP and personal income per capita.
Both regions grew at higher than average rates: Region F grew at 3,69 times and Region E at 1,31 times the South African average. Other similarities are urbanisation rates of 59 per cent of the South African rate in both regions and below-average functional urbanisation rates (79 per cent of the national average in Region E and 86 per cent in Region F) in 1989.
Moving down the rows of Table 12, the regions again diverge for the remaining indicators (except education level). After growing faster than average in population between 1970 and 1980, Region E fell to 75 per cent of the national growth rate during 1985 to 1990. For Region F the trend was reversed: the region grew more slowly than average in the earlier period but 7 per cent faster than average in the later years. Other indicators reveal that population structure is more favourable in Region F than E, although differences may be narrowing slowly. In Region E females comprised 55 per cent of the population in 1989 compared to 44 per cent in Region F, while the dependency ratio in Region E was 2,8 compared to 1,6 in Region F. Similarly the economically active population included only 26,4 per cent of Region E's population but 38,5 per cent of that of Region F. The participation rate was only 49,4 per cent in Region E relative to 66,3 per cent in Region F, and the informal sector participation rate was more than twice as high in Region E (31,8 per cent) as in Region F (14,8 per cent). Working in the opposite direction and reflecting the higher density in Region E are the medical indicators which are substantially better in this region. Educational levels are roughly equal between the two regions.
Region E is therefore poorer and more densely populated, with a larger share of females and dependents in the population and a more extensive informal sector. Taken together, these indicators depict a regional economy likely to be dominated by low-wage labour. The dual nature of Region E's economy does provide some grounds for optimism, since the transportation and communications infrastructure, as well as health and education, are better developed. The Region F profile is more encouraging in growth and employment prospects, but with significant qualifications. Growth has been concentrated both sectorally and spatially, and is particularly vulnerable to changing market conditions. Table 1 shows that the personal income per capita in 1985 of the Transvaal portion of Region F was only R2875, making it one of the poorest of the 'white' areas in the country. In Region E, on the other hand, Natal had a personal income per capita of R4223. Therefore, despite rapid industrialisation and growth in Region F, the population in all areas was relatively poor in 1985.
Inequality within former development Regions E and F exhibits somewhat different patterns than for the country as a whole. Table 13 allows comparison of inequality among the former homeland areas of Region E with inequality among non-homeland areas.
By coefficient of variation the largest inequality in GGP per capita is for all districts, with a coefficient of 1,3213, which is extremely high. For non-homeland magisterial districts the coefficient is 0,5074 and for homeland districts 0,4740. This implies that the greatest source of intraregional variation is between the two groups and that within each group variation is approximately equal. At the national level, variation among homelands was greater than variation across provincial areas. What the magisterial level data reveal compared to those of the national level is the existence of pockets of poverty within non-homeland areas of Region E. The incidence of poverty is therefore greater in, but not limited to, former homelands. At the national level, this poverty is obscured by the level of aggregation to subregions rather than magisterial districts.
Another important result from Table 13 is that urbanisation rates are twice as variable in the homeland as in the non-homeland districts. Policies aimed at the urban poor will therefore have a widely varying effect on the homeland districts.
Table 14 provides the results of an analysis of variance which calculates the effect of residing in a non-homeland versus a homeland district. Here among-group variance is variance between homeland and non-homeland districts, and the residual is variance within the homeland group. The advantages of the non-homeland districts are shown in the table as the 'Natal premium'. Premiums for all but the population density variable are significant at the 1 per cent level. Positive premiums are found for participation rate, urbanisation rate, economically active population shares, GGP per capita and male absenteeism. Negative premiums appear for population growth rate, unemployment rate, percentages of the population that are black and/or female, and the dependency ratio. As mentioned, it is noteworthy that for all but the population density variable the homelands are significantly worse off than the rest of the region. Indeed, the GGP gap between Natal and the former homeland districts is R669 or 85 per cent of the average homeland GGP per capita of R781.
Inequality within Region F is shown in Table 15 for the more limited set of variables available for this region's magisterial districts. Overall, the region is more unequal with respect to these variables than the country as a whole and this is particularly true for the non-homeland areas. Moreover, looking at changes over time it is clear that the region has become even more unequal.
Coefficients of variation for all the variables in Table 15, except population growth rates, have increased for the entire region. However, changes over time vary for the homeland and non-homeland groups. Non-homeland inequality has diminished with regard to economically active and participation rates. Homeland inequality has risen along with inequality across the entire region with the exception of the urbanisation rate, which has become less unequal over time in the homelands.