It is generally acknowledged that there is no sufficient, exhaustive and elaborate empirical examination of the quantitative impact of policies pertaining to import demand and economic growth in South Africa. In order to arrive at conclusive, sagacious and applicable policies on the economic growth potential of an economy, it is imperative to evaluate, empirically, whether envisaged economic growth rates and employment creation are feasible, given the socio-economic circumstances.
The fundamental question of the constraint or rather effective constraints to high economic growth rates, measured by gross domestic product, has always desired urgent attention but has been neglected. There appears to be strong reasons to believe that the South African economy, like other middle-income developing economies, is subject to a "powerful balance of payments constraint that effectively aborts the growth process before it is able to deliver rising per capita incomes" (Industrial Strategy Project, 1995 :49 ).
Furthermore, although this issue is widely recognized, there has been little systematic analysis of this important question. Many writings which, implicitly or explicitly, note the foreign exchange shortage as adversely affecting the economy's growth capacity have tended to focus and give enormous emphasis on exports and export expansion as a means to eradicate this economic dilemma. However, together with exports the demand for imports clearly determines the behavior of the trade account of the balance of payments as a whole. Consequently, this paper intends to consider one important aspect of the balance of payments constraint, namely, the import performance and import demand elasticities.
The study derives the import demand function and applies the recent time-series techniques to modeling economic time-series. Prior to the empirical model, the paper describes the behavior of imports. This section examines the cyclical and trend behavior of import performance since the beginning of the 1970s. The study also briefly looks at the relationship between import of capital goods and investments in South Africa. The geographic origin of imports by regional trading blocks is also discussed. That is followed by an extensive literature survey conducted on import demand elasticities in South Africa and trade elasticities in general. The import performance and import demand functions were studied in an economic policy context and the analyses were in some cases restricted by data constraints. Import behavior patterns and empirical results of the import demand models are discussed and international comparisons are drawn.
There are a few basic points that emanate from the overall discussion. In the import performance section, it can be concluded that labour intensive commodities have the largest share in total imports; there is a very steady, insignificant decline in import penetration ratios and these have increased lately and that import of capital goods is positively correlated to investment.
The description of studies shows that the demand for imports is largely influenced by economic activity as compared to relative prices. Some of the results are shown in the appendices and discussed in text, where comparisons are made between the results of different studies and the main findings of this study. Precisely, the major finding is that, as other studies concluded, the propensity to import with respect to income is more significant than the price elasticity of demand for imports.
The import performance findings combined with time-series estimation results raise doubts to envisioned employment creation levels and economic growth rates in South Africa. This is questionable because South Africa's imports have been on an increase whilst exports have not performed well. From the time-series point of view and based on estimation results, the current economic strategies should also address the import demand question or foreign exchange and domestic economy development if the projected growth rates and employment levels are to be achieved.