Notes and Comments An I ntertemporal Calculation of Effective Rates of Protection for South Africa *(1) MERLE HOLDEN AND PAUL HOLDEN *(2) THE PURPOSE OF THIS STUDY is to calculate rates of effective tariff protection for South Africa for two points in time. This is of interest for a number of reasons. Firstly, there has so far been no published calculation of effective rates of protection fo r South Africa. Secondly, it is important to know how nominal rates of protection change over time compared with changes in effective rates. Frequently, the structure of nominal protection is so complex that the ultimate protection accorded to industry can not be ascertained. It is therefore advantageous to have some knowledge of effective tariff rates in evaluating and rationalizing commercial policy. Finally the determination of how effective rates have changed through time is an important first step in as sessing the influence of effective tariff rates on resource allocation, an area where there remains an important gap in the effective protection literature. Economists have long realized that nominal tariff rates do not provide an accurate indication of the influence of protective policy on resource allocation. However, it was only in the 1950s that the concept of effective protection was conceived and a satis factory analytic foundation remained undeveloped until the mid-1960s. Since then the theory and application of effective protection has commanded the attention of an increasing number of economists, effective tariff rates have been computed for many countr ies, both developing and developed, and the underlying theory has been extended. The theory of effective protection provides an understanding of how nominal tariff rates affect a country's pattern of production. Effective protection can be most simply defined as the proportionate change in the value added of a process due to the tariff structure. A simple example should serve to illustrate the difference between nominal and effective rates. Let us assume that a country can manufacture shoes at a world price of $10 a pair; $8 of inputs such as leather and buckles are required per pair, a nd the value added by labour and 1 975 SAJE v43(3) p371 capital in the final stage of manufacture is $2 per pair. To simplify, assume that the prices of the non-labour and non-capital inputs are equal to their world prices. The government of the country now decides to place a nominal tariff of 20 per cent on sh oes, and 10 per cent on other traded goods. This raises the price of shoes to $12 and other inputs to $8.80 per pair. This means that the shoe manufacturing process can now spend $3.20 per pair on labour and capital. The theory of effective protection infe rs that given a nominal tariff of 20 per cent on the final good with a nominal tariff of 10 per cent on other traded inputs, the effective tariff protection given to the shoe manufacturing process is 60 per cent. The effective rate of protection f j can be defined as the proportionate increase in value added in an industry due to the tariff' structure, - relative to what the value added would have been in the absence of tariff protection. Then where V' j is the value added at domestic market prices and V j is the value added at world prices. A number of assumptions underlie the model which is used to calculate the effective rates of protection for South Africa. International prices of both imports and exports are assumed to be fixed. In other words, the demand for South African exports and the supply of imports are both perfectly elastic. This is the standard small country assumption. In the calculations the prices were normalized to unity. The input-output coefficients are assumed to be fixed for both the intermediate goods and the primary fac tors. No exchange rate adjustment is made. It is assumed in addition that tariffs are the only form of market distortion. This is made necessary by the lack of comprehensive price survey data in South Africa which would enable a comparison to be made between domestic and world prices for each indus try. Thus an incomplete picture is obtained of the true protection granted to each industry. At the worst this implies that the calculations made reflect only effective tariff protection rather than the total effective protection granted to each industry. In the years for which effective rates were calculated, however, it is probably safe to say that non-tariff protection was not at a maximum, so the incompleteness of the picture is not as great as it might otherwise have been. The final assumption is that despite input-output coefficients being fixed, value added per unit of output of each industry rises with output. This implies that the supply of primary factors is less than perfectly elastic. Two formulae are used to calculate effective rates of protection for South Africa. The first, the simple Balassa formulation, *(3) is: 1 975 SAJE v43(3) p372 Where f j is the effective rate of protection of the j th industry; X' j is the value of the output of the j th industry at domestic market prices; X' ij is the domestic value of the i th input used in the j th industry; T Dj are the indirect taxes (including tariffs ) levied on the 223