Skip to content

The New Economic Geography of a SADC Free Trade Area

Literature › Works

Hess, Simon Peter (2004). The New Economic Geography of a SADC Free Trade Area. Master of Economics thesis, Rhodes University. Rhodes University, Grahamstown.
Type thesis
Scope Post-1994 South African economy
Method descriptive quantitative
Period studied 1970–1999
Regions SADC, Southern Africa, South Africa, European Union, Mercosur
Themes manufacturing industrial policy, comparative theory, methods data sources, transition post apartheid, labour markets wages
Citation key hess2004

Argument

Using the new economic geography (NEG) framework, the thesis asks whether a SADC free trade area will pull manufacturing towards South Africa at the expense of smaller members. It computes locational Gini coefficients on UNIDO employment data for 28 ISIC industries in SADC (1970-1999) and compares them with the EU and Mercosur (1980-1999). Concentration rose to the mid-1990s and then eased, and much of the rise was towards peripheral countries rather than South Africa, so effects are industry-specific. The thesis concludes that mass polarisation towards South Africa is unlikely, given wage gaps and preferential access to overseas markets, provided intra-regional transport costs fall quickly.

Evidence and method

UNIDO (2003) and TIPS database manufacturing employment by ISIC rev 2 industry for 11 SADC countries (6 benchmark years, with nearest-year substitutions), plus EU and Mercosur data. Locational Gini coefficients, South African employment shares, location quotients, and a simple regression on average firm size as a scale-economy proxy. A large literature survey of NEG theory and empirics, and secondary sources on SADC, SACU, COMESA and bilateral trade agreements.

Findings

  • South Africa dominates regional manufacturing output and GDP, which drives the polarisation fears motivating the study. (p. 17)

    South Africa’s overwhelming economic and political dominance within the region, to which it contributes 77 percent of regional gross domestic product (GDP) and 84 percent of manufacturing value added (MVA)

  • Average SADC locational Gini rose from 1980 to 1990, rose only marginally to 1995, and then fell, indicating early dispersion. (p. 122)

    From 1990 to 1995, the Gini coefficient increased marginally by 0.0085 and appeared to peak at this level. This could potentially have been a turning point for the region as, by 1999, the Gini fell by 0.011

  • Including South Africa in SADC did not cause mass relocation of industry, contrary to the fears of initial members. (p. 122)

    contrary to the expressed fears of initial members of SADC, the inclusion of South Africa did not lead to the mass re-location of industry.

  • Textiles and apparel became concentrated in peripheral countries as South African employment fell, reversing their 1980 status. (p. 115)

    Both the textile and wearing apparel industries changed from being two of the least concentrated industries in 1980 to the being the most concentrated in 1999.

  • The Gini is a relative measure, so a high Gini can reflect concentration in small countries, as with tobacco, rather than a pull to South Africa. (p. 105)

    the Gini will be higher for a small country with a high degree of concentration of a particular industry, even though a larger country may possess an overwhelming majority of the industry.

  • Six industries were pulled towards South Africa over the whole period, while eight concentrated in the periphery. (p. 141)

    Industries that have been pulled towards South Africa throughout the duration of the analysis are food products, printing and publishing, industrial chemicals, petroleum refineries, miscellaneous petroleum and coal products, and electrical machinery.

  • Scale-intensive industries tend to be agglomerated in SADC. (p. 145)

    Eight of the 12 industries agglomerated above the average in SADC were also in the 12 industries most subject to scale economies.

  • Cross-bloc comparison: Mercosur has the lowest aggregate concentration and the EU concentrated fastest, but all three show dispersion trends in the late 1990s. (p. 157)

    Mercosur, was found to have the lowest aggregate level of concentration with a Gini of 0.08 in 1999. This compares with Ginis of 0.28 for the EU, and 0.22 for SADC.

  • Large wage differences between South Africa and the rest of SADC may explain early dispersion, as does preferential access to foreign markets. (p. 151)

    the South African average labour cost in urban areas was estimated to be US$ 1.89 in 1999 (McCarthy, 1999:388). Although this figure decreases to US$ 1.08 in the former homelands

  • Policy conclusion: intra-regional transport costs (tariffs and non-tariff barriers) must fall quickly if the periphery is to gain. (p. 159)

    it is imperative that transport costs (both in the form of tariffs as well as NTBs) are reduced as quickly as possible in order for the peripheral countries to gain significantly.