Externally constrained growth: Testing the applicability of Thirlwall’s law in South Africa

ABSTRACT

This study tests whether the performance of South Africa’s trade balance, that is, the country’s export and import performance, can effectively explain the economy’s growth rate. Formally, the study tests the applicability of Thirlwall’s law to the South African economy. The law states that the estimated growth rate $\left({\mathrm{y}}_{\mathrm{B}\mathrm{t}}\right)$ of an economy is proportional to the growth rate of exports $\left({\mathrm{x}}_{\mathrm{t}}\right)$ divided by the income elasticity of imports$\left(\pi \right)$. The South African Reserve Bank (SARB) quarterly data from 1960 to 2009 for exports, imports, the exchange rate, export prices, import prices, and the economic growth rate is used for regression. The study runs an autoregressive distributed lag (ARDL) model in the presence of structural breaks and after adjusting for structural breaks and finds that growth in the South African economy can be shown to be trade balance constrained. This finding has policy implications with regard to trade promotion and strategies and strategies to enhance growth performance by increasing the overall competitiveness of the South African economy.

KEYWORDS:

• Open economy Macroeconomics
• economic growth
• export promotion
• competitiveness
• trade balance

JEL Codes:

• F41
• F43

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 The following South African Reserve Bank data series are utilized in this study. The volume of imports from the SARB series 5034 is used. For the growth rate of exports, the SARB time series 6013 is used. For the price of imports, SARB series 5035 is used and for the price of exports SARB, series 5031 is used. For the nominal exchange rate, end of quarter data from SARB series 5339 is used. For the income proxy, expenditure on GDP in real terms from SARB series 6045C is used. To calculate the rate of GDP growth percentage change SARB series 6006 is used.

2 As above, the following SARB data series are used for the volume of imports SARB 5034, for the growth rate of SARB 6013, for the price of imports SARB 5035, for the price of exports SARB 5031, for the nominal exchange rate SARB 5339, for expenditure on GDP in real terms SARB 6045, and for GDP growth SARB 6006. Various stationarity tests indicate that in the general the growth rates of imports and income, as well as changes in the real exchange rate are stationary after first differencing. All the data is seasonally adjusted.

3 The period from 1970 to 1972 is not included as the income elasticity of demand for imports is insignificant at 5 percent level of significance.

4 A study conducted by (Kabundi, 2014) showed evidence for South Africa that net exports, boosted by a weaker real effective exchange rate, improved the trade balance in the period 1994–2011.