Master of Science (MS)


Agricultural Economics

Document Type



The objective of this thesis was to examine the validity of the Export-Led Growth (ELG) hypothesis in nine Southern African countries using annual data for the period 1980-2002. The thesis used time series econometric techniques to test for the causal linkage between exports and economic growth in Southern Africa. Dynamic econometric models were estimated to test for time series properties: unit root (ADF and PP tests), cointegration (Johansen’s procedure), and Granger-causality (Likelihood Ratio test-LR). The results of the unit root tests show that most of the series are stationary in first differences (series in levels have unit root—I(1)). Co-integration and causality between exports and economic growth were tested and compared using two types of bi-variate vector autoregressive models: models without exogenous variables VAR (p), and models with exogenous variables VARX (p, b). The results of the co-integration tests on both types of bi-variate models show that all three Granger-causality alternative models fit the ELG study for Southern Africa (stationary models; integrated but not co-integrated models; and Error Correction Models). In both types of models, the direction of causation (unidirectional or bidirectional) between GDP and exports was tested using a SUR system of equations by computing the LR test. Without exogenous variables, the ELG hypothesis is found to be valid in Lesotho and Swaziland, and, with exogenous variables, it is valid in Botswana, Lesotho, and Swaziland, implying that expanding exports can contribute to economic growth, poverty reduction, and job creation in all three countries. This research reveals that, even though most countries have adopted export-friendly policies, the long-term impact of such policies is yet to be observed for most countries.



Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Hector O. Zapata