This dissertation investigates the Granger-causality between exports, imports, and economic growth in 32 developing countries. The role of the imports variable in the investigation of exports-output causality is emphasized, enabling one to test for the cases direct causality, indirect causality, and spurious causality between export growth and output growth using the definitions provided by Hsiao (1982). On the technical side, the Augmented-Dickey-Fuller and Phillips-Perron tests are used to test whether exports, imports and per capita real GDP are cointegrated. Furthermore, tests for breaking trends are carried out using a modified version of the procedure advanced by Perron (1989) and Zivot and Andrews (1992). Depending on the results from these tests, vector autoregressive or vector error correction models are specified for each country. The election of lag lengths is done by using the Schwarz criterion. Following Poskitt and Tremayne (1987) a procedure for testing the decisiveness (strength) of (Granger-) causal findings is developed. Finally, the results are compared to those from a benchmark model, that is, a bivariate VAR model of export growth and output growth. The empirical results indicate the following. First, breaking trends occur frequently in macroeconomic time series. Failure to test for breaking trends may lead to the possibility of spuriously cointegrating relationships. Second, exports and per capita output were not found to be cointegrated in most cases. Third, there is only weak evidence for a positive Granger-causal relationship from export growth to output growth--the only two cases being Thailand and Pakistan. Fourth, the findings may lack decisiveness, and additional data may prove useful to draw more conclusive inferences. Fifth, results from the benchmark bivariate VARs indicate additional cases of positive causality from export growth to output growth. However, these cases should be regarded as spurious since the findings have not been confirmed when the real imports are included. Finally, there is no evidence of indirect causality from exports to output growth via the real imports variable. Thus, no causal evidence was found for the "exports as a foreign exchange source" hypothesis.
| Identifer | oai:union.ndltd.org:uottawa.ca/oai:ruor.uottawa.ca:10393/9467 |
| Date | January 1995 |
| Creators | Atukeren, Erdal. |
| Contributors | Bodkin, Ronald G., |
| Publisher | University of Ottawa (Canada) |
| Source Sets | Université d’Ottawa |
| Detected Language | English |
| Type | Thesis |
| Format | 204 p. |
Page generated in 0.095 seconds