英语论文网

g yearly dummy variables which pools together the samples from different years and firms. The dummy variables will also capture any measurement errors that might arise. We also estimate robust standard errors that controls for heteroskedascticity of unknown form.

 

The results of a particular year were obtained using the simple OLS estimation and results reported robust standard errors to correct for variations across firms. For all the years Pooled OLS estimation were done and used robust standard errors to correct for variations across countries. To correct for bias that might arise due to variations over time yearly dummies variables were used. Results for a particular year and all the years excluding the natural resource sector were also conducted. Additional simple OLS regression analysis and pooled OLS were also done and reported for South Africa's trade with regional blocks -Africa, Developed Nations and Rest of Asia excluding services sector for 1990 and 2000. The results obtained from the simple OLS estimation for 1980, 1990 and 2000 with and without natural resource sector and also results from the pooled OLS with and without natural resources as presented in Table 5 below. The signs of the coefficients are what we are interested in since theory does not say much about the size of the coefficient.

 

The results revealed that the coefficient of capital stock and unskilled labour are positive and statistical significant for all the three periods as shown by the t-test, meaning that a rise in overall capital and unskilled labour endowments enhances exports that are relatively capital and unskilled labour intensive. The coefficient of skilled labour is negative, meaning that a rise in the overall skilled labour endowment is unlikely to enhance exports. This might be explained by the factor that South Africa is not highly endowed with skilled labour relative to its trading partners. Excluding natural resource services did not change the results as capital and unskilled labour remained positive, while skilled labour remained negative. This shows that South Africa's capital intensity in world trade is not necessarily driven by the natural resource sector as suggested by other studies. For the three periods the three factors explained 30 and 50 percent of the variation in South Africa's trade as shown by the R-Squared ratio and this indicate that the are other factors not included that help explain South Africa's trade.

 

Results from the pooled OLS which looks at all the combined years, but controlling for variation over time reveal that capital and unskilled labour coefficients are positive, while for skilled labour is negative. All three coefficients are statisticaly significant and the factors explain about 40 percent and 50 percent of the variations in South Africa's total trade and trade excluding natural resource services respectively. The coefficients of most of the yearly dummies are positive and significant, meaning that the use of these dummies was correct and it corrected for the variation over time (in the table of results yearly dummies are not reported). These results reveal that in world trade patter本论文由英语论文网提供整理，提供论文代写，英语论文代写，代写论文，代写英语论文，代写留学生论文，代写英文论文，留学生论文代写相关核心关键词搜索。