The Impact of Ghana’s Oil and Mineral Resources on Its Economic Growth.pdf

1、This paper sought to look at the long run effects of Ghana’s oil and mineral industry on its economic growth. T he pivotal role of institutions in determining the effects of natural resources has been widely acknowledged

2、 in the literature. Mehlum，Moene and Torvik for example have demonstrated using regression analysis that the resource curse is strongly present in countries with relatively weaker institutions but is barely present in co

3、untries with strong institutions. Their distinction between strong and weak institutions is buttressed by the far reaching ways in which institutional arrangements encourage manufacturing and production to take place in

4、an economy as against allowing corrupt practices that inhibit economic efficiency.
　　Other authors have also gone a step further by analyzing the effects of institutional design as a component of institutional quality

5、. For the purposes of this paper a dummy variable DEM was used to account for governance. The priori expectation is that it has a positive effect on economic growth. Various resource abundance variables were used in the

6、regressions namely total natural resource earnings as a percentage of GDP，oil receipts as a percentage of GDP and gold export earnings as percentage of GDP . Also unit root tests were performed using the Dickey Fuller an

7、d Phillip Perron tests . These were performed due to their consistent results when dealing with a small sample size. It was noted that most of the time series data had stationary properties. In running the regressions，fi

8、rst an OLS regression was run where the dummy variable was omitted and in subsequent regressions it was addeD.This was to ga uge the effect this variable had on the regression. The Durbin Watson test for autocorrelation

9、consistently showed result s between 1.5 and 2.This implied the absence of autocorrelation.
　　To be certain I performed the Durbin Watson alternative test whose null hypothesis is the absence of serial correlation . B