Measuring the Effectiveness of Econometrics Methods in the Study of Productivity in Syria: Comparison between Cointegration and OLS

Abstract

In line with the increasing use of econometrics applications in economic research, one could easily notice that the conditions of these applications are, to some degree, neglected which has serious complication on the results attended. This study sought to discuss the differences related to the use of different econometric methods in estimating the relationships among economic variables. To accomplish this aim, the study estimated productivity function in the case of Syria for the period 1985- 2010, twice, in the first time by applying the rather new econometrics method of cointegration, and in the second time by the Ordinary Least Squares OLS. The empirical results showed high importance of physical capital in enhancing productivity level in Syria. While, the contribution of education human capital was negative and in consistent with relevant theories but in line with results of similar studies in the case of Syria. Heath human capital variable was significant with the expected sign in most models. The study found complicating and inconsistent results from the two methods, and that models estimated by OLS suffered of several econometric problems such as autocorrelation, hetrosekdasticity, and structural instability of coefficients. Whereby cointegration models passed such problems.