Effect Of Comesa FTA On Trade-Three Empirical Approach

The aim of this paper is to estimate the effects of formation of the COMESA FTA on the bilateral trade of African countries over the long run (2000-2013) and the short run (2000-2003). Using the most recent data from 20 African countries, the paper uses three different methodologies [parametric; semi-parametric; and Non-parametric approaches] in order to achieve the following sub objectives:1) to investigate whether COMESA FTA leads to trade creation between member countries; 2) to investigate if there is a difference between long run and short run effects of COMESA FTA on trade; 3) to investigate whether there is consistency in results between Parametric, semi parametric and non-parametric approaches.Over both the long and short run, the following are the estimation techniques used under the mentioned methodological approaches:Parametric approach in this study involves the use of the gravity model estimation; Non Parametric approach employed matching estimation technique; and Semi Parametric approach involved the combination of the Difference-in-Difference estimation technique and the matching estimation technique. The study uses these three methodological approaches because of the following reasons:1) Parametric techniques have always been the mainstream estimation technique in trade literature and looks at the average changes in trade across the period; 2)Non parametric techniques control for the parametric technique’s unrealistic assumption of linearity between variables by allowing for non linearity and also uses comparable groups for estimation; 3) Semi parametric techniques look at the average changes in trade across the time period and also accounts for non linearity between variables, while using comparable groups for estimation.Firstly, the impact of COMESA FTA on bilateral trade was analyzed over the long run, covering the period from 2000-2013. With the use of the three methodological approaches, the study found that both Parametric and Non Parametric estimation techniques yielded a positive and significant impact of COMESA FTA membership on export flows. The Semi Parametric estimation technique however yielded negative and significant results. This different result (negative) from the semi parametric technique is due to the downward bias that occurs when endogenous regressors are used in semi parametric estimation techniques; a negative correlation between variables of interest and omitted variables is the cause of the negative bias as observed in our study.Secondly, the impact of COMESA FTA on bilateral trade was analyzed over the short run, covering the period from 2000-2003.With the use of three methodological approaches, the study found that both Parametric and Non Parametric estimation techniques yielded a positive and significant impact of COMESA FTA membership on export flows. The Semi Parametric estimation technique however yielded negative and significant results. This different result(negative) from the semi parametric technique is due to the downward bias that occurs when endogenous regressors are used in semi parametric estimation techniques. This downward bias is caused by a negative correlation between the variables of interest and omitted variables. The short run results are therefore in line with the long run effects.In line with the objectives of the study therefore, the paper concludes:1) the formation of COMESA FTA leads to trade creation amongst the members of the free trade area; 2) there is no difference between the long run impacts and the short run impacts of COMESA FTA on trade; 3) there is consistency of results in all three estimation techniques over the short and long run.The study also concludes that in terms of Trade literature, Parametric and Non parametric estimation techniques give consistent and reliable results whereas Semi Parametric estimation techniques give unreliable results.