While the application of the GMM techniques helps in addressing the issues of simultaneity and endogeneity in dynamic panel models, the approach also suffers from instrument proliferations, which has the potential to over fit endogenous variables to produce biased coefficients and weakens the test of instrument validity (Roodman, 2009b)