However the transformed error term is

However, the transformed error term is correlated with since both contain . In disparity to a static model, ordinary least square on the first differenced data in a dynamic model generates inconsistent parameter estimates since . Note that for all h≥2, t=3,…T. This opens up the possibility of using instrumental variable IV estimations using the lagged variables as instruments. Following this fact, Anderson and Hsiao (1982) proposed IV estimation using as instrument for since . Blundell and Bond (1998) argued that if the dependent variable is close to a random walk, the difference GMM performs poorly because the past levels convey little information about the future changes. This makes untransformed lags to be weak instruments for transformed variables. Thus, to increase the efficiency we orthogonality moment conditions that for all i and t. This approach was outline in Arellano and Bover (1995) by transforming the difference of the instruments to make them exogenous to the fixed effects. Hence, the validity of the assumption is that changes in the instrumental variables are uncorrelated with the fixed effect. If this assumption holds, then is a valid instrument for the variables in levels since . Difference and system-GMM estimators are used when N>T. The Nickell (1981) bias disappears in large T panel indicating that the shocks to the country׳s fixed effect has shown by the error term will decline with time and the correlation of the lagged dependent variable with the error term will be insignificant (Judson & Owen, 1997; Roodman, 2009). Dynamic GMM solves the problem of endogeneity than in the static and OLS models that do not allow the use of internally generating instruments. Also, all the variables from the regression that are not correlated with the error term (including lagged variables) can be potentially used as valid instruments (Arellano, 2003; Baltagi, 2005; Han, Phillips, & Sul, 2013; Horváth, Hušková, Rice, & Wang, 2015; Wooldridge, 2002); and it CA-074 Me Supplier accommodates a situation where ‘T’ is smaller than ‘N’ in order to control for dynamic panel bias (Baltagi, 2005; Baum, Schaffer, & Stillman, 2007; Bond, 2002; Roodman, 2009) as in this study. Furthermore, the study employs dynamic panel model using system-GMM in this study because it has advantage over difference-GMM in variable that are ‘random walk’ or close to be random walk variables (Arellano, 2003; Baltagi, 2008; Baum, Schaffer, & Stillman, 2007; Han, Phillips, & Sul, 2013). Since the model specified in this study encompasses macroeconomic variables which are known in economic for the presence of random walk statistical generating mechanisms; the technique also produces more efficient and precise estimates compared to difference-GMM by improving precision and reducing the finite sample bias (Baltagi, 2008). In the literature, performing the tests of over-identifying restrictions - whether the instruments, as a group appears exogenously; either Sargan or Hansen J statistics or both are used. Sargan statistic is reported for one-step non-robust estimation which minimized the value of the one-step GMM criterion function. Further, Hansen J-statistic is reported for one-step robust estimation and for all two-step estimation and minimized the value of the two-step GMM criterion function and Contractile ring is robust. Based on this criterion, this study uses the Hansen J-statistic test to account for the over-identifying restrictions. Also, the time-series data employs by the study are subjected to unit-root tests using Levin, Lin $Chu t* test and Im, Pesaran and Shin W-stat.
Results and discussions
Conclusion Finally, the study concludes that exchange rate volatility has significant effect on private consumption in Sub-Saharan African countries. The implication of this is that majority of individuals cannot afford to consume imported goods. This may be associated to high level of poverty lingering within the SSA countries. However, if exchange rate volatility continues without being controlled; the long-run effect on private consumption may be severe. Then, it effects would be disastrous to economic growth having established in the literature that private consumption contributes a larger proportion of the real GDP (Mankiw, 2012) and a sustainable increase in real GDP leads to economic growth. Hence, this study therefore recommends that government should embark on policy that will discourage exchange rate fluctuations; make local currency stronger; reduce export competitiveness and imports cheaper which can cause the trade deficit to widen further, eventually weakening the currency in a self-adjusting mechanism.