Panel Model Analysis of Impact of Climate Change on Crop Yields in Nigeria: A Fixed Effects Approach

Abstract

The impact of changes in climatic variables (mean temperature and rainfall) and some non-climatic variables on the yields of seven cash crops production in Nigeria was estimated using the fixed effects panel data approach based on a balanced panel of seven crops, for a period of twenty-nine years (1981-2009). The results of the pooled Ordinary Least Square (OLS) show that the impact of climatic factors on crop yields was significant while the fixed effects panel data estimators’ shows that changing climatic factors have no significant effects on crops yields. However, the non-climatic factors were generally significant for both the pooled OLS and the fixed effects estimators. But the coefficients of estimates from the pooled Ordinary Least Square (OLS) were generally larger than that of fixed effects model. A comparison of the  three fixed effects estimators shows that in terms of  efficiency, first difference performed better  than the Least Square Dummy variable (LSDV)  and within estimation methods as it reported a more  smaller value of Standard Error of Estimate (361631, 361631 and 278988)  and absence of autocorrelation (2.097). The Least Square Dummy variable (LSDV) estimators  performed better in terms of the amount variability in crop yields accounted for by the climatic and non-climatic factors in the model as it reports a higher  R² values (72 %, 30% and 18%) for LSDV, within and first difference estimators respectively. The intercept values of the seven crops used in the study were statistically significant. From the study, non-significant of the climatic variables suggests that the predicted increase in temperature and precipitation have virtually no effects on yields of the selected crops. For controlling for unobserved heterogeneity and omission variable bias, the study, shows that the fixed effects model is more appropriate than the pooled OLS. Since all the intercepts values of the seven crops were statistically significant, this shows that the impact of climatic factors and the other variables on crop yields are not significantly the same. The study therefore suggest  the need for crop specific mitigation or adaptation policies against country level or a national level policy as this may be ineffective.