Abstract
System GMM techniques have been found to be significantly superior to short panels in situations where T is short and N is big in prior studies. Although the system-GMM coefficients are short-run coefficients, while access to land is critical to long-term food security too, However, the elasticity or responsiveness of food security as a result of access to land and other predictors in the long run is not well understood in Ethiopia. We used the system GMM method as an estimation strategy for the study. This study addresses this issue by computing the long-run GMM coefficients from its short-run GMM estimates. In addition, year-dummy control for time variations of the dependent variable across the panels is also estimated and used. Since the dependent variable is expressed in natural logarithms and the explanatory variable is a dummy (1/0) measure, According to the findings, farm size, TLU, heads completed primary education, adult equivalence, one-year lag of annual food consumption per adult equivalence, number of household parcels, households’ distance to the main road (Kms), heads age, households’ distance from the market center (Kms), and female-headed households are both short-run and long-run complements of food security. Results also showed that household farm size (acre) has a larger positive effect on food security in the short run (0.179%) than in the long run (0.076%). As far as we are aware, no research has been done in Ethiopia or anywhere else in Africa that uses System GMM to compute the long-run GMM coefficients. Thus, in addition to answering the issues being addressed, the work also makes methodological contributions.
Keywords