Vulnerability to multidimensional food insecurity: empirical evidence from wheat farmers in the Arsi Zone of Ethiopia

Although food security is multidimensional, most previous studies have measured households' vulnerability to food insecurity via unidimensional indices. To address this gap, this study assesses households' vulnerability to multidimensional food insecurity (VMFI) by employing the vulnerability-as-expected poverty technique for 422 randomly selected wheat-producing households from three districts of the Arsi Zone. Before households' VMFI was estimated by the three-stage feasible generalized least squares approach (FGLS), the multidimensional food security (MFS) index was developed by using principal component analysis (PCA). The MFS index was then calculated to be 0.4825 by the geometric aggregation technique to summarize the PCA results, and the descriptive summary indicated that 45.02% of the respondents were food secure. The three-stage FGLS revealed that 57.11% of the respondents were vulnerable to multidimensional food insecurity. Moreover, 33.41% of the respondents had stable food security, 45.5% were chronically food insecure, 9.48% experienced transitory food insecurity, and 11.61% were vulnerable to multidimensional food insecurity. The results of the probit model, which is employed to identify the determinants of VMFI, confirm that contact with agricultural extension agents, family size, and participation in off/nonfarm activities reduce VMFI, whereas the dependency ratio, credit utilization, shock experiences, and living in lowland areas increase it. Accordingly, it is important to adopt policies and strategies that enhance household interaction with agricultural extension agents and develop an efficient early warning mechanism to build a resilient food security system. To reduce household VMFI, it is also crucial to promote financial education and provide special support to drought-prone areas by developing drought-tolerant agricultural technologies and reducing agricultural rain dependency.