Resilience Index Measurement and Analysis (RIMA) Survey in Afghanistan, First Round, 2023

A two-stage cluster random sampling approach will be applied to draw the sample of households to be interviewed. The first stage selects an equal number of clusters per province. The selection of cluster will be done with Probability Proportional to Size. In the second stage a fixed number of households per cluster will be randomly selected who participated in wheat cultivation package, livestock protection, cash for work, unconditional cash transfer, poultry, small farm equipment, summer crop and home gardening interventions. The resilience will be measured by statistically testing the differences in means of key outcome variables between beneficiary and non-beneficiary (control) households. Control households will be randomly selected among those households that did not receive any assistance (neither from FAO nor from other organizations) in the project period. The sampling frame of control household will include only households that comply with the targeting criteria of FAO’s interventions, in order to ensure that beneficiary and control households (1) have similar socioeconomic conditions; (2) are located in the same areas. For beneficiary households, the sampling will be based on the complete lists of beneficiaries of the different projects analyzed. Using these lists, a random sample of households will be drawn. For control households, a long-list of comparably vulnerable households in the pre-selected sample provinces would be obtained, based on the eligibility criteria of the three projects. A random sample of households will be drawn from these lists, and the process will be documented. Household will be surveyed in the 34 provinces. 30 percent oversampling is considered because of the successive rounds of surveys to tracking the same households to ensure offset of attrition rate. The total household population target which will be assisted by each intervention would be used as a weight.