Ethiopia Pilot Baseline Survey
I am Sana Khan and I work at Innovations for Poverty Action (IPA) w
here I’m responsible for coordinating the randomized impact assessment of the Ethiopia Graduation Pilot. We have now completed and analyzed a baseline survey administered to all potential program participants. It has enabled us to assemble a clear picture of the people the program sets out to help.
The baseline survey actually consisted of two different types of surveys per household: a household survey administered to the person most knowledgeable about how money is spent in the household and two adults surveys—one administered to the person targeted by the project and a second to any other adult in the household. After completion of survey work, the individuals were randomly assigned into treatment and control groups through publicly held lotteries in each community. This randomization process ensures that the difference in outcomes between the two groups can be attributed directly to the program. For various reasons, 75 households were dropped from the research and we now have an overall sample size of 925 households – 458 in the treatment group and 467 in the control group.
The preliminary analysis from the baseline survey allows us to paint a picture of our overall sample:
• The average household size for this group is 5 members with 2.2 adults and 2.8 children.
• The average age of the head of the household is 38 years.
• The majority of the household heads in the sample are male (60 percent) and less than half are literate (40 percent).
• The housing and living conditions data indicates that each house has an average of two rooms.
• Access to protected sources of drinking water such as wells, hand pumps, and taps is rather high (62 percent), but still at least 14 percent of the people use unprotected sources like rivers, ponds, canals and lakes.
• Similarly, access to dry or pit latrines is very high (86 percent), but we are unsure about actual usage.
• The two main sources of energy for cooking include firewood & chips and dung cake. A smaller but still substantial percent (19 percent) of the sample also uses charcoal. Finally, kerosene lanterns serve as the main source of energy for lighting.
The preliminary analysis also shows that the two main sources of income for our sample are the government safety net/food for work program and self-employed agricultural activities (82 and 80 percent respectively).
The households report a variety of animal/livestock holdings. The most common animals include: Chicken (56 percent), Cows (49 percent) and Oxen (47 percent). A lower, but still significant, portion of the sample reports owning sheep, bee-hives and goats.
The primary assets owned by the households include farm tools (80 percent), donkey/bull cart (34 percent), radio (33 percent) and jewelry/watch/clock (18 percent). Both the livestock and the primary household asset ownership reflect the agrarian nature of the society in which our sample resides.
Finally, as in Honduras and Peru, our initial results also indicate that formal financial services among eligible households are limited. Only 1.6 percent of the household reported having a bank or savings account, and only 7.4 percent of the sample had one or more outstanding loan at the time of the survey.
