This was created in partnership with Environment for Development .
This study looked into tree-growing behavior of rural households in Ethiopia. With data collected at household and parcel levels from the four major regions of Ethiopia, we analyzed the decision to grow trees and the number of trees grown, using such econometric strategies as a zero-inflated negative binomial model, Heckman’s two-step procedure, and panel data techniques. Our findings show the importance of analysis at the parcel level in addition to the more common household-level. Moreover, the empirical analysis indicates that the determinants of the decision to grow trees are not necessarily the same as those involved in deciding the number of trees grown. Land certification, as an indicator of tenure security, increases the likelihood that households will grow trees, but is not a significant determinant of the number of trees grown. Other variables, such as risk aversion, land size, adult male labor, and education of household head, also influence the number of trees grown. In general, the results suggest the need to use education and/or awareness of the role and importance of trees and point out the importance of household endowments and behavior, such as land, labor, and risk aversion, for tree growing. Finally, we observed that, while tree planting is practiced in all four regions covered,there are variations across regions.