The Impact of improved maize varieties on farm productivity and wellbeing: evidence from the East Hararghe Zone of Ethiopia

Abstract

The aim of this study is to measure the impact of improved maize varieties on farm productivity and smallholders' wellbeing using data collected from the East Hararghe Zone of Ethiopia. We combined propensity score matching method with endogenous switching regression to estimate the impact on the welfare of farmers and we applied the stochastic frontier corrected for sample selection to measure the impact on farm productivity. The results show that adoption of improved maize varieties leads to significant gains in wellbeing and improves farm productivity. ARTICLE HISTORY

Impact of soil and water conservation practices on household vulnerability to food insecurity in eastern Ethiopia: endogenous switching regression and propensity score matching approach

“…However, most impact assessment approaches using non-experimental data (not randomly assigned) fail to capture observable and/or unobservable characteristics that affect adoption and outcome variables. For instance, instrumental variables capture only unobserved heterogeneity, but the assumption is that the parallel shift of outcome variables can be consider as a treatment effect (Ahmed et al 2017;Kabunga et al 2012;Shiferaw et al 2014). In contrast, using regression models to analyze the impact of a given technology using pooled samples of users and non-users might be inappropriate because it gives the similar effect on both groups (Ahmed et al 2017;Kassie et al, 2010;Kassie et al 2011b).…”

“…For instance, instrumental variables capture only unobserved heterogeneity, but the assumption is that the parallel shift of outcome variables can be consider as a treatment effect (Ahmed et al 2017;Kabunga et al 2012;Shiferaw et al 2014). In contrast, using regression models to analyze the impact of a given technology using pooled samples of users and non-users might be inappropriate because it gives the similar effect on both groups (Ahmed et al 2017;Kassie et al, 2010;Kassie et al 2011b). A methodological approach that overcomes the aforementioned limitations is endogenous switching regression (ESR), which is the most frequently used common method to analyze the impact of a given technology (Abdulai andHuffman 2014, 2014;Ahmed et al 2017;Asfaw et al 2012;Di Falco et al 2011;Jaleta et al 2018;Kabunga et al 2012;Kassie et al 2011a;Shiferaw et al 2014).…”

“…In contrast, using regression models to analyze the impact of a given technology using pooled samples of users and non-users might be inappropriate because it gives the similar effect on both groups (Ahmed et al 2017;Kassie et al, 2010;Kassie et al 2011b). A methodological approach that overcomes the aforementioned limitations is endogenous switching regression (ESR), which is the most frequently used common method to analyze the impact of a given technology (Abdulai andHuffman 2014, 2014;Ahmed et al 2017;Asfaw et al 2012;Di Falco et al 2011;Jaleta et al 2018;Kabunga et al 2012;Kassie et al 2011a;Shiferaw et al 2014). In this paper, we employ parametric ESR with non-parametric PSM technique to reduce the selection bias and assure consistent results by capturing both the observed and unobserved heterogeneity that influence the outcome variable as well as the adoption decision.…”

“…Then the variable included in the second stage questions capture both absorbed and unabsorbed heterogeneity in estimation procedure ESR (Jaleta et al 2018). To address the heteroskedasticity arising from the generated regressors, the standard errors in questions (2a) and (2b) are bootstrapped (Ahmed et al 2017;Jaleta et al 2018;Shiferaw et al 2014) Based on the above context, comparing real and counterfactual scenarios of expected values of the outcomes of adopters, the average treatment effect on the treated (ATT) is obtained. Similarly average treatment effect on the untreated (ATU) also can be calculated by comparing the expected values of the outcomes of non-adopter in real and counterfactual scenarios (Khonje et al 2015).…”

“…According to Khonje et al (2015), Shiferaw et al (2014) and Ahmed et al 2017, ESR models have a very strong exclusion restriction and the falsification test may not be adequate to confirm identification. Thus, results may be sensitive to selection of instrumental variables.…”