Irrigation infrastructure and farm productivity in the Philippines: A stochastic Meta-Frontier analysis

• Published in: World development Vol. 135; p. 105073
• Main Authors: Bravo-Ureta, Boris E., Higgins, Daniel, Arslan, Aslihan
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 01.11.2020; Elsevier Science Publishers; Pergamon Press Inc
• Subjects: Agricultural production; Analysis; Beneficiaries; Bias; Crops; Development strategies; Farm income; Farmers; Farms; Heterogeneity; Households; Impact analysis; Impact evaluation; Infrastructure; Infrastructure (Economics); Irrigation; Livelihood; Philippines; Production; Productivity; Propensity; Rice; Rural areas; Rural development; Scarcity; Selectivity; Selectivity bias; Statistical analysis; Stochastic Production and Meta-Frontiers; Stochasticity; Technology; Water; Water scarcity; Philippines; Productivity; O13; Irrigation; Stochastic Production and Meta-Frontiers; Selectivity bias; Q12; Q15; Philippines; Impact evaluation; D24; Rice
• ISSN: 0305-750X; 1873-5991
• DOI: 10.1016/j.worlddev.2020.105073

•Improving irrigation infrastructure can have significant impact on output and yields while being pro-poor.•Improved small scale irrigation systems can shift the frontier, but farmers need complementary extension support.•Infrastructure and institutional support can be effective in improving water equity issues.•Additional attention is warranted to reach female farmers and to promote improved management overall. Irrigation is a lynchpin of rural development strategies and a key input to improving productivity and farm incomes, the key source of livelihood for the majority of the world's poor. Limited land and growing water scarcity mean that establishing systems to maximise the benefits from every drop is pivotal. In this paper, we analyse the impact of a canal irrigation project for smallholders in the Philippines, focusing on rice, one of the world's most water intensive crops. We contribute to two strands of literature by combining impact evaluation and efficiency analysis methods. Using a dataset for 714 treatment and 440 control farm parcels, we apply Propensity Score Matching and a selectivity-corrected Stochastic Production Frontier to handle biases from both observable and unobservable variables. We then analyse technical efficiency (TE) and frontier output using a shared Stochastic Meta-Frontier. We find that the project had a statistically significant impact on frontier output but not on TE, suggesting that improved irrigation technology increased beneficiaries' production potential, but it did not improve TE likely due to insufficient training and input access. Thus, beneficiaries were unable to take full advantage of their improved production potential, highlighting the need for suitable complementary support in future projects. Heterogeneity analysis reveals that the main beneficiaries were downstream parcels, smaller parcels, those located in the poorer project district, and farmers with lower education, all implying a pro-poor impact. Finally, we find that female-headed households benefitted less from the project, suggesting the need for additional support in future interventions.