Abstract

Investment in agricultural water management has been one key poverty reduction strategy in developing countries. In Ghana, irrigation development for livelihood support, which dates back to the 1960s manifested in the development of formal irrigation infrastructure, starting with the rural savannah and coastal regions. This study applied a propensity score matching (PSM) approach and regression analysis (ordinary least squares and switching regression) to ascertain that, pro-poor irrigation investment in the rural savannah region of Ghana is justified due to significant irrigation contribution to consumption expenditure per capita in farm households. The results also show differences in the impacts of irrigation access on household consumption expenditure per capita due to differences in the methodologies employed. The gain in household consumption expenditure per (GH¢) using the different methodologies are as follows: Propensity Score Matching approach (GH¢ 24.90, GH¢28.30), ordinary least squares regression (GH¢ 5.40), and switching regression (GH¢23.70). Thus, the range of estimates of irrigation’s impact on household consumption expenditure is positive (GH¢ 5.4 to GH¢ 28.3). The differences in the magnitude of these estimates are ascribed to the underlying assumptions and robustness of each of these methodologies employed in the study. 

Key words: Irrigation, consumption expenditure, propensity score matching, ordinary least squares regression, switching regression, Ghana.