Economic analysis of water harvesting technologies in Ethiopia

Rainfall shortage and variability constrain crop production of smallholder farmers in Ethiopia and climate change may even aggravate this problem. An attractive method to mitigate this is water harvesting. This thesis examines the economic aspects of water harvesting by exploring optimal water use and the impact of water harvesting using micro-econometric analyses of cross-section and panel data collected from Ethiopian farmers in 2005 and 2010.

In the first empirical chapter, the study estimates marginal values and elasticitiesof harvested water in the production of three vegetables to determine whether water allocation is economically optimal. The results are mixed, although the estimated marginal product values between onions and tomatoes show that farmers on average allocate water economically across these two crops.

The descriptive data show that the share of irrigated land is lower at larger farms. Because farm size may increase in the future, it is interesting to investigate what determines the share of irrigated land in relation to farm size. A random-effects tobit model is appropriate to estimate this relationship. The result shows that access to both credit and markets, farm-size, region, aridity, and plot distance to water source all affect the share. Encouraging water harvesting requires flexible and effective variables that will work also for larger farms.

Despite its weather-risk reducing advantage, the average disadoption rate of water harvesting technology between 2005 and 2010 was as high as 42%. To find out why farmers disadopt, binary choice models are estimated to investigate the factors that cause disadoption. Based on the estimation results, it is concluded that increasing availability of plastic sheets and labour-saving equipment (water pumps), easier market and credit access, and the cultivation of perennials can reduce disadoption.

The last empirical chapter focuses on the relation between water harvesting and fertilizer use. Due to weather risk, farmers may limit the use of purchased fertilizer, thereby continuing to grow a high share of low-risk and low-yield crops. To establish whether harvested water encourages fertilizer use, two variants of random-effect models are estimated. The results strongly support the idea that water harvesting technology induces fertilizer use, indicating that water harvesting can increase fertilizer use- and hence crop yields- in Ethiopia.

The concluding chapter discusses the results against the background of the research objective: what are the economics of water harvesting at micro level?