Abstract
In Bangladesh, HYV boro rice is produced using mainly groundwater irrigation in winter (dry) season. A market has evolved for buying and selling of irrigation services. One- fourth crop share has been the dominant contract payment system for groundwater irrigation in the past. Recently, more rational contract payment systems are emerging. This study examines groundwater irrigation markets in Bangladesh and aims to answer questions relating to the different contract payment systems for irrigation water, choices of contract payment systems, contract choices affecting rice production efficiency and factors determining water price in irrigation markets. Using a multi-stage random sampling technique, primary data were collected from 960 farm households spread over 96 villages located in 48 upazilas in five major boro rice growing regions. Descriptive and inferential statistics, multinominal probit and translog stochastic frontier regressions models were employed to achieve the main objectives. Fixed charge, crop share and fixed machine charge were the common contract payment systems, which covered 58%, 24% and 18% of the sample pump irrigation systems, respectively. In fixed machine charge, a buyer brings diesel to operate water seller's tubewell on payment of service charges per unit of land being irrigated. The major determinants of choosing the crop share payment were the years of experience in irrigated farming, interest rate of loan, access to electricity, flood and drought problems, and lowlands endowment. The same factors also significantly influenced the choice of fixed charge payment. Water buyers preferred fixed machine charge and fixed charge payment than crop share. The technical efficiency in boro rice production varied among the payment contract choices. The technical efficiency was the lowest in crop share, which indicates the potential to increase rice production and reduce cost by shifting away from crop share payment system. The statistically significant determinants of technical efficiency were soil type, education level of household head, kinship relationship between buyers and sellers, and asset endowment of a farmer. As expected, the sandy loam soil had significant negative effect, while kinship relationship and education level had significant positive effect on technical efficiency. The large effect of education on efficiency was striking. As compared to crop share, a higher technical efficiency was observed in fixed charge system. The technical efficiency was the highest in the fixed machine charge payment system. Across farm categories, technical efficiencies were relatively higher among small and medium farmers, compared to marginal and large farmers. The major determinants of water price in groundwater markets were irrigation payment system, density of tubewells, the number of water buyers, amount of remittance per household, lowland area, and fuel price. The imputed water price was the highest in crop share and this was the lowest in fixed payment system. Rice yields were lower, while irrigation and labour costs were higher in crop share as compared to the other two systems. The fixed charge system is less viable because sellers may refuse water supply at peak season, buyers may ask more water than needed, higher price of diesel, erratic electricity supply, unfriendly relation between buyers and sellers, and limited cash flow of water sellers to buy fuel. Through market-based mechanism, fixed machine charge system minimizes irrigation water use and thereby benefits water buyers, sellers, and the society at large.