Wednesday, August 19, 2015

Mitigating last-mile gaps in irrigation

Mihir Shah has a nice reminder in the Indian Express where he advocates to "push irrigation not dams". Pointing to large numbers of dams constructed with disproportionately low irrigation coverage realized, he argues in favor of a participatory and multi-disciplinary approach to the development and management of irrigation structures. 

This is not dissimilar to last-mile gaps elsewhere - schools where no learning happens, hospitals which do not cure, toilets which are not used, no-frills bank accounts which are not transacted, and so on. At a fundamental level, it is also a reflection of how weak state capability comes in the way of achievement of transactional outcomes. 

In case of irrigation, the incentives associated with major irrigation projects and minor irrigation works are grossly mis-aligned. In fact, it strikes as disturbing that irrigation sector stands out with contracting which is divorced from outcomes. Irrigation structures like large and small dams and water harvesting units, by themselves contribute little to irrigation, unless complemented with canals and field channels. 

Contractors though are more interested in the former. They find the single-location, regular construction work associated with dams far more easier and attractive than the right-of-way acquisition problems and other transaction costs that characterize widely-spread canals and channels. In case of larger projects, the dam and canals are given as separate contracts, and it is common place to find the dams and other large engineering structures in place without the canals. 

Similar incentives drive government stakeholders. Irrigation departments, most often interested only in contracting out works, too push projects, making unrealistic irrigation coverage estimations. In cases of projects done with assistance from Government of India, state governments routinely make over-optimistic irrigation coverage claims to get project approvals. The tortuous process of acquiring right-of-way for canals and field channels, which involves engaging and negotiating with local land owners and resultant transactional challenges, is severely constrained by weak state capability. All this, coupled with the problems of siltation and maintenance for canals and channels, makes dams the primary objective and irrigation a distant and secondary objective. 

The physical salience of the dam or water harvesting structure being dominant and its relative ease of site acquisition, compared to the long-drawn process of getting right-of-way clearance for long-winding canals and channels, makes everyone more interested in the dam instead of the canals. It is therefore no surprise that dams have been built with vast difference between the promised and actually delivered coverage. In fact, an audit of the original estimate of the irrigation potential and the realized coverage from all major and minor irrigation contracts is most likely to reveal a scandal as big as anything we have seen. 

While there are no easy answers to the problem, one possible strategy would be to package construction contracts as irrigation contracts rather than dam construction contracts. Instead of constructing a dam, the terms of reference in the bid should be clearly redefined as "creation (or stabilization) of 5000 Acres", with levels of water access. This would require much closer engagement among the contractor, irrigation officials, and the beneficiary farmers, and far more rigorous project preparation work. The project reports for each structure would have to evolve bottom-up, capturing local requirements and practical considerations, so as to ensure that outcomes remain at the center of the project.

The risks associated with not getting the coverage estimations right can be mitigated by appropriate safeguards with respect to upstream water availability. I am not aware of any state which have procured even minor irrigation contracts through irrigated land coverage tenders. 

No comments: