Gradualism or Big Bang

One of the first national efforts to use a participatory approach in irrigation management was the Philippines, beginning in 1974. That was the year that the then president, Ferdinand Marcos, amended the charter of the National Irrigation Administration (NIA) to enhance its financial separation from the central treasury. NIA was charged with keeping its own accounts, and becoming financially autonomous over the long-term. The financial pressures to keep administrative costs low encouraged the NIA to shift management control of irrigation systems to the users.

While the concept was radical, the approach was gradual: Those systems already under local management would no longer become incorporated into public management following state-sponsored rehabilitation. And in the larger systems currently administered by NIA, any new construction works would be supervised by farmers, with some level of cost-sharing between farmers and government.

To enlist farmer support for taking on a greater, and more onerous management role, NIA employed a cadre of social organizers who were specially trained for this task. The process of organizing the farming community was slow but over the long term, was quite effective. The water user associations thus formed were able to negotiate successfully with the NIA for demand-driven improvements, and the same associations have functioned well in O&M and in collecting irrigation service fees which are shared by the association and the NIA.

The process of slowly organizing water user associations which are limited first to the tertiary levels of the system and then eventually federate at a higher level is an example of a gradual approach. This approach can be contrasted with the ìbig bangî approach taken by Mexico and Turkey where very large irrigation systems are transferred overnight to management by the users. Not only is the process far quicker, but it is also broader (larger areas) and deeper (more functions): the farmer association hires their own irrigation technicians to carry out O&M. Instead of joint management with the public irrigation agency, the relationship is one of a water wholesaler (the government) and a water retailer (the water user association). The government continues to regulate water use, but does not interfere in the routine management of the irrigation system.

Can farmers manage large scale irrigation systems? The answer from Mexico and Turkey is a clear, ìyesî. Legally constituted farmer associations have entered into contractual agreements with the state water agency specifying the mutual rights and obligations for providing water (the stateís obligation) and maintaining the infrastructure (the farmersí obligation). In both Mexico and Turkey a farmer management board oversees technical managers who actually operate the irrigation system and who are paid by the association. The fact that the engineers who operate the system are paid by the farmers who receive the water provides a completely different incentive structure than is the case when a government bureaucracy manages the water.

The basic vision of PIM, at least in my view of it, is this type of management control by the users. This approach does not mean that farmers have to work as managers in addition to farming; rather it means that the farmers control the managers, and that the managers are accountable to the farmers they serve.

It is interesting, and perhaps important, that this degree of user control is found only in countries that have adopted ìbig-bangî policies of rapid transfer. Is there a real correlation between the rapidity of transfer and the ìcompletenessî or degree of management transfer? It may be that under a gradual process of encouraging farmers to take on incremental management functions, and encouraging the public agencies to gradually relinquish management control, that there is too much room for maneuver. Perhaps the vested interests have too much opportunity to find new ways of holding on to the status quo. Revolutions, but their very nature, are both rapid and deep; evolution is slow. I would like to suggest that the process of irrigation management transfer has a direct impact on the results, and that in this sense, a rapid process may be more likely to have greater impact than a slow process.

Your comments and reactions are welcome, and will be printed in the next INPIM newsletter.

- David Groenfeldt, World Bank Economic Development Institute