Shivakoti, Ganesh P., Douglas L. Vermillion, Wai-fung Lam, Elinor Ostrom, Ujjwal Pradhan and Rober Yoder (Eds.) Asian Irrigation in Transition: Responding to Challenges.

Shivakoti, Ganesh P., Douglas L. Vermillion, Wai-fung Lam, Elinor Ostrom, Ujjwal Pradhan and Rober Yoder (Eds.) Asian Irrigation in Transition: Responding to Challenges. New Delhi: Sage, 2005, 528 pp., hardback.

This book is a selection of papers from a workshop on "Asian Irrigation in Transition" that was held at the Asian Institute of Technology in Bangkok in 2002. The other institutions involved were the International Water Management Institute, the Workshop in Political Theory and Policy Analysis at Indiana University, and the Department of Political Science at the University of Hong Kong. It has twenty chapters that include case studies from Pakistan, India, Nepal, Sri Lanka, Thailand, China, Taiwan, Vietnam, Indonesia, and the Philippines.

The second chapter by Randolph Barker and Francois Molle provides a historical overview of Asian irrigation development. Dam construction grew rapidly in the three decades after 1950 and peaked around 1980. The reasons included a decline in world cereal prices together with the exhaustion of the most favorable sites. Moreover, the green revolution aided by the expansion of irrigation had ended the threat of famine even in South Asia, the poorest part of the continent. In the faster-growing economies of East and Southeast Asia, agriculture was no longer a large share of the economy.

While canal irrigation growth has largely ended, this is not true of groundwater irrigation, which has continued to grow very rapidly till the end of the century. Barker and Molle do not inform us whether this too has peaked. Owing to the common pool problem, there has been massive overinvestment in pumps. Water scarcity has grown, albeit at very different rates in different regions. The absence of new sources of supply and the projected increases in demand from the growth of population and industry imply that water will have to be allocated better.

Barker and Molle state that the conventional wisdom among policy makers and academics that most Asian irrigation systems are poorly managed may be misleading; efficiency may be higher than what is assumed once return flows are taken into account. However, the remaining chapters mostly take it for granted that poor irrigation management means that there are large potential gains to be made by improving management and that these can only be achieved by means of more meaningful user (read "farmer") participation.

There is no doubt that some Asian systems, especially in the poorer countries of South Asia, are badly managed and that irrigation bureaucracies engage in rent seeking. However, as some chapters in this volume point out, the adverse consequences of badly timed and uncertain water allocation may be limited by the fact that canal water recharges groundwater. Farmers use the aquifer as an on-site storage facility and can time water applications as needed. It is true that this comes at the expense of pumping. On the other hand, there may be considerable expense involved in inducing farmer participation and monitoring in a reformed, less centralized, allocation system.

Most chapters deal with institutional reform, usually understood as reducing the role of the irrigation bureaucracies and increasing that of farmer organizations, known as Water User Associations (WUAs). The bottom line that emerges from this book is that it is unclear whether anything much has been achieved. Reforms have mostly been directed at greater farmer participation in management at the lower levels of distribution in order to reduce costs for the government agency, leaving the upper levels firmly in bureaucratic hands. And even where they have gone further, it is not clear whether productivity has gone up. There seems to be only one exception to this dismal pattern, and that is the case of the much-cited Gal Oya system in Sri Lanka discussed by Norman Uphoff (chapter 3). We do not learn, however, why it succeeded while all other reforms in South Asia have apparently failed.

Reading this book was a thought-provoking but unsatisfying experience. Several questions came to mind. First, just how inefficient are the bureaucratically managed systems that are undergoing reform experiments? The chapters in this book do not quantify this seriously. They do not even cite references to the literature about such quantification. One suspects that the underlying reason for this is that poorly managed systems do not collect the basic data on the timing and quantum of water flows and crop productivity that are essential to such evaluations.

The second question that follows naturally from the first is: How much of a productivity gain can be expected from such reforms? Without an answer to the first question one could hardly expect answers to the second, and sure enough, we mostly do not get them. We are left to conclude that, since the dog did not bark (except in Gal Oya), nothing very much has happened as yet.

The third question arises naturally from the fact that one of the major motivations for governments to reform irrigation systems is budgetary: To reduce the level of government subsidies. Is the efficiency gain from a reform likely to be large enough to cover the loss of the subsidy that farmers receive? If not, then why expect farmers to incur the costs of the collective action necessary to take over governance in addition to the operating costs of the system? This question is not tackled squarely by any of the authors or even raised by the editors even though it is surely central to the reform process.

With this in mind, I find myself somewhat skeptical of Vermillion's (chapter 16) characterization of the Andhra Pradesh reform that was initiated in 1997 as "empowerment with accountability." The reform was lubricated by a loan from the World Bank. Water charges were more than tripled from Rs 60/acre to Rs 200/acre and the recovery rate went up from 54% of target revenues at the old rate to 70% of target revenues at the high rate. By the second year after the reform, the irrigated area in the state had increased by 10-15% and yields by 20%. This was apparently not attributable to rainfall and the increase in irrigated area "occurred mainly in the tail of canal commands."

Does this improvement hold up over the longer term? A survey of 300 farmers in six WUAs conducted in 2003 gives the following picture. Average acres irrigated per farmer in the tail of canal commands jumped from 2.61 in 1997-98, the year preceding reform to 3.54 the following year, but then stayed between 2.69 and 2.79 for the next three years before crashing to 0.57 in 2002-03 as a result of drought. Average acres irrigated in the tail of tank (a small reservoir) commands, by contrast, fell from 1.43 to 1.12 between 1997-98 and 1998-99, remained between 0.99 and 1.05 for the next three years, before falling to 0.52 in the 2002-03 drought year. So, these survey data, albeit from a small sample of WUAs, do not support the idea of any consistent improvement, although in the reform year, there was a significant difference between canal and tank performance, possibly due to works carried out in canals using loan funds. They also cast doubt on the state-level numbers for Andhra Pradesh, whose source is presumably official.

In conclusion, one is left with the feeling that much more data on the technical parameters of the system are needed before meaningful evaluations can be made. Social scientists attempting to study irrigation are not going to make much headway if they stick to "soft" issues. This issue of missing information may also cripple bottom-up reforms based on farmer participation. As long as water control at higher levels remains opaque, farmers may not be willing to make the contributions necessary to induce good governance.

E. Somanathan

Indian Statistical Institute, New Delhi