Last fortnight, the Supreme Court issued a diktat to the government to implement the scheme to interlink rivers. The directions are straightforward. The government shall set up a high level committee of ministers and other representatives on interlinking of rivers; the committee shall meet "at least once in two months"; in the absence of any member the meeting shall not be adjourned; the committee shall submit a biannual report on actions to the Union Cabinet, "which shall take final and appropriate decisions in the interest of the country as expeditiously as possible and preferably within 30 days from the matter being placed before it for consideration."

Without getting into the obvious matter of judicial overreach, let us take a careful look at what interlinking is all about, and what the decision will imply. The fact is that transfer of water from one river basin to another is not, per se, either a novel or an untested idea. Every irrigation project involves such transfer at some scale. The question is what this particular idea of linking rivers implies.

The term river linking has come from the idea floated by irrigation engineer K L Rao way back in 1972. He proposed the construction of a grandiose Ganga-Cauvery Canal, which would divert floodwaters of the Ganga near Patna for about 150 days in a year to river Cauvery some 2640 km away in the south. This idea captured imagination, as it seemed to state such a delicious proposal: take excess water from the Ganga to the water-deficit and stressed areas of Tamilnadu.

A pilot, Captain Dinshaw J Dastur, suggested a variation: construction of garland canals, one for the Himalayan watershed and the other for the Western Ghats. This idea was also appealing, simple and essential. Long-distance irrigation projects then spawned a huge water bureaucracy. In 1982, the National Water Development Agency was set up to study and implement the project to first link peninsular rivers and then Himalayan rivers. Its objective is based on the same simple concept: there are floods in some parts, droughts in the other, so if we link the rivers, we all will be happy.

For obvious reasons the NWDA's proposals were, government after government, studied, considered and buried. But not forever.

 •  President Kalam, please listen
 •  Cart before the horse
 •  Full section: Interlinking

But for equally obvious reasons the agency's proposals were, government after government, studied, considered and buried. But not forever. In early 2000, the Supreme Court and government got back into the game. The court ordered the government to speed up implementation of the project and set the deadline of 2016 for its completion. The National Democratic Alliance then in power quickly announced the setting up of a task force for linking rivers. It was to complete some 30 river links in two years, adding some 1000 km of canals. This task was a non-starter.

The next government came to power and while the concept appealed, better sense prevailed. Interlinking was found technically unfeasible and costly. But the water bureaucracy did not give up. In 2008, the National Council of Applied Economic Research produced a study volume, in which it explained in simplistic terms that the project would cost Rs.444,331.2 crores at 2003-2004 prices. But this investment would lead to rich dividends in terms of increased household income and prosperity for all, it stated. The report would have gone unnoted but for the Supreme Court, which has bought this line and ordered the government to obey or face contempt.

The question still is: what does this project imply, given that a massive number of irrigation projects on the government's wish list remain incomplete? First, it implies the notion that there is huge surplus of water in river basins. This assumption is flawed. Most river basins today are overextended in usage, and in most regions tension is growing between old rural users of surface water and new industrial and urban users. The Mahanadi basin, which would be linked to the Godavari is a classic example of this error. As Down To Earth explained in a recent investigation, there is little unallocated water in the basin (see "In deep water", Down To Earth, February 16-29, 2012).

The second assumption that floodwaters can be channelised is equally erroneous. The fact is when one river is in spate so is the next river, and transferring water would require huge storage facilities. Construction of large reservoirs has massive environmental impacts not considered in the scheme. Many irrigation projects are stalled on this count. More importantly, the government's track record in resettling people displaced by such projects is abysmal.

The third assumption is that India will gain from investment in irrigation projects, and this is indeed true. But it is equally true that the current challenge is to ensure that the projects already built and commissioned, are kept operational. The 12th Five Year Plan working group clearly states that priority is bridging the growing gap between the irrigation potential created and utilised.

The idea of interlinking rivers is appealing because it is so grand. But this is also the reason it is nothing more than a distraction that will take away precious time and money from the business at hand. The task is to provide clean water to all and to use the resource with efficiency. This agenda needs our attention. Indeed our obsession.