WATER theft has increased on channels taken over by local farmer organisations compared with those managed bureaucratically, according to an analysis of the governance reform in the country`s vast Indus irrigation system.

The theft was more prevalent along channels where large land holdings are situated upstream, the World Bank research study finds. It describes the devolution of irrigation management in the Indus basin as a `felicitous combination`.

The study shows a unique data set on both landownership and influential positions held by every water user in each of the 3,923 watercourses (outlets) along 448 channels throughout Punjab. In addition, it also gives information on about 220,000 individual farmers.

In 2016, the World Bank commissioned a survey of 4,294 outlets on 470 irrigation channels distributed across 24 of the 49 administrative divisions of the Punjab irrigation department.

The study compares changes in water discharge along channels whose management was taken over by locally elected farmer organisations to changes that occurred in channels that remained centrally managed. Water theft increased by more in the former case, leading to a large redistribution of wealth.

Decentralisation also increased water theft more along channels with a greater preponderance of large landowners at the head, which suggests that investment in de facto political power can sometimes more than offset changes in de jure political power brought about by institutional reform.

The authors of the study say that `while our evidence is not favourable to the centralisation effort in the Indus basin inasmuch as it did not deliver on its promise of a more equitable and efficient distribution of canal water, it would be premature to throw out the reform baby with the bathwater`.

Successful decentralisation will likely involve directly addressing power asymmetries along the irrigation system, such as by giving tail-enders exclusive control over farmer organisations.

The World Bank research took the advantage of a partial governance reform in the world`s largest canal irrigation system the Indus basin watershed. During the last decade, in an effort encouraged by the World Bank, the management of several large sub-systems in Punjab was transferred from the provincial irrigation department to farmer organisations organised at the channel level.

Assessing how this shift from bureaucratic to local control affected rent-seeking in the form of water theft along a channel, the study notes that in such quota-based systems, users have a strong temptation to bribe local officials to look the other way as they use various means to illicitly enhance their water entitlement. Invariably, such water theft benefits farmers at the head of the channel, where water is first to arrive, at the expense of farmers at the tail.

While decentralisation strips authority of unelected irrigation department bureaucrats, farmer organisations may also be subject to capture by these upstream elites and, hence, may sanction as much (if not more) water theft than the irrigation department functionaries they replace.

Given the locational asymmetry, corruption and theft are concentrated at the head of a channel. However, theft induces rent-seeking by coalitions of gainers (farmers at the head) and losers (farmers at the tail), each with varying degrees of political influence.

Under irrigation department control, lobbying effort is directed `over the head` of the local official involved in the corruption whereas, under decentralised control, it is directed towards farmer organisations.

The centrepiece of the analysis of Pakistan`s irrigation reform is an administrative database maintained by the Punjab irrigation department and consisting of readingstaken from water discharge gauges installed at the head and tail of each channel of the entire system.

The data arguably provides an objective measure of water theft along a channel. Moreover, water discharge data is available over the years from 2006 to 2014, a period encompassing significant devolution of irrigation management to farmer organisations.

In a comparison research paper, the authors study the allocation of canal water in the presence of rent-seeking farmers and corruptible irrigation officials with career concerns. Using data from several hundred distributaries in Punjab that were not subject to irrigation reform, the study found that under bureaucratic control, the extent of water theft is substantially affected by the distribution of political power along a channel, where political influence is relatively concentrated at the head of a channel, water allocations are more favourable towards the head as reflected in both the canal discharge and the land value differential between the head and tail.

In empirical results, the authors find strong evidence of an economically important decrease in the relative allocation of water to the tail of a channel once a farmer organisation becomes operational.

The Indus basin irrigation system, which accounts for 80 per cent of Pakistan`s agricultural production, lies mostly in Punjab, wherein it encompasses 37,000 kilometres of canals and irrigates about 8.5 million hectares.

From Indus, Jhelum, Chenab, Ravi and Sutlej riversissue a dense network of main canals, branch canals, distributaries, minors and sub-minors, ultimately feeding 58,000 individual watercourses in Punjab alone.

Formally launched with the passage of the Punjab Irrigation and Drainage Authority Act of 1997 by the provincial assembly, irrigation management reform in the Indus basin, and specifically the devolution to water user`s associations, was strongly encouraged by the World Bank.

Administratively, Punjab`s irrigation system is divided into 17 circles. As part of the reform, Area Water Boards were established at the circle level to promote the formation of farmer organisations covering every water channel within the circle. The famer organisations were themselves tasked with the operations and management of distributaries and their off-taking channels.

In particular, a farmer organisation is responsible for monitoring the rotational system to ensure equitable allocation along the distributary, for mediating and reporting water-related disputes among its irrigators, and for collecting water taxes to fund canal operations and maintenance.

Starting from the universe of 2,902 irrigation channels in Punjab, dropping cases that either had zero discharge at the head throughout the 2006-14 period or in which the overseeing farmer organisation included a larger branch canal (three organisations in all), leaves 2,860 channels.

Of these, 1,007 are covered by 394 farmer organisations and 1,853 are not covered by them. •