Water level and irrigation 2012=

A silent revolution? Agricultural groundwater use in Ferghana Valley


And now, field notes from Central Asia. This is based on fieldwork in Ferghana Province of Uzbekistan in August 2010. I wanted to understand various aspects of agricultural groundwater use. Conventional wisdom is that private groundwater extraction is prohibitively expensive, especially when compared to free canal water and hence private investment in groundwater would be conspicuous by its absence. But my brief fieldwork showed that a lot of investments were indeed taking place in the private domain. Excerpts from my field notes:

“We found three types of private groundwater extraction mechanisms on farmers’ fields. Of these, open shallow dug wells (Type I) and shallow tubewells (Type II) tap shallow groundwater at depths of not more than 8 to 10 m. These operate with either electricity (preferred) or petrol pumps. Most of these pumps are made in China and are small, portable and as cheap as USD 25 to USD 30 per piece. Total investment for this kind of shallow dugwells and tubewells ranges between USD 100 and USD 400. The third technology is that of deep electric tubewells (Type III) which are usually 120-180 meters deep and tap aquifers at depths of 70 to 90 m. These are fitted with high capacity electric submersible pumps that yield 12 to 20 litres of water per second. Capital cost of investment is high and estimated at USD 20,000 per tubewell. There is no public finance for these tubewells and farmers invest their own savings.

Type I and Type II technologies are used by kitchen garden owners and their preferred crops are grapes, tomatoes and cucumber – the latter two are usually grown in green houses. We found shallow tubewells (Type II) to be more widespread than shallow dugwells (Type I). Shallow dugwells were mostly found in somewhat rocky terrain of Ozbekistan district of Ferghana Province, while shallow tubewells were found all throughout Oltariq and Ferghana districts of Ferghana Province. As per one estimate, almost 50% of kitchen garden owners in Oltariq district have shallow tubewells. As per another estimate, around 25% of all agricultural land in Ferghana Province is under kitchen garden. While we do not yet have a robust estimate of the number of shallow wells and tubewells in the region, our preliminary field visits suggests that they are more common than generally thought.

These groundwater extraction mechanisms are essentially very shallow structures, use small capacity pumps (less than 1 l/s discharge) and do not need a separate electricity connection – they are always connected to the household electricity grid and monthly electricity bills averages between USD 7.00 to USD 10.00. Hours of pumping varies from less than 50 hours to 250-300 hours in a year. Grape is the preferred crop and kitchen garden size varies from 0.10 ha to 0.25 ha. While these shallow wells and tubewells do not account for the bulk of pumped volume of groundwater, they do have a large impact in terms of livelihoods benefits. The most oft quoted reason for farmers’ adoption of shallow wells and tubewells is that it gives them assured and reliable water supply, which canal water does not always give.

Type III technology (electric deep tubewells) is almost always owned by orchard farmers with land holding varying from 4 ha to upto 40 ha, or more. These are almost exact replicas of government deep tubewells and may cost anything between USD 15,000 to USD 25,000 for installation. These are fitted with large capacity pumps (20 KV to 45 KV) and receive electricity bills of upto USD 500 or more per month. They operate for 2000 to 4000 hours in a year and serve anything from 4 ha to 70 ha of land. Crops grown are peaches, apricots, grapes, apples and flowers. Farmers with land under cotton and wheat seldom invest in these deep tubewells. These deep tubewells are more widespread than generally believed. In one village called Eske Arab in Oltariq district of Ferghana we found 25 farmers who owned deep tubewells. A government official had earlier told us that there would be a maximum of 100 private tubewells in whole of Ferghana Province. However, given that there were 25 reported from just one village in one district of Ferghana province, our guess is that these numbers would far exceed 100.

Then there are government irrigation wells and drainage wells. These provide irrigation to a substantial number of farmers and kitchen garden owners. There are some 2500 of these in Ferghana Province and is an important source of irrigation. These are operated by government employed tubewell operators cum electricians and farmers get water for free. These are mostly used to grow cotton and wheat. We talked to a few farmers who depend on these public tubewells for irrigation and they seemed content with the quality of service received. The condition of most of these public tubewells was satisfactory.

To sum up, shallow groundwater extraction mechanisms are adopted by kitchen garden owners, while deep tubewells are adopted by farmers who grow orchard crops (and not cotton and wheat) and farmers who grow cotton and wheat depend on government owned deep tubewells. All these categories of farmers and farms also draw water from canal, so most of groundwater use in this region, is in reality, conjunctive use. This short fieldwork covering 3 districts in Ferghana Province shows that private investment in groundwater may be more widespread than is generally known. This is not surprising and has indeed been the case in most countries where groundwater use has peaked over the years, but has largely remained outside the ambit of public knowledge and discourse. This is precisely why Llamas et al. (2008) has termed the slow but sure ascent of groundwater use as a “silent revolution”.

Regulatory Commission or Captured Regulation?


During last month’s blackout in India, my husband and I were happily vacationing in the Himalayas. Of course, there was no electricity, but then, we weren’t expecting any either up there in the pristine hills. So we were oblivious of all the drama surrounding grid failure down in the plains.  On our return, our tax consultant told us that last date of tax filing has been extended by a month, thanks to a massive grid failure on the 31st July when no one could file their returns.  What an unintended happy consequence of a mess that is India’s power sector!

It was precisely to manage this mess that Electricity Act of 2003 stipulated that independent electricity regulatory authority be set up in every state. And most states complied. Just after the recent power crisis, I noticed that many newspapers were blaming ineffective regulation by these very regulatory authorities as one of the reasons that pushed our power systems on the brink on those fateful nights in late July. And, I believe, rightly so.

As a part of a project on energy-irrigation nexus, we have been looking at the role that these electricity regulatory authorities have played in streamlining methods for calculation of agricultural electricity supply. This is one of their many roles. Agricultural electricity is an Achilles heel of our power sector – not so much because electricity is free or subsidized, but because meters were removed from tubewells decades ago with the result that no one had a clue about agricultural electricity consumption. Then, as a part of the sector reforms program in the 1990s, former state electricity boards were unbundled and asked to work on commercial profit making principles. They were also told that should they provide free or subsidized electricity to the farmers at the behest of the state government, they would be compensated for the same by the government. Now, imagine a situation where Electricity Company does not know how much electricity goes to agriculture and on top of that government commits to compensate them for all the electricity they provide to the farmers. What will they do? Indeed, what will anyone do under the circumstance? They will inflate their estimate of electricity supplied to farmers in order to claim larger subsidy from the government. And this is where the regulatory commissions come in. They are supposed to provide checks and balances and ensure that the electricity utilities do not claim inflated amounts of subsidy in the name of the farmers.

And how well are the electricity regulatory authorities functioning in this regard? We found that Punjab SERC has taken its mandate more seriously than most. They have over the last 10 years almost forced the state electricity utility to do a better job of energy accounting through fixing of 10% sample meters on tubewells, third party audits, commissioning studies for estimation of correct estimation methodologies etc. Karnataka SERC seems to have floundered, with the result that agricultural energy consumption estimation methodologies are very problematic. And some states like Madhya Pradesh does not even have proper records available online for us to understand its current state of affairs.

Why is there such variation in performance of regulatory commissions when all of them have the same mandate? I am told by many experts that this is a classic case of regulatory capture. Most of these commissions are headed by retired Power Secretaries. To me the conflict of interest is so glaringly clear that I am surprised that no one thought about barring ex-bureaucrats from assuming such regulatory roles immediately after retirement. Looks like the recently launched and much discussed Shunglu Commission Report has done precisely this. To quote:

“It is also recommended that an individual having worked in any capacity with the state government during immediately preceding five years should not be eligible for appointment as a Regulator in that state. Similarly, the Regulator should not take up further employment with the concerned state government on relinquishing office.” (Chapter 4:105).

Will this recommendation be accepted? I doubt it. After all, when has the ruling elite (and our bureaucrats are the elites of the elite) agreed to reduce their own power?

Making groundwater public property is a bad public policy choice


Today’s Times of India has published a news article on its front page, lauding a government proposal to make groundwater public property and engaging village communities in its management. The link is here.While this sounds reasonable in theory, there are several reasons why this is a poorly thought out idea.

First, it makes the assumption that once in public domain, groundwater will be better managed in the greater public interest. But where is the evidence for this? The newspaper, in another related article, said that this is being done in several villages in Andhra Pradesh.  I believe, they are referring to Andhra  Pradesh Farmer Managed Groundwater Systems Project (or APFMGS in short). While APFMGS was indeed successful in involving farmers in managing groundwater through proper water accounting and demand management, but  our recent field visits show that this success was  limited to the duration of the project and has fizzled out since then. Not surprising , this is often the fate of most projects, especially projects aimed at ‘demand-management’ as opposed to ‘supply-augmentation’.

Second, this proposal also says that village panchayats should be involved in groundwater management. For those who think that village panchayats are capable of doing this, or will do a fair job of it, must indeed be living in Gandhian utopia of Gram Swaraj  than Ambedkar’s reality of villages as center of oppression against the poor and the lower castes. I ask again, where is the evidence that involving village panchayats in management of such a precious resource — a resource on which lives  of millions of farmers depend, will be any better than the status quo? Or that such one-off initiatives can be sustained? Indeed, if anything, there is a lot of evidence to show that performance of Panchayati Raj Institutions has been dismal all throughout out India.

Third, there are other countries in the world like Mexico and Spain which have made groundwater public property years ago and have since then tried to manage it through laws and regulations. Ours and others assessment of these initiatives clearly demonstrates the implementation challenges. And these are countries with five times our per capita GDP, less than 1/10th  our numbers of wells and tubewells and several times better implementation capabilities. Here are links to papers from Spain and Mexico  highlighting the challenges and limited successes so far.

Fourth, several states in India have tried  implementing groundwater laws. This has almost always led to corruption and rent seeking. Our work in West Bengalshows that officials in charge of giving permits to farmers acted arbitrarily  and were often accused of seeking bribes. Ramamohan’s work in Andhra Pradeshalso highlights similar problems.

Our contention is that as an idea, this is not bad. But given the implementation challenges, the net result of making groundwater a public property may not be any better than the current regime. And if more government control becomes a channel for bribes and corruption, then the very people that this move aims to protect, will be even worse off than they are now. So, if such ‘direct’ regulations do not work, what does? Global evidence of managing groundwater shows that indirect measures often work better than direct interventions. In the context of India, the most important lever for managing groundwater is through managing the energy-irrigation nexus. It could be  done either through rationing or pricing of electricity. Various states in India are indeed trying to do this and they would need support in their endeavor. Making groundwater a public property will hardly help — it is neither here, nor there!

Metering in West Bengal: Of economics and politics


In those good old days (that is good old days for the Bengalis!), there was a saying that what Bengal does today; India does tomorrow. But thanks to decades of bad governance, this is not true anymore. Indeed, there is hardly anything that India can learn from Bengal – except perhaps how to manage agricultural electricity supply. In my previous post, I had written about how in the 1970s and 80s, all states in India had decided to remove meters from irrigation tubewells. Most of these states, including West Bengal, then started charging farmers a fixed fee for electricity use. This policy made sense at that time because tubewells were few and far between and the cost of meter reading was higher than the revenue generated from it. Besides, an unintended benefit was the proliferation of informal groundwater markets which then became the main conduit through which poor and marginal farmers got access to irrigation.

However, unmetered electricity supply precipitated a crisis in the electricity sector. By early 2000s it was widely recognized that agricultural tubewells – which now stood close to 10-12 million, needed to be metered for the sake of proper energy accounting. Energy accounting is the Achilles heel of India’s power sector. But by then, there were strong farmers lobbies and vested interest within the electricity departments which resisted any attempt at metering. The Electricity Act of 2003 made metering mandatory, but to no avail. West Bengal is the only state which has been able to meter agricultural tubewells. It started its metering program in 2007 and by now almost 90% of all tubewells in the state are metered. A 3ie funded project helped us evaluate the impact of metering of tubewells on groundwater use. Here is the link.

Why was West Bengal able to meter tubewells when other states failed? The answer, our research shows, lies in the domain of both economics and politics. The electricity utility in West Bengal, by continually raising flat tariff had made it so high, that by 2007, most farmers realized that electricity bill under a metered tariff will be much cheaper than the flat tariffs they were paying. In contrast, the other states had kept their flat tariffs so low (or even free) that farmers saw absolutely no benefit in switching over to metered tariff. And that West Bengal had all but 100,000 electric tubewells, as against 1.1 million in Punjab also helped. But then, why was Bengal able to raise flat tariff over the years, while other states could not? The answer lies in the politics of groundwater. West Bengal, unlike Gujarat or Punjab, never had a strong farmers lobby agitating for access to groundwater. This was because the way the Left Front government had co-opted the only farmers group in the state – the Krishak Sabha. The contrast is clear if we compare Gujarat’s Bharatiya Kisan Sabha with Bengal’s Krishak Sabha – something I did a few years ago. Here is the link.

The upshot is that West Bengal now has a reasonably good agricultural electricity governance regime – a regime that other states might as well emulate if they want to avert the crisis that they have been plunged into. Once again, Bengal sets the precedence. Yay!

Electricity use in agriculture: Deliberaterly messy data


Yesterday I wrote about inconsistent data on pumps and tubewells in India. That inconsistency, I think, stems from sheer apathy and lack of coordination among data collection agencies. It’s frustrating, but not quite as frustrating as dealing with agricultural electricity consumption data which is not only messy, but made messy on purpose. Indeed, two policies are responsible for the deliberate misreporting of electricity consumption in agriculture. One is the indirect subsidy payment regime in which the state government reimburses the electricity distribution company (DISCOM) an amount equivalent to the electricity that they supplied to the farmers. Second, the policy decision taken some 25-30 years ago to remove meters from all agricultural tubewells means that no one, not even the DISCOM, really knows the amount of agricultural electricity consumption. The electricity company has all the incentive in the world to claim a higher subsidy than they actually provided to the farmers and in the absence of metering, it is not easy to pin them down.

This is where the State Electricity Regulatory Commissions (SERCs) come in. Their role, among other things, is to examine the veracity of subsidy claims of the DISCOMs.  In Punjab, the utility and the SERC have been engaged in a game of cat and mouse since the early 2000s – the regulators constantly upbraid the utility for submitting exaggerated claims and reduce it by 5 to 15% every year, while the utility keeps trying to justify its claims. Punjab SERC pretty much forced the Punjab State Electricity Board (PSEB) to install meters on 10% of agricultural tubewells and now use these meter readings to approve any subsidy claims.  Karnataka SERC, on the other hand, does not seem to be able to convince its DISCOMs to do a better job of energy accounting. Here, most DISCOMs grossly exaggerate the number of electric pumps and therefore, end up claiming a much higher subsidy for providing electricity to farmers than they actually do. SERCs have not yet been able to get the DISCOMs to do a decent job of tubewell census – let alone sample metering.

While all SERCs have the same mandate as per the Electricity Act of 2003, some SERCs do a better job than others. Why this is so is an interesting research question. Regulatory capture by the state comes readily to my mind as an explanation. But this needs more systematic analysis — something I intend to do in the future.

How many wells and tubewells in India? No one really knows!


With all the brouhaha about unsustainable groundwater use in India, one would have expected that there are robust estimates of number of wells and tubewells. But, sadly, there aren’t!  While working on a project on energy-irrigation nexus, I started looking at the data on wells and tubewells more closely than I had ever done before. In doing so, I realized that there were vast discrepancies in the numbers depending on the source of data. I was intrigued. Stuti Rawat – a colleague of mine and I decided to investigate further. We looked at four sources of government data: Minor Irrigation (MI) Census, Agricultural Census, Input Survey and State Electricity Boards (SEB). We compared these numbers for four time periods (mid 1980s, 1990s, early 2000 and mid 2000) making sure that the enumeration year of the sources compared did not differ by more than a year.

Our findings took us by surprise. We found that numbers of wells, tubewells, and electric and diesel pumps varied between the sources and varied widely. While a ten to twenty percent difference in figures enumerated may be due to definitional differences and time lags, but differences that are as much as 40% or more, raise questions about the veracity of the data. For example, while Input Survey put the number of diesel pumps at 13.1 million in 2006-07, Agricultural Census showed only 4.5 million diesel pumps 2005-06 – almost a 3 times difference! Similarly, in 2000-01, Agricultural Census and Minor Irrigation Census had reported 6.3 million and 4.2 million diesel pumps respectively– a difference 50%. Estimates of electric pumps are no better. In 2000-01, Input Survey reported some 18 million electric pumps, while Agricultural Census and Minor Irrigation Census reported only 10 million electric pumps. Data from the SEBs were even more divergent and almost always 30-50% higher than estimates from other sources with the exception of few states. We found that wide divergence in data is the norm and convergence is the exception.

Since much of this data collection takes place at the state level, inconsistency in data is indicative of poor data collection machinery in that state. We tried to rank states in terms of data consistency and found that there are three states which were doing better than the rest. Any guesses? Well, these are Gujarat (but then of course!), Punjab and Haryana. And surprise, surprise, Bihar did so much better on the data consistency front in the 1980s and since then, it has deteriorated. And data in some states has always been inconsistent. West Bengal is pretty much on top of that list (why I am not surprised?), so are Uttar Pradesh, Karnataka and Rajasthan.

I think that such glaring data inconsistency in minor irrigation is symptomatic of overall decline in standards of statistical database in India – something that worries me deeply and ought to worry our policy makers even more.

Electricity and groundwater: if you can’t price it, ration it!


Several major Indian states- Punjab, Karnataka, Tamil Nadu and Andhra Pradesh all provide farmers with free electricity. Unfortunately, these also happen to be the same states that face threatening levels of groundwater depletion.  A popular opinion is that giving farmers’ free electricity leads to over-pumping and waste of both electricity and groundwater. Pricing electricity in a way that reflects the scarcity value of both is an oft-cited solution. This reasoning seems to make sense, not just to the economists but also to the general public.  I thought so too, until I started doing research on the energy-irrigation nexus in Punjab and Karnataka.

Surprisingly, what I found was that pricing is not the only mechanism that signals scarcity. Tight rationing of electricity seems to have pretty much the same effect. Take Punjab for instance; in the kharif (rainy) paddy season, electricity is supplied for only 6-8 hours/day. This is not enough for farmers to irrigate their crops and forces them to use diesel generators to extract groundwater. My team and I surveyed 250 farmers in three districts of Punjab. The data that we collected showed that 15-20 % of farmers’ irrigation requirements in the paddy season are met by using diesel. The cost of pumping using diesel is close to Rs. 6.00/unit, while the total cost of energy (including free electricity and diesel) comes to around Rs. 1.12/unit. So, even though electricity is free, energy is not and in the end farmers in Punjab end up paying for pumping groundwater. That farmer’s pay a certain cost for pumping groundwater is a good thing – it gives them the incentive to make efficient use of both groundwater and electricity. The opportunity cost of ‘wasting’ electricity is equivalent to the cost of using diesel and can be as high as Rs. 160/hour (~USD 3/hour) or upto INR 25,000 per acre (~ USD 500/acre). Such high opportunity cost means that farmers are wary of wasting that electricity. We found that all farmers use branded pumps from reputed companies and most have adopted laser leveling on their fields. Both these investments are aimed at making more efficient use of groundwater.

In Karnataka, the energy-irrigation nexus plays out a little differently. Unlike Punjab, here, use of diesel pumps is an unviable option because of hard rock hydrogeology and deep water tables. In the 1990s, farmers got 16-18 hours of electricity supply, and low voltage at that. Since then, there has been no improvement in quality and supply has further whittled down to a mere 3 to 6 hours a day. In response, more and more farmers are leaving their land fallow. We surveyed 215 farmers in the Kolar and Tumkur districts of Karnataka, and found that farmers on average keep 20% of their land fallow and 25% of their land un-irrigated because of their lack of access to both groundwater and electricity. Average land holding size is roughly 5 acres – so this equates to one acre of fallow land. What this basically means is a lost income of INR 23,000 /acre of land (the average gross value of agricultural output in Karnataka). This means, here too, like Punjab, the opportunity cost of ‘wasting’ electricity is very high. We also found widespread adoption of drip and sprinkler irrigation to make more efficient use of available groundwater.

So what does it all mean? It means that we need to re-visit the conventional wisdom that farmer’s waste electricity because it is free. In a political economy where pricing of electricity is a sensitive political issue and avoided by politicians, rationing offers a second best solution to the problem of groundwater over-extraction while invoking much less resistance. In other words, if you can’t price it, ration it.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s