In January 2006, UK-based P G Economics reported that Bt cotton had on balance performed better in India than non-Bt counterparts, despite higher input costs. The findings were contested heavily by NGOs, who for some years now have been vocal in showcasing failures of Bt cotton crops - or put more accurately, cotton varieties with the Bt trait - around the country.

The criticism was predictable; the Bt Cotton controversy (or debate if you like) is a vexing one that increasingly wears polarised colours. Proponents of Bt praise the benefits of the technology, while a range of pro-sustainability NGOs charge that Bt has failed farmers and that its environmental risks far outweigh the benefits. But the debate itself not so simplistic. In a number of areas related to cotton crop failures -- the science of Bt, farmers experiences on the ground, positions taken by NGOs on transgenics and spurious seeds -- views, research and outcomes are fiercely contested. There isn't a clarity that gives watchful citizens the sense that they understand what's really going on.

Dr Ronald Herring
(Picture credit: Cornell University)

Ron Herring is Professor of Government at Cornell University, New York and also Director/Convener of the Program on Nature and Development there. He teaches political economy and political ecology. His current work includes state property in nature and politics of genetically engineered organisms (on which he is editing a special issue of Journal of Development Studies, Transgenics and the Poor). His articles have appeared in Frontline, Times of India, Financial Express, Economic and Political Weekly, and other Indian publications. He has also been consultant to the U.S. State Department, World Bank, UNDP, and other international organizations.

Professor Herring has a particular interest in Bt Cotton in India. He has been a regular visitor and has spent extended amounts of time conducting research. He is also currently one of the advisors of Devparna Roy, a Ph D student in Development Sociology at Cornell University whose forthcoming dissertation is on Bt cotton. Herring is skeptical of the NGO-led environmental movement in India. He argues that the NGOs are out of touch with Indian farmers. Subramaniam Vincent talked with Professor Herring for his sense of the news reports of Bt Cotton failures in A.P. and Maharashtra in 2005 and 2006.

Was Bt technology busted?

First, the science of Bt technology and its impact on crop success and failure have been misunderstood. Bt is a trait; it is not a cotton crop variety. The Bt trait causes a particular protein to be expressed; the protein kills a small class of bugs, most importantly the American bollworm, which often devastates Indian cotton. That is all there is to the technology: one gene, one protein. Varieties into which the Bt gene has been inserted are as susceptible to all the variables in cotton cultivation that non-Bt varieties are – micro-climates, sucking pests, soil conditions, rainfall -- including human factors such as farmers' familiarity with cotton in general and the variety in particular, spurious seeds, adulterated pesticides, delays in input arrival, and credit shortages. Any and all can cause disaster.

State governments, other ministries in Delhi, and farmers in villages are all outside the control of the Genetic Engineering Approval Committee.

 •  Chor Bt aur Bt Chor
 •  Illegal seeds to be uprooted
 •  Centre's no to Bt cotton

Cotton is a very risky crop. If there is a varietal mismatch between the agronomic conditions and variety, then independent of the Bt-pest equation, crops may fail. Drought resistant varieties are valuable in drier areas but may yield less; high-yield varieties often require more water and will not do well if water is not available; some seed varieties need early watering that farmers may not understand, depending on where they got their seeds and whether the seed vendor knew enough to tell them – a good example is Bollgard MECH 184, which has wilting problems without early water but yields well with early watering. And so on. This is why three hundred or so varieties of cotton are planted in India; this number is necessary to adapt to different regions with different agro-ecological conditions.

Mahyco Monsanto Biotech Limited had the first three Bt varieties approved for cultivation. Most cotton breeders felt these were not the three best possible varieties, but this was the choice, made a decade ago. Testing and development time is quite extensive. Three varieties (Bollgard MECH 12, 161 and 184) could not be expected to cover the entire range of agronomic variation of India. It was a start; some varieties did better than others, and all were costly. Each new variety of cotton requires biosafety testing for government approval, so there will be a time lag before the big seed companies reproduce the diversity of Indian cottons with the Bt trait added. Farmers and seed companies would like to see more varietal choice, opponents of the technology would like to see it curtailed. Seed companies would move faster if MMBL's fee were lower. In recognition, MMBL has lowered prices to meet the competition of other seed companies.

In Gujarat, farmers have adapted Bt varieties for their own conditions and created hybrid concoctions. These hybrids are cheaper and more locally adapted; many farmers believe they work better than the approved Bt varieties, and they are spreading rapidly.

A Gujarat seed producers association has data indicating that 1.1 million packets of farmer-bred Bt seeds of various varieties shipped from Gujarat to Punjab last year. This shows the interest in Bt in farmers, even though this interest may not be for the licensed/legal varieties. In fact, the Gujarat seed association folks feel that Bt must be delicensed so that seed producers can use the working varieties in the public domain and continually adapt and monitor them for quality control so that they can make a legitimate business out of them, as opposed to the sword of illegality always hovering over their seeds.

But what puzzles me is the continuous drum-beat of "Bt cotton has failed." If so, why do sales of MMBL's official seeds keep increasing? Why have other seed firms licensed this very expensive technology from MMBL? Why do farmers used saved, traded and locally-crossed illegal varieties of Bt cotton in ever increasing numbers and acres? Surely the NGOs don't want to argue that Indian cotton farmers do not know cotton!

Are companies guilty of marketing seeds without approval anywhere (in India)? This seems to be one element of opposition to Bt from NGOs.

India once had a heavily regulated official seed sector: certified seeds came with advice and extension help in theory. During the central planning period, seeds were central to agricultural development. And like much in liberalization, the state's role has declined in recent years. It is more market, caveat emptor, less state-led and state-certified.

Seeds that spread from farmer to farmer are in any event protected by the Seed Act of 1966 that makes it the right of farmers to exchange or sell or save seeds. So spread of seeds from farmer to farmer cannot be regulated unless the seeds are transgenic, and then the GEAC (Genetic Engineering Approval Committee) gets into the act under provisions of the Environmental Protection Act. I.e., genetically engineered seeds are regulated much more than other kinds of seeds. And there is good evidence that farmers are cheated by people claiming to be selling seeds of a variety they are not, especially when there is a shortage, and certainly Bt seeds after the rush to obtain them in 2001. Many farmers were duped, thinking they had bought Bt seeds but received only knock-offs.

Navbharat did register NB 151 as a hybrid with the state of Gujarat, as required by law. And it sold well. It was only later -- when it was discovered that this variety was transgenic -- that Navbharat was enjoined from distribution pending biosafety tests and approval by the GEAC. Ironically, Monsanto/Mahyco BLtd has been the most regulated of all seed companies. I've heard no reports of their selling spurious seeds, though spurious seeds are a huge problem through small seed/cum/credit and pesticide wallas. It is true that MMBL varieites were smuggled back into A.P. after the government refused to renew the license for these varieties; farmers were willing to pay more to get them, and they returned, but were at that point illegal in Andhra. Farmers prefer certified seeds, but are willing -- see Devparna Roy's work -- to take on more risk for greater yield and lower pesticide costs of the "Navbharat variants" -- what Roy calls "loose seeds."

Lets talk about the environmental risk of Bt. You seem to agree here that there is uncertainty in transgenics and Bt.

The crux of the issue is the opposition to the transgene in cotton. Many plants produce their own insecticides. Among plants there is a lot of gene flow. There are examples of "super-weeds" from conventional breeding: "Johnson grass" is a good example, and one that burdened my teenage years with a hoe. This is a perfectly natural cross between jowar and prairie grass, but it is a very destructive weed. (I would stress that superweeds need no genetic engineering to arise. Traditional plant breeding results in characteristics that can be shared with wild relatives, just as genetic engineering does. One difference is that there is more ecological monitoring of transgenes than products of other forms of plant genetic modification in breeding.)

Transgenic flow involves implanting a gene from one species into another to alter the crop species. The primary unknown with transgenics is how much horizontal gene flow from the transgenic variety into a wild relative of the agricultural crop will occur. And if gene flow does occur, will the offspring be viable and will they have some fitness advantage which makes them outcompete species?

What is the risk for cotton? Cotton pollen does not travel very far, and there is evidence that wild relatives do not become transformed to superweeds, but it is conceivable that wild relatives of cotton could obtain the Bt gene and thus be less susceptible to bollworms and similar pests. (See references)

What about a moratorium on GE crops/Bt Cotton then?

If there were no other existing externalities, (like pesticides), then that would be all right. But we already use pesticides heavily, so a moratorium on Bt Cotton is hard to justify. There is no doubt that in India, as in many other countries, Bt cotton is typically sprayed (with pesticides) much less than conventional cotton. Farmers are adopting Bt cotton at such a rapid rate mostly for reduction of pesticide cost and pollution of their fields, water and bodies. Virtually all conversations with farmers indicate this to be a major consideration; the forthcoming Ph D dissertation of Devparna Roy covers these issues systematically for Gujarati cotton farmers, based on extensive field interviews.

Think about a bio safety regime this way. They are expensive. Continuous monitoring of any ecosystem is expensive. The greater threat comes from invasive species. If I have a dollar, I would spend it to monitor invasive species and pathogens. The US says invasive species cost it 138 Billion Dollars in 2004. If you are a state concerned about welfare, then at the margin, where is the regulatory dollar best spent? The threat of transgenic contamination from cotton is relatively small compared to other threats. It is not that there may not be bad effects. But development policy is always about the utility of the marginal dollar: where should it go? Clean water? Nutrition? Orphan drugs? Rural clinics? Or biosafety regimes unlikely to work?

Other transgenics may be a different story. The transgenics debate must not put all plants in the same category. Each transgenic must be looked at separately. For Bt Cotton, the results that say Bt Cotton has failed have not compared isogenic varieties in their studies. The studies need to look at identical varieties of cotton with and without the Bt trait. I know of only one study that has done this for India, and it concludes that the best results in cotton come from a Bt variety (MECH 162 in this case) using integrated pest management. (See references)

The technology works, in India as in other countries; it's a question of fitting the technology into the right varieties, and on this score the creative work of farmers themselves seem to come up with good solutions – the "illegal" varieties of Bt cotton (Luxmi, Rakshak, Vasatch, etc.) – all many Navbharat variants. These "variants" are illegal offspring of the original Navbharat 151, a tremendously successful Bt variety bred by Dr D B Desai and his colleagues at Navbharat Seeds in Ahmedabad. The male plant is Bt, the female a local variety; the hybrid is made by techniques farmers in Gujarat have been using for a long time: the world's first hybrid cotton H 4 came from Gujarat. Farmers find re-crossing the variants to retain the Bt trait to be both profitable and useful for fitting varieties to local agronomic requirements. Vikas Chandak and Anil Gupta have done a pioneering study of how farmers engage in this participatory plant breeding of useful (albeit illegal) transgenic cottons. (See references)

One reads a lot about "the failure of Bt cotton". The question has to be disaggregated and treated empirically. There are certainly spurious seeds in the industry: some "failure" may be due to fake seeds pretending to be Bt. Farmers may believe they used Bt, but has anyone tested their seeds? Second, if farmers believe the government is going to give compensation for crop failure, you can safely predict more reports of crop failure. I haven't seen any evidence that the technology has failed, but there are many reports of crop failure in specific locations of specific varieties.

What is a transgenic refuge?

A refuge is any host plant (non-Bt crop, potatoes, oats, sorghum, and some weeds) not producing Bt proteins or not being treated with conventional Bt formulations. The purpose of the refuge is to supply a source of susceptible insects that may mate with resistant insects of the same species emerging from nearby Bt crop fields.

If the refuge is large enough, susceptible insect numbers should outnumber resistant individuals, and in theory, a greater number of the next generation offspring will be susceptible to the Bt crop.


Towards the end of 2005, reports from NGOs showed that small farmers have violated the refuge requirements for Bt Cotton. The refuges that needed to be there were not there. Isn't this an example of regulations for transgenic failing to work?

Refuges with a non-Bt crop are felt to be necessary so that a population of bollworms not exposed to Bt can continue to breed, slowing development of resistance to the Bt toxin. There are lots of reports of farmers ignoring this advice. Refuges are costly, especially for small farmers.

So, will resistance develop quickly? Perhaps, but there is so far no evidence of it. One advantage of India is that great biodiversity and small size of fields together make for a wide variety of alternative hosts for bollworms, even if no one plants refuges. Refuges are a burden especially for small farmers; there are variable reports of compliance, but if I had to guess I would not expect much compliance over time. But I’m not sure that’s a huge problem; I'm also not a biologist. Again, the science on this is incomplete. Moreover, a lot of money is made in proliferating biosafety regimes globally, and public opinion supports caution. So we will have a lot of regulations formulated for transgenics that may or may not be necessary, and many will be ignored by farmers: just as the testing process for the first Bt cottons was ignored by farmers growing Navbharat 151 under the radar screen of government.

Again, the problem with regulation of transgenics in general is that it is expensive and likely to fail – like controls on software, pornography, drugs, etc. Governmental institutions in India – and in other countries as well – seem incapable of policing seeds, even if the costs were justifiable. The number of illegal Bt varieties under cultivation already exceeds the legal varieties.

In 2001, the GEAC issued a burn order in Gujarat for illegal Bt cotton crops, but could not enforce it. State governments, other ministries in Delhi, and farmers in villages are all outside the control of the Genetic Engineering Approval Committee. For example, some years back, the Government of Maharashtra announced release of Bt Seeds in the state even before the GEAC approved Mahyco-Monsanto’s Bt varieties, just as the Government of Gujarat had refused to defy its farmers by uprooting and burning the Bt cotton crops.