as the introduction of transgenic (GMO) cotton seeds to India in 2002 the beginning of the renaissance of the country’s then struggling cotton industry? After years of farmers losing crops to tobacco budworms, cotton bollworms and pink bollworms, costing billions of dollars a year in losses, Monsanto developed insect-resistant Bt cotton in the early 1990s. The engineered crop has become widespread since its commercial release in China and the United States in 1996, followed by its introduction to India in 2002.
Within just a few years, India’s troubled cotton industry had done a 180, emerging as one of the world’s largest producers of GMO cotton, as exports boomed, helping to fuel India’s rapid rise as an emerging nation. But not everyone accepts this version of events. Agricultural biotechnology critics maintain that the success of Bt cotton was more smoke and mirrors than science, a story deceptively promoted by the beleaguered agricultural biotechnology industry and its supporters.
The Bt cotton debate was reignited this year following the publication of contrasting scholarly analyses, one challenging the success narrative and several others defending it. The latest volley of criticism was launched in March when Indian entomologist K. R. Kranthi and Washington University anthropologist Glenn Davis Stone wrote a scathing analysis of Bt cotton success claims in Nature Plants, an article widely disseminated by the global media. Reviewing 20 years of data, the authors claimed that, counter to the scientific consensus and established literature, the dramatic success of India’s first (and only) GMO crop was largely hype, and may have even been a failure. According to Stone in a press release put out by Washington University in St. Louis:
Yields in all crops [in India] jumped in 2003, but the increase was especially large in cotton,” Stone said. “But Bt cotton had virtually no effect on the rise in cotton yields because it accounted for less than 5% of India’s cotton crop at the time.
Now farmers in India are spending more on seeds, more on fertilizer and more on insecticides …. Our conclusion is that Bt cotton’s primary impact on agriculture will be its role in making farming more capital-intensive — rather than any enduring agronomic benefits.
That led to a rebuke by long-time scholars in the field. In June, GLP published a detailed critique by plant geneticist Deepak Pental, who wrote:
The article’s authors claim to have carried out ‘a new analysis of unprecedented scope, time depth and detail’ on cotton cultivation in India to find the real reasons behind the doubling of yields between 2000 and 2006, followed by yield stagnation. While the avowed goal of the analysis is to set the record right on the contribution of the Bt trait to cotton cultivation in India, the real purpose of the report is to cast doubts on the utility of GE technologies.
Then in early May, four scientists at the South East Asia Biotechnology Center in New Dehli weighed in with their own take down in the open access Cold Spring Harbor Laboratory Publication bioRxiv, concluding:
This study [Kranthi and Stone] conspicuously ignores positive shifts that occurred with Bt adoption at reduced real cost of production in all states resulting in large welfare benefits netting out increased cost of cultivation. [The fallacy] associated with increasing yield trends even before [the] introduction of Bt cotton as claimed by Kranthi and Stone does not stand [up to] scrutiny of increasing yield trends from 2002-03 to 2009-10, with some years showing significant yield dips due to drought [only] to bounce back …. in 2017-18. The ignorance of drought impact tends to attribute the yield reduction entirely [to] the failure of Bt technology.
Most recently, in a letter originally published in Nature Plants, agricultural economist Matin Qaim, who has been writing about the impacts of Bt cotton in India since its introduction, jumped into the fray:
Kranthi and Stone’s attempt to analyze long-term effects of Bt cotton is laudable, as the effects of the technology can change over time due to evolving pest populations and other dynamics. However, their claim that Bt contributed little to the yield increases observed in India between 2002 and 2008 is unconvincing ….
Strong arguments on both sides. What do the facts say? Let’s separate the cotton from the sharp ends of the boll.
What is Bt cotton?
Bt cotton is an insect-resistant transgenic crop (GMO) designed to combat the bollworm. It was created by genetically altering the cotton genome to express a natural, non-pathogenic microbial protein from the bacterium Bacillus thuringiensis that is found in the soil. Bt in its natural and transgenic forms has been extensively evaluated and found to be safe to all higher animals tested. Bt has been used as an insecticide in organic farming since the middle of the 20th century.
Traditionally, pesticides have been used to combat the cotton bollworm. However, in developing nations like India, the expense of using large amounts of pesticide is often too high for marginal farmers. Bt cotton was developed with the intention of reducing the amount of pesticide needed for cotton cultivation, thereby reducing production costs for farmers, environmental impact, and the pesticide exposure of applicators, often women and children.
Numerous independent studies have attributed anywhere from 14-30% of the cotton yield increase in India to the cultivation of Bt seeds. Five years after the introduction of Bt cotton, a professor at Jawaharlal Nehru University and visiting fellow at Centre de Sciences Humaines, New Delhi would write in the Wall Street Journal about India’s recently flagging cotton production: “By 2007-08, India became the largest producer of cotton with the largest acreage under Bt cotton in the world, pushing China into second place.” Many scientists and news organizations cited the surge in production of Indian cotton as one of the clearest GMO success stories.
Despite its resounding success, Bt cotton seed is more costly than non-transgenic (but lower yielding) varieties, making it a target for some critics who are skeptical of crop biotechnology. One of those longtime skeptics is Washington University professor Stone.
Stone is part of a cohort of scholars and activists, including Indian-philosopher Vandana Shiva, which fervently believes that the Indian Green Revolution that dramatically reduced hunger and is credited with saving more than a billion lives was a failure. As far back as 2012, Stone challenged a plethora of independent studies confirming the success of India’s Bt cotton crops and the resurrection of the nation’s once-threatened cotton industry.
But the data seem to disagree. Within a decade, Bt cotton accounted for more than 95% of all cotton cultivation in India, as yields increased and pesticide use decreased.
During that same period marking a 55% rise in yields, overall use of insecticides remained below absolute levels from 2003, while per- hectare usage dropped precipitously.
Glenn Stone, however, looked at the data from a cultural anthropology perspective and saw more hype than substance. Writing in his influential paper in 2012, “Constructing Facts: Bt Cotton Narratives in India,” Stone maintained, “We simply cannot say how Bt seed has affected cotton production in India.” The “triumph narrative” of Bt cotton in India, he claimed, “flows mainly from economists and the biotech industry (and its academic allies)” in “industry-journal authentication systems” (peer-reviewed journals), which “serve the interests of their constituent parties.” The arrangement is a “cosy alliance between GM manufacturers and ostensibly independent researchers,” he added.
Jump to 2020, and Stone, joined by K. R. Kranthi, the former director of India’s Central Institute for Cotton Research and now the head of a technical division at the Washington-based International Cotton Advisory Committee, reemerged as a sharp critic of Bt cotton—though the success narrative appears even stronger now. Since 2012, water usage has dropped sharply in Indian while Bt cotton yields have continued to climb, and are at or near historic highs, up more than 150% since the early 2000s.
Despite these numbers, Kranthi and Stone argued that “the largest production gains came prior to widespread [Bt] seed adoption and must be viewed in line with changes in fertilization practices and other pest population dynamics.” They also cited the pink bollworm’s evolved resistance to Bt insecticide and the threat posed by other pests that are impervious to the insecticidal power of Bt cotton.
Qaim found these arguments lacking, however. Building on previous scholarship, the agricultural economist explains that, when other relevant factors are accounted for, Bt cotton did indeed boost crop yields in India.
The agronomic and socioeconomic effects of insect-resistant Bacillus thuringiensis (Bt) cotton in India have long been debated. In their recent Perspective article, Kranthi and Stone  used 20 years of data to analyze associations between the adoption of Bt cotton, crop yields and insecticide use, claiming that Bt technology had little yield effects and did not produce any enduring benefits.
Here, I argue that the methods used by Kranthi and Stone are not suitable to make statements about causal effects, so their conclusions are misleading. As earlier studies showed [2–7], Bt cotton has contributed to sizeable yield gains and important benefits for cotton farmers and the environment. Kranthi and Stone’s attempt to analyze long-term effects of Bt cotton is laudable, as the effects of the technology can change over time due to evolving pest populations and other dynamics.
However, their claim that Bt contributed little to the yield increases observed in India between 2002 and 2008 is unconvincing, as this part of their analysis looks at the same time period that was also analyzed previously by other authors with more precise microlevel data and better methodologies [7,8]. Kranthi and Stone use simple graphical analysis to compare time trends for Bt adoption, fertilizer use and yield at national and state levels.
Comparing the graphs, they find a stronger correspondence between the fertilizer and yield trends than between the Bt adoption and yield trends. Thus, they conclude that the observed yield increases were primarily due to the higher use of fertilizer and other inputs, and not to Bt technology. The problem is that such a simple graphical comparison of time trends is inappropriate to analyze causal effects. Crop yields may increase because of more fertilizer or because of better pest control through the adoption of insect-resistant Bt varieties. It is also possible that some farmers decided to use more fertilizer because of Bt adoption. Many other factors, such as changes in irrigation, other inputs and technologies, agronomic practices, training of farmers or simple weather fluctuations may also affect cotton yields and broader socioeconomic benefits.
In principle, Kranthi and Stone acknowledge these complexities but they do nothing to control for any of the confounding factors. Previous studies used microlevel data and more sophisticated statistical techniques to control for confounding factors and possible bias, hence leading to more reliable effect estimates. Kathage and Qaim  used panel data collected between 2002 and 2008 from over 500 randomly selected cotton farms in four states of India. They used statistical differencing techniques and controlled for the use of fertilizer, irrigation, pesticides, agronomic practices and many other factors, including location and time trends, to deal with selection bias and cultivation bias.
Results showed that—after controlling for all other factors—Bt adoption had increased cotton yields by 24%, farmers’ profits by 50% and farm household living standards by 18%, with no indication that the benefits were fading during the 2002–2008 period. The same data also revealed that chemical insecticide quantities declined by more than 40% through Bt adop-tion, with the largest reductions in the most toxic active ingredients previously sprayed to control the American bollworm [9–11].
There are not many other examples from India or elsewhere where a single technology has caused agronomic, economic and environmental benefits in a similar magnitude.Against this background, Kranthi and Stone’s statement that “the surge in yields has been uncritically attributed to Bt seed” is not correct. Of course, there are other factors that contributed to the observed doubling of yields between 2002 and 2008 but the 24% estimate by Kathage and Qaim is the net effect of Bt technology after controlling for other factors . Using longer-term data but inap-propriate methodologies to challenge earlier results, as Kranthi and Stone do in their article, is not convincing. Bt cotton has increased yields through better pest control and has benefited adopting farm-ers in India and several other developing countries [12–14].
This letter was originally published in Nature Plants and has been republished here with permission. Nature Plants can be found on Twitter @NaturePlants