Wilhelm Gruissem reviews “A Meta-Analysis of the Impacts of Genetically Modified Crops”

Abstract: What has long been suspected is true: genetically modified (GM) crops do have real benefits for the environment and for the economic well-being of farmers. A meta-analysis of peer-reviewed journal articles and other literature not published in journals reveals that the adoption of GM crops reduces pesticide input and increases crop yields and farmers’ income. The results confirm earlier and smaller studies and therefore are not unexpected. But they are particularly welcome for significantly informing the public debate on GM crops.

How reliable is the Klümper, Qaim Meta-Analysis? I may have missed the essential critique – but in the open-access literature this commentary gets it about right. Here we have a Swiss reviewer at ETH Zurich on the work of German researchers:

Why can this study be trusted? The authors focused on herbicide-tolerant and insect-resistant crops (maize, soybean, cotton) for which a large number of original peer-reviewed impact study reports were already available and that have also been discussed widely in the non-peer reviewed literature. They searched not only in the ISI Web of Knowledge and Google Scholar for their analysis but also used the EconLit and AgEcon Search databases because their contents are tailored to economic studies. The keyword search was unbiased and designed to retrieve peer-reviewed and gray literature reporting both positive and negative impacts. This approach was different from previous reviews of GM crop impact that were limited to peer-reviewed literature only and therefore may be been skewed toward positive results. It is often studies without such peer review, for example, [11], however, that influence the public debate and therefore detract from fact-based decision-making processes.

When dealing with a large dataset on GM crop literature, effect sizes and influencing factors are important considerations because they allow a quantification of the extent of GM crop impact rather than estimating only whether or not an impact was observed. Supported, as they are, by rigorous statistics, the results of the meta- analysis reported by Klümper and Qaim [10] convincingly show that average agronomic and economic benefits of GM crop production are significant and sizeable. Although the review was limited to insect-resistant and herbicide-tolerant maize, soybean, and cotton, the impacts are likely to be similar for canola and sugar beet, which are now grown on large acreages as well. There was no evidence that studies funded by industry had any influence on impact estimates. Studies reported in the peer-reviewed journals trended toward a higher yield impact of GM crops than the average resulting from the meta- analysis [10]. This is perhaps not unexpected because non-reviewed (gray) literature published by nongovern- mental organizations that was included in the meta- analysis typically has a negative bias.