ON NOVEMBER 4th voters in Colorado rejected a ballot initiative that would have required special labels for foods made with genetically modified (GM) ingredients. As The Economist went to press, voters in Oregon seemed likely to say no to a similar proposal there, though the count was not complete. Regardless of the outcome, however, the referendums indicate the strength of feeling generated by GM crops: the Oregon vote was the costliest ballot in the state’s history. By chance, the day before the poll saw the publication in PLOS ONE of the largest review yet conducted of the crops’ effects on farming. It concludes that these have been overwhelmingly positive.

The review in question is a meta-analysis. This is a statistically rigorous study of studies, rather than a mere summary of the literature. Its authors, Matin Qaim and Wilhelm Klümper, both of Göttingen University, in Germany, went through all examinations of the agronomic and economic impacts of GM crops published in English between 1995 and March 2014. This provides a near-complete survey. Most studies of the subject have been published in English, and the widespread adoption of such crops began only in the mid-1990s.

Commercial genetic modification for crops comes in two forms. One makes them resistant to insect pests. The other confers tolerance to glyphosate, enabling farmers to spray their fields with this herbicide and kill off all the other plants (ie, the weeds) in them. As a consequence, the study found, herbicide-tolerant crops have lower production costs—though this was not true for insect-resistant crops, where the need for less pesticide was offset by higher seed prices, and overall production costs were thus about the same as for unmodified crops. With both forms of modification, however, the yield rise was so great (9% above non-GM crops for herbicide tolerance and 25% above for insect resistance) that farmers who adopted GM crops made 69% higher profits than those who did not.

Many poor countries eschew GM crops, fearing they will not able to export them to areas which ban them, notably the European Union. This has a big opportunity cost. Dr Qaim and Dr Klümper found that GM crops do even better in poor countries than in rich ones. Farmers in developing nations who use the technology achieve yields 14 percentage points above those of GM farmers in the rich world. Pests and weeds are a bigger problem in poor countries, so GM confers bigger benefits.

In debates about GM the methodology of studies has often generated as much controversy as the crops themselves. Drs Klümper and Qaim have done something to moderate these controversies, too. Though some studies they include were not peer-reviewed, and a few of the early ones did not report sample sizes, limiting their value, the data they used for the meta-analysis—which include conference papers, working papers and book chapters as well as work published in academic journals—may correct for perceived publication bias, the tendency of journals to publish only the most dramatic findings. This large body of evidence enabled the authors to control for possible differences in matters other than whether a crop was modified or not, such as fertiliser use. They also found that who pays for a study does not seem to influence its results.

Dr Klümper and Dr Qaim conclude by expressing a hope that their work “may help to gradually increase public trust in this promising technology”. To judge by the heat generated in Oregon and Colorado, that may take time.

(Photo credit: AFP)