Article originally published by American Enterprise Institute.

Executive Summary

Genetically modified (GM) crops are plants in which DNA has been altered in a way that does not occur naturally through plant breeding. Genetic engineering transfers selected individual genes within or across plant species to produce plants with targeted characteristics.

Some have questioned whether GM crops have been responsible for increased crop yields. Comparing US yields to European Union yields (where GM crops are banned) provides evidence that GM technologies have increased crop yields. Agricultural yields have increased over the past several decades. But, such increases are not fait accompli. Rather, they result from the development of new technologies. Banning yield-enhancing technologies means that food crop production will be lower than would otherwise be the case, and more water, land, and other inputs will be needed to increase global food production.

Introduction

Genetically modified (GM) crops are plants in which DNA has been altered in a way that does not occur naturally through plant breeding.¹Genetic engineering transfers selected individual genes within or across plant species to produce plants with targeted characteristics. Golden rice, for example, is a GM plant that produces more vitamin A (which can reduce blindness in susceptible human populations) than conventional rice. For other crops, genetic engineering produces plants that are resistant to insects, viruses, and certain herbicides. The technology can greatly reduce or eliminate the need to use insecticides and fungicides to control insects and viruses, while herbicide-resistant traits increase the efficacy of weed control.

Over the past 20 years, genetic engineering has most often targeted crop varieties that resist the effects of glyphosate herbicides. A non-glyphosate-resistant crop (and all weeds) will die if sprayed with a glyphosate. However, a glyphosate-resistant GM crop will not die when sprayed with a glyphosate, while other plants (i.e., weeds) will die. The GM approach to weed control is much more effective than the traditional use of selective and nonselective herbicides. In some cases, several GM traits (e.g., insect or virus resistance) have been “stacked” into a single crop variety.

GM seed varieties were first introduced in the United States in 1996 and are currently available for corn, cotton, soybeans, sugar beets, canola, and alfalfa. Although GM adoption rates were initially modest, 100 percent of US sugar beets, almost 100 percent of Canadian canola, and more than 90 percent of US corn, cotton, and soybeans were planted with GM seed in 2016.²

The high adoption rates might appear surprising given that GM seed varieties are two to four times more expensive than conventional seed varieties. Agricultural producers, however, generally adopt practices that increase profits through per-unit cost reductions or increased yields. Cost savings associated with GM crop varieties have accrued through reductions in the use of agricultural chemicals, mechanical tillage, labor, and perhaps, increased yields. Nonetheless, the European Union has prohibited the use of these technologies.

A recent New York Times article by Danny Hakim titled “Doubts About the Promised Bounty of Genetically Modified Crops” discussed the impact of GM crops in the United States and Canada by comparing crop yield outcomes in those countries with outcomes in the European Union, which has banned GM technologies.³ Hakim asserts that although the United States has been using GM crops for two decades, yield trends between the two regions have not changed. Several other reports appear to echo similar concerns about an apparent lack of GM technology benefits.⁴ However, those assertions run counter to research that indicates GM technologies can increase yields through increased plant populations and reduced rotational impacts at the lower end of yield distributions,⁵ as well as improved pest control.⁶

Various issues regarding GM technologies have been extensively debated since their introduction in the United States 20 years ago. The importance of technological change in agriculture has a long research history, and many innovations have been the subject of similar assessments.⁷ Zvi Griliches investigated the impacts of research expenditures on the development of hybrid corn (a plant modification that occurs through the selective movement of DNA material contained in pollen across plants within a species) and related innovations.⁸ He estimated that the social rate of return on public and private funds used to develop hybrid corn technologies was substantial. Yujiro Hayami and Vernon Ruttan also examined the impact of technology on agricultural productivity. They argued that a continuous sequence of innovations can yield increased agricultural output even in regions where agricultural production is not the focus of the research that led to the innovations.⁹

The issue of yield outcomes resulting from GM crop technologies is complex, especially when comparisons are made across countries and regions. Hakim used data from the Food and Agricultural Organization of the United Nations (FAO) to claim that there has been no difference in per-acre yield trends among the United States (corn and sugar beets), Canada (canola), and the European Union since the introduction of GM crops. However, the “promised” benefits of GM crops could have little to do with yield increases.

For example, E. D. Perry, G. Moschini, and D. A. Hennessy observe that GM technologies have reduced mechanical tillage operations with resulting reductions in CO2 emissions, soil compaction, soil erosion, and production costs.¹⁰ While GM corn varieties have reduced the use of many production inputs (agricultural chemicals, labor, etc.), most weed and insect problems were effectively controlled before the use of GM seeds—albeit, with the use of more toxic chemicals, labor, and machinery inputs.¹¹

Consequently, perhaps we should not expect large increases in corn yields because of GM technologies even though these technologies reduce the amounts of other inputs being used by farm businesses and lower per-unit costs of production.¹² Nonetheless, it seems reasonable that if a technology increases the efficacy of weed and pest control and reduces soil compaction, yields should respond accordingly.

Hakim’s analysis is flawed in its assessment of the yield impacts of GM technologies for at least three reasons. First, Hakim chooses to only visually interpret the data (through linear trend lines on graphs) rather than conduct and report the results of more careful statistical evaluations. Second, the period Hakim selects to make the comparisons is both arbitrary and too short for valid trend analyses. Third, Hakim assumes that yield trends are necessarily linear in levels (that is, yields grow at a constant rate from one year to the next). All three of these issues generate errors that systematically bias Hakim’s results and conclusions.

Original article can be found here.