CSPI Documents Library
Return to CSPI Homepage
National Opinion Poll on Labeling of Genetically Engineered Foods


     The United States and several other countries are currently debating whether to require food labels to disclose the presence of ingredients that were developed using genetic engineering. To help inform that debate, in April 2001, the Center for Science in the Public Interest ("CSPI") commissioned an opinion poll of 1,017 American adults. Respondents were asked questions about what information should be provided on food labels and how they might react to various labels statements. Consistent with previous surveys, this poll found that 62% to 70% of respondents desire labeling of genetically engineered ("GE") food. The survey found that most Americans also desire labeling for many other currently unlabeled food processes, such as whether crops were sprayed with pesticides (76%) or imported (56%). The desire for labeling of GE foods, however, was strong for a modest percentage of respondents. Seventeen percent of those surveyed picked GE food labeling (out of four choices) as their top priority, and only 28% of respondents would want GE labeling if it added $50 or more per year (about 1%) to their family’s food bill.

     The survey found that consumers’ attitudes and purchasing behavior would be affected by GE food labels. About 30% of consumers stated that GE-labeled foods were "not as safe" as or were "worse" than identical foods without such label information. In addition, 40% to 43% of those surveyed would buy products labeled "genetically engineered," while 52% of consumers would choose a product labeled "does not contain genetically engineered ingredients" over a product labeled that it does "contain" such ingredients. In other words, the poll indicates that many consumers would favor non-GE foods because straightforward label statements about GE or non-GE implies to them that non-GE foods are better and safer than comparable GE foods, even though most scientists and the U.S. Food and Drug Administration say that currently marketed bioengineered foods are just as safe as other foods.


     Agricultural biotechnology and GE foods are controversial issues, with concerns ranging from health to ecological disturbances to corporate power. One of the most contentious issues is labeling of GE foods, with the U.S. biotechnology and food industries opposing it and many nonprofit groups critical of agricultural biotechnology supporting it. The European Union, Japan, Australia, New Zealand, and several other nations require or plan to require foods containing more than insignificant amounts of GE ingredients to be labeled. Moreover, polls of American adults have found that a majority of respondents say they would like GE foods to be labeled.

     CSPI is a nonprofit consumer advocacy organization that has worked for three decades on food safety, nutrition, and other issues. CSPI is supported by foundation grants and the more than 900,000 subscribers to its Nutrition Action Healthletter; it does not accept industry or government funding. CSPI is well-known for its campaign to obtain Nutrition Facts labels, opposition to the food additive olestra, studies on the nutritional value of restaurant foods, and opposition to misleading food labeling and advertising.

     Recently, CSPI started a Project on Biotechnology. CSPI recognizes benefits from biotechnology, but has called for stricter regulation to ensure safety and bolster public confidence. To better understand public attitudes regarding the effects of labeling of GE foods, CSPI commissioned an opinion poll to explore labeling more thoroughly than several previous polls. The telephone poll, conducted by Bruskin Research (Edison, New Jersey), was a nationally random sample of 1,017 adults and was conducted from March 30 to April 1, 2001.


Consumer attitudes regarding the labeling of foods produced with GE and other technologies

     Many consumers desire information on food labels about how foods and their ingredients were produced. A strong majority wanted foods containing GE ingredients to be labeled, 62% in one question (Question #2) and 70% in another (Question #3). To put those response rates in a larger context, the survey asked about the labeling of other technologies. Seventy six percent of consumers wanted labeling for crops grown using pesticides (Question #3), 65% for crops grown using plant hormones (Question #3), and 56% for crops that were imported (Question #2). Remarkably, 40% of respondents said that they would like products containing cross-bred corn to be labeled (Question #3). Cross-breeding of corn (and every other crop), of course, has been used to improve corn for decades and would have to be listed on every product containing corn. Those results indicate that many consumers would like more information about how their food is being produced — be it through biotechnology, pesticides, importation, or even traditional breeding. Genetic engineering is one of several processes that many consumers say they would want to know about. One explanation might be that consumers, few of whom have ever lived on farms, want labeling for any process with which they are not intimately familiar. The public needs to be better educated about where food comes from, whether or not foods are labeled with process information.

     When asked if they could choose only one piece of information (out of four choices offered) to add to a label, 17% of respondents said they would add information indicating that a food was genetically engineered (Question #1). In contrast, 31% of consumers would add information about whether the food contained minute quantities of pesticides, and 16% did not specify any of the four choices offered. That is another indication that labeling of genetically engineered foods is a high labeling priority for only a small core of people.

     Another possible way to measure the strength of a person’s desire for labeling is to determine how much a person would be willing to pay for that information. Indeed, the U.S. Department of Agriculture has reported that segregating GE and non-GE crops and food ingredients to allow for accurate labeling would result in higher prices.(1) The survey found that 44% of consumers would pay "nothing" and another 17% would pay $10 per year on top of their family’s current annual food bill for GE food labeling. Only 28% were willing to pay $50 or more (Question #12). The average family of 3.16 people spends approximately $5700 per year on food.(2) Thus, only 28% of consumers were willing to increase their average annual food bill by 0.9% to obtain information about GE-ingredients. More significantly, of the persons who said that their highest labeling priority was genetic engineering (the 17% from Question #1), 50% of those people would pay nothing or $10 per year for that labeling. Similarly, of the people who believed that labeling genetically engineered foods should be required (the 62% from Question #2), 56% would pay nothing or $10 per year for that labeling. Although as many as two-thirds of consumers may desire labeling of GE foods, few appear willing to pay more than a small amount for that information. Many people who want GE-labeling, however, may feel strongly that they should not be the party who pays for the costs of labeling. If GE-labeling were required, consumers might or might not have to pay any additional costs for GE labeling, depending on how much of any labeling costs were passed on to the consumer.

     CSPI’s survey found that if labeling is required, which foods should be labeled depends considerably on how much of the engineered ingredients are in a given food (Question #11). When asked which foods should be labeled as "genetically engineered," 61% agreed that a whole food, such as a tomato, should be labeled, and 53% said that a processed food with a major ingredient (such as Wheaties made with GE wheat) should be labeled. 42% agreed that a multi-ingredient food (such as a frozen dinner) with a minor ingredient from a GE crop (such as corn starch) and only 38% agreed that a highly processed foods, such as soybean oil, should be labeled. Thus, the desire for labeling is strongest for whole GE foods and decreases significantly if the food contains either none or only small amounts of a genetically engineered ingredient. As might be expected, higher percentages of consumers for whom labeling of GE foods was their first labeling priority wanted labeling of processed foods with minor GE ingredients or without any engineered molecules than the survey group as a whole (54% versus 42% and 53% versus 38% respectively). Among the people who answered in Question #2 that they wanted labeling of GE foods, however, only 74% said that a tomato should be labeled and only 47% said that soy oil should be labeled. Those inconsistent answers may be an indication that for some people in favor of GE-labeling, the expressed desire for labeling is not strongly held. The inconsistency may also be due to the particular wording of the different questions.

How labeling might affect consumer attitudes and behavior

     The CSPI survey explored how label statements indicating the presence or absence of GE ingredients might affect consumer attitudes and behavior. When asked whether corn flakes with a GE label were better than, worse than, or the same as corn flakes without such a label, 30% stated that the GE-labeled food was worse and only 12% stated that it was better (42% stated that it was the same) (Question #9). Conversely, when comparing corn flakes labeled that they did not contain GE and corn flakes without such a label, 35% stated that the non-GE product was better and only 8% considered it to be worse (42% stated they were the same) (Question #8). Thus, 77% of consumers believe that foods without GE ingredients are either the same as or better than the same products without such a label. The survey also found that 40% of the people who said that foods labeled "made from genetically engineered corn" were better than or the same as foods not labeled (Question #9) also answered that they would not buy genetically engineered fruits or vegetables (Question #4). Therefore, a percentage of consumers who think GE foods are the same as or better than unlabeled foods still would not buy a labeled GE food.

     When asked whether foods labeled as containing GE ingredients were just as safe as, not as safe as, or safer than similar products without such a label, about 30% of consumers said that the labeled product was not as safe. Only 7% said that the GE-labeled product was safer (about 33% said the labeled product was just as safe) (Question #10). Thus, in addition to perceiving that GE-labeled foods are not as good as foods without such a label, about one-third of consumers also perceives that GE foods are not as safe as foods without such a designation.

     The actual language used to label a GE food could affect consumers’ perceptions of food safety. However, the survey found no significant difference between the perceived safety of products labeled as containing "genetically engineered wheat" and labeled "wheat developed with biotechnology" (Question #10). In both those cases, 6% to 7% of respondents said the food was safer and about 30% said the GE-labeled food was not as safe as an unlabeled version of the same food, while about 33% said the foods were the same. When asked about labels stating that the purpose of the genetic engineering is to reduce pesticide use, 21% of consumers answered that the food was "safer," a three-fold increase from a label that only stated "contains genetically engineered wheat." However, the same percentage of consumers (28%) still found the food with the label stating "reduces pesticide use" to be "not as safe." Therefore, consumers do not perceive any difference between the terms "genetic engineering" and "biotechnology," but an explanation of a benefit from the genetically engineered food affects some consumers.

     Finally, the survey asked respondents about purchasing GE-foods. When consumers were asked if they would buy foods labeled as being from crops made with genetic engineering, only 40% to 43% said they would buy those foods (Questions #4 and 5). At the same time, 40% to 44% of consumers stated that they would buy foods labeled as being from crops made with cross-bred corn. When asked to choose between two otherwise identical foods where one is labeled that it contains GE ingredients and the other is labeled that it does not contain GE ingredients, 52% of consumers said they would buy the non-GE food, whereas only 8% would buy the GE labeled food (37% did not care which food they brought) (Question #7). Thus, there is a preference among consumers for foods that are labeled that they do not contain GE ingredients, which is consistent with how consumers would react to labels indicating that foods do or do not contain GE ingredients.

Possible limitations of the survey results

     Although this survey provides valuable information about consumers’ attitudes and behavior regarding food labels indicating the presence or absence of genetically engineered ingredients, the responses may have been affected by several factors beyond the survey’s control. The extent of recent news coverage, the type (size, location and wording) of labeling that might be employed, and many other factors might affect how a person responds to specific questions. That is particularly the case when most respondents (55%) are not very or not at all familiar with the subject matter. In a telephone survey, individuals not familiar with biotechnology are asked to make spur-of-the-moment decisions about a complex subject that they may never have thought about previously. In addition, telephone polls reflect a person’s answer to the particular question asked, which may or may not be the same as that person’s buying behavior at the supermarket. For example, consumers may answer a generic question one way but act much differently when they consider purchasing a product for which they have brand loyalty. Similarly, although information may be on a label, a consumer may not read the information before purchasing the product. Thus, many consumers may continue to buy exactly what they have always purchased in the past, whether or not it contains GE ingredients.

     The survey measured consumers’ views on affirmative labeling for GE ingredients but did not attempt to analyze consumers’ attitudes concerning lack of labeling of foods containing GE-ingredients, as is now the case. Consumers might express strong feelings about such products. Also, the survey did not attempt to explain inconsistencies between different answers to similar questions. For example, 15% of the people who said in response to Question #2 that they wanted labeling for GE-foods did not answer that any of the foods in question #11 should be labeled. For all those reasons, further polling and focus groups would be valuable in increasing understanding of consumers’ knowledge, attitudes, and behavior regarding the labeling of foods for the presence or absence of genetically engineered ingredients.


     When asked about adding information to food labels, a majority of respondents to CSPI’s survey say they want information about numerous matters, not just whether a food was developed with genetic engineering. A small core of consumers considers information about GE foods to be highly important, but most consumers, including a majority of those for whom GE labeling is a high priority, are not willing to pay very much for that information. These results indicate that mandatory labeling would be useful to some people and that people need to be better educated about where food comes from. It is important that policy makers who are considering requiring GE labeling also consider the costs that would be borne by industry and consumers.

     The survey found that approximately 40% of consumers believe that GE-related labeling reflects upon the quality and safety of the food, even though many scientists and regulatory agencies have found no such differences for current products. Therefore, policy makers would have to identify a labeling system, including language, prominence, and disclaimers, that would be informative to consumers but not lead them to think that non-GE foods are safer than other foods, and that GE-containing foods are less safe than non-GE foods, when that is not the case.


1. According to a 1999 Economic Research Service study, there would be 12% premium over the farm price for corn and soybeans to segregate nonbiotech varieties of these commodities. The average preliminary cost to the U.S. grain handling system of segregating nonbiotech corn was estimated to be $0.22/bushel and the cost for nonbiotech soybeans was $0.54/bushel. See "Biotechnology: U.S. Grain Handlers Look Ahead" by William Lin, William Chambers, and Joy Harwood, in Agricultural Outlook, April 2000.

2. 1998 data on amounts spent on food were obtained from "Food Spending by U.S. Households Grew Steadily in the 1990s" from USDA’s FoodReview, Volume 23, Issue 3 (2000). Average family size data was obtained from "Age and Family Structure, by Race/Ethnicity and Place of Residence" by Carolyn C. Rogers in USDA’s Rural Minority Trends and Progress.