Yes, you read that headline right. We should support labeling “GMOs”¹ because of science, but not any science about the process of genetic engineering or agriculture or pesticide use. No, the science I’m referring to is psychology. The current political and cultural environment of the United States² is heavily polarized on several major issues where science is an important factor (e.g. climate change). I don’t want genetically engineered foods to become one of them. Opposition to labeling “because the science doesn’t support it” plays into a narrative that I believe will increase emotional and cultural reasons to support or oppose crop biotechnology regardless of the evidence.

I’ve always cared a lot about how scientific information is shared (a very early post here criticized how scientific information was being presented in popular media). But last summer I started caring even more and realized that there’s actually science on this topic! Dan Kahan studies something he calls “cultural cognition”. The short version is that we process information about risk using shared values and it affects how we judge the relevance and correctness of new information. If an issue becomes culturally polarized (as it is with climate change), it’s very hard for new information (new science, etc.) to actually be useful in understanding or driving policy. He looked at whether or not GE foods were currently polarized last year and concluded they weren’t. Yet. Further, he argues labels might more easily allow “bad actors” to attach emotional meanings that pollute public discourse. That has long been about how I felt about labels: if we put on vague, useless labels won’t that just make it easier to fear-monger?

But I’ve changed my mind. It took a while. During the election last year, I spent a lot of time talking about, writing about and arguing about California’s Proposition 37 to label genetically engineered foods. Most of the arguing was probably useless³. I wasn’t going to change anyone’s mind and most of the people I was arguing with weren’t using arguments that were going to convince me. Gaythia Weis was an exception. She stayed well away from what I frankly consider absurd claims and argued on the basis of public risk perception. I’m going to mangle it horribly, but one point she made was that by aligning with the idea that information should be hidden from the public, scientists and others will be perceived to be less neutral. Sure, the science “doesn’t say” there is any (health, etc.) reason why you need to know that your foods have GE foods in them but that isn’t the only reason people want labels. But I resisted this argument.

The last six months or so, especially after a short piece by Ramez Naam, I’ve turned around to the idea that supporting labels for genetically engineered foods is the simplest way to reduce the risk of cultural polarization for the large segment of Americans who just don’t give a damn about GE foods. The short version of Naam’s piece is that by resisting labels, we feed paranoia about agricultural technology and that can only reduce future support for agricultural technology. He argues labels are inevitable. Wouldn’t it be better if industry and government created labels that aren’t likely to increase risk perception (as Kahan fears)? Including whether or not a food may be GE in the usual ingredients list seems far less likely to support fear-mongering than any of the state laws now under consideration. Why not support that and short-circuit the entire argument?

Naam cites some science on risk perception⁴ to support his position but the main piece of psychological science that convinces me opposing labeling is the power of narrative and explanation to how we remember and process information. An awesome review article I read last year summarized a study about how people need a story for why some fact makes sense. In one study, participants read newspaper articles about a warehouse fire. The first article mentioned that the fire was started due to some inflammatory materials stored negligently in a closet. Participants were then given a follow-up correction article that clarified the closet was empty but didn’t provide a replacement narrative for the cause of the fire. When questioned, the study subjects still believed the fire was caused by the (non-existent) contents of the closet because that explanation made sense and there wasn’t a plausible alternative. For GE foods and labeling, I’m seeing the same problem. Why are “GMO” labels bad? What is the compelling narrative that explains why you (the skeptical, low information voter) don’t actually need or want labels for GE foods? Let’s list some arguments I’ve seen against labeling:

• They are a warning label. Americans are misinformed on GE foods so that if we label, they will be prone to believing every scare story about them and it will force GE food off the market. Telling people they are ignorant and stupid does not seem like a winning narrative. It certainly doesn’t seem to work on other topics (risks of vaccination, climate change, etc.)
• They aren’t scientifically necessary as there are no (increased) risks from GE foods or GE crops that non-GE ones don’t have. As Ramez Naam notes, there are myriad pieces of information we include on food packaging that don’t tell us anything about the risks of food or methods of production. Moreover, as Gaythia Weis pointed out to me numerous times, not all reasons for consumers to want labeling are related to the health or safety issues of GE foods. In the US, GE foods are largely used in the “industrialized” food supply (i.e. corn and soy) and some consumers want to avoid that⁵.
• You can always buy food labeled organic if you want to avoid GE foods. Telling people they should buy more expensive food that they might perceive as “being for rich people” seems like a poor idea. There are few (no?) other labeling differences between “organic” and “conventional” foods. Worse, many Americans (a majority depending on polls) believe that organically produced foods are healthier to eat and safer for the environment. Telling people who might want to avoid GE foods to eat organic confirms a narrative about the relative risks of organic vs GE foods!⁶
• Those who support labeling are opposed to GE foods and GE foods (such as Golden Rice) can save lives! This story is somewhat compelling. However, it’s really complicated, hard to tell and not terribly relevant to Americans. The GE foods we eat in the United States are not particularly useful to consumers and don’t save lives (any more than non-GE equivalents).
• Labeling GE foods will increase food costs to consumer. This is the one story that seems to ring true to consumers. Looking at how the California labeling battle played out, I suspect that the claims of increased costs sunk the initiative. The pro-labeling forces did themselves no favors by repeatedly claiming there were absolutely no costs if the initiative passed. It just wasn’t a credible claim⁷. But is opposing transparency merely on (probably small) costs, worth it?

Now, consider the main point of the pro-labeling forces: the right to know. You have a right to know what’s in your food. The more responsible pro-labeling groups are fairly careful to stay away from outright claims that GE foods present actual health or safety risks. Instead, they might claim they aren’t tested enough (which, depending on how you view voluntary testing, is arguable). But the unknown risks are implicit in the argument. You have a right to know. They (big food companies, big agribusiness, unregulated regulators) don’t want you to know. Why is that? What don’t they want you to know?

If you’re a low-information voter, who doesn’t spend much time researching issues, who depends on proxies like the opinions of groups you support, this argument, even if it’s barely an argument, is pretty compelling. The reasons why you wouldn’t want to know aren’t obvious and just aren’t very convincing.

So am I going to vote for Washington’s initiative to label genetically engineered foods? I don’t know. I think a lot of what the campaign in favor says is subtle fear-mongering. Many of those who support it, but aren’t “officially” part of the campaign, say and publish things that are simply untrue. The law itself is poorly written. It requires front of package labeling — far too much like a warning label. Worse, due to political and legal limitations, there are many exemptions that make the law all but useless for its stated purpose — for example, if you want to avoid all GE foods, don’t ever eat out. Given the nature of federal and state law, I even think it’s possible it will be immediately tied up in jurisdictional conflicts and not implemented. It’s just not a good law.

On the other hand, I think labeling is inevitable. Perhaps losing in one state will force the food industry to voluntarily label, ideally with something better than the laws proposed so far. Or alternately, go to the FDA and ask for regulations to be made for labeling them. While there are people for whom voluntary labeling will be seen as just another part of the conspiracy, I think for most voters (remember, the low interest, low information ones) it will make the “what aren’t they telling us?” question disappear. That story just isn’t as powerful with a label. Many (labeled!) food additives have “scary” claims made about them, but very few people care enough to take them all that seriously. Labeling lets us stop arguing over “GMOs” which really aren’t the most important question as far as sustainable agriculture goes.

GE labels might mean we can talk about better labels. I want to talk about how we balance high yields against other values⁸. Without the missing labels to trigger fears of unknowns, we can talk about the complexities of agriculture without it all being Monsanto acting in the shadows. If farms use an herbicide-tolerant crops, I want the question to be when and why applying the chemical herbicide is the most appropriate weed control method, not just whether or not the crop was genetically engineered. When we start talking about pest resistance, I want the conversation to be about all methods of pest control and how you encourage and support the least harmful ones, year to year, field to field. Instead we talk about whether or not Bt proteins (but only in “GMO” food) can hurt mammals. I want to have those conversations, but it gets immediately bogged down “is it GE or not?” which isn’t that interesting.

I’m still pretty conflicted, but I now think it’s foolish to defend a lack of transparency. A lot of what we label is pretty arbitrary and we only do because, at some time, it seemed necessary for consumer trust. At this time, labeling genetically engineered foods appears necessary for consumer trust. Sure, it’s not “supported by the science” but politics and regulation aren’t only about science. They are also about values and belief systems and right now opposing labels makes it look like the food industry — and anyone else who opposes labels — has something to hide. You don’t fight so hard just to hide nothing … so what exactly are we trying to hide?

Yesterday, the USDA released an update on their analysis of the unexpected presence of an uncommercialized transgenic wheat in an Oregon field. There’s been a lot of unjustified attraction to the idea that since we haven’t figured out what happened yet, it must be sabotage. Unlike actual cases of anti-GM activism, there’s no direct evidence. When a lab is blown up by anti-biotech activists, it’s obviously a malicious act. Labs don’t just blow up for no reason. If a field trial goes up in smoke unexpectedly, it’s pretty obviously a malicious act. Fields that are behind locked fences with restricted access don’t light on fire spontaneously.

But this is pretty different. Unexpected seeds do actually appear in fields all the time. This is inevitable given imperfect seed cleaning, etc. Most of the time no one cares or even notices. In this case, if the farmer had tried to kill the wheat with a different herbicide, he probably wouldn’t have sent it off to be tested. Now obviously biotech seed trials are more heavily controlled than regular commercialized seeds. When the trials were ended, fields and equipment were cleaned thoroughly. Everything is inspected and documented. But humans do make mistakes and the seed line was available at dozens of places around the country for several years. It’s not impossible for some to have gotten somewhere it shouldn’t have and even possible no one noticed till now. There are literally hundreds of commercial wheat seed lines in the United States.

But they’ve tested some, of course. Monsanto, ironically doing a better job of public reassurance, has been testing commercial seed lines and publishing the results. Last week, they said they had tested enough varieties to cover 60% of the commercially planted acres of white wheat wheat in the Pacific Northwest and found no presence of the unexpected trait (white wheat is the variety the farmer in Oregon planted). The USDA’s press release doesn’t mention anything so specific which is unreassuring and vague. How many varieties have been tested? If there is some commercial seed line that has a small percentage of its output containing the transgene, it would only take a small bit of mixture. I wish the USDA would have been more specific about what they have tested.

Still, given how much has been tested, the “contamination”¹ is a mystery. Perhaps it is sabotage. But how would it happen? As Andrew Kniss noted on twitter yesterday², how would the saboteurs get the seeds? Monsanto starting shutting down the program in 2003. The last fields were cleaned up in 2005. Presumably Monsanto keeps the seed line alive somewhere, but likely that’s pretty heavily locked up. So did activists break into a Monsanto facility, find just the right seeds, then go pick a random field in Oregon to “contaminate”? Alternately, did some activists steal some seeds literally a decade ago, keep the line alive by planting and harvesting, then contaminate the field this year³?

I find the sabotage idea laughable at this point. I do not accept absence of explanation as proof of malicious intent. As far as I know there are literally hundreds of wheat lines that could have been the source. In my previous post, I clearly don’t consider this discovery unexpected or even particularly scary, but I do think it’s scary that some have so quickly jumped on malicious intent as an explanation. Many in the pro-GM side of the debate do not take the concerns of others very seriously. Escape of traits is a concern. Jumping to sabotage as the explanation rather than accepting that we don’t know everything is wrong. It reinforces the perception that some groups pushing biotechnology are arrogant and unconcerned with real risks and possibilities.

Some who oppose biotechnology spin fanciful stories about the risks of genetically engineered crops. Many of their stories depend on reasoning much like this. Without evidence, many anti-GM activists will accuse researchers or writers of being in the pay of Monsanto or wanting to “destroy the earth”. They don’t have positive evidence. They don’t need any in their mind because it’s the only explanation left in the complicated and confusing world they see.

Without evidence, a claim of sabotage here is the same. There is a world of possibilities for what happened. I’d rather a slow and careful investigation than conspiracies and accusations of malice. Instead of encouraging calm and reasoned discussions of values, risks and science, sabotage suggestions will just create a shouting match of competing conspiracies. That won’t do any good at all. To most people, Monsanto and large agribusiness surely look more malicious than activist groups who are trying to make the world better. Can we let the USDA and Monsanto continue their work? No one wins if evidence goes out the window.

This will hopefully be a short note, but something bugs me a bit about the current kerfluffle over one farmer finding a small number of wheat plants that have a Monsanto-origin trait for glyphosate tolerance. Even though we still don’t know how it got there (it’s been eight years since any field trials and it was never commercialized) or even how broadly spread the trait is in commercial seed, we’ve got entire countries stopping some wheat imports, others pushing conspiracies that it was intentional sabotage¹ and GE labeling groups using it to support their cause². But all this points to the fact that we aren’t having a reasonable discussion about the actual risks involved in plant modifications. The trait found was one that makes wheat tolerant to the herbicide glyphosate. It’s been successfully used in corn, soy and other crops. It’s safety record is quite good.

But Monsanto isn’t the only organization to develop glyphosate-tolerant wheat. Some researchers at Washington State University have developed non-transgenic glyphosate tolerant wheat, using mutation breeding (thanks to Steve Savage for mentioning it on twitter). They even patented it (most likely the university decides how to license it for development in commercially available seed). What is really different about this trait over the transgenic one? Fundamentally, the risks are similar. The major risk with herbicide tolerant crops (beyond overuse of an herbicide selecting for resistant weeds) is that the modified crop can interbreed with wild relatives, passing on the herbicide tolerance which might establish in the wild. This appears to be possible with wheat. Researchers are thus somewhat wary of widespread use of herbicide-tolerance traits in wheat (thanks to Bill Price pointing this out on Biofortified).

However, under the United States regulatory system (and most other nations’), only the transgenic wheat undergoes greater legal scrutiny. What would it mean for non-transgenic glyphosate-tolerance traits to be found unexpectedly in a farmer’s wheat crop? Practically it would mean the same to the farmer (difficulty controlling volunteers), but legally it would be very different. There would likely not be any talk of suing WSU for letting the trait escape or would Korea or Japan be suspending wheat imports. I’m not sure we would have even heard about it.

The worse part, to my mind, is that this inconsistent regulation means that non-transgenic versions of a trait may not be subject to the same scrutiny and care. With current transgenic traits, farmers often agree to use the seed responsibly. For example, with Bt corn, farmers are required to plant refuges³. There’s little in the regulatory system that would consistently require that for non-transgenic traits. The reason it works with transgenic ones is that the seed seller (e.g. Monsanto) has to get regulatory approval which mind end up requiring farmers agree to use it how the regulator wants. Fortunately, most seed sellers try to manage their products well even if non-transgenic (for example, Clearfield products are non-transgenic herbicide-tolerance traits that BASF at least recommends good practices.) I’m not sure what the answer is here, but it’s clear that we don’t consider non-transgenic traits as having similar risks at the regulatory level, even though they can.

Another part of this question is the expectation that there be 100% no “contamination” of seed lines with unexpected traits — if that trait is transgenic. But fundamentally this is impossible. We can make it unlikely it will happen (thru careful equipment cleaning, buffer zones in trial fields, etc.) but we can’t make it impossible. The focus on this being “GM wheat” distracts from reasonable investigation of how and why it happens and what is a reasonable way to control it. If it hadn’t been a transgenic version of the glyphosate tolerance trait (and even worse, Monsanto’s), would it have even been news? How many resources could the USDA have put behind the investigation to find out where the trait entered the seed supply? Would they have quietly tested likely varieties and, when they didn’t find them, throw it up as a fluke? Would there be greater pressure to have management practices (and regulations!) to prevent spread of possibly risky traits into wild relatives on the assumption that seed will occasionally be “contaminated”?

I’m not sure I have any answers. It just bothers me that (so far) a very low risk discovery is being spun into apocalyptic proportions. Grist worries about “GMO wheat infecting the food supply” which is clearly absurd. Would they worry so much if it was a trait derived from mutation breeding, but still with all the same practical risks to farmers and the environment? Our current regulatory system suggests few would. Maybe we can get past the idea that everything “genetically engineered” is dangerous while treating non-GE as safe. I still hope we can regulate the outcomes of using particular crops or crop systems, rather than the methods that developed them.

The headline was so promising: Anti-GMO Grass-Roots Effort Gains Ground in U.S. I hoped it would be about just how strange the alliances in the anti-GMO movement are. We have groups like the Environmental Working Group or the Union of Concerned Scientists along side organic industry boosters, organic food companies and off to one side (but louder and shriller than most) groups that claim dire risks from GMO foods. It’s a topic that deserves more attention.

Instead, I found a piece that repeated — almost without challenge — misleading claims about GMOs as given by anti-GMO activists. Most of these claims are at best half-truths. Let me be up front: I can’t write about everything I think is misleading in this piece. There’s just too much. But, there are a few ideas that are commonly repeated uncritically which are far more interesting than this piece lets on.

This piece is a slightly edited version of a rebuttal I wrote for Keith Kloor’s Collide-a-scape blog. Thanks to Keith for offering me the opportunity to get this to a wider audience. He also helped me tighten this up quite a bit from the original draft. Please have a look at the conversation on Collide-a-scape.

Probably the most common claim of anti-GMO activists is that GMOs are untested and unsafe. Charles Benbrook, an organic proponent (why is a proponent of organic and opponent of GM merely labeled an “agricultural policy expert”?) is quoted: “science just hasn’t been done”. That’s just not true. There are hundreds upon hundreds of studies. Genetically engineered crops are some of the few foods tested before they come on the market and all the data is on the EPA’s website. This testing is not typically the case for non-biotech foods. In the 1950s, the kiwifruit was introduced to the US without testing. We’ve since learned that it’s allergenic. More recently, new celery varieties have resulted in contact rashes in workers due to increased amounts of natural toxicants called psolarens (similarly, new potato varieties sometimes develop excess toxic solanine).

But there’s a kernel of truth behind the claim. David Schubert (who is given as a biologist without noting his association with noted anti-GMO campaigner Jeffrey Smith) is very careful when he says: “no significant safety testing is required by FDA.” That is absolutely true. All safety testing done on GMO crops is voluntary. But even though it’s voluntary every company has complied! The lack of mandatory testing does unnecessarily worry people: what good is the FDA if it can’t even require safety testing for food? This is perhaps why the American Medical Association recommended mandatory safety testing, despite also saying current GMO foods are safe and labeling is unnecessary. Pre-market testing should probably be mandatory, but it’s just not the case that GMOs are untested before going to market.

GMO foods are also subject to repeated studies by scientists after they come on the market for toxicity and other risks. Overwhelmingly, these studies show no problems for human health. You wouldn’t know that from this piece, since it chose to mention problems found by a small minority of researchers, often heavily criticized. The piece wiggles around the topic by noting it’s hard to prove a causal link in humans. The reality is links aren’t even proved in animals.

But what about those “superweeds”? And that toxic Roundup? There’s context missing here too. Roundup is indeed toxic — to most plants and marginally to some animals — but it is one of the safer herbicides we have. While the term “superweeds” is not one scientists use, preferring phrases like “herbicide resistant”, they are a real problem. We have seen an increase in weeds resistant to glyphosate (the active ingredient in Roundup). But what this piece is not telling us is this is normal and expected: any herbicide used heavily for years will naturally select for weeds that can survive it. The most important missing context though, can be found in this graphj, which shows cases of herbicide resistant weeds, grouped by site of action.

One herbicide deserves special mention. Atrazine (a member of the triazine family) doesn’t harm corn, so it’s been used for decades to control weeds in corn fields. Unsurprisingly, that class of herbicides has had significant weed resistance for much longer than Roundup. The point here is emphatically not to deny the problem of herbicide-resistant weeds. That is a real problem. But it’s not a problem with genetically engineered crops per se, but with management practices. Roundup Ready crops could have been banned ten years ago and farmers would still have weeds resistant to herbicides. You just wouldn’t hear about them from anti-GMO campaigners.

There’s a real story in the alliance of anti-GMO groups and how, in the United States, they are using the political system, instead of vandalism (as in Europe) to advance their agenda. This story isn’t it; instead, it prefers to repeat myths and exaggerations about GMO foods in a one-sided manner.

There are some good questions about GMO activism springing up in the U.S. For example, why do organic food growers care so much about labeling GMO ingredients when transgenic crops are not allowed in organic products? Why is the anti-GMO campaign only now gaining traction in the United States? What’s behind that? Those would be interesting questions to explore a story.