Choking the Black Swan: GM Crops and Flawed Safety Concerns

Failure of precuationary principle

Despite the fact that the technology behind genetically modified crops has been around as long as Commodore 64 and been shown to be safe in hundreds of studies, anti-GM activists continue to spread misinformation.

Recently, a paper on the precautionary principle in relation to genetically modified foods has been making rounds in the anti-GM social media circles. One of the authors is statistician Nassim Nicholas Taleb, who has previously written books such as The Black Swan on the impact of low-probability events. The other two authors are physicist Yaneer Bar-Yam, and politician-philosopher Rupert Read. They attempt to develop an improved version of the precautionary principle in an effort to undermine the usage of GM crops.

What can a thinly veiled anti-GM paper written by a physicist, a politician and a statistician teach us about the risks of genetically modified foods? Unfortunately, it is just more of the same illusionary sophistry common among anti-GM activists.

Failing to understand basic biology

Their first anti-GM argument boils down to a common misunderstanding of basic biology:

Top-down modifications to the system (through GMOs) are categorically and statistically different from bottom up ones (regular farming, progressive tinkering with crops, etc.) There is no comparison between the tinkering of selective breeding and the top-down engineering of arbitrarily taking a gene from an organism and putting it into another.

First of all, both traditional plant breeding and genetic modification uses the same general approach: they introduce genetic variation and then pick the variants that suits human needs.

Traditional plant breeding does this by looking at the phenotype of the plant, cross plants and then pick the variant that has most of the trait of interest. During this process, thousands of genes are shuffled and altered into new combinations. These large-scale genetic changes, which have the potential to create toxic plants or plants harmful to the ecosystem, occur with low precision and they are essentially unknown to the breeder as he or she only has the phenotype to examine. The breeders can, if they want, release the new variant that has been created the very same day without any regulation. This method is usually very slow and can take many years and decades. It can be sped up by using chemicals or radiation to flood the plant genome with mutations and then select the plants that have the required properties. After that, breeders outcross the newly created variant with the wild-type to get rid of irrelevant mutations (selecting for the trait at every generation). Using chemicals or radiation in this way does not make it into a GM crop according to regulatory authorities.

Genetic engineering plants, on the other hand, result in small and precise changes that extremely well-known. Before being released, GM crops go through 10+ years of toxicological and ecological testing, from small-scale to field tests. Genetic engineering allows researchers to use genes from evolutionarily distant organisms, whereas traditional plant breeders are limited to genes from closely related organisms.

Here is a table summarizing the differences between the two methods.

Traditional plant breeding Production of GM crops
What is the size of the genetic changes? Genetic recombination causes thousands of large genetic changes. Adding or modifying one or a few genes qualifies as small genetic changes.
How precisely are the changes done? Very low precision because all breeders have to look at is the phenotype. Extremely high precision because scientists can use biotech techniques.
How well-known are the genetic changes? Unknown since you are only looking at the phenotype. Highly characterized, both by virtue of the techniques involved and because of regulation requirements.
How long does it take? Decades, unless you mutate seeds with chemicals or radiation (which still qualifies as traditional plant breeding by regulators). Very fast.
When can they be released? Right when they are made. No government regulation. After 10+ years of intense toxicological and ecological testing.

Second, genetic engineering of plants is not based on “arbitrarily taking a gene from an organism and putting it into another”. The usage of the word “arbitrary” here is not only false, it is also an emotionally manipulative weasel word. The choice of genes and species are chosen with great scientific care. There is nothing “arbitrary” with it.

The bottom line is this: if you accept conventional crops despite the risks, then you must also accept GM crops because they are safer.

Monsanto fearmongering

Predictably, the authors quickly descent into anti-Monsanto fearmongering, apparently oblivious to the fact that counterparts exists in the traditional plant breeding industry:

What people miss is that the modification of crops impacts everyone and exports the error from the local to the global. I do not wish to pay—or have my descendants pay—for errors by executives of Monsanto. We should exert the precautionary principle there—our non-naive version—simply because we would only discover errors after considerable and irreversible environmental damage

This same risk can happen by conventional plant breeding. Shuffling thousands of genes in an unpredictable and unknown way can cause the emergence of allergens, toxins and plant varieties that are ecologically harmful. There is nothing in this risk that is specific to GM crops. Traits can spread to wild relatives no matter if it was traditional or biotechnological methods that were used to create these traits in the first place. Even horizontal gene transfer from other organisms into plants occur in nature, so cross-species gene transfer is not a risk that is specific to GM crops.

The authors claim that they do not want to pay for the “errors by executives of Monsanto”. What about the errors caused by the executives of multi-national corporations that deal in seeds from traditional plant breeding? The risk of exporting local errors to a global systemic problem does not depend on the method used to produce a plant variety.

Flawed risk analysis

The risks of releasing genetically modified foods such as golden rice has to be weighed against the risks of not releasing them. Golden rice can provide children in developing countries with beta-carotene (a precursor to vitamin A) and thus prevent the serious consequences of vitamin A deficiency (VAD). According to the WHO:

—> Between 1/4 and 1/2 million children become blind every year because of VAD.
—> Half of them die within one year after going blind.

So what do the authors have to say about the risks of not releasing genetically modified foods? They dismiss it out of hand:

Invoking the risk of “famine” as an alternative to GMOs is a deceitful strategy, no different from urging people to play Russian roulette in order to get out of poverty.

Preventing half a million children per year from going blind is the same as urging poor people to play Russian roulette to get out of poverty? GM crops are more safe than their conventional counterparts. If their view was accurate, then providing people in developing nations with seeds from traditional plant breeding would be equivalent to a genocide of poor people.

In reality, they have been cognitively confounded by their own analogy. Russian roulette is psychologically associated with a high risk of failure. No such evidence has been presented by the authors (or anyone) that this is the case for GM crops.

In the end, the authors have clearly demonstrated that they do not care about biology, medicine or rational risk analysis. They have negligible knowledge of molecular biology, plant breeding and genetic engineering. It does not matter how much knowledge they have of statistics. If your model is based on flawed premises, then the application and conclusion of that model is going to be flawed. Garage in, garbage out.

The main message here is abundantly clear:

GM technology causes smaller, more precise and more well-known changes than traditional plant breeding, and they are much more stringently regulated. Thus, GM crops are safer than conventionally bred crops. If you accept conventional crops, then you must also accept GM crops. Thus, the Russian roulette argument does not work.

Categories: Debunking Anti-GMO, Debunking Misuse of Statistics

Tags: , , , ,

11 replies

  1. Excellent article. I have just finished reading Charles Wheelan’s very good book, “Naked Statistics” which is a great introduction for people (like me) whose knowledge of statistics isn’t strong. He made the exact same point about flawed models that you did and also used the phrase “Garbage in, garbage out”. I was always under the impression that that had originated in computer programming.

  2. Excellent rebuttal. Before the Taleb PP paper came out, there was a massive blow up on his FB page when he started talking about GMOs. Anyone with a differing opinion was banned. If you pointed out his lack of knowledge in biology, you were called a shill then banned.

    • And if you admitted that you were an educated person, with a knowledge in biology, you were castigated and usually banned, unless you could argue without referring to your knowledge of biology, which is silly.

  3. Hi, I’ve here seen the information you post here that GMO crops undergo 10+ years of years of intense toxicological and ecological testing. From my understanding the FDA does not require that. Can you link me to something so I can read about how this process takes place ?

    • In the post, I wrote:

      Before being released, GM crops go through 10+ years of toxicological and ecological testing, from small-scale to field tests.

      Regulation of GM crops varies between geographical region of course, but in general it takes over 10 years from discovery to commercial launch. This involve small-scale trials, field trials of different sizes, ecological and toxicological testing and so on. One survey that shows this is the Phillips McDougall survey from 2011 that can be found here. On average the time period from start to commercial launch was 13.1 years (the span was from 7 to 24 years) and the cost was 136 million USD.

      The FDA writes in their FAQ that:

      4. How is the safety of food from a genetically engineered plant evaluated?

      Evaluating the safety of food from a genetically engineered plant is a comprehensive process that includes several steps. Generally, the developer identifies the distinguishing attributes of new genetic traits and assesses whether any new material that a person consumed in food made from the genetically engineered plants could be toxic or allergenic. The developer also compares the levels of nutrients in the new genetically engineered plant to traditionally bred plants. This typically includes such nutrients as fiber, protein, fat, vitamins, and minerals. The developer includes this information in a safety assessment, which FDA’s Biotechnology Evaluation Team then evaluates for safety and compliance with the law.

      FDA teams of scientists knowledgeable in genetic engineering, toxicology, chemistry, nutrition, and other scientific areas as needed carefully evaluate the safety assessments taking into account relevant data and information.

      FDA considers a consultation to be complete only when its team of scientists are satisfied with the developer’s safety assessment and have no further questions regarding safety or regulatory issues.

  4. There is a very major flaw, here. If one has a need to assess genetically modified foods, one should also have the same approach to conventional and organic in order to come up with a fair evaluation. I am sure, due to the lack of research being done on Organic farming practices, that Organics would defintely fail, as we know there are chemicals used that cause reactions for individuals.


  1. Another Week of Climate Disruption News – May 11, 2014 – A Few Things Ill Considered
  2. The Hypocrisy of Pseudoscientific Cranks: Response to Criticism | Debunking Denialism
  3. Anti-GMO Statistician Nassim N. Taleb ‏Responds to Criticism (Sort of) | Debunking Denialism
  4. Taleb’s Crock Pot: Whip Statistical Theory and Rhetoric To a Fine Agitprop | Sacred Cow Chips
%d bloggers like this: