Anyone who teaches Food Law & Policy knows that “genetically modified food” will get all the students participating and bring out some passion in the classroom.  As the National Academies of Sciences pointed out in their 2016 report, Genetically Engineered Crops (see pages 48-51), public opinions are very strong and range from intense opposition (on the view that genetically modified food is “extremely risky”) to strong support (“overwhelmingly beneficial”) even though, as NAS pointed out, most Americans nevertheless do not know much about genetic engineering as it relates to agriculture. 

Even the dispute over the term used to describe these foods reflects some of this tension.  FDA prefers the more precise phrase “food derived from genetically engineered plants” — because there are many ways to alter the genes of a plant, including selective breeding.  (See guidance, here.)  It points out, too, that “non-GMO” and “GMO-free” are usually misnomers, because O is for “organism” . . . and these are usually put on foods that don’t actually contain organisms in the first instance.  But the agency is fighting an uphill battle.  Consumer usage — although patently unscientific — drives the train.

This leads to the challenge of figuring out the right role for the public, when policymakers synthesize the science on genetically engineered food and set public policy.

The National Academies of Science — which was tasked with conducting a broad review of genetically engineered crops (assessing the evidence for purported negative and positive effects, among other things) and preparing a report directed at policymakers — engaged in a highly consultative process with the public (see pages 39-42) even though it was focused primarily on assessing the science.  Why?  It takes the view (based on a 1996 report from the National Research Council on Understanding Risk: Informing Decisions in a Democratic Society) that a “purely technical assessment of risk could result in an analysis that accurately answered the wrong questions and was of little use to decision-makers.”  (See page 47 of the NAS report.)  The NRC report, NAS notes,  “outlined an approach that balanced analysis and deliberation in a manner more likely to address the concerns of interested and affected parties in ways that earned their trust and confidence.”  And that is what NAS set out to do.

Even though the NAS report ultimately concluded that “no differences have been found that implicate a higher risk to human health safety” (page 19), the presentation was nuanced, and in places the committee called out gaps in the evidence and raised questions for future consideration.  Consequently the committee took flak for not going far enough — for a report that contained “far more equivocation than the data justify.”  And some blamed this on the public’s involvement.  “Science is not democratic,” critics wrote.  “The citizenry do not get to vote on whether a whale is a mammal or a fish, the temperature at which water boils, or whether the number “pi” should be rounded off.”

Into this frying pan jumped Edward Rubin (Vanderbilt) and Joanna Sax (California Western), who tackle FDA’s use of guidance to regulate genetically modified food in Administrative Guidance and Genetically Modified Food.

Let me confess right off the bat: I love this article, which was published this year in the Arizona Law Review.  It’s a robust piece of administrative law scholarship combined with a meticulously researched piece of food law scholarship, and it is easy on the eyes and brain (clear as a bell).

As Professors Rubin and Sax point out, guidance documents provide a way for agencies to “regulate” without exposure to the crucible of notice and comment procedures, let alone judicial review.  As an administrative law point, the use of guidance (without notice and comment) is acceptable if the rule in question is interpretive and lacks binding force.  FDA has been faced with a conundrum: there is substantial scientific consensus on the safety of GM food, and yet the public is deeply skeptical; Rubin and Sax note that many do not agree with the conclusions scientists have reached and do not understand that these conclusions represent a consensus.  This would, undoubtedly, make rulemaking an extremely contentious process.  Instead, FDA regulates through guidance documents that lay out a “voluntary consultation process” for GM foods.  This, in turn, Rubin and Sax argue, is a de facto premarket approval process — even though the scientific community views these foods as safe, and foods ordinarily do not go through preapproval.

The meat is in section III (“Guidance as a Means of Regulation”), which argues that the distinction between interpretive rules and legislative rules doesn’t work.  First, it doesn’t work to focus on whether the agency is actually interpreting because, after all, even legislative rules must interpret the statute they implement.  And second, there is officially binding . . . and there is binding anyway.  As they write, “in legal doctrine or a law school classroom, there may be a great difference between disobeying and annoying a government agency, but in our real, regulated world, that difference might not be as evident.”

Rubin and Sax argue for a different approach.  You will need to read section III to absorb their justification for the approach (which satisfies “both the instrumental and normative goals of notice and comment procedure, and the countervailing goal of rulemaking flexibility”) but it boils down to asking: (1) is this an agency pronouncement to the general public, (2) would requiring notice and comment provide the agency with information that might otherwise be unavailable to or ignored by the agency, and (3) does the rule “activate” people’s desire for participation?  (The latter will of course be true if the topic is controversial.)  If the answer is yes, then informal notice-and-comment rulemaking should occur.  Needless to say, this test leads Rubin and Sax to conclude that FDA should have addressed GM food in regulations rather than in guidance.

Bottom line:  highly recommended, including the hilarious digression (page 31) on the etymological origins of “guidance” as a noun that takes the indefinite article.