The deadline for comments on the U.S. Food and Drug Administration’s Produce Safety Rule is Nov. 15, 2013. This was probably supposed to be the easiest and most straightforward of the rules to implement the Food Safety Modernization Act (FSMA). It may be that FDA fell into the trap of a well-worn groove in designing the Produce Rule. There are precedents for their approach from both Democratic and Republican administrations. However, this is the fourth time that manufacturers, processors and larger handlers of produce have attempted to use, or successfully used, food-safety legislation or state power to achieve multiple economic agendas – even when this would seem unnecessary, given their tremendous economic strength in the marketplace. In commenting on one of those attempts, I wrote: “It doesn’t take much time or experience to realize that Democrats act like they have never met a regulation that they didn’t like, and, if the regulation is ineffective, dysfunctional and complicated, they may like it even more. The Republicans claim they have never met a regulation they liked, except in practice, when it is demanded by an industry to give a gloss of procedural protection (biotech for example) or provides a barrier to competition protecting the largest companies in an industry from competition.” My view in this article is: a plague on both their corporatist houses. Different strands were contributed for different political reasons, but they were woven into a political, not a science-based, approach, which said that farmers are ultimately responsible for all produce safety problems; therefore, food safety is achieved by telling farmers what to do. Furthermore, that a farm can be treated in exactly the same way as an individual closed-system manufacturing plant, rather than recognizing the reality of farms being open to, and part of, both a larger polluted environment and a natural ecosystem – and often cooperating with other farms in a social agricultural community, or human ecosystem, if you like. It’s like blaming shellfish harvesters for the pollution entering Puget Sound, as well as the trucker who did not maintain a cold chain, the chain of stores that improperly stored the shellfish, and the ill restaurant worker who accidentally contaminated a seafood salad bar. What is more appropriate for regulating shellfish seems to be discourage  harvesting during the warmer months (those not spelled with Rs) , or when public health agencies detect concentrations of human pathogens. The three previous attempts to regulate produce, in this manner, were: (1) Use state power. This includes (a) the Leafy Green Marketing Agreements (first California, then Arizona) under the state equivalents of the 1937 Agricultural Marketing Agreement Act, creating industry-run state agencies under the Departments of Agriculture authorities; (b) the use of the federal law for the single-state (California) production of almonds under a mandatory Marketing Order, with a rule issued by the USDA’s Agricultural Marketing Service, and (c) changing a state law to incorporate food-safety practices for a commodity, which the Florida tomato industry did in 2008.

(2) Incorporate a national food-safety enabling section for produce marketing agreements and orders in the Farm Bill. (3) Create a national Federal Leafy Greens Marketing Agreement (or even Order), which is dormant but still pending. We are now on version (4), achieving similar goals under FDA’s proposed Produce Rules as an interpretation of FSMA. I was often quite naive about some of this. When the California Leafy Green Handlers Marketing Agreement was being implemented, the organizing procedures were being run by the chief counsel for the California Department of Food and Agriculture. He was a wonderful man who could have come straight from central casting and had to shepherd food corporation vice presidents into acting like a state agency. When I was trying to challenge him about the misuse and potentially illegitimate use of a progressive-era law to enable processors and handlers to regulate farmers (and not themselves) on food safety, he cut me off before I could finish the question because he thought he knew where I was going and said, “It’s not a problem. They are exempt from anti-trust under the specific language of the Act.” I hadn’t even considered anti-trust issues, but, sure enough, both the federal and state versions of the 1937 Acts (as amended) specifically exempt Market Orders and Market Agreements from anti-trust legislation and actions. What a deal! I found the food-safety theater aspects of FDA’s proposed Produce Rule on irrigation water quality, using indicator bacteria, to be an abysmal misuse of regulatory authority that would fail to enhance food safety except by accident, but would wipe out farms or eliminate major crops as options to be grown by many farmers. However, similar rules in the state programs also have a genuine legal and functional role. It shows a kind of science-inflected good will, which calms down the large buyers. It does not present the liability problems of having detected actual human pathogens like pathogenic Salmonella, O157:H7 and other pathogens in water (or other inputs) and, having documented this, potentially used the water (or other inputs). It’s better to be regulated by food-safety theater and not document intentional decision-making informed by genuine hazards; these may be dealt with in private. Here is another strand of the rope to hang farmers with:

The so-called Good Agricultural Practices, or GAPs, came in under the Clinton administration. I may be the only commenter who actively dislikes them. They were the soft-sell for “improving food safety on farms” by telling farmers what to do – without quite ordering them to follow other people’s guesses. As checklists of concerns, they can be useful. As specific demands for detailed practices, they can be worse than useless by giving a false sense of security to everyone. GAPs are not guarantees, even if they are treated roughly as guarantees in the marketing of produce. Global GAP Certification is a more rigorous and certified version of voluntary Good Agricultural Practices. A former secretary of the California Department of Food and Agriculture found his family-owned produce operation subject to a Salmonella detection recall (on peppers). The month before, they had been Global GAP Certified. The farm involved in the cantaloupe Listeria 2011 outbreak had received an extremely high rating from their third-party food-safety inspection. Litigation, liability, and now criminal liability adhere to actual contamination of food in commerce. How much of GAPs and certifications and inspections are just more food-safety theater? USDA and private groups did come up with a farm self-assessment food-safety software program, available from the USDA or, that may be more useful to some farmers than GAPs. A science-based question: Are there any studies documenting an improvement of safety after the implementation of the California LGMA? Answer: No. All right, then, are there any studies documenting that farms following GAPs and FDA Guidances improve food safety? Answer: No. FDA Guidances are hybrid animals. They are not regulations; farmers are not required to follow them, but not following them may increase potential liability. It’s not that GAPs and Guidances may not ever help. But the assumption behind assuming they will help is that farmers are not taking food safety seriously in the first place. Furthermore, the farmers who have already adopted following GAPs and Guidances may have done so to document food-safety practices they were already using in order to sell their produce to end-users. The difference in their food-safety concerns and practices before and after becoming GAP-certified may be nonexistent. Their ability to sell their produce, not their ability to implement food safety, may be the only change.

The assumptions behind the economic benefits to be derived from the FDA’s proposed Produce Rule are (approximately): “Farmers are idiots. Following the proposed Produce Rule will force them to improve their food-safety practices over an incompetent base level of inadequate food safety. Therefore, all food-safety practices, including those already used, including actual competent existing food-safety programs, can be counted as an economic benefit solely due to the proposed rule being implemented and to the supposed reduction in illnesses, all of which are accounted for as due to compliance with the Produce Rule.” Checklists are fine. Let’s recall that it was Curtis Lemay and a few other surviving top pilots who developed the aviation checklist so that the experimental planes that became the B-17 could safely be flown. Four of the best pilots in the Army Air Corps had just been killed or maimed in test flights. Safety checklists were developed by expert pilots that could protect every plane and crew before flight. Some 60 years later, similar checklists were developed by some hospitals for preventing respiratory infections in patients at risk after surgery, using the aviation industry as a model. I have never heard of a farm GAP or a produce Guidance whose development follows this model – where the persons with the greatest experience and competence, who have to implement, live with, and work with a safety program, design the program, hitting everything essential for survival. In this case, farmers with demonstrated expertise in both farming and delivering safe produce. There are major corporations that do this and small truck-farms that do this. Farmers have never run the farm food-safety agenda. You would think someone would notice, even as an academic observation. Enter the FDA, writing a Produce Rule implementing FSMA against this background. There seemed to be models, all of which claimed to be science-based produce-safety programs. Along with the models they were using, defective as these may or not be, they picked up some of the baggage that informed those models and came along for the ride. Briefly, this baggage includes: (a) the war on organics (still a cultural factor after all this time); (b) fear of competition from local production; (c) protection of the largest processors and handlers; (d) offloading as much liability and regulation as possible onto produce farmers; (e) disregarding systemic off-farm issues that affect farm safety, particularly urban waste streams and CAFOs. Therefore the proposed rules:

(a) Violate the specific language of FSMA, overriding the Organic Farming Production Act on compost and manure use and disregarding microbial ecological control of pathogenic micro-organisms. (b) Impose regulatory costs that will impact all farmers’ profits, but the largest farms the least harmfully, and, according to FDA’s own analysis, drive many farms out of business with complex facilities rules that will destroy food hubs and CSAs as common practices; (c) Neither regulate fresh-cut under the Produce Rule nor specifically regulate fresh-cut under any other rule (unlike sprout production); (d) Completely misuse food-safety statistics to imply, for example, that farming practices cause human-transmitted norovirus outbreaks or that amoebic contamination only found in foreign imports are due to U.S. farmers; (e) Fail to regulate inputs to farms from CAFOs, urban greenwaste, sewage sludge or ash, etc., according to their health hazards or risk before farm purchase. Some of the worst outbreaks in U.S. history have been the direct result of farmers or processors trying to improve food safety in handling on-farm or processing after delivery of raw produce. The farms and processors involved followed the advice of “the best and the brightest” to make their operations safer and ended up in disasters. Trying to improve food safety led to a food-safety failure. I am alone, perhaps, in finding that acquiring new equipment, including for specific food-safety reasons, or taking over old equipment new to a farmer or processor, is a specific hazard point that has to be controlled for. This can be made more dangerous when a farm or processor is expanding production and under market pressure to meet higher production levels. It seems to me to be obvious in the outbreak record. You might think that this record would result in a certain humility or caution in telling farmers exactly what to do, especially in those areas on which FDA’s expertise is most lacking, from planting to harvest. Yet this is exactly the area that the proposed Produce Rules spend most of their time on, with the greatest negative impact on farming and the fewest potential returns in actual improvement of food safety. The errors in the drafting of the Produce Rule propagate, populate and intensify in the other proposed rules implementing FSMA, including preventive control rules, foreign facilities, foreign imports, intentional contamination, third-party inspections, and others. Who among the congressional drafters of the legislation anticipated that the exceptions and exemptions for U.S. farmers under FSMA, such as the Tester amendment, would apply across borders to foreign producers with the same dollar values and distances? Yet that seems to be the interpretation.

Or that FDA would have no problem with sewage sludge for growing produce, so long as EPA’s complicated calculations on cumulative use are followed, and that this would open the door for importing foreign produce grown with sewage sludge for U.S. consumption (even if not legal for their own consumption)? Or that the failure to specify produce-caused outbreaks shall be a guide to drafting the Produce Rules would open the door to conflating outbreaks caused by farming practices (minimal in the record used), with handling and processing contamination (large), point of consumption (largest) and human-to-human transmission, all under the rubric of “FDA-regulated produce”? Most of the food processors I have met and worked with over the past 30 years are far more negative than I am about FDA’s role. One of my all-time favorite clients, a major West Coast food processor, could never refer to a former FDA commissioner – whom I liked – by his name, only by his full title: “That idiot, David Kessler, who thinks he knows more about running my business than I do.” It goes downhill from there. Actually, way downhill. Most farmers I have worked with are even more caustic about regulators. Some of them voted out the then-Speaker of the House when the Republicans gained control of the House. It goes downhill from there. And this is just the background farm culture. Few of them have had the pleasure of a uniformed FDA inspection by someone trained in regulating medical devices. Therefore, in my view, it is in FDA’s own interest to tear up the proposed Produce Rule and start again, aiming for statistical improvement in food safety by a few clear and simple rules that are not based on food-safety theater or corporatist favoritism, but instead follow FSMA’s direction in being actually science-based, that build on FDA’s existing strengths and competence, and are open to change with new information.

  • Michael Bulger

    If testing for E. Coli indicator bacteria is not justified in Daniel’s eyes, would he rather farms be required to test for pathogen strains? The test for indicator bacteria is available for around $2 a test, and can be done on the farm without the need for a lab, intensive labor, or expertise. Testing for specific pathogen strains would be hugely expensive in comparison. That appears to be FDA’s reasoning that led to the proposed rule’s requirement that water contacting ready-to-eat products be at least clean enough to swim in.

    • Daniel B. Cohen

      “Safe enough to swim in, safe enough to irrigate with” is a great rationale. I first saw it in an Arizona Extension pamphlet. If we were looking for the viral contaminants that cause gastric illness in ocean and freshwater swimmers, it might make some sense. Maybe not — because the recreational swimming standard was based on actual viral illnesses and actual fecal indicator counts and their correlation, to set a standard for closing beaches when there would be x % of ill swimmers per hundred swimmers. “X” to be chosen by the State authorities for their decision making. The viral illnesses are not relevant to produce outbreaks, and the numbers of illnesses per 100 that have ever been set by states are totally unacceptable for food safety in produce when bacterial pathogens are the issue.

      Unfortunately indicator bacterial tests, such as generic E. coli, can fail to detect pathogenic bacteria such as O157:H7 when present in produce, or can be high enough to trigger non-use of irrigation water when not present. They are most useful to farmers in a closed established system, such as a well, where once having been cleared for say potable use (or irrigation use) you use them for routine monitoring of system integrity. In this case, however, broader indicators may be more useful than, say, generic E. coli.

      If you are talking about cost to farmers, then you have to look at the cost of losing your farm, bankruptcy from liability and now jail time as an alternative possibility. And farms routinely do ask for salmonella and O157:H7 tests for purchased animal composts, from their suppliers usually. But it is not clear that farmers should be paying for water quality tests rather than, for example, irrigation districts or public health authorities.

      When O104 hit an Arizona farm and caused an outbreak and recall in fresh-cut bagged produce, the root cause investigation found two potential human sources combined with “an unusual weather event.” If irrigation systems routinely monitored their canals, the cost per acre for testing would be miniscule. A more typical contamination event is dredging and other work that disturbs contaminated soil banks. Even monitoring after known disturbances — such as unusual rainfall or dredging would give good safety information, testing for actual pathogens.

      Many fresh-cut processors have pathogenic bacterial (and other microbial) testing requirements and systems that go back to farm inputs such as water and — at least for organic production — compost. The liability to a brand is huge. This is done privately, and its not clear who is paying for the additional, and functional tests.

      If you go to the FDA Produce Rule website at and look up my comment, just read my interview with Will Daniels at EarthBound Farms for an example. Their data alone is equivalent to a thousand root-csuse investigations. Most processors have extensive and unreported information. There was supposed to be a joint federal-private initiative to look at what could be learned from the private data while maintaining confidentiality.

      You can also look at onion association comments in the same FDA Docket. Do you really think that taking 3/4 of a million acres out of dry bulb onion production in one region alone (Magic Valley, the lower Snake River)due to an uncorrelated generic E. coli test is cheap? And that the onions from other countries that replace it will be safer?

      Cattle production now involves testing for 6 STEC E. coli and the demand for this routine daily testing has reduced costs for this panel of tests. On the other hand the proposed rule will only allow certified labs using older epidemiological methods. So costs will be higher than you think, for generic E. coli, for weaker data.

      We’re looking at very cheap genomic testing that could not only detect pathogens but also “fingerprint” them in a way that the ultimate source of contamination can be found, and possibly be held liable in turn.

      I hope this helps.

  • Michael Patoka

    Mr. Cohen suggests that the FDA improperly puffed up the estimated benefits of the produce rule by trying to connect human-transmitted norovirus outbreaks to on-farm contamination of produce, and thus implying that safer farm practices under the rule would eliminate a large number of norovirus illnesses.

    But that is simply not true: the FDA explicitly says that its estimated benefits would *not* capture any illnesses due to norovirus or other pathogens that are potentially under-represented in the
    outbreak database, even if some small portion of those illnesses may in fact be linked to produce contamination on the farm. (p. 85 of the agency’s economic analysis) If anything, it’s a conservative assumption on the part of the agency, not an attempt to exaggerate the rule’s benefits or a “misuse” of food safety statistics.

    In his official comment, Mr. Cohen points to a table on p. 371 of the FDA’s analysis to show that the agency included norovirus in its calculations, but that table encompasses all FDA-regulated
    food products, including seafood, processed foods, eggs, etc. It was not in any way implying that *produce* was responsible for a large number of norovirus cases. The FDA only attributes five known pathogens to produce contamination in its analysis (Cyclospora, E. Coli O157:H7, Hepatitis A, Salmonella, and Shigella), as it makes clear on pp. 63-65.

    FDA’s economic analysis:

    • Kevin

      When the FDA effectively regulates (eliminates) pesticides, herbicides, and GMO crops in the food supply is when I will believe the FDA is anything but the small farm/business oppression department of giant agribusiness, with no redeeming social value. Does the FDA have cost/benefit analysis of pesticides versus mentally impaired and ADHD children? Does the FDA have data on nutritional deficiencies due to glyphosates in food plants? Does the FDA have cost/benefit analysis on elimination of pollinator insects? Mr. Patoka’s picayune criticism is a perfect example of misdirection and obfuscation as he attempts to discredit the article with a pointless debate on obscure and meaningless statistics. The relevant facts, which Mr. Patoka ignores, would be the rise of competition from small/family farms and the agribusiness response of eliminating the competition through “safety” regulations. No amount of drama and faux concern over a few inevitable sickness cases will hide the truth or justify the oppression carried out by the FDA. Clear identification of the source of food products is the only regulation that is actually required to assure safety, but this is not convenient for giant agribusiness. Thus, we have this Rube Golberg exercise whose only practical result is reduced food safety through elimination of alternatives to giant agribusiness products. The stupidity of the regulations and their obvious purpose is well explained in the article.

  • Michael Patoka

    I appreciate your response. Still, I’d like to point out: the calculation that was responsible for bringing the estimate up to 2.3 million annual illnesses was done by (1) estimating the percentage of all CDC-reported illnesses due to known pathogens that were linked to produce during that time period, and then (2) applying that same percentage to a CDC estimate for the number of illnesses due to *unidentified* pathogens (those that are just emerging, or difficult to identify through testing), in an attempt to account for this epidemiologically significant category of illnesses, which are likely much more prevalent than those due to identified pathogens. (pp. 63-65 of the analysis)

    Some may take issue with this calculation for any number of reasons–all of these estimations are affected by significant uncertainty, as the FDA acknowledges–but it does not incorporate norovirus statistics, or suggest that they are linked to on-farm contamination of produce. Any norovirus illnesses reported to the CDC (along with cases of seafood poison, vibrio, and other non-produce-related pathogens that appear on p. 371) would be in the denominator, as part of the total number of illnesses *out of which* the FDA tries to derive what portion was linked to produce.

    For example, on p. 63, the FDA concludes that raw produce other than sprouts was responsible for 5.41% of illnesses, after dividing produce-related cases (4,293) by all CDC “Identified Cases” (79,347–this is the number that encompasses *all* kinds of foodborne illnesses, including norovirus and pathogens due to meat and seafood).

    That is why I think it’s not accurate to suggest that FDA puffed up the benefits by incorporating norovirus statistics.

    • Daniel B. Cohen

      Michael Patouka deserves more details on this, which follows. It may interest others.

      Please see: Appendix A to the Analysis of Economic Impact, Table 138 “Estimated Number of Illnesses Attributable to FDA-Regulated Foods” page 371. First column “Agent”, in alphabetical order “norovirus”.

      Reading across from “norovirus”

      Second column: FDA cases (2003-2008) 119.

      Third column “CDC Identified cases” 24,570. Fourth column “Percentage attributed to FDA produce” 0,50%. Fifth Column “Estimated Annual Foodbourne Illnesses” 5,461,731. Sixth Column “Estimated Illnesses Attributable to FDA Produce (due to norovirus): 26,453.

      This is summed and included in the TOTAL: 5,051,481.

      Yet the head epidemiologist statistician at CDC confirmed to me there are no known norovirus illnesses due to produce contaminated on-farm. The presence of these norovirus illnesses in the final calculation was a red flag to me to me that their methodology was flawed.

      The second footnote to the table explains the unidentified illness calculation, where there are no FDA cases or CDC identified cases. They estimate 13.16% as “Attributable to FDA produce” and multiply by 38,392,704 “Estimated Annual Illnesses” to get 5,051,481 unidentified estimated illnesses attributable to FDA . That’s out of the 5,741,212 total.

      I would argue that this is a significantly worse statistical move than including norovirus — which is trivial in comparison but still a red flag — 88% of the total illnesses used are a guestimated percentage of a guestimated “unidentified” category which includes both etiology/pathogen unknown and unreported illnesses estimates combined.

      This move is especially unjustified in the context to which FDA is using it: to justify regulating farmers on their on-farm practices from planting to harvest.

      Neither the Economic Analysis nor the Qualitative Assessment of Risk documents were peer reviewed before the produce rule regulations were drawn up. This will happen AFTER the close of the comment period on the proposed rule November 15th 2013.

      • Michael Patoka

        I really do appreciate all your explanations, Mr. Cohen, especially since this is a very difficult and technical subject to discuss.

        At the risk of belaboring the point (which I’m sure we already have!), I’ll just say one more thing that may help clarify things: the last column in that table is labeled “Estimated Illnesses Attributable to FDA __Products__”, not “FDA Produce” as you suggested in your last reply. So, the 26,453 figure you mentioned (estimated illnesses due to norovirus) refers to *all* FDA-regulated food products, which extend far beyond produce. This is reflected both in the title of the table, and in the repeated references to “FDA-regulated foods” or “FDA products” in the surrounding text. After all, “seafood poison,” “scombroid” and “vibrio” are also among the pathogens listed in this table, and the FDA is certainly not implying that these are related to raw produce. As you can see on pp. 63-65, where the specific pathogens relevant to produce are identified, norovirus is not among them (nor are these others). That’s the only point I’m trying to make.

        • Daniel B. Cohen

          Different context though. I started with the “FDA has estimated 2.3 million illnesses due to on-farm contaminaton of produce” and chased down the references in the references to Appendix A; then followed the numbers back up the chain to the prologue and the QAR descriptions. Its actually true that “the FDA has estimated…” etc but it wasn’t a good estimate and it came from the tables in Appendix A to the econ report. So your point is well taken; when they want to they do talk about actual produce-associated illnesses. I agree. Even then they have no methodology for discriminating in the CDC data between planting-to-harvest, later steps on farm, processing or any other point on the delivery chain up until restaurants and facilities. The CDC only recently started to even ask reporting doctors or hospitals “do you know where the produce was contaminated”> And usually the answer is either left blank or they say “no”. You have good analytic observations; keep on the food safety issues if you can and have time. It will be interesting to see the peer reviews that were promised, if they ever come out.

  • Joanna Reuter

    Well said, thank you for this. As a CSA farmer, I wholeheartedly agree that the proposed produce rule is garbage and needs to be scrapped. Since I don’t actually expect that to happen, I’m at least hoping for an increase in the gross income level for the full exemption. I’m glad to see that your analysis points out many of the absurdities in the rule that I also honed in on; at times I thought I had lost my mind when I realized what the rule was actually saying. And thank you for pointing out the absurdity of placing all of the blame on the farmer.

    My comment on the rule isn’t yet showing up on, but it is reposted on our farm blog.

  • Nick

    This is classic FDA: create more regulations that harm small businesses in favor of the big giants. When are they going to learn that the best way for the federal government to help us is to get out of the way?