Our piece, Without Clear Proof, Industry Suffers From Mango Recall And Is Left To Defend Itself, featured a letter from Dave Westendorf of Bay Area Produce, San Clemente, California. He is trying very hard to look at the Splendid mango situation and find solutions that might help the industry in the future. He was kind enough to share his thoughts with us once more:
Thank you very much for publishing my letter regarding the Splendid mango recall and for your thoughtful response. I have had many people call or email their thoughts, and I’m coming away with several observations:
1. Given the fact that foodborne illnesses are extremely hard to trace and that our food safety agencies’ main responsibilities lie in protecting the public, some insurance mechanism must be made available to companies like Splendid to cover their losses resulting from a recall. Product liability insurance is readily available, yet in most, if not all situations, it does not cover the expense of actually taking the product out of the supply chain. Perhaps United Fresh or Western Growers could underwrite such policies. While I’m not a fan of expanding our government, how about some mechanism through PACA? After all, the USDA subsidizes crop insurance for farmers. Why not recall insurance for produce distributors?
2. Health agencies should not be allowed to publicly implicate a specific grower’s product without definitive proof of contamination. Nothing more than a general warning should be issued, and even then, other possibilities should be thoroughly researched and eliminated before a product is implicated. Early in this investigation, a California health official told me that California was only following Canada’s lead and could not tie the California outbreak to any specific mango label. Who is to say that the 125 people who were infected with Salmonella Braenderup consumed it in a mango salsa containing items like green onions, cilantro, chili peppers, or orange juice? After all, those who will remember the Salmonella Saint Paul tomato contamination will recall that tomatoes were never the carrier, and the ultimate source was eventually proven to be chili peppers. The collateral damage done to the tomato industry was incalculable, yet tomatoes were never the culprit. Perhaps the same thing will be proven in the mango situation, albeit a little too late for Splendid’s sake.
3. Sad as it may be, Splendid might not be facing the losses that they are currently incurring if they had not issued a ‘voluntary’ recall and had waited until a recall was mandated by one of the state or federal health agencies. Splendid did the morally right thing but will pay dearly for its integrity and moral code. Some other mango shippers have proudly announced that their product is not contaminated and therefore will not be recalled. Splendid could have also pointed to every independent analysis of their product to date and said the same thing. Companies need to think twice about issuing a voluntary recall unless their product is specifically proven to be contaminated with the suspect bacteria.
4. All companies involved in the distribution of fresh produce need to look at their individual business models to determine where changes need to be made to protect themselves from a Splendid Products-type of situation. Grower and retail supply contracts need to be modified to account for recall liability and placing it where it belongs, at the proven source of contamination whether that be the grower, distributor, wholesaler, purveyor, or retail level.
5. In discussing this recall situation about Splendid, a shipper friend of mine suggested that perhaps there should be a protocol implemented by the FDA/CDC whereby when a product like mangos or tomatoes is suspected in a salmonella/listeria/etc., contamination that a ‘stop sell’ order be issued where one or more shipper/grower(s) who ‘may be’ a source of the contamination would quit selling, advise their customers to also quit and hold the product until testing can confirm or deny contamination. When the contamination or lack thereof, becomes certain, then either sales could resume, or a recall notice would be implemented.
I appreciate the work that the health agencies do, however when a contamination source is not verified and a shipper/grower is forced into a ‘voluntary’ recall, he suffers serious financial penalties through possibly no fault of his own. If no contamination is ultimately proven, or another source is found, the shipper/grower is left holding the bag with no recourse against the government agency that falsely accused him of handling contaminated product.
I’m sure we haven’t seen the end of this saga, but hopefully lessons will be learned and some good things might come from it.
Bay Area Produce
San Clemente, California
We certainly feel the angst that Dave is expressing and appreciate his efforts to think through to a solution. We think he raises points well worth thinking about:
1) Insurance on Recalls
The need for this is obvious, and efforts have been made to get private companies to do this and to see if the government would get involved. Recall insurance is sold. But in many cases, it is inadequate — covering expenses such as stamps to mail out recall notices but not fees charged by retailers to remove product from the shelves — or triggered by events, say the finding of a pathogen on a product that often don’t apply. For example in the spinach crisis, the whole industry had to deal with a mass recall — even though no pathogen was ever found on the product of most producers. Since recall insurance is often not triggered by a recall but by a condition such as the finding of a pathogen, recall insurance, even when one has it, often won’t pay. One can piece together policies that seem to provide the proper coverage, but they are not as ironclad as one would like and they are expensive.
It is a tough thing to make happen. From a public policy point of view, what you don’t want to do is create a situation where companies want to wait for a government-ordered recall. Yet, if you don’t do that, then you are insuring against a “choice” that a company makes, and this creates the fear that if the cost of recalls can be shifted to a third party, there will be a lot more recalls.
For that matter, even if such insurance only worked in the event of a government-ordered recall, what constrains the government from ordering recalls but aggressive efforts by the companies implicated to prove the recall unjustified and the knowledge of government officials that their acts can crush companies and cost jobs. If these were no longer concerns, government officials might well order many more recalls and companies might well acquiesce. Also many of these costs are retail dictates; as it is, many shippers think them excessive. Well, what would hold down these costs if retailers knew an insurance company was going to pay them?
On the government side, the thought is that one who suffers from bad weather is blameless. So flood insurance or insurance against crop failures is acceptable. Neither the government nor the citizenry are at the point of yet saying that pathogen distribution is also without blame. Correctly or incorrectly, officials look at situations such as the Jensen Farms Cantaloupe matter and see the producer at fault. They don’t want to subsidize his recalls; they want to make a recall as threatening as possible to his business so he will be incented to do everything possible to avoid a business-threatening disaster such as a recall.
2) Ban on implicating specific grower without proof
Here the industry needs to be careful about what it wishes for — it just might get it. The primary goal of public health authorities is to minimize illness among the population. If it believes a foodborne pathogen is threatening the population, the public health authorities will, of course, want to warn the population. This is pretty reasonable. People would be rather upset if the public health authorities knew of a threat and didn’t identify it. So the question then becomes: Should the FDA make a general warning — “Don’t eat spinach” — or a specific warning — “Don’t eat a specific brand or lot.”
There is little question that the interests of the industry lie in making the announcements as specific as possible. This was the lesson of the spinach crisis and, during the salmonella Saintpaul crisis, the big effort was to continuously get the FDA to limit its area of interest.
There exists some question as to the effectiveness of this. During the Jensen farms cantaloupe incident, sales collapsed for cantaloupe in general despite the FDA being very specific in its announcements. Still, it seems logical that more limited announcements are better for the industry than broader announcements. Indeed the entire Produce Traceability Initiative was founded in this belief. It was Bruce Peterson of Wal-Mart fame that noted that every day at Wal-Mart there were hundreds of recalls, but most were uneventful because the manufacturer quickly supplied lot numbers that were to be recalled and this constrained the impact of the recall.
If anything the goal should be that instead of implicating Splendid, we could define a series of lot numbers that meaningfully distinguish between lots and just recall those lots.
We have addressed the issue of lots in pieces such as this and this. The problem is that if lot numbers simply signify some quantity run on a belt, they are meaningless for food safety purposes. We need to know that lot 100 is differentiated from lot 99 and lot 101 by some procedure of significance to food safety — say a sanitization of the line. Otherwise being able to identify problems with specific lots is of little value. It is significant that when Splendid Products issued its recall, it did not list specific lot numbers, just PLU numbers. In fact, it is not clear why it listed PLU numbers at all since they are not specific to the mangos being recalled.
Other than not having pathogens to begin with, the ability to identify breaking points in production or processing so that we can say that lots differ meaningfully is probably the best way to limit the impact of recalls.
3) Limiting Voluntary Recalls
From a business perspective, delaying might well save money — although only at the season end or if the lots are meaningful and one can say that the new product is definitively not contaminated.
If the season is ongoing and the lots not meaningful, one would just have to recall the presently available product.
Of course, if there is a real risk, then failing to recall could have financial consequences when one gets sued because someone dies. It could have reputational consequences if retailers think you are the company that won’t do the right thing and, of course, if someone gets really sick or dies, the executives who made this call have to live the rest of their lives knowing they may have caused a severe illness or death.
4) Supply Contract Modification
There is little question that everyone should review their business and consider what would happen in the event of a food safety problem. Although having liability placed at “the proven source of contamination” sounds like a great idea — in the vast, vast, vast majority of cases the cause of contamination is never “proven” — plus such clauses would likely lead to endless litigation as people tried to get a court to declare who was at fault. Plus, perhaps fault is sometimes shared. A pathogen is always unacceptable but may have less impact if the cold chain is always perfectly maintained. A farm may be the “cause,” but the washing facility may have failed to remove it.
Besides most vendors have little choice but to sign what customers demand and then pray they don’t get called on to have to execute.
5) “Stop Sell” Orders
It is an interesting idea but testing cannot confirm or deny the existence of a pathogen on a wide basis. Otherwise every shipper could do test and hold, and we would never have a food safety outbreak. We dealt with this issue in an interview titled A Closer Look At Finished Product Testing.
The problem is that pathogen contamination is often seemingly random. You can take a spinach field and do a random sample and it will find no pathogen. Do the same field a second time you will find one. Then the third time, it will be clean again.
It is too expensive to do statistically significant testing — and even statistically significant testing is no protection against Black Swan events.
This is all beside the point that logistically doing such testing in thousands of stores, warehouses etc., would be difficult and unfair to consumers and trade buyers who would be left losing shelf life and providing warehousing while waiting for results.
Over the years, we have felt that many things could be done to make it less likely that government would cause damage, include crushing businesses and loss of jobs by erroneously fingering particular businesses or commodities. We also thought that steps could be taken to reduce the impact of even correct advice.
For example, government epidemiologists could have to submit their case before a panel of independent epidemiologists for a review. There have been cases where government asked for a recall and then backed down when persuaded their epidemiology was in error — but few companies have epidemiologists on call to fight such things.
Government epidemiologists could be required to publicly sign such recall orders so their reputation is on the line.
Instead of declaring that there is a risk and a recall is required, or declaring a recommendation not to consume, any risk could be put in perspective. If the FDA was to announce that there is a .00056% chance that one’s fruit or vegetale is contaminated with a pathogen, and a .000021% chance one might get ill from eating this crop, consumers might view such announcements in perspective.
Many thanks to Dave for trying to think through a better way for dealing with such issues.