The Lettuce Has a Passport Problem

The nationwide cyclospora outbreak linked to iceberg lettuce is a reminder that the American food supply chain is less a chain than a vast, refrigerated diplomatic arrangement. FDA officials said they were confident that a Taylor Farms farm in Guanajuato, Mexico, was the source; the July 17 recall covered thousands of packaged salads containing iceberg lettuce and other vegetables, distributed through major restaurant chains. Fortune’s report is worth reading not because it supplies a tidy culprit, but because it shows how little tidy certainty an investigation can honestly offer.

The key distinction is unglamorous and essential: a product recall, a confidence assessment, and a confirmed positive sample are not interchangeable things. FDA later retracted a faulty July result that had purported to detect cyclospora in a border-collected Taylor Farms lettuce sample. Mexican authorities, meanwhile, said they had found no positive samples or other red flags at the farm. The FDA investigation was still ongoing. This is what careful food-safety reporting looks like: evidence moving forward, then occasionally backing over its own shoelaces.

For consumers, the practical message is neither eat nothing green nor the recall is ancient history, pass the Caesar. Check recall notices when one is active, follow instructions for affected products, and avoid turning one outbreak into a theory about every bag of lettuce. Cyclospora illness is serious enough to merit attention; it does not require a nationwide audition for panic. The larger question is how the system identifies trouble before a recall has to become dinner-table news.

FDA inspectors reached the Guanajuato site in mid-August—about a month after the recall announcement and, according to the report, for the first time since 2019. That timing does not establish that an earlier inspection would have prevented this outbreak. Inspections are not force fields, and the article explicitly does not show that reduced inspections caused the contamination. But it does expose a mismatch between the scale of imported food and the government’s ability to inspect the places producing it.

Foreign-food-site inspections were nearly 35% below their 2019 level, and fell 17% to 1,140 in fiscal 2025 from the preceding year. The 2011 Food Safety Modernization Act calls for inspection of more than 19,000 foreign sites annually. FDA’s historical peak was 1,700, itself less than a tenth of that statutory ambition, according to the Government Accountability Office. As former FDA official Michael Rogers put it, “the agency just can’t achieve what is required”.

That sentence needs a footnote the size of a produce crate. Former regulators characterize the statutory goal as unachievable and inadequately funded; GAO has said FDA has not identified a suitable annual target. The inspection workforce has more than 420 people, with 10% to 15% of positions chronically vacant, while international inspectors may spend half or more of their working time traveling. Then more than half the staff handling travel bookings were eliminated, leaving inspectors to do more of that logistical work themselves. Somewhere, an outbreak-prevention strategy is trying to expense a flight while waiting on hold.

The administrative lesson here is not that every imported leaf needs a federal chaperone. FDA itself argues that inspections are only one tool among border checks, importer oversight, sampling, remote monitoring, and other forms of surveillance. Given the volume of food crossing into the United States, that is plainly true. A system designed around occasional, labor-intensive visits cannot by itself observe a continental supply network in real time.

But the alternative tools need to exist in working order, rather than as a PowerPoint ecosystem. The produce traceability bar-code rule envisioned under the 2011 law was due to take effect in January and was delayed again after industry pushback. The administration has also proposed a pilot of one-day inspections, though officials cited in the report say these are not equivalent to conventional inspections, which may take a week or longer. Speed can be useful; speed without enough time to examine records, practices, and conditions is mainly a calendar achievement.

More than 480 food recalls had been announced in the current fiscal year, exceeding four of the previous five years, according to FDA figures cited by Fortune. That number does not prove the food system is becoming less safe; recalls can also reflect detection and disclosure. Still, a useful food-safety system should be able to answer basic questions quickly: where did this lettuce travel, what else shared its path, and who needs to know? Parasites are indifferent to jurisdiction, budget cycles, and regulatory postponements. The rest of us have chosen to make those things relevant.