The FDA Can Pull Almost Any Tainted Food From Shelves — Except the One Babies Live On.

The recall notice went out over a weekend. By the time Nara Organics agreed on June 13 to pull every lot of its whole-milk powdered infant formula from American shelves, three babies in California, Washington and Pennsylvania had already been hospitalized with infant botulism and treated with an antitoxin. None of them died. But the order of events is the part worth sitting with: the illnesses came first, the recall came after, and the recall happened because a company said yes — not because the federal government could make it.

That last fact tends to surprise people, because the conventional wisdom is that the Food and Drug Administration was handed real recall power years ago. It was — for almost everything except the single food a newborn may consume to the exclusion of all else. When Congress built the FDA a mandatory-recall hammer in 2011, it pointedly left infant formula out. The Nara outbreak is the latest reminder of why that carve-out should not survive the next scare.

What actually happened

The federal account is straightforward. The FDA and the Centers for Disease Control and Prevention contacted Nara Organics on Friday, June 12, after epidemiologists tied three cases of infant botulism to babies who had consumed the company’s formula. The agency recommended a recall. The next day, the company agreed to pull all of its Nara Organics Whole Milk Organic Powdered Infant Formula. The three infants had been hospitalized and treated with BabyBIG, the antitoxin used for infant botulism; the illnesses had begun weeks earlier, in April and May.

To Nara’s credit, it moved quickly once contacted, and it pulled product even though, as of the recall, the formula itself had not tested positive for Clostridium botulinum. The link was epidemiological — a pattern in sick babies, not a positive lab result on a can. The company chose the cautious path, which is the right one.

But strip away the cooperation and look at the machinery underneath. The strongest tool the federal government used here was a phone call and a recommendation. The recall was, in the FDA’s own language, voluntary. Had Nara said no, or dragged its feet, the next steps would have been slower and murkier than most parents assume — because the modern recall authority Americans think protects the formula aisle was written to cover almost every other shelf in the grocery store instead.

The hammer Congress built, and the hole it left

For most of the FDA’s history, the agency could not order a food recall at all. It could inspect, warn, seize and publicize, but pulling a dangerous product depended on the manufacturer agreeing to do it. The Food Safety Modernization Act changed that in 2011, giving the FDA the power to compel a recall when a food is contaminated enough to cause serious harm and a company won’t act on its own. That was the hammer — the thing that finally let the government say “off the shelf” and mean it.

Read the FDA’s own guidance and you find the catch: that mandatory authority “applies to foods other than infant formula.” Formula was carved out, left under an older section of the law that predates the modern framework. That provision does obligate a manufacturer to recall when the government determines its product presents a risk to health — the duty is real, not optional. But the streamlined enforcement machinery built in 2011, the expedited order to stop distribution and the fast hearing to back it up, was designed for spinach and peanut butter and ground beef. It was not extended to the one product a baby may eat and drink and live on, with no substitute, for months.

So in practice the government’s first and strongest move on formula is to recommend, and then to count on a yes. Most of the time it gets one, because the legal and reputational exposure of refusing is enormous. The problem is that a safety system you only trust because nobody has tested it is not really a safety system. It is a streak.

The honest counterargument

There is a serious case on the other side, and it deserves a straight hearing. Nara cooperated within a day. Mandatory recall authority is a backstop that the FDA rarely fires even where it has it, precisely because companies almost always comply. And there is a sharper version of the objection: faster recall power would not have saved these three babies anyway. The contamination was caught after the fact, the product never tested positive, and the warning sign was a cluster of sick infants. If detection is the weak link, the argument goes, then arguing about recall authority is fighting the last war.

Take all of that seriously. It is mostly right, and it still misses the point. A backstop you rarely need is worth having for the day a company decides its inventory is worth more than a precaution — and the formula aisle is the last place in the food supply where you want any daylight between “the government has concluded this is dangerous” and “it is gone.” The detection critique actually cuts toward the same conclusion. If the system keeps discovering formula problems only after babies are in the hospital, then the recall step is the one stage left where speed is still possible, and it should not hinge on a corporate signature.

This keeps happening

What makes the carve-out hard to defend is that the pattern is not new. The FDA was investigating a separate cluster of infant botulism tied to formula only last November. And the country has been here before in a far bigger way: the 2022 shutdown of an Abbott plant in Sturgis, Michigan, over contamination concerns helped trigger a nationwide formula shortage that left parents driving across state lines to find anything to feed their children. Each episode ends the same way, with reassurances that the lessons have been learned and the system tightened.

Then the next one arrives. It is worth remembering that this agency is being asked to keep that promise during a stretch of real internal turmoil — a period that saw its commissioner resign after a year of upheaval and a broader argument about whether independent safety judgment is even welcome at the agency, captured in our colleagues’ case that Marty Makary was pushed out for being too independent. A food-safety net is only as strong as the institution holding it, and this one keeps re-stitching the same hole while the building shakes. It is the same reactive posture that let a parasite the U.S. eradicated decades ago resurface in Texas before the response geared up: wait for the problem to announce itself, then scramble.

Why it matters now

The reason formula deserves the strongest rules in the food code, not the weakest, is simple and physical. A parent who hears about contaminated lettuce can buy a different brand at the same store an hour later. A parent whose baby depends on a specific formula cannot improvise a replacement between feedings, and an infant has none of the resilience an adult brings to a foodborne illness. The stakes are categorically higher, which is exactly why the safety posture should be categorically stronger.

The fix is not complicated and it is not about punishing Nara, which did the right thing. Congress should write infant formula back into the modern recall framework — the fast order, the enforceable timeline — and pair it with the kind of proactive testing that catches contamination before it reaches a bottle rather than after it reaches a bloodstream. A recall that arrives after three babies are already in the hospital, on terms the manufacturer is free to set, is not protection. It is a favor. The food a baby cannot live without should not depend on one.