This is the story of how a jar of peanut butter became evidence in a federal investigation—and how the owner of a food-processing company received a sentence comparable to one for murder.
When people hear “peanut butter,” most picture a labeled jar on a supermarket shelf. The very thing that starts a school sandwich. But Peanut Corporation of America—PCA—hardly sold any peanut butter at retail under its own name. Its business worked differently: a plant in Blakely, Georgia, produced peanut butter, peanut granules, and peanut ingredients that then disappeared deeper into the food chain. These semi-finished products went into cookies, crackers, energy bars, candy, ice cream—products no end consumer would ever connect to a small factory in southern Georgia. The affected brands included Keebler, Austin, King Nut, and hundreds of others. In all: 3913 products. PCA was not the face on the package but the pipe through which peanuts flowed into the national food system, and that is precisely why contamination at a single facility managed to spread so far.
The scale of that branching network was not immediately apparent. The first signs of an unusual increase in salmonellosis cases appeared in the fall of 2008. Epidemiologists at the CDC began piecing the chain together: the strain of Salmonella Typhimurium found in patients across different states was genetically identical. The first clustered cases dated back to September 2008. By April 2009, when the outbreak began to subside, it had spread across 46 states, the number of confirmed cases had reached 714 people, and nine deaths were linked to the outbreak. The median age of those affected—16 years—means the disease hit children and teenagers especially hard: the people most likely to eat peanut crackers and snack bars.
The FDA inspection at the Blakely plant ran from January 9 to 27, 2009. The facility manager at the time was Sammy Lightsey. What the inspectors documented reads like a guide to “everything you must never do in food production,” followed point by point.
The roof leaked: gaps up to a foot long gaped in it, and rainwater fell directly into the packaging area, where shipment-ready batches of nuts stood beneath them. There were holes in the walls large enough for rodents to get inside; mold grew on the ceilings and walls. Inspectors recorded one live cockroach and several dead ones in a washroom cluttered with equipment. Raw peanuts were stored beside products ready for shipment—without a physical barrier, without separated flows—and at a glance it was already hard to tell one from the other. The roasting process, which the plant considered its pathogen-kill step, had never been validated, while temperature records were incomplete: several circular charts did not even have dates on them.
But the central piece of evidence was not the cockroaches or the leaks. The central evidence was on paper. PCA’s internal laboratory tests repeatedly detected Salmonella in its products—not only the Typhimurium strain, but Anatum as well. After a positive result, the company did not stop the line or disinfect it. Instead, it could order another test—from a different laboratory or on another batch—and, once it got a negative result, ship the product to customers. Put bluntly, PCA played the lottery with microbiological tests until it drew the answer it wanted. This was not negligence in the ordinary sense, and not “someone forgot to wash their hands.” This was a system in which a positive test for a deadly pathogen did not cancel a shipment—it was an inconvenience to be worked around through retesting.
What still gets me is that detail: not the absence of testing, but its presence—and the conscious disregard of the results. PCA knew. And shipped anyway.
Here, it is necessary to understand the legal reality in which the FDA existed before 2011. The agency had the right to inspect food facilities. It had the right to respond to violations—to issue warnings, initiate recalls, refer cases to court. But it did not have the authority that seems self-evident: to require food producers to draw up and follow a formal preventive risk-control plan in advance.
The difference is fundamental. The FDA operated on a response model: an outbreak occurs—an investigation begins—the source is identified—action is taken. A plant could operate for years without a federal inspection if nothing triggered one. State inspections were conducted by the Georgia Department of Agriculture, but their scope and depth did not allow them to uncover what the FDA later found. The system resembled a fire department without a fire alarm: it knew how to put out fires, but only learned of one after smoke was already pouring from the windows.
PCA exploited that blind spot. The company was not required to present a federal regulator with a plan for preventing microbiological risks; it was not required to document corrective actions after pathogens were detected; it was not required to prove that its processes killed Salmonella before the product left the plant gates. All of this sounds wild—and sounded wild even then—but that was the law. The core food law on which the FDA relied had been adopted back in 1938 and had been patched up ever since, but never rebuilt. It assumed that the manufacturer was responsible for safety and that the regulator checked after the fact. When a manufacturer turned out to be unscrupulous—and PCA was exactly such a case—the system discovered it too late.
The recall of PCA products began on January 13, 2009, and became one of the largest in the history of the American food industry. Thousands of products were pulled from shelves across the country. Twenty days later, when FDA inspectors made public Form 483, the recall had to expand to all products from the plant—including those manufactured two years before the outbreak. Companies that used PCA ingredients were forced to recall their own products, even if they themselves had done nothing wrong. The reputational and financial damage spread in ripples, like a stone thrown into a pond—except the pond turned out to be the size of the national market.
The criminal prosecution of PCA owner Stewart Parnell became a precedent the food industry had never seen. Federal prosecutors charged him with offenses connected to fraud and the shipment of products known to be contaminated. The key evidence was those same internal documents: test results, emails, shipping records. The papers showed that PCA management knew Salmonella was present and chose to send out products on the basis of repeat negative tests—even though the lines had not been cleaned between tests.
In September 2014, a jury found Parnell guilty on 71 counts. In September 2015, the judge handed down the sentence: 28 years in federal prison. It was the harshest sentence ever imposed in the United States in a food-safety case. Parnell was sixty-nine, and twenty-eight years was, in effect, a life sentence. Other company managers, including the owner’s brother, also received prison terms. PCA went bankrupt and ceased to exist.
I am not a lawyer, and I will not pretend to understand every nuance of the sentence. But the mere fact that the head of a food company received a term comparable to sentences for violent crimes says something about how the court assessed the degree of intent. This was not an accident. This was not ignorance. It was a choice—to ship a product knowing it might be contaminated, for profit and delivery deadlines.
The Food Safety Modernization Act—FSMA—was signed by President Obama on January 4, 2011. It became the most sweeping reform of American food law in more than seventy years. The law rewrote the FDA’s operating logic: from “respond to an outbreak” to “require preventive controls.”
What exactly changed? FSMA required most food facilities to develop and carry out written risk-control plans—documents in which a producer must identify potential hazards (microbiological, chemical, physical), define critical control points, and set out corrective actions. The FDA gained the authority to demand these plans, verify that they were being followed, and—for the first time—the right to mandate a product recall if a company refused to do so voluntarily. Before FSMA, a recall was formally a voluntary act by the manufacturer; the FDA could pressure, persuade, threaten litigation, but it could not order one.
It would be inaccurate to say that FSMA was the direct and sole consequence of the PCA case. The law had been in the works for years, advanced by different forces, and shaped by other outbreaks and scandals. But the PCA story became the case that turned an abstract debate about reform into political inevitability. Members of Congress cited testimony from relatives of those affected. Documents from Blakely appeared at hearings. The peanut disaster gave reform a face—and that face was frightening enough to get the law through both chambers of Congress.
FSMA does not guarantee that outbreaks will stop. No law can do that—the food system is too vast, too complex, too dependent on human beings at every stage. But the law changed where effort is applied: the regulator is now required to look at what a manufacturer does before a product reaches the shelf, and the manufacturer is required to prove it controls the risks rather than merely hoping for the best. The difference between “show us that you are preventing contamination” and “we will inspect after someone gets sick” is a difference paid for by 714 sick people and nine dead, whose names in this story remained in the shadow of Stewart Parnell’s.