When a trick or treat turns deadly

We are acutely attuned to danger, but that does not mean we accurately evaluate risk.

One of my children says that Halloween is their favourite holiday, even better than Christmas, because it’s all about community — getting out in the local area and meeting your neighbours. Gifts are doled out liberally. Children experience kindness not just from friends and family, but kindness from strangers, too.

But we have been warned that not all strangers are kind. “That plump red apple that Junior gets from a kindly old woman down the block,” cautioned The New York Times in 1970, “may have a razor blade hidden inside.” Newsweek in 1975 went further: “In recent years, several children have died and hundreds have narrowly escaped injury from razor blades, sewing needles and shards of glass purposefully put into their goodies by adults.” And the nationally syndicated Dear Abby column warned in 1983 that, “It’s Halloween time again and time to remind you that . . . somebody’s child will become violently ill or die after eating poisoned candy.”

After reading such alarming warnings, a young sociologist named Joel Best became curious. Best was interested in crime and social problems but puzzled by these stories. “I could not for the life of me figure out what the reason for poisoning Halloween treats might be,” he told me. A razor blade in the plump red apple from the kindly old woman? Perhaps someone in The New York Times had read too many of the Grimm Brothers’ fairy tales.

When Joel Best told his friends he thought these stories might not be real, they were outraged. Of course they were real! So Best started to look for data. How often were children poisoned or maimed by evil strangers peddling treats? That is not easy to say: there is no crime of “Halloween sadism”, so crime statistics wouldn’t track it. Instead, Best looked at newspaper reports through the decades. He figured that whenever a child was seriously hurt by a Halloween treat, that story would be covered in the first few days of November.

There were a couple of odd incidents.

In 1959, a dentist, William Shyne, handed out laxatives to neighbourhood children — who had no interest in eating the unpleasant-tasting pills. Shyne was charged with “outraging public decency”. In 1964, Helen Pfeil of Long Island gave a “treat bag” to a couple of teenagers that turned out to contain steel wool, dog biscuits and a pack of ant poison. This was unwise but Pfeil assumed the teenagers wouldn’t eat dog biscuits or ant poison, and of course they did not.

But if children were dying because of some horror-movie murderousness, surely Best would be able to find a specific named child suffering this fate? He could not. There were lots of reports of scares and pranks, many of which turned out to be hoaxes perpetrated by children on their parents, imitating the crime their parents dreaded, but no serious harm.

There seemed to be one awful exception. In 1974, eight-year-old Timothy O’Bryan collapsed and died at his home near Houston after eating a tube of powdered candy that turned out to have enough cyanide in it to kill three grown men. His five-year-old sister Elizabeth and three other children were also found to be in possession of cyanide-laced candy tubes — thankfully, none of them had yet sampled the deadly treat.

This would have been the stranger-danger nightmare every parent had been warned about, except that the poisoned candy came from Timothy’s father Ronald, who had prepared it and handed it out to his children and their friends. Ronald had taken out insurance on his own son’s life eight days before Halloween and had called to collect his money just hours after Timothy’s death. O’Bryan had believed he’d get away with murdering his own son — after all, kids are poisoned by strangers all the time, aren’t they?

The world is full of threats which loom large, even though they perhaps should not. How should we think about them? Amid a vast academic literature on risk perception, one idea that stands out is “Risks as Feelings”, a 2001 article by psychologists Loewenstein, Weber, Hsee and Welch. They argue that our perception of risk is governed not just by estimates of the likelihood and severity of harm — which themselves may be erroneous — but by factors such as “vividness, immediacy, background mood”. For most of us, risk isn’t a probability — it’s a feeling. Our emotions are inextricably entwined with our evaluation of the dangers we face.

Halloween candy is the perfect example of this process misleading us. The fact that the scenarios seem to have been dreamt up by Brothers Grimm should make them risible, but instead make them compelling. The horror is real, the danger non-existent.

Not every case is so clear-cut. My friends are sending their daughters off to university, and are swapping stories of young women being surreptitiously injected with an incapacitating drug at a bar or nightclub. Predatory men pose many risks to women, but this prospect is particularly unsettling. Still, the logistics of this crime seem formidable; doctors tell me that it would be no easy task to inject someone with a well-calibrated dose of sedative without being detected.

Several academic studies have tried to gather data on the prevalence of “needle spiking”. In a recent article in Policing and Society, sociologists Nicole Westmarland and Melanie McCarry presented the results of freedom of information requests to every police force in the UK. They discovered hundreds of reports from women who complained of being pricked with a sharp object, but no hard evidence that anyone had successfully been injected with a drug.

Another study examined suspected needling cases in Paris. The researchers found 171 cases in which people (usually, but not always, young women) had reported being assaulted with a needle. Physical evidence of a jab was quite common, but no evidence was found of drugs being injected.

This is very patchy data and there are various ways to interpret it. One explanation is that there are people frequenting the nightclubs of Paris who have started spitefully jabbing the arms of young women with whatever sharp object is available, not in order to drug them but in order to frighten them. As with the murderous Ronald O’Bryan, the dreadful story may have inspired the crime.

We are acutely attuned to danger, but that does not mean we accurately evaluate risk. And in a world that is still full of perils, it will never be easy to figure out where true risks lie, and when we are being frightened by shadows.

At least Halloween offers us some clarity. Several children are killed each Halloween by cars, and Halloween is much the riskiest night of the year for such accidents. Compare that with the number of children ever killed because a stranger poisoned their candy: not one.

© 2026 The Financial Times Ltd. All rights reserved. Please do not copy and paste FT articles and redistribute by email or post to the web.

This Financial Times article was legally licensed by AdvisorStream