Selfies Kill More People Than Sharks — Which Tells You Nothing

The Comparison Nobody Can Resist

In 2015, a handful of news outlets noticed that more people had died taking selfies that year than had been killed by sharks. The headline wrote itself. It has been rewritten every year since, usually under a photo of someone leaning over a canyon with a phone at arm’s length.

The underlying research is real. A 2018 study in the Journal of Family Medicine and Primary Care counted 259 selfie deaths worldwide between October 2011 and November 2017. A broader 2022 analysis in the Journal of Travel Medicine found 379 between 2008 and 2021. That works out to 30 to 45 deaths a year depending on which period you pick, and both counts come from news reports, so they miss deaths nobody wrote up. Sharks, by comparison, kill somewhere between five and ten people annually according to the International Shark Attack File.

So selfies really do out-kill sharks, by a factor of five or more. That comparison sounds like it means something. It means almost nothing, because both numbers are being divided by the wrong population.

The Denominator Is Absurd

Back in 2014, Google estimated that Android users alone were uploading 93 million selfies a day. Adjust for iPhones and a decade of growth and a conservative figure today is around 100 million selfies per day, or roughly 35 billion a year.

Call the death toll 50, a round upper figure that leaves room for undercounting. Fifty deaths against 35 billion photos works out to one fatality per 700 million selfies. (The 35 billion is today’s volume; the death counts span years when far fewer selfies were taken, so the true per-selfie rate is a little higher. At this scale it changes nothing.) Nobody makes decisions at that scale. If you took a hundred selfies a day, every day, you would need about nineteen thousand years to expect a single death. By that reckoning the selfie is one of the safest activities ever invented, safer than sitting in a chair.

That conclusion is just as wrong as the shark one, and for the same reason. The 35 billion includes every bathroom mirror, every restaurant table, every airport queue. None of those people are at risk. The deaths come from a tiny, specific slice of the total.

Where People Actually Die

The 2018 study broke its 259 deaths down by cause. Drowning came first, then transport (mostly people standing on or beside railway tracks), then falls from height. Add fires, animals and electrocution and you have covered nearly the whole list. The 2022 study found the same pattern, with falls from height, transport and drowning as the top three causes, and one victim in three a traveller.

The victims were young. Mean age in the 2018 study was 22.9. Just under three quarters were male. And they were concentrated geographically in a way that has little to do with phone ownership: India alone accounted for 159 of the 259 deaths, followed by Russia, the United States and Pakistan. India’s rivers, cliffs and rail network, combined with a large young population and a strong culture of group outings to scenic spots, produce most of the world’s selfie fatalities.

Every one of these deaths shares a feature that the global statistic erases. The person was standing somewhere that would have been dangerous without the phone. A ledge over a waterfall. A rock at the edge of an overlook. A railway line with a train in the background. The selfie did not create the hazard. It added distraction, a step backward, and a reason to be closer to the edge than a person would otherwise stand.

The Chain of Probabilities

To estimate a risk that means anything, the denominator has to shrink to selfies taken at genuinely hazardous locations. Nobody counts those, so this is an estimate: if one in a thousand selfies is taken at a cliff edge, a waterfall, a rooftop, a railway line or fast water, that is around 35 million hazardous selfies a year. Fifty deaths against 35 million gives roughly one death per 700,000 edge selfies.

The calculator splits that into two steps. First, something goes wrong: a slip, a lost footing, a step back too far, a wave or a current. Given how the deaths break down, a rate of about 1.4 mishaps per 100,000 edge selfies (one per 70,000) is consistent with the reported death toll. Second, the mishap kills you rather than injuring you. Falls from height and immersion in fast water are unforgiving, but most slips end in bruises or a broken wrist. A fatality share of one in ten fits the sources.

Multiply those together and each selfie taken at a hazardous spot carries a fatality risk of roughly one in 700,000. Small, but a thousand times the one in 700 million that the naive global number implies.

Your Actual Risk

The two sliders that matter are how many edge selfies you take per visit and how many such visits you make a year. Somebody who takes five photos at a lookout on ten hikes a year is at about 1-in-14,000. Not trivial, but a fraction of the annual risk of driving to the trailhead.

Push the sliders to their maximum, thirty edge selfies on a hundred trips, and the figure climbs to about 1-in-240. That profile describes a small group of people: outdoor influencers, rooftop photographers, and the young men who make up most of the fatality reports. For them the “selfies are dangerous” warning is accurate and the calculator agrees.

Everybody else, the person taking a selfie in front of a cathedral or at a wedding, has a risk of zero. No shark comparison changes that.

What the Comparison Gets Wrong

The shark comparison persists because both numbers are surprising in opposite directions. Sharks seem terrifying and kill almost nobody. Selfies seem harmless and kill a few dozen people a year. Placing them side by side produces a satisfying reversal.

But the pairing also hides the one fact that would actually keep people alive. Selfie deaths are a location problem, not a phone problem. India, Russia and several national parks in the United States have responded by fencing overlooks, posting “no selfie zones” at dangerous river banks and railway crossings, and stationing wardens at the sites that produce repeated fatalities. Mumbai alone designated sixteen no-selfie zones in 2016 after a string of drownings along its seafront.

Those measures target the denominator that matters: the few thousand square metres of the planet where standing with your back to the drop is lethal. The global statistic, divided across eight billion people, points at nowhere in particular.

Sources

• Bansal, A., Garg, C., Pakhare, A., Gupta, S. (2018). Selfies: A boon or bane? Journal of Family Medicine and Primary Care, 7(4), 828-831. 259 deaths in 137 incidents, October 2011 to November 2017. Mean age 22.9, 72.5% male, 159 deaths in India. Leading causes drowning, transport, falls.
• Linares, M., Santos, L., Santos, J., Juesas, C., Górgolas, M., Ramos-Rincón, J.M. (2022). Selfie-related deaths using web epidemiological intelligence tool (2008-2021): a cross-sectional study. Journal of Travel Medicine, 29(5). 379 deaths, one in three a traveller. Falls from height, transport and drowning the top three causes. India 26%, USA 10%, Russia 9%.
• Lamba, H. et al. (2016). Me, Myself and My Killfie: Characterizing and Preventing Selfie Deaths. Carnegie Mellon University and IIIT Delhi. 127 deaths March 2014 to September 2016.
• Google (2014). Android statistics presented at Google I/O: 93 million selfies per day on Android devices.
• International Shark Attack File, Florida Museum of Natural History. Annual unprovoked shark attack fatalities, typically 5 to 10 per year worldwide.
• Mumbai Police (2016). Designation of 16 no-selfie zones following drowning incidents at Bandra Bandstand and Marine Drive.