Ice cream sales and drownings rise and fall together. Does ice cream cause drowning?

Two things moving together is a real pattern you can measure. It still doesn't tell you which one, if either, is causing the other.

Ice cream sales and drownings through the year

Each line is shown as a share of its own busiest month, so the two shapes can be compared.

Line chart of an illustrative year: ice cream sales and drownings both stay low in winter, climb through spring, peak together in July during the summer months, and fall again by December. SUMMER 0% 50% 100% Ice cream sales Drownings JanFebMar AprMayJun JulAugSep OctNovDec
Ice cream sales Drownings Summer months

What's really connecting them

Diagram: hot weather causes more ice cream sales, and hot weather causes more swimming, which causes more drownings. There is no arrow between ice cream sales and drownings; they only move together because of the weather. Hot summer weather the hidden third factor people want something cold more people go swimming More ice cream sold More drownings no cause here they just move together

When two things move together, something else may be moving both of them. Ice cream doesn't make anyone drown, and drownings don't sell ice cream. Hot weather does both: it sends people to the freezer aisle and to the pool, lake and beach.

This hidden-third-factor trap shows up everywhere. Kids with bigger shoes read better, because older kids have bigger feet and more practice reading. Countries that eat more chocolate win more Nobel Prizes, and both go with being wealthy. When two trends line up, ask what else could be driving both before believing that one causes the other.

What two things have you seen rise and fall together, and what else might be moving both of them?

More pairs that move together without one causing the other
These move togetherThe hidden factor
Ice cream sales and drowningsHot weather
Kids' shoe size and reading levelAge
Firefighters sent to a fire and the damage it doesHow big the fire is
Chocolate eaten per person and Nobel Prizes per person, by countryNational wealth
Hospital visits and deathsHow sick people already are
How can you tell if one thing really causes another?

The strongest test is an experiment: randomly split people into two groups, change one thing for only one group, and see whether the outcome changes. Random assignment spreads hidden factors like weather, age or wealth evenly across both groups, so they can't explain the difference. When experiments aren't possible, researchers compare cases where the suspected hidden factor is the same, such as comparing ice cream sales and drownings only on days with the same temperature. If the link disappears, the hidden factor was doing the work.

Related exhibits

Sources: The monthly lines are an illustrative pattern, not measured data, drawn to match the well-known summer peak in both; CDC death records show July has more drowning deaths than any other month. Chocolate and Nobel Prizes: F. H. Messerli, "Chocolate Consumption, Cognitive Function, and Nobel Laureates," New England Journal of Medicine, 2012.