H025

If nothing bad happened, was the danger real?

When prevention works, the disaster never shows up, so the danger starts to look made up. Then people stop preventing it, and the danger comes back.

Measles cases in the United States per year
U.S. measles cases per year Before the 1963 vaccine: about 3 to 4 million cases a year. 2024: 285 cases. 2025: 2,289 cases. 2026 through October 1: 3,887 cases.
On a normal scale, today’s cases are too small to see next to the pre-vaccine years
Measles before the vaccine
400–500 deaths
every year in the U.S.
Plus 48,000 hospitalized. Measles was so common that most people caught it as children
After success
Eliminated, 2000
no longer spreading in the U.S.
A generation of parents grew up never seeing a case
Y2K, January 1, 2000
$308 billion
spent worldwide on fixes
Almost nothing failed, and many people decided the whole thing had been hype

When prevention works, it erases its own evidence. Before the measles vaccine, 3 to 4 million Americans caught measles every year and 400 to 500 died. By 2000 the disease was eliminated in the U.S., and for many families it became something from history books. As vaccination slipped in some communities, measles returned: 285 cases in 2024, 2,289 in 2025 with three deaths, and 3,887 in 2026 by October 1. Early in 2025, 96% of cases were in people who were unvaccinated or whose vaccination status was unknown.

The same thing happens with levees, fire codes, and computer security: the better they work, the less there is to see, and the easier they are to cut. “Nothing happened” can mean the danger was never real, or that the prevention worked. Before deciding which, ask what would have happened without it.

Try it somewhere else

A school's IT team blocks thousands of attacks a year, and there hasn't been a breach in five years.

Is the security budget wasted, or is it the reason nothing happened?

Was Y2K real, or overhyped?

People still disagree, which is the paradox in action. Computers that stored years as two digits could read 2000 as 1900. Governments and companies spent an estimated $308 billion worldwide, about $134 billion of it in the United States, fixing systems before the date changed. Some failures still slipped through: computers handling data from U.S. spy satellites crashed for about two days, and in Sheffield, England, 154 pregnant women were given wrong risk results for Down syndrome.

Skeptics point out that some countries, such as South Korea, Italy, and Russia, spent much less and also saw few problems. Supporters answer that the quiet New Year is exactly what successful repairs look like. Once prevention has worked, there's no clean way to prove what would have happened without it.

Other places the paradox shows up

A levee that holds for decades starts to look like wasted money. A town that hasn't had a big fire in years questions its fire inspections. A company that hasn't been hacked trims its security budget. In each case, success is invisible: when prevention works, the result is that nothing happens.

Related exhibits

Sources: U.S. Centers for Disease Control and Prevention: measles before the 1963 vaccine (an estimated 3–4 million cases, 48,000 hospitalizations, and 400–500 deaths a year), elimination in 2000, and annual case counts (2,289 in 2025; 3,887 in 2026 as of October 1, 2026). CDC, Morbidity and Mortality Weekly Report, “Measles Update — United States, January 1–April 17, 2025”: 285 cases in 2024, three deaths in 2025, and 96% of early 2025 cases unvaccinated or of unknown status. Y2K: Wikipedia, “Year 2000 problem” ($308 billion worldwide, about $134 billion in the U.S.; satellite and Sheffield failures; skeptics' comparison with low-spending countries).