Sometime this week, in a spreadsheet kept by the Centers for Disease Control, a line was crossed: the United States has now recorded more measles cases in 2026 than in all of 2025. That sentence needs one piece of context to land. Last year was already the country's worst measles year since the early 1990s, back when my parents were in elementary school. So this is two record years in a row, with the overwhelming majority of cases in unvaccinated people, and with the country's kindergarten vaccination rate sitting below the level that keeps the virus contained.
When I debuted on this beat in March, I promised to follow the mechanism — the verbs, not the nouns. This is the follow-up I hoped I wouldn't be writing so soon. But the numbers insist, so let me do what this column exists to do and walk through exactly why this is happening, and exactly how it ends.
The place to start is a number called R₀, the basic reproduction number: the average number of people one sick person infects in a population where nobody is immune. Seasonal flu runs around 1.3. The original pandemic coronavirus was near 3. Measles sits somewhere between 12 and 18! That's the highest of any human virus we routinely encounter. I had to read the next part twice when I first learned it: the virus can linger in the air of a room for two hours after its host has left, so one case in an unprotected room doesn't cause a few more cases, it causes a dozen-plus.
That ferocity is why the herd immunity threshold, the vaccination level at which chains of transmission sputter out, sits near 95 percent for measles, higher than for anything else. It's also why measles is always, everywhere on Earth, the first disease to come back when vaccination slips. It isn't the deadliest pathogen we track. It's the most sensitive instrument we own for measuring a society's immunity: the canary, as my predecessor on this beat called it, in the coal mine of public trust.
Now the part of the arithmetic that took me the longest to understand, which is that averages lie. An average is what you get when you smooth everyone together into a single number, and with immunity, the smoothing hides exactly the thing that matters. The country's kindergarten coverage right now is 92.5 percent, which sounds close to the 95 we need. So why does a country that close to the line keep setting outbreak records? Because immunity isn't spread evenly. It clusters, since the decisions that produce it cluster. A county here, a school network there, a religious or ideological pocket somewhere else sits at 70 or 60 percent while wealthy suburbs sit at 99. The virus never meets the average. It meets the pockets, one at a time, like landing strips.
And the case maps of both record years look exactly like that. Not a national wave, but clusters: an undervaccinated community catches fire, the virus burns through its unprotected people for weeks, then a highway trip or a family reunion carries sparks to the next pocket. Between clusters, months of calm. The calm isn't recovery. It's just the time the sparks spend traveling.
None of this is new biology. Measles in 2026 is genetically the same enemy the vaccines defeated a half-century ago, and the vaccine is still one of the most effective ever made: two doses, ninety-seven percent protection, for life. What changed is everything around the biology, and regular readers can recite the list with me. A decade of eroding confidence. Pandemic-era disruption that millions of children's vaccine schedules never caught up from. An internet that spreads doubt faster than corrections. And, as I wrote in March, a policy year in which the machinery that sets vaccine defaults was handed to its skeptics, while the country stepped back from the international system that watches outbreaks for everyone. None of those causes is a virus. All of them have case counts now.
I said the arithmetic explains how this ends, and it does, because there are exactly two endings and both are fully available.
The first ending is the historical normal, and I want to state it plainly, without drama: the outbreaks keep growing until measles re-establishes endemic transmission, meaning the virus circulates here all the time again instead of arriving from outside. The United States loses its elimination status, joining the continental backslide that began when Canada lost its own last fall, and a disease our grandparents' generation defeated becomes a routine childhood risk again, with the hospitalizations and the rare, terrible complications that pediatric textbooks had started treating as history. Countries have walked this road before.
The second ending is just as proven: the pockets can be closed. Every recovered elimination in history, and there are several, used the same toolkit, and none of it needs inventing. Granular data to find the gaps. Trusted local messengers, meaning pediatricians, pastors, school nurses, because trust turns out to live locally, in specific people families already know. Catch-up campaigns that meet families where they are. School-entry systems that make vaccinating the easy default again. No breakthroughs required. The science was finished fifty years ago. The unfinished part is entirely human.
What I find clarifying about the arithmetic, honestly, is that it doesn't care which ending we choose. R₀ has no politics, and thresholds don't negotiate. A society gets exactly as much measles as its immunity gaps can feed, not one case more or less, and two record years in a row means the gaps are growing. I don't know which ending we'll choose. I do know we've changed these numbers before; a record low had to exist for these record highs to be measured against. The next kindergarten coverage figures arrive with the new school year, and when they do, I'll sit down and do this arithmetic again.
