These tentative conclusions come from three different sources: expert opinion; research studies; and overall data on cases, hospitalizations and deaths. Let’s take them one at a time:
1. Expert opinion
Because the evidence is still emerging and sometimes inconsistent, experts do not read it precisely the same way. Still, I have noticed the broad outlines of a consensus in recent interviews:
“Any difference is a small difference,” Dr. Aaron Richterman of the University of Pennsylvania told me.
“On a per-infection basis, I don’t think it’s changed,” Jennifer Nuzzo, the lead epidemiologist for the Johns Hopkins Covid-19 Testing Insights Initiative, said. “The risks to vaccinated people haven’t really changed.”
“It really is in the ‘I don’t know stage,’” Dr. Paul Sax of Brigham and Women’s Hospital said.
Dr. Robert Wachter of the University of California, San Francisco, said: “I treat as if it might be a little more severe, but I don’t know that it matters that much. I can’t think of any public health decision that hinges on whether it’s the same or 20 percent worse.”
Dr. David Dowdy of Johns Hopkins pointed out that most people naturally assume Delta is more severe — because it is new, scary and more contagious. Given this bias, he advises assuming it is not more severe absent good evidence. He thinks it’s equally likely that Delta is slightly more severe or slightly less severe.
2. The research
The studies evaluating Delta’s severity have been frustratingly inconsistent. “There has been mixed evidence about whether it’s more severe or not,” Dr. Rebecca Wurtz of the University of Minnesota told me. “The final word is not yet in.”
Some research — like a study from Scotland — has found that the people who contract Delta get sicker on average than similar people who contracted an earlier version of the virus. Other evidence says otherwise. As Ariana Eunjung Cha of The Washington Post wrote:
A recent technical paper out of Britain suggested the Delta variant does not cause more serious illness than its predecessors, but the analysis did not specifically break out children. David Rubin, a researcher at Children’s Hospital of Philadelphia who has been studying U.S. hospitalization data, said that in recent weeks, 1,200 to 1,400 children were inpatients at the peak, and while those numbers may be large, the rate of hospitalization remains the same as in the past at 0.8 to 0.9 percent.
Dowdy notes that the more alarming findings tend to receive more public attention, but that doesn’t make them more likely to be right.
3. The topline data
The broad Covid data is probably the biggest reason to doubt that Delta is meaningfully more severe than earlier versions of the virus. The share of Covid cases leading to hospitalization seems to have held steady or even declined in recent months.
That’s true in Britain, including among children too young to be vaccinated. It’s true among both vaccinated and unvaccinated Americans in places that publish detailed numbers, like the state of Utah and the city of Seattle. And it’s true among the full U.S. population: Total cases have risen almost fifteenfold since their June nadir, while hospital admissions have risen about sixfold.
By themselves, none of these comparisons prove much, because the mix of who gets Covid can change over time. But if Delta were significantly more severe, you’d expect to see some signs that hospitalizations or deaths were rising faster than new cases. Those signs are hard to find.