Back in May, I noted that the CDC was providing a very low “best estimate” of the COVID-19 fatality rate to modelers. Its “case fatality rate” was 0.4 percent; it also thought that 35 percent of infections were asymptomatic, which implied an infection-fatality rate of just 0.26 percent. This was far lower than most other estimates at the time, and the agency ignored requests for an explanation from me and other journalists.
Well, they’ve updated the site: They now put the infection-fatality rate at 0.65 percent, which is much more in line with other sources, if fortunately below the 1 percent rate many estimated months ago.
They fixed it just in time for new evidence to throw off the calculations again!