This is quite the assumption.
The COVID Data Tracker Vaccination Demographic Trends tab shows vaccination trends by age group. As of June 24, 87.5% of people ages 65 or older have received at least one dose of vaccine and 77.4% are fully vaccinated. Over half (65.7%) of people ages 18 or older have received at least one dose of vaccine and 56.2% are fully vaccinated. For people ages 12 or older, 62.8% have received at least one dose of vaccine and 53.3% are fully vaccinated.
You’re right that about 45% of Americans are fully vaccinated, so far, but about 2/3 of U.S. adults have had at least one vaccine dose, and about 63% of all eligible people have had one dose.
We should focus on those who are adults, because children don’t really have much choice one way or the other.
So far, we’ve got evidence that at most about 1/3 of American adults don’t want to get the vaccine. I don’t assume that’s true, because there are lots of people who have accessibility problems in getting the vaccine, as mentioned in one of the links above.
It would be interesting to see the vaccination %age uptake for those on Medicaid, because I would think that would be among the group of people with the biggest healthcare accessibility issues, in terms of geographic location, ability to get time off from work, or even having transportation access to somewhere with vaccines.
Going to take a few snapshots of the CDC vaccination data, just for convenience.
Here’s the age breakout:
I will note that those age 65-74 have higher takeup than those over age 75. This can reflect mobility issues for the very old… as well, they may be more likely to have health issues that preclude them getting a vaccine.
Breakout by sex is boring (about 5 percentage point difference, and that could be partly age-related…as women do live longer)
Racial breakout has numbers that don’t tie up with age breakout or overall %ages… because they have only 60-70% of the racial info on those vaccinated. If the racial data were more often collected in some states, then there would be a bias to the results just due to geographic differences, so I wouldn’t rely on these data for anything.