So you want to make a point about Americans’ skewed perceptions of race?
Step One: Don’t use a racist chart. Even if your heart is in the right place. Even if you really, really didn’t mean to be racist while making what you thought was an anti-racism point.
Step Two: Educate yourself on what “race” means in the context you’re discussing. At the very least, educate yourself on what “race” does not mean.
For example: Muslim is not a race.
Corollary to Step Two: Footnotes are not magic wands. Tacking on a note that reads “Assumes respondents define Muslim as neither Black, White, or Hispanic” does not magically transform Muslim into a race. It also doesn’t make sense as an assumption. And it doesn’t explain the presence of “Muslim” in the righthand pie chart, which purportedly shows the actual US population mix.
When you find yourself adding a footnote to explain away a problem with your chart, stop and consider fixing the chart, re-evaluating the information, or not using a chart to make the point in the first place.
Step Three: Avoid racism by omission. According to these charts, including the “actual” population mix chart, Asian Americans do not exist. Neither do Native people or Pacific Islanders. The sources for these charts are given as Gallup, IPSOS Mori, and US Census, so it seems safe to assume that the US Census is the basis for the righthand chart, but that’s clearly not true because of these omissions and because the “white Hispanic or Latino” and “non-white Hispanic or Latino” categories are ignored, with some Hispanic people possibly being counted twice, in the White wedge and in the Hispanic wedge, which might explain in part how on earth that chart totals 100%.
Corollary to Step Three: Don’t use “Average American” when what you mean is “Average White American.” Not only are large groups of people omitted from the results in these charts, they are clearly omitted from the sample who took the survey(s) or we would see them represented in their own estimates of population mix.
Not that I’m convinced that the lefthand chart actually reflects the results of the survey described. I could not find such a survey on Gallup or IPSOS Mori, which could just mean a search fail on my part, but unless the person who tweeted (and created?) these charts responds to the queries about specific sources, we can’t really determine where along the way these mistakes crept in. The closest I can find on Gallup is a series of questions asking people to estimate racial composition where they live, which is very different from estimating the US population. Plus, the numbers don’t match up because the questions were framed as “many, some, only a few, none,” not estimates of percentages.
Step Four: Honestly consider your motivation for making this point in the first place, especially as a white person. Is this tweet, for example, really about furthering the discussion about race and racism in the United States or is it about feeling superior to ignorant racist Americans? It clearly and ironically fails on both counts.