President Donald Trump speaks during a news conference at the White House, Tuesday, July 28, 2020, in Washington. (AP Photo/Evan Vucci)
There it was for all the world to see, an August 28 headline from The Hill: “Poll: Trump approval rises among black, hispanic voters amid convention.”
From the looks of it, you would think the headline made sense, given what The Hill said:
“Twenty-four percent of registered black voters in the Aug. 22-25 survey, which included the first two days of the convention, said they approve of the job Trump is doing as president, while 76 percent said they disapprove.
That is up 9 points from the previous survey conducted Aug. 8-11, where the President received 15 percent support among this group.
The survey found support among Hispanic voters also grew by 2 percentage points, from 30 percent in the last poll to 32 percent in this most recent survey.
Forty-four percent of overall registered voters in the Aug. 22-25 survey approved of Trump’s job as president, a 1 percentage point dip from the last poll.”
As with any poll, it’s all about the sample, and from the looks of this latest poll, it appears that it is once again the example of a poll that grossly underrepresents the Black and Latino electorates.
For this specific poll, it asked only 343 Black voters and 401 Latino voters. There is also no mention of language for Latinos.
So if one were to take 24% of Black voters who strongly supported the President in this poll, that would be 83 Black voters. And if one took the 32% of Latinos in this poll who strongly supported the President, that would be 128 Latinos.
That’s not a national poll, that’s something you do on Facebook.
Blaring such decisive conclusions from such a small sample with a much larger sample of white voters is just irresponsible reporting by The Hill.
We are also fairly certain that The Hill forgot to cite recent polls that sampled 1,310 Black voters in (Florida, Georgia, Michigan, North Carolina, Pennsylvania, and Texas) and 1,552 Latinos (in Arizona Florida, Colorado, North Carolina, Pennsylvania, and Texas). Those two polls cited the following:
- “There is a noticeable bump in support for the Democratic ticket among Latinos, going from +33 [Joe] Biden to +48 Biden, following the [Kamala] Harris announcement.”
- “[African American] support for Biden over Trump following the Harris announcement grew by 11 points, going from +68 to +79. Those under 40 say her selection makes them somewhat or much more likely to vote (82%).”
Granted, we are very aware that Biden and Harris have issues with Black and Latino voters and that Trump still has support in communities of color, but from a purely polling standpoint: which poll would you think might be more reliable—the one that used Black and Latino voters as a subset or the one that actually polled Black and Latino voters as a primary focus?
That is all.
