"Trump will probably get a lot more than that. What percent of blacks is Ben Carson getting? I’m guessing that Carson is getting a lot of the black vote right now. Carson is a good man and smart, but I don’t think he has the appropriate experience to be president. And when it comes down to the election, Carson’s supporters are going to realize that and vote elsewhere, most likely Trump."
The past numbers are so hard to get a handle on, and they originate from exit polls which I don't trust a lot. If wikipedia is to be believed ...
Black Hispanic
2008 McCain ... 4% ... 31%
2012 Romney ... 6% ... 27%
If Trump is in fact cutting in to those numbers even by 10-15% there are some reasons for (D)ummies to worry ...
Last election was 127 million voters,
Blacks should be ~12.2% = 15.5 million total votes
06% Romney would be just under 1 million
10% would be 1.55 million
15% would be 2.33 million
20% would be 3.10 million
It looks like he could pinch 1 million away, perhaps 2 million. Of course this is the popular vote only. Where it really counts is in the States electoral votes. If he can grab these in: OH, MI, WI, PA, NC, IN, ( that's 6 of the 12 swing states, the others are: IA, FL, VA, CO, NM, NV ) and flip them to "red", then we'll be talking victory.
If Trump or anyone got 1 million black votes that voted (D) in the last election, it would equal 2 million new voters because of (D)-1 Million and (R)+1 Million.
In a close popular vote 52-48 the effect would be devastating as 1.5% changes sides to 50.5-49.5.
The vote switching effect is a big deal to the parties. It is the main reason that Reagan went from 50.8% in 1980 to 58.8% in 1984.
This will get really interesting when hard numbers on Hispanics comes out ( which heavily influences the other six swing states ). A similar 5-10% change has a magnified efefct as they are 16% or more of the population at present. In theory, Trump *could* actually win this thing just be picking off a chunk of these two voting blocks.