Already did that. Here, from the Census Bureau website, is the breakdown by state of the two groups - percentage of African-American, and total population in millions:
MA - 5.4 percent, 6.3 million
NJ - 13.6 percent, 8.5 million
RI - 4.5 percent, 1.1 million
HI - 1.8 percent, 1.2 million
AL - 26 percent, 4.4 million
Ark - 15.7 percent, 2.7 million
LA - 32.5 percent, 4.4 million
MI - 36.3 percent, 2.9 million
WV - 3.2 percent, 1.8 million
WY - .8 percent, .5 million
Note that the researchers were careful to lump states together, to make it impossible to look for variations within the high-ownership states and low-ownership states.
MA - 5.4 percent, 2.3
NJ - 13.6 percent, 4
RI - 4.5 percent, 3.7
HI - 1.8 percent, 2.6
AL - 26 percent, 8.5
Ark - 15.7 percent, 5.5
LA - 32.5 percent, 11.2
MI - 36.3 percent, 9.9
WV - 3.2 percent, 2.2
WY - .8 percent, 1.8
So with this simple bit of analysis, the high-gun ownership rate states of West Virginia and Wyoming have LOWER murder rates than the four low-gun ownership states. Hence the need to lump them with the four Southern states - to hide another possible finding of that study, one the PC folks at Harvard would NOT want to discover.
MA - 9.3 percent, 2.3
NJ - 8.5 percent, 4
RI - 11.9 percent, 3.7
HI - 10.7 percent, 2.6
AL - 16.1 percent, 8.5
Ark - 15.8 percent, 5.5
LA - 19.6 percent, 11.2
MI - 19.9 percent, 9.9
WV - 17.9 percent, 2.2
WY - 11.4 percent, 1.8
So, with the exception of West Virginia, the poverty rate has a fair correlation to the relative murder rate.
In other words, there are many, many factors that better model the murder rate than gun ownership.