With a sample size of 395, given the intensity of feeling around general elections, all this really does is show that America is staggeringly not racist.
If this was a full scientific study, I'd agree. Lets keep in mind that it isn't, and it's just a reflection of what was actually said.
It isn't fair to draw conclusions from this such as "Alabama is a racist!?!?", but the data do still say something.
If data are to "say something", they must do so in a statistically rigorous sense. I'm not going to speculate on statistical significance in this particular case.