I've been procrastinating this post for a while. Mostly, the problem is hard. So here are some thoughts.
Back when I was in grad school and studying for my comps, the big debate on racial politics was the concept of "symbolic racism." The notion is that expressions of old-fashioned, KKK-style racism are no longer considered socially acceptable, but racism remains, and those who hold racist beliefs have to find more subtle ways to express them. Advocates of the "symbolic racism" theory basically claimed that negative attitudes towards "welfare," for example, are expressions of racial animosity. The mental image that many people have of a "welfare recipient" is of an African-American. And, if someone secretly holds the old-fashioned, racist belief in stereotypes of laziness, then our closet racist will oppose "welfare" out of the belief that it is just a cash payment to supposedly lazy minorities. This is a bit of an over-simplification, but if you want to read what I read in grad school, try Divided by Color, by Don Kinder and Lynn Sanders.
It isn't hard to see why the claim upsets people, particularly on the right. The "symbolic racism" argument comes pretty close to saying that conservatism is a smokescreen for racism. If you want a good counterpoint to the Kinder & Sanders book, try The Scar of Race, by Sniderman & Piazza.
Race has a complicated relationship with partisan politics today. The vast majority of non-whites are Democrats, a majority of whites are Republicans, and a regular theme of political dispute is whether Republicans' opposition to "welfare" is racially motivated, or simply based on more of an Ayn Rand view of economics.
How can it be sorted out? The problem has always been one of "falsifiability." Buzz-word alert! Buzz-word alert! A claim is "falsifiable" when the following condition is met: hypothetically, if the claim were false, evidence could be found. That's not the same as being false. It is a fancy way of saying "testable." If I accuse you of being secretly racist, then I can always come up with an argument for what you really mean because the whole point is that I challenge your sincerity.
So, consider the common Democratic claim that the Republican Party is just plain racist, and everything they claim to believe is just a smokescreen. How could we test that?
Here's a thought experiment. Imagine a candidate running for the Republican presidential nomination who has absolutely no commitment to conservative or libertarian principles or policy, but amps up the racism. How would that candidate fare in the Republican Party? If Republican opposition to "welfare" and such policies is based on ideological principle rather than affect towards non-whites, then that candidate would lose. Badly. On the other hand, if Republicans are just plain, old racists, that candidate would do pretty well.
Unfortunately for the social scientists, Donald Trump isn't exactly the perfect vessel for this experiment. Does he have any demonstrable commitment to conservative or libertarian principles? Not by a long shot. In the past, he has advocated massive tax increases on the wealthy, single-payer healthcare, etc. Does he say racist things? Obviously. I'm not going to bother documenting them because, well, I'd like to finish this post before November.
There are several problems with treating Trump's victory as a vindication of the "symbolic racism" theory. First, while he has no historic commitment to conservative principles, he has performed ritual conversions. It is hard to find examples of him committing conservative apostasy since he decided to run as a Republican. Thus, a gullible conservative who doesn't research Trump's political history might not know that Trump is a fake conservative. Trump's victory over real conservatives like Cruz and Rubio might not mean that his supporters prefer racism to real conservatism. It might just mean they bought into his con.
Second, Trump's support within the Republican Party has not been primarily among movement conservatives. As other political scientists have documented elsewhere, Trump's support comes not from evangelicals, nor the business community, but among the lower-income voters within the Republican Party who oppose cuts to Social Security, etc. So, the voters who are supposedly "symbolic racists" haven't been the ones supporting Trump. Those voters have been supporting real conservatives, like Cruz. The ones most in favor of redistribution are the ones supporting Trump, who is the most open to redistribution. If their support for Trump is motivated by the, as it turns out, correct observation that Trump is the most supportive of redistribution, then it isn't about race. It's about policy, and the divisions within the Republican Party are between those who support redistribution, voting for Trump, and those who oppose it, voting for Cruz.
Then, of course, there's immigration. Disentangling negative attitudes towards immigrants and a desire not to depress citizens' wages by increasing the labor supply? Again, we're back to the question of sincerity.
We can't necessarily draw any conclusions about "symbolic racism" from Trump's success among Republicans. What does Trump's takeover of the Republican Party mean? Figuring that out will take a long time.
What to look for: eventually, we will get the results of academic surveys, and in particular, the American National Election Studies survey, and the Annenberg survey. The question will be as follows: what is a better predictor of Trump support-- policy positions or affect towards minorities?
I can't wait to get my hands on those data!
(Yes, "those." See my related comments here.)