Whether it is primarily due to migration, mass media, federal policy, or the homogenizing effects of global capitalism, the cultural and political differences between the North and South have become less pronounced in recent decades. But it remains unclear just how different Southerners remain from their fellow citizens in other parts of the country. White Southerners’ attitudes on the issue of race remain a question of particular interest and importance.

This subject has been coming up with increasing frequency. Across the South, monuments to the Confederacy have been coming down—though not everywhere in the region—and with some serious pushback. Alabama, for example, passed legislation protecting such monuments in May. A white nationalist rally in defense of Confederate statues occurred in Charlottesville, Va., that same month. A similar rally is apparently in the works in a few weeks.

On the other hand, if support for Trump is a reasonable proxy for racial attitudes (I vacillate on whether it is), we might say that white Southerners are moving in a different direction compared to the rest of the country. That is, in some Southern states, Trump underperformed compared to recent Republican presidential candidates. This is one reason Trump was able to win the election, despite not actually increasing the Republican share of the white vote: He gained white votes in states where it mattered, but held steady or actually lost white support in states where it  mattered less.

Whereas Mitt Romney won 68 percent of the white vote in North Carolina, Trump won only 63 percent of that demographic group. Trump also performed worse than Romney among whites in Virginia. Of the Southern states for which we have exit poll data from both years, only Florida showed an increase in the Republican share of the white vote.

Reniqua Allen raised the geographic distribution of racial views in her recent New York Times piece, “Racism is Everywhere, So Why Not Move South?” She correctly noted that black Millennials are moving to states like Texas and Georgia, and exiting states like New York. Allen suggested that the greater opportunities for blacks in the South outweigh the higher levels of racism in the region, which may not actually be much higher.

I am not qualified to comment on how black Americans’ experiences with prejudice varies across the country, and am not suggesting that this is something that can be measured by anonymous public opinion polls. But given recent events and Allen’s article, it is worth considering whether surveys show a significant geographic divide among white Americans when it comes to race.

To begin answering this question, we can turn to the most recent American National Election Survey. I have mentioned before that the 2016 survey had a trove of very good questions relating to race. By disaggregating whites in the North and South, we should be able to discern if there is still a significant difference in attitudes, as this survey had a sufficient number of white respondents to allow us to disaggregate the data by region and maintain confidence that the results are correct; the survey included 853 Southern whites and 2,185 non-Southern whites.

There are too many questions about race on that survey for me to discuss all of them in detail here, but a few stood out as worthy of extra scrutiny. We might expect, for example, for Southerners to be particularly hostile to affirmative action, both because of the GOP’s history of using that issue in the South (such as Jesse Helms’ infamous “white hands” commercial) and because the comparatively large African American population in most Southern states makes affirmative action more likely to actually impact Southern whites.

Yet when it comes to affirmative action, the difference in opinion is rather small. In both the North and the South, white support for affirmative action in university admissions is weak. Among Northern whites, about 15 percent favored “allowing universities to increase the number of black students studying at their schools by considering race along with other factors when choosing students.” In the South, this number was 12 percent. That is very low, but scarcely lower than what we find among the white Yankee population.

The ANES similarly asked respondents whether they believed whites were actually suffering negative real-world consequences because of affirmative action policies. Specifically, they were asked, “How likely is it that many whites are unable to find a job because employers are hiring minorities instead?” Once again, the results show greater concern about affirmative action in the South, but the difference is quite small in each category, and very few whites in any region think it is “extremely likely” that affirmative action keeps whites from finding jobs.

We see greater geographic variation, however, in the answers to the vaguer question, “In general, does the federal government treat whites better than blacks, treat them both same, or treat blacks better than whites?”

In both regions, a near majority of respondents declared that the government was racially neutral. And when that category is combined with those whites that think the government is actually anti-white, we are looking at a large majority of the white population in both regions. As I’ve written before, this suggests that discussions about white privilege have had a meager impact on the actual attitudes of most white Americans.

Although we should not overstate this difference, it is notable that whites in the South were more likely to think the federal government has a pro-black bias than a pro-white bias; this is not true outside the South. In fact, almost one third of Southern whites believe the government treats blacks better, compared to about one fifth of whites elsewhere.

To be clear, the belief that the federal government favors African Americans is not necessarily rooted in anti-black prejudice, nor is opposition to affirmative action—though feelings on those questions will certainly be correlated with feelings of prejudice. So what about questions that more closely approximate what social scientists now call “old fashioned racism”? Are Southern whites more likely to openly state negative beliefs about blacks? We may question the accuracy of polls using such questions, given that social desirability bias may cause many whites to answer dishonestly. But unless we expect dishonesty to vary systematically in the North and South, we can still use them to look for regional differences.

When looking at these kinds of questions, we again see differences, and in the expected direction, but the gap is small in each case. When asked about black Americans’ work ethic, about four percent of non-Southern whites described blacks as “lazy”; among Southern whites, the percentage was about five percent. On a similarly-worded question asking whether blacks are violent or peaceful, about eight percent of Southern whites described blacks as violent, compared to about four percent of non-Southern whites.

Given the rather modest regional differences on most of these questions, it is possible that the gaps that exist are driven by other factors (such as different average levels of educational attainment), rather than the lingering memory of Jim Crow and de-segregation. But regardless of the causal mechanism at work, there does appear to be a gap. The question is how much it matters.

There are various ways we can interpret these data. Depending on your hopes and expectations, these numbers may be distressing or encouraging.

White Southerners who are frustrated when described as racists by their fellow citizens to the north can note that the Mason-Dixon Line is not a clear divider when it comes to racial attitudes. This may also be a relief to people concerned that the North-South divide is once again becoming dangerously wide.

White nationalists who hope the American South will be a vanguard of a white racial awakening will be discouraged to see that Dixie is, on most of these questions, only marginally different from the rest of the country. Anti-racists, who may want to blame the South for all of America’s racial divides, will be frustrated to learn that it is not just reactionary white Southerners holding racial progress at bay.

These descriptive statistics from one survey are of course not definitive. Nor do they adequately capture the lived experience of race relations in the South or elsewhere. But these data do provide one further indication that North-South divide on racial attitudes, although real, is narrow.

George Hawley (@georgehawleyUA) is an assistant professor of political science at the University of Alabama. His books include Right-Wing Critics of American Conservatism, White Voters in 21st Century America, and Making Sense of the Alt-Right (forthcoming).