A poll worker and voters in Columbus, Ohio, in 2016John Minchillo / AP

Contra Barack Obama, there is a white America and a black America. There are also varying versions of Latino and Hispanic Americas across different regions of the country. There are robust, enduring differences in belief across races and communities about just what America’s identity should be and how politics are experienced, and they in turn create the political reality of the country. Partisan politics arise from these differences, and exploit them. And these differences might be structural, informed by the basic fact of human geography, a geography itself built on the fact of American apartheid.

A new poll conducted by the Public Religion Research Institute (PRRI) and The Atlantic delves into these differences, and finds distinct racial outlooks on the most fundamental issues of American identity and values. This poll reaffirms previous findings from PRRI and The Atlantic of deep racial differences in policy preferences across a number of issues, and also confirms the picture of deep structural barriers to the ballot for black and Hispanic voters, barriers that played a role in the 2016 election and the 2018 midterms.

The PRRI/Atlantic poll, a random survey of slightly more than 1,000 people taken in December, reveals major differences among racial groups on some of the basic questions about what makes America America, and what makes Americans so. A strong majority of white respondents—59 percent—think that speaking English is a very important part of being American. The majority of black and Hispanic people do think that speaking English is at least a somewhat important component of Americanness, but almost a tenth of both groups think it’s not important at all, while only 2 percent of whites feel that way. Interestingly, black and Hispanic respondents are much more likely than whites to say that belief in God is a very important component of a specific American identity.

When it comes to the basic philosophical underpinnings of what an American nation means, the poll also finds sharp distinctions among races. Seventy-six percent of white people and 74 percent of Hispanics think that civil liberties such as freedom of speech are very important pieces of American identity, while only 61 percent of black respondents feel so. More than half of black respondents think that a belief in capitalism isn’t a very important part of that identity, while good majorities of both white and Hispanic people think that it’s either somewhat important or very important. Forty-five percent of Hispanic respondents said that racial, ethnic, and religious diversity make the country much stronger, compared with 32 percent of whites. And white respondents are most likely to say that diversity makes the country weaker, or to be ambivalent about the idea of diversity altogether.

These differences equate to real political differences. Black people overwhelmingly identify as Democrats, and the small but influential sliver of black conservatives who identify as Republican appears to be diminishing, as the increasing influence of Trumpism and the alt-right of the modern GOP have made the Republican Party more and more openly hostile to black voters. Other polls also show black voters increasingly concerned about racism. White voters might be moving in the other direction, and philosophically seem to be deprioritizing the importance of diversity in favor of an embrace of capitalism, nationalism, and individual liberty. And according to the researcher Adrian Pantoja, an analyst with the Latino Decisions political-opinion research firm, Latino voters have increasingly made opposing the GOP agenda a top political priority in the age of President Donald Trump.

The profiles of the Republican and Democratic parties have shifted accordingly. In the 2018 midterm elections, Democrats elected one of the most racially diverse incoming classes of legislators since Reconstruction, and a diverse field of potential presidential contenders revolves around a multifaceted policy debate that’s heavily influenced by progressive ideas. Republicans have had to shape their party around explicit appeals to white voters and their anxieties, and have had to build an electoral strategy that can promote low overall turnout and stoke white grievances. In short, Democrats have cultivated an image as the party of racial and cultural pluralism, while Republicans have rejected pluralism as a viable strategy.

The emerging policy strategies within the parties also reflect these racial differences. As the Fight for $15, a global political movement advocating higher minimum wages, has significantly affected liberal organizing, and as $15-minimum-wage promises have become a potential litmus test in the 2020 Democratic primary, the PRRI poll indicates that 81 percent of black people favor raising the minimum wage to $15 an hour, with 60 percent of them strongly favoring the law. Majorities of both white and Hispanic respondents also favor the minimum-wage increase, but not at the numbers or with the fervor of black people. With gun control a major partisan wedge issue, black and Hispanic respondents are also much more likely to strongly favor stricter gun-control laws than whites. Interestingly, black respondents are the most likely racial group to support providing pathways to citizenship for unauthorized immigrants currently living in the United States.

As racial identity and partisan identity more closely align, and as turnout becomes a more and more important metric for assessing political fortunes, the PRRI poll also finds that people of color are markedly more likely to report having faced notable barriers in electoral and civic participation than white respondents.

Six percent of black people and 7 percent of Hispanic people reported that they or someone else in their household did not have the proper voter identification the last time they went to vote, a small proportion but one much higher than the 1 percent of white respondents who said they’d had similar problems. Nine percent of black respondents and 8 percent of Hispanic respondents said that they or someone in their household had trouble finding their polling place, again as opposed to zero percent of white respondents. Nineteen percent of black people and 14 percent of Hispanic people said they’d had to wait in long lines, as opposed to 9 percent of whites. And the percentages of black and Hispanic voters who reported harassment or an inability to take off work, for themselves or for household members, was significantly higher than the percentage of white voters who said the same.

These results support findings in a June 2018 PRRI/ Atlantic poll, in which “across just about every issue identified as a common barrier to voting, black and Hispanic respondents were twice as likely, or more, to have experienced … barriers as white respondents.” And they come after a 2018 midterm election in which claims of voter suppression and major civil-rights lawsuits came to define elections in places such as Georgia, Kansas, and North Dakota.

Notably, the 2019 update also polls respondents on their conception of the controversial potential addition of a citizenship question to the census. Advocacy groups have opposed the question on the grounds that it might really be used to further an anti-immigration agenda and imperil the residences of unauthorized immigrants—or that people in Hispanic communities especially might believe that the census would lead to illegal immigration checks, and would refuse to answer or take the survey. The PRRI/Atlantic poll finds that this suspicion has merit, and might already be affecting the census’s effectiveness. On the one hand, 40 percent of Hispanic respondents think that a potential citizenship question would be used for checking individuals’ immigration status as opposed to counting the population. Black respondents, on the other hand, are the least likely to buy the stated rationale for the question, and only 17 percent believe that it will be used for the sole purpose of counting the population. These numbers track with the Census Bureau’s recent approximation that more than 600,000 households would fail to complete the census because of the question.

In light of the national conversation around both the census and voter suppression, both of which concern the basic shape of American democracy and the role minorities play in the future of America, these results from PRRI and The Atlantic shed new light on fundamental questions. They illustrate that the so-called demographic destiny of America is one that would look radically different should the country become majority nonwhite sometime in the next 20 or 30 years. The data indicate that white, black, and Hispanic voters have markedly divergent ideas on what exactly makes the American identity, and they also indicate that these differences are enforced and entrenched via spatial and social segregation. But the data also cast some doubt on the political prospects of that demographic destiny. They show that black and Hispanic voters are more likely to be carved out of the political process, and that those efforts are perhaps aided by the existing regime of segregation. In all, they show that the competing visions of America, as separated by race and region, are indeed competing, and that they are the chessboard upon which all politics is played.

This project is supported by grants from the Joyce, Kresge, and McKnight Foundations.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.