Brynn Anderson / AP

When Kamala Harris laid into Joe Biden in last night’s debate, she knew what she was doing. In a speech earlier this month, the former vice president had reminisced about being able to work with die-hard segregationists when he was a young senator from Delaware. In sharply criticizing those remarks, and his past position against school busing, Harris was trying to convince the Democratic Party’s most reliable voters—African Americans—that Biden isn’t as strong on civil rights as many of them thought.

In early polls of the 2020 Democratic presidential field, Biden has held a strong lead among African American voters—an outcome that, to some, might seem surprising when two black senators, Harris and Cory Booker, are also running. While progressives on Twitter were appalled by Biden’s speech, many in the Congressional Black Caucus downplayed the former vice president’s comments or defended him outright. Nor did the dustup hurt Biden at the grassroots level, either. In a recent report from South Carolina, a key primary state where nearly two-thirds of Democratic-primary voters are black, CNN found that support for the former vice president was holding steady.

These are signs that the controversies that blow up on social media haven’t changed the fundamental political calculations of African American voters—calculations whose savviness and complexity are only dimly understood and often entirely ignored by campaigns, political commentators, and the voting public alike. The historically large and diverse 2020 presidential field provides a unique opportunity to observe the process by which the black electorate settles on candidates.

Like other voters, African Americans make judgments about a candidate’s character and policy positions. Because descriptive representation matters, black voters do show more interest in black candidates—but historically, that hasn’t always translated into electoral support in the end.  

The long history of discrimination against African Americans, at the polls and elsewhere, has shaped voting behavior in more distinctive ways. Black voters have been highly pragmatic: They have typically favored candidates who are known quantities over fresh faces with shorter records, and in primary campaigns, have deliberated long and hard over who’s most likely to win in the general election.

All of these factors help explain Harris’s attack on Biden last night. Polls show that African Americans intensely disapprove of Donald Trump. Past experience suggests that they will support Biden—over Harris, Booker, or anyone else—if they conclude the former vice president is best equipped to beat the incumbent. And their support could well prove decisive. But Harris just made Biden’s task harder.

Since the 2016 election, the white working-class voters who broke for Donald Trump have been media darlings, the subjects of extensive reporting and analysis. But in a Democratic-primary campaign with a fractured field, the choices that black voters make are far more consequential—and receive far less attention.

Exit polls showing that more than 90 percent of black voters now support Democratic candidates in midterm and general elections fuel the myth of a monolithic black vote and conceal the diversity of opinion within the African American community. The partisan skew is easily explained. While 31 percent of African Americans identify as liberal, 42 percent identify as moderates, and 22 percent as conservative, nearly all of these voters have ascertained that the Republican Party is less interested in federal protections of civil rights than the Democratic Party. Indeed, African Americans’ enduring quest for stability and certainty on civil rights helps explain a cautious voting pattern in primary races as well. When a candidate has had a long relationship with black America, the risk of any unpleasant surprises on civil-rights issues is lower.

Still, other factors come into play. In the past seven presidential-primary campaigns without a Democratic incumbent, black voters analyzed the political landscape in quite different ways.

In 1984 and 1988, Jesse Jackson’s presidential campaigns received 77 percent and 92 percent of the black vote, respectively, signaling that black voters were most interested in descriptive representation, believing that a candidate with firsthand experience of being black in America would prioritize a program of racial equality. As the civil-rights legend John Lewis more or less said at the time, these voters knew Jackson was unlikely to secure the party nomination. Yet, as the political scientist Katherine Tate has documented, they understood how a strong showing in the primaries would give him leverage to incorporate civil-rights and economic-security measures into the party platform—a strategy that bore some fruit. In subsequent campaigns, the assumption that black voters naturally will support black candidates proved inaccurate; this was clear from the largely overlooked presidential runs of former Senator Carol Moseley Braun and civil-rights advocate Al Sharpton in 2004.

In 1992, Bill Clinton emerged from the field and secured the nomination by winning black voters, particularly those clustered in the southern states. Crucial to his appeal was a cultural fluency that he could deploy as a native of the region. The approach wasn’t as effective elsewhere in the country; black voters in New York, for example, gave just as much support to California Governor Jerry Brown.

Eight years later, Clinton was highly popular among African Americans, who overwhelmingly supported his vice president, Al Gore, a heavy favorite in the primaries. In 2004, black voters largely coalesced around John Kerry, who made the case that he was more electable and offered the best chance to unseat George W. Bush, who was deeply unpopular among African Americans. (John Edwards, a senator from North Carolina, still managed to split the black vote in a couple of southern states by making regional appeals.)

In 2008, black Democrats were notably slow to warm to the newcomer Barack Obama, who had to convince black voters that his historic presidential bid was viable before they would peel away from the establishment favorite, Hillary Clinton. In 2016, she performed admirably with black voters, though Bernie Sanders, with his more expansive policy agenda, was able to win the majority of those under 30.

This behavior isn’t capriciousness run amok. It does reflect the interplay of a varied set of electoral preferences. In the five consecutive contested primaries since Jackson’s heyday, black-voter choice wavered between choosing the culturally familiar candidate (Bill Clinton), the well-known establishment candidates (Gore, and Hillary Clinton in 2016), the candidate who stressed his electability (Kerry), and one candidate who is African American (Obama). Not coincidentally, the candidate that won the black vote in each of these primary campaigns also won the nomination. Which candidate won and why was the result of the black electorate’s careful analysis of the political moment.

Seeing the overwhelming margins for Democratic candidates in presidential elections, some commentators portray African Americans as blindly adhering to groupthink—in essence, as simply doing what they’re told—as if they were incapable of making up their own minds. Yet there are good reasons why, even in Democratic presidential primary campaigns, large majorities of black voters do end up supporting the same candidate.

Political solidarity is not an innate characteristic of black America; it is a survival tactic that adverse experience has reinforced time and time again. Chattel slavery and Jim Crow paid little mind to the specific talents, abilities, and aspirations of each black person. Simply being a member of the race was sufficient to be disenfranchised and oppressed. Black Americans have been made acutely aware that their individual fates are linked to the well-being of the whole group. As political scientists have documented, social norms in the community encourage African Americans to look out for one another and ostracize those who don’t. During presidential campaigns, politics is a major topic in what the political scientist Melissa Harris-Perry has described as “everyday talk”— the lively conversations that occur in African American common spaces such as barbershops, salons, churches, neighborhoods, and now, black Twitter. After all this deliberation, African American voters typically end up backing the same candidate—which maximizes their electoral power.

What does this mean for 2020? Like everyone else in the party, black Democrats are trying on candidates for size. Biden, who served eight years as Obama’s vice president, benefits from being a known quantity. Polls suggest his pitch that he’s the most electable of the bunch has been resonating. Then again, Harris and Booker have raised the most money from African Americans, suggesting that black political donors have been doing with their dollars what members of the broader black electorate haven’t been ready to do with their votes. How long these trends hold up depends not just on the candidates’ performance, but on the outcome of the millions of social interactions occurring each day within black America.

Last night, Kamala Harris tried to steer those conversations in a new direction—and gave her own campaign an air of viability. It was a shrewd move because, when black Democrats finally settle on a candidate, they hasten the decision of the party as a whole.

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