Twelve years ago, Florida and Ohio were swing states. Obama campaigned often in both and won them. He had relatively easy victories in Iowa (six percentage points) Michigan (10), Pennsylvania (five) and Wisconsin (seven). He finished ahead of former Massachusetts governor (and current Utah senator) Mitt Romney in a ton of small counties in Iowa and Wisconsin.

In 2020, Joe Biden also won 51 percent of the national popular vote. But the way Biden got there was significantly different: He lost Iowa, Ohio and Florida. He barely carried Michigan (three points), Pennsylvania (one) and Wisconsin (one), and did much worse compared to Obama in many small counties in the Midwest. He received only about 35 percent of the national vote in rural areas and among White people without college degrees.

But Biden won in Arizona and Georgia, two states that Obama had barely contested. He finished well ahead of Donald Trump in Colorado and Virginia, states that had narrowly gone for Obama. Biden lost Texas by only six percentage points, compared to Obama’s 16-point defeat in 2012. People who live in suburbs and White voters with college degrees backed Biden at higher levels than they had Obama.

You probably knew some of that, but it’s worth stating explicitly: Barack Obama of Harvard Law School and Chicago did better with rural voters and White Americans without degrees but worse among White college graduates than Joe Biden of Scranton and Syracuse Law. Obama did better than Biden in the fairly White Upper Midwest; Biden ran ahead of Obama in very multiracial Arizona, Georgia and Texas.

Those somewhat counterintuitive results are probably not because of particular campaign tactics of Obama or Biden, but instead broader changes in American politics. The urban-rural divide (people in cities voting Democratic, those in small towns backing Republicans) was huge in 2012 and has grown since. College-educated White voters now lean Democratic while their non-college counterparts are overwhelmingly Republican. It’s likely that Obama and Trump, the defining political figures of the 2010s, helped cement these fissures.

Republicans are now dominant in statewide elections in Florida, Iowa and Ohio. On the other hand, Arizona, Georgia and Virginia have become more Democratic. Wisconsin is perhaps the most closely contested state, electing a Democratic governor but a Republican U.S. senator two years ago. These state-level dynamics are in part explained by demographic factors but not fully. Ohio has a lot of White people without college degrees who live in rural areas — but so does Wisconsin.

Harris can’t reverse trends that have been developing for more than a decade in a three-month campaign. So I expect her strategy will be trying to reprise Biden’s victory from four years ago: huge margins in urban areas; strong support in suburbs and among White voters with degrees; deficits in rural areas and among White voters without college degrees that are a bit less than those of former secretary of state Hillary Clinton’s 2016 campaign.

Harris is likely to concentrate her time and campaign funds in Arizona, Georgia, Michigan, Pennsylvania and Wisconsin. Harris’s first major campaign stop as the presumptive Democratic nominee was in West Allis, a suburb in the Milwaukee area; her second will be in Atlanta this week. Harris isn’t likely to compete that hard in Florida, Iowa or Ohio.

It would be awkward if her campaign said this directly, but I assume they will try to win college-educated voters, people living in cities and suburbs, and the growing ranks of nonreligious Americans by even larger margins than Biden did. In my view, the segment of the electorate where Harris is positioned to gain ground is best understood not as women or White women, but White women and men who live in cities and suburbs and aren’t progressive on every issue but are culturally liberal and wary of Republican stances on abortion and LGBT rights.

Biden won about 73 percent of Asian, Black and Latino voters combined in 2020, significantly less than Obama won in 2012 (81 percent). That fits more with our stereotypes of Biden and Obama. Perhaps Harris can match Obama’s minority support levels. After all, there is a lot of excitement among African Americans, particularly Black women, about her candidacy. She is also the first Asian American major-party presidential nominee, and perhaps that also draws new support.

But there is a growing body of evidence, both in election results and polling, that Latinos and Black men in particular are more open to voting Republican than they were a decade ago. Democrats other than Biden are also doing worse among voters of color. If Harris gets around 95 percent of the Black vote and Black turnout matches White turnout (both of those things happened in 2012), I would be surprised. I would bet on her support among voters of color being closer to 73 percent than 81. And she can win at 73 — Biden did.

Overall, Obama’s 2012 coalition was more electorally efficient than Biden’s eight years later: Obama won 332 electoral votes with 51 percent of the national vote, Biden just 306 with that same 51 percent. In even the most optimistic scenario for Harris, she wins 319 electoral votes. (That’s everything Biden won plus North Carolina, where polls have shown a close contest.)

But a 2024 Democratic presidential candidate probably isn’t going to win lots of small towns in Wisconsin or the state of Ohio, no matter who they are or how they campaign. Voting patterns have changed too much. Biden was never going to be the icon among Democrats that Obama was. But Biden’s electoral path is probably the party’s best hope this November.