Whether measured by campaign advertising, candidate visits, organizational effort or nervous obsessing over poll results, Michigan, Wisconsin and above all Pennsylvania have moved to the top of the priority list for both Vice President Kamala Harris and former President Donald Trump – just as they have in seemingly every recent presidential election.
Trump won the presidency in 2016 by stunning Democrat Hillary Clinton to win all three states by a combined margin of about 80,000 votes. President Joe Biden won back the White House in 2020 by recapturing all three states by a combined margin of around 260,000 votes.
Since Harris took over at the top of Democratic candidate in July, the candidates have spent more money in advertising in Pennsylvania than anywhere else, with Michigan ranking second and Wisconsin fourth, according to data provided to CNN by AdImpact, an advertising tracking service. Pennsylvania, Michigan and Wisconsin rank first, second and fourth as well in the amount of advertising the campaigns have reserved through November (with only Georgia intruding as number three on both lists.)
Bob Shrum, the long-time Democratic strategist who now serves as the director of the Center for the Political Future at the University of Southern California, said the three Rust Belt battlegrounds have remained pivotal in presidential elections for so long because they encapsulate so many of the entrenched divisions that now define American politics – between, for instance, urban and rural areas and white-collar and blue-collar voters. “They reflect the polarization,” Shrum said.
In a clear statement of their priorities, the campaigns have spent nearly $120 million more on ads in the three big Rust Belt battlegrounds than they have in the four Sunbelt states they are contesting (Georgia, North Carolina, Arizona and Nevada). After camping out in Pennsylvania for her debate preparation, Harris is appearing in all three of the big Rust Belt battlegrounds again this week. Trump is holding a town hall in Flint, Michigan, on Tuesday.
Other political operatives point out that the historic tendency of these three states to vote the same way in presidential elections functionally makes them the nation’s largest swing state at a time when the other biggest states lean reliably toward one party or the other (California, New York and Illinois toward Democrats; Texas, Florida and Ohio toward Republicans).
Michigan, Pennsylvania and Wisconsin are not carbon copies. But they do share enough common characteristics that the long-time Democratic strategist Tad Devine argues they should be thought of effectively a single state – what he calls “Mi-Pa-Wi.” Each of them is less racially diverse than the nation overall, according to data from the Census Bureau’s American Community Survey. Whites account for about three-fourths of the population in Michigan and Pennsylvania and roughly four-fifths in Wisconsin. Although their Latino communities are growing, Blacks remain the largest minority group in each of them. The three states are also slightly older than the nation overall, with seniors accounting for about one-fifth of the population in each. None have many immigrants, with residents born abroad accounting for only about 7% of the population in Michigan and Pennsylvania and just 5% in Wisconsin. All three have seen minimal population growth in recent years.
At a time when education has become an increasingly powerful predictor of political allegiance, the three converge, with about one-third of their adults holding at least a four-year college degree – just slightly below the share in the nation overall, the Census found. The median income just slightly lags the national average in Pennsylvania and Wisconsin, and trails by a larger margin (about 10%) in Michigan. All three are big manufacturing states that have seen substantial job loss in that sector since 2000, but have also seen employment in it increase by about 20,000-30,000 jobs since Biden took office, according to Bureau of Labor Statistics data.
In their religious orientation, they are very similar too: White Christians, who generally lean Republican, comprise about 55% of adults in Pennsylvania and Wisconsin and just over half in Michigan, according to newly released findings from the Public Religion Research Institute. Meanwhile, voters with no religious affiliation, who have become a staunchly Democratic group, represent about one-fourth of the population in all three, PRRI found.
While the big three Rust Belt states look similar on all these measures, in some other respects Wisconsin, on paper, should be the most difficult state of the three for Harris this year. Not only is the minority share of the population smaller in Wisconsin than the other two, but Whites without a college degree (the core of the modern GOP coalition, especially in the Trump era) cast about three-fifths of the votes there compared to about half in Michigan and Pennsylvania, according to calculations from Census data by William Frey, a demographer at the non-partisan Brookings Metro think tank.
Heavily White and blue-collar small town and rural areas, which have moved toward the GOP almost everywhere, also constitute a much bigger share of the vote in Wisconsin than in the other two. In a six-category geographic measuring system, devised by the non-partisan Center for Rural Strategies, small metros and non-metros cast nearly 50% of Wisconsin’s votes in both 2016 and 2020, compared to about 30% in Michigan and 20% in Pennsylvania each time, according to results provided to CNN by Tim Marema, the center’s vice president and editor of its Daily Yonder website.
Conversely, Democrats don’t have as strong an asset in Wisconsin’s largest metro area as in the other two states. The county centered on Milwaukee is only about half as big as the counties that encompass Philadelphia and Detroit, and doesn’t provide Democrats nearly as large a vote advantage, particularly with turnout there lagging in recent years; simultaneously, while Democrats have steadily gained ground in the suburban so-called WOW counties outside Milwaukee, Republicans still win those three big counties (Waukesha, Ozaukee and Washington), by bigger margins than almost any other major suburbs north of the Mason-Dixon line.
One last factor makes Wisconsin on paper less attractive to Democrats: unions only represent about half as much of the private sector workforce in Wisconsin as they do in the other two states, according to federal figures.
Election results over the past few decades have frequently reflected the tougher demographic and geographic challenge that Wisconsin presents for Democrats. While Wisconsin provided bigger margins for Barack Obama than Pennsylvania in his two races, Democrats won by more in the Keystone State in 2000, 2004 and 2020. In 2022, Wisconsin was a much tighter squeeze for Democrats than the other two: Democrat Tony Evers won the governor’s race there by a significantly smaller margin than fellow Democrats Gretchen Whitmer in Michigan and Josh Shapiro in Pennsylvania, and Republican Sen. Ron Johnson narrowly held his seat (while the GOP was simultaneously surrendering an open Senate seat in Pennsylvania).
Yet this year, strategists in both parties consider Wisconsin the best bet for Harris, and the toughest challenge for Trump, among the big three Rust Belt battlegrounds.
One reason is the enormous growth in Wisconsin’s second largest city, Madison, which is at once the state capital, the home of its flagship public university, and a burgeoning biotech and finance center. Even as Dane County, which encompasses the city and its suburbs, has added population faster than any other Wisconsin county, it is also becoming more blue: The Democratic share of the vote in Dane County has increased from 70% in the 2016 presidential election to 75% in the 2018 gubernatorial and 2020 presidential elections, to 79% in the 2022 governor’s race, to 82% in the hard-fought 2023 state Supreme Court election that revolved around the issue of abortion rights.
The other key to Democrats’ success in Wisconsin is that the party has remained competitive in smaller places, a key priority for party chair Ben Wikler. According to the Center for Rural Strategies, a larger share of residents in Wisconsin than in Michigan and Pennsylvania live in mid-sized cities and Democrats run better in such communities than in the other two states. These include places like Eau Claire, Appleton, La Crosse and the “BOW” counties (Brown, Outagamie and Winnebago) centered on Green Bay.
Pete Giangreco, a Democratic consultant who has worked in Wisconsin, said the party’s continuing competitiveness in the Green Bay area is symbolic of its broader reach in the state – and key to its ability to win it. “The Green Bay market is kind of the bellwether of every [Wisconsin] election, [and] it’s been trending more our way,” he said. “It’s because there has been a lot more college-educated people and people of color moving there. You’ve had a lot of people leave Milwaukee and leave Chicago and go up there. That’s not like Scranton or Wilkes-Barre in Pennsylvania, which used to be Democratic, [but] are really going the other way.”
Michigan recently has been the strongest of these three states for Democrats. Not only have Democrats run better there than in the other two in every presidential election since 2000, but the party now controls all four of the elected statewide constitutional offices (three of which are held by women), both US Senate seats and both chambers of the state legislature. (Democrats are defending an open US Senate seat this year.)
But this year, operatives in both parties consider Michigan a closer call for Harris than Wisconsin. “Michigan has some weird things going on,” said GOP pollster Gene Ulm, who is working in the state. “There are some x factors there.” Michigan’s principal “x factor” is anger in its large and previously Democratic-leaning Arab American population over the Biden administration’s support for Israel’s war in Gaza. That unhappiness has spilled over more broadly among young people on college campuses. Michigan Democrats must also contend with a concerted push by Trump to convince auto workers that the Biden administration’s efforts to accelerate the transition to electric vehicles (which Harris supports) will destroy domestic jobs. And while Biden was the candidate, Democrats faced the risk of depressed turnout among Black voters, especially younger men.
But Democrats have significant offsetting assets in the state. The focus on abortion rights since the Supreme Court overturned Roe v. Wade has turbocharged the party’s advance in white-collar suburbs around the state, particularly among women: exit polls in 2022 found that Whitmer won nearly 70% of college-educated White women. As in Wisconsin, Democrats have also retained their competitiveness in Michigan’s many mid-sized cities better than they have in most states; also as in Wisconsin, Michigan Democrats traditionally have run slightly better with working-class Whites than they do in most places. With Harris replacing Biden, organizers are cautiously optimistic as well about reinvigorating turnout among Black voters. “There was such a lack of enthusiasm over these last few months, a shot in the arm that maybe we have a chance is palpable for folks,” said Branden Snyder, a senior adviser to Detroit Action, a grassroots group that mobilizes working-class and younger voters of color in Michigan. “And there are people who are apolitical who are excited about the prospect of a Black woman being the president of the United States.”
That leaves Pennsylvania as the consensus toughest of the three states for Harris. It’s also the state analysts generally consider the most likely to provide the 270th Electoral College vote for the winner in November. (Wisconsin played that role in both 2016 and 2020.) “To me, the tipping point is Pennsylvania,” said Giangreco. “If we win Pennsylvania, [Harris] is going to be president. It’s really, really hard to see where you win Pennsylvania and you lose Michigan or Wisconsin. It’s not going to happen.”
The same broad trends reshaping the political landscapes in Wisconsin and Michigan are evident in Pennsylvania. As in the other states, Democrats are gaining in white-collar suburbs, especially in the Trump era: Biden in 2020 won the four big suburban counties outside Philadelphia by nearly 300,000 votes – over 100,000 more than even Clinton did four years earlier.
But, as in other states, Democrats have been concerned about the risk of depressed turnout and some gains for Trump among Black voters in Philadelphia. And Republicans have built an imposing and enduring advantage among the state’s large population of non-college educated White voters.
Exit polls in 2020 showed that Trump won working-class White voters in Pennsylvania by a larger margin than he did in the other two states. The Center for Rural Studies’ typology likewise shows that Trump ran much more strongly in mid-sized and smaller communities in Pennsylvania than he did in the other two states. That may be at least partly because the loss of manufacturing jobs over the past generation has been even more severe in Pennsylvania than in the other two battlegrounds. “Pennsylvania has just become a much better state” for Republicans, said Ulm, noting how the Democrats’ lead in Pennsylvania voter registration has shrunk since 2020. “Places that used to be Democratic bastions aren’t anymore.”
Dante Chinni is founder and director of the American Communities Project, which has developed another well-respected classification system to sort the nation’s political geography. Chinni says that in Pennsylvania (as well as the other two states) many of the places where Trump runs best are what the project calls “Middle Suburbs” – middle-income places outside urban centers predominantly filled with White working-class voters, like the blue-collar counties around Pittsburgh.
“These were union strongholds…and in the past that’s meant they were Democratic,” Chinni said. “But they’ve shifted. They’ve become Trump-y. Usually he runs up vote in tiny places. But these are Trump’s most reliable concentration of dense votes.”
Republican pollster Patrick Ruffini said he’s dubious that Harris is as strong in Wisconsin as most public polling now shows. And he believes all of this year’s unusual factors keep Michigan in play for Republicans despite the Democrats’ inherent demographic strengths there. But he agrees Pennsylvania is Trump’s best chance to peel off one of the big three Rust Belt states.
As an older White Catholic with roots in the state, Ruffini said, Biden benefited in Pennsylvania’s eastern areas from a “regional strength” he displayed throughout the Northeast. “I’m skeptical Harris can recreate this,” Ruffini said. “Moreover, the issue environment might be worse for Harris in Pennsylvania than elsewhere due to the fracking issue and lingering sympathy [for Trump] from the Butler, Pennsylvania, shooting.”
Democrats, however, see the possibility of Harris amassing huge margins in the Philadelphia suburbs with voters who support abortion rights and reversing some of the turnout decline that Biden suffered among Black voters in the city itself.
“Nirvana is getting 2016 numbers in the city of Philadelphia and 2020 numbers in the Philadelphia suburbs,” said Geoff Garin, a long-time Democratic pollster. “If you do that…your arithmetic in the state starts to be pretty good.”
For any of these three states to vote differently than the other would depart from recent political history. For Michigan and Pennsylvania to diverge would depart from a much deeper trend.
All three states were cornerstones of what I labeled in 2009 “the blue wall”: the 18 states that ultimately voted Democratic in all six presidential elections from 1992 to 2012. The three states moved out of the blue wall together in 2016 when they each went narrowly for Trump and then together moved back into the wall in 2020 when they each broke narrowly for Biden.
In all, Michigan, Pennsylvania and Wisconsin have now voted the same way in every presidential election since 1980 except in 1988, when Wisconsin backed Democrat Michael Dukakis and the other two went with Republican George H.W. Bush. The three states have even elected the same party to control their governorships in every election since 1994 – except in 2014, when a Democrat ousted a Republican in Pennsylvania while GOP incumbents won reelection in the other two.
The convergence between Michigan and Pennsylvania stretches back much further.
It probably didn’t seem particularly noteworthy in 1856 when Pennsylvania backed Democrat James Buchanan for the presidency and Michigan supported John C. Fremont, the first nominee of the newly formed Republican Party.
But Michigan and Pennsylvania would not vote for different presidential candidates again for the next 76 years, when Franklin D. Roosevelt won Michigan and Pennsylvania stuck with incumbent Herbert Hoover in 1932. Eight years later, the two states split again, when Roosevelt comfortably won Pennsylvania and Michigan narrowly went to Republican Wendell Willkie.
But once again, those New Deal-era splits proved the exception to a remarkable run of convergence. Michigan and Pennsylvania have again voted for the same candidate in every presidential election since 1940 except one: in 1976, Michigan supported its native son, Republican President Gerald Ford, while Democrat Jimmy Carter won Pennsylvania.
Add it up and Michigan and Pennsylvania have diverged in their preference just four times in the 42 presidential elections since Republicans and Democrats became the principal alternatives in the American political system. The two states even moved together to reject the two parties in 1912 when they were among the mere six states that backed former President Theodore Roosevelt’s third-party bid.
This trio of states has arguably become the most consistent tipping point in American politics. Michigan, Pennsylvania and Wisconsin have voted for the same presidential candidate in 17 of the 26 elections since 1920; the candidate who swept these states won 15 of those 17 campaigns. In the nine elections since 1920 when they split their vote, the candidate who carried two of these three states won seven times. Over the past century, the only presidential candidates who swept these states and lost were Democrats Al Gore in 2000 and John Kerry in 2004; the only candidates who won two of them and lost were Republican Thomas Dewey in 1948 and Democrat Hubert Humphrey in 1968.
The history of convergence between these three states – and especially between Michigan and Pennsylvania – is no guarantee they won’t tilt in different directions in November. As Garin pointed out, “Given how close the results have been in these states in the last couple of elections … they could be pretty darn close to one another and still have different outcomes.”
Nor is there any guarantee that the candidate who wins more of them will win the White House: Harris faces a real risk she could win both Michigan and Wisconsin and still fall short if Trump wins Pennsylvania, Georgia and North Carolina.
The one thing that is certain: whatever the outcome in November, the burly battlegrounds of Michigan, Pennsylvania and Wisconsin – the political state of “Mi-Pa-Wi”– will be pivotal in it.