Arlando Monk is an increasingly rare find in presidential politics: a voter whose choices matter.

The Black entrepreneur lives in Wisconsin, one of seven expected battlegrounds in the 2024 presidential race. He is registered to vote but not sure he will bother. He has not decided between former president Donald Trump or President Biden, if those are the major-party options.

“If it’s between them, I’m going to say this: Trump was hilarious. He was hilarious,” said Monk, 43, who lives in the Milwaukee area. Biden, meanwhile, has not delivered the change he expected, leaving Monk unsure. “I would say, it’s kind of up in the air.”

If U.S. presidents were selected through the principle of “one person, one vote” that governs legislative races, the ballots of undecided swing-state citizens such as Monk would be worth just as much as the other 150 million or so Americans who are expected to vote next year.

But that is not the system handed down from the nation’s founding fathers, who opted for multiple winner-take-all contests that give greater power to smaller states. The electoral college was supposed to moderate the passions of what Alexander Hamilton called the “general mass,” which he worried could fall prey to candidates with “talents for low intrigue and the little arts of popularity.”

That 18th-century system – which is unlike anything used by the United States’ 21st-century democratic peers – has aged in surprising ways. Premised on the idea that states should each choose electors who would then select a president, the system increasingly distorts the democratic process as partisan divisions grow along geographic lines.

Advances in technology, meanwhile, allow campaigns to calibrate their outreach to only the most persuadable voters. The upshot is that a tiny segment of the population will get an outsize say in who leads the United States. And the will of the majority may not even prevail.

Once rare, the frequency with which the electoral college has skewed the overall result has increased: The “general mass” – now called the popular vote – has been won in two of the past six contests by someone who lost the White House. In both cases, the Republican candidate benefited.

At the same time, the count of swingable states has narrowed. The 2024 presidential campaign is likely to target a smaller share of Americans than at any point in the modern era, despite massive increases in spending due to online fundraising, a Washington Post analysis found.

During the last election, just 10 states and two congressional districts were targeted by Republican or Democratic nominees’ campaigns. It was a precipitous drop from the 26 states on average that were targeted each year between 1952 and 1980, according to a forthcoming book by political scientists Daron R. Shaw, Scott Althaus and Costas Panagopoulos. The research is based on internal campaign documents, interviews with campaign leaders and media reports.

The Washington Post’s analysis found that just 1 in 4 Americans lived in such areas in 2020, down from roughly 3 in 4 in the earlier period. If the major parties do not contest Florida in 2024, as is widely expected, only 18 percent of Americans would live in battlegrounds.

The targeted voters in the decisive states should expect a barrage of communications – on their phones, in their mail, from their friends, family and colleagues, over radio, television, streaming services and social media networks. The rest of the country’s citizens will find themselves on the sidelines, watching the news or the occasional live candidate event with a diminished voice in their own futures. Their vote will still count but is unlikely to decide anything.

“It’s now getting to the point where you are probably talking about 400,000 people in three or four states. That is what it is getting down to,” said Joe Trippi, a Democratic strategist who has worked on presidential campaigns since 1980. “It does mean that more and more people feel that they don’t have a say.”

The geographic sorting of Americans along partisan lines helps explain why. Most of the country resides in red or blue states where the outcome in a two-party race is not in doubt. If a candidate wins California by one vote or 1 million votes, the electoral college outcome is the same, giving candidates no incentive to campaign in places where the outcome can be predicted. Thirty-three states – including giants such as California, New York and Texas – have voted for the same party in each presidential election dating back to 2000.

But geography is not the only way Americans have found themselves excluded. That’s because voter opinions have hardened in recent decades, and technology has improved to allow candidates to target only the individuals they need to reach, sometimes even distinguishing between members of the same household.

“We are in an era of politics where data makes campaign strategy highly sophisticated and specialized, which creates smaller and smaller universes of what we call ‘gettable voters,'” said Republican strategist Mike Shields, one of the architects of his party’s data infrastructure. “Instead of using increased spending to target a broader swath of voters, you have more and more money focused on specific voters that campaigns think will be decisive in the electoral college.”

The shrinking campaign is aided by the fact that many voters are consistent from election to election in their choices about whether to vote and which party to back.

An analysis for The Washington Post from Grassroots Targeting, a Republican data firm, found that the number of voters who split their ticket between parties in their presidential and federal legislative votes has dropped sharply since 2000 in 7 of 8 major battleground states. No more than 3 percent of voters in the 2020 presidential election split their tickets in all those states. In Nevada, the share was just 0.1 percent.

TargetSmart, a Democratic data firm, did another analysis of swings in partisan vote share for the 2012, 2016 and 2020 elections. Despite some double-digit outliers, the average swing in targeted states between elections was close to 5 percent. In the 2020 election, the second in which Trump was on the ballot, turnout rose but the number of voters who changed their party preference was especially small as a wide majority remained locked in.

“It is pretty clear that not all voters are equal,” said Shaw, the political scientist, whose book “Battleground: Electoral College Strategies, Execution, and Impact in the Modern Era” is due out next year. “It’s not just the system. It is the system and the voters it is operating on.”

The data science campaign

The great irony of the shrinking campaigns is that they have never had more money to spend. Advances in online campaign fundraising and changes in campaign finance rules to allow unlimited donations have unleashed billions of dollars in new donations. Political ad spending on the 2024 campaigns is expected to approach $11 billion, a nearly fourfold increase over 2016, with about 1 in 4 dollars going to the presidential race, according to projections by AdImpact.

The vast majority of the ad and organizing dollars is focused on media markets and outlets where the voters who will decide the election reside, as campaigns seek to infiltrate the communities, households or iPhones where the few identified swing voters such as Monk spend their time. The evolving science of figuring out where those few people are has become a growing obsession of the political strategist class, and is the central purpose of Biden’s initial $25 million in advertising this year.

As it stands, the campaign data science cannot predict exactly who will vote or how those who do vote will mark their ballots. But it can create probabilities for each voter on those metrics, crafting digital dossiers that are used by everyone from the campaign bosses to grass-roots volunteers. Those dossiers are then used to hone small universes of potentially persuadable voters who get the most attention. With each election, explains Trippi, the ability to narrow the size of these groups improves.

“We are reducing it and reducing it,” he said, adding that the effort will probably become even more precise once artificial intelligence tools become more involved.

Michael Whatley, the chair of the Republican Party in North Carolina, said his party will focus on less than 2 percent of the votes cast in the swing state.

“We’re looking at 5.5 million votes that will be cast in North Carolina. It’ll be down to 100,000 voters who are undecided going into the election that we’re going to be targeting and communicating with,” he said. “We can target them.”

Ben Wikler, the Democratic Party chair in Wisconsin, said he begins with the assumption that both major parties in his state will get 48 percent of the vote, though the identity of those voters changes from cycle to cycle depending on turnout.

“We look for inconsistent voters who will probably vote for a Democrat if they vote and then we look at ticket splitter or swing voters,” he said.

At the Biden campaign, the early betting is that the outcome of next year’s election will again be very close, with a margin once again hinging on tens of thousands of votes in a few states. Former Secretary of State Hillary Clinton, who won the 2016 popular vote by 2.9 million votes, or 2 percent, could have won the electoral college if about 80,000 people in Michigan, Pennsylvania and Wisconsin had voted differently.

In 2020, about 45,000 votes in Georgia, Arizona and Wisconsin could have changed the outcome of that race, even though Joe Biden won the popular vote by more than 7 million.

Before 2000, the outcome of the electoral vote had matched the popular vote in every election for more than a century. But in recent years the aberrations have become more common, with mismatches in 2000 and 2016 that enabled the popular vote loser to claim the White House.

In the past two elections, Republicans have had a distinct electoral college advantage. That’s because the tipping-point states that won the election were more Republican-leaning than the country as a whole. In 2020, Wisconsin, the tipping-point state, was 3.5 points more Republican than the country, the highest advantage since 1948.

The shifts have only magnified the need for campaigns to ignore the bigger blue states where Democrats tend to rack up large margins. Biden volunteers are being trained to use a smartphone app that allows them to directly enter information to the party voter file about their friends, family and people they meet from key states to help improve the targeting. Early organizing efforts have focused narrowly on Black, Latino, young and female voters in Arizona and Wisconsin, groups that they now think could be decisive.

Persuading people to participate in the election at all may prove more important than winning them over from the other side. That means campaigns are likely to focus on voters who are among the least politically engaged.

“The people who will decide this election don’t want to participate in it,” said one Democrat who requested anonymity to describe planning for next year.

One person close to Trump, who is not authorized to speak publicly, said most of the campaign’s resources if he wins the nomination will focus on the same seven states that Biden has identified as early targets – Wisconsin, Michigan, Pennsylvania, Nevada, Arizona, Georgia and North Carolina. All were vigorously contested in 2020. If things go well for Trump, the campaign could expand into Virginia and Minnesota.

In some formerly swing states, such as Florida and Ohio, Republicans are unlikely to spend much money. And even in the states up for grabs, the focus will be narrowly defined.

“The reality is: The swingy areas aren’t that swingy anymore. You win an election by going with your base vote plus a swing universe. The swing universe is smaller now because people just put on their uniforms and go vote,” said Justin Clark, the 2020 deputy campaign manager for Trump. “You’ve got to dial in the message in terms of what people care about. It’s a lot of data work, focus grouping, targeting. Really, really digging in and finding voters. You’re in an election where everything is on the margins. You have to find that group of 10,000 votes.”

Clark said that in the 2020 campaign, Trump and his team looked for particular issues in key states to move groups of voters, such as the Lumbee Indians in North Carolina, a population of about 55,000 that the Republican National Committee recently opened an office to target. “You’re talking about raw numbers of voters in the thousands that are winning and losing presidential elections. You’ve got to find those votes and move them. You’re talking about crazy small margins here,” he said.

The legacy of 1787

At the root of this dynamic is the antiquated assumption at the heart of U.S. presidential politics – that regular people cannot be trusted to directly select America’s leader. Enslaved people, Native Americans and women had no vote in the original U.S. Constitution. The White, landowning men who did were seen as susceptible to their own passions and selfish interest. Electors, by contrast, were expected to be enlightened and above the fray.

“The people choose the electors,” James Madison, a drafter of the U.S. Constitution, said during the debate. “This can be done with ease and convenience, and will render the choice more judicious.”

To this day, courts have ruled that electors, chosen by state ballots and selected with the help of campaigns they are intended to support, have the ability to change their mind before selecting a president, unless otherwise barred by state law.

Throughout history, the number of electors who have defied their state’s will has been small, peaking at 10 in 2016 who defied their instructions to support Trump. In 2020, the Trump campaign sought to overturn his defeat with a strategy that focused on submitting alternate sets of electors that would back him in states where he had lost the popular vote.

The electoral college system came with tricky ramifications, including a built-in bias toward smaller states. The population of California, the most populous state, is more than 67 times the population of Wyoming, the least inhabited. But California only gets 18 times the representation in the electoral college of Wyoming. Put another way, the vote of Wyoming residents is worth 3.7 times more than Californians in presidential contests.

In recent elections, Democrats have dominated in many populous states such as California and New York. Republicans have won by smaller margins in large states such as Texas and Florida, while racking up wins in many small, rural states. The overall effect has been to give the GOP an edge. University of Texas political scientists found that in a 50-50 popular vote election, the Republican had a 65 percent chance of winning in recent elections.

Polling data has consistently shown that a majority of Americans oppose the electoral college, and in a 2019 Gallup survey one of the most common concerns among those opposed was that “the winner of the popular vote doesn’t always win the election.”

The United States has periodically considered switching to a popular vote system. In 1969, the House voted overwhelmingly to do so. But the measure was blocked a year later by a filibuster in the Senate led by segregationist Southerners. Today, despite an effort to get states to honor the national popular vote, there is considered to be little prospect for change given the country’s polarization and the high barriers to amending the Constitution.

Just how the electoral college process plays out in 2024 could hinge on whether possible third-party campaigns by candidates such as Robert F. Kennedy Jr. and by the centrist group No Labels gain traction.

A serious third-party campaign next year could scramble the Republican and Democratic battleground map and lower the barriers to victory.

“In a three-way dead heat, it’s 34 percent to win,” Kennedy’s campaign manager, Amaryllis Fox, explained in a recent live stream for supporters.

Distortions between the electoral college and the popular vote can be even sharper in cases where the presidential choice is split among three or more presidential contenders.

Abraham Lincoln won 40 percent of the popular vote in 1860 when he ran against Stephen Douglas and two other candidates. He got 59 percent of the electoral college votes. Bill Clinton won 69 percent of the electoral college votes in 1992 against two rivals, while winning only 43 percent of the vote.

A campaign with three or more serious contenders could also yield an inconclusive result: In cases where no candidate wins a majority of the 538 electoral college votes, the Constitution punts the election to the next Congress, with each state delegation getting a single vote. Republicans currently control a slim majority of House delegations.

For Monk, such considerations are unlikely to play a role in his decisions about the coming campaign.

“A lot of people have given up their faith in voting,” he said about the disconnect he feels with the nation’s leaders, who have not yet delivered what he wants on issues such as student loans and spending on overseas conflicts. “A lot of things haven’t changed.”

But it is precisely his indecision and relative lack of interest that will make him so important, subject to a campaign effort larger and more sophisticated than anything the nation has previously experienced.