America has never dangled on a knife’s edge for this long

Introducing the EKG of American politics.

Sep 24, 2024

This election, in case you haven’t caught on, is going to be a close one.

Yeah, yeah, I can hear you thinking, 2016 and 2020 were close. We’ve seen close elections before.

That’s true! But there’s good reason to believe November’s match-up will be even tighter than those exceptionally tight contests. In fact, in no presidential election in modern history has polling been as consistently close as this one. Not in 2000, 2016, 2020, you name it.

There are several ways to make this point. Let’s start with the national polls. As CNN’s Harry Enten has noted, in all 15 presidential races between 1964 and 2020, there was always at least a three-week period where one candidate led the polls by five points or more. In 2024, not a single day has gone by when either candidate held an average five-point lead.

At the state level, the race is within 2.5 percentage points in all seven battleground states, according to the FiveThirtyEight polling averages. (Trump +0.4 in North Carolina, +1 in Arizona, and +1.3 in Georgia; Harris +0.6 in Nevada, +1.4 in Pennsylvania, +2 in Wisconsin, and +2.4 in Michigan.) In 2016, the polls on Election Day only showed three states (Nevada, North Carolina, and Pennsylvania) within that margin. In 2020, only two (Georgia and North Carolina).

In terms of percentage of Electoral College votes that went to the winner, 2016 and 2020 ranked as the 13th and 14th closest elections in U.S. history, respectively.1 But if today’s polling averages bear out, Harris would win the election with 276 electoral votes to Trump’s 262 — which would rocket up the list to be the 4th closest Electoral College margin in history.2

Numbers 1, 2, and 3 are not exactly happy company to be in. The 1824 election was decided by the House because no candidate won a majority, with a shady agreement known as the “Corrupt Bargain” sealing the deal. 1876 was decided by an electoral commission amid rampant allegations of fraud; ending Reconstruction was the price for settling it peacefully. And then the 2000 election, of course, involved a chaotic recount that stretched for weeks, with litigation going up to the Supreme Court. (It’s also worth noting that in all three of these elections, the popular vote winner did not end up in the White House.)

And that’s not even the closest 2024 could plausibly end up! If you take Nevada (where she has the slightest lead) out of Harris’ total, she would end up with exactly 270 electoral votes, the precise number needed for victory. (By comparison, in each of the last four elections, including 2016 and 2020, the winners ended with more than 300 electoral votes.) Only once in presidential history (in 1876) has a candidate squeaked by with the exact amount of an Electoral College majority. (In 2000, George W. Bush ended with 271 electoral votes, one above the threshold.)

A hypothetical Trump win at that margin is also not difficult to conjure.

The possibility for chaos, gridlock, and discord on the level of 1824, 1876, or 2000 should not be underestimated.

But what makes the closeness of 2024 even more remarkable than that of its predecessors is that it comes so soon after the closeness of 2016 and 2020. And of 2012, 2008, 2004, 2000, 1996, and 1992.

I was curious if the U.S. has ever gone through a longer stretch of close presidential elections. So I graphed it. We haven’t:

I’m defining “close” here as “neither candidate winning more than 53% of the presidential popular vote,” which has not happened in any of the last eight presidential elections (and is all-but-certain not to happen again in 2024, the ninth).

America has never before gone eight consecutive elections meeting this criteria, dating back to the birth of our current two-party system in 1856.

The era that came closest was the Gilded Age, America’s other extended stretch of intense political competition. In the seven elections between 1876 and 1900, no candidate even managed to reach 52% of the popular vote. The streak ended in 1904, when Theodore Rosevelt notched 56.4% of the vote. But, then, it should be noted, three more razor-thin elections followed — meaning 10 out of 11 consecutive elections were 53-47 or closer. (2024 will make 11 out of the last 12 for us.)

Of course, though, the presidential election isn’t the only thing that will be close in 2024. The Senate currently stands at 51-49, after being 50-50 last election; the next Senate will likely mimic one of those two breakdowns (although with control expected to flip to the other party). After the 2022 election, House Republicans controlled a four-seat majority, just as House Democrats did after the 2020 election. 2024 is all but certain to look similarly (again, though, the slim majority might land in the hands of the current minority party).

How historically aberrant is it for control of all three institutions (the House, the Senate, and the presidency) to be so close for so long? I graphed that too. I call it, for reasons that will become clear in a moment, the EKG of American Politics:

This graph — which charts each party’s percentage of the presidential popular vote, percentage of Senate seats, and percentage of House seats in every election since 1876 — effectively let’s you explore the heartbeat of American political history. (Interactive version available here.)

You can see the huge Republican spike in the 1860s and ’70s: that’s the Civil War and Reconstruction. (In fairness, for part of that, most of the Democratic Party had seceded from the country.) You can see the huge Democratic spike in the 1930s and 40s: that’s FDR and the New Deal.

And, then, as time goes on, you can watch as all six lines — which once bounced around with abandon — slowly begin to cohere around the 50% mark. Even in the last few cycles, you can see the spikes get smaller and smaller: the blue Biden spike, smaller than the red Trump spike, which is smaller than the blue Obama spike (which was an impressive victory for our time but still clearly outmatched by the landslides of previous eras).

You can also, again, see the last time this happened — the Gilded Age — when the lines briefly constrict around 50/50 before eventually returning to trading off landslides for both parties.

For now, though, the era of landslides is over — replaced by an era of nail-biters.

There’s a lot you can glean from this data. The Princeton political scientist Frances Lee has written extensively about these trends, including in her book “Insecure Majorities.” Lee argues that the increased political competition of our modern day explains our increased congressional gridlock — since both parties are less incentivized to strike compromises with the other, when their chance at a majority (and, thus, a better deal) is always right around the corner.

I also think it helps explain the increased bitterness and emotion with which we talk about politics now — if every election is 50/50, if the country is always on a knife’s edge, the stakes feel (and, maybe, are) higher in every election. Then, multiple that by eight consecutive elections — three straight decades of being told we are in the most important election of our lifetimes — and it can help understand why the emotional dial of our politics seems to have been turned up (hopefully) about as high as it can go.

Personally, the graph also makes me feel better about the accuracy of the polling we’re seeing this election cycle. 2016 and 2020 were close elections — but, as noted above, the polling wasn’t especially close in either cycle. This feels like the first time polling has caught up to the highly contested nature of our modern politics — which means, if the 2024 election is (as I expect it will be) another nail-biter, the polling will end up having been more accurate than it has been in several cycles. The closeness makes the polling more credible, in my opinion.

The data is also, in a way, a tad deceiving — do you see those two blue lines that hover in the 60s for most of the second half of the graph? That shows the late-20th-century Democratic stranglehold on Congress, which remained almost completely unbroken from the 1950s to the 1990s.

I don’t mean to say it’s incorrect, of course — that stranglehold really did happen — but it can give an impression of pure progressive control that didn’t really exist, since the two parties were not yet polarized by ideology. At the time, plenty of conservatives were still Democrats and plenty of liberals were Republicans. So although the Democratic Party did maintain consistent majorities in Congress, the party was much less homogenous than it is now — so governing still required a great deal of haggling and horse-trading, just within a party coalition (or between the liberals or conservatives in both coalitions) instead of across the more standardized party lines we know now.

Towards the beginning of this Democratic streak, in 1952, the Polish-born pollster Samuel Lubell penned “The Future of American Politics,” a book in which he introduced the metaphor of America’s political solar system.

The solar system, he said, consists of two parties: a Sun Party, the majority, around which everything else orbits, and a Moon Party, the minority, which can do little more than reflect what the Sun shines out. At the time he wrote, the Democrats were the Sun and the Republicans, the Moon.

“Each time one majority Sun sets and a new Sun arises, the drama of American politics is transformed,” he wrote. “Figuratively and literally a new political era begins. For each new majority party brings its own orbit of conflict, its own peculiar rhythm of ethnic antagonisms, its own economic equilibrium, its own sectional balance.”

But, in the decades following, something strange happened. The parties eventually began to polarize by ideology, with liberals congregating in the Democratic tent and conservatives joining the Republican fold. Not long after, a new Sun did emerge — the Reagan Revolution of the 1980s — although it didn’t rise quite as high as the Democratic Sun that preceded it, or last nearly as long.

And then, not long after that, instead of a new Sun arising, the solar system just stopped spinning. Of the last 12 presidential elections, Democrats have won six and Republicans have won six. In the last 30 years, control of the Senate has flipped four times and control of the House, five times. No party has maintained united control of government for more than three election cycles.

In the 2024 election, it is entirely possible that the House, Senate, and White House could all flip party control, something that hasn’t happened since 1952 (and something that has never happened, as would be the case this time, in conflicting partisan directions).

America is without a Sun and Moon — which means the “orbit of conflict,” “peculiar rhythm of ethnic antagonisms,” “economic equilibrium,” and “sectional balance” of both parties (not just one) suddenly must concern us, since the two trade the place of prominence so frequently. Some days, it feels like the psychological weight of paying attention to all those overlapping zones of conflict is beginning to crush the nation.

Control of American politics has never sat this close to a knife’s edge for this long; if you’re wondering why things have begun to go a little berserk, that’s probably as good a place as any to start.

The winners in these elections should have won the same number of electoral votes, but 2016 technically counts as closer because two faithless electors defected from the winner (Donald Trump).

I’m counting Nebraska’s 2nd congressional district in Harris’ column here, which after this week appears to be a much safer bet.