A Brief History of Electoral College Bias

GOP had big edge in 2020, although it has bounced around in the postwar period



— The tipping point state in a presidential election is the state that gets the winning candidate over the 270 electoral votes needed to win the presidency.

— In the last two presidential elections, the tipping point state was Wisconsin.

— In the postwar era, the tipping point state has frequently tracked closely with the national popular vote, although in 1948 and 2020, the tipping point state was clearly right of the nation.

— The Electoral College is still likelier than not to continue to have a Republican bias in 2024, although it’s not outlandish to imagine a scenario in which there would be a Democratic bias.

Tipping point states over the years

A little over a month ago, we wrapped up a series that examined the relative “lean” of each state in presidential elections since 2000. One state that we paid special attention to in that series was Wisconsin. In 4 of the previous 6 presidential elections, the Badger State has been decided by less than a percentage point.

OK, so Wisconsin is usually a competitive state, which isn’t very surprising — so where’s this going?

Well, Wisconsin is also notable because it was the “tipping point” state in the last two elections. Essentially, the tipping point state is the state that pushes the winning candidate over the 270 electoral votes required to become president. Using the 2020 election as a template, let’s imagine all the states — and the District of Columbia — on a continuum, ranging from most Democratic to most Republican. Starting out with DC, which gave Joe Biden an 87-point margin, we’ll put 3 electoral votes in the Democratic column. The bluest state is Vermont, which went to Biden by 35 points and is also worth 3 electoral votes. So it is next, bringing his cumulative total up to 6 electoral votes. After repeating this several more times, we’ve added Pennsylvania, which brings Biden up to 269 electoral votes. Though we are sticking purely to Electoral College math for the purposes of this article, a 269-269 tie would have kicked the election to the House — Republicans, controlling a majority of delegations, would have almost certainly voted to keep Donald Trump in power. So Biden needed at least one more electoral vote.

Moving along the continuum, the next most pro-Biden state was Wisconsin, which he carried by less than a percentage point. With the addition of Wisconsin’s 10 electoral votes to the aggregate total, Biden is now at 279, enough to claim the presidency. In this sense, Arizona and Georgia, the two closest Biden-won states, are essentially just “gravy.” Although they can hardly be described as insignificant electoral prizes, Biden could have won the presidency without them.

As an aside, the concept of the tipping point state is not related to the order in which states are called on Election Night, although the 270 threshold is key in both scenarios. In 1976, for example, ABC News declared Jimmy Carter the winner after Hawaii’s 4 electoral votes bumped his total up to 272. Similarly, on Election Night 2008, several outlets were able to call the presidency for Barack Obama at 11 p.m. sharp, or immediately after polls closed in four deeply blue Pacific Coast states that were collectively worth 77 electoral votes.

With that distinction nailed down, Table 1 considers the tipping point states for Electoral College winners in postwar presidential elections, as well as their deviation from the national popular vote each year. On Table 1, positive margins are pro-Democratic and negative margins are pro-Republican (which is not a commentary on the parties, but just makes comparison easier).

Table 1: Tipping point states in postwar presidential elections

Source: Calculated using Dave Leip’s Atlas of U.S. Presidential Elections

In 1948, California was several decades away from taking on the navy shade of blue that we know it for today. As Harry Truman surprised political observers to win a full term as president that year, he carried California by less than half a percentage point, which put him over 270 electoral votes on our continuum. But Truman’s national popular vote margin over Republican Thomas Dewey was considerably more comfortable — he prevailed by just under a 50%-45% vote, or 4.5 percentage points. With that, the Electoral College would have had a 4-point pro-Republican “bias” in the context of 1948.

In presidential elections since 1948, the tipping point states have tracked closer to the national popular vote. In the 1950s, Republican Dwight Eisenhower came out on the winning end of two bouts with Democrat Adlai Stevenson. Though Eisenhower’s 1956 victory was more comfortable than his 1952 win, as he increased his popular vote margin from 10.8% to 15.4%, he had a less efficient coalition his second time around. In 1952, Michigan, the tipping point state, was slightly more Republican than the national vote, while in 1956, Florida was about a point more Democratic than the nation.

In 1956, Missouri, then a prime bellwether state, was notable because it was the sole Eisenhower-to-Stevenson state, although the Democrat carried it narrowly. But by 1960, Missouri seemed to regain its swing state image. In another narrow result, it backed the winning Democrat, John F. Kennedy, and was the tipping-point state that election — its result mirrored the national vote closely. Four years later, Washington state was the tipping point state in Lyndon Johnson’s landslide. The Electoral College had a Democratic bias in 1964 in part because the Deep South, where resentment over Johnson’s civil rights legislation was apparent, acted as something of a “sink” for Republican nominee Barry Goldwater’s votes.

During his first — and unsuccessful — run for the presidency, in 1960, Ohio backed Richard Nixon over Kennedy. But in both of Nixon’s successful elections, 1968 and 1972, Ohio was crucial. In his close 1968 win over Democrat Hubert Humphrey, it had a 1.6-point pro-Republican bias. Then, four years later, under a completely different electoral environment, Ohio remained the tipping point state: as the then-incumbent Nixon won in a 49-state landslide, Ohio assumed a 1.6-point bias in the Democratic direction.

In 1976, 11 states were decided by less than 2 percentage points, the most of any election that Table 1 considers. The tipping point state that year was Wisconsin — on the ABC Election Night footage linked to above, Wisconsin happened to be called right before Hawaii came in, pushing Carter over 270 in their count. Four years later, the tipping state was just south of Wisconsin. As Ronald Reagan ousted Carter, the state where he was born (Illinois) had a nearly 2-point pro-Democratic bias.

From 1984 to 1992, the Electoral College had a narrow but persistent Republican bias. In 1984 and 1988, the key state remained in the Midwest, as Michigan was the tipping point state. Then, in 1992, Tennessee provided Bill Clinton’s 270th electoral vote — this may also have vindicated his selection of Al Gore, a Tennessee senator, as his vice presidential pick, as Gore was considered a somewhat unconventional pick at the time. But by 1996, the modern coalition of red and blue states was beginning to align and the tipping point state that year was Pennsylvania, which, unlike Tennessee, is a present-day swing state.

This brings us to 2000. Florida, which was hotly contested — both in the context of pre-election politicking and post-election legal fights — gave the presidency to George W. Bush even as he narrowly lost the popular vote. Though the Electoral College had a slight (but obviously critical) pro-Republican bias in 2000, it is worth noting that Gore was an especially strong candidate for Florida — in relative terms, no 21st century Democratic nominee has come within 2 points of matching his performance there. In 2004, as Bush won the national popular vote by a 51%-48% margin, the Electoral College’s bias shifted to favor Democrats: John Kerry’s 2.1-point loss in Ohio, the tipping point state that year, was less than his 2.5-point popular vote loss. Still, in 2004, Ohio was within half a point of the national vote — tipping point states in the Obama and Trump eras would see higher deviations.

In both of Barack Obama’s elections, Colorado was the tipping point state — and in each year, it was about 1.5 points more Democratic than the nation. Last week, in our inaugural 2024 Electoral College ratings, as we outlined our expectations for a competitive election, we placed Colorado in the Safe Democratic column — so it is not well-positioned to be the tipping point state again.

This is of course cold comfort to Democrats, who lost in both 2000 and 2016 despite their candidate winning more votes than the Republican nominee nationally, but the Electoral College actually had a Democratic bias in the 2004-2012 period.

Despite his complaints the Electoral College was biased against Republicans, Donald Trump clearly benefited from the system in his own coalition. In 2016, Hillary Clinton essentially traded working class voters in states like Iowa and Ohio for college educated white voters in states like Georgia and Texas — the former group flipped against her after supporting Obama twice and while she got closer in the latter group, it wasn’t enough to win their electoral votes.

As Trump was ousted in 2020, data-minded Republicans could take some comfort in the fact that the Electoral College’s bias continued to drift in their favor. In a sense, we’ve come full circle on Table 1: in both 1948 and 2020, Democrats won the national popular vote by 4.5 points but only carried the tipping point state by about half a percentage point. Indeed, the 3.8-point pro-Republican bias of the Electoral College was the highest since 1948. In 2020, Biden patched up Democrats’ Midwestern “Blue Wall” by narrowly returning Michigan, Pennsylvania, and Wisconsin into the blue column, and he also managed a couple of historic flips in Arizona and Georgia. But all those states were more Republican than the national popular vote — that had not been the case for those Blue Wall states in the Obama era.

All told, and based on the Electoral College apportionment that will be in place in 2024, Biden did better than his national margin in states and districts that accounted for 226 electoral votes. He carried an additional six states — Arizona, Georgia, Michigan, Nevada, Pennsylvania, and Wisconsin — where his margin was smaller than his national margin.

A hypothetical for 2024

What would have to happen for the Electoral College’s bias to move back in the Democratic direction? Well, we actually don’t have to look very far back in history: in short, the 2024 election would have to look more like the 2022 midterms. Last year, Democrats underperformed, but still won statewide contests, in California and New York. Meanwhile, Democratic gubernatorial candidates in Florida and Texas lost statewide by margins that were much greater than Biden’s 2020 deficit in their states. At the same time, many of the states that were marginal in 2020 did not shift right in 2022 — in fact, Michigan and Pennsylvania looked bluer last year than they did in 2020.

Part of why the Electoral College has been biased against Democrats in recent elections was because states like California and New York are home to millions of “wasted” Democratic votes. In other words, Biden gets the same number of electoral votes if he carries California by 15 points or 30 points, although a weaker margin might have consequences for down-ballot races and be indicative of broader national problems.

Hypothetically, let’s imagine a world in which Biden is winning the national popular vote by Hillary Clinton’s 2-point margin from 2016, as we see a continuation of some of the trends we saw in 2022 (blue states getting a little less blue, red states getting a little redder, etc.). Perhaps Biden also bleeds more votes to third-party options than his Republican opponent, which also slightly reduces his national margin. In this scenario, the 226 electoral votes noted above where Biden overperformed his national margin remain more Democratic than the nation.

However, let’s also assume that Arizona and Georgia — two states that have trended Democratic in recent years — continue to shift, backing Biden by 2.5 points apiece. Meanwhile, Michigan and Nevada simply maintain the way they voted in 2020, backing Biden by 2.8 and 2.4 points, respectively.

All of a sudden, the Electoral College has a slight Democratic bias, as Nevada becomes the tipping-point state at a 2.4-point Democratic margin, with the state putting him at 274 electoral votes (for the purposes of this illustration, we don’t even need to consider what might happen in Pennsylvania and Wisconsin).

To be clear, our baseline expectation is for the Electoral College to continue to have a Republican bias in 2024. But we just wanted to show that it’s not impossible for it to flip toward the Democrats.