“Better dead than red.”

Cold War conservatives used to say that, and they meant it: Better to suffer nuclear annihilation, they believed, than endure a “red” victory. And by “red” they meant anyone on the left of the political spectrum – from the Soviet communists to their “fellow travelers” in the U.S.

It’s strange, then, that America’s conservative party, the Republican Party, is now the red one. There’s no ideology behind the switch, no founding decision by either party to claim a color as its own. The whole red-state/blue-state vocabulary is much newer than most people assume. It dates almost exactly from the 2000 election.

And in many cases, it fails to capture the reality of politics in America.

When nobody could agree on a color scheme

Television networks needed to distinguish two candidates on a map, and for decades they just picked whatever looked good that year.

In 1976, NBC colored GOP presidential candidate Gerald Ford’s states blue and Democrat Jimmy Carter’s states red. Four years later, that same scheme meant Republican presidential candidate Ronald Reagan’s landslide victory painted the map almost entirely blue, which is a strange thing to picture now, given how thoroughly Reagan became the patron saint of the red party.

That wasn’t just a one-network quirk, either. The Washington Post was still assigning red to Democrats and blue to Republicans in graphics used during the 2000 campaign.

If you go looking for the decision that gave American politics its colors, you won’t find one. You’ll find almost a quarter-century in which nobody had settled on today’s version.

Bush v. Gore

That inconsistency ended in 2000 with the fight over Florida’s presidential votes, though nobody at any TV network decided it should.

The networks covering the presidential contest between Republican George W. Bush and Democrat – and sitting Vice President – Al Gore happened, that year, to converge on red for Bush and blue for Gore. NBC’s Tim Russert is often credited with popularizing the phrase “red states” during the campaign, though the exact origin is disputed.

Under normal circumstances, none of this would have mattered much. Four years later, some other combination might have taken hold, the way it always had before.

But 2000 wasn’t normal.

The election stayed unresolved for five weeks while Florida’s votes were counted, contested in court and counted again. Night after night, on every channel, Americans watched the same map, colored the same way.

By the time the U.S. Supreme Court ended it and Bush was certified the winner by 537 votes, the color scheme – red for Republican Bush and blue for Democrat Gore – had become familiar enough to stop looking like a choice at all. Nobody argued anyone into it; the map just stayed in front of the country long enough to seem inevitable.

The labels started doing more work

Once red and blue attached themselves to the parties, they didn’t stay confined to maps.

By the 2002 midterms, journalists were already writing about “two Americas: red and blue,” as though the country had sorted itself into two coherent cultures overnight.

By the 2004 campaign, the vocabulary was everywhere, carrying connotations that nobody had assigned it on purpose. Red became associated with rural America, religiosity, the South and the interior West; blue with cities, more secular electorates, the coasts and college towns.

There’s truth buried in that shorthand: American political geography really has sorted itself along those lines – more than it used to, at least.

But the language hardened fast, turning a graphics convention into shorthand for identities it was never built to describe. Within a couple of election cycles, pundits were talking about red America and blue America as though they were two different countries with two different sets of values, rather than two loose clusters of states that happen to vote differently.

What the map doesn’t show

I’m a political scientist, and here’s the problem I see with this crude way of presenting voter preferences. A state decided by 2 percentage points gets colored exactly the same solid shade as a state decided by 30. Pennsylvania in a close year and Wyoming in a landslide show up looking identical, even though one of them is split down the middle.

The map doesn’t lie about who won – it just can’t show you how close it was. The result, after 20-some years of maps built that way, is a picture of the country that looks more neatly sorted than the electorate actually is.

The red-blue convention didn’t invent regional political differences. Those are real, and they’ve been growing for decades.

What the convention does is magnify them. A standard election map colors each state by who carried it, not by the margin – a state won by 4 points gets the same solid shade as one won by 26. That’s not a limit of the medium; color could run on a gradient just as easily. It’s a choice, and one that tracks the actual mechanics of the election: The Electoral College is winner-take-all, so a map showing only the winner in each state is reproducing the same either/or logic that decides who gets the electoral votes.

A third color exists for exactly this problem. Princeton’s Robert Vanderbei built the first “Purple America” map after the 2000 election, shading each county somewhere between true blue and true red according to its vote split. Landslide counties are rare, so most of the map came out some shade of purple. The technique got wider attention after 2004 – Vanderbei’s map was Newsweek’s “click of the day” on its website, though it never ran in the magazine itself – and others built their own versions afterward.

But purple never became the convention. It turns up in academic work and the occasional data-journalism piece built to correct a flatter map – but rarely on the graphic behind a network anchor, and almost never in a candidate’s own description of the country.

By the time anyone offered the correction, red and blue were no longer just colors on a map. They were the language.

A reputation never quite earned

America’s conservative party ending up red is a decent trivia fact, the kind of thing that makes for good conversation at dinner. The real oddity runs deeper than that irony, though.

An arbitrary graphics convention, repeated for five weeks during a contested recount, hardened into a vocabulary that Americans now use to describe political geography, culture and identity, as if the country had always divided itself this cleanly into two camps.

It never did.

Andrew Latham does not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.