A Third Party Doesn't Have to Win to Remake the GOP

Buffalo, New York, August 1848. The heat hadn't broken, and neither had the men filling the fairgrounds outside the city — several thousand of them, more than had ever assembled for a party that did not yet fully exist. They came from three factions that despised each other on ordinary Tuesdays: the old-line abolitionists of the Liberty Party, the Barnburner Democrats who'd bolted their own president rather than swallow slavery's westward expansion, and the Conscience Whigs who couldn't stomach nominating a slaveholder one more cycle. They nominated Martin Van Buren, the man the regular Democrats had just spurned, and sent him out under a banner that read Free Soil, Free Labor, Free Men. He took ten percent of the national vote that November — more than any third-party candidate in the republic's history to that point — and carried not one state.

By the arithmetic of a losing campaign, Buffalo was a rout. By the arithmetic of the next decade, it was one of the more consequential things to happen in American politics before the Civil War. Within six years, a dozen Free Soilers sat in the House. Six more, including a Massachusetts lawyer named Charles Sumner, sat in the Senate. The party that ran Van Buren dissolved by the mid-1850s. Its voters and its platform did not; they walked into a new coalition called the Republican Party and brought the whole argument with them.

I start there because the current chatter among anti-Trump conservatives — Tucker Carlson lending airtime to the idea, Marjorie Taylor Greene breaking publicly with a president she once defended without qualification, reports of strategists gathering to sketch what a conservative alternative to Trumpism might look like — keeps getting covered as either fantasy or catastrophe. Fantasy, because third parties don't win, everyone knows that. Catastrophe, because a fractured right hands Democrats the White House by default. Both readings undersell what third parties in this country actually do, which is rarely win and almost always matter.

The structural case against them is not weak, and it deserves saying plainly before it gets answered. The Electoral College and winner-take-all voting were not accidents; they were built, and are maintained, by the two parties that benefit from them, which is why ballot access laws in most states are a genuine obstacle rather than a technicality. The spoiler effect is real and it is not a myth invented by scolds — Ralph Nader took 97,488 votes in Florida in 2000, George W. Bush won the state by 537, and whether or not Nader actually cost Al Gore the presidency, millions of Democrats believed he did, and that belief shaped Democratic strategy for the next two decades. Rational voters respond to that fear by abstaining from their true preference, which is precisely why third parties tend to top out at protest-vote percentages and stay there. My grandfather voted for George Wallace in 1968 and would not discuss it after 1972, which taught me more about the ceiling on protest voting than anything I read afterward in a textbook.

But the Wallace campaign is itself the answer to the argument it seems to support. Wallace won five states and 13.5% of the national vote running against desegregation, and he did not need to win the presidency to change one. Richard Nixon watched that number, built the Southern Strategy explicitly to absorb it, and spent the next generation governing on terms Wallace voters had set from outside the party that eventually swallowed them. The GOP didn't defeat that third-party movement. It annexed it.

The Tea Party ran the same play forty years later without ever filing as a party at all. Rick Santelli's rant from the floor of the Chicago Mercantile Exchange on February 19, 2009, produced a conference call the next day with roughly fifty activists, and within four years research on congressional voting found that members of the Tea Party Caucus behaved statistically like a distinct right-wing third party operating inside the Republican conference — voting together, against their own leadership, often enough that political scientists could measure it. Grover Norquist extracted tax pledges from candidates who feared that bloc more than they feared Democrats. In June 2014, a little-known economics professor named Dave Brat used it to unseat the sitting House Majority Leader, Eric Cantor, in a primary nobody outside Virginia's 7th District saw coming. None of that required a Tea Party ballot line in a single general election.

That is the mechanism now in motion, and it is why the Carlson-and-Greene story is not really about whether either of them ends up on a ballot in 2028. It is about whether enough of the party's donor class and voting base come to believe a credible exit exists, because belief is the only thing that has ever done the work in these episodes — not the vote count, the fear of the vote count. A GOP that has to spend real attention managing the possibility of defection governs differently than one that assumes its right flank has nowhere else to go, whether or not the flank ever actually leaves.

None of this means a Carlson-adjacent third party wins a state, or should be mistaken for one. The Liberty Party men in 1848 understood something the loudest third-party enthusiasts today keep forgetting: fusion, not purity, is what converts a protest into leverage — working with a major party's own factions rather than running a spoiler against them for the satisfaction of it. Van Buren's people got Sumner into the Senate by 1854 precisely because they stopped insisting on running their own candidate everywhere and started trading support for commitments. A movement built purely to punish Trump voters for their choices, with no theory of what it wants the Republican Party to become afterward, will do what Ross Perot's Reform Party did to itself: make noise for one cycle and leave no institution behind that outlasts the men who founded it.

The file on this dates to Buffalo and runs through Wallace's five states to Cantor's primary night in 2014, and it says the same thing every time. The threat was never that the third party would win. It was that the major party would start behaving as though it might.