As a left-leaning white kid who grew up in what turned out to be a dying Southern city, going from a fully-integrated Montessori elementary school in a neighborhood named after the white flight it later suffered to an East Memphis middle school whose history includes a massive expansion of its boarding program for Little Rock families resisting integration in the 1960s, I’m exactly the type of person you’d expect to point to the premature end of Reconstruction with the Compromise of 1877 as the point at which it All Went Wrong.
When I left the United States 18 years ago, I did so for a variety of reasons. One of them was the politics of the early aughts: the five-to-four appointment of George W. Bush as President, the rise of right wing nationalism in the wake of nine-eleven, the ensuing neocolonial expedition to Iraq, and the fair re-election of Bush after said expedition. That one was the last straw: at that point it was obvious to me that American politics was lost, and that it would end up in a place by the 20s very much like the place it has ended up. And now you can buy T-shirts publicly proclaiming your desire to Finish Reconstruction.
Finish Reconstruction, in my mind, isn’t just an exhortation to tip white supremacy and fascism in general back into history’s dustbin, although it is also that. The Reconstruction I’d like to see finished is arguably more expansive than what we’d have ended up with without the Corrupt Bargain whose 150th anniversary we’ll celebrate next year1. Forty Acres and a Mule was a dead letter essentially as soon as it was written, so reconstruction was never going to be about reparations. The political enfranchisement of freedmen was uneven, and without redistribution of land the feudal conditions of the antebellum South would have persisted to the 21st century regardless of the outcome of the 1876 election.
But what would finishing reconstruction actually look like? This being 2026, I decided to ask Claude (Sonnet), mainly as a test of how good Claude is at writing background briefs for alternate-history fiction. What it came up with was plausible and intellectually entertaining, but probably not the basis of something I’m a good enough writer to actually do well. But the four2 scenarios it came up with, each with its own historical branchpoint and focusing on one aspect of postwar enfranchisement, are interesting enough to write about as an experiment on their own.
Aside from the fact that Claude is altogether Not Bad at testing out alternate history scenarios and analyzing the plausibility of certain branches not taken, the most important thing I learned from this little experiment is that change is always gradual, which can be difficult to remember looking back at the past couple of decades of Anglophone cultural change. Except for one scenario, a Hollywood romp we’ll get into more below, even fairly large changes in the history books lead to more modestly different futures than I’d have expected. This is certainly caused in part by generating an alternate history with a model trained on the real one: it doesn’t really want to give up the dates, facts, and mesh of causality among them that are at the core of its being, unless it has to.
But it’s probably also Not Wrong. “Delete Cambridge Analytica and the mid twenties are all roses” is a comforting thought if you happen to have a single-use time machine, but it ignores the force and the inertia of three hundred million personal histories treated in aggregate. Or in other words, I admit surprise that the Form of the Destructor is one particular washed-up if well-known D-list reality television personality, the individual steps to whose rise seem improbable in isolation and not credible in combination. But the fact and the form of the destruction? I pretty much called it two decades ago.
The first three scenarios I looked at in this experiment branch out from three points in the history of the post-Civil War south, and ended up focusing on three aspects of integration (political, cultural, and economic) in isolation.
1877: The most intuitive way to finish Reconstruction is to finish Reconstruction, so we started with this one: the Compromise of 1877 is never struck, Reconstruction simply continues, and the disenfranchisement machinery (poll taxes, literacy tests, grandfather clauses) gets blocked federally through the 1880s. Black Southerners vote, hold office, and accumulate political capital across the Deep South for the next century and a half. The catch, as noted above, is that political Reconstruction without economic Reconstruction produces tenant farmers who vote, so the social structures of the antebellum South persist in different ownership. The divergences compound outward from there and blur a bit 150 years on, but end up with a measurably but not dramatically more racially equal United States, though, perhaps paradoxically to the time traveler from today, moreso in the former Confederacy than outside it.
1883: The next branchpoint starts much smaller: the Supreme Court decides the Civil Rights Cases differently, upholding the public accommodations sections of the Civil Rights Act of 1875 five to four instead of striking them down with one dissent. This is a plausible3 change to the timeline. It does nothing to attenuate Southern racism, but by removing one specific legal permission (discrimination in public places), it does inhibit the assembly of Jim Crow apartheid. This slight shift in enforcement isn’t even a norm change, at least not over the first generation. The first years of the 20th century see the same race riots and massacres, but without the unambiguous backing of the state, these are perceived differently. In this scenario, it’s Methodist clergy that speak up first, and as the Protestant church is more dominant in early 20th century America than in the early 21st, the needle moves on social integration. This story surprised me the most, I think: it makes the case that what moves history is not the large legislative moment but the slow erosion of social license through an argument made in the dominant institution’s own language. It also led to the least recognizable present America. For example: what does the music of the 21st century sound like where the Black musical vernacular didn’t have to be transcribed by white jazz and rock musicians through the first half of the 20th century, because the musicians themselves were getting their own work recorded and heard?
1927: What if the Great Flood intersected with America’s love of muckraking to produce an actual redistribution of land? The flood happens, as it will — the river does not consult political arrangements. A fictional agricultural economics reporter with the Minneapolis Tribune named Vera Haas, sent downriver to determine when grain will start flowing to the mills again, finds herself at the federal relief camps. Shocked by a scene in which a man is held back from evacuation from the disaster zone at gunpoint, at the instruction of a private plantation owner who sees him only as “my laborer”, she stays to write about it. Her series gets picked up by the AP and runs nationally, reaching (and shocking) a far wider audience than the Black press could reach. The political uproar leads to an expanded Flood Control Act of 1928, and a resettlement provision tucked into Section 7 becomes the legal basis for conveying title for tens of thousands of Black tenant farming families. This was fifteen acres and a federal loan, after a decade of litigation, in an actively hostile environment: not the original promise, but not nothing. Running the clock forward on this one felt like a confirmation of the insight from the 1883 scenario: redistribution did vastly improve the lot of the families who managed to hold on to the land — not a foregone conclusion in a legal and social environment stacked against them — but without gradual cultural change, left a much smaller dent in the future than I’d have thought.
For entertainment value, though, perhaps due to my fascination with the history of nuclear weapons, nuclear energy, and the general they-did-what dodginess of essentially everything about the 50s and 60s, my favorite branchpoint involves a Mark 7 Atomic Demolition Munition, deployed to Columbus AFB in Mississippi for a flood control study on the Tallahatchie River.
In September 1959, following years of Brown v. Board resistance much more severe than in our timeline that had metastasized beyond the ability of Federal law enforcement to contain, two Mississippi Air National Guardsmen deployed at the base, and sympathetic to the segregationist/neo-secessionist cause, conspire to cause this device to go missing, delivering it to their local paramilitary. While the complicated relationship between the Mississippi Guard and the White Citizens’ Councils had been largely ignored to this point, this crossed a line. Strategic Air Command began its recovery attempt immediately; following armed resistance to a recovery team in Columbus, Eisenhower invokes the Insurrection Act. What begins as a strictly limited US Army support for the recovery operations spreads to five states throughout the Deep South. By the time the device is finally recovered in a Memphis garage in July 1960, the re-occupation of the South is an uncomfortable fact on the ground.
Being closer to the present, this branchpoint has more recognizable politics than the others, though a de facto re-occupation of much of the former Confederacy, initially under Strategic Air Command control, scrambles things somewhat. Hubert Humphrey4’s politics and home state are a better fit for this moment, so he ends up as Kennedy’s 1960 VP pick at the Democratic National Convention, nominated in Los Angeles two days after the bomb is secured. Kennedy survives an assassination attempt, the shooter is taken alive, and the subsequent trial runs the chain of command straight back to the paramilitary network, which ends the constitutional argument about what the federal government is doing in the former Confederacy. Reconstruction Authority legislation passes in 1964, the pendulum swings a bit so Nixon wins in 1968, but without a Southern Strategy available to him, he’s mostly remembered as the moderate he tried to govern as, and at heart was. The military-industrial complex is applied to threats both domestic and foreign, and the split attention to the Cold War and its proxy battles changes the milestones of our history somewhat. The first Moon landing, with Apollo 10, happens on 16 October 1972, and is understood in this timeline as a national statement of a particular kind: we did this while also doing the other thing, and we are still here. The astronaut corps is altogether more integrated: the lunar lander on Apollo 19, in 1978, is piloted by a Black astronaut.
Without Nixon and Reagan, the culture war that comes out of the other end is organized not around race per se but around reconstruction itself. It’s a fundamentally more intellectual argument: did Eisenhower’s choice (comma “invasion”, comma “liberation”, comma “betrayal”) in 1959 taint the legal and moral basis for everything that followed? Was it not obvious that the domestic intelligence apparatus built to manage the conflict would eventually be turned against the left (by Nixon, of course)?
By the 21st century, the central cultural artifact of this argument is an essay by Bill Kristol in the Weekly Standard, on the occasion of the formal end of Reconstruction in 2001. “Ike’s Ghost” argues the constitutional case against the original action and then declining, with what critics variously describe as intellectual honesty or bad faith, to say what that argument implies about the fifty years that came after. It has been in continuous argument for a quarter century, and remains fuel for memes to the present day.
This is the most American outcome on the list.
Did I learn anything, beyond the fact that writing alternate history with an LLM is a new and entertaining form of interactive fiction? Probably not. Are the histories presented and summarized here plausible? Within the bounds of suspension of disbelief customary in fiction, yes. Does it feel like vibecoding? Yep. Am I going to do it again? Also yep. It’s a more engaging way than random-walk doomscrolling Wikipedia, and leads to largely the same outcome (me staying up to all hours of the night reading Wikipedia).
Do we need to finish Reconstruction?
Absolutely.
(1): presumably with gladiators frolicking around in the ditch in front of the Lincoln Memorial where the Reflecting Pool used to be, if this year’s anniversary celebrations are any guide.
(2): by four, I mean five; the timeline branching on 6 January 2021 invokes too deep a sense of lost opportunity even to write about here.
(3): In addition to Harlan, the sole dissent in our timeline, the four additional votes for the flipped majority here were Matthews, Woods (both seated 1881), Gray, and Blatchford (both 1882). Matthews would go on to write Yick Wo v. Hopkins and Gray Wong Kim Ark, both unusually expansive equal-protection and citizenship rulings for the era.
(4): hey, look, Minneapolis again: I swear I did not do this on purpose!