Book Reviews

Are We Always on the Verge of Civil War?

All that Civil War II talk is overblown—but that isn't the only sort of political violence to worry about.


If you define civil war loosely enough, the second one started as soon as the first one ended. While Reconstruction was in progress, multiple states were under military occupation, Southern insurgents embraced guerrilla tactics, and the Klan and similar groups carried out a racial terror campaign that killed thousands of African Americans. Yet another civil semi-war followed on that one's heels: In 1877, just as Reconstruction was ending, a great railroad strike broke out in cities across the United States. The ensuing months saw riots, repression, shootings, troop deployments, and dozens of deaths; a socialist group briefly seized power in St. Louis. A striker quoted in The Pittsburgh Leader declared that these clashes "may be the beginning of a great civil war in this country."

The railroad strike was just one of many violent labor battles of the late 19th and early 20th centuries, from the Bay View Massacre of 1886 to the Paint Creek Mine War of 1912–1913. That episodic class war intersected with another set of episodic hostilities: a decadeslong conflict that the historian Richard Maxwell Brown has called the Western Civil War of Incorporation. In that frontier fighting, he wrote, the "conservative incorporating authority" of large corporate interests and the state battled small ranchers and farmers, discontented wage-workers, and sometimes outright outlaws; the former were often allied with the Republican Party, while the latter were more likely to be Democrats or Populists. These skirmishes stretched from the 1850s through the 1920s, and some of them are legendary: the Mussel Slough War of 1880, immortalized in The Octopus; the Cochise County War of 1881–1882, immortalized in My Darling Clementine; the Johnson County War of 1892, immortalized in Heaven's Gate.

Those were not the last great spasms of bloodshed. There was Red Summer in 1919, when whites invaded black neighborhoods across the country, killing hundreds. There was another wave of labor-capital combat during the Depression, complete with steel companies stockpiling poison gas. There were the riots of the 1960s, and there were the bombing campaigns of the '70s. ("During an eighteen-month period in 1971 and 1972," Bryan Burrough writes in Days of Rage, "the FBI reported more than 2,500 bombings on U.S. soil, nearly 5 a day." Though most of those did not kill anyone, prompting Burrough to call them "exploding press releases.")

There were locally focused fights too, like the 1898 coup in Wilmington, North Carolina, when white supremacists overthrew an elected biracial government, killing dozens and possibly hundreds of people in the process. Or, on a less overtly political level, those prohibition-fueled gang battles for territory and market share that have periodically seized a city—Chicago in the 1930s, Miami in the 1980s.

When violence wasn't actually breaking out, you still sometimes could feel it simmering. In Georgia in 1968, Hancock County elected a predominantly black government—the first county to do so since Reconstruction. Not everyone in the area was happy about that. The ensuing decade saw sporadic harassment, arson, and gunfire, and there are locals who haven't stopped harboring suspicions about the plane crash that killed black activist John McCown. At one point, tensions were so high that both the county seat's all-white police force and a club formed by McCown started stockpiling submachine guns. The governor, an up-and-comer named Jimmy Carter, eventually stepped in and helped negotiate a disarmament pact. But in another timeline, they might still be telling tales of the Hancock County War.


If you think it a bit inflated to call such conflicts civil wars, I think you're right. No, not even the Western Civil War of Incorporation—that's a handy historical framework and a wonderfully evocative phrase, but it does not describe a war in the conventional sense of the word. There is a reason why that four-year fight from 1861 to 1865 is "the" Civil War. Those other mêlées pale beside it.

But you should remember this long history of armed conflict when people issue overwrought warnings that America is headed for a new civil war. Those Civil War II discussions tend to conflate two different scenarios: one where this ambient political violence intensifies, and one where Red/Blue polarization evolves into an actual war with two formal sides. The first is certainly plausible; the second is not. (And then there are the pundits with their own private definitions of "civil war" that apparently do not require violence at all. Last year, New York Times columnist Charles Blow published a column headlined "We're Edging Closer to Civil War." About a dozen paragraphs in, he allowed that "this new war will be fought in courts, statehouses and ballot boxes, rather than in the fields.")

Plausible or not, that fear of a full-fledged civil war has been growing for several years now. It has become especially intense since the riot at the U.S. Capitol on January 6, 2021. As the first anniversary of that eruption arrived, two books surfaced to greet it. Barbara F. Walter's How Civil Wars Start and How To Stop Them is a tour through recent research on political instability. Stephen Marche's The Next Civil War takes a more fanciful approach, sketching out four stories of possible paths to Civil War II and a fifth in which the U.S. breaks up peacefully. Walter comes across as someone who has read a lot of serious social science; Marche comes across as someone who has read a lot of Vox articles. Yet Walter's arguments leap far past anything demonstrated by her scholarly evidence. And Marche, whose method gives him permission to make huge leaps, somehow keeps forgetting to jump all the way to a civil war.

The problem with Walter's book—well, one of the problems—is best seen in her discussion of the Polity Project, which rates countries on a scale from -10 to 10. If you come in at 6 or higher, you are a liberal democracy; if you hit -6 or lower, you are an autocracy. Walter informs us that the countries in-between, dubbed "anocracies," are more likely to see a civil war break out; the United States, she adds ominously, has recently slipped into that middle zone. She offers several reasons to expect a country that is democratizing but not yet democratic to be more unstable, and she buttresses her case with examples from Indonesia to Ethiopia. She weaves such an enticing argument that you might miss the moment when she casually concedes that every one of her real-world precedents involves a country rising, not falling, into anocracy territory.

In fact, she offers just one example of a backsliding country that even came close to civil war: Ukraine around the time of the Euromaidan protests. That's our single precedent: a country with a very different history than the United States, in a very different geopolitical context, that didn't actually fall into a civil war. Suddenly the science is looking a little threadbare.

The science looks even less impressive when we learn a bit more about the Polity Project's precise-sounding numeric scores. When the U.S. recently dropped to a rating of 5, Walter breathlessly tells us, it was our "lowest score since 1801." So according to this oh-so-scientific measurement, the entire period from 1801 to 1865, when vast swaths of the population were not just kept from participating in politics but were held as human chattel, was more democratic than America today. And this scale is supposed to tell us something useful about the risk of civil war? Garbage in, garbage out.

Marche is less interested in crunching numbers and more interested in spinning scenarios. There's one where federal inspectors shut down a bridge they deem unsafe, and a sheriff defies them by reopening it, and the sheriff becomes a right-wing folk hero, and one thing leads to another, and pretty soon we've got militias facing off against the Army. In another, an incel assassinates our first female president. Or what if a drought hits the Midwest while a hurricane destroys New York City? Or a dirty bomb irradiates the U.S. Capitol?

Marche tries to justify his story choices by citing various experts and by mentioning the occasional historical precursor, as when an Arizona sheriff backed a county government that had opted to fix a bridge before the federal government had completed an environmental impact study. (Yes: Marche replaced a sheriff who wanted to repair a bridge with a sheriff trying to keep a bridge from being repaired. I guess the original setup wouldn't let him have the sheriff say things like "Anyone can choose to use the bridge or not to use the bridge. You know, my grampy always told me life is unsafe.") While it's easy to think of ways that these stories aren't likely, I don't want to dwell on that. Marche isn't trying to convince us that one of these specific stories will happen; he's arguing that something could happen. These are just illustrations.

So it's striking that even then, they tend not actually to end with a second civil war. The hurricane story is apocalyptic, but in a Mad Max way, not a Chickamauga way. At the end of the sheriff's story, militias are mustering and swaths of the country are under military occupation; it feels a lot like…Reconstruction. And no, Marche is not using a looser definition of "civil war" that would include Reconstruction or Red Summer. He explicitly invokes the Civil War as the sole civil war in U.S. history, even as he offers outcomes that don't look much like it.


Needless to say, a Red Summer rerun would still be extremely unwelcome, even if it isn't Civil War II. My point here isn't that we have nothing to worry about. It's that we should worry about the right thing. You won't find the right remedy if you've made the wrong diagnosis.

When I say the wrong diagnosis, I don't just mean this tendency to conflate political violence in general with a full-scale, 1860s-style civil war. I mean a tendency to treat political violence and partisan polarization as two sides of the same coin, as though the next war will be fought between the Fox and MSNBC audiences: the Red and the Blue replacing the Blue and the Gray. It is possible, I guess, to conjure unlikely scenarios where this happens, maybe centered around a disputed election. (It's odd that none of Marche's stories begin that way.) But political disorder is not encumbered by the structural factors that have allowed just two parties to dominate American elections. Our hatreds, our loyalties, and our violence point in far more directions than just two. And while Walter and Marche both understand that, that isn't true of everyone prone to pending-civil-war chatter.

Think of the riots of 2020. These often come up when right-wing figures predict that we're heading toward a Red/Blue civil war. (Here's a Townhall columnist writing that summer: "There is good reason to believe our nation is heading for a bloody civil war given the myriad forces supporting rioting, looting and terrorizing innocent people…") Yet that year's violence doesn't fit onto a partisan map very well, given how much of it was directed at Democratic municipal governments. When liberals fret about Civil War II, meanwhile, they're more likely to cite the Capitol riot of January 2021. That one did at least involve the partisans of one party trying to prevent power from passing to the other party. But even there, the pro-Trump unity masked a lot of competing agendas and mutual distrust—and the latter has increased considerably since then, as the fear of federal infiltrators drives groups that were already distrustful even further apart. The Capitol mob doesn't look like the seeds of a unified army; it looks like a temporary alliance that almost immediately disintegrated.

That isn't the only misdiagnosis floating around. Walter's book opens with another recent moment of violence, or rather would-be violence: the time a small group of anti-lockdowners conspired in 2020 to kidnap and kill Michigan Gov. Gretchen Whitmer. Walter describes the conspirators as white nationalists, but at least one of them participated in that summer's Black Lives Matter protests. Not that they were all pro-BLM: Another one reportedly tried to start a fight with some anti-racist protesters. This group just wasn't united by a racial ideology.

They weren't united by partisan loyalty either: The cabal included both a guy with Truckers for Trump lawn signs and a guy who said Donald Trump should be hanged. What united them was violent opposition to Whitmer's COVID policies. And under the surface, even that didn't unite all of them: At least a dozen informants and at least two undercover FBI agents were part of the plot, sometimes taking significant organizational roles.

Not content to misdiagnose, some analysts offer remedies that don't even fit their diagnosis. When Walter gives us advice on how to avoid a civil war, one of her first suggestions is to restore trust in America's elections—and one of the ways that she'd like to do that is to adopt automatic voter registration. This may well be a good idea on its own merits, but it makes absolutely no sense as a way to restore trust in elections. The people who worry the most about voter fraud are overwhelmingly likely to believe automatic voter registration makes fraud more likely.

Walter half-acknowledges this: The "far right" won't be assuaged, she tells us, since its "vision of a white Christian nation depends on disenfranchising minorities." But she waves that aside, declaring that the reform would "earn the support of moderates." That's the sort of thing you write when you have no idea what the debate over election integrity looks like in mainstream conservative circles.

If anything in these books makes me anxious about the future, it's moments like that one—not passages that argue compellingly that we're heading toward war, but passages that display the blindness that makes civil conflict more likely. Political violence has been a part of the American experience since the colonial days, and no doubt it will be with us for years to come. If we are to withstand it, let alone transcend it, it'll help to have a grounded sense of what those guys on the other side of the barricades believe.

How Civil Wars Start and How To Stop Them, by Barbara F. Walter, Crown, 320 pages, $27

The Next Civil War: Dispatches From the American Future, by Stephen Marche, Avid Reader Press, 256 pages, $27