The sort of civil war that Walter and Marche worry about wouldn’t involve red and blue armies facing off on some battlefield. If it happens, it will be more of a guerrilla insurgency. As Walter told me, she, like Marche, relies on an academic definition of “major armed conflict” as one that causes at least 1,000 deaths per year. A “minor armed conflict” is one that kills at least 25 people a year. By this definition, as Marche argues, “America is already in a state of civil strife.” According to the Anti-Defamation League, extremists, most of them right-wing, killed 54 people in 2018 and 45 people in 2019. (They killed 17 people in 2020, a figure that was low due to the absence of extremist mass shootings, possibly because of the pandemic.)

Walter argues that civil wars have predictable patterns, and she spends more than half her book laying out how those patterns have played out in other countries. They are most common in what she and other scholars call “anocracies,” countries that are “neither full autocracies nor democracies but something in between.” Warning signs include the rise of intense political polarization based on identity rather than ideology, especially polarization between two factions of roughly equal size, each of which fears being crushed by the other.

Instigators of civil violence, she writes, tend to be previously dominant groups who see their status slipping away. “The ethnic groups that start wars are those claiming that the country ‘is or ought to be theirs,’” she writes. This is one reason, although there are violent actors on the left, neither she nor Marche believe the left will start a civil war. As Marche writes, “Left-wing radicalism matters mostly because it creates the conditions for right-wing radicalization.”