[ Question ]

Models predicting significant violence in the US?

by kdbscott2 min read25th Oct 20206 comments


Changing Your MindForecasting & PredictionPracticalWorld Optimization
Personal Blog

Do you have a model predicting >1% chance of significant political violence (>5,000 deaths) in the US in the next year? Pls share.

Also welcome: models predicting <.1%

Metaculus has been walking around 1-5% probability of a US civil war before July 2021. While 1% is the lowest probability one can give on Metaculus, rumor has it some reasonable people put more than 1% probability of significant political violence (SPV) happening.

I am[was] confused about this, and want[ed] to get a better sense of what models lead people to assign credence on this order of magnitude.

[it turns out this exercise led me to no longer feel disbelief, but I'm preserving it below. See Changing My Mind for my update]

My model borrows a bit from Samo Burja's civil war typology and goes something like: "while there's some probability of small-scale violence (e.g. a few hundred deaths from scattered violence), there simply doesn't exist the mechanisms for SPV". In my mind, mechanisms include:

  • Military: e.g. military vs. government (coup), military vs. military (split along e.g. leadership lines), or a weak/disorganized military that doesn't withstand a revolutionary offensive. I don't see any of this happening with the US military (e.g. <.01%). ThirdEyeOpen on Metaculus outlines cases where the military 'sides' with the democrats or republicans, which blows my mind. Is this a blind spot for me? Is there some small chance that any coherent part of the military defects in this way?
  • State (proxy wars): Both sides of a conflict having support from states. I guess there's some 'foreign interference' in our politics already, but if I'm Russia I really don't want to get caught supplying significant amounts of money/training/arms to insurgent groups in the US.  That feels like much more risk (to Russia) than the incremental gain. But maybe there's some situation where it's worthwhile? Seems pretty close to WWIII, but I guess that's tail risks?
  • Insurgency: I currently have this as a large bucket that spans from haphazard terrorism to organized attempts to replace the government. It seems like the US might experience some things on this scale, but how far? I’m having a hard time imagining violence worse than e.g. The Troubles. What am I missing here? Maybe it's easier to kill more people these days? Is the main concern effective terrorism from a few actors, or more like thousands of insurgents? Because organizing insurgents requires a bunch of stuff...

Changing My Mind

So I started to read about insurgencies, specifically looking for the kinds of qualities that give insurgencies 'oomf'. I felt confident that US extremist groups didn't have these qualities, and if I kept reading about insurgencies (e.g. by skimming articles in Wikipedia's list of revolutions and rebellions) I would find them. I was roughly right, but in the process of articulating these qualities and digging into the specific messy ways insurgencies go, my curve of ways-things-could-go-in-the-US started to change:

from this...
...to this
(where SPV lies further to the right)

Looking back, I think my model was something like:

  • There are a bunch of prerequisites to SPV.
  • We are missing too many prereqs and they are too difficult to reach from here. It would be like threading 6 needles in a single go.

Now my model is something like:

  • Prereqs can snowball, and the US has some proto-prereqs.
  • The political violence nerds think it's possible (if unlikely)

I'll add an answer below for more detail on my model.

Feely feelings

N.B. I'm not sure why I included this section... maybe I think other versions of myself would benefit from knowing that the inside experience of writing/researching/mind-changing is a feeling-heavy process.

There were at least three occasions where I sat down and tried to write down the gears of my own model so that I could make it abundantly clear why 1% chance of civil war was loony. The first two times I got increasingly irritated, like: why am I doing this? This is such a waste of time, it's just so obvious that this is not something to worry about. I'm having a bad experience, and also this is probably just making my life worse by taking up time and making me worry about things that don't need to be worried about. I'm just at a complete loss about what would cause people to believe civil war is possible, it feels like the burden of proof is on them. Why do I need to build these complicated models myself?

"Because a few reasons" went the response "but mostly I'm feeling a kind of feeling... a feely feeling... I'm not sure what it means but I have a suspicion that following it will lead to some cool discovery like:
  a) a useful implicit model
  b) the reason why people believe there might be SPV
  c) actually changing my mind
And regardless, I'll get a better sense of what's up with this feeling."

Often times the feeling was demotivating, so the first two times I gave up / respected the feeling. The third time it led me to drafting the post above, and then shifted (like the felt shift in focusing) when I was reading about insurgencies.


New Answer
Ask Related Question
New Comment

4 Answers

some models for things getting worse

(I attempted to rank this list and the sub-lists from stronger to weaker models)

  1. Some pre-insurgency qualities
    1. More protests
      1. correlate with more conflict[1]
      2. create more opportunities
        1. for violent-leaning people to find each other and become more radicalized
        2. to evolve more virulent ideology
        3. to become better organized
    2. Already exist plenty of resources & training
      1. Highest guns per capita
      2. Lots of people with military experience - e.g. to source more weapons, to train recruits, and to fight effectively
    3. Shifting overton window
      1. More political polarization
      2. Non-negligible support of political violence[2]
      3. More and larger protests (involving both far right & left)
      4. Trump
      5. More mass shootings, hate crimes[3]
    4. Ideologies
      1. Disenfranchised populations[4]
        1. On the right, the decline of WASP power, e.g. via shifting demographics and culture[5]
        2. On the left, decline in economic power of the 99%.[6]
      2. Distributed media makes it harder to control the narrative, and more likely that extremists find each other.
    5. Possibility for local support[7] i. Some limited coordination with far-right groups among local law enforcement[8], where it’s possible this could lead to a festering insurgency in rural areas where local law enforcement is unwilling to step in[9].
    6. Financing maybe easier these days (crowdsourcing, crypto).
  2. Appeal to authority
    1. ACLED has the US on its list of 2020 conflicts to worry about[10]
    2. David Kilcullen is the kind of person who might know and has recently written a couple articles highlighting that characterize the US as in a pre-Mcveigh moment (May article) and an incipient insurgency (June article).
  3. Black swan: we don’t have much data on insurgencies / SPV in developed countries, but developed countries haven’t existed for long. We might just not know what it looks like.
    1. I don’t know if I can make a strong case for it being impossible for civil wars to emerge from developed countries.
  4. WMDs: maybe it’s easier to kill a lot of people these days, so it might only take a few actors to cross my arbitrary >5k deaths SPV threshold.

  1. Ward et al has ‘high-intensity conflictual events’ (protests, fighting, killings) as the second-highest correlated variable with higher probabilities of conflict / civil war. ↩︎

  2. Voter Study Group found that 21% of Americans thought that violence was at least a little justified if the [opposing party] won the 2020 election. This study also found an increase in the tolerance of violence since 2017. ↩︎

  3. In 2018, the most recent year the FBI reported data. Also my inner Steven Pinker compels me to note that the overall violent crime rate has been declining steadily ↩︎

  4. I currently think the most-likely-to-foment-insurgency ideologies are about disenfranchised populations, in large part due to the Ward et al having ‘Excluded Population’ as by far and away the highest correlated variable with conflict. Ward meant Excluded Population to mean “excluded from political access to the state”, which I understand to be groups that cannot vote, or are otherwise feel they are being deprived of political power like the shia in iraq or hutu in rwanda. ↩︎

  5. The “Ideologies of Rebellion” section of this article covers some adjacent far-right ideologies. They often seem to orbit around a decline in WASP power, as the author of this thesis makes a (biased) case for. I wonder if given more opportunities to evolve, some violent version of this ideology could garner support in more than 5% of the population (where 5% is a wild guess for the level of local support at which fighting an insurgency becomes difficult). ↩︎

  6. While Occupy fizzled, maybe some violent iteration of it could snowball? Seems pretty unlikely to me. ↩︎

  7. I currently model local support as important for sustaining an insurgency, from reading e.g. How Insurgencies End and Guide to the Analysis of Insurgency. ↩︎

  8. See the “Far-Right Links with Law Enforcement” graphic in this CGPolicy article. There’s a history of this, see e.g. this retired sheriff helping to defend Clive Bundy’s ranch from federal officials. ↩︎

  9. See the ‘Rebel Opportunities’ section of this Just Security piece for a brief case. ↩︎

  10. Conflicts where “violent political disorder was likely to evolve and worsen” ↩︎

some models against things getting worse

  1. Negative correlates: Country qualities that negatively correlate with conflict[1]
    1. Strong democratic institutions. Maybe because it makes lots of trusted non-violent avenues for change[2]
    2. Wealth[3]. Maybe you’re less likely to risk dying if you can meet your needs well enough with the current system
    3. Political representation[4]
    4. Being a developed country[5]
  2. Military: I think the overwhelming majority of groups would not want to fight the military, from PR risk[6] and dying risk[7]
  3. Ideology: Hard to get people to rally behind a specific extremist cause
    1. The ideology of the extremist right-wing is actually pretty varied and sometimes contradictory.
    2. I think the extremist left-wing is similarly varied: from a strong central government (socialist/communist/environmentalist) to ~anarchists (who thus far have been the only violent ones[8])
    3. The Voter Study Group actually found that tolerance of violence correlated negatively with (one measure of) partisanship
    4. Less of a stomach for violence (a la Steven Pinker)[9]
  4. Financing: Would be hard. If a group gets labeled as a terrorist organization you really don’t want to be associated with them financially[10]
  5. We’re still missing a lot of insurgency qualities[11] (this can also be used as a list of red flags if any of these crop up)
    1. High levels of political violence[12]
    2. Organized, violence-endorsing groups
      1. With significant membership (say >50,000)
      2. Publicly claiming responsibility for specific violence, e.g. assassinations of political leaders
      3. With popular-ish ideology
      4. With charismatic leadership
      5. Attempting to garner popular support
      6. Low rates of defection
    3. Institutions supporting violent groups (e.g. town or state or foreign governments, churches, unions, wealthy individuals/organizations)
    4. Economic gradients towards supporting or joining insurgents
    5. Insurgents attempting to claim & defend territory from the government
    6. Insurgents being supported by foreign groups (governments, terrorist orgs)

  1. according to Ward et al’s model ↩︎

  2. I think a parliamentary democracy would probably be better, but still ↩︎

  3. Ward et al used infant mortality rate to track this ↩︎

  4. ‘Excluded Population’ (large slices of the population excluded from political access) is by far the biggest factor that predicts conflicts in their model. I think political representation is the rough opposite, and that the US is doing pretty on the front, compared to e.g. 55 years ago when plenty of folks couldn’t vote. ↩︎

  5. I could only find one instance (although RAND says there are two) of something approximating a civil war in a developed country since 1945: The Troubles in Ireland. That’s out of >127 civil wars that killed at least 1,000 people. Fearon and Laitin: “for any level of ethnic diversity, as one moves up the income scale, the odds of civil war decrease, by substantial factors in all cases and dramatically among the most homogeneous countries. The richest fifth is practically immune regardless of ethnic composition” ↩︎

  6. Going up against the most respected US institution is rough if you need recruits and the support of locals. ↩︎

  7. My current guess is the US military would be especially effective at counterinsurgency in the US: shared language & culture with the locals, better command & control (compared to e.g. cooperating with foreign militias), and probably less political quagmire due to fewer governments at play. Although politics could make things very hard, e.g. blowback when fellow Americans get caught in the crossfire. ↩︎

  8. The Portland protest shooting is the only far-left death in the past 20 years according to New America. There’s also plenty of ~anarchists that don’t fit cleanly in a left/right bucket, like the Michigan folks. ↩︎

  9. While the Voter Study Group has some fraction of voters feeling violence is ‘justified’, it’s not clear what this means. The steady decline of violent crime still feels pretty compelling. Perhaps the definition of ‘violence’ is shifting away from ‘killing people’ towards ‘punching people’? While people might feel it’s justified, would anyone actually commit violence? ↩︎

  10. I’m pretty unsure, but it would probably fall in ITAR/OFAC violation territory, which involves million dollar fines, frozen assets, and decades in prison. Banks are allergic to people/orgs remotely associated with terrorism, because the Treasury can invoke §311 of the Patriot Act to cut the bank off from the financial system. Oh and you might lose nonprofit status. ↩︎

  11. See e.g. the CIA's Guide to Analysis of Insurgency or RAND’s How Insurgencies End ↩︎

  12. to get to the same per capita rate as The Troubles we would be losing ~50,000 people per year to political violence (Troubles had ~250 deaths per year in the 70s with a population of ~1.6m, scale that up to a 328m population and you get ~51k). Though many other insurgency conflicts had lower deaths per capita. ↩︎

It does depend on your reference class. European political systems have two big differences from the US

  • mostly parliamentary
  • fewer guns

Chile, Argentina, Brazil and Mexico have political structures closely modeled after the US constitution (though each is different). They also have similar guns per person. This reference class increases your base rate. Mexico, per capital, probably crosses the threshold (but has 1/3 America’s income and is a young democracy).

I still don’t see p>1%. But I didn’t do the math out.

Also, one of the shootings at CHOP/CHAZ in Seattle and one shooting in Portland were clear incidents of PV. The Rittenhouse incidents count. Two separate vehicular manslaughter incidents should be included. But they are such rare events and COVID has everyone on edge. No trend yet IMO.

Metaculus 2020 U.S. Election Risks Survey doesn't give >1% for >5000 deaths, but I think it is justified to infer something like that from it:

While large-scale violence and military intervention to quell civil unrest seem unlikely, experts still judged these possibilities to be far from remote. Experts predicted a median of 60 deaths occurring due to election-related violence, with an 80% confidence interval of 0 to 912 fatalities that reflects a high degree of uncertainty. Still, the real possibility of violence is a notable departure from the peaceful transitions that have been the hallmark of past U.S. elections. Results indicate an 8% probability of over 1,000 election-related deaths — suggesting that while widespread sustained clashes are unlikely, this possibility warrants real concern. Experts assigned a 10% median prediction that President Trump will invoke the Insurrection Act to mobilize troops during the transition period.

2 comments, sorted by Highlighting new comments since Today at 4:30 PM

Baseline rates of deaths from political violence are a bit tricky to pin down, but a process of elimination on homicides (YMMV, depending on which political acts you normalize as, on balance, apolitical. For various reasons, I tend to exclude DV, and theft, but include racism and police violence as political) gets us to 1k to 5k per year for the past decade, so, assigning an expectation of 1% for breaching 5k for the next year seems low.

We are likely operating on different definitions what constitutes political acts

Good point - I'm thinking political acts along the lines of violent protests, terrorism, and insurgencies. I can see how police shootings could be included there. The spirit of what I'm going for is how much change to expect, so e.g. deaths above and beyond what you would have in an average year