This is great. I don't understand something, though: if the resulting disbursement through strategic demolition works to break these nodes, why doesn't the same effect occur when Katrina prisoners were disbursed?
Great question, the difference is strategic (permanent) versus accidental dispersal.
HOPE VI targeted the highest-concentration hubs and scattered everyone simultaneously. Maintaining network ties is much harder when all your co-offenders relocate to different neighborhoods at once. Chyn found children from the most concentrated projects (>70% poverty) showed the largest gains.
Katrina was a random geographic shock. It scattered everyone regardless of network position, then let them return. Kirk also did not measure the dispersal window; he measured what happened after return migration began. That shows the alternative to targeted demolition: as networks “heal,” reincarceration rates rise.
The contrast is clearer when Chyn compares demolition to voluntary relocation programs like Moving to Opportunity. MTO gave families vouchers to move, but their friends stayed behind in the projects. 57% of relocated MTO youth returned to visit. Weakened ties can heal, but severed and salted ones cannot.
Also: much of this elaborates on the theme that crime freezes housing development and undercuts housing prices. The general point is obvious but you add lots of fascinating detail. And the suggestion that crime aversion is the main driver of NIMBYism is fascinating and plausible.
This is framed as a housing policy solution, yet the policy prescription is actually in a completely different policy space: police procedure. Nothing wrong with that: being e is certainly relevant to housing. But it puts me out of my comfort zone in analyzing the trade-offs. How does this impact civil liberties, due process of law, and the rights of people who are socially adjacent to criminals but themselves innocent? Are there downsides there? I take the Fourth Amendment seriously.
There's a prima facie disanalogy between the pandemic problem where "contract tracing" developed, and the crime fighting space. If you are ordered to self-isolate because you were in touch with a disease carrier, the reason you're a risk is physically clear and not morally discreditable. People might feel violated if the government is ordering them around and curtailing their liberty when they didn't do anything wrong, but it's cleanly technocratic. But now I'm picturing a scenario where your brother has just been murdered, and the contact tracers identify you as a likely revenge killing risk and put you under house arrest. I'm not sure if that could actually happen under your proposal, but if so, it seems wrong.
I don't understand the implementation of this as well as I would like. What agencies are tasked? What do they do?
You're right that proactive policing involves tradeoffs the advocacy space underweights. But we already restrict rights based on statistical risk. Restraining orders, bail conditions, and domestic violence gun prohibitions all act before conviction. The question is which risks justify which restrictions.
On the revenge killing scenario: if someone's brother was just murdered and they're flagged as high-risk for retaliation within the 125-day window, temporary restrictions on gun purchases seem less like a violation to me than waiting for the next body. No house arrest needed, and we can use both carrots and sticks, as in the READI program, where the city offered transitional jobs and mental health services through outreach workers.
The civil liberties tension is real, which is why I scope narrowly to superstar cities where the gains from unlocking frozen land are largest. Those who consider the surveillance tradeoff intolerable will find housing cheaper elsewhere. But the current equilibrium protects incumbents' tenure at the expense of potential residents' freedom to live where they work. Implementation details depend on jurisdiction, but they follow from who you're optimizing the city for.
Aside from the constitutional dubiousness, do you think the people who are participating in this retaliatory violence are walking into gun stores and filling out paperwork? I'd bet money that this population is largely already legally prohibited and would be better targeted with stop and frisk style enforcement coupled with actual sentences than with point of purchase restrictions.
Correct, felons themselves are mostly already prohibited. The targeting logic is one step removed: the girlfriends and mothers who are legally eligible and do walk into gun stores. ATF already tracks straw purchasing patterns; the network analysis identifies which legal purchasers are funneling to high-risk nodes.
It's interesting that the laws are already in place and carry significant penalties, yet are simply not enforced. This is very much in keeping with my own observations regarding gun laws in that they're seldom used against actual criminals or in a way that would indicate that preventing violence was the goal.
Incredible as always Calipers
This is great. I don't understand something, though: if the resulting disbursement through strategic demolition works to break these nodes, why doesn't the same effect occur when Katrina prisoners were disbursed?
Anyway, fascinating piece.
Great question, the difference is strategic (permanent) versus accidental dispersal.
HOPE VI targeted the highest-concentration hubs and scattered everyone simultaneously. Maintaining network ties is much harder when all your co-offenders relocate to different neighborhoods at once. Chyn found children from the most concentrated projects (>70% poverty) showed the largest gains.
Katrina was a random geographic shock. It scattered everyone regardless of network position, then let them return. Kirk also did not measure the dispersal window; he measured what happened after return migration began. That shows the alternative to targeted demolition: as networks “heal,” reincarceration rates rise.
The contrast is clearer when Chyn compares demolition to voluntary relocation programs like Moving to Opportunity. MTO gave families vouchers to move, but their friends stayed behind in the projects. 57% of relocated MTO youth returned to visit. Weakened ties can heal, but severed and salted ones cannot.
Makes sense, thanks for the explanation!
How do we send Hurricanes exclusively to high crime neighborhoods? Perhaps Boyd can finance a study or two?
This is good, but let's not involve the organization that was supposed to do the original version of contact tracing. ;)
Regarding the topic, your contagious disease analogy is brilliant.
Also: much of this elaborates on the theme that crime freezes housing development and undercuts housing prices. The general point is obvious but you add lots of fascinating detail. And the suggestion that crime aversion is the main driver of NIMBYism is fascinating and plausible.
This is framed as a housing policy solution, yet the policy prescription is actually in a completely different policy space: police procedure. Nothing wrong with that: being e is certainly relevant to housing. But it puts me out of my comfort zone in analyzing the trade-offs. How does this impact civil liberties, due process of law, and the rights of people who are socially adjacent to criminals but themselves innocent? Are there downsides there? I take the Fourth Amendment seriously.
There's a prima facie disanalogy between the pandemic problem where "contract tracing" developed, and the crime fighting space. If you are ordered to self-isolate because you were in touch with a disease carrier, the reason you're a risk is physically clear and not morally discreditable. People might feel violated if the government is ordering them around and curtailing their liberty when they didn't do anything wrong, but it's cleanly technocratic. But now I'm picturing a scenario where your brother has just been murdered, and the contact tracers identify you as a likely revenge killing risk and put you under house arrest. I'm not sure if that could actually happen under your proposal, but if so, it seems wrong.
I don't understand the implementation of this as well as I would like. What agencies are tasked? What do they do?
You're right that proactive policing involves tradeoffs the advocacy space underweights. But we already restrict rights based on statistical risk. Restraining orders, bail conditions, and domestic violence gun prohibitions all act before conviction. The question is which risks justify which restrictions.
On the revenge killing scenario: if someone's brother was just murdered and they're flagged as high-risk for retaliation within the 125-day window, temporary restrictions on gun purchases seem less like a violation to me than waiting for the next body. No house arrest needed, and we can use both carrots and sticks, as in the READI program, where the city offered transitional jobs and mental health services through outreach workers.
The civil liberties tension is real, which is why I scope narrowly to superstar cities where the gains from unlocking frozen land are largest. Those who consider the surveillance tradeoff intolerable will find housing cheaper elsewhere. But the current equilibrium protects incumbents' tenure at the expense of potential residents' freedom to live where they work. Implementation details depend on jurisdiction, but they follow from who you're optimizing the city for.
Aside from the constitutional dubiousness, do you think the people who are participating in this retaliatory violence are walking into gun stores and filling out paperwork? I'd bet money that this population is largely already legally prohibited and would be better targeted with stop and frisk style enforcement coupled with actual sentences than with point of purchase restrictions.
Correct, felons themselves are mostly already prohibited. The targeting logic is one step removed: the girlfriends and mothers who are legally eligible and do walk into gun stores. ATF already tracks straw purchasing patterns; the network analysis identifies which legal purchasers are funneling to high-risk nodes.
It's interesting that the laws are already in place and carry significant penalties, yet are simply not enforced. This is very much in keeping with my own observations regarding gun laws in that they're seldom used against actual criminals or in a way that would indicate that preventing violence was the goal.
That sounds reasonable. Thanks for the stimulus to think in new ways about the interactions of crime, networks, and housing prices!