An incentives based "problem homeless" solution

A cost / benefit analysis

We spend $10B annually on just homeless shelters today, at least another $3B on HUD and ESG stuff, and I'm sure there's more. San Francisco by itself spends about $1B a year on homelessness, and it is manifestly doing nothing.

There's something like 150k "problem homeless" who are chronically homeless and on downtown streets in the entire USA. $13B / 150k is $86k per person per year. Similar relevant comps for California and NYC are psychiatric hospitals ($350k-$550k annually), nursing homes ($110k-$170k), prisons ($75k-$150k) and developmental disability facilities ($100k-$300k).

Is that amount doing anything? Ha! Look around you.

What if I told you that we could essentially eliminate “problem” homelessness nationwide for half this per capita expenditure?

We all know the problems - not enough shelters, shelters are terrible places to be, and even when they’re not terrible most homeless people don’t want to use them, because they have “no drug” policies.

The big problem with current homelessness solutions is that people want to help, but moralize and put a lot of conditions on that help, and this hasn’t been working. Drug addiction is too endemic, and moralizing and requiring conditions doesn’t work to actually solve the problem. Our downtowns are hostile wastelands. All types of crime are high, and property crime especially.¹

So what’s the solution?

Listen to the frickin’ market! What’s the *actual* problem? A bunch of people want to do drugs, but illegal drugs are expensive (and dangerous), so homeless people forego rent and commit property crime to have enough money for drugs, and refuse to use any of the existing homeless options that might take them off the street. In the process, they make our downtowns unusable, increase property crime stratospherically, and generally crap things up, no matter how much money we throw at the problem.

What we need is Wirehead City.

What is Wirehead City?

We use BLM land deep in the deserts of Nevada to create a big tent city, much like Burning Man.

• Within Wirehead City, drugs and alcohol are both legal AND free.

• Food and water are supplied to you in public canteens.

• You can leave Wirehead City at any time, but you have to walk 20 miles to the nearest town, then take a bus to wherever you’re going.

Basically, legalize everything and put free tents, drugs, food, and booze for anyone who wants out in the middle of the desert. All free! You just need to self select to living in the middle of the desert hundreds of miles away from all the productive people.

There's no way out except walking for 20 miles and then catching a bus somewhere else. But all your friends are back there! Plus free drugs and booze! Also, are you *sure* you'll be able to score drugs back in SF or NYC or wherever, especially now that everyone you know lives in Wirehead City too? Better turn around and stay, to be safe.

Right now, we spend tens of billions and barely help or ameliorate any homelessness at all. Wirehead City will have homeless people from every city in the USA voluntarily flooding in, in entirely self-directed ways. You probably don’t even need to offer bus tickets, they’ll figure out bus fare themselves!

THAT’s the power of incentives.

Nobody wants the current homeless solutions - there’s no demand, because they’re not offering anything homeless people want. If you offer something homeless people actually WANT, the problem solves itself.

And there's no existing residents in the middle of the desert to be negatively impacted or initiate local NIMBY wars against it on federal BLM land. Sounds like a win to me.

Oh, and the cost is trivial relative to the benefits.

Warning - math: skip if you don’t care, but I’m adding up the numbers and making sure they make sense.

Legal wholesale opiates are actually dirt cheap, even extremely heavy users can be zonked out of their minds on $5 a day if it's not illegal. Wholesale cheap alcohol is a little more expensive, maybe it would have to go up to $10-$15 a day for somebody who drinks a liter a day of vodka? Maybe throw in another $55 a day for all the other drugs.² So $75 a day covers drugs. Tents are cheap, let's say $500 a person gets them started with tents and blankets and whatever other minimal infrastructure. Food and water is probably $30 per person per day. We're clocking in at a little over $105 per person per day, plus a $500 one-time expense, for $39k per person per year, or $5.8B, for a ~$7B savings. Prison in California, by the way, costs around $150k per year, vs the $39k a year in Wirehead City.

Let's say selection effects and free drugs crank 150k to 1.5M Wirehead City people willing to live in the middle of nowhere for free drugs. Man, now we're blowing $55B and losing money! Or are we losing money...because a lot of these additional Wirehead citizens would have been in prison costing $75-150k a year, or doing crimes on our streets. Per Scott’s recent post on prisons, the median person who ends up in prison (which is probably a decent proxy for an average Wirehead citizen) does 6 property crimes and 1 violent crime per year when they’re not in prison. Another great point he raises - often less than 1% of people are responsible for the overwhelming majority of crime, with 1% of Swedes responsible for 61% of violent crime, and with 327 individual shoplifters responsible for 1/3 of all the shoplifting in New York City.

What would you like to bet that most of those criminal overachievers will be Wirehead Citizens?

1.5 million citizens is roughly half a percent of the adult population in the US. That half percent will undoubtedly be one of the most criminally concentrated slices of American humanity. A super majority of our “power law” peak criminal candidates will probably be citizens. Imagine the immense declines in crime in every single city in the US wrought by creating Wirehead City!

2023 property crime reached $26.6B in combined property losses. How much of that do you think will be eliminated when most of these people are in Wirehead City, and don’t need to steal to get a fix? Let’s be really pessimistic and say only half, for a $13.3B savings.

According to the National Institute of Justice, violent crime costs us $671 billion annually! Once again, a big chunk of these people will be in Wirehead City, and NOW they have a very big incentive to NOT be violent, because if they get violent, they lose their nice lifestyle with free daily drugs and alcohol with all their buddies. Obviously, reducing violent crime by any reasonable amount, say 10%, more than pays for the entirety of Wirehead City ($67B saved in violent crime more than covers the $55B cost, and that’s before you get to the property crime savings or any other benefits).

So not only will Wirehead City reduce crimes in all the rest of the US, it will likely reduce crimes in an absolute sense. That’s the power of incentives!

What about violence and law enforcement?

Great question! Drugs are legal, violence shouldn’t be. People coming in will be thoroughly searched and metal detected, to avoid weapons.

This is a lower impulse control and fairly drug addled population, so let’s say we need law enforcement on the higher side - the US average is 2.1 per 1k, but let’s say we need more than twice as much, and put it at 5 officers per 1k. Hey, let’s make sure we’re *really* overpoliced - after all, this population probably needs it. Let’s make it 10 LEO’s per 1k, 5x the US average and more than anywhere in Europe. Let’s pay them $300k fully loaded, to make sure we can staff that many and that they’re happy to be there. That brings it to 15k LEO’s total and $4.5B in law enforcement expense.

If we throw in doctors, nurses, food service workers, admins, and support staff at similarly generous salaries we add another $4.4B. Let’s throw in another $1.6B in cost, just for fun. I’ll call it “contingencies.”

Well, we’re up to $68B total, or ~$45k per person in Wirehead City. STILL a huge savings over the $75-$150k per capita prison cost, the current $86k per capita from current homelessness initiatives, and the $100k-$550k costs of psyche hospitals, nursing homes, or developmental disability centers.

And this is STILL more than 100% covered by a 10% reduction in violent crime and the property crime savings, as it’s at $68B annually (versus $81B+ benefit from very pessimistic estimates of 10% violent crime and 50% property crime reduction).

We should note, given the power law of crime-commission, and given that all of those people as Wirehead citizens are going to be heavily incentivized to reduce violent crime (or lose their free drugs) AND extraordinarily heavily policed (with 5x the typical police per citizen), AND have no weapons, we are likely going to see an aggregate drop in murders and violent crime of more than 50-66% across the entire nation. Which is worth $335-$442B annually.

Don’t buy that we’ll get much reduction in property or violent crime, despite eliminating the largest motivation for property crime and aligning incentives? I think we’re winning even in that case. $68B annually is a BARGAIN for actually solving homelessness in every major city in the US. The economic growth alone of suddenly having usable downtowns in the most expensive real estate in the country would likely offset that cost.

An additional benefit - many lives saved.

Since the opiate crisis was “solved” by more or less telling doctors “you need to prescribe 10x fewer opiates, or we’ll take your license,” overdose deaths are up to 100k people per year, as addicts can’t get safe, legal opiates, and all street opiates have become fentanyl due to the lower cost and higher concentration leading to much easier and more profitable smuggleability. Fentanyl has much higher overdose risks than other opiates.³

Before fentanyl and before doctors were forbidden from dispensing safe and legal opiates, overdose deaths were at ~20k per year. 100k overdoses per year is the biggest cause of death for people under 40, and it’s ~80k incremental over what you’d expect.

Since Wirehead City will be dispensing legal, pharmaceutical opiates of known strength, overdose deaths will go way down, likely to the 20k former baseline.⁴ Of those 80k incremental people, some of them will straighten up and leave Wirehead City and get jobs and have kids and be productive members of society at some point. That’s all marginal additional economic and societal productivity that is currently being thrown away every year. That’s also 80k incremental lives saved *per year.*

At the current “$9M per human life saved” valuation (and these are mostly young people), that is an additional $720B in value unlocked annually by Wirehead City.

A golden age unfolds

In the meanwhile, all the downtowns in every major city? Spotless.

Crime in every major city? Plummeted to 1/3 the usual levels. The money you’re spending on Wirehead City police is 9x offset by just the reduction in crime and police officers needed in every major city in the US!

All the productive people who have jobs and would like to use their own downtowns for commerce and recreation? It's actually possible now!

City downtowns bloom in a flourish of gentrification.

Crack houses become trendy restaurants. Boarded up convenience stores become fancy craft brewery drafthouses. The economic growth from these things happening in every major downtown also offsets the ~$62B yearly cost of Wirehead City.

And everyone is happy! Both sides are “winning!”

The homeless people have free, safe drugs, the productive people have usable downtowns and craft breweries, the pressure is off in the prison system and we can imprison genuinely violent offenders at a higher rate. We enter a golden Natufian age of bliss and harmony on all fronts.

Let’s just recap the costs and benefits, to really see them side by side:

Zero out or reduce whichever you don’t think are true, the benefits still massively outweigh the costs, and “4x higher benefits than costs” is a pretty conservative estimation

Arguably, the benefits go up to 22x the costs, if you include the lives saved, the likely full magnitudes of 50-66% violent crime reductions and a 66% property crime reduction, and take into account that those crime reductions would let us reduce expensive urban police in the rest of the US.⁵

Let’s not forget that more than 80% of police hours are wasted on overhead and traffic stops, and this is while rapes have a ~30% closure rate, murder and aggravated assault a ~50% closure rate, and property crime a 10% closure rate, so any reduction in crime in US cities allows us as a society to reapportion the remaining police hours sensibly and make progress on our dismal crime closure rates.

For the gentrification and productivity estimations, along with a quick FAQ covering common questions and objections, see here.⁶

I think it’s worth briefly pointing out that I’ve used maximally generous, “always err on the side of higher headcount, salaries, and expenses” methodologies here, and the benefits are still 4x-22x the costs.

But what about the non-problem homeless?

1. They’ll also benefit with the problem homeless gone - downtowns are safer, and property and violent crime is lower.

2. The money saved with this cheaper, more effective per-capita solution leaves more money available to help non-problem homeless.

3. The GDP growth, lower LEO expense, and other benefits also frees up money we can allocate towards non-problem homeless or other public goods.

4. Public sentiment and willingness to fund initiatives will likely be higher when our downtowns are actually usable and there aren’t problem homeless making public life difficult.

Why aren’t we doing this?

The primary reason NOT to do this is moral high-handedness about not wanting to give slackers free drugs. But the current “solutions” to this (prisons and current homelessness initiatives) cost more than Wirehead City would cost, do NOTHING to ameliorate the problem homeless, leave our downtowns unusable, leave our prisons overcrowded and extremely expensive, and leave at least 80k incremental people dead annually. It seems to me like we can get a LOT of benefits on a lot of fronts simply by relaxing one “we shouldn’t give drugs to slackers” moral opinion.

Yes, public opinion is a hurdle to overcome. But if we trial this for just one city, and show the before / after of actually eliminating problem homeless, greatly reducing crime, and having a usable downtown, it’s a good bet that people will come around and embrace the practical benefits for all cities.

As we’ve seen in this analysis, giving people free drugs is actually an overlooked and extremely under-rated sorting mechanism that we can use to separate positive and negative externality populations, concentrating and amplifying positive externalities and productivity in our large cities, which are now - and always have been - the primary engines of economic growth. Then you get usable downtowns, lower prison populations, much lower urban crime, smaller urban police forces, and every other benefit on top of it.

If you think so too, spread the word, link people to this post when homelessness comes up, get it in front of policy people. Ideas have to get out there and be debated and compete in the marketplace of ideas to have a chance at becoming reality.

If you don’t think the benefits outweigh the costs, I look forward to hearing why in the comments.

Credit where credit is due - I came across the original Wirehead City idea here, and all credit is due to George Hotz (yes, noted hardware hacker and entrepreneur geohot), I’ve just expanded his idea by putting some numbers to it and articulating the incentives argument.

1. Property crime losses were only $15.8B in 2020, and are now $26.6B in 2023

2. Wholesale legal Adderall is 25-50cents per 20mg. Wholesale surgical cocaine in sterile saline is about $50 a gram. A gram is a LOT if you’re injecting it. All these prices are before “immense volume” discounts that we could almost certainly get. And citizens can allocate their $75 a day however they want from the menu available, and barter or transact with other people, there’ll be a whole Wirehead economy.

3. People overdose on fentanyl for three primary reasons:

1. It's the only illegal opiate around because it's much more compact to smuggle and is cheap - one kilo can serve an entire city for a few weeks to a month.

2. The difference between "feeling good" and "OD" are 1-2mg, which is tiny and hard to eyeball, and there is no such thing as standardized strength and potency in the illegal markets.

3. Fentanyl contamination in coke and speed, or "chocolate chip cookie" effects of clusters in otherwise cut opiates kills people, because drug sellers bag other drugs on surfaces with fentanyl, or don't mix well enough, etc.

In other words, the vast majority of the deaths are accidental, and involve unknown quantities and strengths, poor practices and mixing, and fiddly dosing.

4. Similar “safe supply” programs in Canada where they pass out free heroin or dilaudid to homeless addicts to use those instead of fentanyl reduced overdoses 5.5x (nonfatal, because they had no fatal overdoses during the program).

Lew et al. The impact of an integrated safer use space and safer supply program on non-fatal overdose among emergency shelter residents during a COVID-19 outbreak: a case study (2022)

In the 4 weeks preceding the program, the rate of non-fatal overdoses was 0.93 per 100 nights of shelter bed occupancy. During the 26 days of program operation, there were no overdoses in the safer use space and the rate of non-fatal overdoses in the shelter was 0.17 per 100 nights of shelter bed occupancy. The odds ratio of non-fatal overdose pre-intervention to during intervention was 5.5 (95% CI 1.63–18.55, p = 0.0059).

https://link.springer.com/article/10.1186/s12954-022-00614-8

5. Just off the cuff, there’s 700k police officers nationwide, and if crime was at 1/3 of current levels everywhere outside of Wirehead City, we could get rid of around 400k of them, which at $100k fully loaded would be another $40B in savings.

6. Gentrification value calculation:

Took current SF commercial vacancy percent, current average commercial rent, assumed vacancies would be filled at average rents, extrapolated percent of GDP to Los Angeles, Seattle, Portland, New York City, and Washington DC GDP’s.

Then given those vacancies would now be filled with businesses, assumed SF business tax receipts would increase by the vacancy percent, then divided by average business tax rate to get total annual incremental economic activity.

Productivity value calculation:

Assumed GDP’s in San Francisco, Los Angeles, Seattle, Portland, New York City, and Washington DC would increase by 1% via the productive people not having to dodge syringes and human feces and deal with constant car break ins, shambling fentanyl zombies, etc.

Don’t buy either of these (admittedly lazy and quick) estimates? Zero them out! Still worth it! And remember, any reasonable expectations of reductions in violent and property crime are something like $450-$460B a year in benefits. ($20B from property crime reduction, from today’s $26.6B, and $420B in violent crime reduction, from today’s $670B).

What is $50-$100B annually comparable to? We already waste about this amount on totally trivial, valueless things:

• Oil and gas subsidies and tax breaks, $50B

• State and local tax breaks and subsidies for stadiums: $50B

• Foreign aid: $50B

• Annual upkeep, utilities, and security on unused federal buildings: $50B

• Improper or fraudulent Medicaid, SNAP, or unemployment payments: $100B

• Maintaining unneeded and underutilized military bases: $50B

Quick FAQ and frequent objections:

What if even more than 1.5M want to be Wirehead citizens? This would be GREAT news! Given the benefits are between 4x-22x the costs, this indicates that every incremental Wirehead citizen is a massive win for productivity, crime, and policing for the rest of the US, and that there’s a lot of headroom such that the sorting driven by each incremental Wirehead citizen opting in is almost certainly likely to be net positive, up to at least 10x the 1.5M estimate. Also, need I remind you that both sides are happy in this schema? Each marginal immigrant to Wirehead is happy to go, and the rest of the US is happy that they selected into Wirehead citizenship, because of all the positive externalities for the people remaining in the rest of the US.

What about medical care? For practical reasons, we should also be dispensing free Narcan, syringes, antibiotics, psyche medicines, and birth control to whoever wants it (still rounding error costs). If somebody has a serious medical issue, there can be an ambulance service to the nearest clinic or hospital that allows them to skip the 20 mile walk. But let’s not try to gloss this, a lot of Wirehead residents are going to be dying. This is an unhealthy population with unhealthy habits and a lot of comorbidities. A lot of them are dying *today,* they’re just doing it distributed across the urban downtowns of the US, and now they’ll all be concentrated in one place. But at least they died relatively happier, surrounded by friends, and not going through withdrawal. Would they choose that, versus dying in an alley somewhere while going through withdrawal? Almost certainly.

What about body disposal / funerals? Cremation machines cost $100k and use $10 worth of fuel to cremate somebody, that’s rounding error expense wise. Friends at Wirehead are allowed to gather and do whatever funeral services they wish privately.

What do we do about babies born in Wirehead City? I'm tempted to rejoinder with "whatever we do TODAY when homeless people or addicts have kids," which I would bet is “nothing.”

But, this IS an opportunity to do better. I suggest some sort of formal "Pregnant? We'll get you clean and give you a nice hospital birth" sort of program where they can put the babies up for adoption if they want to go back to Wirehead City, or keep them if they stay in the rest of the US, that’s publicly messaged within Wirehead City.

This does give me the chance to trot out the fact that opiates are actually a pretty effective birth control in most primates, and that they substantially reduce female human fertility. Also, we’d be passing out free birth control, but of course adherence will be less than ideal in our populace. We could also do something like “mandatory IUD’s” before admission - at the least, we could offer free IUD’s for anyone who wants them.

What about whatever the nearest-to-Wirehead town is, aren’t they going to be pissed? I lean towards “maybe not” because we’ll likely recruit heavily from people in that town for the $300k cherry police jobs, and we can message this strongly. But sure, they might be pissed. This will always piss off *somebody,* but doing it this way minimizes that, because the nearest town to actual Black Rock City is Gerlach, with 100-200 population, and the nearest town to Wirehead City is going to be similar in population. If they’re really pissed and we wanted to make them happy, we can “stuff their mouths with gold” as Aneurin Bevan famously put it - you could literally give every citizen of the town $100k each and that’s still rounding error. It's worth noting that their town population, real estate, and economy is going to boom significantly with Wirehead City employees and ancillary businesses and services drawn to the area, so the town is almost certainly going to end up pro Wirehead City by default.

What about sewage and waste? Scaling up from current Burning Man numbers, we’ll need to have about 20-35k porta-potties and an emptying crew that goes around emptying them at least once a day. So that’s $35M for the potties, another couple million for a bunch of septic pump trucks, and probably another couple million a year in salaries and expenses. Rounding error. Also might be worth it to build some actual sewage treatment plants at those numbers.

By creating Wirehead City, aren’t we guaranteeing most of them will never “get clean?” Yes, but this is already the state of affairs. People who self-select into treatment and strongly desire to get clean only have a ~30% “getting clean” rate for alcoholism, and an ~18% rate for opiates, and both of those are in “non homeless” populations. People who are forced into treatment programs as an alternative to jail have a 6% “getting clean” rate for alcoholism, and I couldn’t find numbers, but it’s probably a safe bet opiate rates are around 5x lower too in those populations.

Another thing to consider - for those who actually WANT to get clean, it becomes much easier - you leave Wirehead City and go anywhere else, and you’ve cut off all the “bad influence” friends and contacts in your life and made it MUCH harder and more expensive to score. Consider the fact that Wirehead City will put many drug dealers out of business in most cities in the US. It’s a clean break on all fronts, and would probably increase the success rates for people leaving Wirehead City with the intention of getting clean.

Also, any money or programs targeting these “Wirehead emigrants” will likely be noticeably more successful due to those factors, whereas any money or programs now are at minimum 70-80% wasted.

18% cite: Hser, Y (2007) Predicting long-term stable recovery from heroin addiction: Findings from a 33 year follow up study. Journal of Addictive Diseases. 26(1), 51-60.

30% and 6% cite: White et al. (2012) "An analysis of reported outcomes in 415 Scientific Reports, 1868-2011"