Bottom line, up front: homelessness isn't about alcohol and drug treatment, or mental health, or better weather, or moral fiber. Ultimately, it's just about having enough housing. I'm going to illustrate this with an anecdote, then going to illustrate it with the data, and finally talk about how to attack the problem.
First, the anecdote.
My story is about the guy who lived across the street when I was growing up in San Francisco. He was a Russian painter, and his name was Bogdanoff.
Bogdanoff wasn't a landscapes-and-portraits painter, mind you. He was a make-your-walls-a-different-color painter, a blue-collar guy. Bogdanoff was a great chess player - he'd routinely wipe the floor with my dad. Bogdanoff's people had fled the Tsar for San Francisco in the early 20th century. They were called the "milk-drinkers," because they were famous for drinking lots of milk, and no alcohol.
Ironically, Bogdanoff was never sober. Each morning, he'd walk to Pete Chiotras's grocery down the block, and he'd buy a case of beer for breakfast. Usually it was Mickey's malt liquor, but sometimes it was Budweiser instead.
Bogdanoff is dead now. He's been dead for a quarter-century, because he drank himself into an early grave. It's guys like Bogdanoff who become homeless these days. If Bogdanoff were alive in 2021, he'd be on the street. Whatever it was in his brain that made him an alcoholic would've gotten even worse. He'd be a public nuisance and a public health hazard, sleeping in a doorway and shitting on the sidewalk.
But Bogdanoff didn't end up on the streets, because there was a lot more room for error when I was a kid. It was cheaper and easier to keep a roof over your head, so Bogdanoff was mostly a hazard to himself.
Housing costs are the elephant in the room when you talk about homelessness. You have to have a non-stupid housing market so even dysfunctional alcoholics like Bogdanoff can afford to have a roof over their heads.
Now, let's go into the data.
It shouldn't surprise you where homelessness is the worst, because it reads like a list of places where housing is stupidly expensive: Los Angeles. San Francisco. Seattle. New York. Places where nobody can afford to live, unless you moved there 20 years ago. Raw data is here.
Now we'll go back to the other end of the spectrum, the cities where there's the fewest homeless. The big cities on the list are Detroit, St. Louis, Miami, Houston and Atlanta.
Pause for a sec. Think of what these cities are like. Detroit and St. Louis are post-industrial messes, and Miami, Houston and Atlanta are all Sun Belt sprawl.
Let's compare this list to the common reasons given for why there are so many street people in LA, San Francisco or New York.
- "We need better-funded mental health systems." Nope. Texas, Florida and Georgia's mental health systems are run by cheapskates. (It's page 101 of the PDF.) The state of California spends $187 per capita on mental health. New York spends $278. Georgia? $72. Texas? $51. Florida? $47.
- "We don't have enough substance abuse treatment." Not really. Houston has similar alcoholism and overdose rates as LA, but 1/6 the homelessness.
- "We need competent local governments who can connect the homeless with jobs and train them." Nope. Detroit and St. Louis's local governments are incompetent, decadent, and dysfunctional, and they can barely keep the lights on.
- "It's because there's good weather." (A California excuse.) Nope. Miami's a tropical paradise and it doesn't have a homelessness crisis. New York has a massive homelessness problem and NYC weather sucks. Hard.
- "They're all from out of state." (A California excuse.) No. Four in five Los Angeles homeless have lived in LA for at least five years.
- "The homeless need more moral fiber." If you think people in Miami, Houston and Detroit have more moral fiber than in LA or NYC, I have a bridge to sell you.
Now, I'm not saying that these things can't help. They absolutely can help, but only at the margin. The best way to reduce homelessness is to just to have enough homes, so that even drunks, drug addicts and schizophrenics can figure out how to make their rent.
So you need to build more housing. But that takes years. But what can we do in the meantime?
1: Make it profitable for private developers to build more rent-controlled housing.
The most effective tool for this at present is called the density bonus. This means, if a developer builds rent-controlled low-income housing to their private building, they can build more market-rate apartments than the zoning would otherwise allow.
But the details of your density bonus program matter a lot, and it requires some pretty careful tailoring to the local market. For example, LA's Transit Oriented Communities program works great in neighborhoods like Koreatown because land is relatively cheap and the zoning is fairly liberal. In contrast, Santa Monica's downtown density bonus program is not generous enough, so many developers opt for smaller projects rather than putz around with the arcane bureaucracy of density bonus programs.
Regardless, the bureaucracy should be made cheap and fast, and density bonus programs should be streamlined and expanded to bring more apartments online faster.
2: Allow the conversion of existing hotel space to apartments.
There's a ton of old motels strung along main roads which are barely viable as hotel space, but are still viable as apartments. You probably know the ones I'm talking about, even if you've never stayed in one. Here's an example from south LA, another from East New York, Brooklyn, and another from North Oakland in the Bay Area. A lot of these are being used as de facto apartments already, rented out to people on the edge of homelessness. These conversions should be made legal and encouraged so the people living there get tenant protections, and the owners get incentive to invest in fixing their places up.
In the past, this type of conversion was normal. All the old hotels in the Tenderloin, San Francisco, Downtown LA and the Bowery were converted to bargain-basement apartments.
You, dear reader, probably wouldn't live here if you had better options. But you'd certainly take it over a tent or living in your car.
These types of conversions are making a comeback in the present. One of the clever things that the state of California did during the start of the pandemic was to buy up these types of motels and do quickie conversions into affordable housing. This initiative, called Project Homekey, has worked really, really well, creating thousands of new affordable apartments at 1/3 the cost of traditional affordable housing.
There's an added bonus: these old motels tend to have decent-sized parking lots, so there's a ton of flexibility to expand if necessary. The example I showed in Oakland has a 3900-square-foot parking lot in front.
That's enough space to put another 12 apartments in the parking lot if the zoning allowed it, even if you limited yourself to cheap, fast 3-story wood construction.
The drawback of this is simple: many of the owners just aren't willing to sell.
3: Temporary shelters on underused land.
It's a gigantic pain in the ass to build permanent homeless shelters, because they get nitpicked to death by nosy neighbors. Because of that, you end up with people camping on the streets. Absent a major reform to allow the construction of cheaper, faster shelters, you need to have temporary ones. And to build temporary ones, you want to start thinking creatively about using existing land.
Los Angeles, of all places, has a good example of this in North Hollywood. There's a small, awkward-shaped strip of land between the Orange Line busway and the Chandler Boulevard which is too small to be used as parkland, and too awkwardly-shaped to be developed commercially. It was just sitting there, empty. And there, the City built a few dozen prefabricated tiny homes on that awkward piece of land. They're nothing fancy - 64 square feet, a couple twin beds, and communal sanitary facilities - but they have doors with locks, and they provide access to treatment services.
If you look closely, plots of land like this are everywhere, especially close to freeways. This is a good example, at 120th and Western, in unincorporated LA County. It's two acres of state-owned, vacant land that's just sat empty there since the 105 was built. It doesn't even have a parcel number from the LA County Assessor, but it's there.
These projects could and should be massively scaled up. And if you build mobile home parks or tiny homes for the poor in large numbers, you get economies of scale.
4: Safe parking lots.
Homelessness isn't black and white. It's not just an apartment or a tent. There's also a large number of people who are just plain poor, and living in a motor home or in their cars because the rent is too damn high.
In LA, it's about a quarter of the homeless who live in their vehicle.
Anyone who has a car in a major city knows how much of a pain it is to find somewhere to put it overnight - and the stakes are even higher if you live in your vehicle. Getting towed means you might be sleeping rough, maybe for good. But if you're not going to get towed, at least you can get one thing off your shoulders.
The Veterans Administration campus in West LA is a good example of how this might work, in the absence of sufficient shelter space - the VA provides safe parking spaces for homeless veterans to park RVs, and provides access to treatment services.
It's not a substitute for shelter space or enough housing, but it's better than living on the edge and a small backstop.
5: Make Section 8 universal like food stamps and Medicaid/Medi-Cal.
Section 8 is a federal government program which provides rent assistance to the poor. People with Section 8 pay a fixed percentage of their income for rent, and the gov pays the rest. It's a pretty standard part of the social safety net, like food stamps or Medicaid. But there's one critical difference. Congress has never fully funded Section 8, so maybe one in five people who're eligible for Section 8 rent support actually get it. It's kind of shitty and Darwinian, because it means that there's a massive waiting list for Section 8 vouchers - often, it's years-long.
It would cost about $60 billion a year to fund Section 8 for everyone in the US. This sounds like a ton of money, but it isn't lot of money for the Feds: it's 1/20th of what the federal government already spends on health insurance for the poor and the elderly.
These are only temporary fixes, though.
But all of these things can only really help things at the margin. Unless you have enough apartments for everyone, you're not going to be able to fix the problem. Our laws, and our governmental system, just aren't designed to fix the homelessness crisis. It's quite the opposite. And that's why places with high taxes and lots of government infrastructure like LA and New York are failing at the crisis - while Detroit, Houston and Atlanta are doing okay.