Where should new cities go?

A harder problem than it sounds like

Some people have grown bored or frustrated with the YIMBY conversation and think that instead of dealing with YIMBY/NIMBY politics, we should build whole new cities from scratch. I’m not against this; I have spoken favorably about the California Forever plan to build a new small city in Solano County, and I wish them luck.

That said, the people who are most enthusiastic about this concept strike me as falling into one of two partially overlapping buckets.

One is people who think, wrongly, that the “new city” idea solves the political problems around YIMBY/NIMBY fights. Here it’s useful to look at the actual history of California Forever, which involves a ballot initiative that had to be withdrawn, another withdrawn initiative, and what now seems to be a protracted engagement process with the local elected officials. Which is to say that whatever the merits of the proposal (and I think they are considerable), it is, at the end of the day, yet another real estate development plan that has to run the procedural gauntlet.

The other is right-wingers who are aware that regulations are economically harmful and that housing is the single largest sector of the economy and that therefore anti-density regulations are really bad, but who also acknowledge that there is right-wing culture war upside in opposing YIMBYs. This is how Donald Trump winds up being a candidate who is “protecting American communities from excessive Federal overreach and preserving local decision-making” while also proposing new “freedom cities,” where somehow none of this will be a problem.

But conservatives really need to think about this like a normal economic policy program. It’s true that faced with a costly regulation, there is often some second-best workaround. But that doesn’t mean the cost has gone away. The most valuable place to build housing is where people most want to live. Telling them “go live in a freedom city instead” is better than them being homeless, but it’s still central planning and it’s still a massive distortion and economic loss.

What are cities even for?

The very first cities in Mesopotamia are essentially administrative centers for despotic systems of exploitation. But for most of recorded history, cities have been intimately connected with transportation logistics.

A city typically exists where there’s a port. Or at the fall line of a river. Or near an important river crossing. Later railroads come into the picture: A city might occur where a railroad intersects with a navigable river or where multiple railroad lines converge.

One reason these transportation logistics hubs are important is that they create demand for non-farming labor operating the logistics. Another reason is that being a transportation hub, it is by definition a place that it is relatively easy to bring food to in order to feed non-farmers. The transportation hub is also a logical place for the concentration of administrative functions since it’s a relatively easy place to get to. And when most people are walking and most goods are carried on carts, there is an extreme premium on being located right by the hub. Boats and trains move dramatically faster than a person on foot, so there’s a strong incentive to get really close to the hub.

And then, of course, big cities built out local train networks to create subsidiary hubs of activities around the metro or subway stops.

But how does this apply to modern times?

I’m sure there are parts of the American coast where one could build a new port. But there will already be people living there who probably don’t want a giant new port built. You could override them, of course. But if you’re willing to override local control, there are tons of existing coastal communities that already have robust housing demand and could become large metropolises. Whether we’re talking about Portland, Maine or Santa Barbara, California, it’s not hard to find places where this could work, but now we’re just re-inventing YIMBYism and the point of “new cities” was to not do that.

Also, why would you build a new port? Increasing America’s port capacity is a good idea, but our existing ports have very low productivity due to deliberate anti-growth policy choices that even the Trump administration supports, thanks to his cultural affinity for longshoremen.

The larger issue is that transportation technology has really changed. You can’t build an airport in the middle of nowhere and expect a city to grow up around it. The beauty of airplanes is that they allow direct connections between arbitrary locations. Even worse for new cities, cars allow the same thing. All across America you can find little clusters of commerce near highway exits. But these don’t become “towns” in the traditional sense of a town growing up at a crossroads, because the people who work in the little clusters of commerce drive to work. If you had one of those highway exit commerce hubs, but for some reason everyone had to walk to work, then there would be a bunch of housing right near the hub. And then, because a bunch of people lived there and didn’t have cars, the commerce hub would start sprouting services aimed at serving the neighborhood, not just people driving by. If your “random cluster of commerce near the highway” exit is close enough to an existing population center, it may grow quite large.

But this is suburban sprawl, not new cities. If you take I-10 out of San Antonio, you go through suburbs like Leon Springs and Boerne. By the time you get to Nelson City, the highway exit is more like a random cluster of commerce. But if it gets bigger over time, that will be an outward extension of the existing metro area, not a new city.

That’s fine. I don’t have a problem with the ongoing outward expansion of San Antonio, just like I don’t have a problem with the idea of upzoning in-demand coastal communities so they can turn into large cities. But the “new city” concept keeps collapsing into old concepts rather than offering anything genuinely new.

Another idea I’ve heard is that a new university could perhaps anchor a new city, updating the concept of the transportation hub for the 21st century. I’m a bit skeptical, given the falling number of teenagers and the rise of AI, that anyone is going to be interested in founding a new university. But beyond that, we know what a city anchored by a university looks like — it’s a college town like Ann Arbor or Charlottesville. These places are very nice, but they have their own NIMBYs and growth control politics. Of course, you could stipulate that your new college town won’t have those things. But why won’t it? We do have some relatively growth friendly college towns like Knoxville and Austin, which is great. But this again suggests that we don’t really have an alternative to doing YIMBY politics. You can do it in the college towns we have or you can try to start a new one somewhere and face the same problem.

The paradox of emptiness

Politically, one problem is that while it may seem like a place with fewer incumbent residents would have fewer NIMBY problems because there are fewer NIMBYs, it never seems to work that way in practice. The thing about cities is that the people who live in them have mostly opted in to living in cities. By contrast, the people who live in small towns or rural areas have opted into living in small towns and rural areas.

The California Forever proponents are finding that even in “the middle of nowhere,” there are, in fact, people living. And while the NIMBYs of Solano County may be less numerous than the NIMBYs of Palo Alto or San Francisco, they are if anything more skeptical of density and urbanism, because these are people who live in Solano County.

Of course, there are places in the United States that are genuinely devoid of people.

But an uninhabited wilderness or natural landscape is precious. Why would you specifically encourage new housing development there? It seems perverse. There are a bunch of western members of Congress, including Mike Lee and Steven Horsford and Ruben Gallego, who are interested in opening up federal land for housing development. This makes sense to me, but to the best of my knowledge, they’re all basically talking about extending the suburbs of existing cities.

Which makes sense. The cost-benefit of building more suburbs of Salt Lake City or Provo seems much better than trying to plop a new city down in the Grand Canyon. Right now, federal land is used for a mix of preservation (a minority of it) and natural resource exploitation (ranching, mining, timber, etc). Reasonable people will disagree about how best to strike a balance around shifting some of that to new housing. But I think anyone who looks at it will reach the conclusion that the low-hanging fruit here is non-conservation land near existing in-demand communities. If there are available parcels really far away from anything, that’s where you want to think about doing more nature preservation, not creating a new city.

Which isn’t to say that suburban sprawl is the be-all, end-all of American housing, just that the alternative to sprawl is infill. Positing a new city doesn’t actually resolve very much, and it also doesn’t offer the same benefits as letting people live where demand is highest.

The only way out is through

The version of tech-right fuzzy thinking on this that I found most comical was Mic Solana and Palmer Luckey recently getting gassed-up about the idea of turning Guantanamo Bay “into Liberty City, an American Singapore of the Caribbean … a gilded mega fortress broadcasting capitalism’s bounty on the coast of a failing communist state … any Cuban who makes it there — by foot, boat, or bike — gets asylum and a work permit.”

This sounds great to me… but it’s just Miami!

Or at least it was for most of the Castro regime’s lifespan. But then Trump decided he wanted to end America’s historic generosity to people fleeing Cuban Communism. The Biden administration, thinking conservatives might be sympathetic to the plight of those fleeing failed leftist states in Latin America, created the CNHV parole program to do another version of this. Again, Trump killed it. I was never naive or ignorant enough to believe that American Hispanics had dogmatically pro-immigration views. But I have been genuinely surprised over the past ten years by the lack of backlash in the South Florida Cuban-American community to the Republican Party abandoning its historic commitment on this specific issue. If Solana and Luckey think the new Republican position here is wrong, I would love to hear them say that and encourage them to encourage Trump to rethink his approach.

Displacing this to Guantanamo to turn it into a lib-owning scheme rather than just “I actually disagree with Trump about this” is fine for the sake of a post, but it doesn’t actually answer any of the questions about uncontrolled migration flows.

When people start to show up and they have nowhere to live, what happens? Who’s administering the tent city? Who’s bringing in the food? What if there’s a cholera outbreak? Who are the police? I guess the idea is to take advantage of Guantanamo’s extra-territoriality to make this a libertarian paradise where there are no rules and regulations. But is Donald Trump going to allow products built with low-wage refugee labor to be imported into the United States?

I love the idea of conservatives rediscovering that it’s good when people are able to leave failed states. But it’s also true that chaotic influxes of refugees create a lot of logistical problems. None of those problems are made easier by displacing them onto a remote military base with no relevant infrastructure.

The cities we have

Speaking of infrastructure, the new city concept that resonated most with me was Arpit Gupta’s point that even more people could live between New York and Philadelphia, because there’s already a ton of infrastructure and people nearby.

But of course, if you take that red circle seriously, what he’s proposing is basically that we should do massive YIMBYism on the towns of Princeton, Plainsboro, West Windsor, etc. I completely agree with this. Not only is existing housing demand in this area already quite high, it already has a major university and a solid commuter rail connection to New York City (and the infrastructure needed to create one to Philadelphia would be minor). The local community, though, would murder you if you tried to do this. In principle, I think the New Jersey state legislature should just steamroll them. But again, we’re just back to YIMBYism.

The one original thing I can say here is that just down the road from Princeton is the city of Trenton. Trenton is a smallish, poor, high-crime city with bad schools that is surrounded by a bunch of richer communities. We have, unfortunately, a lot of these places in the northeast (Baltimore, Camden, etc). These places are interesting, because like the affluent suburbs that surround them, they are already reasonable commutes from high-paying jobs. And they already have urban infrastructure and street grids. They also tend to have initiatives like this Trenton program to demolish vacant properties, so I think infill and investment would be reasonably welcome, if you could address the safety and education fundamentals. It seems like it should be possible to work out some kind of deal that involves a “surge” of state efforts on these issues in exchange for pro-building regulation and a commitment not to downzone once conditions improve.

Thinking the details through is hard (a post for another day), but it leaves me with the conclusion that the best place for “new” cities is probably in the footprint of existing cities that are distressed for one reason or another.