The case against municipal fragmentation. By Matthew Yglesias

Read time: 9 minutes

A quick bit of housekeeping before today’s column: This reader column was originally scheduled to run last Friday, but we rescheduled following the debate.

This week, we’re traveling to Portugal, and between that and the 4th of July, Slow Boring will operate at a slightly slower pace for the week. Don’t worry, we have great content lined up, including some of the results from the reader survey, a piece from Ben looking at the complicated history of osteopathic medicine, and an old fashioned mailbag. Plus tomorrow, July 2, I’ll be doing a chat for paid subscribers at 3pm Eastern.

My big recommended reading this week is the Economic Innovation Group’s report “The American Worker: Toward a New Consensus.” There’s a lot of nuance and back and forth in it, but I think this chart is the most importance piece. It makes the point that the problem of “wage stagnation” is something that we mostly left in the past a long time ago.

The question of why, exactly, real wage outcomes were so dismal in the 1972-1993 era is very interesting. It probably has to do with oil shocks, with the difficult fight against inflation, with the massive entry of women into the labor force, and maybe something to do with the sheer size of the Baby Boom cohort. But it can’t be the case that NAFTA, trade with China, and “neoliberalism” are to blame, because it happened before that stuff!

Good news this week: We’re saving lives with “unusable” kidneys, the first big commercial contract for geothermal energy was awarded, and people are feeling fewer negative emotions.

Comment of the week from JCW, on elite misinformation: I wrote a book adjacent to this topic, Radiation Evangelists, about the early development years of radiation therapy in medicine, and I have an add-on to this column: a lot of times people doing this kind of misinformation have functionally managed to talk themselves into believing what they are pitching.

Most of the early radiation innovators that I wrote about ended up dying of cancer or other radiation-induced maladies. I expected the story to be one where people who didn't know better died of something they didn't understand. But what I found in the research is that users recognized--and documented!--the risks more or less immediately. It's just that they then proceeded to talk themselves into alternate explanations. A lot of patients were harmed as a result, but the therapists bore the worst of it; more or less an entire generation of men and women who were enthusiastic early adopters ended up dead in a pretty painful and awful way.

All of which is simply to say that I think "elite misinformation" is an even harder problem than this column suggests, because motivated reasoning is a hell of a drug. Even well-meaning humans armed with reasonable information are highly prone to talk themselves into believing wrong stuff, and they will do that EVEN WITH the counter information right there in on the table. And someone who has lied to themself first is hard to disabuse of a notion, because 1) they do not "know" that they are lying, and 2) admitting that they are wrong now carries a component of shame and disappointment to go along with the embarrassment.

It's just a really hard problem.

Our question this week is from Drew Kudlow: Almost every American metropolitan area has dozens of separate municipal governments, a couple county governments, and a ton of school districts, parks districts, water districts, etc.

Is this degree of municipal fragmentation a problem? New York City is less fragmented than say, Los Angeles or the Bay Area, and NYC seems generally more capable when it comes to addressing big challenges. Should reformers try to push for more consolidated cities?

This is a complicated question to address, because there is a multiplicity of fragmentations. One is that sometimes the geographical units of government are very small, either in area or in population or in both. Another is that sometimes governing authority is very fragmented. If you live in Ingram, TX, for example, some of your local government functions are performed by the City of Ingram, others by Kerr County, others by Ingram ISD, etc.

I think that there’s really just one problem with spatial fragmentation, though it’s a serious one: The more fragmented your geography, the more likely you are to undersupply housing.

The fragmentation of governing authority, by contrast, seems very bad to me, because it makes it much too hard for citizens (and the media) to actually track who is responsible for what.

The New England governance model is in one sense highly fragmented, but in another appealingly transparent. The key feature of the New England form of government is that towns (and cities) “fill the map.” With the exception of uninhabited areas of northern Maine, every inch of New England is within the boundaries of an incorporated town, just as every inch of every state is inside the boundaries of a county.

In this system, counties essentially serve a pure judicial function. Hancock County, Maine has a prosecutor and a sheriff and a courthouse and a jail, and that’s what the county does. Everything else is handled by the town, because even if you don’t live “in town” (as indeed most people do not), everyone lives inside the municipal boundaries of a town. And I think the big virtue of this system is that it’s pretty easy for the residents to understand. If you live in Massachusetts, you’ve got your “state stuff,” your “local stuff,” and then you do jury duty at the county level. That’s it.

Beyond being easy to understand, I think the “teeny tiny towns” model has some governance benefits. The main one is that there is a pretty direct link between taxes paid and services received. Since the town is small, you get what you pay for in a more literal sense. This means people are sometimes more willing to tax themselves in exchange for good schools, parks, libraries, and public safety. But also that people are more likely to notice and get mad if the local government is squandering their money on bullshit.

The big downside of fragmentation is on housing, which is how Massachusetts manages to be nearly as under-housed as California despite dramatically worse weather.

People who like to argue on the internet always overindex on principled accounts of housing policy disputes versus institutional ones. The economic benefits to the Commonwealth of Massachusetts of creating abundant housing throughout the Greater Boston area would be quite large. There would also be some downsides, primarily worse traffic and also various issues about school capacity and other infrastructure. If you instead consider the proposal to create abundant housing just in the town of Wellesley, that has basically the same downsides for Wellesley residents as creating abundance everywhere. But not only are the economic benefits more modest, the regional benefits don’t necessarily accrue to the residents of Wellesley. Some Wellesley residents would be YIMBYs if you ask either the Wellesley-specific or the metro-wide question. Some would be NIMBYs either way. But “metro-wide housing abundance has benefits that exceed costs, while Wellesley-specific housing abundance does not” is a perfectly coherent worldview. In fact, that’s where the whole idea of Not In My Backyard comes from. Concentrating governance powers in small units generates more small, self-interested thinking, which is probably narrowly good from the standpoint of running a cost-effective parks department, but it’s bad for land use policy.

The other kind of fragmentation that matters is the actual fragmentation of authority. Grow SF published a useful report earlier this year showing that even though San Francisco has the superficial attributes of a “strong mayor” city, in practice most executive agencies that people care about — notably police, planning, and public works — are actually in the hands of commissions whose members are mostly appointed by the city council. Combine that with an independent school board, and the mayor is not really in charge of the bulk of city government.

San Francisco is a bit of an outlier here, but it’s an outlier primarily in terms of having the external appearance of being highly consolidated.

In Los Angeles, lots of local governance functions are performed by LA County. The school district is independently elected. Because Los Angeles is the name of both the city and the county, just from looking at the name of something like “Los Angeles Department of Water and Power,” it’s not obvious whether that means the city or the county (in this case, it’s the city). The municipal boundaries of the city do not line up well with a commonsense understanding of what is and is not part of Los Angeles. And city agencies — including the aforementioned school district and water department — do actually provide some services to some unincorporated areas.

I think the main problem with this form of governance is not that the units are too large or too small, but that the landscape is too confusing. People don’t realize how powerful the five-person LA County Board of Supervisors is. A tiny random city that nobody knows exist, like Bell, can become the locus of horrible corruption scandals. A common pattern that you see in San Francisco and most other cities is that the mayor is consistently more moderate than the council, since the mayor is a visible figure who citizens hold responsible for outcomes. And when you sap the office of power, special interest groups that pay tons of attention wind up capturing everything.

To my mind, the problem underlying a lot of bad decisions here is a kind of dual fallacy:

The idea that a system with more elections is “more democratic”

The idea that a small unit of government is “closer to the people.”

If you image life 200 years ago, when it was hard to obtain either information or entertainment, this was maybe right. It was challenging to communicate with people across long distances, but also people didn’t necessarily have anything better to do than listen to some local politician talk about stuff. The modern world is totally different. The odds are extremely high that you have read more about Jamaal Bowman’s primary challenger, the top elected official in Westchester County, than you have about county politicians where you live.

Under these circumstances, you want to emphasize relatively centralized forms of power with relatively few elected offices instead of separately elected boards and complicated nested hierarchies. I think city-county mergers are often a good idea, and when a county is so large that you feel like you need to sub-divide it (LA County is a good example here), you ought to actually draw a line and break the county up rather than having subordinate units. Then the mayor/executive of the county should have the lion’s share of authority over what happens.

I don’t actually mind if the units get very small (as they are in New England), but you then need to think seriously about what is and is not a question of local concern. A park or a library is supposed to serve a local constituency. People who live in Richmond don’t really benefit if Alexandria has good city parks, and they don’t really suffer if Alexandria has bad city parks. This is a local government function. But when you treat land use regulation as a purely local concern, you in a very predictable and systematic way undersupply housing relative to what would be best for the state as a whole. And you repeat that error across the entire state. That’s bad, and it’s a reason for state legislatures to take control of land use policy.