The "Great Bones" of Rust Belt Cities

Can quality Rust Belt neighborhoods lure residents back?

A view of St. Louis’ Central West End neighborhood. Source: stlouisneighborhoodsguide.com

Hey there! I’m one week into the paid subscriber launch I kicked off last Monday! I invite you to sign up to continue getting unlimited access to hundreds of articles and the ability to comment and have a dialogue with the growing CSY community. And if you want to remain a free subscriber, that’s fine! You’ll still receive as many as two posts a week. Join now!

I went to St. Louis over the weekend, and I was reminded how much I love the way St. Louis neighborhoods look. The city has wonderful vernacular architecture that leads to beautiful neighborhoods at a human scale. That’s not true everywhere, since St. Louis has lost a lot of character through abandonment and demolition, and the scars of highway construction exist throughout the city. But St. Louis is perhaps the quintessential city with “great bones”.

If you’re not familiar with St. Louis, take a look at the image above, and this one below:

A view of homes in St. Louis’ Shaw neighborhood. Source: stlouisneighborhoodsguide.com

Over the years I’ve heard urbanists of all types say that Rust Belt cities were primed for a comeback, someday, because they have “great bones”. But how can we define “great bones” and make it a strategy for growth? This is an especially good question given the work-from-home era we live in now, and where we work and live aren’t necessarily connected anymore.

As for me, I view “great bones” as the quality of the built environment of a city. It’s subjective, yes, but there’s some agreement among people about community quality. Given a chance, people will choose neighborhoods (and I think we all choose neighborhoods or communities, not entire cities or metro areas) that provide us with the housing we like, good schools for our children, parks, shopping amenities, social gathering spaces, and a welcoming environment. The desire for public transit, walkability and multimodal accessibility, and the mix of housing will vary with each person, as would the level of public-facing or private-facing view wants to give.

I laid out my so-called “Big Theory” of American urban development several years ago, and brought it up again last month. I think how people look at whether a place has “great bones” or not falls somewhere into the Big Theory framing. Essentially, people prefer the kind of built environment of particular times in history, and the infrastructure and amenities that come from them, and I think they come in a general order. The strongest preferences are for new places among the public because they’re, well, new. Contemporary housing and commercial development designs, upgraded roadway and utility infrastructure that’s not in danger of deterioration. There’s also a preference for much older places that have a development character that’s difficult to replicate in our cities today. Lastly there’s a vast middle type of built environment. It’s new enough to be missing the character of older places, and old enough to be missing the contemporary comforts you might want.

For me, a good proxy for assessing development quality and character is looking at U.S. Census housing data. A table that’s been included in the Census and American Community Survey data for decades is Table S2504: physical characteristics of occupied housing units. I look specifically for Census estimates on when housing structures were built, and that gives me a sense of what kind of “bones” a place may have.

After going through St. Louis over the weekend, I revisited that table to get a sense of whether there’s some truth that the core cities of the Midwest’s largest metros indeed have “great bones”. I gathered the data on the number of occupied housing units in the region’s largest cities and categorized their housing in three buckets, by age: 1) housing built in 1939 or earlier; 2) housing built between 1940-1979; and 3) housing built since 1980. Here’s what I found:

The Midwest core cities I pulled data for, each being the core cities of metros with one million or more residents in the region:

Chicago Cincinnati Milwaukee

Detroit Kansas City Louisville

Minneapolis/St. Paul Columbus Grand Rapids

St. Louis Indianapolis Buffalo

Pittsburgh Cleveland Rochester

Looking at the table, Midwestern core cities have substantially older housing than the nation overall. The Midwest has three times as many homes built before 1940 as the rest of the U.S. Similarly, it has nearly half as many homes built since 1980 as the rest of the nation. Of course, not all old homes are good homes; in fact, many old homes aren’t good homes. This doesn’t take into account the quality and condition of housing, the location, or many other factors. However, if you’re looking for the character and charm that older homes and neighborhoods offer, you can find it in the Midwest. And again, no one’s building pre-WWII neighborhoods anymore.

Let’s take a look at how the cities stack up by the general preferences as I see them. This first table shows the percentage housing units in Midwest core cities constructed in 1980 or later:

If newer neighborhoods is what you want, Columbus and Indianapolis is it. A slight majority of Columbus’ housing stock was constructed since 1980, and Indianapolis isn’t far behind at 44 percent. However, there’s been very little new housing construction in Detroit, Buffalo, Rochester or Cleveland, where fewer than one in six homes is under 45 years old.

Now, look at this table to see the breakdown by city, ranked by the percentage of pre-WWII homes:

Buffalo tops the list, with 62 percent of its housing stock built prior to 1940. St. Louis is right behind with 56 percent. At the bottom of the list are Indianapolis, Louisville and Columbus, each of which have undergone city/county consolidations in their history to expand (Indianapolis, Louisville), or had an aggressive annexation policy (Columbus). In all cases, more expansion means more land for new homes and new communities.

Lastly, let’s rank the cities by their percentage of homes in the midrange, built between 1940-1979:

Detroit stands out as the city with the highest percentage of mid-century homes on the list, at 56 percent. Detroit’s numbers are slightly deceiving, however; in 1960 Detroit hit its peak number of housing units at 515,000. Today’s 256,000 is half the total of 65 years ago, with the bulk of the loss coming from the abandonment and demolition of the city’s pre-WWII housing stock. St. Louis has the lowest percentage of mid-century homes, at 26 percent.

The general age of housing stock tells me quite a bit about the makeup of a city. One reason St. Louis has the “great bones” of older housing stock is that it’s had the same city boundaries since 1876, and experienced it boom years in the first quarter of the 20^th century. What comes with that is the street network, the mix of uses, and the public transit infrastructure we associate with older cities. It’s not always good, but it’s there, and that can’t be said for every city.

Detroit’s concentration of mid-century homes could be considered a detriment. As I mentioned earlier, perhaps the least preferred housing type is in the 45-60 year-old range, the first and second generation of suburban housing. Detroit is full of neighborhoods with small single-family homes that once accommodated auto factory laborer families, but don’t necessarily meet contemporary needs. Yet there’s also an opportunity. So much has been lost and so little replaced in Detroit, there’s a chance to rebuild in areas that lost the housing but still retain the pre-WWII character in terms of lot size, scale, and access.

Meanwhile, cities like Columbus, Indianapolis and Louisville could still be chasing suburban dreams. Indianapolis and surrounding Marion County consolidated in 1970, with the exception of a handful of exempted suburbs. The consolidation meant that overnight Indianapolis more than quadrupled in size, going from 80 to 365 square miles. Louisville and surrounding Jefferson County did the same in 2003, going from 62 to 390 square miles. Columbus’ annexations have added some 90 square miles to the city’s size since 1970.

In other words, they were able to consume their suburban areas, ultimately changing how their cities would be viewed.

Again, I believe there’s value in preserving what remains of pre-WWII housing stock and their neighborhoods, and capitalize on their quality as well as scarcity to attract newcomers to aging cities. Given the prices of homes in the nation’s least affordable cities, often without the pre-WWII connectivity that people enjoy, it could be an option worth exploring.