This data is from 2019, but I imagine that things would look pretty similar today and that it might even be a little more pronounced. The dataset from the above article looked at how many people have cars in a given area (a darker dot = fewer cars) and then plotted this against population density and income per capita.

Here's what that looks like for the regions of New York, Boston, Los Angeles, and Houston (data from 2013 to 2017):

What is fascinating about these charts is that they show two different correlations. In dense and transit-rich cities such as New York and Boston, car usage is most closely linked with population density and not with income. The dark dots form a horizontal line near the top.

However, in the case of Los Angeles and Houston, car usage is instead most closely linked with income and not with population density. The dark dots form a vertical line near the left -- the lowest income per capita.

So what does this tell us?

It tells us that if you design a city to broadly require a car, then you are likely to sort people based on those that can afford a lot of car and those that cannot. On the other hand, if you design a city around transit, then you are likely to instead create a place where both the rich and poor get around in similar ways.

There is also evidence that the latter is being increasingly viewed as more desirable. 2017 was the first year in the US where high-income young people (ages 26 to 33) drove less than low-income young people. Presumably these high-income people had choices, and so I tend to view this as a preference.

As a whole, this is surely a good thing for our cities. But now I think we need to be careful not to allow density and walkability to become the new luxury that only the rich can afford.

Here is an interesting chart from the New York Times explaining the disproportionate impact that highway and urban renewal projects have had on non-white families in the US. The x-axis is the non-white population share in 1950. And the y-axis is the percentage of displaced families that were non-white. What this means is that the diagonal dotted line through the middle represents a kind of racially balanced displacement.

However, as you can tell from the graph, displacement from 1950 to 1966 was not balanced. In Providence, for example, only 3% of families were non-white in 1950. But these families represented 31% of the ones displaced for renewal projects. In Philadelphia, about 18% of families were non-white, but here they represented 71% of those displaced.

I don't think that this will be news to a lot of you. "Urban renewal" is a loaded term in American urbanism. But the article does do a great job of taking you back through time in cities like Houston, Chicago, and New York. The article is also by Adam Paul Susaneck, who is the founder of Segregation by Design. If you're interested in this topic, I would encourage you to check out his website.

Maybe it's confirmation bias, but I continue to feel like there is a groundswell of interest in trying to improve housing supply and overall affordability. The YIMBY movement continues to gain steam. Here are are few excerpts from a recent M. Nolan Gray article where he calls for an end to zoning as we know it today:

In nearly every major U.S. city, apartments are banned in at least 70 percent of residential areas. San Jose prohibits apartments in 94 percent of its residential areas. The most a developer can build in these zones is a detached single-family home.

The results speak for themselves. Houston builds housing at 14 times the rate of peers like San Jose. And it isn’t just sprawl: In 2019, Houston built roughly the same number of apartments as Los Angeles, despite being half its size. Since reforms to minimum-lot-size rules were put in place in 1998, more than 25,000 townhouses have been built, overwhelmingly in existing urban areas.

To be clear, Houston has made its share of planning mistakes. But, free of zoning, the city can constantly remake itself. That Houston is now one of the most affordable and diverse cities in the country is no accident.

The relationship between housing affordability and constraints on development is a well-documented one. If you want more affordable housing, you generally want fewer, rather than more, constraints on delivering new housing.

But as I was reading through the article, I couldn't help but think more specifically about the relationship between sprawl and affordability. Because it is also true that, for a variety of reasons, the former has tended to help the latter (or at least coincide with it), which is why Gray felt it was important to say "and it isn't just sprawl" when talking about Houston.

Part of this relationship has to do with the fact that expansionist development tends to be of the low-rise stick-built varietal, which is a relatively cost effective way to build. Whereas the higher density infill stuff tends to be built using more expensive materials like reinforced concrete. But of course there are many other factors at play, including lower land costs.

So I think one really interesting question is this one here: To what extent could we break this relationship between sprawl and affordability with what Gray is advocating for? In other words, how cheap could we make new infill housing in our older cities if we were to greatly loosen zoning controls or possibly even remove them all together?

I don't know the exact answer, but I know that directionally it would be better.