Fixing One of America's Worst Designed Cities
A Texas-Sized Intervention
Last week, I published an article about the single most important element in creating good cities. If you haven’t read it yet, please do so here. In it, I teased that I would show how to improve a place by narrowing an overly wide Right of Way through building housing in the middle of the road. This isn’t necessarily a novel idea, but it sounds so wildly impractical given competing interests in managing public space, and the paralyzing inertia of the status quo, that many dismiss it as little more than a quixotic dorm room musing.
I assure you, my Optimistic friends, my goal is not mere discourse-generation, nor to demonstrate how graphically adept I am at placing lines and arrows. It’s to shift the Overton window of what’s possible for our streets. In this spirit, I hope you read this piece not as an amusing thought experiment, but as a serious urban policy proposal.
To any public officials, Departments of Transportation workers, planning staff, think tankers, or other interested parties: this is a sincere invitation to reach out and work towards effectuating this proposal. You can email me here. I would love to collaborate!
So, what city shall we choose as our preliminary sandbox? It needs to be somewhere sizable enough that the intervention would have a real impact on hundreds of thousands of people locally (with the potential for successive interventions), and serve as a beacon for other cities around the country. Much as I love small towns, much of the good work being done there broadly goes unnoticed, so it couldn’t produce the hoped-for ripple effect.
Over the last few years, I’ve had the great privilege of traveling around North America, talking with public officials, touring new developments, and meeting built environment professionals who are working to make their regions better. From walking every block in a neighborhood to driving the periphery of metropolitan areas, I’ve seen hundreds of cities from many different angles. But one trip has stuck with me.
When time came to visit Dallas, I was genuinely excited. It’s a place with an outsized personality and one of the most dazzling, lively cultures in the country. Surely the heart of Texas would have these values embodied in its buildings and streets, right?
For all the spirit of its people, the city does them no honor. As I drove endless mile after endless mile, equally unchanging and immiserating, disappointment weighed heavy on my heart. Here was a city with no discernible character. Without any soul. For a region home to nearly 9 million people, I’m hard pressed to think of even one building—even one structure—widely known outside of the area, favorably or otherwise. Its most famous place is a stretch of asphalt where someone was killed. Jerry World (AT&T Stadium) is the only structure that comes close, but this is a deeply unsatisfying conclusion; nearly every city of even middling repute has a modern stadium. Besides, it has little architectural merit apart from its awe-inducing size. Few people visit outside of game days as it’s nearly 20 miles from downtown and there’s not much to do. Great civic structures are largely absent, and for those few that do exist, they’re subsumed by the bleakness of their surroundings.
Union Station might, in another context, be a proud and genial host to newcomers and returnees alike. Instead, it sits sequestered in a pocket of Downtown Dallas that no one visits beyond conventions. And even then, attendees are usually dropped off in the morning, and whisked away in the evening. The convention area is itself a pocket within a pocket: ensnared by train tracks, seas of asphalt (both in parking lot and street form), and a noose-like tangle of overlapping freeways. If one was made to scurry in like a rat upon entering the new Penn Station, then Union Station seems designed to freight people in as indifferent cattle.
Some may disagree with my perception of the city, or with the idea that Dallas is the most worthy place for this intervention. That’s fine, but I won’t be dissuaded.
The only other city I seriously considered was its neighbor to the southeast. While Houston suffers from many of the same ailments, it’s land use policies are generally far superior. Plus, the city is infilling at an impressive rate. There is less obvious need for external prodding, and in fact, much praise is owed for the progress that’s been made. Though Houston is larger at the city level, when we scale out to the metropolitan area, Dallas is home to half a million people more, and is growing faster. Sometime in the next decade, it will likely surpass Chicago as the U.S.’ third largest metro area.
One of the most economically productive regions on Earth, Dallas is also one of the saddest urban environments. It seems like the perfect place to start.
Let’s Build A Better World, Together
If you’re reading this, you care deeply about why the world looks the way it does, and how to make it better. So do I!
If you enjoy thinking critically about how we design, build, finance, and steward the places we live, and how we can practically create the built environment we all dream about, I’d be enormously grateful for your support of the ideas, projects, and essays we share here.
Thank you so much for reading. I hope you enjoy this piece!
Ask locals where you should go to get a real sense of the city, and you’ll basically get three answers (beyond stadiums and the site of JFK’s assassination): Highland Park Village, Bishop Arts, and Deep Ellum.
Highland Park Village is an outstanding retail center, arguably one of the best in America, but it’s not a place where most people can gather and feel welcome. Hermes, Tom Ford, Jimmy Choo, Oscar de la Renta, Rolex, and Goyard are just some of the offerings. Bishop Arts has more going on for a broader crowd, but as a neighborhood it’s a little too small and too removed from the heart of the city for the level of impact we’re aiming for.
That leaves us, conveniently, with Deep Ellum, which has the most potential of the bunch. Here’s what it has going for it: some semblance of identity, local stores, and historic architecture. People already come here to shop and eat, so new habits wouldn’t have to be trained. Although it’s mostly one-story structures today (with some exceedingly vapid 5-over-1s starting to appear), the infrastructure could easily support far higher densities.
Most importantly, selecting Deep Ellum aligns with Step Two of last week’s framework for creating a virtuous cycle: concentrate efforts in the city’s core to maximize impact. It’s adjacent to downtown, and is, by Dallas Standards, highly walkable (about 4 linear blocks).
The neighborhood is gridded, which bodes well for our purposes for a few reasons. First, cars that might otherwise have traveled down our soon-to-be-narrowed Right of Way can easily be rerouted to parallel streets if they don’t appreciate the newfound intimacy of the driving experience. Second, it’s a lot easier to build in rectangles than amorphous shapes. Grid are nothing if not rectilinear.
Most of the action in Deep Ellum is oriented along its east-to-west streets, principally on Main Street. As Main Street is already fairly successful, it’s not the best place to start our intervention. Luckily, Commerce Street, just one block south, makes a lot of sense. Curb to curb, Commerce is 50 feet wide. It’s Right of Way (or interface between both building facades—thanks Andrew!) is 80’, with 30 dedicated towards two sidewalks.
While not as egregiously vulnerating as other streets in the city, Commerce is an ideal case study because it’s potential upside is so high. The street network here is tighter. Redeveloping it holds the promise of weaving the fabric of a truly holistic neighborhood—perhaps the first of its kind in Dallas.
Instead of promenading along a linear row that unceremoniously ends at a highway on one side and a railway on the other, building up Commerce would invite a deeper intra-neighborhood interaction. Pedestrians could do all manner of exotic things, like turn left and right on street corners to see what’s beyond the main row. Side streets could be unlocked as valuable urban real estate rather than languishing as tucked-away parking lots and back entrances for industrial uses.
Commerce is also a good case study because the city has said it’s a good case study—Dallas has dedicated $30 million towards redesigning the street, converting it from a one-way to a two-way, along with other infrastructure improvements. If the city has already committed real dollars towards enacting change, why not do something truly transformative? En garde.
First, to orient ourselves: here’s the layout of Deep Ellum. Downtown sits just to the west of the image, beneath I-45, while Fair Park, home to the State Fair of Texas and the historic Cotton Bowl Stadium, lies less than a mile to the east. On the merits of location alone, Deep Ellum should already be one the most dynamic urban neighborhoods in the country.
Zooming in to a stretch between South Malcom X and South Walton, things start to get interesting. We can see an alley on the lower part of the aerial with an interface of only ~25’ wide. Even before rightsizing this Right of Way, we’ve found another narrow street for our soon-to-be-cozy neighborhood!
5 lanes of traffic, however, are not so cozy. Let’s go about fixing that. Each lane is about 10 feet wide, giving us 50 in total. We really only need two lanes here, so we can recapture 30 feet. The sidewalks are currently around 15 feet wide. I think we can trim 5’ on each side while still keeping them accessible and pleasant (10’ in total). Now we have 60 feet to play with.
To align with the city’s plan for a two-way street, the northern lane beside our new row of buildings will run westbound, while the southern lane will run eastbound. That’s 20’ gone. We’ll also need two new sidewalks, so there’s another 20 accounted for. That leaves us with 20’ of depth for our row of buildings. While shallow, this isn’t impossibly small. If we’re going to be ambitious—and why propose such an radical idea here, only to get cautious on the finer details—I’d want this to be as fine grained as possible rather than just dropping in a few wide buildings.
This block of Commerce is 450 feet long. At 20 feet per building, we can fit ~22 along this stretch. With the 10 feet left over, I propose a pedestrian lane through the middle of the block. This will break up the scale, make the walk feel more welcoming as opposed to relentless, and create two additional corners for retail—the best sort of commercial frontage.
We should take care not to be overly prescriptive. While I think two intimate Yokochos would instantly create one of the most compelling places in America—and offer a cheap first step for burgeoning small businesses—22 new retail units, even at just under 400 square feet each, might be challenging to lease up in a neighborhood that already has a lot of commercial space.
To help us get around this, zoning should permit some degree of flexibility: if the market prefers ground floor residential, office, or even a workshop, we should let that emerge organically. That would be a more honest reflection of the people of Dallas, and enliven these quarters.
Above the ground floor, we’ll have two levels of residential. At four stories, we would need to add an elevator, which would increase costs that might be hard to recoup at this scale. Plus, the lanes could feel uncomfortably claustrophobic by Texan standards, given the tight 2:1 building height-to-street width ratio. So, three stories it is.
There are a few ways we can break down the buildings. The simplest would be to make every unit a studio. This would add 44 apartments at around 350 square feet each. People might not want to spend all of their time in apartments this small, so they would naturally spill into the cafes and small shops below, creating unprecedented vibrancy. I like this.
Others might not. Having some diversity in the unit mix is worthwhile to attract different sorts of residents—especially if it’s as desirable as I’m certain it would become. The easiest way to do this would be to combine some of the studios on the second and third floors into two-bedroom duplexes. If we did this in 11 of the buildings and kept the rest exclusively as studios, we’d have 33 residential units in total.
As the street evolves, modest one-and-two story buildings would, if allowed, grow upwards on the north and south sides of Commerce. If we opted for wider lots, say 30’, we could fit 15 buildings on either side. Narrow plots don’t allow for soulless mid rises predominate, and critically, they keep the scale human. These are deeper lots (100’), so they might be able to support three times as many units per floor as our modest mid ROW buildings, at least, with larger unit sizes. Rather than run through the full design exercise here, let’s assume these wider buildings rise five stories (the extra width and unit count make the economics of going taller feasible) with four stories of residential above the ground floor.
On just the block bounded by South Malcom X and South Walton, we might have nearly 400 apartments.1 This works out to about 123 units per acre. Yes, that’s a lot of housing, but it’s discreetly dense due to the attractive urban design, unlike the oppressively dreadful block through new product in the neighborhood that feel denser than they really are. Nearly all of these, bad buildings with bad urbanism, can be redeemed somewhat by the fact that they have brought more people into the neighborhood, enabling its quasi-walkability and a far more sustainable development pattern than elsewhere in the metroplex. This foundation has allowed us to dream big for Deep Ellum.
The only drawback from this intervention is that the sidewalks might be too narrow to plant a proper row of street trees. Ten feet should be wide enough, but if it’s not, we can place pots and planters as freely as our hearts desire. Amsterdam offers an excellent example of how to do this well.
This transformation isn’t just about aesthetics, identity, or culture. If the math doesn’t work, then this exercise is nothing more than an empty fantasy. So, how do the economics of our proposal look? Quite good.
For a more empirical perspective, let’s start with the assessed values of existing properties. According to the city’s Appraisal District, the combined market value for this stretch of Commerce is $6.7 million. (Three lots couldn’t be counted because they’re grouped with properties on Main Street and couldn’t be separated.) The total tax rate, including school, hospital, college, county, and city levies, is 2.23%. Multiplying the market value by the tax rate means this block is only generating ~$151,000.
What could the city/county collect if we built in the Right of Way? Some quick math is in order. Based on neighborhood comparables, we can estimate the retail/first floor rents at about $36 per square foot annually ($1,100 per month in rent), our small studios at $1,200 a month, and the duplexes at $2,100 (both much lower than market). Valued at a 5% cap rate, each studio building would be worth $588,000, while the duplexes would be worth $537,000.
Before getting carried away, let’s make sure our plan actually pencils out. For a rough viability test, I assumed prevailing hard costs of $200 per square foot, plus $50 per square foot in soft costs and interest. This should be more than sufficient.
Arriving at a land cost is trickier. How do you price roads that currently have no commercial value? Admittedly, this is a bit tenuous, but I pulled a nearby land comp: 2926 Commerce, right in the middle of our block, is a 6,000-square-foot lot currently used for parking. Its market value is $281,250 — about $47 per square foot. Given its location and use, that’s a fair basis for our lots. At 400 square feet each, the land would cost just under $19,000 per lot.
Of course, the city might ask for more in exchange for ceding this Right of Way for such an unconventional project. That’s fine. Let’s double it for the sake of the argument. At $38,000 per lot, the numbers still hold up beautifully. The buildings remain fabulously profitable, with strong margins and very attractive going-in cap rates. This gives me confidence that we can build this as imagined, with plenty of room for the inevitable mistakes that come with pioneering something new.
How do these building valuations translate into tax dollars? They nearly triple the total assessed value of the block, adding more than $275,000 a year in property tax revenue. But that’s just the beginning! What about the new buildings we expect to rise on either side of the existing Commerce Street?
As mentioned above, these buildings will contain larger units that command higher rents. Assuming an 85% lot coverage ratio (after factoring out building service and circulation space) I’m estimating there will be two one-bedrooms at 575 square feet each (renting for about $1,600 per month) and one three-bedroom at 1,200 square feet (renting for around $3,800) per floor. While these may seem small by Sunbelt standards, there are almost no three bedrooms on the market, which inhibits families and larger groups of roommates looking for cheaper rent from living in a certain neighborhood. Starved for options, I think these would lease up swiftly. If not, the third bedroom could easily serve as a home office for a different set of renters.
The retail space on the ground floor would also be larger at about 1,400 square feet per building. These could be split into two bays of roughly 700 square feet each if the need arises, but for simplicity, we’ll assume they stay as single 1,400-square-foot spaces renting at $30 per square foot. This is slightly lower than the smaller mid–Right-of-Way units. While not every one of the 30 buildings would have the exact same floor plan in practice, it’s simpler to assume a standard layout for our purposes.
Running the same valuation analysis for these new, wider buildings, I’ve adjusted a few assumptions: land value increases to about $500,000 per lot (to reflect buyouts of existing building owners) and construction costs increase by about 25% to account for the larger, taller structures. Even with these changes, the margins remain positive, though understandably tighter due to higher land costs and construction expenses.
Here’s where the wow-factor comes in: building valuations on this block alone would rise to more than $170 million dollars, up from just $6.5 million today. That would add nearly $4 million in annual property tax revenue to the city’s coffers, not to mention the untold millions in sales taxes that would be generated from transforming this into one of the most desirable neighborhoods in Texas. It may well even give the West Village a run for its money! Deep Ellum would finally begin to scratch the surface of its true potential, leveraging its proximity to downtown and its identity as a creative hub.
Where would the obligations shake out for this stretch of Commerce? It’s hard to say, exactly. While the city would save on road maintenance, there would be higher demand on utilities. Those costs wouldn’t come close to outweighing the added revenue. It’s a consideration, but hardly a dealbreaker.
Why would Dallas ever say no to this plan? Well, because it’s so different from the status quo. (And let’s be honest — they don’t need a New Yorker telling them what to do!) I’m proposing microunits in a city famously known for the exact opposite. And you may have noticed: I’m not proposing any parking for this neighborhood. Nearly 400 units, and not a single parking spot. Whoops.
Don’t clutch your pearls just yet. There are still dozens of surface parking lots in the neighborhood, plus several garages. And remember, this is a case study about giving space back to people, and taking it away from cars. Should the need arise, the city could easily build municipal garages with the increased tax revenue (below grade in some places, above in others). Or, an even more outlandish idea, they can invest in bus and bike lanes.
If this all feels a bit too much, critics are welcome to buy or rent one of the millions of single family homes across DFW instead of throttling the core of one of the most economically productive areas on Earth. But they need not worry too much. This plan only works in places with high enough land values and proximity to other valuable uses to even be worthy of consideration. By this accounting, more than 99% of the metroplex would be safe from this sort of development.
Does this mean people in Dallas would accept such conditions? In city after city around the world, people have proven to more or less be the same with respect to the built environment. Is it really so hard to believe that less than 0.01% of the region, maybe 800 people, would leap at the chance to live in a beautiful, walkable, culturally vibrant neighborhood, just like so many hundreds of millions do elsewhere? I don’t think it requires a particularly active imagination to presume so.
To recap how I would fix one of America's worst designed cities:
Start by identifying a core neighborhood that has wide Right of Ways. Concentrate building efforts in the middle of one block close to other productive uses, and only grow outwards after the first phase is complete.
If property values rise faster than the work can be completed, that’s a good sign, it means the intervention is working. In that case, prescribe tight lot widths to guard against large, horizontally sprawling structures that kill street life, and let the market to do much else as it pleases on the extant street walls.
Work can continue on mid-street development more or less insulated from market forces, since this land isn’t currently a commercial asset.
Our 80’ Right of Way on Commerce Street in Deep Ellum would turn into two 30 foot wide interfaces, where each the east and the west street would have 20 feet of sidewalks, and 10’ for driving. In another proposal, I think we can make these interfaces even narrower, where each building is 30 or 40 feet deep, and the interfaces drop to 20 feet.
The result? A block that today has no housing would gain nearly 400 units, more than 800 residents, and dozens of businesses.
The $151,000 in annual property tax revenue today would grow 25-fold, to more than $3.8 million, and that’s not counting the millions more in sales taxes that would flow to the city and state.
And most importantly, if less tangibly, Dallas would begin to craft a proper urban identity for itself.
By fixing just this one stretch of one street in one small neighborhood, we could start to reverse the fortunes of an entire city’s urbanism. And maybe a meaningful extent of a continent. There are certainly other neighborhoods in Dallas even more in need of the this kind of transformation, but that’s the point. We shouldn’t focus on exurban stroads where land is cheap and new development isn’t viable. Infilling interfaces is all about maximizing expensive, underutilized neighborhoods. We can’t do this on state roads, nor can we dream of it in places considered too precious to touch. The point is, we have to start somewhere that has a real chance of achieving success, and then keep pushing the idea forward once it finds traction. No plan, however enlightened, can work if it’s starting conditions are not well suited to its aims.
In this spirit, another road I’ve identified nearby is Cesar Chavez, located west of the 45 in the heart of Dallas. Its Right of Way is 110 feet wide, with 90 feet curb to curb widths. A deeper case study here is merited, as the potential and scale are even greater than on Commerce, though the neighborhood is not as felicitous as Deep Ellum. I’ll be working on this with Aaron Lubeck and Steve Mouzon over the coming weeks and months.
While we’ll do our best to advocate for the construction of buildings in the middle of the Right of Way, if our efforts fail, Dallas could do a lot worse than simply pedestrianizing, planting a tree canopy, and promptly becoming Beziers.
Put buildings in the middle of the road. If we can reclaim one street, we can reclaim the whole city. Let’s start here, and not stop.
With dreams of a better tomorrow,
Coby
33 apartments in the middle of the street, plus 12 units per 100’ depth structure, times 15 buildings per side, times two sides, equals 360 units. 33 plus 360 is 393, or exactly almost 400.
I’ve fantasized about this concept for a very long time, thanks for putting it to words! I’m American born and raised, lived in the status until I was around 26 - I live in Tokyo now but American roads always felt comically wide to me.
Visionary as always! I'm assuming numbers for this development style would pencil in colder (Rust Belt) markets?