The concept of Transit Oriented Development (TOD), relies on the thinking that if you build enough housing/retail/whatever close to transit, people will be less likely to drive. Though these policies can produce some good results, they usually ignore the true nature of the problem.
This week, the Houston City Council adopted a new set of Walkable Places Code Amendments, basically a huge TOD-centered reform program which will allow more land use flexibility within a half-mile of metro station platforms or certain designated streets. Tying land use reform to availability of transit has also been the approach of Senate Bill S.237 in Vermont, which allows reduced parking mandates within a half mile of a transit stop, and to Oregon's 2019 senate bill SB 10 which proposed to allow greater residential density within certain distances of transit lines. In all of these policies, the proximity to transit is the crucial factor in justifying the reform--apparently based on the assumption that when transit appears, people ride more and drive less. This is commonly accepted as fact, but I think the relationship is not clear. TOD has been around for over 20 years, but in most of my experience in visiting these things around the country, the TODs themselves do not really do much to reduce driving if the surrounding context remains convenient for cars. And in most cases, for every million spent on a TOD, there were probably $5 million more spent in the area around it for the benefit of driving and storing of cars.
But of course this is the way it has been for generations; since the 1950s we've spent spent billions of dollars on infrastructure that ensures that motor vehicles are the fastest, most convenient way to access anything. We continue to invest this way every time our zoning codes require more parking or new roads to accommodate the drivers traffic engineers imagine will show up with every new development. A TOD approach like Houston's sort of reverses that, but only in the small areas of the city where the paltry transportation network cast a shadow. What does this mean for the areas where transit doesn't exist? And who chose where transit lines were placed in the first place--why should those areas be the only ones that get land use regulation reform with this set of amendments?
Furthermore, what happens in the future when someone proposes a new transit line--how will local homevoters respond? The old NIMBY trope of "I support changing the zoning to allow more density, but only when we have enough transit" could become really problematic here--if the transit line facilitates the density, then are we putting the chicken before the egg? Or the cart before the horse? If Houston's new reforms allow flexibility within a half mile of any transit line, could some bus route planner now draw a bus route through the lowest-density rich neighborhood in town, thus magically making it eligible for massive land use reforms? Conversely, could some city manager cancel 80% of the bus routes, effectively wiping out any chance for reform? It all feels kind of iffy and based on the wrong premise.
But the main thing I want to look at in all this is the widely-held assumption that mere proximity to transit, especially light rail, actually makes people ride it. I propose that this way of thinking misses the point of how and why people travel, and thus never really addresses the over-driving problem it seeks to solve. Travel happens because people want to access destinations--the store, their work, their kids' school, the park. People choose to access their destinations by foot, by car, by bus, or by whatever other mode is available to them based on time, convenience, and cost. TOD might make transit, and nearby destinations, slightly more convenient, but as long as most of our transit systems suck and TOD reforms are limited to highly specific areas, the rest of the destinations are probably going to remain most conveniently accessible by car. So people are going to keep driving.
As an example of this, see the Lincoln light rail station below, located on the west side of interstate 25 south of Denver. It's a pretty standard TOD development with transit-branded apartments ("Lofts at Lincoln Station") and just a dash of commercial activity sprinkled in for good measure. It even has some employment areas with a corporate medical office on the left and an engineering firm on the right:
TOD is supposed to encourage people to take transit. But how many places do you think the light rail from Lincoln station can go? Looking at the image, compare the amount of light rail infrastructure with the amount of single occupancy vehicle infrastructure, including space for storing vehicles. Where (and how) are the people living in these apartments or the townhomes behind them going to go for groceries or a new desk chair? What mode do you think they'll choose for most of their trips? What about the people who work at the medical place or the engineering firm--what transportation mode do you think they use to get down here? Even with TOD, TOD-parking reductions, and TOD-allowed higher density, this whole landscape is still designed to encourage people to make the choice to drive. While the light rail line might provide a person living here access to 1000 destinations in 30 minutes, the highway system immediately adjacent can provide access to 10,000 destinations in the same half hour. What's the big deal about TOD if you probably still need a car to live there?
Another example is the Tyvola Station in Charlotte, NC. To be fair, this isn't exactly a TOD, because the station is in the middle of a road. But it's a good example of how the existence of transit might not necessarily mean people will ride it:
How many people shopping at this big box mall on the right take the light rail? How many people living in the apartments behind use the light rail to get to their work? There is practically more parking space in this image than anything else. Now, a former resident of this area pointed out that this complex is pretty old, and development has filled in to some degree around Scaleybark station two stops to the north, but I remain skeptical. The rail network provides a fraction of the access of the roads, the roads widths themselves are unlikely to change, and re-subdividing the land along this corridor to be more conducive to real walkable places could take decades. Driving is going to be the most convenient choice here a long time.
To be fair, we should look at a development which appears to have filled in and built up considerably (or maybe the density was already there--I don't know Dallas). Here is SMU/Mockingbird station on the north side of that city:
Even with all this density, how many people are really travelling to or from this place by train rather than car? Regardless of whether you build out the surface lots, the highway and connector roads will always provide easy access and the structure parking will always provide convenient storage. Accessibility by car is baked into both the right-of-way and the adjacent land use. And, as usual, City regulations had a hand in it--according to a 2002 ULI report on this project:
"The developer estimates that he had to build $6 million worth of excess (structured) parking for the project. While the city allowed the developer to build only 1,600 spaces (2,200 were required; 1,400 have been built thus far) by granting a mixed-use parking reduction credit, it refused to reduce parking further to reflect transit’s proximity."
Finally, in 2018, the average daily ridership at this station was 3,021 passengers while the average daily traffic on the road immediately to the south was 41,895.
The final example is an image of Bad Cannstatt, Germany, probably the kind of place most people are imagining when they get excited about TOD in the first place. This is a semi-urban area on the way out of Stuttgart, with above-ground light rail and lots of shopping and residential nearby:
So why does this look so different from the American examples? First I'd say that there's significantly less surface parking, and significantly more residential space. But also if you zoom in on the map, it is clear that there is also less land devoted to motor vehicles even within the right-of-way. Vehicles are allowed, but they have to share space with buses, bikes and rail tracks, and they don't necessarily have priority. Cars are still convenient, so people still drive. But in this context they are probably less convenient, so people choose other modes. Transit works best when it replaces the physical space once dominated by cars.
If we want fewer people to drive, or we want to make density more viable, we need to reallocate space in the right-of-way, not dilly dally around with excited talk about the the 1999 innovations of TOD. Real change to reduce driving will mean mean removing lanes, removing parking at destinations, or some type of congestion charge for motor vehicles taking up road space. We get the traffic we build for, and in most places, including around TOD areas, we still build for driver traffic first. Once you make driving less convenient, demand for more density will appear naturally. Right now we seem to be operating on the backwards idea that we'll allow density and then hope people stop driving.
It's great that Houston and Vermont are taking these steps to reform their land use regulations, and hopefully it will spur some much needed land use reform. But we should be wary of TOD in the one hand distracting attention away from the ongoing road building and the still-required parking minimums in the other. Making access to destinations via transit slightly more convenient through TOD does not make accessing those same destinations less convenient by motor vehicle. Our land use regulations need reform, but we can't forget to change our right-of-way management rules as well.