Transit-oriented development (TOD) has been growing in popularity for many years, particularly in large metropolitan areas. With this development method, a rail line is built out to a less populated area, then large, mixed-use residences and businesses are constructed in close proximity to that rail stop. The idea is that now you have manufactured a dense neighborhood where people can live without needing a car and use the rail line to commute to work.
Strong Towns is often assumed to be in favor of this style of development since it has many of the features we discuss and advocate for like mixed-use buildings and compact, walkable neighborhoods. But this assumption is misguided.
We cannot manufacture productive places like a cargo cult, simply by adopting some of the characteristics of other productive places we’ve visited or heard about; we have to drill deep to uncover why those places were actually productive. It wasn’t just because they were walkable or built more compactly, it was because they developed over time, through trial and error, with resilient, bottom-up practices, rather than a top-down, megaproject, debt-based mentality. And those traditional development features are certainly not present in the transit-oriented development model, any more than they’re present in the manufactured “town squares” and “entertainment districts” popping up all over the place these days.
To get to the heart of why transit-oriented development is so problematic (and to hear about a better alternative), let’s turn to a response Chuck Marohn, President of Strong Towns, wrote several years ago when a reader asked what he thought about TOD:
Build-it-and-they-will-come transit is no better than build-it-and-they-will-come highways. Both are a risky gamble with sketchy results, particularly for local governments. Transit-oriented development is the transit-advocate’s response to highway strip development in the same way that the early planned New Urbanist developments like Seaside were a response to greenfield suburban development. I’m sympathetic, but this isn’t the answer.
Instead of transit-oriented development, we should have development-oriented transit: Identify places where things are happening now and then connect them with the lowest level of viable transit possible. Make sure those places allow the next increment of development by right (without extensive permitting). This will ensure that the transit is viable and that it supports that next level of growth and expansion.
When that next level of growth and expansion happens, everything moves up a notch. Upgrade the transit to the next level — from jitney to shuttle bus, from shuttle bus to city bus, from city bus to streetcar, from streetcar to light rail, from light rail to subway — and repeat.
Yes, this is less “efficient” than getting out there and building the light rail to the parking lot now, but only if we judge efficiency on the project level. It would be cheaper on a unit cost basis if we just built everything right now and let the development fill in afterwards. If we judge efficiency on a macro-financial level, however, it is more efficient to allocate scarce resources to endeavors that are proving their success than to those that may be successful but, thus far, are unproven. This isn’t to say you won’t find a successful TOD out there. But the few that do exist tend to serve as a cherry-picked template for what other TOD advocates strive to achieve, rather than a proven method of success.
Like most post World War II development, the affluence of the United States created a development approach where we could transform our places to “perfection” overnight without the difficult, iterative process that has always shaped healthy, strong places. What we gain in instant gratification we lose in strength and resiliency.
Great places need a train less than a train needs a great place. Build the place first and transit becomes the logical, inevitable next step. No more transit-oriented development schemes. What we need is development-oriented transit.
(Top photo source: Fairfax County)
The Discussions are further valuable to the subject
In Japan, rather than build parking lots around transit stations like we do here, they build shopping centers, housing, and office buildings. The transit line creates demand for these buildings, and these buildings create demand for transit. Consequently, Japan is one of the few countries that doesn’t subsidize their trains. Their land is also more productive compared to our low density residential neighborhoods.
Japan’s model is also how our railroad towns were built back in the day.
I think what we’re doing wrong today is we subsidize low density and then we have to subsidize high density also in order to attract development and get the tax revenue to pay for the low density neighborhoods. It’s a big mess.
So I’d say TOD isn’t the problem, it’s just a symptom of the problem and an attempt to prevent it from growing into something worse (municipal bankruptcy).
- Lewis Thorplanner Avocado Moose • 3 years agoPart of our issue is that we have the tendency to break everything into bits and require that anything remotely profitable goes to the private sector, while the risk and unprofitable things (like building infrastructure) are burdened to the public sector, then everyone complains that the trains aren’t “run like a business”. This makes it really tough for transit agencies to do the kind of value capture that they do in other countries. In New York, the transit agency got sued by private bus operators and lost, meaning they could not serve school students and “compete” with private bus companies. So they have to have special runs that are part of a numbered route, but happen to start near a school at the right hour, are not labeled with anything to do with the school and are open to all riders, although who would want to ride a bus with 50 unsupervised 8th graders.•Share ›
I’ve always had a problem with this strong town’s concept because it treats modes of transportation as if they are on the same continuum. But there’s really 2 continuums. The flaw of cars in cities is that they don’t thru-run (like a bus does). Cars have to be stored at each destination rather than on some marginal property in the periphery of the city. So the increments are not: bike > car > van > bus > light rail. They are: “Things that need to be stored” (bikes > cars > bigger & bigger parking decks) and “Things that thru-run” (taxis, buses, trains).
If we don’t provide thru-running transportation options (buses & rail), by default almost everyone will drive. And if a city is of sufficient size, then it starts getting paved over with parking lots destroying quality of place, walkability, and tax base.
This is not an endorsement of TOD, but the idea that we need to hang around & wait for bus ridership to reach some arbitrary level while at least 90 of our top 100 cities have gigantic parking craters downtown is outrageous. The fact that we’ve paved over our most valuable property is a big clue that we’ve messed up our transportation mix.
I’ve tried to calculate the cost to Milwaukee of woefully underserving the active and public transit needs of the metro (manifested in 55,000 downtown parking spaces). It’s runs easily into the 9 figures annually, just from lost tax base in the greater downtown area from surface parking (not including the billions spent on urban highways or their devastating effects on the communities). This is while we continue to cut transit service to the bone on a system that subsists on only a minuscule ~$150M annual budget. Just by doing rough math, it’s pretty clear that there is a huge need for improved thru-running transportation service in the core areas of most of our medium to large cities.
- Zvi Michael • 3 years agoInteresting perspective – I would perhaps make the distinction between ‘things which add structure’ to a place (certainly rail, and also bus service if well designed) versus those which do not: ubiquitous road access and surface parking. On street parking can block public transport priority and leave buses stuck in traffic, and massive parking lots essentially create barren waste-lands….3 •Share ›
- Charles Marohn Mod Michael • 3 years ago“…the idea that we need to hang around & wait for bus ridership to reach some arbitrary level while at least 90 of our top 100 cities have gigantic parking craters downtown is outrageous.”What’ stopping you from filling in the parking craters? Why is an expensive, all-in transit investment essential for doing that? Is there an advantage to keeping your options open and flexible as things start to recover and stitch themselves back together? (And related – how is someone to know which set of parking craters are ready to receive development that justify the all-in transit investment?)1 •Share ›
- Michael Charles Marohn • 3 years ago • editedRe: What’s stopping. – I think we’d all love to see the parking lots downtown filled with gracious townhomes and low-rise retail or offices. Unfortunately, even in a “dead” downtown surrounded by highways with pretty low quality of place, parking lots are still quite valuable and thus it’s not economical to build the first several increments of development.
Re: Why transit: our dilemma is that we’ve got about 100k downtown workers of which about 2/3rds drive. Hence the need for 55K parking spots (not everyone works at the same time or every day). We can either decrease the number of workers or decrease the percent that drive. We’ve actually been doing pretty well at decreasing the percent of people that drive by getting people living near downtown, but it turns out any parking capacity is simply gobbled up by more jobs. (not the worse problem but it doesn’t fix the sea of parking that’s draining the public coffers).
Re: waiting. That’s what we are doing and it’s what we have been doing since about 1958. The fact that every medium to large city without robust transit has a giant parking crater downtown is pretty compelling evidence that downtowns have unique transportation situations that require thru-running transportation solution, lest it turns over to parking.
re: which ones. I’d do the math. If a parking lot value rises into the low single digit millions/acre, we’ve priced out the first several increments of development (townhomes, low rise etc.).If the best case for these downtowns is to be filled with 20+ story buildings with attached block sized parking garages, that model isn’t working. But that’s the framework we’re pushing cities toward when we starve them of thru-running transportation options. I’ve tried to do the math – our parking lots downtown cost us on the order of 100s of millions per year in lost tax revenue. Locally, we contribute about 10M per year to the bus system, with the rest of funding from fares and the scraps that trickle down from the feds. It’s pennywise, pound foolish.2 •Share ›
- Charles Marohn Mod Michael • 3 years agoOkay. Let me reflect back to you what I’ve heard you say and you can tell me if I’m understanding you.1. We have so much empty unused space — parking lots — that it’s not economical to start building modest structures on them.2. We need to handle ~66,000 auto commuters a day and so we need the parking.3. We’ve been incrementally getting more people to live downtown but, darnit, when we do it just creates more jobs and that means those large unused spaces (see #1) are used for parking. Catch 22.4. Unless we make a major investment in commuter transit — thru-running transit — we’re stuck with our only viable option being 20+ story buildings and that’s not working.5. You are losing money on parking lots and spending not much on a bus system. This represents a major lost opportunity (although it’s unclear to me what dots you’re connecting there — I agree, although I suspect for different reasons).Is that what you think? Have I represented your thoughts correctly?•Share ›
- Michael Charles Marohn • 3 years ago • edited3. Should be, we’ve gotten development on former brownfields within the 2 mile walkshed of downtown. To date, there’s been little on downtown parking lots. But, yep that’s the general gist.•Share ›
- Michael Michael • 3 years ago • editedHere is how I would approach it incrementally. In the last 20 years, we’ve reduced the budget of the bus system by about 75M and lost 75K riders. So we have a pretty good idea that at the margin, it cost $1K to get/lose a rider, of which riders directly fund about 1/3 of costs. So it’s about 667/yr of public spend to capture ridership at the margin. (It’s probably cheaper, for various reasons because we’ve essentially made the system a “coverage” system for social equity reasons, rather than prioritizing ridership… but 667/yr can be our conservative estimate.)At the same time, the lots/decks are choking out development on about 40% of the surface area of downtown and while reducing quality of place over all. I don’t know what “quality of place” is worth, but based on the existing development, that land would be generating greater-than-100M in tax revenue annually if developed similarly. So we can estimate that 55K spots has a cost to the public at least 100M or 1.8K per spot, in lost tax revenue relative to being developed similarly. So there’s a back of the envelope that going heavy on parking in our transportation staregy has been negative ROI at the margin.So now we’ve got to start fixing it. Some of those downtown spots are more costly to the public than others, such as ugly surface lots at gateways to neighborhoods. So those are our first targets. We know americans will only walk a 0.25 mile to a bus stop, so we take out our map, look at where the worst offenders are in comparison to routes, then we invest incrementally in the routes most likely to support development there. Adding 1 bus per hour to a route costs about 1.3M/yr. That should add about 1300 riders. We keep running the math quarterly until either the parking is gone or more likely, it becomes more expensive to capture that ridership than to absorb the opportunity cost on the parking lots. Like almost anywhere our strategy has been to mine the value out of downtown. We think we’re saving a few bucks by underserving the needs of downtown, but we end up devaluing our region’s biggest assets.(As an asside, we’d never make half of downtown a rail yard or bus depot or helipads or runways or whatever, regardless of how useful we find them, but we’re culturally desensitized to making half our cities parking)2 •Share ›
- Marven Norman Michael • 3 years agoWe know americans will only walk a 0.25 mile to a bus stop, so we take out our map, look at where the worst offenders are in comparison to routes, then we invest incrementally in the routes most likely to support development there.Actually, they’ll walk a bit farther (link: https://www.accessmagazine…. and most people can bike three to four times as far as they can walk in the same time. Adding those two assumptions should certainly improve your results quite a bit.1 •Share ›
- Charles Marohn Mod Michael • 3 years agoI’m sorry, but I got stuck on your first paragraph. It’s $1k to get/lose a rider. Um…. no. 😦 That’s not how complex systems work. Thinking about it as a linear relationship is comfortable from an advocacy standpoint, but that’s not how the world works.The reality with transit systems is that, as the network improves, the cost to get the next increment of rider decreases. That observation is actually the best argument for a non-incremental approach.The problem with it is that it assumes we know and understand where success will happen, the form it will take, the characteristics and needs of the people who will be there. Our outcome becomes rather binary — either total success or failure — and I’m not confident enough in my ability to predict the future to believe I know exactly how it should be done, in the absence of decades of incremental changes to co-build it with the community.•Share ›
- Michael Charles Marohn • 3 years ago • editedCome on, Chuck. That’s a pretty silly cop out. Mass transit is the only area of transportation & land use where we don’t use math? This isn’t any more complicated than choosing between a street or a road. Right now we’re using the battering ram that is tens of thousands of SOV cars (the stroad of vehicles…), when maybe we could be using the ball pein hammer of a few targeted bus routes. If we did the math.You see, there are very real spacial differences between transportation modes that we can observe rationally, empirically, and heuristically. Rationally: an average 40-foot bus does 31 passenger boardings/hour and can thru-run the CBD while being stored in low value property on the periphery. Conversely, 31 cars takes up about a quarter acre no matter what lot we put them in. Obviously, the value of thru-running is dependent on the value of land in the service area.Empirically: Knowing that thru-running is an attribute of value, we can try to “do the math”. My results are pretty convincing. It’s worth making incremental investments in routes that run from areas likely to sustain higher ridership through downtown parking wastelands.Heuristic 1: Every US metro area over, say 1M people, without robust transit networks has a huge build-up of parking in their downtowns. While cities with robust transit networks have comparatively smaller ones, if any. This same pattern plays out across the rest of the 1st world in medium to large cities.
Heuristic 2: Milwaukee, like most rust belt cities, had development across its entire CBD before we bankrupted our transit networks (topic for another discussion…). Within a decade of the transit failing, every easy-to-tear-down structure was torn down for automobile parking. So we can observe repeatedly across eras & continents, that there’s a strong relationship between transit level of service, automobile use, and parking. All of which point to the need to do some pretty basic calculations if we’re going to be prudent stewards of our collective wealth, make small iterative investments if/when appropriate, then react to the results. Unlike a billion dollar highway interchange, we can add a bus for a couple years and reverse course when needed. My bus schedule gets cut every 6 months…see more1 •Share ›
- Michael Michael • 3 years ago • editedOne more thing regarding your “binary” outcome. Suppose you live, work, send your kids to school, in a city that is blanketed in surface parking lots/decks on its most valuable property downtown (that is, the land not already covered by urban highways). Suppose there’s a growing list of local employers that would like to move onto those downtown lots, but most ultimately don’t because the development model that pencils out is $350/finished square foot space because they first need to buy an existing parking lot (priced as a profitable business, not undeveloped land) and then put their building stacked on top of an enormous parking garage.https://www.google.com/maps…This is the binary outcome we’ve created for these cities by so obviously flubbing the transportation mix: either a surface lot or a cataclysm of development.•Share ›
- Wells Michael • 3 years ago • editedHere’s a perspective on transit and TODs that may fill in some gaps for Michael and moderator Marohn. It’s derived from Calthorpe and Fulton’s planning philosophy in their tome “The Regional City” which I take as an advanced application of New Urbanism: Commute systems create more demand for commuting than they can handle. Light rail is an anti-commute system. To optimally fill LRT seats in the reverse-commute direction and in both directions off-rush hours, mixed-use development (TODs) must occur at all stations, particularly those at the far ends of lines radiating out of city centers.In short, the larger and more complex transit systems with transfers between rail and bus and between bus lines made convenient: the greater the mixed-use TOD potential.Example: BART runs 10-car trains during rush hours, but 4-car trains, less than half full at other hours. Yet BART proposes a 2nd Trans-bay Tube to increase rush hour capacity even as highways suffer with traffic congestion likely to worsen. With such widespread TODs, BART could run 4-car trainsets all day and more travel demand would created and met during off-rush hours in both directions. More people would have their needs met closer to home, leaving more time for off-rush hour travel.To disagree with the notion of transit evolving from bus to rail. High capacity rail is more suitable for long-distance trips at relatively higher speeds than the standard 40′ bus NOT suitable for stop-n-go circuitous routes in traffic. Half of such routes would be better served with 20 to 30-passenger low-floor low-emission hybrid or battery EV paratransit van sized vehicles seniors, disabled and all transit patrons need, yesterday.GM & Ford owe municipalities a new paratransit van. The new model van could complement existing 40’ bus lines more suitably arranged as Express or BRT during rush hours when higher capacity is necessary. The short wheelbase, new standard bus/vans are more suitable frequent service through both higher and lower density districts as the simplest system applicable for major transit centers with rail connections. (edited)see more•Share ›
- Wells Charles Marohn • 3 years agoMarohn, please read my reply to you and Michael on the last long discussion on TODs V DOTs.•Share ›
The Overhead Wire is correct that “Transit-Oriented Development” and “Development-Oriented Transit” are misused in this article. TOD typically refers to designing development to take advantage of existing transit and DOT typically refers to building transit into a relatively undeveloped area to induce development. Good communication requires using terms correctly.
It’s also true that both TOD and DOT can be appropriate under the right circumstances. The first railroads were built by giving land to the railroad companies that they could sell at a profit once the train station made these locations valuable. Likewise, some streetcar lines were built by developers who bought cheap land outside a downtown and then sold it for a profit once the streetcar made it more valuable as a place to locate homes and businesses.
Finally, “what prevents you from filling in the surface parking lots that detract from your downtown?” Our upside-down system of property taxation is one culprit. Owners who develop their land are punished with higher taxes — not just the year that they develop, but each and every year that the development adds value to the property. Whereas building owners who allow their buildings to deteriorate are rewarded with lower taxes. And those who own vacant lots (or parking lots) pay much less property tax than their more responsible neighbors even though their land requires the same amount of street, sidewalk, water pipes, sewer pipe, utility lines, etc. Thus, the property tax encourages land speculation — and surface parking lots are simply a convenient and lucrative form of land speculation.
Some communities are turning their upside-down property tax rightside-up by reducing the tax rate applied to privately-created building values and increasing the rate applied to publicly-created land values. This makes buildings cheaper to construct, improve and maintain. It also makes it more expensive to hold valuable downtown land vacant (or grossly underutilized). And, by reducing the speculative holding of vacant and underutilized land, this system helps keep land prices more modest while avoiding the booms and busts created by speculation, The end result is more affordable and more compact development. If your community wants to move in this direction, I would be happy to provide assistance.
“Transit-oriented development (TOD) has been growing in popularity for many years, particularly in large metropolitan areas. With this development method, a rail line is built out to a less populated area, then large, mixed-use residences and businesses are constructed in close proximity to that rail stop.”
This is not true.
This “Strong Towns” reframing of an already understood concept is confusing.Transit-Oriented Development is building development that serves existing transit corridors in cities. There’s already a rail or bus line, focus on building around it. Development-Oriented Transit is building a transit line into the weeds and figuring out how to get something to pop up. I think the post makes more sense if you switch the terms. For more information on TOD, feel free to visit http://ctod.orgwhere there’s plenty of literature on the subject down to the minute detail. There’s even a 600 item research archive that was curated over a decade.
I realize I’m probably never ever going to agree with Chuck on this subject (and lots of other subjects as well), but I can hope that the basic concepts are understood so they can be discussed within the same frame.
Either way, it’s the development pattern that will determine. A walkable place doesn’t need transit absolutely. Connecting it to other walkable places is great, but the development is the key.
American cities have developed completely dependent on sprawl, federal subsidies, state tax benefits to large corporations, and a growth ponzi scheme. By removing these factors, and allowing development to work, we will see sustainable communities emerge.
My issue with TOD is that we often allocate scarce transit resources to these big flashy projects instead of using those resources to build out or beef up existing transit resources in areas that need them, and could help create the urban environments desired. I don’t think that building a light rail system, then throwing up three story suburban condo buildings around the stations solves any problems, and quite probably creates more than it solves.
The article uses the term TOD to refer to the large scale developments built at once around new transit lines. That’s the cargo cult/build t and they will come model that has problems.
But the article’s alternative inst’ well written. It would seem to suggest building anywhere whether there is transit or not and then as an afterthought trying to add transit. It sounds the same as building development without considering road capacity. Just build anywhere and the roads and transit will have to follow. don’t worry about where the roads and transit are now.
instead Cities/Towns should decide where incremental development is appropriate. One factor should be considering where transit already exists. Not creating an entire new town center but changing the zoning around bus or train routes. Or thinking about transit access before a new facility is built.
It’s simply not true that all successful places were built by incremental development without planning. There are a number of successfully planned places, including most streetcar suburbs, Washington D.C., and even many districts in European cities like Eixample in Barcelona. Even Paris was extensively reworked according to a master plan in the mid-19th century, with good results. A *bad* plan will fail; and some adaptation is always necessary so an excessively strict plan will fail; but substantial planning can work and is even often necessary.
The problem with unplanned incremental development in the modern US is there is a wide range of density which is too dense for cars but not dense enough for walking. The only places I’m aware of that have crossed over that barrier naturally are in the Los Angeles area, and that has been only under extraordinary economic pressure. As a result most walkable development in the US is still just restoring areas that were originally built to be walkable, which is great but there’s not nearly enough of it, especially in areas that were much more sparsely populated 100 years ago like the West Coast and the Southwest. But with planning development and transportation concurrently, it because much more practical to jump that boundary. Transit-oriented development doesn’t always work but there have been a number of successes and a successful technique shouldn’t be rejected because of a ideology based on pre-automobile conditions.
This was a fascinating read. My first reaction was “isn’t this an ahistorical view? isn’t this how streetcar suburbs were formed?”
On further reflection, (some) streetcar suburbs are case studies on why Transit Oriented Development might not be a good long-term strategy. My streetcar suburb neighborhood in the Greater Boston area is a good example.
Constructed from 1910 to 1920 the area was built very quickly to a finished state, and grew rapidly. ~12% per year over 20 years. Doubtless this made the investors very wealthy. In it’s 3rd decade growth fell off because of WWII and a myriad of other factors, and the neighborhood began the long march towards decline. _Everything_ went bad at around the same time, the buildings, the roads, the underground infrastructure. For a variety of reasons the original streetcar lines got ripped out and replaced with bus service. The neighborhood seemed to hit its nadir sometime in the 1980s~90s suffering from the same city center exodus as the rest of Boston. It’s only recently with huge growth in nearby employment centers that there’s been pressure to invest again. With the restrictive zoning of a town that went through truly rough times we’re getting the outcome that you’d expect, huge upward swings in land values, with little change in the properties developed on them.
There just might be something to this Strong Towns idea after all… Whether you call it incrementalism or gradual money or something else.
I had a discussion with Chuck in which I expressed my severe doubt about the applicability of his notion of incremental development. I ended up concluding that he was discussing only a special case, where the good bones of past urbanism still exists. In the rest of America, where almost all development is accessible only by auto, I doubt his ideas apply. The scale of the dysfunction is so extreme that incrementalism clearly can’t work. Rather than see incremental change for the better, I think much more likely will incremental descent into gridlock.
At the same time, I recognize Rachel’s point. I toured the area being redeveloped around a suburban BART station, and was horrified to not find anything resembling a neighborhood. Only big condo buildings… That too can’t possibly work as TOD–far too many car trips will be necessary. I think this was a failure of planning. What’s needed are complete neighborhoods. Whether Chuck is correct that they have to develop organically, or whether enlightened planning can set the stage for neighborhood retail, only time will tell.
I think the term TOD was misappropriated in the first place. What the description of TOD describes isn’t transit ‘oriented’, it is transit ‘accessible’. Transit is available if you like, but you don’t need to use it. You’re paying for parking and abundant road infrastructure right in your rent, so you might as well use it, right?
– TODs are too often accompanied by generous parking ratios, which are typically required by code.
– Oftentimes, this parking availability is particularly dense near the transit.
– TODs often have a superblock type configuration, but with poor ped/bike penetration of the blocks. They often lose in these areas to driveways and parking for cars
– The transit that is offered at TODs is too often at the edge of the development.
Real TOD would have strict parking maximums, so that residents and workers locating there WILL use the transit or the ped/bike, because that’s what is available. People expecting a car centric lifestyle wouldn’t be living here, because the location wouldn’t serve their lifestyle.
Real TOD would center the transit station at the heart of the density, with the built environment being its most ped/bike friendly nearest the station, with more car oriented development that is inevitable happening farther away to reduce conflicts between modes.