Atlanta Boston Urbanism Washington DC

Defining Clear Standards for Transit-Oriented Development


» A new report attempts to quantify the relative merits of development near transit. What value can this tool bring for planners?

Transportation and land use are inextricably linked. Building a new rail line may expand development; new development may expand use of a rail line. The direct connection between the two makes differentiating between cause and effect difficult to measure. Transportation planners frequently make the argument that a new investment will produce new riders, for example, but whether those riders would have come anyway is not a simple question to answer. There is no counter-factual.

Nevertheless, planners have invested decades of considerable work in the pursuit of transit-oriented development (TOD), under the presumption that clustering new housing, offices, and retail will result in rising transit use and, in turn, reduce pollution, cut down on congestion, and improve quality of life. There remains some controversy about the effectiveness of TOD investments in actually increasing transit ridership, but, at least in my mind, the success of certain areas over others has as much to do with the manner in which developments are designed as the mere fact that there is construction adjacent to a rail or bus station.

For example, the considerable success of Arlington, Virginia in attracting riders to the Washington Metro, as compared to Rosemont, Illinois’ interaction with the Chicago L, is likely due to the fact that the former prioritized walkable construction immediately adjacent to subway stations while the latter put the rail line in the median of a highway, separated buildings from the station by hundreds of feet, and minimized pedestrian amenities. Getting the design of new development around transit right is often just as important as the transit itself in terms of attracting ridership.

If design matters, what has been missing has been a tool that offers empirical insight into the benefits of specific development interventions in terms of their effect on growing transit use. To fill the gap, a new tool for measuring TOD quality has recently been introduced by the Institute for Transportation and Development Policy (ITDP). It holds potential value in terms of defining the appropriate measures for creating effective TOD, but it needs further development to be useful in aiding the creation of best-practice development designs.

The TOD Standard

ITDP’s TOD Standard replicates the BRT Standard the organization finalized this year. Like the U.S. Green Building Council’s LEED, both are scoring systems meant to offer a quick measurement that allows projects to be compared with one another on a variety of relevant criteria.

ITDP argues that this tool allows planners, developers, and the public to assess proposed or existing projects on a wide number of measures. The report aims to identify “what constitutes urban development best practice in promoting sustainable urban transport.” In other words, the goal of the tool is to determine which development projects would do the most to positively expand public transportation usage.

From the perspective of planners who should be actively promoting urban design that increases transit use, the Standard’s recognition of “development that is pro-actively oriented toward, rather than simply adjacent to, public transport” is encouraging. If the system works, it could be an potent way to make the connection between transit and development more explicit and, if used by municipalities or developers to design projects, it could eventually result in expanded ridership.

Best practices are identified across eight design categories — walking, cycling, the transportation network, accessibility to transit, a mix of uses, density, connections to existing employment centers, and changes in parking and road use. Within those categories, 24 criteria can be analyzed individually and then combined into an overall score in which developments rate from -50 (terrible) to +100 (excellent). The tool could be used to determine how well a completed development compares to best practice, or to identify areas for potential improvement.

Vienna Town Center Plan

Testing the Standard’s Effectiveness

ITDP’s tool is designed for large-scale projects within 800 meters (one half mile) of transit stations that implement at least 20,000 square feet of new construction on an area of four or more square blocks. The tool is not meant to measure the TOD effectiveness of existing districts (which are supposed to be included in a future revision of the Standard).

To evaluate just how the tool works, I chose three large TOD projects in East-Coast cities to compare their relative advantages, and inputted project data into the Standard for comparison. The three projects I selected are the Lindbergh Town Center project in Atlanta, which is fully completed (photograph at the top of this post); the NorthPoint project in Cambridge, just outside of Boston, which is partially completed; and the Vienna MetroWest (aka Metro Town Center), outside of Washington, D.C., which is in planning (the plan is just above this section). The latter project is the now down-scaled version of the Vienna plan.* Each constitutes a major development program located immediately adjacent to a transit corridor.

The data that I inputted into into the tool produced the results seen in the following table. As the last row indicates, the MetroWest project scored most poorly, with a rating of 3 out of 100.** Lindbergh did much better, with a score of 39, and NorthPoint best, with a score of 56. It should be emphasized that neither NorthPoint nor MetroWest are completed, so upon actual construction, the final TOD scores could be significantly different.

Evaluating TOD Based on the ITDP TOD Standard
Note: These figures offer a “sketch” computation of each project’s TOD score; the score should be treated as a general figure, not a fully accurate measure of each project’s characteristics.
CriterionPotential PointsMetroWest, Vienna, VANorthPoint, Cambridge, MALindbergh, Atlanta, GA
Sidewalk access0-1000
Complete crosswalks0-500
Driveway interference1010
Active frontage101105
Permeable frontage2011
Cycle network0-10-3
Cycle parking at stations2200
Cycle parking in buildings2???
Cycle access in buildings1???
Intersection density2-1021
Small blocks5030
Prioritized pedestrian connectivity3111
Max walk distance to transit00-20
Avg walk distance to transit5555
Mix of uses10101010
Access to food1100
Affordable housing4022
Residential density10497
Non-residential density5000
Connections to existing urbanism10044
Distance to existing jobs5555
Off-street parking15050
On-street parking5???

Assessing the Standard’s Value

The tool was simple to use and its results make sense intuitively: Whereas the MetroWest project is poor urban design from the perspective of encouraging transit use, the other two are far more oriented toward the nearby rail stations. Hypothetically, if the projects were all proposed for the same site, the tool would allow decision makers to make a quick quantitative comparison between the designs and identify the best project for public transportation riders. This could offer a clear benefit in terms of, for example, choosing a winning team for the contract to develop a publicly owned site. Rather than rely on “subjective” comparisons of the aesthetics of site designs (a comparison that too often devolves into a question of individual architectural taste), the tool quantifies the physical.

The Standard could also play a useful role in improving the ability of developers to design their new transit-adjacent buildings most effectively by highlighting where plans fall short in comparison with best practices.

Yet, as beneficial as it could be, the Standard does not appear to have been developed with a clear research methodology to back its scoring system. Why is the active frontage criterion worth 10 points, but the amount of shade on nearby streets only worth 2? Perhaps I am wrong, but my sense is that residential and commercial density are the overwhelming influencers of transit use, yet those criteria only account for a quarter of the score. ITDP does not appear to have conducted a real-world analysis to demonstrate whether certain elements are more beneficial in terms of attracting transit use. Rather, the tool seems to have been created using a common sense approach, which is not as good as one might hope for a “Standard” that is explicitly designed to provide an empirical scoring system for measuring TOD effectiveness.

ITDP’s TOD Standard, though, remains a draft; it will be revised over the next year based on public and expert input. It would be beneficial if those revisions attempted to incorporate evidence about the relative effectiveness of the various criteria in terms of growing transit ridership.

Even so, for those who are already familiar with the basic principles of transit-oriented development, ITDP’s scoring system will do little more than reinforce already-acquired knowledge. Every urbanist knows quite well that good TOD requires pedestrian connectivity, a mix of uses, and bike parking — those goals might as well be imprinted on the foreheads of most people trained in planning. At most, ITDP’s guidelines may highlight slight differences between individual projects, but a quick comparison of the site plans of Vienna’s MetroWest and Cambridge’s NorthPoint is enough for most even unexperienced planners to make out which one is designed for transit, and which one isn’t.

So what added value does the Standard bring? Like WalkScore, it provides an “objective” number that can be used by non-planner decision makers to help them determine which projects would best fulfill the policy objective of maximizing transit use. The Standard must be refined, however, to focus on making that number into something that’s genuinely reflective of best practices.


* I recognize that the Vienna project profiled here may not be the project that is built; the original plan for the site envisioned much higher densities just next to the station. For this comparison, though, I wanted to pick a project that I hypothesized would score poorly using the ITDP tool, and this revision fit the bill.

** Because I could not locate data on bike access to buildings or the area devoted to on-street parking, the documented scores were really out of a total possible score of 93.

Image at top: Lindbergh City Center in Atlanta, from Cooper Carry Architects; below: Revised Vienna Metro Town Center Plan, from Paraclete Realty

12 replies on “Defining Clear Standards for Transit-Oriented Development”

“… the Standard does not appear to have been developed with a clear research methodology to back its scoring system.”

The way to do this would be to look at existing developments, run a model with the intent of finding which aspects contributed most to existing developments’ successes and failures, and then work off of that. Better I’d imagine that, for this to be effective, different scores for different kinds of development would have to be made (office-oriented, shopping-oriented, neighborhood-ish). Although that might seem to go against the whole idea of mixed use, large shopping and office districts do tend to congregate at nodes; there’s nothing that says those nodes have to exclude other uses, but there’s a big difference between someplace like Arlington and a dense, gentrified neighborhood.

My big concerns with this, then, are similar to those of the BRT report—the numerical scoring may not be well-aligned with actual performance measures, and the presence of a checklist does not ensure the planning process will be done right (a checklist for how to plan TOD—first make sure you’re near transit, then make sure you’re willing to be dense, etc—might be more useful and prevent superficial TOD elements from raising a development’s score). Similarly, the bar is so low for transit-oriented development in this country (much like bus infrastructure) that truly mediocre developments may be awarded silvers and bronzes when they deserve no praise at all.

This is a positive contribution, but as its authors recognize it’s a work in progress.

I am very familiar with the Northpoint site as it was decades ago (horrendous), and with the Vienna site today. To me, a key issue with the Northpoint site is the taming of Monsignor O’Brien Highway. Something missing from the scoring system is minus points for a suburban-style highway that cuts through the supposedly urban area. If anything, the “prioritized pedestrian connectivity” measure seems to reward sites that funnel car traffic into big highways (and make the byways pedestrian-only) rather than distributing traffic across a grid. I would de-emphasize that measure, or eliminate it, and give a lot more points for intersection density.

This is a thoughtful review of our TOD Standard v.1.0. We greatly value this feedback. Our tool is based on existing best practices and knowledge about efficient integration of land use and transport – something that cities in the US and Europe already do well – and translating it into simple principles which can also be easily understood and implemented in the cities that are seeing rapid urbanization and where we also work, in Latin America and Asia. As you have already pointed out, it is currently a pilot/draft which we are making publicly available in order to get feedback from professionals working in this space, so we would be pleased to receive any and all comments.

We are also currently creating a dataset of scored developments to be used in refining our scoring system, so we would be pleased to receive any data collected on developments tested. Thanks in advance and please do keep an eye out for the revised version in 2014!

Mariane Jang
Institute for Transportation and Development Policy

What came to mind when you described supposed TOD around a transit line built into the median of a highway is that the alignment itself is “Transit Relegating” rather than “Transit Oriented”. If the development inherits from the corridor, that would make it TRD ~ Transit Relegating Development.

Once again a scoring system is deviced without taking into account the development’s potential impacts on the quality of the transit service, especially in respectfully considering the repercussions to the “geometry of transit”, as Jarrett Walker puts it. At least it has better potential to avoid a critical flaw of the LEED-ND scoring system. LEED-ND disproportionately awards transit coverage at the expense of total daily rides actually serving the development. LEED-ND’s SLL Credit 3, can allow, for example, project teams to double count transit rides by splitting a stop into two stops to distend the service in the development – improving occupants’ access to service, yes, but eroding the quality of service. Planning for LEED-ND points can potentially conflict thus with effective stop spacing and alignment geometry for an overall alignment. By the sheer fact that this system rewards shorter peak-time commute duration, this potential of a development to erode transit service is better guarded against, but, unfortunately, not completely addressed.

It also bugs me that a “high capacity” service can outweigh frequent bus services in a metro area without taking an actual look at total daily rides serving the development (which at least LEED-ND can reward a project for). It seems we could come up with a more mode neutral way to consider the value of transit. We could actually expand the applicability of the system to greater urban areas while being more definitive of the metrics at the same time, so that effective transit planning can actually lead to development that can support the shift from the auto.

We just need to get transit planners in the room next time a scoring system is devised.

I find the scoring somewhat confusing. In many cases, I see “potential points” of 0, or of 2, yet a project which apparently lacks the desired features has a score of -5 or -10.

If a maximum of 2 points are possible, how is it possible to subtract 10? This makes no sense.

The entire scoring system appears not to function in an intuitive, or even defensible, manner. I’d recommend going to a zero-based system with no negatives, and spelling out the criteria for each level of points achieved. A maximum of 100 points makes sense; what doesn’t make sense is for the MetroWest project to receive a score of 3. This only occurs because of the deductions far in excess of possible points, and suggests that efforts made for MetroWest shouldn’t have been made at all. Just removing the “excess negatives” revises the score into the mid-20s. Still not good, but certainly better than a strip mall near a bus stop.

Lindbergh center is not complete. It still has great development potential around it and the potential of two light rail lines connecting to the area.

I’ve been wondering what, if anything could possibly be done to mitigate the effects of the situation with Rosemont station outside of Chicago. It sounds like somebody needs to study what could be done to make the situation more pallatable. Once upon a time there was supposed to be a PRT link between that station and the airport station on the WC commuter rail line. Had that gone through, it quite possibly could’ve lead to some kind of expansion that just might’ve helped matters there at least a little. I know there are a lot of people who are very prtial to the particular mode or modes of transportation they advocate to the expense og all others but from what I’ve been able to see, Rosemont station and the area around it would be just ripe for for a peole mover network, a PRT network or some kind of monrail netwrk which could and should feed passengers directly to the Blue Line.

Is this website’s staff no longer active? There has not been a new article since July 15, 2013, or 90 days, or 3 months.

This sounds a great plan in assessing the suburban design. But could this scoring system also be applied to an underground transit or would there be a separate scoring system to assess those? Since dvelopment around the areas are fast approaching how can this score system cope up with them? Every when are they updating this to know if this would give the most accurate data?

Leave a Reply