My first brush with transit oriented development – the first time it really dawned on me as to how it worked – was at Kew Gardens station in London. I went looking for a train station, but I found a transit village. I was hooked.

Walking from the train platform to the entrance to the Kew Botanical Gardens (more than a quarter mile away-my reason for being there in the first place) is everything urban living should be. There is of course the rail service and train station itself, and outside of that a Tesco Express (small grocery store) and a combination of retail, including a coffee shop, restaurant and local businesses in stately row along the Station Parade. These stores serve many of the community’s needs. Surrounding this are rowhomes, single-family homes and flats, all in a mixed-density and walkable environment. Interestingly, it is exactly one quarter mile from the tube station to Kew Road, the edge of the neighborhood.

I spend a lot of time visiting transit villages. In fact, I live in one in Minneapolis. Some are better than others, and I’ll admit, mine (the 38th Street station) needs work. A good transit village (not just a train station) is a artful balance of design and density. In my observations, I cannot help but wonder if we are selling ourselves short in terms of density.

My colleagues Jon Commers, Peter Musty and Michael Lander just released a quick snapshot of what should make up a good transit-oriented development. You can see from reviewing this, they break TODs in to three mixed-use scenarios, the Village, the Town and the City, and all are quite dense. Among other uses, the Village contains 1,650 housing units, the Town 4,000, and the City 6,000 units within a quarter-mile radius of the station.

How does that compare with my favorite TODs? Let’s look around. The Center for Transit Oriented Development (CTOD) has a great database with, among other things, housing unit counts within a half-mile of station areas across the country. However, these figures are for 2000, so they provide a great baseline but don’t include units at TODs that have been built since. For example, Englewood outside Denver has 856 households in the CTOD database. Add to that the 438-unit apartment project by Trammell Crow and you get 1,300 units. The Mission station on the Gold Line in Pasadena is a great little village, and has 2,505 according to CTOD. Add the 67 units at the Mission Meridian condo project and you are just shy of 2,600. Downtown Plano in the Dallas metro had 819 units in 2000, and adding the 463-unit Eastside Village and another 500 units still planned may get the unit count up to 2,000. Orenco Station, the beloved TOD in suburban Portland, Oregon, was just in the news because its residents are still driving to work. Indeed, a major reason may be that not all of the 1,800 units are within a half-mile, much less a quarter-mile of the light rail station.

You get the picture. Across the country, most half-mile areas around light rail stations have fewer than 3,000 units within a half-mile of stations, which is in all likelihood far below the threshold of 1,650 units within a quarter-mile promoted by my colleagues for the baseline Village. Don’t get me wrong, these are good TODs, but they need additional intensity.

Not all transit villages are selling themselves short. Older TODs like the Francisco station in Chicago have over 7,000 housing units within a half-mile. As well, the Rosslyn-Ballston Corridor, a favorite among TOD policy enthusiasts in Arlington, Virginia, does a little better. In 2000, those station areas had between 2,500 and 9,000 units each, with more added this decade. Much of this development (well over 10,000 housing units, not to mention 15 million square feet of office) has occurred in the past 30 years following a conscious policy decision to intensify development around stations.

What station areas in the Rosslyn-Ballston corridor may be lacking in terms of design is more than made up for by very high ridership and modal splits, and the fact that Arlington County gets a disproportionate amount of their real estate taxes from the intense uses along the corridor – good policy indeed!

That is why I believe the Twin Cities, among others, is selling itself short. We are planning two additional light rail lines here in the next six years. Both lines will add a little over 30 rail stations that previously did not exist. The planning for each corridor includes a general vision of intensification of land uses around each station area.

The average number of new housing units around each station area is around 1,000, give or take, by 2030. The total varies by station, but overall the 30 new stations may accomodate 30,000 units. That seems like a lot, but it is just 10% of the 300,000 new households forecast in the Twin Cities in that time. Considering the lack of other road and freeway investments planned in that time, why not go for 25% of the total? Or one-third? Increased intensity around each station will accomodate more of the metro area demand for housing and increase transit usage, reducing strain on roads and lowering VMT. It will also result in more support for retail and other uses to create complete transit villages. For example, it takes a population of 10,000 to support a small grocery store. 1,000 doesn’t make much of a dent.

Local residents aren’t always nuts about more density, but if we are investing billions in transit, we should be maximixing the bang for our buck. I will delve deeper in to this in the next few months, and I welcome any conversation. The bottom line is, we are selling ourselves short if we just build train stations. We should be encouraging the sort of density required to create true transit villages.