Edge Cities

Alon has a trio of posts up on the question of centralization versus dispersal of populations. His second discusses the formation of Edge Cities as a result of limits on CBD growth, and suggests that their formation is negative due to the burdensome commute they place on lower-income individuals who live on the “wrong side of town.” I share his concern, but think the problem is more intractable than a simple question of supply and demand.

So let’s look at two very different cities, Chicago and Houston.

Chicago is somewhat unique in that despite being a fairly traditionally-zoned city, in practice it historically has no height limits downtown. Transit service to the Loop is probably as close to perfection as can reasonably be achieved in the U.S., with multiple overlapping rail services geared toward destinations both near and far. And yet, Chicago has Edge Cities. Rosemont is familiar to anyone who’s flown into ORD, but Schaumburg is probably the most prominent. There’s also a smaller cluster around the 88/355 interchange. There’s no similar areas on the south side, although there is a UPS DC that’s large enough to warrant multiple dedicated express bus routes.

Houston, with no zoning or height limits, is even more unique. Granted, there are a variety of subdivision regulations and building codes which shape development in the Bayou City. And while there are no height limits in principle, the City’s reaction to the Ashby high-rise project creates a sort of “soft” height limit that discourages high-rises adjacent to affluent SFR. But this doesn’t apply in the CBD, where the land is platted and space is plentiful. In fact, nearly a third of the land bounded by 45, 59, and the Bayou is used for surface parking lots. The only thing holding back the supply of Houston CBD office space is demand.

And yet, Houston has edge cities. Boy, does Houston have edge cities.

Alon slags Houston’s nerd scene, but overlooks the fact that Houston is one of the best cities for engineers. Mechanicals are particularly well off thanks to the robust process market. Exxon’s Baytown facility alone pushes through more barrels of oil than every refinery in the state of New Jersey combined. Houston is also, of course, the national center for oilfield engineering. This raises wages for everyone, even Civils, because in lean times, the oilfield outfits will poach employees from the other disciplines.

Where are Houston’s engineering companies located?

A reasonable dataset can be created by looking up ENR’s list of the top 500 Design Firms and plotting the location of the Houston office for each one. Do this, and you get a weighted centroid of Houston Engineering Employment that sits slightly to the west of 77042.

Why is this? As near as I can tell, causality runs in the following order.

1. A certain percentage of people will move out as far as possible while maintaining a bearable commute. In post-1970 America, the stated rationale is often “to find a good school district,” but this appears to be a worldwide impulse. Wendell Cox has flown to a great many world cities for the sole purple of driving to the periphery and proclaiming that yes, there be single family homes here.

Likewise, in post-1955 America this sprawl takes the form of freeway- and tollway-fed subdivisions, but in the past this desire to get out was fed by streetcars, elevated trains, and steam roads. As late as 1953, the developers of Pennsylvania’s Levittown put in a new train station to serve the community. We generally don’t tend to think of Bill Levitt’s mega-plats as TOD, but the stop is still out there, begging to to disagree with you.

2. Having moved out to the sticks, outer suburban residents will suggest moving the office. Everyone is familiar with the sort of urbanite who moves into a loft in an industrial district, then immediately begins complaining that the warehousemen next door don’t keep the same hours they do. This is essentially the suburban version. And since the type of people who prefer suburbs comprise a large chunk of middle and upper management, they often get their way.

3. The process iterates. To put it in Houston terms, a West Loop or Greenway office enables you to live in Sharpstown or Memorial. A Beltway office lets you to move to Cinco Ranch or Mission Bend. An office on Park Ten lets you move out to Katy or Fulshear (which may eventually also be branded as part of Cinco Ranch).

4. Eventually, a limit is reached. This occurs when it is no longer possible to move the office any further out without cutting off younger workers who prefer to live in the core. Houston appears to be reaching this point. The reverse commute on Interstate 10 now generates worse traffic than the traditional commute, at least inside the 610 Loop.

Speeds on IH-10

Data from Houston Transtar; 3-month rolling average of Monday speeds as of 01/05/2015.

Since this limit is reached when the commute for residents of gentrified, inner-core neighborhoods (the young worker base) becomes as unbearable as that of the outer suburbanites, it follows (almost by definition) that crosstown commutes from the “wrong” side of town become exceedingly long. This imposes a disproportionate burden on lower-income residents.

But what can be done?

Urban Growth Boundaries Don’t Work

UGB imposition is effective at curbing sprawl, preserving farmland, and the like. But it doesn’t eliminate the desire for management to move the office out to the ‘burbs. Indeed, this appears to be stronger than the desire to live there. In the western suburbs of Portland, Oregon, an Intel campus and various high-tech companies along US 26 give way to farms. In Portland’s southeast, a substantial agglomeration of light industrial sits just across the river from wilderness.

Nor can this be solved by zoning. Lower- and middle-income residential tends to be a money-loser for cities, who have to provide good schools, miles of small-diameter utility lines, and a police and fire staff large enough to deal with squabbling families in tinderbox houses. Office and industrial, on the other hand, is a cash cow. Given a limited quantity of land to allot, the financial returns to most cities will simply be too great to stick to residential on principle.

Crosstown Highways Do

Outer-suburban office and manufacturing clusters don’t tend to concentrate near rail stations. So if you’re stuck with a bunch of car commuters, highway expansion may be the least worst option. Urbanists may bristle at the thought of a social justice argument for more highways, but a debate over freeway removal in Dallas has circled back to that point several times.

Mass Transit Can

It’s possible for mass transit to serve crosstown commuters from poorer suburbs. The gold standard here is purpose-built rail lines, on the model of the Grand Paris Express. A reasonable middle is radial rail lines which go underground to move quickly through the core. DC Metro is a good example; the ride through from Branch Avenue or Largo Town Center to suburban employment in Silver Spring or Bethesda isn’t any worse than the slog from Jacintoport to the West Belt along IH-10, which two of my coworkers do.

The worst example of crosstown rail connectivity is Portland, Oregon. MAX light rail drops from a design speed of 55mph on most radial alignments to approximately 8mph through the downtown core, a design choice made to help “activate” downtown screetscapes by having passengers wait for trains at curbside. This adds up to 20 minutes each way for the H1B crowd, which is stuck reverse-commuting to their Westside tech jobs from affordable housing east of 82nd.

Rail isn’t the only option, of course. Portland’s TriMet historically had several peak-hour bus lines which ran directly to the hospital complex on Marquam Hill, although these non-Jarrett Walker-approved routes have tended to fall off the map in recent years. And Houston’s “reimagined” transit system will soon add a 15-minute all-day express route which runs direct from Downtown to the Memorial City area, with a single stop at Northwest Transit Center.

In fact, while much is made of rail bias, 45-foot commuter motorcoaches which directly link low-income suburbs with suburban employment may be perceived as higher-status than a rail-station “shuttle bus,” which is almost invariably the sort of cutaway-chassis affair that is usually associated with paratransit. If it’s good enough for Google, it oughtta be good enough for you.

5 thoughts on “Edge Cities”

  1. I read somewhere that when a company moves its HQ, in something like 85% of cases, it ends up moving closer to the CEO’s home. So there’s probably something to your theory. Also, the difference between a cutaway and a 45 foot motorcoach is at least as big as the difference between rail and bus, in terms of comfort and desirability.

  2. I’d argue one big problem with edge cities is their geographic location on the periphery does not allow them to capture all possible workers in a metro area. Now, for many smaller metro areas this isn’t necessarily a problem. But you referenced Chicago, my hometown and current residefnce, and in particular, Schaumburg, the classic definition of a edge city. About 20 years ago Sears moved their corporate headquarters out there (in nearby Hoffman Estates). They soon after lost almost 50% of their workforce. Because when they were headquartered in Sears Tower downtown, they attracted workers from throughout the Chicagoland region. After the move, it became difficult to attract workers, particularly those in the south suburbs and many parts of the city. This is because of geography – located some 30 miles northwest of downtown Chicago and 10 miles from the nearest Metra station, it proved very difficult for Sears to retain its workforce.

    In my own profession as a transportation planner for a major engineering firm, me and my colleagues are all located downtown to take advantage of talent that lives throughout the region. Many other industries, particularly in tech and finance are similarly located and this bears out: downtown Chicago’s office vacancy rates are among the lowest in the whole metropolitan area.

  3. UGB imposition is effective at curbing sprawl, preserving farmland, and the like. But it doesn’t eliminate the desire for management to move the office out to the ‘burbs. Indeed, this appears to be stronger than the desire to live there.

    I disagree with your interpretation here. All that proves to me is that land is cheaper on the outskirts. Here’s a counterexample I happened to read about recently.

  4. You mentioned that the EB congestion in the evening being worse than WB is a shift away from the historic norm; do you have data to support this? A similar pattern can be observed on SW US59 inside the loop. I hypothesize this isn’t driven by an imbalance in suburban vs urban commuters, but the fact that in the evening, drivers are not always headed home. Potential reasons for influx of non-resident traffic into the loop on weekday evenings: school (UH-D, Rice, St. Thomas), patronizing restaurants/bars, working at restaurants/bars, Memorial Park/Buffalo Bayou, etc

Comments are closed.