Friday, November 03, 2006

Comparing Atlanta, Dallas, and Houston

Today's post is a little different from our usual at Houston Strategies. It's an email dialogue I recently had with Steve sparked by the Chronicle article on downtown Atlanta's struggles. This is also a much longer post than usual, but I think anybody curious about how cities evolve in both good and bad directions will find it pretty interesting. Hopefully a broader dialogue can continue in the comments.

First, one quick side note: don't miss Christof's analysis of Metro's planned intermodal terminal north of downtown. Some excellent detailed analysis. My thoughts on it from earlier this year are here.

On to our dialogue, starting off with Steve:

Houston should feel VERY fortunate that its downtown vacancy rate isn’t similar to that in Dallas and Atlanta. Also, downtown Los Angeles’ vacancy rate, though similar to Houston’s, is much higher than in non-core locations. Some thoughts on this:
  • Central Houston has remained central geographically to the region’s growth, unlike in the other two cities, where growth has been very lopsided to the north (an outgrowth of racial bias in my opinion).
  • The wave of corporate consolidations in energy and finance / banking seems to be over for the moment in Houston. It did have a significant negative impact on downtown and other urban core office markets here. It’s still occurring in Dallas and Atlanta, apparently. Not that we couldn’t still see more of it here though.
  • Dallas and Atlanta both have strong freeway access to their downtowns, just like in Houston. Atlanta is particularly notable for its giant freeway “connector” through downtown. However, with such employment decentralization (and of course residential decentralization), residents from outer suburbs – still the premier destination for upwardly mobile families – just have to go through too much congestion to access downtown, from the outer suburbs all the way in. In Atlanta (and Dallas too), the first-ring suburban employment centers suffer the same problem. It just drives employers further and further out. Upscale residential densification in the core helps to slow this a little maybe, but upwardly mobile families with school age children are still the employees you have to please – they’re often your most valuable human resources. Freeway extension and expansion seems to alleviate this problem at best temporarily if regional growth is strong. Plus, the more you accommodate cars coming into downtown, the more your parking demand…
  • The highway problem means you’re left with transit solutions as one of your few remaining options. We are especially fortunate that METRO covers most of Harris County, including all of the unincorporated area. Harris is a very large county geographically, so that means METRO can put park-and-rides and HOV / HOT lanes pretty far out, and the service in most corridors is really exceptional. The weak spots are Fort Bend and Brazoria Counties, which really aren’t very far out. The Woodlands has its own commuter bus service. The transit providers in Dallas and Atlanta are limited to much closer-in coverage. For example, in North Texas , McKinney, Frisco, Allen, Flower Mound, etc. – where enormous middle and upper class residential growth (and now employment growth) is occurring – are outside DART’s reach. Houston’s downtown would be much, much worse off without park-and-ride service.
  • Dallas and Atlanta have significant low-income, high-crime neighborhoods within and adjacent to their downtown areas. Houston’s comparable neighborhoods are a bit more removed from downtown for the most part, helping make downtown a relatively low-crime area. We obviously still have a homeless issue to deal with however.
  • Many employers (a U.S.-wide phenomenon) are looking for a different physical building type than what you find in a downtown. They want larger floorplates to accommodate a high employee density. This means they also need a high parking ratio, even with transit access. Lower lease rates too – trophy buildings / real estate are often now considered to be a negative financially. Pretty hard to satisfy such needs in a downtown. The more a region is dominated by such companies, the less likely is downtown office space to be in high demand. San Jose is a great example of this.


Steve, this is a fascinating article. Thanks for the heads up. I had no idea Atlanta was hurting that badly. 25% of their crime downtown? Ouch. And AT&T will probably lay off a bunch of BellSouth employees.

I pretty much agree with all your very well thought-out points. The lopsided growth is one I've mentioned before but people don't seem to recognize what it does to an old core city (also see: Detroit). Having only one freeway to the popular northern suburbs from downtown is extremely limiting - Houston's downtown has many more feeder freeways from desirable suburbs. Obviously MARTA didn't save downtown Atlanta as commuter rail supporters would have you believe. You're right that it's good Metro's area is big, but they should be trying to figure out how to make it bigger, at least for commuter services from Montgomery, Ft. Bend, and Brazoria.

Metro's announcement of tolling the HOVs will help downtown Houston. Obviously one set of users will be top executives that can afford the toll no matter what it is, and if they have a reasonable commute, they're a whole lot less likely to look at moving their office to the suburbs.

Can you explain to me the floorplate/employee-density issue? Maybe with a couple good/bad building examples?


OK, prepare yourself for a non-compact response…

I do have to wonder how Atlanta’s “downtown” is defined…are they including adjacent residential areas that may have much higher crime than the CBD proper? I don’t know.

Atlanta actually three radial freeways serving the northern suburbs, and two of these have spurs that split off to get more coverage. However, because the desirable residential areas have vaulted so far out, the “opportunities” for congestion if you’re going downtown are just too numerous. Really, how likely is it that a freeway in a high-growth corridor will be congestion-free for a 25-30 mile distance (or longer) during peak hours? Especially when you get a lot of intervening employment centers along the way (see the Georgia 400 going north out of Atlanta plus other examples). More than likely, you’ll hit the brakes at least two or three frustrating times. Short of building mega-facilities (minimum five through lanes each direction) for several 30-40 mile radii from downtown, plus the associated intervening circumference loops, it’s hard to see how you avoid this problem for solo drivers looking to use a “free” facility. And the mega-facility idea is obviously extremely unlikely under current funding constraints. Plus with mega-facilities you have to have mega-ramps at major interchanges and mega off-ramps to surface thoroughfares, plus lots of separations to minimize weaving, since otherwise each create choke points and congestion…you can see the sort of perpetual-motion nature of the problem. Not to mention where do all these cars coming into town from mega-facilities park? Structured parking has gotten very expensive.

HOT can address this for the highly paid executives that you mention, and that’s certainly important. At some point though, you have to wonder if even some affluent commuters would balk at the tolls required an HOT facility to keep it congestion-free for that long a distance: what if the 30-mile price during peak hour reaches $8 or more each way? Then their alternatives are (a) going back to the old congested freeway or (b) taking transit (ostensibly on the HOT facility). However, MARTA barely goes outside the Perimeter 285 freeway and DART has similar limitations as I’ve already mentioned, so they’re going to be fairly minimal help to their downtowns. People have to drive through miles of congestion just to get to a park-and-ride station (and hope the parking lot isn’t full). Plus you’ve always got a market segment that won’t take transit regardless of quality of service (I don’t have much sympathy for those folks unless they truly have special needs that prevent their using transit).

You’re right about METRO needing to expand – the lack of service to Fort Bend, Brazoria, Montgomery, and maybe even Galveston counties will become a real problem for downtown employment, as well as other urban core areas that are or should be served by METRO commuter buses. Even with HOT expansion, in the end, good commuter transit becomes essential.

The floorplate / employee density issue isn’t necessarily one that I have in-depth expertise in, but I can tell you that certain industries seem to prefer low-rise, low-cost office buildings where they can have “cube farms” – fewer employees have enclosed offices and window views. Some have mid-rise buildings or taller but prefer to be isolated in a campus environment. It may be a cost thing in some cases, for others it might be related to their business and innovation model. Silicon Valley is dominated by these low-rise users and you see similarities in Dallas and Austin with their high-tech employers. And in general, across many industries, you’re seeing an increase in employee density within buildings, primarily to save on occupancy costs. Skyscrapers typically have smaller floorplates, part of which are taken up by elevators and HVAC, etc., so it’s harder to accommodate this density. Plus, the more employees in a building, the more parking demand, unless you can increase the transit share (not easy). Most of the high-rise buildings built up through the 1980s have parking ratios of 3.5 spaces or less per 1,000 sq.ft. of leaseable space. Most new office buildings in the suburbs have ratios over 4.0 and as high as 6.0. Not to mention employees at suburban buildings typically don’t have to pay “extra” out-of-pocket for parking – not good policy perhaps but it is the current market reality. All of these things make existing dense urban centers less competitive for such office users.

In Houston, many of our energy firms seem to be able to make this work in high-rises, which benefits our urban core office centers since that’s what they offer. Law firms and finance / banking also do well in high-rises and can afford to pay top dollar for Class “A” space (to a point). Another reason downtown / Uptown / Greenway Plaza etc. are fortunate.

To conclude this long response, I think we should step back and look at the “big picture.” The real root of the downtown / urban core problem is that upwardly mobile, middle and upper-middle income families with school age children – a minority of the market to be sure, yet still so important because of the emphasis we place on raising our children – just don’t seem to find the residential product they want in the urban core, the inner suburbs, or even the middle suburbs. They want the outermost suburbs, for the most part. I know there are some exceptions, but the relatively few closer-in neighborhoods considered “acceptable” for such consumers tend to be very expensive. In low-density places like Houston, Dallas, and Atlanta, with tremendous affordable single-family detached housing stock even in their urban cores / inner suburbs which should provide the needed physical product, this problem is particularly dismaying. The geographic distances between urban employment / cultural core and desirable fringe become huge very quickly. And middle-income suburbs seem to cycle into undesirability after only 20-30 years of existence (if that). Short of radical changes in the perception (and reality) of public schools and crime in our inner / middle suburbs, it’s hard to see how this problem will go away in the foreseeable future.


I basically agree with all your arguments. I see what you're saying about 3 northern radial freeways, but they all collapse to one well before downtown. That makes commercial office centers farther north at major intersections and before chokepoints to be much more attractive. MARTA/DART won't help. In my mind, the only option is HOT lanes with express bus and vanpool service from many points in the far suburbs. But even that, I think, is only good for holding on to existing downtown employers - and probably won't attract new ones when there is cheaper, more accessible, newer office space in the nice suburbs closer to workers.

I might guess that part of the attraction of low-rise is lower construction cost and easier surface parking without the expense of giant parking garages, not to mention the annoyance of winding your car up and down them every day. It's a good point that Houston is helped by a hefty dose of very big companies, inc. F500 HQs, and they want a lot of contiguous office space, which drives them to skyscrapers instead of just low-rises (see the recent Chevron deal downtown). BP has announced they want to add 2000 employees at Westlake - you don't do that with low rises. Places like Dallas, Atlanta, and Phoenix have more midscale companies (like Silicon Valley), and I think they're perfectly happy with low-rises.

I very much agree with your second to last paragraph. I think Houston is helped tremendously by our limited zoning/permitting regulations, which allows the townhome phenomena - an attractive housing product in the core, at least to households without children (a rapidly growing market, to be sure). Most cities, I think, have a hard time enabling teardowns/townhomes because of powerful neighborhood NIMBYs and the governmental mechanisms to stop them (zoning/permitting regs). As I explained to Joel the other day, the flip side of our core flexibility is that we don't have the downtown housing demand of other cities because there are so many affordable townhomes available. In other cities without that product, people are driven to condos downtown - one of the few places they can get built.

The aging middle suburbs are problematic, but they're also a source of affordable housing. As Reason has noted, it's much better to allow unfettered new, upscale construction and then let aged residences become affordable than to mandate new affordable home construction, which tends to pretty much shut down almost all construction. I think the best a city can hope for is a vibrant but upscale core, affordable middle suburbs, and upscale outer suburbs - with employers in both the core and the far suburbs, and commuters going both directions to get better freeway utilization/balance. I think the outer affluent residents/employers is a given (i.e. it's inevitable in almost every healthy metro), the question is whether a vibrant core can be maintained (see Dallas, Atlanta, Phoenix, and Detroit) - and that, in my mind, depends on good HOT transit to hold on to the legacy employers that are already there. I think the young would like to live in a vibrant urban core (and would be ok with a reverse commute), but if it's not there or not affordable, they'll end up in these suburban town centers like The Woodlands and Sugar Land.

And, of course, schools are the linchpin - but I don't see any imminent breakthroughs there.


A few more quick points, and then I think we need to draw this to a close, before we’re overwhelmed!

I agree with your HOT assertion, the point about new employers, and about the large block users. The only caveat is that if there is a glut of high-quality office space downtown and there is rapid employment growth, you can sometimes get new smaller employers downtown by default, because that’s where immediate space needs can be best satisfied, often at very affordable lease rates if occupancy is really low. How sustainable such occupancy is becomes questionable though.

Absolutely second the townhome / no zoning point. Have a large set of neighborhoods within the core that are attractive to childless households helps the core employment centers too, to an extent. Reverse commuting certainly helps utilize our transportation facilities more efficiently too. Interesting that the Katy Freeway outside Loop 610 actually achieved overcapacity conditions in both directions from this phenomenon.

The affordable housing benefit of aging suburban neighborhoods is helpful, true. And I am generally opposed to regulatory measure requiring affordable housing allowances, inclusionary zoning, etc. The problem is that while the location of the employment core doesn’t really move, the desirable outer ring does – sometimes disturbingly fast. It would be of great benefit to core employment centers to be able to house a greater share of affluent families with school-age children in the inner and middle suburban area. That would still leave room for a lot of affordable older neighborhoods. It makes me wonder, once a single-family suburban neighborhood becomes undesirable from a schools or crime standpoint, can it re-emerge as desirable for affluent families? While we have some subdivisions in our inner suburbs that were formerly in “free fall” and have recently stabilized (Westbury is a good example), they’re still not considered desirable for families with kids in junior or senior high, and in some cases not even for those with elementary school kids. Can expanding the roster of private / charter schools be part of the solution?


It's not just schools and crime - affluent families want newer, larger, more modern homes, and those are much more affordable in the far burbs, along with better schools and lower crime. There are places like Bellaire and West U that get the teardown/rebuild biz, but it's very pricey. Better schools would help, but they can only do so much. More charter schools might help, but even then, most charter schools are focused on disadvantaged kids (as they should be) rather than the affluent - so I'm not sure they would accomplish what you're looking for.

> the desirable outer ring does (grow) – sometimes disturbingly fast

Not as fast as you might think. Check out one of my earliest blog posts on this topic.


Last points – yes I agree about the tear-down-and-rebuild vs. buy an existing old home and live in it vs. new production home price trade-off – I actually thought of that after I had already sent my last reply. This economic reality definitely factors into it too. Not to mention that newer subdivisions are often more in tune with consumer tastes regarding amenities (playgrounds, walking trails, water features, etc.) that often can’t be retrofit into older subdivisions.

I also agree about the mathematical property of ring area – more sq.mi. of area for less radial distance the further out you go. The problem is that school districts are “lumpy” and once their perceived reputation goes downhill (can happen in a matter of 5-10 years), you’ve suddenly lost a big chunk of real estate with appeal to a certain demographic. This applies to specific high schools also, but on a smaller geographic level.


At 8:57 AM, November 03, 2006, Blogger zip77077 said...

This is a great post, thanks for putting it up.

My wife and I have just gone through the decision making process of where to move. It was tough and we are still not 100% happy even though our new home is under construction. The factors that influenced us the most were: cost, commute time and convenience. The cost and commute differences between inner loop and outer suburbs cancelled each other out. Convenience was the back breaker for in-town living. Day to day activities like grocery shopping, getting to schools, etc. just seem to take too much time in town due to congestion. Plus when you add in flooding issues and deteriorating infrastructure the outer 'burbs are very attractive.

At 9:28 AM, November 03, 2006, Anonymous RedScare said...

I find 77077's complaint about in-town congestion being a problem very interesting. I found the opposite to be true. Because suburban developers only build streets for their subdivisions, not being required to upgrade arterial streets, the congestion in the suburbs was far worse. Examples include FM 1960, Mason Road, etc.

Inner city street grids tend to carry more traffic in a less bottlenecked manner. I never encounter congestion while running errands, unless I have been sentenced to the hell that is the Galleria.

The cost and commute factors are easy to agree with, however.

At 11:46 AM, November 03, 2006, Anonymous Mr. Safroni said...

Great Post, very interesting reading. Virtually all of the core problems seem to have been addressed with the exception of several Elephants in the room that no one wants to address. One is the irony of the outer ring development, i.e. the energy issue. The larger homes/lots further and further out are being built at a time when energy prices are spiking and will continue to be a problem. It's as though we are planning and developing a "future" that will never truly exist because the costs will overwhelm the buyer. Serious initiatives need to address the need to re-develop the "outer loop/inner ring" subdivisions. Serious attention needs be paid to good legislation which would incorporate land use planning regulations for future developments such that Developers themselves would be held to new, stricter, energy efficient and environmentally friendly rules of development. Presently, the Developers are proceeding along with a format more appropriate to the 1950's, i.e. cheap Gasoline and Electricity. If they were held to a 21st Century standard whereby the costs would be adjusted to account for both the environmental and energy footprint of the individual unit and the layout of the Development as a whole, we would begin to see Develoopments that more appropriately address the realities of the 21st Century and beyond.

The other Elephant in the room (no political analogies intended) are the socialogical problems of school and crime. The outer rings deteriorate in a 20 year cycle because people move to them to escape "the Others" only to find that within 20 years, "the Others" have followed them. This trend is accelerated by new HUD policies whereby HUD builds and licenses Section 8 housing in the new rings bringing "the Others" with them. Shortly thereafter, property values fall and school districts fail and the cycle begins anew. I know of no solution to this problem. The European solution has been to build gov't housing far beyond the outer rings and encourage re-development of the Inner Rings, but that solution, while it has many advantages, seems best suited to Europe where land space is severely limited. It is an answer we might want to start looking at here in the U.S. however. When coupled with strict limits on taking land out of agricultural use for the purpose of building new housing we could at least make re-development more attractive to the Developers.

At 12:16 PM, November 03, 2006, Anonymous Robin Holzer said...

I appreciate the wide-ranging discussion of issues in this post a lot. I agree that great connectivity -- not only for car access but also transit, pedestrian, bikes, etc. -- is critical for sustained and sustainable city growth. I agree also that rising energy costs make the connectivity issue more pressing.

However, I am struck by many data-free generalizations, both in the post and the comments.

Anecdotally, my "in-town" experience is similar to RedScare's. I am 3-5 minutes from 4 grocery stores, 6 drug stores, and a dozen convenience stores. I can conduct all of my daily tasks without getting on a freeway and the convenience is outstanding. Heck, I can walk for several of my errands. But I would never assume that every "in-town" neighborhood enjoys these benefits, nor would I assume that some further-out neighborhoods don't.

The Houston region is comprised of widely disparate places. There are suburban neighborhoods inside the loop and urban neighborhoods outside the loop. There are good schools and bad schools both inside the urban core and in the farthest suburbs. There are crime challenges of varying flavors all across the region.

Tory and Steve both talked about "upwardly mobile families with school age children" and the attractions of new suburbs. Anecdotaly, I see a LOT of these families in my urban neighborhood, shopping at Whole Foods, etc. HISD is now expanding the local middle school as a result of gentrification and changing demographics.

But we shouldn't make policy based on anecdotes. It should be easy to pull U.S. census data for the region and see where these families live. Personally, I would expect that families with children make up a larger share of total residences in the outer 'burbs, but that there are more of these families in absolute numbers in the central city, simply because there are so many more people in central Houston.

Further, the Urban Land Institute estimates that anymore, only 1 in 4 households includes dependent children. As a result, the impact of families with children on regional growth patterns is diminishing significantly.

Finally, neither school quality nor crime rates are defined by geography. As Jane Jacobs explains, vibrant streets with engaged neighbors and activity throughout the day deter crime, whether in a central urban neighborhood or further out. Schools with a stable population and engaged parents fare much better than schools with high turnover and disengaged parents, independent from socioeconomic status or urban/suburban location.

It seems to me that most of these issues demand more than an urban/suburban heuristic. So while I do appreciate the discussion, I hope none of us run off to make policy on this basis!

At 3:09 PM, November 03, 2006, Blogger Tory Gattis said...

Great comments! A few random thoughts:

-In my experience, congestion for errands is worse in the burbs than in the core. Fewer freeways and underdeveloped roads from their previously "rural" status. Parts of west Houston, esp. The Galleria, are definitely an exception.

-People will take energy into account with their purchasing decisions. No need for govt mandates. I think gas is now less than 5% of household income - i.e. not big enough to overhaul people's lifestyles. Buy a smaller car with better gas mileage, sure. Give up homes in the suburbs, not likely. Employers will move out to them (or allow telecommuting) way before there is some sort of mass abandonment of the burbs.

-I do see more *affluent* families in the core too, but I still think they're a tiny fraction of the ones in the burbs. Plenty of families in the core, but fewer of affluent category.

-I think it's a fair rule of thumb that the farther burbs have lower crime and better schools. The Jacobs observation is a little archaic, based on dense village neighborhoods of the 1950s, usually ethnically oriented. The gated community with their own guard or constable replaces "eyes on the street" in modern times.

-the 1 in 4 households has children stat: makes sense, just take a hypothetical couple. 20 years as singles (2 households), 20 years married with children, 20 years empty nesters. That's 80 total household-years, of which 1/4 is with children. The thing is, the mass majority lives in a rental apt for the first phase, in a suburban house for the second, and just retires in-place with that house for the third (despite a few anecdotes about empty nesters moving into the city). Those apartments don't take much space compared to a house, so the spatial demand for residential is still overwhelmingly tilted toward suburban, even if families with children are only 1/4 of the households. I'm not saying the urban core is not hot (it clearly is), it's just a small but very visible piece of a much bigger residential picture including absolutely vast suburban areas.


Post a Comment

Links to this post:

Create a Link

<< Home