Let’s just go through his latest:

I do not buy into the idea that Tysons Corner is a lot like central Arlington.  Last night I did some field research and walked through the major developed parts of Arlington (Rosslyn aside) in one fell swoop.  I’ll put the rest under the fold…

As I had expected, I could do the whole walk on little more than one street, Clarendon Blvd. (it is called Wilson going the other way), with but a bend at Fairfax Drive.  I never had to cross a major road, run across a road, come near a major highway, or circumnavigate a major shopping center.  The main artery was straight enough to be followed by a single line Metro line throughout.  I was never more than seven or eight minutes from a Metro stop, if that.  And if I had done this walk thirty years ago, while the buildings and shops would have been very different, and many fewer, the physical geography of the walk would have been the same.

I have written not that Tysons is a lot like “central Arlington”, but that in terms of land area and employment the two places are quite comparable. If they were a lot alike, the Tysons redevelopment plan would not need to incorporate the many infrastructure improvements that are a part of the plan. To say Tysons isn’t exactly like Arlington right now is like looking at a before and an after picture and noticing some differences.

I don’t know why Tyler thinks linearity is so important. It isn’t. Connectivity is important, and that’s what the introduction of a grid system within Tysons is meant to address. A look at the Tysons plan does show, however, that it will wind up more or less as linear as the Orange Line Corridor, with 123 and then 7 providing the spine (with a bend where they meet).

The Beltway does create an unfortunate barrier across Tysons, and that is the most difficult design aspect of the project. But as unpleasant as freeways are, the Beltway won’t be a killer.

In contrast look at Tysons Corner (you have to type in “Tysons Corner” yourself).  The whole area is carved up by major roads, including three significant highways, one of which could be called massive.  Try crossing Rt. 123 at Tysons Corner or try crossing Rt.7. Even some of the “small” roads on this map are harder to cross than is the main Clarendon/Wilson thruway in Arlington.  It’s not just the roads and the overpasses; crossing or circumventing either major shopping center is a daunting experience.  Furthermore very little is laid out in a line and thus the presence of Metro stops (right now there aren’t any) would not cover the area nearly as well as they do in central Arlington.  Tysons is more like a large box with distant extremities protruding, all laid on top of some multi-level and impassable thick bones.  Overall there is plenty of this, except it’s usually much busier.  There’s also plenty of this.  Making Tysons denser in residential terms, whatever its virtues, won’t eliminate those barriers and in some ways the current plan will make them worse.

I am beginning to wonder if Tyler has actually looked at the redevelopment plan. It does not naively assume that density will magically solve the current pedestrian unfriendliness of the road network. It actively changes the road network.

How? As I mentioned above, the plan will create a system of streets creating a grid off of the main arterials. This will allow local traffic to get around within Tysons without using the main roads, it will increase connectivity (allowing pedestrians to get around more easily than they can now, amid the network of superblocks), and it will allow pedestrians and cyclists to move through Tysons without needing to walk along 123 or 7, should they wish to avoid them. The one exception is in moving across the Beltway, as mentioned before, but if one wishes to avoid walking along the arterial there, Metro is always an option, as are local circulator buses.

The streets, including the main arterials, will be redesigned to be more pedestrian friendly. Broad sidewalks will be built on either side, and where roads exceed a certain width, medians will be built (to accommodate Metro as well as pedestrians).

But Tyler, do note: the whole point of the redevelopment plan is to rectify the problems you identify. Next:

Now let’s turn to the debate.  When Ryan Avent writes: “Tyler seems to approve of the fact that a local planning board will dictate the size of buildings which can be built [at Tysons]” I would offer a different narrative of what I believe, also citing my comment on Matt’s blog.

“We made past mistakes, we won’t institute congestion pricing or other congestion ameliorations, local government is a cesspool of rent-seekers and homeowners, and so we’re stuck for the foreseeable future, on top of which the public choice critique means that even apparently sensible deregulatory pro-density plans will in practice be turned into additional subsidies for suburban growth, the latter observation in fact being derived from a broader point frequently offered up by Matt Yglesias in a variety of other contexts.

Believing the above paragraph is not well described as “favoring regulation and subsidy.”  I think it is also a deeper understanding than:

“Let us build more densely in the most congested areas and it will work out for the better, even though road policy is terrible and lobbyists will de facto control all plans.”

For many years privatizers and deregulators have been criticized for moving too quickly, before the right conditions for reform are in place.  China has been praised over Russia, etc.  Some privatization and deregulations have indeed backfired and maybe this one would too, unless it is done properly and that means done in conjunction with good roads policies.

I wish Tyler would unpack this a bit. There’s a lot going on here, and I can’t say I understand entirely what he is getting at. Let’s first note, as I did above, that roads policies are changing to accommodate planned land use changes. Let’s then note that the question at issue is a very simple one. The original plan was to allow significant density in Tysons as part of the redevelopment. Planners charged with creating zoning rules for the area are now saying that that level of density is too high, and that tighter limits should be placed on building sizes. In neither case will builders be forced to build up to the limits, or indeed, to build at all. But they will not be allowed to go above the zoning limits. This is bad policy, and it’s the kind of thing one might expect a libertarian to oppose.

No one is going to be out there forcing developers to fill up the land with tall buildings, and no one is going to be out there forcing residents and businesses to take space in what buildings are constructed. If current densities are, actually, the ideal, then developers will see that they’re likely to get a bad return on their investments and they’ll take a pass. But if that’s not the case, and developers are anxious to take advantage of the economic potential of the area, then zoning rules are going to be imposing a real, and perhaps large, economic cost on developers, on potential tenants, and on Fairfax as a whole.

Random points: Crystal City tried residential density and it didn’t work nearly as well as Arlington.  It’s a dead zone.  The earlier attempted dense development of Skyline Drive also stalled and was beaten out by Tysons.  Or look at the new (and failing?) complexes on Rt.29 and Gallows Rd. and Strawberry.  In 1989 I moved into a tall apartment building, right at Tysons, which had stores on the ground floor so residents would not have to drive to shop.  I was delighted but within six months all those shops had closed for lack of interest.  At the risk of sounding like Gustav Schmoller, each case really is different, just as Tysons is different from Arlington.

Crystal City is in Arlington. It would be difficult to find urbanists who would defend Crystal City as originally built; it is of a piece with much of 1960s sprawl and context-free urban redevelopment, much of which is loathed by urban planners and city residents alike. Density alone won’t solve anything. All the same, Arlington south and east of 395 isn’t what I’d call a failure. It is home to thousands of workers and jobs, it is redeveloping rapidly, and it accommodates a lot of mobility without the kind of traffic you see in Tysons. And there, too, the infrastructure is being improved to better accommodate density.

Density can’t be context free. A tall building with shops on the ground floor, completely detached from any kind of neighborhood context, isn’t likely to support a thriving retail environment. Does Tyler think that the shops at the base of tall buildings in New York or Washington serve the occupants of those buildings and no one else? There is a network effect. One building with shops can’t provide that many options, and so any resident of the building requiring more than the most basic things will go elsewhere (by car, in Tysons). Two buildings with shops can offer more variety, making the shopping area more attractive and drawing a larger share of local residents. Four offer more variety still, offering more of a draw for local residents. Ten buildings together can offer a very diverse set of retail opportunities, all within easy walking distance. At that point it is much more attractive to walk around in the local area for a meal or an essential from the grocery or convenience store, or a drink, or a movie than it is to drive off to some other shopping area. The local shops therefore attract most of the business of local residents, and they thrive.

Matt Yglesias wrote:

…why on earth isn’t the libertarian take on this that we should permit high density construction and let the market decide what happens?

When it comes to the current Tysons plan, it is not “the market deciding.”  It is a mega-plan with road widening, the bane of progressive pro-environment, pro-urban advocates, and also massive subsidies for growth and not just density of growth.  More generally, when road policy is so politicized, it is never the market deciding in any case.

Call me odd, but I’m not opposed to urban (or suburban) planning and in fact anyone who recognizes the ongoing existence of public roads has to end up in the same place.  I might add that postwar Germany did a good job of such planning.  Tysons Corner is not, right now, doing a good job of it.  You can believe that whether or not you’re a libertarian.

Look, the road improvements added to the latest plan, along with reduced densities, are among the main reasons urbanists are upset by the proposed changes. The fact that this is a “mega-plan” involving significant infrastructure changes does not mean that planners are directing the development of the real estate. They aren’t, and Tyler should recognize that it’s as troubling to have planners setting artificial caps on potential density as it would be for them to be forcing construction up to a maximum height.

Beyond that, it’s nice to see that Tyler appreciates the essentially non-market nature of much of the planning process. I hope he’ll take this religion that he’s found to his libertarian counterparts, who seem to think that suburbia (structured, as it is, around public roads) emerged magically from the workings of the market. This is what Matt and I are trying to get Tyler and others to notice. In real life, all planning and land-use choices involve complicated and value-laden decisions to be made by government officials, while in libertarian land, only the land-use choices producing dense, walkable streetscapes involve complicated and value-laden decisions to be made by government officials, while the stuff that leads to more pavement is 100% invisible hand.

It is never the market deciding how to create the infrastructure around which to build. I accept that. And so I accept that just because the last 60 million housing units were mostly built around the massive network of highways built by the government, there is no reason the next 60 million housing units have to be mostly built in the same manner. Price premia on housing units in walkable areas suggest that would be a big mistake, actually. But given that there’s an inevitable complementarity between government-produced infrastructure and the development that follows on, there’s no getting around the fact that to satisfy household demand you have to create more opportunities for denser growth. Unfortunately, when governments try to do this, they’re often met with a chorus of complaints from libertarians who reference settlement patterns over the past half-century as evidence that anything different will fail, as if that argument weren’t entirely circular.

I actually think Tyler’s final suggestion, that denser redevelopment take place in areas that aren’t currently so built up, is a fine one. But there are two problems with it. One is that zoning rules prevent it. And the other is that while it’s extremely easy to build speculative highway infrastructure in relatively underdeveloped areas, it is extremely difficult to build speculative transit infrastructure. Authorities will say, based on current densities, predicted ridership is too low, as if people don’t respond to changes in relative costs. This line of thinking is relentlessly hammered home by libertarian transportation writers. Ed Glaeser essentially based his entire series opposing HSR construction on it. It doesn’t make any sense.

Maybe the Tysons redevelopment won’t actually work, but I don’t think Tyler has offered a convincing argument for why it won’t. And if it does fail, I’d wager it will have much more to do with the density reducing measures being considered by the zoning board, that Tyler appears to like, than the fixable problems of the road network.