We can do this the sprawl way, or we can do this the TOD way

Recently I've been examining the decision-making process for "bus rapid transit" in the Tappan Zee corridor, in response to Stephen's insightful question, "Can someone explain to me how the hell an I-287/Tappan Zee BRT could cost ("as little as"!!) $1 billion?" I originally listed four parts to the answer: First, all new busways are vulnerable to the dreaded "empty lanes" attack, which (bonus post) can be part of a broader strategy. Second, you're trying to serve a sprawling population with transit, which is always expensive. Third, the transit proposals all include some part of the "full corridor" from Suffern to Port Chester, while the current road proposal only covers the bridge. Fourth, transit advocates didn't have - and never have had, in this country - the political will to formally propose true dirt-cheap BRT.

I've dealt with the first and third factors, so now on to the second. I want to lead with this quote from our State Transportation Commissioner, Joan "Smart Growth" McDonald, speaking to Kate Hinds of Transportation Nation:

Some of the transit issues, whether it’s BRT (bus rapid transit) or commuter rail, are very detailed issues that need to be resolved with localities, particularly in Rockland County. Where do you site bus rapid transit stations, where do you put parking, if you want to add another lane for bus rapid transit, that would entail property takings, and that will take two to three years to get there, and the costs are between two billion and four billion to build that.

Uh, wait. Parking? Adding another lane? This is smart growth?

The inherent nature of sprawl is that transportation is inefficient. It's even more inefficient for personal cars than for transit; we just notice the inefficiencies less with personal cars because that system outsources the responsibilities of driving, fueling and maintenance onto the vehicle owner - typically a single occupant and an amateur, with predictable results. But transit is inherently inefficient in sprawl.

There are two ways to handle this. The first is simply to go with the form of sprawl and cater to people who want to go from far-flung cul-de-sac tract house to far-flung office park to far-flung strip mall, wasting gas, driver time, and bus wear and tear on snaky little routes guaranteed to make passengers simultaneously nauseous and bored stiff. The second, smart growth way, is to foster Transit Oriented Development. This works by starving the sprawly cul-de-sacs, office parks and strip malls and feeding the denser, pedestrian-oriented mixed-use apartments, townhouses and loft buildings. By doing this you encourage people to move their homes and businesses into the denser developments.

With busways in particular, there is a sprawl way and a TOD way. The sprawl way is to run the buses in a physically separated center lane on a highway out of walking distance of anything you might want to go to, with park-and-ride stations, and to keep the sprawl-inducing 1950s zoning. The transit-oriented development way is to run the buses in a corridor that accommodates commercial and residential development within comfortable walking distance of stations, and to throw out the zoning so that people can build more dense walkable buildings. The prototypical bus rapid transit systems - Curitiba, Bogotá, Guangzhou - all work the transit-oriented development way.

The transit-oriented development way is physically cheap - all you need is a bunch of jersey barriers and maybe a signal priority system for the traffic lights - but it is politically very expensive because it involves (a) taking road space from cars and (b) allowing density and mixed-use, two things that are guaranteed to get crowds of baby boomers all steamed up. By contrast, the sprawl way is politically cheap but expensive to build. This is one reason that the transit-oriented development way was pioneered in authoritarian societies like 1970s Brazil and 2000s China.

In the I-287 corridor between Suffern and Port Chester, there is a highway that is out of walking distance of anything you might want to get to, and there is a parallel series of older roads that run right through downtown Suffern, Spring Valley, Nanuet, Nyack, Tarrytown, Elmsford, White Plains and Port Chester. We've got a campaign that specifically invokes Curitiba (PDF) and a project team that holds Transit Oriented Development Workshops. And now we've got a Transportation Commissioner with a reputation for supporting smart growth. So which way did they choose?

Yes, that's right, they chose the sprawl way (PDF). God help us all.

The habits of successful busways

In response to Stephen's question about why "bus rapid transit" across the Tappan Zee Bridge was projected to cost "as little as one billion," I described the Empty Lanes attack and how sometimes it may even be part of a more nefarious plot. Tonight I'll explain how this led to only full-corridor BRT proposals, even while the road widening had a proposal just for the bridge.

The successful busways of the world fall into two categories. There are thick market busways, built in areas where there already is a thriving bus market, following the Magic Formula for Transit Ridership, and often earning a profit for the operators. Examples of this include Bogotá's Transmilenio, Guangzhou's Zhongshan Avenue line, and our own Lincoln Tunnel Exclusive Bus Lane and Bx12 Select Bus. Then there are former rail lines that are too narrow to fit more than a couple lanes of road, such as Pittsburgh's East and West Busways and Los Angeles's Orange Line.

Busways connecting sprawl with sprawl, without established bus markets and with heavy competition from government-subsidized roads, tend to have a hard time attracting passengers and fostering transit-oriented development. This 2002 post from Light Rail Progress cites Virginia's Shirley Busway and LA's Harbor Transitway as having lost riders, as well as former train lines like the Pittsburgh South Busway. At the very least they take a long time to build ridership, and are thus vulnerable to the Empty Lanes attack.

That makes it all the more mystifying why the Tri-State Transportation Campaign would push so hard for BRT across the Tappan Zee Bridge, and why they would happily throw the "prohibitively expensive" commuter rail option under the bus, so to speak. But it does explain why Tri-State pushed so hard for "full-corridor" BRT, from Suffern to Port Chester.

Work destinations of Rockland residents, according to the 2000 Census. From Tappan Zee Transit Mode Selection Report, Figure 1-4.

Nyack on the west shore of the Zee and Tarrytown on the east shore are both pretty small towns with high car ownership rates, and neither of them has enough residents or jobs to fill a busway between them when there are four parallel lanes of mixed traffic. Tarrytown has the Metro-North station where bus riders could transfer to Manhattan-bound trains, but even that isn't enough of an attractor. The buses could run in mixed traffic on both sides of the bridge, but that would limit their competitiveness with driving. Thus, transit advocates pushed for "full corridor BRT" to try to pick up as many Rockland residents and Westchester jobs as possible. And this, of course, added to the projected cost.

Given all this, the full corridor is pretty essential to the success of BRT. Without the full corridor you might as well not bother. In contrast, the highway widening plans put forth by the DOT were pretty clearly fantasy projects thrown in to make their bridge widening seem cheap by comparison. But because the bridge widening was only for the bridge, it also looked cheap by comparison to the BRT, and that's what allowed Commissioner MacDonald to reject BRT as not "financially feasible.”

Still, a billion dollars seems a lot for thirty miles of bus rapid transit, which we're constantly being told is "much cheaper than rail." So why aren't the proposals cheap?

The HOV Bait-and-Switch

In my discussion of the Empty Lanes attack on busways and HOV lanes, I neglected to mention that it can be part of a larger, more nefarious strategy, the HOV Bait-and-Switch. The sequence of events is the same; it is only the motivation that is different.

In both cases, a road agency builds HOV lanes, but they are not filled immediately. Solo drivers complain about wasted space, and eventually someone makes the decision to open the HOV lanes to all vehicles. This can happen with busways too.

The politicians and bureaucrats who build and maintain the roads usually express some support for the HOV lanes or busways, but that support can evaporate pretty quickly in the face of the Empty Lanes attack. And that's made me wonder if in some cases they hadn't planned it all along.

As far as I know there's never been a smoking gun to show that a highway bureaucrat suggested HOV lanes or a busway. But you'll notice that highway bureaucrats almost never suggest turning an existing lane into an HOV lane or a busway. They always want to build a new one, and somehow that new lane is always on a stretch of highway that they wanted to widen not long before, but ran into opposition on environmental or fiscal grounds. HOV lanes and busways appeal to environmentalists and fiscal conservatives, and the Empty Lanes attack is so predictable that you really have to wonder.

The transit planners are all too familiar with this pattern. It's no wonder that they never want to just open a busway without ensuring there's demand for it from day one.

The Tappan Zee Bridge and the Empty Lanes Attack

A tweet may not be worth a thousand words, but a well-crafted tweet can be worth a few hundred at least. On Monday, Stephen Smith asked, "Can someone explain to me how the hell an I-287/Tappan Zee BRT could cost ("as little as"!!) $1 billion?" With those few words, Stephen put his finger on a major weak point in the "enveloping BRT" strategy practiced by transit advocates. There is no "cheap" option for transit among the alternatives that were considered for the Tappan Zee corridor in the last round of discussions. Every transit option adds at least a billion dollars to the cost. When nobody can even find five billion for "just the bridge" (with a little stealth widening thrown in, naturally), it's too tempting for our austerity-obsessed "car guy" governor to cut the transit completely.

So why are the transit proposals so expensive, especially compared with the car proposals? The answer is a combination of four interconnected factors. First, all new busways are vulnerable to the dreaded "empty lanes" attack. Second, you're trying to serve a sprawling population with transit, which is always expensive. Third, the transit proposals all include some part of the "full corridor" from Suffern to Port Chester, while the current road proposal only covers the bridge. Fourth, transit advocates didn't have - and never have had, in this country - the political will to formally propose true dirt-cheap BRT.

The "empty lanes" attack has probably been attempted on every busway and HOV lane in history. In part it stems from basic physics: vehicles traveling slower can fit closer together safely, but faster ones have to leave more stopping distance between themselves and the vehicle in front. If the bridge managers succeed in speeding up bus traffic, the bus drivers need to leave more space. Drivers stuck in nearby traffic see that space and clamor for the lanes to be opened to them.

If demand for the bus or HOV lanes is at maximum (for example, the Lincoln Tunnel XBL with its steady stream of buses flowing down into the tunnel), drivers can't really say anything. But if it's anything less - say for instance that the project is just getting started - the clamor will start immediately and just keep building until the bureaucrats give in and open the lane to all vehicles.

Note that the bridge itself was pretty empty when it first opened - but there was nothing anybody could do about that. Tearing down the bridge wouldn't bring the money back. So they made the best of it. That's the double standard that all busways and HOV lanes face.

The transit agency could always run a bunch of buses whether anyone's riding or not, but that just opens them up to the related "empty buses" attack, another perennial favorite.

A strong political leader or movement could protect the busways and say "fuck you, it takes time to build ridership, we're not letting you drive in the lane." So a new busway has to have strong political sponsorship, or else it has to have a certain amount of ridership from day one.

It would add almost nothing to the cost of the current proposal - a rounding error - to build a new bridge with three general-purpose lanes and a busway in each direction. But if the busway is not immediately full of buses, car advocates will hit fast and hard with the "empty lanes" attack. I don't see any political leaders ready to stand behind a half-empty busway. The staff at Tri-State knows this well, and I'm pretty sure that's why they didn't object to the "BRT" actually being high occupancy/toll lanes on the bride itself. They don't want an actual busway on the bridge until they know it'll be full enough to stave off the empty lanes attack. So how do you build that ridership? That's where the other three factors come in.

The Tappan Zee Bridge: Money Pit, or Giving Tree?

I want to correct something that I wrote last week. I said, "even though the Thruway spends $30 million a year to maintain the bridge, and that number is rising every year, the tolls bring in $50 million." The tolls are still bringing in $50 million a year, but in 2009, according to Phil Ferguson, then the Finance Study Project Manager, "the bridge requires $130-150 million annually to operate and maintain, which is much more than it generates." Similar figures have been repeated in the media lately.

Given the number of times that we've heard about the high maintenance costs of the current bridge, a natural question is what the maintenance costs of the replacement bridge might be. Surprisingly - or perhaps not - this figure is a bit hard to come by. There's nothing in the scoping packet about it. The only figure I was able to find was on Page S3 of the Executive Summary of the Rehabilitation vs. Replacement analysis, where the "Present Value (150-year) Maintenance Cost" for a slightly larger bridge is estimated at $700 million. I'm not a finance expert and I haven't figured out how to get from that to estimated annual costs, but if we just divide the PV total by 150 years we get $4.67 million a year. Even if that doesn't include the cost of police, tolls, etc., it's still a lot lower than the $130 that's currently claimed, and the $20 million that the Authority paid for maintenance when the bridge was newer.

No wonder the Governor wants to replace it - and probably to spin it off as an independent authority, or even a separate for-profit corporation. The bridge is actually a huge drain on the Thruway budget, and part of the urgency comes from the desire to get out from under that burden. That said, it's really pretty rotten of Upstaters to take from the bridge when it was running a surplus and not want to chip in when it's running a deficit. It's like the Giving Tree or something.

Unreliable projections

Figure 1. NYMTC projected increases in residents and employment.

The new Scoping Packet (PDF) put out by the Federal Highway Administration for the reconstruction of the Tappan Zee Bridge includes this cute little map, showing "projected" increases in thousands of residents (on top) and jobs (on the bottom). The implication is that these thousands of residents will need to get to the thousands of jobs, and they'll all want to drive, and they'll need the bridge to do it. These numbers all come from the New York Metropolitan Transportation Council, essentially an arm of the New York State Department of Transportation.

Figure 2. NYMTC historical and projected bridge traffic volumes.

A similar argument was made in Figure 2 (from the Alternatives Analysis document that has since been scrubbed from the website), showing the average daily number of vehicles that have driven across the bridge for each year since the bridge was built in the 50s. It's a nice straight line, huh? So extending it, as the dashed line does, seems natural. This projected demand is an important part of the State DOT's argument for replacing the bridge. It's also an enticement for people to buy bonds to finance the bridge replacement: the higher the tolls, the more certain it is that the bondholders will be paid back. But why didn't the FHWA use that chart in the most recent Scoping Packet? Instead they used a chart that only went up to 2010 (Figure 3).

Why did they do that? Well, you might notice that the data for the past decade doesn't quite match projections. In fact, if we compare the actual volumes to the projections, as I do in Figure 4, we find that the projections haven't been very reliable.

Figure 3. NYMTC historical bridge traffic volumes.

The bridge traffic volume has actually been pretty stable over the past decade, in contrast to the dramatic increases that were predicted. There are two likely reasons for this. One is that the high gas prices in 2006-2007 discouraged people from driving, and the unemployment from the resulting recession took away the reason to drive from many west-of-Hudson residents.


However, the rate of increase in traffic volume was low even in the beginning of the decade. The other potential explanation is that the congestion on the bridge, and the high rate of crashes caused by the State squeezing an extra lane in, is a turn-off. Rather than deal with this, people will choose to get jobs elsewhere, or not even move to Rockland.

Figure 4. NYMTC historical (indicated in green from 2002-2010) and projected bridge traffic volumes, including recent totals.

If the volume of traffic is self-limiting (and come on, really, why wouldn't it be?) then there is no urgent need to widen the bridge, or even to replace it. If the bridge capacity is reduced, people will either stop moving to Rockland, Orange and Bergen counties, or they'll move to someplace where they can catch the train to Hoboken.

Even if the new bridge is built, most scenarios call for the tolls to be doubled or even tripled, and many call for them to be indexed to inflation. That would serve to discourage use of the bridge. Furthermore, there is a chance that higher energy prices would limit driving. This could mean that the tolls never bring in enough revenue to pay off the bonds.

The future of Rockland, Bergen and Orange

One of the unpleasant possibilities to contemplate if the Tappan Zee Bridge gets rebuilt at double the width is a massive growth in sprawl. Right now, driving to Westchester every morning is a pain in the ass and getting worse every day. When the new bridge opens with its wide lanes and shoulders, traffic should move quickly, even with four lanes in each direction. If traffic gets backed up, you know that Cuomo would use one or both of the shoulders for an extra lane, even though he pretended he wouldn't. Tolls may be double, but they'll pay for a nice easy commute - until drivers get to I-287 at least.

With this easy commute, people will start buying houses in Orange, Bergen and Rockland again, and businesses will start moving offices back to Westchester. The money spent by all the construction workers in the immediate area will boost the economy, putting more people to work. "Homebuilders" will go back to work knocking down trees on the sides of Orange County hills to build more subdivisions. Big box stores will build more locations along Route 59, with nice big parking lots. Maybe the State will get excited at all the toll money coming in and widen the Thruway like it proposed in the study.

You and I may recoil in horror at that scenario - carnage! pollution! wasted oil! wasted space! obesity! bowling alone! But that's the happy motoring vision put forth by the task force under Governor Pataki and Commissioner Boardman in 2006, and continued right through to the present day under Andrew "car guy" Cuomo and Joan "smart growth" McDonald. A lot of it is based on the projections from the New York Metropolitan Transportation Council, like this one for traffic volume on the bridge.



The problem is that these projections are moronic. Sooner or later this will stop. As Kate Slevin predicts, even at fourteen lanes, the bridge will get congested. But before that, a lot of people will not be able to afford to pay ten dollars a day to cross the bridge on top of everything else. Gas prices will rise to five, six or seven dollars a gallon, and that will knock out a lot of other commuters on the bridge. If people can't drive to work, they will take transit, and if the transit continues to suck they will get jobs in the city where they can take the train or the bus in. If they can't do that they will move to someplace with more convenient transit. Others will stop moving to that part of the region, because it won't be a good deal any more.

When that happens, in order to function at all Bergen, Orange and Rockland will need to transition to a lifestyle oriented around transit and walking, which means living in town. How easy will that be, with a brand new five billion dollar bridge that nobody needs anymore, and all the sprawl it created? How easy would it be instead if we spent the money on transit, such as restoring passenger service on the West Shore Line?