Seattle's SR520 toll reduces overall traffic volumes

Famous US political journalist and commentator, Michael Kinsley, writes in the LA Times about the SR520 bridge toll that I reported on last year here (noting the congestion pricing that has since been applied), with this interesting statistic:

Since the toll was imposed, traffic on the toll bridge has dropped about 40%, while traffic on the free bridge has risen 10%. Overall traffic to and from the east side has dropped about 6%. 

Of course, the main point of his article isn’t about the traffic or economic implications about it, he’s being philosophical about it as an example, saying it makes society less egalitarian because people can pay to avoid congestion. I’m not going to engage on that, except to say that I’m sure he’s flown in premium cabins on airlines happily bypassing queues at airport check-in and boarding, and people pay for TV, pay for valet parking, pay for the best seats in theatres etc, and he doesn’t seem to mind that. Indeed, the SR520 toll is to help pay for the replacement bridge.

The interesting result is that a 40% drop of traffic on one bridge (with two lanes each way) has resulted in only a 10% increase on the other bridge (with three lanes each way excluding the reversible two HOT lanes on the I-90 bridge), resulting in a 6% overall reduction in crossings.  The report I blogged about last year predicted a 50% reduction.

What this demonstrates is that the toll has made more motorists think about whether driving is the correct option (I don’t have figures for public transit use) or whether people consolidate trips, travelling less when faced with paying the cost of the infrastructure they use. In either case, it shows that tolling one route whilst leaving another free does not just result in a 1 for 1 diversion, it does mean that there is some trip suppression and some mode shift. 

As I said before, what's particularly interesting is that congestion pricing has been introduced on the SR520 bridge.  The effect of this means different patterns of behaviour change, with more mode shift and trip suppression likely at peaks than other times (and no toll overnight).

For cities considering tolls on one crossing but not another, it may be an  interesting case study because the automatic concern will be gridlocking the other bridge.  In this case, the other bridge also has a tolled alternative in the form of express lanes. The simple response to congestion concerns is that users of either crossing have uncongested options, and a free one.   Washington State has introduced tolls on SR520 for financial reasons, but has cleverly designed it to also maximise efficient utility of the crossing as well.  

Vote No on Initiative 1125 if you support better road pricing

I don't have a particular personal interest in the State of Washington, since I've never actually been there.  However, I do have an interest in good public policy on road pricing.  Initiative 1125, which comes up for a vote on 8 November, is not an example of this.

It is being promoted by activist Tim Eyman, who frankly has taken stands on a number of issues that I would support, since I tend to be in favour of free market oriented solutions and prefer to err on the side of less government intervention over more.   However, as I have written before, I believe on balance, that while i understand his motives, on this occasion, he is wrong.

In September I said Initiative 1125 had good intentions.  I can understand not wanting money collected from road users to be spent on anything other than the roads themselves, but to clumsily only link tolls to the projects they are specifically on limits what can be done to replace fuel taxes over time.  Ultimately, I believe tolls of one form or another, will be applied to all major roads and perhaps all roads, as fuel tax becomes unsustainable.  This initiative would block that.  Similarly, the idea that a road is "paid off" and no longer needs to be paid for, doesn't really bear close scrutiny.

The key problem being that roads always need new capital put into them to be operational.  It is like building a house and assuming you never need to paint it, fix the roof, fix gutters, replace pilings or the like over time.   The setting of tolls should take into account recover of capital for long and short life assets, bearing in mind renewals will eventually be needed for major assets such as bridges.  It is a failure to consider roads like other assets that has seen so many bridges fall into disrepair.  

Beyond that is to consider congestion pricing, and what to do with surpluses from tolls.  The advantages of congestion pricing in terms of improving productivity, reducing delays and emissions are potentially enormous.  Initiative 1125 would stop this because the revenue wouldn't be spent on the roads subject to congestion pricing.  Even if you support using congestion pricing to engage in wider tax reform (e.g. replacing other taxes for money used to spend on road), it wouldn't be allowed.

He could have stuck to the first provision of the initiative (Prohibit state government from diverting gas taxes and toll revenues in the motor vehicle fund or other funds to the general fund or other funds and used for non-transportation purposes”), and I would not have found it problematic, but the rest makes it unacceptable.

The Olympian reports:

This is a typical Eyman initiative that overreaches and attempts to tackle multiple issues with a single initiative. By cluttering I-1125 with multiple requirements, Eyman has torpedoed his own effort. Toll rates should be set as close to the users as possible, not by lawmakers trading political favors in Olympia.

The Yes campaign website says tolls should only be used for "capital costs", presuming capital costs cease to exist at some point.  It seems to suggest congestion pricing is "politically correct social engineering designed to raise the congestion misery index high enough to force you out of your car."  Nothing is more politically correct than running roads in the traditional Soviet style manner of taxing users on an equivalent basis.   However, while I have some sympathy for those who question spending large amounts of money raised from road users on politically driven, financially unsustainable and economically inefficient public transport projects (and road projects), this initiative not only doesn't deliver for supporters of more market oriented approaches to roads, but is positively contrary to it.  Some of what the yes campaign says I agree with, but the ends do not justify the means, and the means are contrary to the philosophy driving the ends.

So while the No campaign looks like primarily a campaign from labor unions, Democrats and environmental groups, it should also be supported by those who believe in more economically rational pricing of roads and indeed may even interfere with privately owned toll roads if they were to be built.

So although I have some reservations about opposing the initiative, I believe those who DO support better pricing of roads, who think fuel tax is a poor way of pricing for road use and think tolls, HOT lanes, congestion pricing and ultimately network wide pricing are worthy tools for the future, should reject Initiative 1125.  Boeing and Microsoft already are opposing it.

Initiative 1125 is being returned by postal ballot by 8 November, it is one of several other ballots on the day.

Tolls on Seattle SR 520 will "halve demand" says study

I wrote recently about the proposed tolls on the SR 520 floating bridge in Seattle.  The bridge is to be replaced, with tolls imposed on the current bridge to help fund the work.

SR520 in blue will be tolled. I90 beneath it will not be tolled

Now the Seattle Times reports on the study compiled to prepare construction bonds for the project which claims that traffic will drop from around 100,000 vehicles a day to around 52,000.  The reduction is due to some suppressed trips, mode shift to public transport and diversion to Interstate route 90.   So the tolling will effectively deal to congestion, with traffic levels not returning to pre-toll levels until 2032.  The tolls will pay back a US$1 billion bond (total cost is US$4.65 billion for the bridge) and generate more than enough revenue to support it.

Curiously, the report also states that if equivalent tolls are placed on the I-90, then toll revenues on SR 520 increase by 38% - because the diversionary route is no longer more attractive.  State Treasurer Jim McIntire is confident that toll revenue will support the proposed bonds.   It's worth also remembering that the proposed tolls in include congestion pricing at peak times, a trend that is appearing more regularly on new toll facilities in the US, and a very welcome one to aid sustainable use of the infrastructure and to maximise revenues.

Seattle's SR 520 floating bridge to have congestion pricing tolls

In December 2011, the existing "Governor Albert D. Rosellini Bridge—Evergreen Point", in Seattle, will have tolls reinstated. The original bridge was built in 1963, with the toll removed in 1979. It is the longest floating bridge in the world being over 2.2km long and it reaching the end of its useful life. It closes in high winds and is at risk of collapse in an earthquake. It currently carries around 115,000 vehicles a day.   Vehicles using the current bridge will be untolled until the new bridge opens (and tolls will be instituted on that bridge).

Seattle's floating bridge (blue) connects the city with Redmond

A new bridge is to be built with six lanes, two of which are for bus rapid transit and HOVs, maintaining the general traffic capacity of the current bridge. It is currently a major bottleneck and curiously features peak travel in the direction AWAY from Seattle during each peak, because it is the best route to Redmond - the home of Microsoft.  It will cost $4.65 billion.

Tolling is intended to be fully electronic free flow using the DSRC system branded "Good to Go" by the Washington State DOT (using the long standing 915MHz technology).  

Vehicles without a "Good to Go" tag will be expected to pay within 72 hours, or receive a bill through their address identified through automatic number plate recognition.   Customers have a choice between having the prepaid tag, registering as a number plate customer, or simply waiting to get a bill in the mail (the latter costs an additional $1.50 per trip).  Outlets for "Good to Go" include online (a tag is sent in the post), retail outlets and customer service centres.  All options, except setting up an account at a customer service centre (where cash is allowed) require credit or debit cards, or bank account details.  Accounts must be opened with a prepaid balance of $30 with top ups required when balances drop below $8.

However, there has been some trouble with the installation of the toll system on the existing bridge.  The Seattle Times reports that Texas based Electronic Transaction Consultants have faced some difficulties, having promoted early sign ups to "Good to Go" earlier this year, but being then it "ran into problems, both with sluggish customer service and with the technical challenges of counting tolled trips and merging those into a new statewide toll-accounting system". It has had $2 million deducted from its contract worth $23 million for the delays.   The company claims it is unusually complicated because it is integrating with "Good to Go" which is run by the state for two other toll facilities.

Besides the point of tolling an existing old bridge to pay for a new one, what else is interesting is that the tolls wont be a standard flat rate.   The rates will vary as follows (for cars with tag accounts):

 WEEKDAYS

midnight-5am                   Free

5am-6am                           $1.60

6am-7am                           $2.80

7am-9am                           $3.50

9am-10am                         $2.80

10am-2pm                         $2.25

2pm-3pm                           $2.80

3pm-6pm                           $3.50

6pm-7pm                           $2.80

7pm-9pm                           $2.25

9pm-11pm                         $1.60

11pm-midnight                  Free

WEEKENDS

midnight-5am                     Free

5am-8am                             $1.10

8am-11am                           $1.65

11am-6pm                           $2.20

6pm-9pm                             $1.65

9pm-11pm                           $1.10

11pm-midnight                    Free

As you can see, it shows some interesting demand patterns.  I wonder if motorists will understand that degree of complexity at first, but the pricing itself is quite elegant.  Timed to manage any bunching of demand either side of the peaks and during the day.  Weekends are more interesting, with the higher rate matching steady day time demand. 

What this means, of course, is that Seattle is getting congestion charging, albeit on the tolls on this one crossing.  It looks sophisticated from an economic point of view, and the bus rapid transit system will offer an alternative for some.   I will look forward to see how the tolling affects congestion on the existing bridge as a prelude to what may happen when the new one is built.

Washington state toll referendum threatens funding for SR-520 bridge replacement

Seattle PI reports that Initiative 1125, a proposed referendum on severely restricting the use of tolls in the state of Washington, would interfere with the funding for a number of projects according to the State Office of Financial Management.

Some of the effects would be:

- To block use of toll revenue from Interstate 90 to help pay for rebuilding the neighbouring SR-520 bridge;

- To have toll rates set by legislation rather than by the State Transportation Commission;

- To eliminate any potential options for congestion pricing;

- To puts at risk $123 million in Federal funding, secured for the S.R. 520 project, because congestion pricing could not be considered;

- To risk blocking the Alaskan Way Viaduct replacement tunnel because it too includes congestion pricing.

I reported earlier on Initiative 1125 and concluded that, on balance, it would be negative:

It is tempting to stop politicians using tolls as an easy tax to divert from motorists, but it also doesn’t make sense to be ridiculously prescriptive in ways that don’t reflect good economic principles. I can understand imposing a restriction on a hypothecated fund to spend money on roads, and for tolling to be related to the costs of the roads tolled, and be spent on the network of the entity tolling. However, to restrict time and purpose of toll revenue spending is unduly restrictive...it would be wrong for this initiative to pass, because it is better to sacrifice improvements in one area to avoid a significant deterioration in the law on tolling in another.

This simply strengthens my view on this. There are sound reasons for putting some constraints around the use of tolls for non-road purposes, this initiative does far more than that and risks damaging the value of tolling for the State of Washington.

Congestion reduction fee in Seattle is not what it sounds like

When I read that King County Council had agreed to impose a “congestion reduction fee” on vehicles in Seattle, I thought it might have been some sort of congestion pricing. It isn’t.

It is a two year vehicle licencing tax of $20, collected to help fund subsidies for the county’s bus service. It is no different than regular annual vehicle licensing fees in other jurisdictions.

The report in the Ballard News Tribune states that a number of measures are to be introduced to contain the bus subsidies, including abolishing free travel downtown.

Congestion reduction though? No. It wont change driving habits, it will have an ever so tiny impact on car ownership, and wont reduce congestion at all. It is simply a straight out subsidy from vehicle owners to bus users, and employees of the bus operation. It is a simple tax, but bears no relationship whatsoever to congestion, or to reduce negative externalities from car use. Let’s not call it something that it is not.

Seattle's toll tunnel to replace untolled viaduct (UPDATED)

Route of the tunnel

Those unfamiliar with the story of the Alaskan Way Viaduct should read this. In summary, this viaduct is a six-lane part of Washington State Route 99 running along the Seattle waterfront. It was damaged in the 2001 Nisqually Earthquake and faced either needing to be repaired or replaced. Replacement was to cost up to US$3.5 billion, but as the viaduct runs along the waterfront of the city, it was seen as a chance to do something different (don’t ask why it wasn’t insured like any privately owned infrastructure). The options narrowed down to a bored tunnel or a surface street option.

Environmentalists argued, remarkably, that it would be better to demolish the viaduct and not replace it, so that traffic would be pushing through surface streets. The presumption is that the resulting severe congestion would simply deter many from driving and so have a positive impact – except of course for those who DO need to use the route (even buses and trucks), and anyone walking, cycling, working or living along the surface route. A clear case of local environmental gains being overridden by those wanting to wage a war against road traffic.  Indeed they even argued that the tolled tunnel would be no better on the streets that just demolishing the viaduct and routing traffic through town.

SR 99 tunnel, two lanes stacked

A postal referendum on the options came out in favour of the tunnel, which is an innovative stacked tube design as seen here. 

So the state has signed a construction contract to break ground on a $2 billion, 1.7-mile tunnel from Sodo to South Lake Union in September, to replace the Viaduct by 2016.

The problem with the tunnel is that as it will be tolled (and be four lanes whilst the viaduct is six lanes), and the environmental-impact statement claims that some 40,000 to 45,000 cars a day as drivers dodge a cost of anywhere from $1 overnight to $5 in the southbound afternoon peak. Yes, the tolled tunnel would have peak congestion pricing to manage demand. The state predicts that by 2030, about 57,000 cars would pass through the tunnel each day under the current proposed tolling plan, compared with 93,000 if the tunnel wasn't tolled at all. The city may gain a waterfront, but will also gain more downtown traffic. The question may yet be raised as to whether downtown congestion pricing might be a useful strategy to limit this, indeed the environmentalists might be better focused on that, now that an effective bypass is to be built.  

Traffic on the highway has been largely static in the last decade, hovering around 110,000 daily trips near Seneca Street, or 63,000 through the Battery Street Tunnel.

According to the Seattle Times, Highway 99 project Administrator Ron Paananen claims that the tunnel project, with a widened Alaskan Way south of the ferry terminal “preserves similar overall capacity to the current viaduct”. The FHWA has also signed off on the option.

Seattle isn’t unfamiliar with tolls, with the following existing projects:

- Tacoma Narrows Bridge;

- SR 520 bridge replacement (tolls starting in December); and

- SR 167 HOT lanes.

Big inner city toll tunnels are a modern solution to getting major traffic to bypass built up areas. They are not cheap, but they are permanent and make an enormous difference to traffic flows and the local environment. I wish Seattle the very best with this project and hope the traffic management and design for the surface routes is good enough to balance the needs for traffic and the desire to unlock the value of the land along the waterfront that will be freed up by removal of the viaduct.

Tolling will be electronic free flow and the pricing will need to be attractive enough to appeal to a majority of the existing users of the viaduct.

Full details on the project, including a cost breakdown are here.

UPDATE:  Useful details provided in the comments below, a key point being that much of the current traffic on the viaduct is entering and exiting the downtown.  The tunnel will be simply a bypass, so in fact it will be providing tolled capacity only for those bypassing, the existing streets will have to manage (with some widening work), begging the obvious question as to whether congestion pricing might be helpful?

Washington State Initiative 1125 - good intentions marred by unrealistic complexity

According to The NewsPaper, there will be a vote in November in Washington State, USA, on Initiative 1125. The referendum will require a yes vote for a negative proposition that “no revenue in the motor vehicle fund or toll fund could be used for non-transportation purposes”. Now it looks on the face of it like a great initiative for motorists, particularly since it avoids diverting such revenues to top up other funds. However, when you get into the detail it contains some flaws.

The proposals are:

- “Prohibit state government from diverting gas taxes and toll revenues in the motor vehicle fund or other funds to the general fund or other funds and used for non-transportation purposes”. 

Now I can go along with that, given it is about specific funds and revenues from road users. The case for diverting such funds for non-transport purposes is weak in an environment of publicly owned conventional toll projects. Indeed, it might strengthen the case for tolls. 

- “Prohibit state government from transferring or using gastax-funded or toll-revenue-funded lanes on state highways for nonhighway purposes”. 

This is not exactly grammatical, but I infer from it that it means not allowing revenue raised from toll lanes or indeed gas tax on state highways to be spent for “non highway” purposes. If it means not spending on other roads, I wouldn’t be supportive of that because there are plenty of instances where it is sensible to cross subsidise parallel routes, feeder routes or bridges over major highways. This is clumsy and its objective isn’t clear.  I am unsure as to what it adds to the above provision, other than confusion.

- “Require tolls to be dedicated to the project they’re paying for, ending such tolls when the project is completed, and only allowing tolls to be used for purposes consistent with the 18th Amendment to the Washington Constitution. Tolls on a project must be spent on that project and may not be diverted and spent on other things (allowing tolls to be imposed on anyone and spent on anything stops them from being tolls and makes them into de facto taxes)”. 

This is where it really gets unstuck. The myth that a road is ever “paid for” and that after the initial capital expenditure has been recovered, no more capital ever need be put into the road. It’s nonsense. The surface, road furniture and the like regularly need replacing and renewal, the base layers will eventually need renewal, and bridges will as well. Only the land and the earthworks for the original right of way are “permanent”, the rest of the road needs regular injections of capital. So this doesn’t exactly reflect economic reality. If it said tolls should be based on recovering the long run costs of capital and operating costs of the road, I’d be more inclined to agree. However, to confine tolls in time does not make sense. The counter-factual is not that “tolls be imposed on anyone and spent on anything”, but that they be introduced on a network basis. 

How would a system like the German truck tolls manage under this? Revenue from the entire motorway network and some highways would need to be directed to the exact roads charged in proportion to the revenue raised? No other utility network is that precise. Rail, telecommunication and electricity network providers engage in pricing that reflects network costs and spend money across a network, not just revenue raised on individual components, although there may be some review of costs and revenues at such levels where it makes sense. I understand the purpose, but I don't think it is economically efficient or reflects economic reality, but rather a widely held misperception that treats roads unlike any other capital assets. 

It is tempting to stop politicians using tolls as an easy tax to divert from motorists, but it also doesn’t make sense to be ridiculously prescriptive in ways that don’t reflect good economic principles. I can understand imposing a restriction on a hypothecated fund to spend money on roads, and for tolling to be related to the costs of the roads tolled, and be spent on the network of the entity tolling. However, to restrict time and purpose of toll revenue spending is unduly restrictive. For example, toll revenue needs to be spent on running the toll system itself, or may be spent on a common toll system for multiple roads, which may be in breach of the third proposal. While I am no great fan of tolling be used to fund public transport, I am also not a fan of tolling being prescribed to specific conditions that would limit it scope inefficiently.

It’s a shame. It is good to have public engagement on these issues, but it would be wrong for this initiative to pass, because it is better to sacrifice improvements in one area to avoid a significant deterioration in the law on tolling in another.

Washington considers conversion of Seattle tidal lanes to HOT lanes

Interstate 5 within Seattle has just over 7 miles of tidal flow (reversible) lanes. The lanes operate southbound towards Seattle in the morning (5am-11am) and northbound in the afternoon (noon-11pm) (but are closed to reduce noise for residents near the Ship Canal during the night). The lanes operate from downtown Seattle to Northgate Mall. They are not tolled, but do have an HOV lane.

Map of I-5 express lanes in Seattle

The Seattle Times is reporting that the State of Washington is considering converting all of the express lanes to HOT lanes. This would mean that vehicles could only access the lanes if there was occupancy of 3+ or a toll is paid. It would require a law change and the report indicated that:

“Tolls might begin in five years, and vary from $1.55 midday to $4.30 in the morning peak or $5.50 in afternoon peak, in 2016 dollars.”

The intention is to raise revenue to support maintenance of the highway, which is apparently urgently needed, but the obvious impact will be to increase traffic on the I-5 parallel non-express lanes (which are only different in that they are part of the main highway and the interchanges along it).

The intention is that any toll would be collected using a DSRC type system, which can easily be implemented. What it shows is that the revenue squeeze for transport authorities is causing them to think how they may better use precious scarce road space, manage it effectively and have users pay for the capital tied up in such infrastructure. This HOT lane conversion is unlike all others which have involved conversions from HOV lanes.

Washington State will need to get this considered carefully, in part to ensure the costs of collection and enforcement can be optimised, but also the economic effects of the diversion onto the uncharged lanes (and the long run benefits of optimised maintenance and a long term programme to improve levels of service on the I-5). However, it will be good for other states to notice this innovative approach, and for the Federal Government to positively support it.