Archive | Sellwood Bridge

Time to reboot the Sellwood Bridge?

A re-evaluation of the whole Sellwood Bridge project, in light of the recent funding defeat in Clackamas County

This defeat this past week by Clackamas County voters of the proposed $5 license fee to help pay for the Sellwood Bridge replacement project has Multnomah County scrambling to decide what to do next. While decisions have not been made, I have a sneaking suspicion that what will happen is certain design elements will be removed, and the bridge will be built without Clackamas County’s contribution. (And despite calls from some quarters for the bridge to be tolled for non-Multnomah County motorists, I suspect that won’t happen). There are well-known reasons that this is likely to occur–too many people in positions of power have too much invested in the current approach.

But given that I’m a blogger, not an elected official, and am thus unconstrained by the reality of public project inertia, I’m going to view the enterprise de novo: If we assume that the Sellwood Bridge wasn’t there (or was no longer open to motor vehicles), and we had $200 million or so to spend on a new Willamette River crossing–what would we do?

Purpose and need

It’s a dirty secret of public project management that by controlling the “purpose and need statement”, a document which describes the overall scope of the project, one can often dictate the final outcome–any idea which lies outside the “purpose and need” statement is considered out of scope, and discarded. The P&N statement for the CRC, for example, effectively rules out many of the sensible design options proposed other than the big hairy multimodal bridge and freeway rebuild that has been selected.

The Sellwood Bridge project’s purpose and need statement says the following:

The purpose of the project, as approved by the project’s Policy Advisory Group, is to “rehabilitate or replace the Sellwood Bridge within its existing east-west corridor to provide a structurally safe bridge and connections that accommodate multi-modal mobility needs.”

The following four major issues define the need for the Sellwood Bridge project:

  • Inadequate structural integrity to safely accommodate various vehicle types (including transit vehicles, trucks, and emergency vehicles) and to withstand moderate seismic events
  • Substandard and unsafe roadway design
  • Substandard pedestrian and bicycle facilities across the river
  • Existing and future travel demands between origins and destinations served by the Sellwood Bridge exceed available capacity

This statement was the result of careful negotiations among stakeholders, so it likely contains numerous political compromises. However, it’s worth unpacking. Some of the stated clauses are obvious items that probably aren’t in dispute (safety, modern design, and sufficient capacity); but two important design criteria are effectively constrained by the purpose and need statement: the what (a multimodal bridge), and the where (“within its existing east-west corridor”, which has been interpreted to mean “must connect to SE Tacoma Street”). In addition, the fact that this project is being run by Multnomah County implicitly constrains the “who”–Multnomah County is effectively chartered to design, build, operate, and maintain the bridge–and as we saw this past week, that has significant implications.

Assuming that these three parameters are variable in our hypothetical reboot, it’s good then to turn to the more important underyling question:


Why are we considering building a bridge, and does the reason justify the cost(s) involved? Beyond the obvious why-does-the-chicken-cross-the-road answer, we need to understand what the bridge is for–which means, first and foremost, understanding who the bridge is for.

Is it primarily for residents of Portland neighborhoods near the bridgehead–Sellwood, Westmoreland, Sellwood, Burlingame, and Dunthorpe? Is it for Clackamas County commuters trying to avoid congestion on OR99E? Should be it be a major east-west thoroughfare capable of rivaling the downtown bridges and the Abernathy (I-205) bridge for capacity and throughput? And what modes should it serve, and to what extent? High-speed auto traffic? Local traffic only (including mixed-traffic busses)? Rapid transit? Bikes and pedestrians?

These questions need to be answered before pondering the next level of questions–and the subsequent design and management choices ought to support the stated purpose of the project. My main objection to the current project is that this goal is not met: The project is declared to be of regional importance and assumes significant amounts of regional traffic; this state of affairs is used to justify the unusual funding technique of taxing residents of a neighboring county. Yet the project is being run by Multnomah County, not by Metro or ODOT or any other agency with a broader scope, and the current design–a two-lane bridge which extends Tacoma Street–is incompatible with these regional aspirations. Of course, this is pretty much a description of the existing bridge (other than the fact that its paid for), but decisions made a century ago need not constrain decisions which will be affecting the region a century hence.

At any rate, once the project’s purpose is given due consideration, then its on to the next three questions:

Who runs the project, and who pays for it

As noted, Multnomah County owns and operates the current bridge, and is running the replacement project. The county owns and operates five other Willamette River crossings–the Sauvie Island, Broadway, Burnside, Morrison, and Hawthorne bridges, and has ample expertise in bridge construction and maintenance.

The other bridge operators in the Portland metro area are ODOT (who own and operate the Boone, OC/West Linn, Abernathy, Ross Island, Marquam, Fremont, and St. Johns) and the UP (Steel Bridge, LO rail bridge) and BNSF (N. Portland rail bridge, Wilsonville rail bridge) railroads.

Both county and state ownership pose issues for a multimodal connector of regional importance. Multnomah County has difficulty with self-funding of projects of this sort–as witnessed by its attempt to pass the hat to its southern neighbors; and may not adequately represent the interests of non-county users. ODOT has its own institutional problems–the agency is fundamentally a highway-building agency; and if a highway is not the intended solution, ODOT’s involvement may be counterproductive. There are many complaints about ODOT’s insistence on an expensive interchange on the current project, and about its involvement with the CRC.

Two other possibilities that have been suggested are Metro and TriMet. Metro, as the regional MPO, has a scope of jurisdiction appropriate for this sort of project, and has a more multimodal culture than does ODOT. On the other hand, ownership and operation of capital projects is something which the agency presently does not do. TriMet does have this sort of expertise in house, and will own and operate the new Caruthers crossing once it is built. TriMet has successfully completed other major civil engineering projects such as the Robertson Tunnel, and numerous smaller viaducts and bridges along the MAX line. However, operation of a bridge which isn’t primarily used for transit operations may be outside of its purview.

Where does it go?

The location of the bridge is another important consideration. The current project was pretty much constrained from the get-go to be a replacement for the existing bridge. In some ways, that makes sense–established traffic patterns will be maintained, Sellwood residents won’t be angered by losing “their” bridge, and NIMBY objections will be minimized. But the location of the present bridge is a major liability for regional mobility, due to the character of SE Tacoma Street–a two-through-lane neighborhood street with low speeds and numerous pedestrian traffic; an inappropriate corridor for high-volume regional traffic or rapid transit.

Unfortunately, the existing geography and land use of the river basin limits the opportunities for crossing. Wildlife refuges on Ross Island and at Oaks Bottom make a crossing to the north problematic. Immediately south of the Sellwood neighborhood one finds a golf course on the eastern shore. South of that, downtown Milwaukie is on the riverfront, but opposite Elk Rock/Dunthorpe and some of the most expensive and exclusive real estate in the metro area. A crossing between Lake Oswego and Oak Grove would be possible, but would be too far south to handle OR224 traffic.

Any crossing would be further hampered by the fact that the Tualatin Mountains hug the west bank of the river, limiting connectivity further west. (South of Lake Oswego/Oak Grove, the Oatfield Ridge limits east/west connectivity on BOTH sides). South of Taylors Ferry Road, the next major east-west thoroughfare which crosses OR43 is Country Club Road/A Avenue in Lake Oswego. And no matter where a bridge is built, there is the issue of OR43 passing through the Macadam neighborhood, a traffic bottleneck.

What should it look like

Finally, there is the question of what the bridge ought to look like. This is probably the easiest question to answer–it will probably look like the proposed current design, with 2 general purpose lanes, the ability so support mixed-traffic rail, and ample pedestrian/bike facilities. A highway-only bridge is politically out of the question. A “green bridge” might arise out of a future transit project, possibly in the LO/Milwaukie area, but a standalone bridge without support for auto traffic is probably a non-starter.


Given all of that–if the existing project constraints were mostly removed, and the region had $200 million to build a southern crossing wherever it liked–where would you put it, how would you fund it, who would run it, and what would it look like? To make this interesting, feel free to imagine other connecting infrastructure projects in the future, if you like.

And if you think that the $200 million ought to be spent elsewhere entirely, or pocketed–with no more vehicular crossing of the Willamette between Oregon City and downtown after the existing bridge reaches the end of its useful life, feel free to say that as well.

Public projects, political capital, and the sunk cost fallacy

How the sunk cost fallacy, particularly applied to political capital, can cause public officials to insist on the timely completion of dubious projects.
A few months ago, I went “meta” and did an article on the Gordian knot of compromises the political process forces on public projects (and on private ones, too; though we seldom get to see the dirty laundry that results from corporate cock-ups). This article, likewise, switches gears from the nitty-gritty of transport projects and politics to the higher-level topic of what forces drive public policy decisions. Today’s article is dedicated to the Sellwood Bridge replacement project and the reaction to the news that Clackamas County voters are unwilling to help fund the project.

Rather soon after the vote was announced, Multnomah County commissioner Deborah Kafoury had this to say:

Now that Clackamas County voters have spoken, we will roll up our sleeves to try to complete this important project without their help. Safety concerns dictate that we must fix the badly deteriorated bridge, the busiest two-lane bridge in Oregon. Replacing the bridge must remain our top transportation priority.

This prompted a few howls of outrage those opposed to the project or its funding scheme (including a regular PT contributor who is a well-known local transit critic), noting that the campaign in favor of Measure 3-372 suggested dire consequences, including possible closure of the current bridge (which is in poor structural condition, and already closed to large vehicles, including trucks and busses) if funding is not secured.

I’m not going to debate here whether the “yes” campaign crossed any ethical lines. Bluffing and FUD (fear, uncertainty, and doubt) are part and parcel of political campaigns, and the “no” side engaged in some rather speculative claims of their own–suggesting that approving a $5 vehicle license fee to help pay for this project would pave the way for many more such levies in the future (“its not just $5” was a popular campaign slogan used by opponents). But it was fairly obvious that Multnomah County was bluffing, somewhat–and that they are not about to permit the existing bridge to close without a replacement being built.

This article takes a look at why.

Sunk costs

There’s a well-known economic fallacy known as the sunk cost fallacy. In the sunk cost fallacy, already-expended (or committed) capital is cited as justification for continuing some course of action, even when that action may be no longer viable. Commonly expressed, it goes like “we’ve already spent $x on this, so we might as well finish it”. Rational analysis shows that the money spent is already long gone, so ought not factor into any decisions going forward–only future costs and benefits should be considered by decision-makers.

Why does the sunk cost fallacy frequently trip up decision-makers? It’s not because they can’t read a spreadsheet. In many cases, it’s because changing direction is tantamount to an admission of error. And in some cases, a decision-maker may have even higher personal stakes riding on a project than just ego or pride–in my professional career, I’ve been on several projects deemed vital to the health of the enterprise–with the clear implication that heads will roll if the project does not run to completion. Many of these projects ended badly, with heads rolling anyway.

Another way to approach the issue is to note that decision-makers are managing multiple different “resources”, only one of which is the money of the organization. Another is their personal political capital–in many cases, a manager (or an entire organization) will need to expend significant amounts of this in order to bring a project to fruition, with the payout only coming if the project completes. And political capital is typically spent all up front–if a change of direction occurs, the political capital is gone.

Yet despite this, a scene that plays out all too frequently in the boardrooms of the world is the executive asking the subordinate if he/she is willing to “bet their job” on a particular course of action. While such queries may be helpful in sniffing out BS (and in providing continual motivation once a course of action is approved), it effectively commits the subordinate to the project–and may cause her to defend the project even when common sense makes it obvious that the whole thing ought to be scrapped.

The public angle

Many aspects of the public sector make the sunk cost fallacy particularly troublesome. For one thing, most public works projects are financially back-loaded: most of the money isn’t spent until concrete starts to pour. Given that, prior to the start of construction, there ought to be ample opportunity to explore options and ensure that the project is really the right thing to do. The Columbia River Crossing is often criticized for the over $100 million spent on planning to date, but that figure is still a drop in the bucket compared to the over $3-$4 billion the project is expected to cost (a figure which many critics think is ridiculously low).

Yet the project sponsors are proceeding as if there is a gun to their heads, effectively declaring the current plans closed to any significant modifications. Project leaders, and their sponsors in Salem and Olympia, speak of deadlines and dates as though their life depended upon it–even though the project hasn’t secured all its funding yet, and is facing harsh criticism from many quarters.

One likely explanation is that many involved in the project–Governor Gregoire, to a lesser extent Governor Kitzhaber, and numerous officials in both DOTs, have bet a lot of political capital on the project, and thus have quite a bit at stake in seeing it run to completion. Many sponsors of the project all have their wants and desires, and have been promised these (ample highway lanes, green elements like light rail and ped/bike facilities)–and are expecting that they be delivered. A significant change in the project’s direction would cause many of these fine upstanding public officials to lose face–and in the case of the elected ones, would provide plenty of ammunition for political opponents. (Both Kitzhaber and Gregoire won election in close races).

In the case of public projects, time is also an important factor for many reasons. If Federal funding is involved, the NEPA process is a major headache; significant changes can add significantly delay. Securing funding requires lots of delicate negotiations, and grants often come with expiration dates–delays in the planning phase can jeopardize the receipt of funds needed for completion. There’s also the factor that politicians like groundbreaking and ribbon-cutting ceremonies, and often would prefer that these happen during their terms in office.

Finally, public-sector projects frequently attract political opposition of a nature not found in well-run private companies. (In a well-run company, disputes over the necessity and/or parameters of a project are swiftly ironed out, and bad-faith participants are removed. That said, there are many poorly-run enterprises where it is not unusual to find employees or entire departments trying to undermine the projects and initiatives undertaken by some other part of the company). But what is often considered a bug in the world of business, is a feature in democratic politics. And political actors often try to throw wrenches in the gears of projects they oppose (or which are associated with political adversaries), and this is considered all part of the game.

One way to do that is to propose major changes in a project mid-course–which is one reason that savvy decision-makers will regard any such proposals offered later in the project with suspicion. Picking on the CRC again–proposed design changes promoted by Portland mayor Sam Adams were pretty-much rejected out of hand by the project committee and by the governors–who all but accused the mayor of offering proposals in bad faith in an attempt to gum up the works on a project whose current direction the City of Portland is adamantly opposed to. And it wouldn’t surprise me to learn that this–along with the whole “iconic bridge” shtick, is exactly what Sam was up to.

Back to the Sellwood

Given that, it’s rather obvious that “we may have to close the bridge if this doesn’t pass” was a bluff. The Multnomah County commission has invested quite a bit of political capital (and planning dollars) into the current project, and thus isn’t going to fold up shop just because Clackamas County voters aren’t willing to contribute to the project. (Likewise, Milwaukie MAX is proceeding on schedule despite all the controversy around its funding, and despite the likely prospect of Clackamas County voters overruling the generosity of their commissioners on this project as well. At this point, no county ordinance has been passed to allocate the funds, so there’s nothing for opponents to refer to voters–but it’ll happen).

In some ways, this is unfortunate. The withdrawal of Clackamas County’s funding share provides an opportunity to reconsider the Sellwood Bridge project (there is much about it that I don’t care for). Asking one county’s voters to help pay for another county’s infrastructure is indeed unusual, especially when the design of the project isn’t particularly geared towards the users from the other county, so I’m not entirely unsympathetic to Clackamas County voters. More on this will come on my next article.
But it appears that instead of taking advantage of this opportunity, Multnomah County intends to press on. To be fair, given the decrepit state of the current bridge, time may be really of the essence–and that significant delays in the project might result in the existing bridge being closed without a replacement in the pipe.

But from here, it looks like there’s a lot more riding on the current bridge project than bikes, busses, and cars.

Should Portland Help Pay for Sellwood?

Mayor Adams is being accused of dragging his feet on the City’s contribution to the Sellwood Bridge replacement.

But let’s step back a bit. Is it the best use of $8M/year of the City’s share of increased gas tax revenues to fund a County facility (we’ll leave aside the question of whether having the County own the Willamette River bridges is rational or not – that’s a question for another post)?

Clearly the Sellwood connection is a vital part of Portland’s arterial network. But what are the other options? Given the large percentage of users of this bridge from outside the City and the County I think tolling might be a very good solution to better allocate costs to users. And that would free up $8M/year for other uses like sidewalks, bicycle facilities (or insert your favorite project here).