The Costly Risks of Mega-projects like the Columbia River Crossing

Evan Manvel

90% of mega-projects go over budget... bridge projects by an average of 34%. On the CRC base of $3.6 billion, that’s $1.224 billion.

The Costly Risks of Mega-projects like the Columbia River Crossing

The Columbia River Crossing highway mega-project is the most expensive public works project in the region’s history. But just how much will it cost to construct this highway builder’s wet dream?

Project backers offer some numbers: $3.2 to $3.6 billion – or 3,600 million dollars – plus billions more in interest costs, meaning more than $6,000 total for every household in the region. Here’s the rub: as jaw-dropping as those numbers should be, they’re almost certainly low – by hundreds of millions of dollars, if not billions.

The CRC is our first attempt at building a mega-project this scale. Such projects have been built around the world for decades, allowing academics to study them. The key findings? 90% of mega-projects go over budget. Projections haven’t improved over the past 70 years, or across 20 nations. Bridge projects go over budget by an average of 34%. On just the CRC base $3.6 billion, that’s $1.224 billion.

Oxford University’s Bent Flyvbjerg is a worldwide expert in mega-projects, who distilled the numbers above. He summarizes how such projects get approved:

there seems to be a formula at work: underestimated costs + overestimated revenues + undervalued environmental impacts + overvalued economic development effects = project approval

Flyvbjerg notes major projects usually having the following characteristics (edited slightly for clarity, emphasis mine):

For those who have watched the CRC plans develop, this sounds eerily familiar. A lock-in to a project concept, little analysis of alternatives, conflicting interests, lots of risks, pushing forward a non-standard design, bad information throughout… it’s as if Flyvbjerg is Carnac the Magnificent, opening the envelope to read, “How would you describe the CRC?”

Here’s what we should learn from the costly errors of other mega-projects: irrational exuberance is common, even standard. Flyvbjerg considers why forecasts are off so consistently, including psychological and technical errors, and finds less data to support those arguments than another: “Political-economic explanations and strategic misrepresentation account well for the systematic underestimation of costs and overestimation of benefits found in the data.”

He argues, “project manager and planners… are busy not with getting forecasts and business cases right … but with getting projects funded and built. And accurate forecasts are often not an effective means for achieving this objective. Indeed, accurate forecasts may be counterproductive.”

Remember the Tram debacle? It was projected to cost between $15.5 and $28.5 million, and ended up costing $57 million. Several people knew the tram would be much more expensive than the original cost estimate, but withheld that information from city leaders. And we all should remember the Big Dig – a $2.8 to $6 billion Boston highway project that exploded into a $22 billion project.

Economist Joe Cortright recently looked at data on recent big Oregon highway projects. What he found was sobering. The largest current ODOT projects will cost more than twice what ODOT projected they would cost. The seven-mile long rebuild of US 20 between Corvallis and Newport was estimated to cost $110 million. We’ve spent $234 million to date and the project is still incomplete. ODOT estimated the Newberg-Dundee bypass would cost $222 million. It's now projected to cost between $752 and $880 million. And the cost of the relatively small rebuild of the Grand Ave-MLK-Highway 99 viaduct has ballooned from $36 to over $71 million.

ODOT’s crystal ball is broken, with very costly consequences. Faulty projections mean other roads go unmaintained, potholes go unfixed, sidewalks go unbuilt, and projects around the state lose out. In this case, it could mean several parts of the CRC mega-project go unbuilt, decreasing its limited functionality, or that we grab key pots of state money to backfill the gaps. The desire to think the CRC cost projection is uniquely accurate is the same hubris mega-project backers have made throughout history.

From Flyvbjerg: “A key recommendation for decision-makers, investors and voters… is they should not trust the budgets, patronage forecasts, and cost-benefit analyses produced by promoters of major infrastructure projects.” We’ve already seen major errors in the project’s traffic projections, errors that pre-date the recession (more on that in a future column).

Perhaps the experience we should learn most from is the Oregon Wireless Interoperability Network (OWIN). There, project managers claimed to “need” $400 to $600 million to meet federal guidelines and build a functional network. Caught short of money, apparently a $49 million upgrade will actually suffice. From The Oregonian: “Legislators say they only now understand previous OWIN managers sold a far bigger project than was necessary.”

Hopefully, that’s a lesson we can take to heart. We can push aside the extremely costly and risky plans from the highway departments, take a fresh look at the problems at hand, and focus our resources more efficiently. We can move forward with an approach that meets the needs of the community, the impacted neighborhoods, and the taxpayer. We did it with the Mt. Hood Freeway and the Prescott Freeway, and we’re doing it now with OWIN.

Being a good steward of our resources, be they financial, human, or environmental, is simply the Oregon way. The current CRC plan is not.

Note: Cortright's work led me to read Flvybjerg and inspired this piece. Cortright's full report - which projects the likely actual cost of the CRC at $10 billion, is a valuable read.

As a side note, the Oregon House Transportation and Economic Development Committee is considering HJM 22 on Monday. The bill is a press release combined with a plea for the Federal government to please send money for the CRC.

  • (Show?)

    For those who want to read more, here are the citations:

    Flyvbjerg, B. (2005). Machiavellian Megaprojects. Antipode, 37(1), 18-22.

    Flyvbjerg, B. (2009). Survival of the unfittest: why the worst infrastructure gets built–and what we can do about it. Oxford Review of Economic Policy, 25(3), 344-67.

    Impresa, Inc. (2010). Financial Analysis of the Columbia River Crossing. Portland, OR: Cortright.

  • (Show?)

    Oh, and full disclosure: I've done some paid work for the Coalition for a Livable Future on the CRC. This is not part of that work.

  • (Show?)

    It's a sad day (years) for Oregon that we will be laying off teachers all across Oregon as we begin (possibly) to fund that boondoggle of a bridge project. What's wrong with us?

    • (Show?)

      I'll buy you a beer if they actually start construction anytime in the next five years.

      I continue to believe that so much of this CRC nonsense is fantasy budgeting.

  • (Show?)

    I support the 12 lane-capable CRC crossing.

  • (Show?)

    Just to be clear, most of the cost of this project isn't the bridge itself, it's the highway interchanges.

  • (Show?)

    Whether the bridge construction actually ever gets underway or not, the amount of money being spent to ever "get there" to begin with-will continue to be spent. The CRC will make it's money whether the bridge ever gets built or not; it already has.

  • (Show?)

    Good post Evan. There's a meta-dilemma in there somewhere, that "Great Things" usually aren't great for everyone. You don't get pyramids in Egypt or Millenium Park in Chicago through open and transparent government. So is it better to have a truly representative government that never does anything Great, or accept some level of corruption in return for a chance at the iconic?

  • (Show?)

    Couple things.

    Don't believe the $49 million OWIN figure. The reporter seriously misunderstood what that was supposed to buy. We'll need to spend a lot more than that.

    I wish more people understood that the CRC is a bad idea because it should be a much lower priority than retrofitting our current bridges to withstand the Juan de Fuca quake. That's the 9.0 plus or minus that's probably coming in the next 100 years (37% next 50). If we don't fix the bridges, at a current cost of $4 billion, both I-5 and 101 will be impassible for an extended period, making it almost impossible for disaster aid to get to millions of people. It would be worse than Japan (except for the nuclear thing).

    Also, as Peter Courtney said on Think out Loud last week, we have 800+ schools that will fall down on top of our kids if the quake hits during school hours. We need to shore them up too.

    Spending money building a huge new facility in the face of this impending disaster would be irresponsible.

    Final point: the traffic projections on which the toll revenue projections are based could be totally off - and are, according to Joe Cortright's analysis. With gas prices rising, people won't be using cars as much as they used to. So why do we need a bigger bridge?

    • (Show?)

      Steve --

      I don't disagree with you about the need to retrofit our bridges.

      I do, however, want to question you about this claim: "I-5 and 101 will be impassible for an extended period, making it almost impossible for disaster aid to get to millions of people."

      There are very few communities in our state that are on "islands", reachable only through I-5 and 101 (or other highways). A few on the Coast, perhaps, but certainly not enough to add up to "millions".

      While it might be difficult, for example, to bring in disaster aid through surface streets, it's not impossible or close to it.

      And, secondarily, I trust that your notion of "worse than Japan" is mere hyperbole. Even excepting the nuclear thing, we're not going to have a tsunami in Portland. A 9.0 earthquake would be bad, but we wouldn't have the widespread clearing of the landscape akin to the Japanese tsunami.

connect with blueoregon