Improving project predictions by managing risk
When asked for a project cost estimate, we typically provide a single dollar figure—but Khalid Bekka labeled that a mistake. In his presentation at the 2009 Minnesota Pavement Research Conference on February 12, Dr. Bekka, an economist with HDR, Inc., said, “If we do it that way, it means there’s a 50% likelihood the eventual cost will be above that number and also a 50% likelihood the cost will be below that number. How many of us would bet our mortgage on a 50-50 chance? We wouldn’t! So why are we willing to do that with an infrastructure project?”
Furthermore, Bekka predicted that, with the funds on the way from the American Recovery and Reinvestment Act of 2009 (the “stimulus package”), we can expect our “due diligence” processes to be carefully scrutinized. “We will be judged on the types of processes we use, the transparency of those processes, and what kinds of risk we assume.” (Ed. note: This prediction became reality a few days later when President Obama appointed a former Secret Service agent with experience in rooting out corruption as chairman of the Recovery Act Transparency and Accountability Board.)
In response to these issues, Bekka explained a process that he has used to manage risk in more than 100 infrastructure projects. The first step, he said, is to state each key variable as a range rather than a single number: “I like what Warren Buffett said: ‘I would rather be approximately right than precisely wrong.’ ”
Bekka emphasized that this should be done for the broadest possible range of factors that might affect a project. As an example, he discussed a 2003 project at Ground Zero in Lower Manhattan. “The MTA showed us their time estimate for digging a tunnel. But we concluded that they should move the completion date out at least two years. One risk that led us to that conclusion was the possibility of archaeological findings. It turned out that, on average, archaeological findings had delayed them an additional year in previous projects.” Figure 2 shows how Bekka and his associates plotted this variable: To achieve a 90% probability of completing the project by its estimated completion date, the MTA needed to push its prediction out 28 months. To be sure all important factors are being considered, Bekka recommended proactively inviting a broad range of opinions to the table as early as possible in planning each project. “Of course you will bring in experienced engineers,” he said, “but you should also invite academics, planners, and the public—even skeptics—and let them all throw darts at your project.”
Bekka’s next step is to assess the risk attached to each factor involved in a project. Figure 3 is an example of how he compares factors in terms of their probable impact on project cost.
Finally, Bekka explained that, once you see where the uncertainty is in your project, you can take steps to mitigate those uncertainties. “Do I need more surveying? More public relations? How can I reduce the uncertainty? The more you mitigate, the more you have due diligence, confidence, credibility, and transparency.”
As a final example, Bekka discussed the ROC-52 project, completed in 2006, in which U.S. Highway 52 in Rochester, Minnesota, was widened. “The major issue was whether to do it in the design-build mode or the traditional way,” Bekka said. “We heard from lots of people that their businesses would be disrupted during the project. Initially, construction plans were projected for 11, 7, or 3 years. With input from local citizens, Mn/ DOT, and some university professors who we hoped would not be tied to a political viewpoint, we did a side-byside assessment of design-build and design-bid-build. What emerged is that, by doing the project with designbuild in three years, there would be a cost increase. But there would also be a tremendous amount of public benefit in terms of reducing the impacts on business, property values, and congestion. It was the difference between a short pain and a long pain. I’m not saying you should always use design-build, but in that specific project, the rate of return completely outweighed the premium in cost.”
For an in-depth look at risk assessment in infrastructure projects, see Highlights of an Expert Panel: The Benefits and Cost of Highway and Transit Investments (GAO-05-423SP) available (accessed 2/22/09). This U.S. Government Accountability Office document summarizes ideas on how to conceptualize, measure, improve, and use information about benefit-cost analysis of highway and transit investments.
— Richard Kronick, LTAP freelancer