Cost Estimates in Capital Projects

Once again we have a large capital project going over budget.  The Johnson Street Bridge project is now some $16 million over budget to achieve what the City said it wanted to do - assuming the narrower channel and lower seismic standard are not counted as a cost.

Cost estimates for these sort of projects should be reasonably accurate and should be much more comprehensive.  Over and over again the estimates are too low, much too low.  It is a systemic thing that has been identified as a global problem by Bent Flyvbjerg.

In public sector capital projects the problem often is optimism bias.   Each step of the way the best case is assumed until there is approval for the project to start.  Once the project starts there is no need to contain costs because there is more or less no danger of the project being cancelled.   In the private sector projects do also start and go way over budget, but they are much more likely to be cancelled.   In the early 1990s the company I worked for in the UK had a lot of contractors working on a private sector project that had cost $200,000,000 to date and then had the plugged pull on it.  

Some of the few examples I know around here where public sector projects were cancelled was in Seattle in the 1970s when most of the freeway projects were cancelled when they were partially completed.  Some of the projects were completed in the 1990s, but off of SR 520 you can still partially completed sections 40 years later.

The problem with cost estimates is that they are written up by engineers, often ones interested in getting the work to build the project.  It is about getting that approval for the project.   Cost estimates rarely offer multiple assumptions of what might happen.  I have not seen one in which there is any accounting for what the worst case scenario is and any measure of the risk of this happening.   Risk management is something that should be part of cost estimates but is not.  

Having seen several thousand appraisals for properties and adjudicated well over a thousand property assessments, I have seen professional appraisals that have been more than 50% higher or lower than the market value for a property.  In estimating costs and values there is a large range possible and people can make the numbers say what they need to.

Can we estimate accurately for projects?  Not always, but for most capital projects it is possible.  Developers can construct buildings in hundreds of millions and make a good profit on it.   There are many P3s that have been completed for the estimated costs.  Companies are willing to do P3s because they can work out if the project will make a profit or not.

 Bridges in the $100,000,000 range are not uncommon around the world.  Every possible assumption of what it would take to build the new Johnson Street bridge has been done somewhere before.   The scope and scale of the possible risks are known but have not yet been accounted for with the JSB project.

The way the project is going here in Victoria, what incentive is there from anyone working on the project to ensure that costs are contained?  If you are a contractor and get paid for work done, doing less means you earn less. The Telus duct problem was a financial problem for the City, but it was added profits for the company doing the work on moving it.

So who is held responsible for making such a huge error on the cost estimate on the JSB?  No one.   Who is held responsible for cost over runs?  No one.   No one has their job or professional standing in jeopardy for making a mistake of the magnitude to date.  You will notice that no one is using the word error or mistake in relation to the increase in the estimate.  I ask you, who is not either a mistake or error?

The combination of optimistic assumptions in the cost estimates and no one really responsible for keeping the cost in control means the almost inevitable result is that the project will cost a lot more.

If in 2010 the City had said the bridge was likely to cost around $110,000,000, the odds of the City being able pass the referendum would have been unlikely.   Putting everything in the best terms all the way along assured the approval of the project.   The danger of an accurate estimate for a new bridge was that the cost of refurbishing the old bridge would have seemed a good option.  I have no faith at all in the estimates that were made for refurbishing the bridge either.  I suspect it could have cost a huge amount as well.   The only advantage of refurbishment is that you could do stop gap measures to keep the bridge going for a a couple of more decades for less.

At a current estimate of around $93 million, and the fear that the federal government funding could be lost, the option that looks the best for the City in a bad situation is to continue with the status quo.   To walk away from the project now would lose the City some millions of dollars and threaten the $37.5 million from the Feds.  Choosing to continue seems like the best option.  The new information of the higher costs has come out just high enough and right after the announcement of the new money from the Feds to make the cost rise look reasonable - people have already said there is no extra cost to the taxpayer from what had been assumed a few weeks ago..   Everything is placed in front of the public to show us that the status quo is the best option.   All the information is laid out so that the only rational and responsible decision the council can take to today is to continue with the status quo.

 But what happens if the final cost from this approach is $110,000.000?  Is it still the best option?  How does the City afford the extra $17 million?  Some people are talking about walking away and starting again.

If one were to walk away now, the City is out the current federal funding and the costs to date.  Some of those costs are not lost such as taking down the rail span and moving the Telus duct.  If one were to start over, spending $60 to $70 million would leave the City roughly in the same place where I think they will be at the end of the project if they continue as now.  I think that is reasonable to assume some federal money would be available for the project in some form.

If one were to completely restart the project would it be possible to find companies willing to design, build and deliver the bridge for a fixed price of under $60,000,000?  This would be a risky move and would leave the City in general and council in specific in a very politically exposed place.  It would take a huge gamble by council to go in this direction.  The current council is not this politically courageous.

The reality of how this project was approved and how it is moving forward means the City is effectively committed to the status quo path and City taxpayers are committed to a whatever it takes to complete the bridge.   My one hope is that the Council develop contingencies for how to fund the bridge should it cost $100 million or $110 million.