There’s been a lot of discussion around the House amendment that puts Seattle taxpayers on the hook for viaduct-replacement cost overruns. The Senate did carry the amendment through to the final bill, though it may not be possible for the state to enforce that provision. Some even wondered what it would mean if other municipalities had to pay for their overruns, as of course they don’t. The whole thing had me wondering, how is it that “most projects” have overruns?
It seems like if you know that most big transportation projects have overruns, you would make a plan for the likelihood of overruns and factor that into the cost. We do exactly this in the software world, and Office, the product I work on, takes as long to complete (three years) and has far more people working on it than a tunnel under downtown, so it seems madness to me that “most projects” could come in over budget if planning for likely overruns was built in to the cost early. Well here comes Danny Westneat:
[A] professor at Oxford University in England has done a compelling series of studies trying to get at why big public-works projects such as bridges, tunnels and light-rail systems almost always turn out to be far more costly than estimated. “It cannot be explained by error,” sums up one of his papers, matter-of-factly. “It is best explained by strategic misrepresentation — that is, lying.” … It started seven years ago, when he published the first large study of cost overruns in 258 mega-transportation projects. He found that nine out of 10 came in over budget, and that the average cost overrun was nearly 30 percent. Rail systems had an average cost escalation of 45 percent.
Emphasis added. It’s worth reading the whole thing.
At least in the case of the this SR-99 tunnel, we have a decent idea of what exactly needs to be done. Bored tunnels have become very common, and the technology for building them is well developed. The price tag of the overall project went down not because of “value engineering” but rather because the streetcars and bus improvements were dropped off, so its not exactly like the politicians changed the estimate to fit the budget, but rather they changed the project. So I don’t know how likely overruns are, but this might be worth thinking back to if they arise.