Government Infrastructure Is Inefficient Everywhere
An op-ed in the Wall Street Journal today indicates that Edwards’ Law of Cost Overruns is an international standard. If a politician says that a project will cost $100 million, it will end up costing $200 million or more.
The WSJ piece by Bent Flyvbjerg and Atif Ansar examines the results of an Oxford University study looking at 245 dam projects around the world. The projects had a “dismal track record” in terms of sticking to their promised budgets. The “actual construction costs of large dams are globally on average 96 percent higher than their budgets,” say Flyvbjerg and Ansar. That means a doubling, which is right in line with Edwards’ Law.
Rachel Maddow has drawn the lesson from Hoover Dam that big government projects are really great. But Flyvbjerg and Ansar describe a more typical government project: “Brazil's Itaipu Dam was built in the 1970s. It cost nearly $20 billion, 240 percent more in real terms than predicted and it impaired Brazil's public finances for three decades.”
I’ve written in detail about the history of U.S. government dam building, which has been chock-full of economic and environmental mismanagement. One reason for large cost overruns is that policymakers lie or conceal. I wrote that the Bureau of Reclamation “began constructing the Grand Coulee Dam with $63 million in funding from Congress, but it later became clear that the agency had a $270 million project in mind.” And I wrote regarding Jerry Brown’s father that “in pushing for approval of the huge State Water Project in California in 1959, Gov. Pat Brown kept throwing out a bogus cost estimate of $1.75 billion, even though he knew it would cost far more, as he later admitted.”
Liberals, such as Maddow, who hunger for big government infrastructure projects would cure their misguided lust by reading Cadillac Desert. Written by an environmentalist, I think it is one of the best public policy books of recent decades.
I’ve written about government cost overruns here, partly drawing from Flyvbjerg’s excellent research on the topic.
