The End of Big Infrastructure Projects, Part 1

Last week, it was announced that phase 2 of the Dulles Rail project, which includes the actual airport, will now cost $3.8 billion, or about $350 million per route mile, excluding interest. This blew away previous estimates that had the project coming in at $2.5 billion, and is a huge increase from the $1 billion and change it was supposed to cost five years ago.

The DC area media, including those astute microeconomists at the Washington Post, are chiming in cost saving ideas, particularly relocating the airport station above ground. But these do little to address what's pushing the cost up - labor and health benefits.

In our 2010s economy, where face-to-face contact, not creativity, is the defining scarce resource, big capital projects are becoming more costly at a rate faster than inflation. And it's not an issue of corruption or politicians, but the simple fact that labor costs, including benefits, are rising faster than inflation. Moreover, cap labor productivity is barely advancing, while operating labor is practically being automated out of existence. This means any major infrastructure project - road, rail, or building, is being forced into a tough financing situation.

Maryland's financing $1.2 billion of its 18 mile, $2.6 billion inter-county connector with revenue bonds supported by 25 cent/mile tolls that are expected to rise in the future. Drivers on the Dulles Toll Road will face even more significant increases to pay off the revenue bonds funding the rail line going down its median. While transit advocates might like this arrangement, the same construction firms working on rail work on roads, and the forces increasing costs are the same - it's not like philosophy or politicians can do anything about the underlying economics.

The result of all this that we will soon reach a point where most infrastructure is just maintenance and "construction" mostly applies to existing roads and rails. This will also be seen in sports stadiums, which are subject to similar economic constraints and also likely to be built in far lower numbers in the future, regardless of who gets elected.