2011 July 09 Saturday
Big Infrastructure Budgets Routinely Overrun
Virginia Postrel (who was much shorter in person than I expected - her mind is tall though) reports on a study that found large public works infrastructure projects very frequently go way over projected budget and deliver far fewer benefits than projected.
“Cost overruns in the order of 50 percent in real
terms are common for major infrastructure, and overruns
above 100 percent are not uncommon,” Bent Flyvbjerg, a
professor of major program management at the University of
Oxford’s Said Business School, writes in the Oxford Review
of Economic Policy. “Demand and benefit forecasts that are
wrong by 20-70 percent compared with actual development are
Flyvbjerg says the projects that most overestimate their benefits and most underestimate their costs get funded. Oh and public passenger rail and urban projects have especially big overruns. Why is that not surprising to me? Because passenger rail advocates tend to be zealots. They don't look critically at their great love. Read the full article for other good points of interest.
Given that politicians are going to continue to lade out the pork spending at our expense what's needed are independent agencies with engineers, accountants, MBAs, and the like who critically analyze infrastructure project proposals and present more accurate measures of costs and benefits.
For example, imagine a mechanism where existing government agencies or non-governmental organizations are chosen by lottery to analyze a big infrastructure proposal (e.g. a light rail project or high speed rail project) of some state or local agency. If the project goes forward anyhow and goes over budget as the independent agency predicts then the independent agency should get a financial reward from the agency that originated the project.
There are problems with my first version of this idea. Like, an independent agency would be tempted to project a cost overrun just out of hope it would get it right and make money. How to incentivize the independent agency to estimate the cost correctly? The closer the final cost is to the independent estimate the bigger the reward? One problem with this sort of a reward system: the better the reputation of the independent agency becomes the less likely it becomes projects will get built so their accuracy can be measured and rewarded.
Update: Also see Reihan Salam on how the Federal Railroad Administration drives up the cost of rail.
Update II: How about China? It is cited by high speed rail advocates as an example of enlightened leadership for its big build-out of high speed rail lines. But fares are so high that few can afford to ride and safety is an issue too. Also, lots of corruption has accompanied the construction of the rails. Passengers end up in buses and other cheaper ways to get around.
The essential problem is that there is no suitable negative reinforcement for those who underbid (or that overcharger later, causing cost overruns). In a very high-trust culture, the problem can largely be avoided since peoples' heuristics are such that they don't even consider underbidding or gaming the system. But in the USA, it's clear that the winning strategy is to get the winning bid by any means necessary and then extract as much revenue as possible.
To some extent though... isn't the solution to have the bid be the bid? I mean, what the fuck is this with 'cost overruns'? You agreed to build X for $Y: I paid $Y, now build it or else you can expect to see me in bankruptcy court after I sue you into oblivion. The issue presumably is that all the principal actors are in far too cozy a relationship and don't actually want to play hardball with each other on a human level.
"Because passenger rail advocates tend to be zealots."
Yup, some are. There's a huge nostalgia factor involved, and a lot of them also feel as though passenger rail gets treated like the red-headed stepchild. The truth is, taking people from one place to another is a business you cannot make money at without some form of government subsidy. Airlines, buses, your personal car, even cruise ships that are not a public necessity all use some sort of taxpayer provided infrastructure.
In an ideal world, nothing would be subsidized and the true cost of every form of transportation would be paid by the consumer. But after WW 2, the government decided that cheap mobility was important enough to the economy that the taxpayers should foot some of the bill. So that's the system we have.
On the other side of the coin, I think that conservatives (Michael Medved comes to mind) who flip out over Amtrak and its $1.5 billion a year subsidy look awfully silly. In the context of a $4 trillion budget, that's not much more than a rounding error, and there are much more serious problems confronting us. Amtrak should probably be de-unionized and sold to its employees, but even the left's favorite bogeyman, Ronald Reagan, left Amtrak alone. He understood he had to worry about the big picture.
bbarlog - "But in the USA, it's clear that the winning strategy is to get the winning bid by any means necessary and then extract as much revenue as possible."
But the study looked at many countries, not just the USA - at least according to Virginia. Does the actual study find that higher-trust countries (like, I presume, Scandinavian) don't suffer these problems?
Sgt Joe - "Airlines, buses, your personal car, even cruise ships that are not a public necessity all use some sort of taxpayer provided infrastructure."
Yes, but these non-rail infrastructures support many, many uses beyond transporting people - in particular commercial freight and recreation use. Amtrak of course shares many rail lines with freight operators - and still loses money. High-speed rail, on the other hand, requires an entire infrastructure unto itself that is useful for no other purpose.
Forget about growing in infraestructure. What I'm wondering at this moment is how the U.S. will continue to just "refurbish" and just keep up the current infraestructure.
To implement Bus rapid transit, all cities need to do is set aside land. They use their domain authorit to map out lanes for future use, and no development can "happen" there. Later on, when there is a need, the city can implement bus rapid transit, like that of Curitabo Brazil.
Infrastrucute costs cannot be monetized with a traditional accounting mindset. For example, suppose there is big reduction in foreign oil. The our transportation system grinds to a halt. A backup public system could be seen as a strategic move in that case. How much do we spend on strategic defense? Other public infrastructures help the lower classes, thus keeping social friction at bay. What is the economic cost of increased social friction?
The lack of ledger accounting on accountants ledgers doesn't mean that costs are absent. It takes a Statesman, people who are in short supply, to show us the way and to cut through the BS.
From an economic point of view, if the money is spent at the local level, by the local people, it will tend to be invested wisely.
no i don't,
My expectation is that infrastructure quality will decline. More money will go to entitlements payments while tax revenue declines due to Peak Oil and other natural resource constraints.
Mass transit is not the only possible response to Peak Oil. I expect mass transit will pay only a small role in adjusting to Peak Oil. Other responses include:
- Live closer to work.
- Move jobs closer to where people live. At the extreme, telecommute.
- Bicycles and electric bicycles.
- Electric cars.
- Smaller cars.