Scott Sumner  

The infrastructure illusion

Kibbe on Venezuela for Milleni... Chuck Norris vs Communism...

In recent years, progressives have increasing advocated more spending on infrastructure as part of a broader strategy of fiscal stimulus. I've argued that their proposals are not well thought out, for a variety of reasons:

1. Fiscal stimulus would not be expansionary; it would simply lead the Fed to tighten money more aggressively, leaving the expected inflation rate in 2018 unchanged.

2. Keynesians often engage in "reasoning from a price change", wrongly claiming that low interest rates are a justification for higher investment spending.

3. Keynesians overlook the fact that the US is not very good at building infrastructure and that our ability to do so is declining rapidly.

Larry Summers has been on the other side of this issue, a big proponent of more spending on infrastructure. Now he seems to be becoming more aware of the drawbacks:

I have an op-ed in the Boston Globe today on infrastructure, addressing the issue of quality rather than quantity of investment. Rachel Lipson, a graduate student at Harvard, and I describe the fiasco that has emerged from what should have been a routine maintenance project on the Anderson Memorial Bridge over the Charles River next to my office in Cambridge. Though the bridge took only 11 months to build in 1912, it will take close to five years to repair today at a huge cost in dollars and mass delays.

Investigating the reasons behind the bridge blunders have helped to illuminate an aspect of American sclerosis -- a gaggle of regulators and veto players, each with the power to block or to delay, and each with their own parochial concerns. All the actors -- the historical commission, the contractor, the environmental agencies, the advocacy groups, the state transportation department -- are reasonable in their own terms, but the final result is wildly unreasonable.

At one level this explains why, despite the overwhelming case for infrastructure investment, there is so much resistance from those who think it will be carried out ineptly. The right response is to advocate for reforms in procurement policies, regulatory policies and government procedures to make the investment process more efficient and effective. This is all clear enough.

Because we are increasingly inept at building infrastructure, the investment schedule shifts to the left, and this reduces market interest rates. Low rates would only be an argument for more investment (public or private) if the low rates were caused by more saving.

In the past I've argued that we need to remove regulatory barriers to building infrastructure, and also allow the private sector to become much more involved (as in Europe.)

Alex Tabarrok has a new post advocating the abolition of historical preservation laws, and I'd argue the same for environmental impact statements. Instead, I'd suggest a 6-week period for any project that formerly required an environmental impact statement, a window of time where the government could purchase the property to protect whatever species was endangered, under the eminent domain laws. Requiring the government to actually compensate property holders explicitly, rather than simply taking away their property rights through regulation, would force policymakers to think much more clearly about the costs and benefits of environmental policies.

I'd also completely deregulate labor markets, including laws that favor unions. Thus the new East Side subway being built in New York should be constructed by Chinese firms using migrant Chinese labor that sleep in those crummy blue and white metal sheds they use in China for worker dorms, and paid Chinese wages. New York would then be able to build up a world-class transport system, on the cheap.

Of course none of this will happen, which is why I'm opposed to massive new fiscal stimulus programs aimed at infrastructure. The most likely outcome in the current environment would be white elephants like the proposed high speed rail in California, which will not be very effective, and will cost so much that other needed infrastructure in California will not be built.

I'm glad to see Summers coming around to the view that the public sector many not be the right people to build infrastructure:

At another level, though, our story may illustrate phenomena that go way beyond infrastructure. I'm a progressive, but it seems plausible to wonder if government can build a nation abroad, fight social decay, run schools, mandate the design of cars, run health insurance exchanges, or set proper sexual harassment policies on college campuses, if it can't even fix a 232-foot bridge competently. Waiting in traffic over the Anderson Bridge, I've empathized with the two-thirds of Americans who distrust government.
Perhaps Summers could look at how they do things in Hong Kong, where a private company runs one of the most successful subway systems in the world:
The Mass Transit Railway (MTR) is the rapid transit railway system in Hong Kong. Opened in 1979, the system now includes 218.2 km (135.6 mi) of rail[3] with 155 stations, including 87 railway stations and 68 light rail stops.[1] The MTR system is operated by MTR Corporation Limited (MTRCL). It is one of the most profitable systems in the world, with a high farebox recovery ratio of 186%.

Under the government's rail-led transport policy, the MTR system is a common mode of public transport in Hong Kong, with over five million trips made in an average weekday. It consistently achieves a 99.9% on-time rate on its train journeys. As of 2014, the MTR has a 48.1% market share of the franchised public transport market, making it the most popular transport option in Hong Kong.The integration of the Octopus smart card fare-payment technology into the MTR system in September 1997 has further enhanced the ease of commuting on the MTR.

. . .

As a successful railway operation, the MTR has served as a model for other newly built systems in the world, particularly in mainland China.

Hmm, I wonder why the Chinese don't look to New York City for inspiration?

PS. Unlike New Yorkers, Singaporeans are smart. They use those cheap blue and white dorms for migrant workers building Singapore's wonderful infrastructure:

Screen Shot 2016-06-01 at 12.05.08 PM.png
This sort of "exploitation" also improves global equality, as Summers understood back in the days when he suggested dumping polluting industries on Africa. Ahhh, the good old days of 1990s neoliberalism. When will it return?

COMMENTS (14 to date)
Philo writes:

Summers writes: "The right response [to regulatory bottlenecks] is to advocate for reforms in procurement policies, regulatory policies and government procedures to make the investment process [in governmental infrastructure] more efficient and effective." The right response by whom? Who is Summers addressing? It would be a waste of my time to "advocate for reforms," because nobody is listening to me (not even my wife; she might say, "Yes, dear, I'm sure you're right, dear," while reading the newspaper or manipulating her cell phone). I think my correct response is passive acceptance (though not "support") of whatever policies are put in place by the federal government.

Then you write: "we need to remove regulatory barriers to building infrastructure, and also allow the private sector to become much more involved . . . ." Nah, we don't need to do any such thing (though it would be better if we did): we can rub along indefinitely doing just what we're doing. As you yourself put it: "Of course none of this will happen . . . ."

But maybe a strong man on horseback will ride in and shake things up!

Matthew Moore writes:

Scott - I understand that if the argument is 'rates are low, therefore investment demand is low, therefore government can fill the gap', this is erroneous reasoning.

But if the argument is 'rates are low, so the opportunity cost of investment is reduced, so we should do more, cet par' isn't that valid? ( assuming technology is paribus, which you don't in you post)

As an individual borrower, I tend to raise borrowing when rates are low. I don't need to know why. Why is it different for local government?

JW writes:

"despite the overwhelming case for infrastructure investment" - huh?

I'm a civil engineer in municipal government and the US has excellent civil infrastructure. People think because a road isn't free of bumps its bad even though they speed down the same road without a thought. How often have you been without clean potable water when you turned the faucet, or had raw wastewater running out of the sewer main in your street? Probably the only time you're out of power is during a major storm with high winds or lightning strikes and its usually restored within 30-45 min.

The only public infrastructure that is deficient in the US is urban drainage/flood control and it appears that as a society we are ok with the level of losses vs. cost to correct in that area.

LD Bottorff writes:

I am certainly glad that someone is taking on the historic preservation tyranny. Alex's arguments are sensible and logical. He doesn't go far enough. Historic preservation is anti-liberty! Because some people value a structure as historic, that should not give the government the power to dictate what someone can do with his property.
Compare this to the issue of abortion; some people value the fetus much more than the woman whose body contains the fetus. Should society be allowed to compel the woman to take the pregnancy to full term?

And Scott, you are absolutely correct about eliminating labor regulations; it makes sense, but it will never happen.

Scott Sumner writes:

Matthew, You said:

"As an individual borrower, I tend to raise borrowing when rates are low. I don't need to know why."

Then you are not typical. Most people borrow less when rates are low (like 2009) then when rates are high (like 2007.)

Your comment about "ceteris paribus" is closer to the mark, but I don't think people fully understand the implications of what that means. Interest rates are low partly because there is less investment demand.

BP writes:

Great column Scott.

I believe it was Matt Yglesias who was one of the first on the left to figure this out and acknowledge it (as usual).

Rajat writes:

This is a great post, Scott. Everything you say goes double for Australia, where local unionised tradespeople get paid huge salaries.

From The Australian newspaper (paywalled) - all figures in AUD where 1USD=1.4AUD:

Carpenters on union enterprise agreements struck under the Queensland Palaszczuk government last year are being paid $160,702, compared with the award of $87,064 for a 55-hour week including overtime and other benefits, an analysis commissioned by The Australian has revealed. Those in NSW are paid $152,002. Wages are higher still in VicĀ­toria, where a carpenter is paid $163,000 for a 51-hour week, figures published by Master Builders Victoria show.

Fazal Majid writes:

The US will have to fix its can't-do attitude to infrastructure if it is to remain a world player, specially when so much late 19th Century infrastructure needs to be replaced.

One issue you don't mention is that infrastructure dollars are allocated based on political representation, specially in the Senate, rather than actual need or payoff. That's how you end up with nonsensical projects like the Bridge to Nowhere in Alaska.

Benjamin Cole writes:

Great post, and obviously the private sector should build and operate infrastructure, under some sort of incentive contracts.

As I recall, US infrastructure is perhaps a cut above some developing nations, and a cut below, when it come to airports. US infrastructure is perhaps the crappiest of the developed world.

That said, there is weakness on the American Right, to only examine boondoggles on the civilian side. National security boondoggles remain off limits.

Given the more than $6 trillion in outlays and incurred liabilities in Iraqistan, this is a shortcoming.

For that matter, the U.S. Naval Institute (military guys) reports the $1.5 trillion fighter-plane project F-35 is no longer stealthy (changes in the radar of adversaries), but that performance has been compromised to try to be stealthy. Smaller payloads, less flight range, slower etc.

The F-35 will get built, no matter. Too many Congressional districts and lobbyists involved. Oh, the airframe for the F-35 will be built in...Turkey.

Why Turkey? So they will buy the plane when built.

The F-35 is no longer about what is the best way to build the plane, or if it is a good plane, but who gets the money to build it. Like all federal projects, the F-35 seems to be taking forever, and nary a plane in service.

Yet, in WWII planes were designed and built and in the air in a couple of years. What the Brits did with the de Havilland Mosquito is amazing. Under wartime conditions! With shortages of labor and steel! One of the best fighter-bombers of the war.

In contrast: "The F-35 took its first flight on 15 December 2006"-- Wikipedia. Sheesh, 10 years later, and they are still test-flying this plane. And how long to get the first copy into the air?

"The JSF development contract was signed on 16 November 1996, and the contract for System Development and Demonstration (SDD) was awarded on 26 October 2001 to Lockheed Martin, whose X-35 beat the Boeing X-32."

So, my reading is that it took 20 years from the development contract to today, but the plane is not yet in production.

Five years for a bridge? Horrors!

It would be nice if the American Right expended scrutiny across the board of federal spending. But I won't hold my breath.

The national security budget (DoD, DHS, VA, black budget, prorated interest on debt) is now $1 trillion a year.

Benjamin Cole writes:

Sorry but add on, from Wikipedia:

The 2015 DoD annual report stated that the current schedule to complete (F-35) System Development and Demonstration (SDD) and enter Initial Operational Testing and Evaluation (IOT&E) by August 2017 was unrealistic, instead the program would likely not finish Block 3F development and flight testing prior to January 2018. Based on those projected completion dates for Block 3F developmental testing, IOT&E would not start earlier than August 2018.[160][161]

The Block 3I software was intended to be a revision of the Block 2B software to run on the updated Integrated Core Processor,.[162] Unfortunately, it resulted in a timing misalignment that reduced stability, requiring reboots in flight.[163] By May 2016 the software had been improved to the point where it only crashed every 15 hours.[164] Reboots were required over a third of the time for ground cold starts.[165]


Okay, so the F-35 will not be ready for serious testing until late 2018, more than 22 years after the design contracts were let. If then. When serious production will commence is anyone's guess.

Meanwhile, many say really cheap drones, even disposables, are the way to go. Well, shut my mouth.

It is those preservationists who are getting in the way.

I predict there wil not be a round-robin in Tyler Cowen, Econolog, Cochrane's blog and who knows how many other platforms about F-35.

BP writes:

Great column Scott.

I believe it was Matt Yglesias who was one of the first on the left to figure this out and acknowledge it (as usual).

Scott Sumner writes:

Thanks everyone, lots of good points.

Lorenzo from Oz writes:

Fun post.

Slight wrinkle: infrastructure is generally a network or part of a network, and the economics of networks complicates stuff.

There seems to be something of a long term pattern of private construction of infrastructure until the negative effects of that build up (e.g. underprovision due to positive externalities not being able to generate income) when there is a shift to pubic provision until the negative effects of that build up (see this post) when there is a shift back to private provision.

So I would add regulation of land use to your problem issues. Land rationing tends to undermine a basic benefit of public provision (capturing increased value via taxation) as it raises the opportunity cost to government of using the land while generating (much more cheaply) increased revenue by driving up the value of the rationed land.

Larry writes:

Yes, the regulatory process is obscenely burdensome. It took 25 years to replace the half of the SF Bay Bridge damaged in the 1989 earthquake, and the cost was 25x the initial estimate!

But I'd add another "regulatory" option. Have the developer buy insurance that if unacceptable environmental damage ensues, pays a claim to the government. Now the insurer will push to ensure that the developer behaves, and the developer will be looking to reduce his premium by adopting better designs and practices. No gvt bureaucrats need be involved.

Comments for this entry have been closed
Return to top