Saturday, July 27, 2013

Infrastructure skeptics take a hit

Via Barry Ritholtz and Mark Thoma comes the following chart by McKinsey:


This chart simultaneously debunks not one, but TWO of the biggest talking points of the anti-infrastructure derpers:

Derp 1: "It's just civil engineers angling for pork!"

This is one I hear a lot. The American Society of Civil Engineers released an "infrastructure report card" that gave the U.S. a "D+". In response, anti-infrastructure people often throw up their hands and say "Of course a bunch of civil engineers want us to spend more money on civil engineering!" But now McKinsey says the same thing. The chart above cites the McKinsey Global Institute's own in-house analysis. If you think McKinsey itself is in the tank for a bunch of gold-digging civil engineers,you're...well, it sounds pretty silly when you say it out loud, doesn't it?

As an aside, the McKinsey report also cites independent figures from the World Economic Forum, the Army Corps of Engineers, and the Federal Transit Authority. Everyone agrees that the U.S. needs more infrastructure spending.

Derp 2: "Oh yeah? What about Japan in the 90s?"

A lot of anti-infrastructure people bring up Japan's infrastructure splurge in the 1990s. Japan built bridges to nowhere (I drove across one, once; "nowhere" is not an exaggeration), concreted over riverbeds and parks and beaches, and generally committed an epic waste of money, labor, and concrete. We'd commit the same error if we spent more on roads and bridges, right?

Wrong. As the McKinsey report shows, Japan still spends much more on infrastructure than it needs to (and this, btw, is after a dramatic reduction in spending in the '00s). That is in stark contrast to most other countries. So this report clearly says that we are not Japan. We are on the opposite side of the optimal point. They need less, we need more.

In conclusion, McKinsey just killed two of the main anti-infrastructure arguments. Increasingly, all the anti-infrastructure people have left is their derp. At a time when interest rates are super-low and millions are out of work, there is no good case against a big blast of infrastructure spending. Let's do it right now.

Update: Numbers! We need to spend an additional $1 trillion (that's 1000 $billion) dollars on infrastructure over the next 5 years. or about 1% of GDP.


  1. I have no problem with government spending on infrastructure, which seems to be a fairly basic function for the most part. If something needs to be built, then by all means build it. However, a few points:

    * I don't understand why an infrastructure deficit means that the federal government should start shoveling money out the door on infrastructure. Isn't most infrastructure a state and local responsibility (or at least, shouldn't it be)? Why should the feds fund this? As an example, there is a dog park here in DC that was at least partially funded with stimulus spending money meant for infrastructure -- why? Seems to me that federal spending also runs the risk of public choice problems, with money being disproportionately allocated to areas with politically powerful members of Congress (e.g. West Virginia with Byrd/Rockefeller) and rural areas instead of cities owing to their disproportionate representation. Lastly, there's a simple efficiency question -- why take money from the states, send it to DC, and then have DC send it right back?

    * Is infrastructure spending a great way to put people back to work? I mean that as a serious question. First off, don't infrastructure projects take a while to plan, do the environmental impact studies, etc. and then finally build? Seems like the increased employment won't be felt until at least a couple of years in the future. Also, can just anyone do a construction job? How low-skilled is this? I think a lot of people have a conception of a bunch of unemployed guys standing around warming their hands by a trash can with a fire burning inside of it and thinking you can just round them up and put them to work. Maybe I'm wrong, but operating some of the equipment used for road and bridge building doesn't strike me as something just anyone off the street can do.

    * In addressing the infrastructure deficit, how much does it logically follow that the solution is more government spending? May at least part of the solution be an increased role for the private sector through, for example, toll roads and airport/seaport privatization? Some infrastructure, such as the road in front of my house, I can only see government taking care of. But I am not sure this extends to all infrastructure (indeed, US freight rail is among the world's best).

    * Before spending more money on infrastructure I would love, love, love to figure out why we get so little bang for our buck compared to other countries.

    Basically, my attitude towards infrastructure spending is this: if something needs to be built, fine, but let's first try to wring out as much of the inefficiency as possible -- which seems to be considerable -- before we start throwing tens of billions of dollars around.

    1. I agree with you that our infrastructure costs are way too high here in the US but would push back on most of your other points. It's true that states and cities are "responsible" for their own infrastructure, but that doesn't exclude the possibility that the Feds could pump more money into their budgets to do things like repave crumbling streets. Cities by and large live at the mercy of credit rating agencies that are pretty much willing to downgrade their rating at the drop of a hat. Like if the a city borrows money to fix infrastructure. States often have rules requiring them to balance their budgets every year meaning complex bonding schemes have to be put together for infrastructure.

      Put it all together and it basically means we could easily have the federal government spend a bunch of "shovel ready" projects, like filling pot holes and repaving streets in cities if we wanted to. But the cities themselves can't really do it all by their lonesome.

    2. Also, to Colin, magnifying lwdl, state and local governments are a general shambles. In theory, the smallest unit (municipalities) should be the most effective and efficient in these matters, but reality disagrees. Partly it is because so many of these projects are too big for one (or many, in some situations) provincial backwater(s) to handle.

    3. It is not obvious to me that cities and states lack the money to pay for infrastructure. Here in DC the local government spent something like $700 million for a new baseball stadium, and appear set to subsidize a new soccer stadium to the tune of $150 million. Even bankrupt Detroit is set to spend $283 million in taxpayer funds on a new hockey arena and phase 1 of California's high-speed rail is set to cost $68 billion per wikipedia (granted, this will be partially funded by the feds). If ratings agencies have so much power, how can states/cities borrow for stadium boondoggles but not for infrastructure improvements for which there is a demonstrable need?

      One other point: ideally reduced federal spending on infrastructure would be accompanied by reduced federal taxation, thus leaving more money at the state and local level to be taxed by those governments.

      Lastly, again I wonder just how truly "shovel ready" many of these projects are. Even something as basic as streetscape improvements here in DC usually require plenty of planning. Indeed, here is one streetscape project (widened sidewalks, new curbs, repaving, bike racks, street lamps, etc) that, according to the DC government, required "several years of planning, preliminary engineering, and public outreach focused on pedestrian, transportation, and landscaping improvements."

    4. Jonas -- perhaps you are correct about the actual project administration, I am not knowledgeable enough about the subject to disagree. But I think the public choice/misallocation angle (fantastic highways for West Virginia and overemphasis on rural needs over urban due to disproportionate GOP/rural influence) seems pretty undeniable, no?

    5. Colin,

      All your objections and caveats are very reasonable indeed.

      That's one of the troubling thing with reality. It has a well-known liberal bias but lots of conservatives can still put up a darn good fight with reasonable objections. Which is why this article by Noah Smith will never convince anti-government types who just know, in their DNA, that governments, always and everywhere, are utterly inefficient (except when it comes to military spending).

      TBH, I do think that infrastructure would have been good use of stimulus funds and could still be. But it is indeed slow.

      I prefer debt forgiveness and helicopter money.

    6. Colin- I agree with your 10:15, that this sort of misallocation is an issue, but I'd only say that this is a problem no matter the scale, and the operative question is whether the misallocation due to big-picture concerns (or the public choice stuff you mention) outweighs the economies-of-scale-abilities of bigger agencies.

      As to you 10:10, sending the revenues back to the states is only a benefit if the states can show that they can spend appropriately and efficiently. It's just the public goods question. If one believes that the states are definitionally better at public goods provision, it's only a hop, a skip, and a jump to 'individuals are better able to provide these things than are states.' At this point, one has entered the (intellectual) on-ramp to the Henry Hazlitt Highway, and that's no place for any man.

      *The Henry Hazlitt Highway has no speed limit (coercion) or maintenance services (obtained via coercion), and the toll must be paid in gold.

    7. Removing corporate lobbying would make for more efficient infrastructure investment.

    8. Anonymous4:47 PM

      Re: the first point:

      Infrastructure needs and the ability to pay for them locally are two different things. Most local and State governments have limits on debt financing that the feds do not have. The cost of money "going" top Washington and coming back is minimal in an age of electronic transfers. Its true that politics influences choices in Washington, but it does at the state and local level as well, so that is really a moot point as long as allocation of resources is politically determined (ie., always).

  2. I think the best use of "derp" is for people who think "Bayesian inference" is some incredibly new thing that supplies a total key to life.

  3. I'm not sure if I understand the concept, but could this agreement b/n McKinsey et al be the result of an information cascade? Are we sure that the same data and methodology isn't being reused? How independent are those independent figures?

  4. I always wondered why Japan didn't spend more on mental and other human infrastructure, let alone scientific infrastructure and research.

    Ok, so they're worried about unemployment in their long long liquidity trap and want fiscal stimulus, but why build bridges to nowhere when so much of your population has the ability to go to college and doesn't (from what I've read on Japan, which isn't a great deal). Why not really beef up instead education and training and pre-school and prenatal care, let alone basic scientific research. And I heard their Wi-Fi was bad, a reason for heavy fax and smart phone use. Why not beef up that infrastructure, become a leader in some of the super Wi-Fi technologies I've read about that can blanket a city.

    Part of this is probably, sadly, the political strength of the construction industry.

    1. And if you're going to build, how about like 10 Dubai sized skyscrapers in Tokyo, so people don't have to impoverish themselves to live there, and more people can enjoy, and become more smart and productive, living in a big city?

      Obviously, politics there. We have the same horrible problem with selfish current property owners doing profound harm to their country to jack up the price of their property.

    2. 10 Dubai sized skyscrapers in Tokyo


    3. And what about Godzilla??!!

      But seriously folks, good point. However, there are a lot of tall buildings in Tokyo. Some of them have amazing technology to withstand earthquakes. But ok, 100 buildings, 1/10th the size of the Burj Dubai, perhaps many on the outskirts with a really cool monorail to quickly get anywhere in town.

    4. But ok, 100 buildings, 1/10th the size of the Burj Dubai, perhaps many on the outskirts with a really cool monorail to quickly get anywhere in town.


  5. Anonymous8:51 PM

    Infrastructure spending = Investment = Higher Multiplier than pretty much any other spending.

    So that is why Taylor thinks you can make austerity work - by substituting social spending with investment (infrastructure).

    What would happen if Japan does the reverse?

    And is it really investment if its a bridge to nowhere? How does it increase capital?

  6. Anonymous3:59 AM

    As for the Feds vs. state, there are some infrastructure investments that require coordination across multiple states (think the interstate highway system, the power grid, or a tunnel between NJ and Manhattan) where the Feds can help.

  7. This is more a narrow technical point than a general issue - but when trying to prove that people other than civil engineers want infrastructure spending, is citing the Federal Transit Administration (who provide cash and advice to transit systems) and the US Army Corps of Engineers (who operate dams, canals and plenty of other water-based infrastructure as well) really a wise step? :P

  8. Most U.S. military spending has been a "bridge to nowhere" for the past three decades. But then they blow them up and build them again. And nobody much complains because, TERRISTS!

  9. Anonymous10:12 AM

    Ok, so the conclusion of this article is just that actually there is a hidden association between the civil engineers and McKensey. The deal could be 50/50 from the benefits for each one; that's easy said as the taxpayer always pays... Spain could be another nice example of "infrastucture wating": . Oh damn, and they don't even have a central bank which could print some money to help!

  10. I generally agree that infrastructure spending in the US is too low. However, having worked on several projects with McKinsey "experts" in my corporate career, I have to say that they are capable of spectacular failures in data presentation and interpretation.

  11. Some infrastructure projects do take time to plan. On the other hand we are in year six of a downturn. Six years ago the right was saying infrastructure took too long to plan. If there had been some aggressive moves on infrastructure six years ago it might have done some good. It is not too late. We do not know how long the downturn is going to continue.

    Fixing potholes does not take a lot of advance planning. Fixing a bridge takes less planning than replacing it.

    Detroit is in an economic mess but vested interests are still opposing Canada injecting billions of dollars of Canadian money (and triggering billions of US Federal money) into the Detroit area by building the new international crossing between Detroit and Windsor. It is a disgrace that construction is not yet underway.

    The Fed could support local infrastructure projects by guaranteeing a market for state and local bonds issued to pay for infrastructure.

  12. Anonymous12:00 PM

    "we need to spend"

    what happens if we don't? America just isn't down with infastructure

    1. This comment has been removed by the author.

    2. By 'America,' are we to presume you mean yourself? Because I think I missed that big nationwide vote we had on the subject. I'm pretty sure we all have an infrastructure project of which we approve. The fact that these projects are not the same does not mean we all want nothing.

      'If we don't,' then our roads and bridges continue to crumble until we begin to resemble the third-world country that our political process became long ago. But who cares because 'FREEDOM!' right?

    3. America just isn't down with infastructure

      Barefoot is best!

    4. what happens if we don't?

      The forest wins and takes back America - starting with Detroit.

  13. 1. I'd like to see data on stocks.

    2. If we're short, one reason might be that the government is spending so much money on other things. Arguably entitlements are squeezing out things like this.

  14. "It's just civil engineers angling for pork!"

    Sometimes it is, sometimes it isnt. Some places are underinvested massively in infrastructure and sometimes it is pork. A system of effectively identifying when to invest in infrastructure needs to be put in place. Removing corporate lobbying should be beneficial in minimizing pork projects.

    "Oh yeah? What about Japan in the 90s?"

    In Japan infrastructure investment was an overall positive for the economy but the problem was a disfunctional monetary system due to the excessive leverage overhang from the 80's which undermined the economy because people deleveraged.

    Dont get monetary and real economic causes of an issue mixed up.

  15. Oh yeah, Noah, well, have you ever hear of public choice? Clearly DOT and the Army Corp and all of those others just have their hands in the cookie jar.

    Okay, so, that's dumb, obviously. But what gets me most worked up about this particular derp is the giant leap from "that source is biased" to "we don't any more infrastructure investment." Russ Roberts argued during a debate (I think with Dean Baker) on Planet Money (quite some time ago now) that there was "no evidence" of any need to invest in infrastructure. Sure, you've got a report by the civil engineers, but they are biased, so that's "no evidence." Not "maybe that numbers a little high because they are biased," but literally no evidence.

    Meanwhile, try driving American highways and city streets and then compare them to, say, Germany. Resurfacing our crumbling roadways would alone add up to a pretty big number.

    1. Russ Roberts goes by the handle of "EconTalker". Maybe he should go by "EconListener"?

    2. EconDerper?

      Perhaps there is too much competition for that moniker.

  16. This would be another best Infrastructure..