U.S. Treasury’s end-of-summer $50 billion bonfire

President Obama is asking Congress to give $50 billion of our grandchildrens’ money to America’s least efficient industries: highway construction and the FAA (nytimes). First, let’s think of the scale of the spending. As U.S. adults seem to have no intention of funding current federal spending or paying off debt, Obama’s spending plan will be paid for by the 50 million or so children who are aged 12 and under. That’s $1,000 per child.

Let’s look at the track record of the sectors of the U.S. economy to which Obama proposes to give this $50 billion. One is highway construction. This is an industry that has developed so few new efficiencies and innovations since the time of the Panama Canal that Americans can no longer afford to maintain the highways that we have, much less build new ones (see the book Traffic: Why We Drive the Way We Do). The relative cost of a highway has gone up so much compared to other things that we might purchase that it is almost insane to contemplate buying more of this product. It would be like someone learning that gold was over $1200 per ounce and saying “Now would be a great time to gold-plate the exterior of my home” or finding out about some $5000 per kilo caviar and saying “I will eat nothing else for the next 12 months”.

Much as I love aviation, as a taxpayer it pains me to see money being shoveled into the mouth of the FAA. Obama promises that some of the $50 billion will go for a “next generation air-traffic control system”. The last time the FAA tried to produce an update for some of its ATC software the result was a project that was more than a decade late when the plug was finally pulled. It cost the taxpayers $9 billion and every line of code had to be thrown out. The agency’s latest project is a change to the way aircraft are registered. For the past 80 years or so, aircraft registrations have not had to be renewed. Owner and pilots notified the FAA when an address change was required. The FAA has decided now to send out U.S. mail reminders to every aircraft owner every three years asking “Do you still live at the same address?” A paper form will be returned by the aircraft owner. (FAA’s official explanation of the process) Figure it costs $300 of owner time and FAA time to work both sides of this paper-intensive transaction, including handling a physical check for $5. There are 357,000 aircraft out there, so that’s about $36 million wasted every year.

How would a private company do this? The FAA already pays contractors about 50 cents every time a pilot connects to a weather briefing service such as DUATS. These services are limited to folks whose pilot certificates have been verified. Nearly all aircraft owners, other than airlines, are pilots who use weather briefing services periodically. If Amazon wanted to get some information out of a customer every three years it might put a note on the Web site saying “Would you mind visiting this link and updating a form?” A private business whose Web site was visited at least monthly by nearly all of its customers would only use paper as a last resort, but the FAA apparently never considered using the Web services that it is already paying for to do the job.

An MIT Civil Engineering professor and I once visited the head of Boston’s Big Dig project, an executive at Bechtel. We showed him some software. He said “Wow, this stuff could save 15 or 20 percent of the cost of the project and a lot of time also. I’m not interested in it.” Why not? “I’m getting paid cost-plus on this project. If we build this highway using the same methods used by the Romans to build their roads, that’s fine with Bechtel.” (The project ended up costing nearly $15 billion or $22 billion when interest paid on bonds is considered. Estimates of the return on the investment range from $0 to $167 million per year, i.e., an ROI of between 0% and 1.1%.)

As the U.S. population trends up to between 600 million and 1 billion (range of estimates for 2100), won’t we need more roads? We can’t afford them so we won’t have significantly more, regardless of what Obama promises. How about public transit? New York City is reducing its public transit offerings; it can’t afford to run trains and buses as well as pay pensions previously agreed to (see nytimes and this article on New York state pensions). The future for Americans may be a combination of walking and videoconferencing. It will make a lot more sense to build dense cities where folks can walk to work, shop, and see friends, rather than trying to create enough additional sprawl for 700 million more Americans. What could we private businesses and citizens do with $50 billion if the government did not take it from them and hand it over to our least efficient industries? At $100,000 per conference room, 500,000 businesses could be set up with amazingly high quality videoconferencing/telepresence systems, thus reducing the need to travel for business meetings. At $1,000 per desktop, 50 million American homes could be set up with moderately high quality videophones, thus reducing the need to travel for social meetings. Given a free choice, it seems inconceivable that private citizens would decide to give their money to highway contractors rather than Cisco, HP, and other innovative companies.

[Comments on the nytimes piece seem rather negative. Apparently people aren’t excited when the government buys stuff that they themselves wouldn’t want to buy. Given that Obama and his advisors are famous for political savvy, I’m wondering why they thought that anyone other than a highway contractor employee would be happy to hear about this proposed spending.]


  1. Fazal Majid

    September 6, 2010 @ 9:41 pm


    The SF writer Charles Stross says IT Project Management is a core competency of modern government (or more precisely, ought to be). I can see two ways the FAA’s project could be done on time and on budget:
    1) write an emulator for the current Burroughs mini-computer based system, and virtualize the whole darn thing. Given how laughable that legacy gear is compared to modern CPUs, even the most inefficient emulator should be able to run rings around the old system. The hardest part would be finding JOVIAL programmers to extend the hard-coded limits in the program.
    2) license the Canadians’ system, which seems to work pretty well. Of course, that would mean swallowing national pride and disappointing some well-to-do government contractors…

  2. Libertardian

    September 7, 2010 @ 5:14 am



    As a citizen of Los Angeles, I have noticed the roads here are deteriorating noticeably. My car has taken a beating in west LA. $50b would not make a dent, seriously, in how far behind the CURRENT roads are. A little off topic, could you comment on Christina Romer’s publication (http://emlab.berkeley.edu/users/dromer/papers/RomerandRomerAERJune2010.pdf)
    that has the conclusion:

    “Our baseline specification implies that an exogenous tax increase of one percent of GDP lowers real GDP by almost three percent.”

    As a libertarian, this paper is golden. We are generally bad with statistics and ideology. It’s very hard to gather support an ideology based on individualism, and even more difficult to gather statistics about that idea, since people are allowed to just follow what’s good for themselves. Statistics kind of become meaningless at that point. Herding cats comes to mind. It seems like the liberal that wrote the paper seems to agree, taxation is bad. Though not much is made of why. The blog post you wrote illustrates that though.

  3. philg

    September 7, 2010 @ 8:22 am


    Fazal: The Europeans have a privatized ATC system and very intensive traffic (roughly same number of flights in a much smaller area), so they probably already have everything that we would need over here.

    Libertardian: No argument that, given current highway contractors and practices, $50 billion would not fix the roads of LA, let alone the highways in the rest of the nation. My point was that some things that were affordable in the old days, e.g., custom-made clothing by local seamstresses, aren’t affordable today. If we assume that the government’s current contractors are as efficient as possible, the only logical answer is that we must consume fewer roads, including in LA (which could gradually be remade into a city of 40 million, mostly living in dense walkable clusters and travel from Pasadena to Santa Monica, for example, would be an annual treat rather than a daily experience).

    Thanks for the Romer paper reference. Certainly the proposed $50 billion spending spree will eventually have to be paid for with taxes. I did not read the paper carefully, but I think that there is an escape hatch for proponents of big taxes increases. The Romers suggest that if those tax increases are done during a period of powerful economic growth and are done specifically to eliminate budget deficits, the effect on GDP might not be pernicious. In the case of taxing to close a deficit business investors might not see it as a big change since they already factored in the fact that the deficit spending would have to be paid for somehow. Right now the federal deficit is almost half of federal spending (1.5 billion deficit compared to tax revenues of just over $2 billion). There is no peacetime precedent for that kind of deficit so looking at historic data may not be helpful. We would need a tax increase of more than 10 percent of GDP to close up the deficit and that kind of tax increase has never been tried in the U.S. (or maybe anywhere?). Separately, there seems to be little prospect of a period of powerful economic growth any time soon. The Japanese government probably ran the same calculation when they ran up big debts in the 1990s. They were counting on a big boom in the 2000s where they’d raise taxes and pay back the debt. Big booms had always followed downturns, so they were sure that the cycle would reverse. Except this time it didn’t.

    Working this out in my brain while typing, I guess that is my main comment on the Romer paper. It isn’t likely to convince anyone who believes that the U.S. economy is cyclical and that what goes down now will inevitably go up later. They think that they can painlessly tax working Americans during the coming good times and they will take confirmation of that fact away from the Romer paper. There are a lot of international counterexamples to the “coming good times’ theory, e.g., countries such as Argentina and Japan that had cyclical periods of growth until they flatlined. The true believers will say that America is different because we have a more diverse society, a special work ethic, or something else that makes us completely unlike Argentina and Japan (the Japanese that I encountered on a Tokyo to Hokkaido road trip didn’t have any trouble telling North from South, for example, which would preclude them from employment by T-Mobile).

  4. Jean-Francois Noel

    September 7, 2010 @ 10:34 am


    The government in Canada has the same reflex to stimulate economy, give money to road contractor. It seems they don’t know how else to spend money in a stimulating way, so even if it’s totally useless they do it. The high ROI highways have been in place for a long time, so now they are just fiddling with exits layout and what not. Sometimes I’m afraid I’ll learn that they are building a second Hoover Dam in front of the Hoover Dam, since it was such a good idea the first time… Obviously now it will cost 50 times more and produce no new electricity.

  5. Douglas Johnson

    September 12, 2010 @ 8:32 am



    I’m in highway construction, and given my political bent I want to believe what you wrote (only had time to skim it), but it’s not my experience.

    Every job I have ever bid comes down to low bidder. In my limited 20 years of experience I have never seen, bid, nor worked on a “cost-plus” project as you described.

    My business is the construction of shoulder rumble strips. We designed and built a machine to do the work 20 years ago, and back in 1990 we led the industry winning projects at $2.00/cut. Today we still lead the industry, but now we bid as low as $0.05/cut. Our first machine could average out one cut every 3 seconds. Today we make 7 cuts in one second. It cost us many, many millions of dollars to get here. We wouldn’t have invested a dime if it were not for competitors nipping at our heels. Because in my experience, the only thing that counts is lowest price.

    That said, I would be immensely happy if we privatized our entire highway system, but I’m just sayin’…

  6. philg

    September 12, 2010 @ 10:25 am


    Douglas: I didn’t mean to imply that most highway projects were cost-plus. The oft-quoted figure for the original Interstate construction was $1 million per mile back in the 1950s. That’s about $8 million today’s dollars, according to http://data.bls.gov/cgi-bin/cpicalc.pl , and I believe it paid for a standard 4-lane highway in a Midwestern state. http://www.fhwa.dot.gov/steam/table1.htm shows that the modern cost of adding a single lane-mile is $10 million. So in constant dollars, expanding a highway from 4 to 6 lanes, for example, would cost roughly 2.5X what it cost to build the 4-lane highway in the first place.

    Manufactured goods, such as appliances, electronics, clothing, etc., have generally fallen in cost due to improvements in design and factory techniques. Even food has gotten cheaper relative to inflation. In http://www.bls.gov/opub/uscs/1950.pdf we see that flour was 49 cents for a 5 lb. bag in 1950. That’s $4.43 in today’s dollars. http://www.totallyfrugal.com/forums/showthread.php?t=7774 has folks buying the same bag for between $1.33 and $2.50. A dozen eggs were 60 cents, which is $5.43 today and I don’t think you’ll see even Whole Paycheck trying to get that much.

  7. Douglas Johnson

    September 13, 2010 @ 7:31 am


    Ah, I see what you are saying, but I’d correct your statement that “This is an industry that has developed so few new efficiencies and innovations since the time of the Panama Canal that Americans can no longer afford to maintain the highways that we have…”

    As I said every job we’ve ever bid (and I’ll guess ninety-something percent of all U.S. highway construction work) is low bid. My niche has seen a 97% drop in prices since 1991. But that was for a new process with all the innovation ahead of it. We went from $2 to $0.05 pretty over 15 years. That isn’t going to happen again. Anything that is technology driven in highway construction has seen something like this, and the major trade shows in highway construction are about nothing more than innovation to drop costs.

    But look at the Illinois Tollroad system, and every other tollroad in the country that has installed some variation of iPass. Part of the promise was that it would get rid of all the toll booth workers and save costs in labor. But employment at the toll road system hasn’t gone down by a single worker.

    As for our non-toll highways, no one can ever lose a job and everyone working for a Department of Transportation office will make more than they did the year before no matter for every year they have been in existence. Job performance, and economic ups and downs don’t matter.

    A second point of course is unions. I don’t have time to look up the link, but highway construction costs are about 8X in NY v. FL. Florida is a right to work state and NY is not. My prices are 1/2 to 1/3 of my competitor’s prices in IL & OH (maybe 1/5 my competitor’s costs in southern CA), but I can’t get a single job there as a subcontractor because we aren’t union. (I can win jobs I bid as a general contractor, but there aren’t many jobs for my niche as a general contractor.) And the unions negotiate their wages not with contractors but with the state. The state then requires that I pay my non-union workers what the state agreed to pay the union workers. In Florida, I could pay my guys about $15/hour, but their company wage is actually $25 per hour. In Nassau County, NY where I just bid last week, I have to pay my guys $76/hour.

    So while highway construction workers are technically in private sector unions, they really aren’t in that they negotiate their wages with the state and so they are really government unions. The politicians that promise to fight for the unions for better wages are really saying, “elect me because when you sit down to negotiate your wages with me I’ll obviously give you more than my opponent who is anti-union.”

    But anyway, to get back to my first point, technology innovation and cost reduction is alive and well in the industry, but the Department of Transportation is essentially 100% tenured work, and the unions are little more than government collusion with no risk to risking wages. In that sense, the proper comparison is the cost to run a department of transportation office in 1950 v. 2010, and the cost to run a university in 1950 v. 2010. I’m guessing similar trend lines.

Log in