Fighter spending gone bezerk?

Fighter Chart

Feast your eyes on this chart released two months ago by the Congressional Budget Office, which came attached with an alarming note on the blog of CBO director Douglas Elmendorf.

“The Air Force has been spending nearly as much annually to purchase relatively small numbers of F-22 fighter jets and Joint Strike Fighters (JSFs) as it did to purchase much larger numbers of F-15s and F-16s during the 1980s,” Elmendorf wrote.


The USAF’s purchasing power is declining dramatically over a 30-year period, according to CBO. Adjusted for inflation, the same amount of money — roughly $6.8 billion in today’s dollars — that was good for 220 jets in 1986 is worth only 80 jets in 2016. In the computer industry, technology becomes cheaper even as the products become more sophisticated. The chart above is another vivid example of how that is not happening in the aerospace and defense industry.


Subscribe to our e-mail newsletter to receive updates.

, ,

20 Responses to Fighter spending gone bezerk?

  1. K 25 August, 2010 at 2:03 am #


    I guess when you have fewer and fewer projects that go on for longer and longer, they gunna cost more and more.

    Another good way to boost costs is to invest all the IR&D money, all the test money, the money to get the production line going and the money to buy most of the parts, and then cancel the program just when you start turning out the product where each plane you make actually lowers the unit cost.

    Besides, we’re not talking purely weapons production here, we’re talking the far more important congressional and senate election boosters.

  2. John S. 25 August, 2010 at 2:13 am #

    F-15s and F-16s can still be purchased today. What would that chart look like if the F-22s were replaced by F-15s, and the F-35s were replaced with F-16s? What operational capability would be lost if F-22s and F-35s were replaced by F-15s and F-16s?

    And let’s dare not open up the can of worms comparing total program costs divided by number of aircraft purchased, vs. unit flyaway costs per additional aircraft purchased. You cancel a lot of aircraft, your per aircraft program costs rise accordingly.

    And the comparison between fighter aircraft and the computer industry is silly. How much would a Intel Core2 processor chip cost today if only 187 were to be manufactured?

  3. Stephen Trimble 25 August, 2010 at 2:19 am #

    I don’t think it’s silly. Aerospace is a technology. Computers are a technology. There’s no fundamental reason why they both shouldn’t be expected to become cheaper and provide better performance as time goes by. Obviously, politics, budgets and inefficiency gets in the way, but those are just excuses, no?

  4. Sven Ortmann 25 August, 2010 at 2:31 am #

    That’s roughly 3.5% annual cost growth in 2010 dollars (so the national average inflation was already removed from the figures).

    This increase doesn’t seem to be too outrageous to me, the problem is a different anyway.
    We shouldn’t compare present and past costs of apples and oranges, but instead look at the waste.
    Gold-plating, poor project management, wrong incentives, pork, politics-driven choice of subcontractors, avoidable delays and the strive for “revolutionary” advances with technologies that still need to be developed (= high risk) are systematic reasons for way too high costs. Only the Japanese are worse (with their ridiculous annual production rates).

    The Swedes could teach us a lesson on cost-efficient military hardware R&D and procurement – perhaps the Israelis, too.

  5. Clovis B 25 August, 2010 at 3:33 am #

    No competition is causing this. We went from having four makers for the F-15 competition – General Dynamics, Fairchild Republic, North American Rockwell, and McDonnell Douglas to two for the F-22 – Northrup and Lockheed.

    Now we’ve got two makers total – Boeing and Lockheed. There are a total of five makers in the West how long will Dassault and Saab/BAe last, another generation?

    Moore’s law and increasing use of composites should be driving the price down, but the aviation industry is shrinking too much and driving the prices up.

  6. Anonymous 25 August, 2010 at 5:39 am #

    Take the 2026 F-35 numbers and the 1992 F-16 numbers:

    50 F-16s for $3B
    $60M per

    80 F-35s for $5.5B
    $69M per

    I’d take an F-35 over an F-16 for a 15% increase in cost.

  7. zeno 25 August, 2010 at 7:02 am #

    hm, not silly but misleading. we won’t be hangin’ around with a F-xx or sort of as we do with a smartphone or an iWhatever or a laptop which all surpass the operational capability of the IBM PC of the ’80. IMHO the point is we are not able/allowed to tell what will be the “civil” fallout value of the military spending.
    all military crap is much more related to science than mere raw materials and manufacturing in a geometric progression. resources are not available in the same way so we have to slash inventory.
    effectiveness first, efficiency if possible, civil/public technology fallout and relevant benefit ASAP

  8. Steve 25 August, 2010 at 11:07 am #

    Elmendorf chose the 80′s by which time all the teen series were in full rate producttion – well down their learning curve. Had he chosen the mid-70′s, or earlier, during their development phase the difference might not be so great. Remember too that the B-1 was developed in the late 70′s then cancelled because we now know the B-2 was being developed – and how many of those were produced??

  9. Gilles 25 August, 2010 at 1:21 pm #

    Numbers built according to Wikipedia

    B-1 Lancer 104 with 66 still in service
    B-2 Spirit 21
    F-117 64, all now retired
    F-22 168

    B-52 744
    F-4 5,195
    F-14 712
    F-15 1,198
    F/A-18 1,480
    F-16 4,400
    A-6 693
    A-10 716

    Very high tech, very expensive and very few has never been the way to go.

  10. Stephen Trimble 25 August, 2010 at 1:51 pm #

    In my point of view, this logic is going in circles. The air force never intended to buy only 21 B-2s and 187 F-22s, for example. But that’s all they were allowed to afford in large part because the cost was so high.

  11. John 25 August, 2010 at 4:42 pm #

    We have to start separating out the research funding from procurement budgets. Programs have to develop actual planes with the technologies at hand. The F-22 is a example of what happens when you combine the two.

    Why not set a price of 100 million for a F-22 replacement/complement and keep the price for 10+ years. With a long term achievable goal manufacturers could build efficiencies and plan for inflation. This would also allow the air force to budget for the long term.

  12. alloycowboy 25 August, 2010 at 7:45 pm #

    What you are failing to grasp is it’s the cost of the advanced avonics and software that is really driving up the cost of the fighters up. But once they go through this software development phase they should be able to roll the software over for other fighters and derivatives.

  13. Stephen Trimble 25 August, 2010 at 7:48 pm #

    Why can’t software and avionics follow the same pricing trends in aerospace as they do in the commercial IT industry?

  14. Anonymous 26 August, 2010 at 6:11 am #

    It’s the equivalent of developing Windows. And then selling it to 3000 people.

  15. Atomic Walrus 26 August, 2010 at 4:05 pm #

    Commercial electronics don’t have to be radiation-hardened, and there aren’t many software makers developing real-time non-cooperative target recognition packages…

  16. Big D 27 August, 2010 at 1:46 am #

    It only looks circular, it’s actually a spiral.

    A combination of advanced technology, shifting requirements, and some gold-plating drives up the fixed costs (and to a far lesser extent, the marginal costs) of production. This makes the total program cost and average cost per plane go up, and those seem to be the only numbers that Congress ever looks at.

    Congress reacts to their (gasp, unexpected!) rise by cutting the number produced, usually together with moving the schedule to the right. But, for some (unexpected!) reason, total program costs go up (just like with the space shuttle, moving the schedule out means paying *all* of the workers more salary regardless of output, and it gives more time for requirements to change and gold-plating to occur). Worse yet, the average cost per plane skyrockets (increase the numerator, decrease the denominator)! Clearly, something is broken in the project, so Congress cuts the buy again, and shifts the schedule to the right some more.

    Finally, after several iterations of this, Congress angrily cancels the program and yells at everybody involved.

    Of course, that’s not to excuse what’s happened with the F-35, which wasn’t going to meet its original cost and timeline estimates to begin with, because the requirements and the manufacturer’s claimed capabilities were out of whack with reality from the very beginning.

    Actually, the tech industry makes an excellent counterpoint to the defense industry. Tech companies make multi-billion-dollar investments all the time in technologies and facilities, and then use them to crank out *massive* numbers of units to pay off those billions in R&D/plant costs. There are billions (trillions?) of microprocessors out there today. There are <100,000 4th-generation fighters on the planet, and if you count only 4.5/5th-generation fighters that have the stealth, sensors, and/or C4I capabilities being put into today's requirements documents, that number drops below 1,000. It's apples to oranges; that said, by comparing them, I do think you can see where the orange began to turn into a lemon.

  17. SMSgt Mac 27 August, 2010 at 3:56 am #

    Let’s keep the cause-effect relationship straight.
    One could not have picked clearer cases of how the technology involved had nothing to do with the ‘escalating’ unit and program costs, and cost drivers than the F-22 and the B-2. It is pure ‘Acquisition 101’.

    First: the F-22. When all the guilty are in the political graveyard, I guarantee it will eventually come out that upwards of 70-75% of the cost growth was due to Congress first dragging out the development phase, and then to keep total costs down cut the quantity buy (which raises unit cost) and then repeating the cycle every time the program had to overcome a technical challenge. After you take out the side-effects of later year cost-of-money and compounding inflation in the interim, I would be surprised if the ‘technology’ challenge drove much more than 10% of the cost rise. I’m still kicking myself over not downloading some cost data that Lockheed failed to secure properly on a public server years back, so somebody has just ‘gotta’ still have the data somewhere. The truth WILL out.

    Second: the B-2. The B-2 was purely a victim of the end of the Cold War ending and the misperception that their utility was only found in the SIOP mission. I blame SAC for failing to recognize the strategic value of precision conventional attack: something they should have known about at least by the end of the 1970′s. Their myopia constituted an abdication of the AF to the short-legged fighter ‘Airpower Dwarfs’ that rose to eventually command it, and to say stupid things like: “there is no difference between strategic and tactical“. Read Carl Builder’s ‘Icarus Syndrome’ for the whole ugly story among others. It’s made AF leadership reading lists, but I think they’ve mostly used it for a doorstop. Cutting the B-2 buy based upon the reduced need for SIOP-dedicated aircraft was simply a case of amortizing the fixed development costs over a fraction of the units the system was built to produce. The proof is that years later Northrop offered a B-2C variant for under $700M each. People forget, that unlike previous systems, the B-2 had to be built with production tooling because of the precision involved, and all those costs were sunk before the first aircraft came off the line.

    I’ll let the other WTFO assertions in this article/thread go for now. They pale in comparison to the quantity/stretching-development as cost drivers.

    P.S. Now some earnest ‘deep’ thinker may be tempted to retort with a claim that stretching development to solve tech challenges is unavoidable: that the weapon system MUST be fully mature before it is produced, and that all the bugs must be wrung out before it is fielded. And they would be wrong. There is a reason that there were about 90 B-52A’s B’s and C’s were built before the first definitive D model rolled off the line. Until the politicization of ‘Defense’ in the Vietnam era, it was understood that it was more cost effective to improve aircraft in production with feedback from the field than to make the entire industrial base sit idle and burn mega-$ waiting for the next engineering release.

  18. DJ 27 August, 2010 at 4:35 am #

    The best article explaining why military technology isn’t getting cheaper:

  19. SMSgt Mac 9 September, 2010 at 1:31 pm #

    RE: “The best article explaining why military technology isn’t getting cheaper.”

    Luttwak’s article at the link was possibly the most painfully disjointed, rambling, and void of facts or logic screed written by an alleged ‘expert’ that I’ve ever read. Methinks that by whatever measure Dr Luttwak can be considered a National Security Policy expert, he can be considered the matmatical inverse on the topics of the practical application thereof and the evolution. history, and drivers of defense technology development.

  20. Mozell Palaspas 4 October, 2010 at 7:29 pm #

    I like your post outlining the current issues with an ever volatile real estate market and the associated interest rates. Even with the lowest conventional (30 and 15 year) rates in history, we are seeing customers interested in Plano real estate rates are moving forward with caution.

Leave a Reply