This post discusses the factors that make buying military transformation so difficult.

Why doesn’t the military get with the program? Why does the government seem so reluctant to buy those shiny new Gen-5 planes and ships? These are some of the questions I had when I was a snot-nosed undergrad aspiring to solve all of the country’s defence problems. I thought we could buy our way out of being a big country with a meagre defence force. I thought the government was stupid because it didn’t (or was too slow) to embrace the technological marvels that were on sale. The truth is, defence acquisition may be one of the most convoluted processes within the government, aside from implementing neoliberal macroeconomic policies.

The convoluted acquisition process

The first thing that’s helpful to understand is the weapons acquisition process. For starters, it’s not as simple as buying the new iPhone. The process of buying a new iPhone often goes like this:

  1. Wait for Apple to release the iPhone X.
  2. Beg mom and dad for money.
  3. Line up at the Apple Store on the release date with other like-minded hipsters.
  4. Get the new iPhone.
  5. Tinker around with the new features.
  6. Get bored of it after 3 months and repeat step 1.

This particular model works in real life because as a commercial company, Apple has a potential 7 billion customers to market to. Furthermore, it is driven by competition from other peer competitors such as Samsung and HTC. Apple derives its profits from sales or perhaps other deals, which it can funnel into research and development to make the next iPhone to market to people. This is the typical commercial company scenario.

In defence acquisition, things are often more complicated. Mary Kaldor [paywall] outlines the defence acquisition process as the following:


The invention of a weapons system is often taken by defence labs, universities, government projects, or even defence industries. At this point, the weapons system may still be a prototype. For the sake of this post, let’s say it’s called the War Machine. The innovation process involves developing the War Machine into an actual product that meets the doctrinal and technical needs of the military. The integration process happens when the military actually starts buying the War Machine and begins fitting it into operational concepts. Obsolescence happens when the War Machine reaches the end of its lifespan or is overshadowed by the newer and shinier War Machine Mk. II.

Now, while it may be tempting to think of the process of buying the War Machine as being similar to that of the iPhone, we need to rethink the typical scenario. Within a military setting, there’s only one buyer — the government, and by extension, the military — and many sellers — the defence industry. This system is called a monopsony. Furthermore, defence industries are typically divided into two broad categories: those who have close ties to the government and those who don’t. What you would have then are coalitions of industries that have close ties to the government and military — the military-industrial complex.

We need then to understand the nature of this relationship. The government only has limited funds — the defence budget, which is often limited as a percentage of GDP. The defence industries have an interest in maximising profit and gaining an edge over their peer competitors but can only do so within the limits of which the government is willing to spend. The defence industries also have a very fussy customer, which is the military, who are often really conservative about their demands. The military wants new stuff, but not “too new” in a way that it may disrupt its service culture and organisational setup.

General conservatism

A baroque cathedral. All bling, no bang. 

This leads us to what Mary Kaldor terms as the baroque arsenal, which stems from this convoluted relationship.

To perpetuate development and production, defence industries need to get orders once the development process is completed. To get orders, they need to persuade the government to give them “follow-on orders”, which usually involve competing against other industries. To compete, a defence industry would need to convince the government that its product has the best things money can buy (as Kaldor writes, the basis of cost of a weapons system is “cost-plus markup”, hence if industries sold cheaply, it would ruin the market). The government awards the order to a defence industry, often suffering cost overruns due to the amount of “innovation”. The result is the shiny new War Machine.

However, militaries usually only want weapons that fit their “dominant scenario”. Imagine an Air Force commander. Their past wartime experience is only good for that particular war — the next war may not be the same. Due to uncertainty, militaries tend to stick to tested ways of fighting. It’s reasonable to do so, since changing proven operational concepts or strategies may result in losing a war. So, if the Air Force’s dominant scenario (accumulated from years of experience in real warfighting) requires speedy fighter jets to interdict the enemy, defence industries, no matter how hard they innovate, will never be able to sell the Air Force that new War Machine if it does not fit their “dominant scenario”. They might opt for an F-35 with advanced radar and stealth since it matches what they’ve been trained to do for years.

But, since there’s pressure on the defence industries to deliver something new all the time and the government needs to keep these industries afloat lest they fail during a time of real crisis, this perpetuates the “baroque arsenal” problem: militaries have increasingly advanced equipment that they’ll never use.

(Robert Farley also wrote about five supposedly revolutionary weapons systems that never happened.)

Why are weapons so expensive?

One person that can help shed light on this is Edward Luttwak. In an article titled Breaking the Bank, Luttwak argued that “the U.S. military needs to stop cramming new technologies into old platforms.” While Kaldor focuses on the relationship between the military industrial complex, Luttwak focuses more on the (1) negative economies of scale due to less production and more customisation and (2) how equipment is used and deployed. He argues that the U.S. military needs to change the way it uses its weapons.

Luttwak brings forth Grosch’s Law, which can be understood as the following:

When the price of a processor doubles, you would expect a squared increase in speed.

In other words, if a processor costs twice as much, you should expect it to have 4x the processing power. This is the case with most network-centric weapons these days. Since they need so much processing power to be able to function within a network-centric setting, the costs of making these processors go up, which affect the end price.

Negative economies of scale

This is not an advertisement for the ROG G752VM


Luttwak’s first argument reminds me of the “gaming laptop” problem. If you’ve noticed, gaming laptops have become increasingly trendy right now despite their somewhat ridiculous prices (ranging anywhere from SGD 1,000 to SGD 4,000). Other “gaming” peripherals are also becoming more popular, especially ones with edgy features. Now, putting aside the sheer processing and graphics power of these beasts, what makes them so expensive?

Tuan Nguyen, in his article about “how gaming hardware needs to grow up“, delivers some salient points.

The more effective heatsinks on server motherboards—that need to handle more demanding environments—costs less to manufacture than the (un)cooler heatsink block of metal seen on virtually all gaming motherboards. This is because the tooling and molds needed to produce more complicated designs also cost more to make.

…gaming means having good graphics and CPU performance to get the best experience out of your games. Gaming means decreased load times and higher clock frequencies. Gaming means optimizing a system to speed up texture loads, and using high refresh-rate displays. It means having low latency and no network lag. It does not mean adding more junk onto PC hardware or making uncomfortable design choices in the name of—ugly—aesthetics.

What the heck does this have to do with weapons? Let’s relate gaming hardware to weapons systems.

As Nguyen rightfully pointed out, customisation adds a lot to the final cost of a product. Since gaming hardware requires special moulds to bring out their edginess, the cost of engineering and developing those moulds adds into the final product cost. Also, you can’t mass-produce these customised moulds since you only have a handful of buyers (rich gamers). If you focus too much on this segment, you end up alienating the rest of the 99% that just want normal laptops.

Luttwak points out the same problem with weapons systems. As militaries want more and more customisation, weapons systems can no longer be mass-produced as they did with muskets and rifles of World War II. More customisation equals more money and research and development. Adding insult to injury, the government can only purchase within the limits of the budget. Industries are also limited in their production by the demand of the government and military. The U.S. military can’t afford, nor even effectively use 10,000 F-35s. Since weapons are more expensive, the government can only afford less, which negatively influences production. This causes a negative economy of scale.

This is further exacerbated by the “vicious cycle” of procurement, which goes like this. This basically reaches a point where we end up with Kaldor’s “baroque arsenal” problem: we have fancy weapons that may be useless.

vicious procurement.png

21st-century technology in 20th-century configurations

Luttwak’s most compelling argument for why weapons are so expensive is the presence of “hidden costs inflicted by the astounding persistence of traditional weapon configurations.” He argues that instead of designing our weapons around technology, we design technology around our weapons.

As an example, let’s revisit the gaming laptop problem. Usually, gamers prefer gaming desktops that are huge and bulky. These designs allow for maximum processing power and graphics, although they do take up a lot of space. This is a typical hardcore gamer setup:

maxresdefault (4).jpg

When we try to pack all that power into a laptop, we face problems. There’s the issue of space. You need maximum processing power, but a processor can only be so small before it starts losing power. The same goes for the graphics card. Desktop graphics cards are larger; you need to miniaturise them (at the cost of performance) for laptops. The screen of a laptop can only reach up to 17 inches. We also need to cram in space for the hard drive and cooling systems. Often, a gaming laptop will pale in comparison with a proper desktop setup simply because we’re trying to cram too much power into such a restrictive space, but it may cost twice as much simply because we’re trying to miniaturise everything and expecting it to perform as well as a desktop.

Now, imagine an advanced fighter. A good piece of airborne radar (Luttwak uses the Active Electronically Scanned Arrays as an example) may cost thousands of dollars, but it costs even more to make them small enough to fit into the nose of a fighter jet. And we need to fit a hundred fighters with these, so the cost goes up even more. When we miniaturise radar, we also lose performance.

Luttwak’s answer to this issue is simple: if we want to fight networked, think networked. Instead of fitting every plane with a radar that works half as effective as a normal radar but is ten times the price, consider “specialisation” within a swarm setting: one plane made exclusively to house radar, while other planes are specialised for firing missiles. It reminds me of the Formics in Ender’s Game. The Queen issues out orders while the drones handle the shooting. That way, we can use radar effectively without spending too much money trying to cram it into the nose of an F-35.

However, radical solutions like Luttwak’s often have to face institutional inertia, which draws back to perhaps the most important impediment to military transformation: military conservatism.


As we’ve seen, there are many explanations as to why militaries are reluctant to adopt transformation. We’ve seen that military equipment can cost a lot of money and that innovation is generally not welcome. Thus, we arrive at one major impediment of military transformation: conservatism. Now, being conservative is not necessarily bad. Playing by the book, especially relying on proven tactics and strategies, helps save lives in battle. Opting to be creative may lose lives and the war. Since militaries have a huge burden on their shoulders to keep the nation safe, they can’t afford to experiment with untested, “revolutionary” technologies.

Another impediment is money (or lack thereof). As we’ve seen, designing revolutionary technologies often results in lower profits for the defence industry, an increase in prices for the government, and it may not be accepted by the military. As such, military transformation is often a rich man’s game and only a few countries can afford it.

Perhaps the way to transformation is having enough money and a military that’s willing to try new technologies.