How Procurement Destroys Armies - Requirements, Risks & Development gone wrong
In war, as in many other fields, it often pays to be better prepared and supplied than your potential competitors. However motivated troops might be, ultimately they probably need stuff to accomplish their mission. Without ammunition, weapons, communication, personal protective equipment, and all the panoply of war, even the finest troops in the world are going to be at a major disadvantage. Through a lot of human history, lords and governments simply required troops to bring their own equipment. But given that there aren't many recruits out there with the financial capacity to bring their own fighter jet, these days the task of equipping militaries largely falls to government procurements.
Which does raise the question: if the task of equipping troops is so important, and has been for centuries, why does it sometimes seem like governments aren't particularly good at it? Why does it seem we're constantly seeing stories of defence projects being delayed? Why has the UK spent billions of pounds and many years on a program that may not deliver a fully-operational infantry fighting vehicle until nearly the end of this decade? How does the German Navy end up with ships beset with so many technical issues, they had to send them back to the manufacturer? And how did so many significant and important projects end up floundering or cancelled? To the point where even the mighty United States Army is relying on what are essentially modernised Vietnam-era vehicle designs to provide its mobile self-propelled guns. It's a complex question, but one which probably starts with an understanding of how weapon systems are designed and procured in the first place. So today I'm going to talk about everyone's favourite topic: procurement. So we'll start with a little bit of background about procurement, and then work our way through the various phases of it and how each of them can go terribly wrong. How you can mess up the requirements phase, how contracting and program advancement can go wrong.
And then how issues like delays, scope, creep, or technical risk can scupper programs even once the contracts are signed and you are well on your way. All this will lead into a quick discussion of the "death spiral", which is a situation where a program has been so delayed or gone so over budget that it often has to start cannibalising itself just to stay alive. Shedding scope and quantity until it has potentially a major impact on the overall force. Then what we'll do is put it all together with a worked hypothetical example. I'll be using a lot of historical and hypothetical examples in this video for hopefully obvious reasons. But I do want to stress that when national examples are used, they will inevitably be simplified and shouldn't be taken to suggest that everything that country does in the procurement space goes wrong.
Or that other countries can't make mistakes as well. Finally, having possibly depressed everyone with stories of taxpayer funds disappearing into the void, I'll explore some suggestions as to how procurement processes around the world might be improved to limit the number of occasions where taxpayer funds are set on fire for little result other than a mild increase in temperature in the surrounding area. But before I do any of that, a quick word from a sponsor.
And today I'm welcoming back returning sponsor, Blinkist. This is an app which takes an impressive collection of more than 5,500 non-fiction books and podcasts and gives you a way to understand the most important things in those titles in just 15 minutes. The works cover topics from economics, to well-being, to politics, to this classic work I've recommended before. A 19th century work called On War by Prussian General Carl von Clausewitz, whose lessons, though old at this point, still maintain pride of place at many military academies around the world.
To Clausewitz, war took place through a dense fog with three quarters, he said, of all those things upon which action in war must be calculated being hidden more or less in clouds of great uncertainty. And he says "more or less" there because even the degree of uncertainty involved is itself uncertain. I feel those observations of that theory are particularly pertinent to the war in Ukraine at the moment and all the commentary going on around it. All view the war through sometimes contradictory, partly shrouded, flows of information.
And so there is a desperate need for everyone to understand the limits of their own picture, and to look for ways to reduce that uncertainty. And if you think friends or family might benefit from this or other lessons in Blinkist's titles, there's now a new feature called Blinkist Spaces. This lets you, friends or family create Spaces where you can add, share and recommend titles across the Blinkist library all in one place.
The key here is that all members of a shared Space can access all titles in that Space whether or not they have a Blinkist Premium subscription. So if you're interested, you can start your 7 day free trial and get 25% off Blinkist Annual Premium by clicking on the link below in the description. Alright, so let's start with some basics: what is procurement and at a high level why can it go wrong? If you've watched some of my videos on defence strategies, you probably know where procuring a new piece of equipment might fit in the overall national security scheme. Buying new stuff ideally happens after you've done a scan of your strategic environment, decided what your national objectives are.
Then I ask myself the question, what are the things I need in order to be able to have that capability? Now a lot of things go into capability, different countries think about this in different ways. But I'm Australian, so I get to push my system on you so we get to talk about getting your FICs. FICs are Fundamental Inputs to Capability, they're like all the things you need to put together in order to have a capability.
I've listed those on screen, but hopefully they are relatively intuitive. It's not just about having a weapon system. If you buy yourself an anti-submarine warfare helicopter, but you don't have the pilots to fly them, an organisation to fit them into, facilities like airfields to base them at, an ability to maintain them, etc. etc. then you don't have an ASW capability, you have a state welfare program for the aerospace sector. For the purpose of this video we are mostly talking about one dot point on that list, the major system.
We are going to be aware of the fact that in the background the military is going to have to be training up personnel and other units, developing organisation, command and management structures, but the bit we're going to be concerned with is government going out and buying them the equipment to actually operate. Because just like a helicopter with no pilot is pretty useless, a pilot with no helicopter is likewise missing a little bit when comes to overall utility. The second thing you should know about defence procurement is that there is a trend overall in industry pricing on major procurements, and that is that in general prices only go in one direction - up. And it's hard to find a time in history where this phenomenon wasn't being complained about. Back in the 1980s Norman Augustine, who was both a defence industrialist and briefly Under-Secretary of the Army from 1975 to 1977 in the US, joked that by 2054 the entire US defence budget would only be sufficient to pay for one airplane. Which would have to be shared between the United States Air Force and the United States Navy, with the Marines getting it occasionally.
This phenomenon is an incredible challenge for many governments, because when costs go up one of the first responses a government might have is to reduce quantities. Which diminishes economies of scale, which drives increased cost. We'll talk a little bit more about that death spiral later on. To illustrate this point what you're seeing on screen is a graph I've borrowed, showing the US dollar cost per a ton of various submarines from the early 1900s through to I believe 2009. At first it looks like a steady upward trend, then you look at the Y-axis and realise that's a logarithmic scale, and those Y-axis lines go 1, 10, 100, 1000.
And before you go thinking this is just a result of greedy Western corporations fleecing the taxpayer, notice that the single most expensive dot on that graph is an enhanced version of the Kilo-class submarine, which is a Soviet/Russian design. The biggest single driver of this is probably the fact that a modern submarine packs a lot more advanced technology and capability into every ton than a World War Two-era one did. Equipment is becoming more capable and complex, and as a result more expensive. But that's also probably not a complete explanation. Nor is it probably much comfort to government treasuries around the world who have to deal with a seemingly inevitable reality: stuff seems to be getting more expensive basically no matter where you live.
Now, it's probably impossible to point to just one reason that projects fail or increase in price over time. And I'm certainly not going to try my luck in terms of covering all of them in one episode. Three that immediately come to mind are firstly, all the mistakes, errors and drawbacks that can happen in the requirement, procurement, contracting process.
That's where I'm going to be focusing our attention today. There are also other factors like misaligned incentives, special interests, failures of competition, that might go some way towards explaining why the structure of the market, or the procurement systems, or the political structures around them don't always align in the way taxpayers might expect. Finally, there's what Australians might call "the dodgy stuff". This is what happens in some countries when what you might argue is basically corruption becomes legal. This is where people with the power of position and knowledge of the system are able to redirect funds from the pursuit of effective and affordable military equipment towards other critical national defence priorities, like mansions and yachts. As I said, we're going to focus on errors in the procurement process itself today.
But if this video goes well, maybe the rest of this potential trilogy will explore some of those other issues. OK, so if you want to understand at least a small sample of things that can go wrong with procurement, it helps to understand the broad process. And here as normal I'm going to commit some doctrinal heresy to simplify things down for YouTube. For the purpose of today think of procurement, the process of getting a thing and putting it into service, as falling into just a few key steps.
The first is developing a requirement. What are you actually going to ask industry to make for you or sell to you? Then there's the process of bidding or contracting. That is how are you going to decide who's going to produce it for you, and on what terms are they going to do it? Then there's the process of developing, fielding and sustaining. So getting the thing designed and refined, put into production, and actually in service with your military, and then maintained in service once it's there. If you are trying to equate this to something members of a previous generation might do over the course of their lives, then think of it maybe like designing a home, picking a builder to construct it for you, signing the contract, getting the home built. And then organising regular maintenance to make sure it all remains liveable.
Meanwhile for members of the millennial generation onwards, home building is basically a process by which landlords construct the properties that you rent from them. Now if everything goes well in a project the final stage is just continuing that sustainment until eventually the product is upgraded or replaced. But let's not get ahead of ourselves, and instead start with the requirements phase.
Because if you ask for the wrong thing up front, the best defence industry in the world may not be able to save you from yourself. In simplified terms, the requirement is where you go out to industry and say, "I would like to buy X." Noting that for the purpose of this video, I am going to be merging the terms "requirements" and "specifications" and using them interchangeably. And an awful lot of military equipment is designed in response to potential or actual requirements. Back in the old days industry often did private initiative developments where they would design a tank or an airplane or something basically off their own bat, using their own money, and then go around the world trying to sell it.
That isn't really as common any more, although it does still happen. But because of how expensive and capital intensive most modern military equipment is, most companies won't develop a new system unless the government is either supporting the development process, or is potentially going to buy it. And when governments set out what they want, usually those requirements in terms of performance characteristics or technical requirements tend to err on the side of being very, very detailed. Government wouldn't really issue a requirement for, "A tank with a big gun that does tank stuff good," for example, and then wait to see what bids come in. And because at the end of the day the buyer is the one with the money, producers have an awful lot of incentive to build what they are asking for. Whether they necessarily think that that is a good idea for a requirement or not.
In saying that there is a phrase that comes to mind, "The customer is always right." And I don't really care what field you are in, retail, law, accounting, engineering, or whether you're a barber or hairdresser who has just had a client walk in and ask for the most hideous haircut possible, you all know the basic fundamental truth. It is very, very possible for the customer to be incredibly, terribly, catastrophically wrong. It's true in IT, it's true in business, and it's very much potentially true in military procurement. Get the requirement wrong and your project may be doomed from the start. Because it may be very, very possible for the engineers to build what you have asked for, but the thing that you asked for be a terrible mismatch with what you actually need.
There are many ways this might happen: your strategic concept may be flawed, your capability need evaluation may be wrong, or it may be a mistake specific to the requirement itself. But if you go down this route you may be in trouble, although an honest contract or a good risk mitigation policy might save you. In the 1950s for example, the US discovered that the Soviet Union was producing large quantities of liquid hydrogen, and concluded (because in America everything is about the Air Force) that all this liquid hydrogen was intended to power a new generation of hydrogen-powered aircraft. So they gave the Lockheed Martin Skunk Works the job of developing a hydrogen-powered reconnaissance aircraft, the CL-400 Suntan. It didn't take a huge amount of development work for Lockheed Martin to figure out that the concept was very difficult to work with, incredibly dangerous, and would require massive overhauls on airfield infrastructure. It also would have notably increased the possibility of base infrastructure spontaneously exploding in spectacular fashion, which is also a drawback.
And so the famous Kelly Johnson basically went back to the government, said that he was convinced that they were "building them a dog," and returned the development funding with some of it unspent. But the more tragic versions of this story tend to occur when a system is fielded despite the requirement being dubious. In the inter-war years for example, the French government put out a lot of requirements and spent a lot of money developing so-called multi-role aircraft. The Amiot 143 you see on screen for example, was meant to be a combination bomber, reconnaissance plane and bomber escort. What they got was an aircraft that was extremely vulnerable by the time World War Two rolled around, and which was inevitably slaughtered every time it didn't fly by night. For their part, the Germans didn't have a long-range heavy strategic bomber when the war broke out.
But they did have a program that might potentially fill that role, the Heinkel 177 Greif. In the end however, they ended up getting a terrible aircraft for a variety of reasons. One of which was the strange German insistence that the long-range strategic heavy bomber be capable of dive bombing. That meant strengthening the airframe to enable dive bombing, and also pushed the design towards a two nacelle system.
Where instead of having four engines mounted along the wings you would mount two engines in each of two nacelles. That would reduce drag and increase diving performance, but also led to minor drawbacks like massive mechanical complexity, unreliability, and the fact that the plane's engines periodically caught on fire spontaneously. We shouldn't let the British off the hook either while we're making examples. The specification that led to the Short Sterling heavy bomber in the inter-war years specified a hard limit for the maximum wingspan of the bomber as 100 feet. The downside to that decision was the aircraft had a lower service ceiling, making it more vulnerable to flak and fighters, and it was also more limited in the sort of bombs that was able to carry.
And to be fair it's entirely possible that the designing engineers knew that too. But in the end designers build to a specification, and specification B.12/36 said 100 feet. The exact reason for the requirement is a matter of historical debate.
With one old theory being that it was designed so that the Sterling could fit into existing hangar infrastructure. I'm personally dubious of that explanation, but it does make a valid point. You can absolutely imagine a government writing a specification that saved money on basing infrastructure, apparently on the assumption that that was much cheaper than replacing aircraft and their trained air crews. And then there are plenty of examples where frankly your guess is as good as mine. For example, in World War Two the Japanese put a bayonet lug for a full-length bayonet on one of their light machine guns. Because nothing aligns with the fire support role quite like adding a component which is only useful if you get within a couple of metres of the enemy.
The key point is that a bad requirement up front can result in you producing a sub-par or even terrible weapon system almost no matter how the development process goes, unless there's a massive change in direction. For example, I'm relatively confident concluding that the Russian Burevestnik program is going to produce a failure. That is not because I have access to some super-secret Russian military secrets, or information, or intelligence. Instead it's just because of what Burevestnik is, it's a program to develop a nuclear-powered nuclear cruise missile. Which I'd argue is yet another entry on the long list of expensive Russian doomsday weapons that have no appreciable strategic or operational utility.
Other existing Russian weapon systems fulfil the nuclear deterrent role, and fill it arguably better than this system ever could. And so the entire requirement is militarily an unnecessary waste. I'm sure the Russian military is very glad that there are no other pressing priorities they could have spent the development funding for this program on instead. Jokes aside, your requirement is also going to help shape the sort of technology that your final product is likely to use and the risks associated.
If you ask the market for an artillery cannon that can shoot a high explosive shell 30 kilometres, they are probably going to use existing, proven, known technology. Ask them for one that can shoot 2,000 kilometres, and there's going to have to be some new technologies tested and developed as part of the program. Tech development is a major sort of risk because it doesn't always pan out the way you're planning.
The real world is not a strategy game where you can look at all the technologies in the tech tree and know that if you put exactly 600 science points into that thing you unlock a new unit with a certain performance characteristic. In the real world new technologies and new developments always come with risks, that they may not live up to their promise, or they may take longer to develop, or more cost than you originally planned for. So in an ideal world, you'd want to be sure that the new performance of new technologies or options justifies the risk and expense involved.
Which brings us to the discussion of gold-plated requirements. Gold plating is what happens when governments ask for so much capability in a given requirement that it borders on massive overkill. Often this comes from a very understandable place, militaries wanting their troops to be equipped with "the very best", to enjoy some sort of massive qualitative advantage that's going to give their troops a decisive advantage over potential opponents. The problem of course is that troops may deserve the very best, but they'd prefer to have something at all in general, and most governments don't have the funding available to produce enough of the very best for everyone. And to be clear, producing more technologically advanced qualitatively superior systems is a massive advantage in general. You can get situations of technological overmatch where a small number of F-22 Raptors can seal club their way through almost any number of 1970s or 1980s-era jets.
But there is a trade-off involved with every additional bit of capability or weight you choose to add to a system. And there comes a point where you may not be getting value for money anymore. Back in 1984 Norman Augustine, who we talked about before, also said that the last 10% of capability drives about a third of cost and two-thirds of problems. Now those figures are purely illustrative, but I think there's a point there. And it doesn't just apply to military procurements, it applies to almost any engineering or service field. Building a car that can do 200-250 kilometres an hour is cheap and easy, some common passenger vehicles are able to do it.
Try and build a car that can carry the same number of passengers or the same amount of cargo that can do 350 kilometres per hour, and you're going to add several zeros potentially to the price. Try and hire a pretty good musician to play live music at your event, it might cost you a couple of hundred bucks and a crew meal. Try and hire Taylor Swift and you might bankrupt the national economy. And the problem is a lot of militaries tend to want the Taylor Swift version, as opposed to the Robo and his guitar rendition. In the past I've used German examples for this phenomenon, but an Indian example probably fits here as well.
Now to be clear I have a lot of good things to say about the Indian military, I'm enjoying working on an upcoming episode on India. But you could argue that India's Project 75 submarines are a good example of gold plating in action. P-75I is a plan for a series of diesel-electric submarines for the Indian Navy that dates back to the 1990s in its origin.
There have been several attempts to move the project forward over the years, and as many false starts. In 2021 there were five foreign providers on the short-list, Germany, Russia, Spain, France and South Korea. If you are looking to have a good, healthy, competitive procurement then having five in the race is a very good thing.
However, the Indian Navy wasn't just asking for any submarine, it was asking for a very, very particular type of submarine. They wanted state-of-the-art counter-measures, state-of-the-art sensors and weapon systems, and an Air-Independent Propulsion system that had been extensively tested on an operational submarine. And they wanted it all on a short time-frame. The problem was a majority of those potential foreign providers didn't have an example of an AIP system in active service on an operational submarine that was already proven. They probably could have bearded a design with an AIP system, but the requirement wasn't for a submarine with an air-independent propulsion system. It was for one that was already operationally proven.
And so for that (and for other reasons) France, Spain and Russia all dropped out of the competition. And the story here is about more than just the AIP requirement, which may have made sense on its own, it's about an overarching tendency for militaries to sometimes ask for the proverbial moon when they are putting out requirements. Staying on India for example, in 2015 the now late Defence Minister Manohar Parrikar, described the requirements for systems for the Indian military to be straight out of "Marvel comic books". And suggested that the technology demanded was absurd and unrealistic.
So let's take that Marvel test and apply it back to P-75I. The submarines weren't meant to just have a proven AIP system, they were meant to have large-scale land attack capability, be fitted with modern missiles, sensors, and state-of-the-art counter-measures. In response to some of these requirements, the Rubin Design Bureau from Russia went public stating that the project was "unrealistic", that "the requirements specified by the Indian Navy and the timeline for the project do not match."
And that no one in the world has such a submarine ready. And the reason that last line is important was because of the short timelines proposed by the requirement. As long as the money is available, you can ask the defence sector to build you Marvel scale quality if you want. But only if you give them the time and budget to do so. Surprisingly enough, no one has a Helicarrier schematic just waiting ready to go. Gold plating can also manifest in the small stuff too.
One example being in unnecessary standards, or unnecessarily tight tolerances. Very roughly you can think of this as the quality level you are asking your manufacturer to provide. Is the component going to require service or replacement after 100 hours? 1,000 hours? 5,000 hours? And before you say, "Well, I just want a high quality component that's going to last until the heat death of the universe," remember that there are costs associated with that. Getting a higher manufacturing grade can itself be expensive.
It can mean you can't use an off-the-shelf component and have to use a proprietary military-grade one from one particular manufacturer instead for example. And unsurprisingly, they are probably going to charge more than your local Bunnings. It can also reduce competition by knocking out potential providers. If there are a certain number of companies that can manufacture to standard A, you tighten that standard and you're going to start knocking out potential bidders. If you want a historical example, you can go back to the American B-47 Stratojet where many of the original subcontractors were, as was relatively common back in the day, automotive manufacturers.
And that may have been fine building aircraft that were relatively simple with wide tolerances during the Second World War. But the new stricter tolerances required to build the B-47 started to make it more and more difficult for those subcontractors to bid into the project. At the harshest end of this spectrum there's a relationship between how long you want something to last and how long it's likely to last in battle. The Germans during World War Two for example famously over-engineered a lot of their equipment. But it's massively wasteful to design something to last 4,000 hours if statistically speaking, it's going to be lost after a fraction of that once it goes to the front line.
If I am building a kamikaze drone for example, I don't need 500 hours of engine service life. A final form of gold plating can be particularly applicable when you are talking about foreign arms sales. Specifically where militaries decide that they don't want something that's available on the market as is, which is already tested, with a solid supply chain and limited technical risk. No, they want that thing, but they can't resist the urge to just tinker a little bit. Maybe it's about all the interesting ways they think that they can improve the original design.
Adding new technologies, swapping out weapon systems, protection packages, the powertrain, a dozen other options besides. And to be fair, militaries do have reasons to differentiate their requirements. If you are a primarily desert country and you are purchasing equipment designed for Western Europe, it makes sense to adjust the original design to be more suitable for desert operation for example. But deep care needs to be taken when you deviate from an existing design, because the moment you do that you are re-introducing risk.
You're taking something that is already proven and already mature you would hope, and turning it back into a research, development, testing and fielding project. That may turn out fine, or it may all go horribly wrong. If you want an example of this in action, you can look at the Ajax project in the United Kingdom.
Ajax is at this point massively behind schedule. But the core vehicle is based on an existing mature platform, namely the Austro-Spanish ASCOD. And in trying to integrate a range of changes and upgrades to the original vehicle, Ajax has found itself, instead of being a mature system, squarely in development hell. The noise and vibrations inside some of the test vehicles became so bad that crews had to be limited to spending something like no more than 100 minutes inside the vehicle at a time, and not driving it faster than 32 kilometres an hour.
(That's 20 miles per hour in freedom units.) Which notably is less than half the maximum speed that can be attained by the base ASCOD vehicle. Now obviously a lot of the upgrades did have value, and if Ajax works as intended it will be a better vehicle than ASCOD. But it doesn't work, and that's kind of the point.
And so sometimes the need for speed or reduced risk pushes nations towards existing platforms. After a couple of missteps in military ship production by the United States Navy for example, the US ultimately opted for the Constellation class for their next-generation guided-missile frigates. And if the silhouette of the ship happens to look familiar, that's not a coincidence. Because rather than develop something from scratch, the Navy is going with an adapted Franco-Italian design, namely the somewhat famous Fregata Europea Multi-Missione, or FREMM. To be built in Wisconsin by a shipyard owned by Fincantieri. But let's just say that we've got the requirement right and industry is going to come back with some potential products that genuinely meet our real needs.
All that means is you've passed gate 1, and you now get to move on to the next fun component: contracting. Because competition, selection and contracting can be a whole process unto itself. If you want an example look no further than the Canadian acquisition of the F-35 for the Royal Canadian Air Force. There the Conservative government originally announced that they'd buy the F-35 in 2010.
A subsequent Canadian government would say that they would pick from one of the many available other options for the Canadian Air Force, and have a competition to decide the new replacement instead. That competition would be won by the F-35 in 2022. And in 2023 Canada would announce that they had signed a contract to purchase 88 of them.
Now in terms of the contracting process itself, there's a couple of things that set many military procurements apart from other civilian examples. I'll talk about the issue of competition more in the future, but you need to understand that, particularly if you're trying to buy domestically, there are often relatively few firms involved that can produce what it is you need. Those large defence prime contractors dominate the market for the same reasons that you won't find small companies active in manufacturing advanced semiconductors. Namely that it requires a huge amount of capital investment, expensive facilities, trained personnel and massive supply chains to produce the most advanced military hardware.
And that helps go some way towards explaining the ways in which military contracting has changed over the course of history. In 1809 for example, the US Congress provided that basically all the military goods the government was buying should be made by open purchase. Basically the government would go out and say, "We want to buy 100,000 muskets or 50,000 horses."
And anyone who thought they could sell muskets or horses to the government would come forward and say, "Here's what I'm offering and how much I'm charging." By 1947, as products are getting much, much, much more complicated than muskets and horses, you still will have to advertise in the American system, but there are now exceptions that allow different methods of contract negotiation. By the late 1980s the vast majority of procurements were being negotiated reportedly under one of those exceptions. So generally speaking you're talking about complicated contracts between a few powerful actors: governments and major defence prime contractors. And like any sort of complicated legal process, there's plenty of ways for it to go wrong. For example, you need to decide how you are going to deal with risk, how you are going to minimise it? What are the accountability and reporting measures? But also importantly, who is responsible if something goes wrong, who pays that price? And importantly, it's possible for this to swing too far in either direction.
If the government takes on too much risk it can end up being responsible for things that aren't its fault, and for pouring more and more money into a failing project that has no real potential for success. At the other extreme the requirements can be so onerous on the contractor that no one wants to take the deal. To avoid getting in trouble, let's go with this totally hypothetical example. Let's just say I go out to foreign companies and say, "Hey, I would love you to work with one of my local shipyards to build your submarine design in my country." And they say, "You know what, OK." Then you tell them they are going to be solely accountable for the quality and serviceability of the submarines that are produced in those local shipyards.
At this point they get a little nervous, they haven't worked with those shipyards before, but presumably if they put their own people in there maybe they can put in place some quality control measures, so they nod and they say "Yes." But then you add the caveat that they will have absolutely no control over those shipyards facilities, those are solely under your local control. And so they are responsible for something they have absolutely no control over. Basically it would be like going to a building company and saying, "Hey, I want you to build me a house. But my cousin Ricky the bricklayer needs some work, so you have to bring him on board as the only bricklayer on the job. If you try to fire him or tell him he can't drink on the work site, then I instantly get my money back.
And despite that, if there are any problems with the building, including the bricklaying, you as the builder are ultimately responsible." Under those sort of circumstances you might want to smile, because you've just successfully outsourced all the risk. The downside in that situation of course is that all the global prime contractors are going to look at your incredibly generous offer and say, "Thanks, but no thanks." The terms included in this initial contract can be immensely impactful for how much the program ends up costing the government in the end.
Particularly where they impact things like operations and sustainment that typically account for a lot more of a project's cost than just buying the original platform. You can end up in situations for example, where you buy a system but don't have the right to modify or upgrade it yourself. Or where you accidentally de facto create a monopoly on the sustainment of the system. The US Secretary of the Air Force has been very clear on the fact he thinks a lot of mistakes were made during the contracting and procurement of the F-35. In particular the US government didn't take over ownership of a lot of the essential intellectual property and data that would enable them to do maintenance of the system themselves. That meant your options for sustaining the F-35 were either companies that had absolutely no idea what they were doing, or the original providers.
Otherwise known as the Apple or John Deere approach to product repair and maintenance. Now that approach to contracting is no longer the done thing, and for the Next Generation Air Dominance it's reported that the US government will be purchasing all of that essential intellectual property and data. But it does highlight the importance of getting contracting correct up front. Because as I've said before, once you've signed that contract you've probably locked in a lot of your future costs. Once you've committed to buying a system, you've also indirectly committed to sustaining it. That, or shoulder the opportunity cost of retiring it early and bringing in a different replacement.
A good contract will protect the interests of both involved, strike a good balance, account for future contingencies, and give you decent odds of a smooth ride going forward. But that doesn't mean this process is over or that you're home free. Because what you have at this point is ultimately a plan. And no plan, as they say, survives contact with the enemy. So once you've signed your contract and you move into development and eventual delivery, fielding and sustainment, you should probably get ready for all the new wonderful things that can now go wrong. The first thing that might happen if it was a competition and at least one company lost is, well, you might get sued.
Military procurements can be very expensive company-shaping contracts. The difference between a company winning and losing can be massive for its bottom line and sustainability. So as you can imagine, companies that lose are sometimes a bit trigger happy on the litigation. And indeed if you're in some jurisdictions (cough, Germany) this step can be depressingly common. Now depending on the circumstances, this may actually be the first delivery delay that your program faces, and it may land within days of you announcing the contract award.
Because in some jurisdictions and under certain circumstances courts may decide to freeze or limit the amount of work that can be done while they consider the matter. That however doesn't happen in all cases. When US firm Oshkosh sued the US government for example, for awarding an 8.66 billion dollar contract to produce the JLTV to a competitor, there was no massive court-ordered delay.
And for those who appreciate these things, I'll just note that back in 2015 after Oshkosh won a production contract for the JLTV, a different company sued at that time to contest their award, namely Lockheed Martin. But let's just move on from legal issues, call it a cost of doing business. Maybe we can write it off as respecting deeply held local customs, like suing people in the United States. The real danger now, as in many other industries, is scope creep. Scope creep is what happens where you try and change what is meant to be delivered after you've already started the process of delivering it. Remember during the gold plating stage where there were all those helpful people involved saying, "Hey, wouldn't it be nice if the system in question had this feature or that feature?" Well, this is now that process continuing, only now we are not in the planning and negotiation phase, we are mid-development and delivery.
As I said, just about any industry is vulnerable to this, although a lot of IT examples come to mind. But if you've ever been part of any project or thing where halfway through someone sticks their hand up and says, "Hey, wouldn't it be great if we just ..." and then inserted a suggestion, you've seen scope creep in action. Several studies and pieces have identified scope creep as a major driver for cost blow-outs on certain military projects.
Ships get heavier and heavier, vehicles have more and more components bolted onto them, and aircraft get re-engineered again and again. Every time potentially increasing cost or bringing about delay. This is almost inevitable if you launch into production before you have a finalised product, deciding that you'll fix it later, or change it mid-production. That approach used to be in vogue in certain places. But that idea of concurrency, trying to develop and produce something at the same time has increasingly gone out of fashion. But even if you manage to avoid scope creep there's still going to be a significant array of risks.
Technology may not perform as expected, either individually or once it's integrated into the wider whole. Your supply chain could be disrupted, other costs could increase. Or you could run into potential political and budget risks. If you have a five year program, but you're operating in a system where budgets only get voted one year in advance, and suddenly your new session of Parliament or Congress decides that your project doesn't get as much funding as your budget says you require, then the emergency belt tightening that follows may be a vicious risk to the program's success. In other words, just existing invites risk when it comes to major procurements.
All of which can be made worse by the fact that it's possible to move too quickly and too slowly when it comes to these sort of processes. Charge ahead full speed, damn the torpedoes and just get the thing built, and you may get a new capability fielded cheaply and in record time. Or you may run into unexpected technical issues where the systems or technology don't work, don't integrate properly, and the final product just is unreliable or a complete turkey. The best case scenario here might be you have to do some re-engineering that is more expensive after the thing is already on the production line. The worst case scenario is you end up pushing a system to the war fighter that is dangerous and gets people killed because this is military procurement, you're not building bloody pianos here.
And if the engine catches fire, the gun jams, or the ejection seat doesn't work, then the consequence is not going to be a note played out of key. But if you slow down too much, delay too extensively, you can be almost guaranteed that you're going to face significant cost overruns and potential project cancellation. Delays are common, but they're not free.
Inflation for one is a non-zero factor, so if you were given a certain budget and time passes, that budget may no longer be sufficient to pay current asking prices. Plus there is a cost to keeping industry and contractors idle waiting for an instruction to move on. If you tell a shipyard to be ready to produce a ship starting in 2024 for example, they are not just going to wait until 2024 to start moving, they can't. They are going to start hiring people, ordering in long lead-time items, clearing the space to make sure that there are no other conflicting orders for the dock space in that period. And they're going to go out to all the other providers and subcontractors as well. The firm providing the sensors is going to have to get ready and stand up their subcontractors.
The firm providing the missile systems is going to have to prepare and stand up their subcontractors. All to make sure that the web of inputs is available when it's time to move. And if you're talking about a really major procurement project, you may be talking about hundreds or thousands of potentially small businesses that have invested heavily in getting themselves ready to take part in that project.
Getting people's security clearances, upgrading their IT infrastructure to be more secure, and maybe upskilling to make sure that they can meet all the needs of this large-scale procurement. So when a delay happens, everyone suddenly can have a problem. Because you can't just fire all the staff that would have been working on this project hoping you'll be able to hire them back again later. You probably can't just undo all those investments, send back all the long lead time items, and then just wait for the actual go order this time. Instead some capacity is going to be idle, firms are going to start taking losses.
And at the extreme, if the project is ever cancelled, some of those firms might take quite extreme losses. So while a delay might give you more time to refine the product for example, it certainly isn't going to be free. Before we move on to the final stage of a project going wrong, after all this talk of scope and delay I wanted to dwell on one point that I think is often missed in coverage of defence contracts. And that is that not every cost increase in defence that you see reported is going to be the result of someone screwing up on a project or a company price gouging. Because if there's one thing I hope has come through here is there's a lot of factors that can drive cost increases or delays, with delays in turn often causing cost increases. Some may be the fault of those managing the project or attempting to deliver it, but others are the result of different decisions made by different people, or they may make perfect sense.
For example, you might read some coverage that defence programs in a certain jurisdiction have blown out by potentially billions of dollars, pounds, or whatever the local currency is. But purely hypothetically, if you dig a little deeper and start looking at the reasons for those increases you might see items like: acquisition of additional aircraft. That is, the program budget has gone up because they are now buying more aircraft as part of the project than they originally intended to.
And because defence prime contractors are not charities, they asked for additional payment for the additional aircraft. Another common factor you might see sneak in is foreign exchange factors. If you're engaging with an American or European firm your contract is probably denominated in US dollars or euros. And if your local currency depreciates against that other currency over the course of the contract, the value of that contract in your local currency is going to increase.
Because you still owe the other side 200 million euros, whether that's a billion Kiwi bucks or 2 billion Kiwi bucks. So there's plenty of ways for scope changes or other factors to increase the cost of a contract without anything necessarily being stuffed up. To illustrate this point briefly, let's talk about Canada's plan to buy 15 Type 26 frigates for the Royal Canadian Navy. The plan here I believe was to start construction this year or next year, with the first deliveries of the ships in the early 2030s and final completion of the program in the 2040s. And when you look at the figures it looks like there has been a monstrous blow-out in the associated cost, so clearly someone has screwed up.
I mean in 2008 the initial cost estimate was about 27 billion Canadian dollars. In 2017 the Canadian Parliamentary Budget Office instead estimated 62 billion Canadian dollars. And more recently I believe that estimate increased to 77 billion Canadian dollars. To start with the initial estimate was probably very optimistic. In the words of Dr Marcus Hellyer it was "very inadequate."
The Parliamentary Budget Office would revise its estimate up after the Type 26 frigate was chosen. And this program seems to be about some generic frigate that could meet the requirements, to this specific frigate. Then the estimate would increase further to the 77 billion figure. But in significant part, that would be driven by the fact that the ship kept getting bigger.
The Type 26 grew from being about 6,900 tons displacement to almost 8,000. And the original Type 26 design was modified to add considerably more bling in certain areas. For example, the ship has a different sensor set and elements of the Aegis Combat System, something the original Type 26 simply didn't.
There are reasons to make changes like that, but they are not free. You may end up with a more capable platform as a result, but there's probably no way to avoid a bottom line blow-out. Now if you manage to avoid these and other pitfalls you'll end up in the happy ending scenario: delivery, sustainment, and potentially your program being a lasting contributor to your nation's national security for decades to come. But if the costs and delays start to mount, you may start to drift towards the other ending.
This is the quantity death spiral, and it's where you don't want to be. Let's say you're now in a very difficult position, you've burned through a lot of the initial program budget, and you either don't have a functional product or you have no hope of delivering the number of requested products on time and on budget. Depending on your jurisdiction and political system, you might be in some serious danger if you're delayed or over budget.
For example, if you're relying on funds that have been voted by your national parliament and more are not forthcoming, then you may just run out of money. If you're not moving quickly enough and your funds have to be voted year by year, your funding may actually lapse, that is it might disappear before you have the opportunity to spend it. And patriotic as they might be, most defence industry workers are not going to turn up and keep labouring away if you can't pay them. So you need to find a way to bring costs down, and quickly. The easiest way might just simply be to cancel the project and walk away. This can look bad because at this point the government may have spent a lot of money to get absolutely nothing in return.
But under certain circumstances it may be the best option available. Another one is for the government to try and save money by looking at another element of the contract, namely the quantity. If you find yourself in a situation where you only have 10 billion dollars left in your pocket, a situation I am sure all of us have found ourselves in at some point in our lives, but you realise it's going to cost 20 or 30 billion to actually deliver on the original scope, then you can go back to the manufacturer and perhaps say, "Hey, I don't want 1,000 of the thing any more, can you give me a price on 100 or 200?" And after taking a shot or two of hard liquor, the prime contractor will come back to you with an estimate on what it's going to cost for 100 or 200 of the system in question. But what they're going to do now is take all of their research and development costs, all their investment in new materiel, all their sunk costs, and instead of spreading those over 1,000 units, they are now going to spread them over 100 or 200.
Plus they're going to lose their economies of scale, and so you are going to see the cost for those units absolutely skyrocket. If you're wondering how the United States managed to get a 2 billion dollar strategic bomber in the 1990s in the form of the B-2, well, part of the answer is you cut a planned order of 132 to 21. In a sense you may have staunched the bleeding but you're not out of danger yet. Because now instead of having 1,000 aircraft in your original plan, you may now only have 100 or 200, which means all of the aircraft that those were meant to replace are going to have to keep flying.
And that can be pretty bad for your overall sustainment costs. If you have vehicles, ships or airframes that are reaching the end of their service lives, where the spare parts may no longer be in production, that were meant to be retired and as a result had their supply chain steadily being retired or unravelled. And now you say, "Hey, I want to fly that thing or sail that thing for an extra 10 or 15 years," then you are going to have a budget blow-out. It's one of the reasons why the US Air Force for example, is so keen to rapidly replace its old E-3 Sentry fleet with the new wedgetails. It's not just about the capabilities of the new aircraft, it's about the fact the old one is becoming ever more expensive to keep in the air. And since you are now keeping the older fleet in the air, you have less money available to buy new aircraft to replace the older fleet, which forces them to remain in the air.
And you can see the death spiral continuing. A similar issue can manifest with companies when they are entering into a death spiral of their own. If you see capital equipment being run far past its service life, or maintenance expenditure being cut in order to save the immediate bottom line, then you may be in trouble because the costs of everything beginning to break down over time are going to mount and mount.
Making it ever more unlikely you will ever get back into a financial position where you can afford to actually replace the relevant capital equipment. For an example see the Venezuelan oil industry. Where a lack of maintenance and reinvestment have caused outputs to steadily fall over time. You can try to stabilise the situation by asking for additional funding.
But if that's not forthcoming you may have to pursue other options. And while they might work they can also have associated costs. For example you could just downscale your fleet, have fewer ships, fewer vehicles, fewer aircraft. That allows you to retire expensive older platforms without actually replacing them with anything. And compared to a functional F-15 or F-22, nothingness has a pretty low upkeep cost. But part of the problem here might be as well as just having a smaller force in general, which may or may not suit your national security objectives, you are also now losing economies of scale on all your maintenance and procurement contracts.
But depending on your situation you may not have any alternatives. The Russian model, for example, doesn't provide any good alternative solution. Russian weapon systems have a general reputation for being cheap when the Russian government buys them. But at the same time, since the fall of the Soviet Union there has seldom been money there to invest in new plants and equipment in the various defence manufacturing firms. Productivity levels are low.
And even when a new piece of equipment is developed like Armata, or Typhoon, or the Sukhoi 57, large-scale production never really seems to follow. The money and capacity simply isn't there to produce things at scale it seems. So defence firms were stuck producing things at low rates, but weren't allowed to release their workforce or cut down their own costs, because the government didn't want the unemployment and wanted to keep that capacity within the industry. So a lot of these firms simply went into debt. And when that debt became too much to handle, well, the government would bail them out.
Either directly or indirectly through methods like restructuring their debt and forcing commercial banks to forgive it. So yes, some of these systems may have low official sticker prices. But that doesn't mean they can't also be quite expensive at the same time. And however you choose to handle the blow-out in time or cost, there is going to be a legacy for it. Because let's say you build up a potential project, encourage industry to invest, and then you cancel it.
Maybe in the case of that program it is the right call, but that decision might leave some in industry holding the bag. Some small suppliers may be wiped out entirely. And so the next time a negotiation comes around, they're going to remember. They are going to ask for higher risk premiums, for higher profit margins, for cost plus contracts that protect them from cancellations or overruns. And they're probably not going to do anything until you put actual money in their hands.
The situation is complex, but if you're looking for reasons that parts of the Western defence industrial base haven't move to pro-actively increase production after February 2022, and have instead waited and insisted for governments to sign contracts before they'll start moving, this may be part of the reason. OK, so let's put this all together. We have talked through the various stages of procurement, some of the things that can go wrong, let's lace it all together with one hypothetical example. For almost a year now the army of Kiwiland has been fighting to throw the last "separatists" out of their South Island territory. These separatists claim to be Kiwilanders who would like independence from the North Island, but an awful lot of them seem to be wearing emu and koala arm patches, use Emutopian tactics and they seem to operate an awful lot of Emutopian surplus equipment.
Kiwiland's mobilised army has been making ground for some time now, but they are desperately short of firepower. Like I said the separatists have an awful lot of artillery, which Emutopia insists they could have bought at any military surplus store. So Kiwi troops are often seen doing things like building makeshift home-made MLRS systems and welding them to the back of their Utes.
What they need is for government to step in and give them large numbers
2023-06-20 12:39
