James Hasik

The British Army pulls Bulldogs from Afghanistan

The Times of London reported today that the British Army will be withdrawing from service the 900 upgraded FV430-series armored troop carriers that it has received from BAE Systems and the Army Base Repair Organisation (ABRO) over the past three years. If this is true, it’s a sharp reversal of plans. Termed Bulldogs, the new FV430s came with a host of new features, including

• appliqué reactive armor (from Rafael) designed to repel hollow charge warheads such as those in RPG-7s,
• a new engine and steering gear for better speed, mobility, and maneuverability,
• air conditioning, which is considerably more important in south Asian deserts than on the north German plain, and
• a remotely-controlled 7.62 mm machine-gun mount

The problem, however, is that the upgrades couldn’t fix the FV430’s fundamental shortcoming: like its American analog, the M113, it’s an aluminum box with a flat bottom that sits about eighteen inches off the ground. As such, its shape traps blast gases from explosions (improvised or otherwise) underneath, allowing easy penetration of the hull at the expense of the occupants.

To be fair, the Pentagon did attempt something similar, two years before the British program started, at the urging of Congressman Ike Skelton's (D-Missouri). In December 2004, as roadside bombings were getting really alarming, he requested that the Army consider hanging more armor off those M113A3s that had been sent to Iraq. As Skelton wrote in his request,

        I understand that the M-113 may not provide the same level of protection as some other armored vehicles currently in use, but they certainly provide better protection than soft-skinned vehicles. Moreover, I believe that the M-113 chassis is robust enough to easily accommodate the additional weight of supplemental armor kits, whereas the HMMWV struggles under the burden and it is causing significant maintenance issues. As you know, broken-down up-armored HMMWVs provide no protection at all.

At the time, this seemed a more than reasonable idea, and by the end of 2005, a civil engineering squadron of the US Air Force (!) had upgraded more than 700 M113A3s at a disused factory in Iraq. Unfortunately, the effort didn’t quite pan out; despite all the extra armor, those M113s are hardly used now. In 2007, the US Congressional Budget Office checked mileage logs, and noted in its report that the M113s were clocking just one-tenth the monthly distance of MRAPs in-country, and one-twentieth that of Strykers. Troops logically prefer to ride in vehicles with steel hulls, more ground clearance, and ideally, v-shaped bottoms.

In consequence, I must say that I was a bit puzzled back in 2006 by the Ministry of Defense’s decision to upgrade the FV430s when the US record with the M113 had been so bad. As I noted in a related column last week, back in March 2004, one M113 (though admittedly a plain old A3 version) held up so poorly against a roadside bomb in Fallujah that the five guys inside all had to be buried in a single casket at Arlington. (See Eric M. Weiss, "5 U.S. soldiers, 2 from Texas, buried as group in Arlington," Washington Post, 19 June 2004.) To follow Congressman’s Skelton’s argument, while upgrading all those FV430s would probably have been better than doing nothing, it certainly wasn’t better than ordering more MRAPs faster, as the program absorbed cash that could have gone to better use. What’s embarrassing is that the MRAP concept was more than proven by that point—as if it hadn’t been proven in South Africa in the 1970s. So, it’s really not clear what the MoD was thinking.

Trying to shed some light on this, the Times reported that a spokesman for the ministry asserted that the vehicles really had been usefully employed, as they were procured for service in Iraq, but were just fairing less well on service in Afghanistan. This is a highly suspect explanation: the roadside bomb threat developed earlier in Iraq, where roads also happened to be much better. Faced with these differing environmental profiles and threat assessments, one would naturally have sent the flat-bottomed, aluminum-hulled, tracked vehicle to Afghanistan, not Iraq. The same MoD spokesman further insisted that all was not lost, as the Bulldogs would now be used for training exercises at British Army facilities in the UK and Canada. This follows an earlier assertion that the BvS10 Vikings withdrawn from Afghanistan—similarly out of vulnerability to mines—would be used for training in the UK. Someone might tell these mandarins not to take us all for fools: Bulldogs and Vikings have awfully high operating costs to be confined to parading about the Salisbury Plan or the BATUS. Surplus sale is far more likely at this point.

It’s hard not to observe that both the Bulldog and the M113A3 are built by BAE Systems Land & Armaments. That’s unfortunate, as it’s a great company with some impressive products, like the RG33, the FMTV (for now), and even the BvS10. That’s right: in another war, against a different enemy, the Bulldog or the Viking might excel at its task. The trouble is that the enemy gets a vote, and over the past few years, the Ba’athists, the Al-Qaedists, and the Taliban have come to prefer mines. Withdrawal of otherwise admirable vehicles as these could very well be the death-knell of flat-bottomed troop carriers throughout NATO. Existing inventories will persist for some time, but the implications for new programs, from the US Army’s Ground Combat Vehicle (GCV) to the US Marine Corps’s Expeditionary Fighting Vehicle (EFV), could be dramatic. If the preceding Manned Ground Vehicle (MGV) project couldn’t proceed with its boxy design concept, it’s hard to imagine how the GCV or EFV could. And the implications for that shift extend across the industry.

Boeing's new bomb, Goodrich's new MEMS, and further thoughts on military-industrial restructuring

Defense Industry Daily reported on Monday that a team led by Boeing’s Phantom Works has received a first-phase contract from the US Air Force Research Laboratory to demonstrate a miniature precision weapon for use on unmanned aircraft. [See “Boeing Team Gets USAF Contract for UAV Miniature Weapon Development,“ DID, 16 November 2009.] As prime contractor during the initial nine-month program, Boeing will develop an all-up weapon, including seeker, avionics, guidance, control, and mission planning systems. As designer and builder of both the Joint Direct Attack Munition (JDAM) and the Small Diameter Bomb (SDB), Boeing was a natural choice for this effort, so this story was almost not news.
Almost, because there is an important underlying message here. Boeing will not be going it alone. As in the JDAM and SDB programs, Boeing will be relying heavily on subcontractors to develop and build large parts of the weapon. In addition to Boeing, according to DID’s story, the team includes five companies. Albeit a small sample set, the distribution of the nature of the companies is pretty wide:

• KaZaK Composites will build the airframe. Kazak is a forty-person company that focuses on composites in aerospace, marine, and military applications. As the website puts it, Kazak “is an active participant in the Small Business Innovation Research (SBIR) program and is always looking for partnering opportunities.” It just found one.
• Ensign Bickford Aerospace & Defense Co. will design and build the warhead. Ensign Bickford is actually three companies: the other two in the portfolio work in commercial realty and pet food ingredients. If that seems strange, doubtless someone in the private equity business has had the same thought. The aerospace and defense group was founded by William Bickford, who invented the safety fuze way back in 1830, and even today basically does energetics and explosives.
• Systima Technologies will provide integration and testing services for the launcher. Employee-owned Systima also has fewer than fifty staff, and concentrates on aircraft and spacecraft launching and separation systems.
• Science Applications International Corp. will provide systems engineering for the seeker and seeker algorithms. SAIC is, of course, a diverse and huge company, with 45,000 staff worldwide. Contrary to its one-time image, however, this program shows how it isn’t just a consultancy. As I highlighted in the last chapter of my recent book, SAIC actually writes software and does rather a bit more. [See James Hasik, Arms and Innovation: Entrepreneurship and Alliances in the Twenty-First-Century Defense Industry (University of Chicago Press, 2008)]
• Mustang Technology Group will provide height-of-burst and radar options. Mustang, like Kazak and Systima, is a roughly forty-person outfit. According to the website, it’s mostly largely former senior engineers at Raytheon and Texas Instruments who decided that they could do it better-and-cheaper.

I covered this dynamic in my case study of the JDAM program in Arms and Innovation: a systems integrator concentrates on, well, integrating the system, and parcels out the design and manufacturing of the subsystems to specialists. Now, the JDAM, the SDB, and this new UAV-bomb-thingy aren’t exactly a Dreamliner, so Boeing’s woes on the 787 program probably aren’t wholly applicable here.
Rather, we should expect to see more of this structure in new programs, and for two reasons. First, as it was put by Carl Avila, the director for advanced weapons at Boeing’s Phantom Works, the interest lies not just in the weapon’s low carriage weight for these rather smaller aircraft, but in its prospect for low collateral damage in a “suburban“ environment. In some ways, low weight—low collateral damage is the new tastes great—less filling of military aerospace. Second, the military procurement budget in the United States will contract in the near term, and new programs will need to show how they can meaningful breakthrough old cost-performance curves if they are to capture funding. Hiring the guys who used to work at Raytheon, but now think they can offer more-for-less than Raytheon (or anyone else), will recurrently be appealing.
There’s an important message here for private equity investors, merchant suppliers of engineering products, and systems integrators alike. It’s obvious that the arms industry’s structure is as heterogeneous as its products. What’s not obvious is that all the myriad organizational forms are equally efficient. A few operations (like Boeing’s 787 line) could benefit from a little more selected integration, or from the acquisition of deep functional skills and assets in a few strategic areas. Broadly, though, the biggest firms just can’t cost-effectively manage the down-in-the-weeds details of military design and production efforts. Consider the story from Tuesday morning: J. F. Lehman & Co.'s sale of Atlantic Inertial Systems (AIS) to Goodrich. BAE Systems sold this outfit to Lehman in 2007 for $140 million in cash; Lehman is now selling it on for $375 million. After pulling it out of its parent company, Lehman more than doubled AIS’s value in two years, and in deteriorating military-economic conditions.
That’s such a healthy return that it may lead one to ask why BAE Systems itself couldn’t realize this value from the property. Even so, heaping criticism on BAE would be a bit unfair here. The company makes nuclear submarines, armored fighting vehicles, and perhaps half of every Eurofighter. The span of control needed to administratively control all those activities is just too great. BAE Systems is a 100,000-person company; AIS is an 800-person company. That necessarily puts the embedded operation well below the fold of top managerial attention. Thus, there may be more underperforming assets in the portfolios of big companies in the industry that either need to be turned around to face the opportunities of the twenty-first-century arms industry, or sold off to someone who can take them there.

Is Oshkosh's FMTV bid really a risky move for the US Army? Part Three: Competitive Comparisons, and the Logic of Open Bidding

Back in September, just after I last commented here on the ongoing dispute over the US Army’s selection of Oshkosh as its new supplier of its Family of Medium Tactical Vehicles (FMTV), I was asked by how much I thought that Oshkosh had underbid. Yesterday, I was asked by someone completely different how BAE could have so clearly overbid. (I’m still to be asked about Navistar, but I’ll not be surprised to get that call.) Could both assertions be true? Could neither be true? While the details of the bids may not be completely public, the essence of two of them has clearly leaked. Referring again to the issue brief written in September by Loren Thompson (“Army Truck Award Looks Very Suspicious,” Lexington Institute, 8 September 2009), we all understand that BAE Systems bid 21 percent below its own current price for FMTV trucks, but that Oshkosh had bid fully 30 percent below. That is, as Loren put it a week ago:

On price, the Army accepted a bid from Oshkosh that is 30% below the price BAE Systems is currently charging for the vehicle -- even though Oshkosh, with no direct experience in manufacturing the product, must turn out a vehicle with the same performance specifications and features, using many of the same suppliers.
        (“FMTV: How The Army Erred,” Lexington Institute, 4 November 2009)

On the surface at least, that makes for a darn good question. Moreover, this issue of alleged underbidding is likely to continually resurface, as the Pentagon attempts to work its way through a flat or declining budget that cannot accommodate all its desires. As it so tries, the military services may find competing, lower bids from new entrants rather alluring. In the long run, however, they’ll only find them satisfying if the companies in question can actually deliver. So, as important as this may be in the long run, what more can we say about this case? We’ll center here on Oshkosh: as a company with narrower scope than BAE Systems, its financial statements and press releases are more revealing of its operations.

To begin, consider Oshkosh’s business portfolio in military trucks at the time of its bid. While Oshkosh recently won the contract for the Army and the Marine Corps’ MRAP All-Terrain Vehicles (M-ATVs), production hadn’t started by the time the bid went in. Oshkosh did have, however, considerable experience building the Army’s Family of Heavy Tactical Vehicles (FHTV), the Marines’ (heavy) Logistic Vehicle System Replacement (LVSR), and the Marine’s Medium Tactical Vehicle Replacement (MTVR). That’s a lot of experience in how to make military trucks. That says nothing about its experience in the rather more competitive (civil) speciality truck and construction equipment markets—for its part, BAE Systems only sells trucks to military forces.

From government contracting records, Oshkosh knew how much BAE Systems was being paid for the trucks. Since this was a build-to-print competition, Oshkosh also had the bill of materials (BOM), which it could break down and cost item-by-item. With those estimates, and its running production lines, the company’s bid team could sketch detailed processes for how to source the parts and build the FMTVs in its own facilities. The team could also estimate how much each step in the process would cost, and identify the rate-limiting steps that merited the closest attention in subsequent process improvement.

All this should be fairly obvious, even if it does make for a dramatic amount of information. But what else can be observed publicly for competitive comparison? Still quite a bit:

Schedule. To start, consider BAE Systems’ recent record in the FMTV program. According to Inside Defense earlier this week, BAE is 850 trucks behind schedule. The Army is said to be less than thrilled, but to understand the challenges of producing the vehicles both at high volumes and with a constant flow of engineering changes recommended by battlefield experience.

All the same, Oshkosh’s current military production programs do not appear to be straining the company. Consider its most recent and urgent award, the M-ATV. Through the end of October, according to its sequence of press releases, Oshkosh had delivered at least 735 M-ATVs, and is now ahead of its own schedule to build 1,000 M-ATVs monthly by December. From that point, the company will finish the Pentagon’s current orders before the end of April, and low-rate production of new FMTVs isn’t scheduled to begin at Oshkosh until April. From there, the FMTV contract calls for 23,000 trucks and trailers over five years, or fewer than 400 monthly at full-rate production. That’s roughly 60 percent of BAE’s peak production this past year. Overall, Oshkosh’s FMTV production would neatly follow its M-ATV production, and at its peak, be considerably less taxing.

Capacity. That, however, is just the story for Oshkosh Defense. With its construction equipment and specialty truck business sharply down over the past two years, Oshkosh as a whole has plenty of space in its factories for military production. Some of its M-ATV work has been assigned to a factory in Pennsylvania that, in better economic conditions, would be producing JLG construction equipment. As I will highlight below, that space needn’t all be used, but it is available. That availability also points to how Oshkosh has plenty of furloughed staff who’d be happy to get back on the line.

Staff. By Business Week’s recent estimate, after Oshkosh’s recent divestiture of Dutch trash truck manufacturer Geesink Norba, the company has about 12,400 staff worldwide. Oshkosh doesn’t indicate how many staff work in which subsidiary, but we can estimate this from the revenues associated with each, if we assume that the ratio of sourcing to in-house work is roughly constant across Oshkosh’s various vehicle and construction equipment operations. By that figuring, and allocating their share of corporate staff, Oshkosh Defense probably accounts half the company’s total employment, as it accounted in 2009 for 49% of the company’s annual revenue. Those (presumably) 6,200 people generated completed work worth $2,594.8 million (see below), or roughly $418,000 per person.

According to today’s Chicago Tribune, Oshkosh has hired or rehired 1,400 staff to build the 6,200 M-ATVs that the Pentagon has so far ordered for about $3.2 billion. That was a bit above the 850 to 1,150 staff that Oshkosh had forecast hiring back in July, but after all, the Pentagon’s intended order has almost tripled since then. Since the M-ATV is close to full-rate production now, it’s likely that the company is fully staffed up for it. Compared to the rest of the business, that’s impressive daily productivity, though it’s very possible that Oshkosh is granting lots of overtime to get the job done.

However, after winning the M-ATV contract, but before winning the FMTV contract, Andy Hove, president of Oshkosh Defense, told Crain’s Detroit Business that the latter program would add only 100 to 200 staff at the company. Many of these, he said, would be in southern Michigan, and that has the sound of program management and engineering positions near Army Tank & Automotive Command (TACOM) headquarters. As production of the M-ATV will end well before production of the FMTV is to begin at Oshkosh, Andy’s comment about additional hiring strongly suggests that Oshkosh will build the FMTV with not more than dedicated 1,500 staff. By comparison, note that BAE Systems produced over 8,000 FMTVs in Sealy, Texas last year with about 2,500 staff. That suggests that Oshkosh’s production staff will be just 60 percent of BAE Systems’ current staff, and in any case, will make up about a quarter of the total staff at Oshkosh Defense. That’s a fairly manageable number.

Finances. Excluding non-cash charges related to discontinued operations at Geesink Norba, all of Oshkosh essentially broke even in its just-ended 2009 fiscal year, earning $3.8 million on $5,300 million in sales. The Oshkosh Defense subsidiary, however, did much better, contributing $403.3 million in operating profit on 2,585.8 million in sales. Actually, Defense at Oshkosh has been on a roll since at least the end of 2007, with consistently double-digit contribution margins (CMs):

($ MM)	1Q 2008	2Q 2008	3Q 2008	4Q 2008	1Q 2009	2Q 2009	3Q 2009	4Q 2009
Revenue	398.4	450.9	489.4	553.3	543.8	590.2	605.4	855.4
Op. Prof.	63.9	59.7	66.5	75.1	73.7	75.0	92.9	161.7
CM	16.0%	13.2%	13.6%	13.6%	13.6%	12.7%	15.3%	18.9%

space
This money is coming from mostly five sources: new FHTVs, new LVSRs, new MTVRs, replacement parts for those trucks, and TAK-4 suspensions for existing MRAPs. Only in the last quarter did significant sales of new M-ATVs begin, and these were responsible for about $100 million in revenue. As the contribution margin actually increased in that quarter, it’s highly likely that Oshkosh is making money on the M-ATV program specifically.

Facilities. According to BAE Systems, the current FMTV factory in Texas has over 800,000 square feet of production space. According to Oshkosh, all of Oshkosh Defense has somewhat more than one million square feet. More space might seem better, but utilized space is costly, so the footprint of a factory is a strong signal of its cost-efficiency. That’s a leading reason that the lean manufacturing movement stresses constant efforts to reduce required floor space. These two numbers, then, lead us to question just how efficiently BAE Systems is using its space, and whether its mastery of automotive production processes are at the same level as Oshkosh.

Supplier management. In March, Oshkosh told Defense Industry Daily that over 90 percent of the BAE Systems’ suppliers in the FMTV program were also suppliers to Oshkosh. Since prices for industrial products are hard to observe, it’s not uncommon for a single supplier to charge significantly different prices to multiple customers for what’s essentially the same part. In my own consulting work, I’ve even seen single suppliers sell the same part for significantly different prices to multiple divisions of the same company. Good supply management ferrets out those differences. It also figures out how to reduce costs common to all the participants in the supply chain, in a win-win way. All of Oshkosh’s suppliers use a common software application for cost-estimation, so expected costs should reasonably conform to parameters understood from the FHTV, LVSR, MTVR, and M-ATV programs. Thus, to answer Loren’s question above, Oshkosh is indeed using many of the same suppliers that BAE is, but it could very well be using them better. As we’ve seen from the commercial automotive industry, good supplier management is sometimes a decision competitive edge.

So, from all this, what can we say about Oshkosh’s bid on the FMTV program? When it bid, the company knew that

• it was making double-digit margins in the analogous FHTV, LVSR, and MTVR programs
• it was having little problem meeting its several schedules, while BAE was having a lot with its one
• it would build, at full rate, less than 60 percent of BAE’s production last year
• it could do so with about 1,500 people, or about 60 percent of BAE’s staff
• it could make do with far less production space
• it understood the cost structures of 90 percent of BAE Systems’ FMTV suppliers

Although we can’t be certain, it’s possible that BAE Systems was bidding on an operation that would retain what was effectively surge capacity that it didn’t need, in both people and space. It could have planned future cutbacks insufficiently aggressively. It could have just plain been paying too much for parts. Whatever the case, what’s fairly clear is that Oshkosh was not an inexperienced bidder. It knew a lot about BAE’s operations before it started sharpening its pencil.

That pencil-sharpening of competitive bidding leads to an important point about the limits of bureaucratic rationality, and the utility of market processes in driving efficiency. In this competition, the Army received three bids, but underneath each of these known prices lie two other figures: each company's expected cost, and the actual cost that each company could eventually attain in production. To use one of Donald Rumsfeld's better-known constructs, the bids were known-knows. The eventual actual costs were unknown-knows, in that one of them would eventually be revealed in the performance of the contract. A priori, however, the expected costs were and remain almost unknowable to a fine degree of precision. The contractors could estimate their costs based on codified experience and even tacit knowledge, but could not be sure in advance that they had it right. Still, if this were so for the contractors, it was all the more so for the Army and the GAO.

In short, to answer the questions that launched this analysis, it’s possible that Oshkosh underbid, but the evidence suggests that this is not likely. It’s possible that BAE overbid, but that’s a matter of what BAE could attain with its existing cost structure and managerial capacity. We just won’t know until Oshkosh starts building FMTVs, and begins reporting its quarterly earnings. At that point, we’ll watch to see whether the contribution margin holds up at Oshkosh Defense. If it doesn’t, the Army still shouldn’t get alarmed unless this one program manages to put Oshkosh into financial distress—and that is extremely unlikely. The FMTV rebuy contract is a fixed-price, build-to-print deal. If Oshkosh did indeed bid too low, then the taxpayers and the Army will benefit, and Oshkosh’s shareholders can take the issue up with management.

This, then, gets to the fundamental trouble that I have with any protest after a fixed price competition. If the protestors are arguing that the Army could not possibly trust Oshkosh to do the job for less, then they are implicitly stating that the competition was effectively a sham. If no one but the incumbent could be trusted at such a low price, what was the point of taking bids? Merely to provide stalking horses? Since Oshkosh and Navistar had nontrivial bidding costs, asking for those bids with no real possibility of winning borders would border on dirty pool.

For this reason, I would advise the government to tread carefully in assailing any assertions that the contractors might make about their future cost structures. The beauty of competitive bidding is that it tends to reveal long-term sustainable prices after repeated rounds, as the losers eventually drop out of the business. Mess with that dynamic too much—even on what better knowledge that the GAO thinks it might have—and competition drops off. Eager entrants like Oshkosh may grow discouraged, and stop bidding. If the government can’t follow market processes with something as widely bought and sold as a truck, then it’s looking squarely back-to-the-future of $500 toilet seats.

Some further explanation about the Stryker question

Ever since the US Army first bought into the LAV-III concept in 1999, there has been a fascinating pathology of Stryker-hating afoot. So I was again reminded today when I received some comments on my last analysis of fatality rates in the Stryker in Afghanistan, and how these might compare to previous experience in Iraq. Perhaps unsurprisingly for the Internet, the comments were a bit off-the-reservation, so I’ll not offer them verbatim. Slightly edited for public sensibilities, brevity, and grammar, he made three claims:

1. You ignore the M113 whose up-armored model has a perfect protection record in Iraq.
2. Sergeant Rabidou and his peers are dying: this isn't anecdotal; it’s reality. Using the slang "kevlar coffins" refers to the ceramic tiles placed on the outside to try to compensate for the thin steel Strykers are made of.
3. Flat-bottomed Strykers cannot accept enough add-on armor to defeat the land mine threat or ever go cross-country at will. Thus, they not effective.

No points, by the way, for guessing who offered this. All the same, since it was all fairly entertaining, I thought that my offer some perspective, which is in short supply on the web these days. Good social science requires keeping a level head about phenomena. This isn’t true even when the stakes are high; it’s particularly true when the stakes are high.

By "perfect protection record" and “up-armored model,” I surmise that he means that no one inside or on an M113A3 has been killed in Iraq to date. That's simply not true. As early as March 2004, five soldiers died in an M113 that encountered a roadside bomb in Fallujah. The whole thing was such a mess that all five were buried in a single casket at Arlington. (See Eric M. Weiss, "5 U.S. soldiers, 2 from Texas, buried as group in Arlington," Washington Post, 19 June 2004.) Including this incident, my research from October 2007 uncovered at least 12 deaths of troops in M113s through September 2006 alone, and in six separate attacks. I cannot confirm that these vehicles were all A3 variants, but they almost certainly must have been. Production of A3s finished up in 1992, and while the Army has over 6,000 of them, it has sent only about 1,500 M113s to Iraq.

Next, the fatalities in Sergeant Rabidou's unit were certainly tragic. I cannot not claim that men in his unit are not dying, nor that they have not died in Strykers. Rabidou's dislike of the Stryker is now a matter of record. It may be heartfelt and based on personal experience, but statistically, it is at best a datum. Without more data against which to compare it, it is thus merely an anecdote, if a colorful one. It may lead us in the right direction, but it is not confirmatory on its own.

Thus, even the spectacularly messy deaths of these aforementioned five soldiers are similarly just a datum. We don't know how big that bomb was; if some crew of insurgents had manhandled a 250-kg aircraft bomb into place for the attack, it could easily have done horrific work on an M113. That still wouldn't provide real comparison to any other vehicle. I do, however, take some direction from the Army's total lack of enthusiasm for M113s in Iraq. According to the US Congressional Budget Office, as early as 2007, M113s were clocking only one-tenth the monthly mileage of MRAPs in Iraq, and one-twentieth that of Strykers. (See "Replacing and Repairing Equipment Used in Iraq and Afghanistan: The Army’s Reset Program," September 2007.) That's a fairly strong indication that M113s are considered too vulnerable to be let outside the wire.

As for the assertion that the Stryker cannot accept enough armor to defeat the land mine threat—particularly if it is to travel cross-country—that’s a matter of comparison and perspective. Even Abrams tanks can be penetrated by bombs and rocket grenades, as at least a few have in Iraq. How much protection is enough protection is a matter of balancing military judgement with technological and economic constraints. I wasn't actually asking that question in my earlier column. I was merely observing that the loss rate for troops in Strykers in Afghanistan this year was not yet clearly higher than the loss rate for troops in Strykers in Iraq several years ago. Thus, whether one finds that loss rate militarily acceptable or not, the referenced article is not news.

As for the point that the Stryker is flat-bottomed, well, that’s a reasonable description. While there is a little sloping on the sides of the current vehicles, only GD's latest concept, the so-called "Super Stryker", shown at AUSA last month, has a v-shaped bottom. All M113s, however, are flat-bottomed. So are all Bradleys, CV90s, ASCODs, and BMP-3s. As I have written repeatedly, it is my view that designing a flat-bottomed armored troop carrier is almost a commercial non-starter today. Only two countries, Germany and Israel, are funding active development programs in such vehicles today, and their respective investments, KMW's Puma and IMI's Namer, are very heavy. What's really bad about the M113 (and the Bradley, for that matter) is that the hull is made of aluminum, and aluminum's shear modulus is less than one-third that of steel. The steel plating on the sides helps, but really doing the vehicle right means building all the structural components from steel ab initio. The prevalence of blast weapons today renders that an imperative.

Lastly, for what it's worth, kevlar is a fiber, not a ceramic. So, Sergeant Rabidou's statement still doesn't make sense to me.

While I do not generally comment on client relationships, I will, in the interests of full disclosure, state that I am not working for GDLS or any parent unit of GDLS on this issue. Of course, if I were, one could expect that I would just keep quiet, at least until enough time had passed that the issues had become fresh again. Rather, I am just an analyst using some of the more basic tools of social science to get at answers about what is working, what is not, and what might in the future. I do have clients who ask about just that, because they're looking to sell kit to armed forces around the world. And I work at this hard because the numbers always matter.

Northrop's OCI-induced sale of TASC, and what the resulting share repurchase plan says about the business

Earlier this week, Northrop Grumman announced the expected sale of its government consulting business, a company that back in the 1990s was known as TASC (The Analytical Sciences Company, for those who remember). Rebranded again as TASC, the division and a few other businesses are being sold for $1.65 billion to private equity firms General Atlantic and KKR. The reason for the deal centers on what industry and government these days call organizational conflicts of interest (OCI). As Chairman and CEO Ron Sugar put it in a press release, the sale “reflects Northrop Grumman’s desire to align quickly with the government’s new organizational conflict of interest standards, while preserving TASC’s unique organizational culture and its status as the advisory services employer of choice.” One couldn’t make the motivations more apparent.

Time was, the idea of the US federal government asking a large company to advise it on whether it should buy that company’s products seemed crazy. Such work was reserved for standalone consultancies; frequently, it was reserved for nonprofit businesses designated Federally-Funded Research & Development Centers (FFRDCs)—places like RAND, IDA, and MITRE. At some point in the early part of this decade, the government decided the conflicts would be manageable, and declined to object to the acquisition of many of the consultancies, such as TASC, by hardware-makers like Northrop. The Obama Administration’s reversal of this position isn’t the first time that the US federal government has gotten religion about OCI. RAND, which also undertakes space advisory business for the DoD, started out in the 1940s as a unit of Douglas Aircraft; spinning it off was strongly favored by the USAF at the time. Another crosstown competitor, The Aerospace Corporation, was formed in the late 1950s as a spinout from the merger that formed TRW from Thompson Products and Ramo-Wooldridge. The former was in the hardware business; the latter was largely advising the USAF on space programs, and the combination made the government nervous.

This sea change has probably been coming for some time; indeed, the current administration may have just made a policy of (re)emerging preference. Back in 2007, when I tried to develop some business at the Pentagon on behalf of another (for-profit, publicly traded) consultancy, folks there told me that they loved the idea of the work that we could do, but preferred to hand it to an FFRDC. I couldn’t complain: it was one clear way of avoiding any conflict, real or imagined. In this case, the divestiture of TASC has been expected for many months because it was one of the problematic cases for all concerned. To provide a bit more detail, I refer to recent analysis by Cai von Rumohr, Gautam Khanna and Mark Hokanson of the Cowen Group. They argue for the deal for perhaps four specific reasons (that I’ve rephrased slightly):

1. TASC's primary customer, the National Reconnaissance Office (NRO), has taken perhaps the toughest line of any federal agency OCI issues, and both Northrop and TASC have lots of business with the NRO.
2. Some of TASC’s business with the NRO is up for recompetition, and Northrop has been worried that it might lose this business for just this reason.
3. In the sale, Northrop transferred to TASC other operations with potential OCI issues to avoid further problems on its side.
4. KKR and General Atlantic face no such OCI issues as owners of TASC.

Von Rumohr et alia separately mentioned the possibility that TASC’s business with the NRO might slow if, as some of their sources in the industry expect, the Defense Department’s spending on NRO satellites slows. That, however, only affects the price at which Northrop would sell the property, not its reasons for selling in the first place.

It’s important to note that handing work to even an FFRDC won’t eliminate all charges of conflicts of interest. When the USAF proposed a few years ago to ask RAND to evaluate some aspects of its handling to the KC-X aerial tanker program, US senator John McCain complained that RAND shouldn’t be trusted to deliver an honest answer. So much of RAND’s business was with the USAF, he said, that RAND had effectively gone native. Knowing RAND’s work, I thought that this charge was a little too strong, but there are other examples of FFRDCs thought too close to their sponsors.

All that said, even if the reasons for Northrop’s sale of TASC are obvious, what’s perhaps more notable is what Northrop plans to do with the proceeds. After taxes and fees, the sale will net about $1.1 billion, and all of that will be offered to repurchase shares of NOC. Indeed, Northrop’s management could be said to be on a bender of share repurchases. It’s a very strong signal that Northrop's management is sending its shareholders: we're offering you all the cash because we are unconvinced that we'll outperform, on the margin, any given mutual fund, if we keep that cash for our own uses.

This tells us that management is rather sanguine—or just realistic—about the prospects for the company. I am sympathetic to that view, as I expect military spending in the US to drop sharply, and more sharply than spending elsewhere in the world. As Dick Bitzinger put it in the Straits Times a few months ago, the market in the US looks particularly grim for many of the charismatic macrofauna in Northrop Grumman’s portfolio. Many of these big systems aren’t likely to be prioritized within the US government’s declining military investment spending. Here, there are both an income effect and a substitution effect: with less money to spend, the new Pentagon leadership will trade towards investment in more numerous and less expensive alternatives to the Cold War’s “exquisite” solutions. Since Northrop does only about six percent of its business outside the US, quickly expanding its sales overseas could be quite difficult. All this adds up to a vulnerable position—unless management is willing to recognize it as such, and earn its returns another way.

So, rather than finding this relative industrial disinvestment alarming, I find it deeply refreshing. Frankly, I'd call the relative infrequency of share repurchases a mystery, if corporate double-taxation, agency problems, and general managerial hubris didn't conspire to help the Man hold the shareholders down. :) Northrop’s management deserves credit for thinking better of all those widows and orphans whose retirement portfolios back their operations.

Some real numbers on the Stryker's performance in Afghanistan

A colleague this morning asked for any quick comments that I might have on Sara Carter's article in today's Washington Times on the Stryker armored vehicle, the US version of the LAV-III. The title, "Armored troop carriers called unsafe for duty," rather telegraphs its intent. Specifically, he wanted to know whether the article could inform any speculation about what will come of the US Army's armored vehicle recapitalization efforts—or, what comes after FCS.

Here are the facts that Sara presents:

1. Since the 5th Stryker Brigade deployed to southern Afghanistan in July, the formation has lost 21 of its 350 Strykers.
2. In that time, the brigade has also suffered "more than" 12 dead and "nearly" 70 wounded.

The rest of the article is color commentary. Doug MacGregor, for example, calls the Stryker "essentially a paramilitary police vehicle... designed to transfer American light infantry down a road." He also thinks that what troops in Afghanistan most need is "is tracked armor, off-the-road capability and a stable platform for large-caliber guns." That’s entertaining, but unhelpful. To begin, the term "light infantry“ is essentially meaningless: all infantry should move down roads with armored protection. Their alternative conveyance should be trucks? Walking across Afghanistan? I won't comment on Macgregor’s cranky use of the term "paramilitary," but I will note that no evidence is offered concerning the suitability of tanks with heavy guns for counterinsurgency missions. We’ve seen some useful employment in Afghanistan of Danish and Canadian Leopard tanks with their 120 mm guns, but we’ve also seen useful employment there of British Scimitar tanks with their 30 mm guns. Since General Dynamics has built other version of the LAV-III with 25 mm cannons, and showed one with a remote 25 or 30 mm turret at AUSA last month, it’s far from obvious that the right approach can’t ride on wheels.

Also intriguing is how Staff Sgt. Daniel Paul Rabidou of the 1st Infantry likes to call the Stryker a "Kevlar coffin." The Stryker is made of steel; its spall liner is made of Kevlar. Since tanks have been called "steel coffins" with some affection since at least the Second World War, I'm not sure why he would choose to call his ride a Kevlar one. In any case, this is what we call an anecdote, and as is sometimes lost on some journalists, the plural of anecdote is not data.

Rather, let's focus on what we do know, working through those few datums we actually have in this article:

1. Assuming replacements and a constant level of violence, losing 21 of 350 vehicles in six months amounts to an annualized loss rate of 12 percent.
2. Similarly, "more than" 12 dead in a brigade of 3500 troops represents an annual loss rate of "more than" 0.7 percent.

Regarding the 12 percent vehicle loss rate, as a marketer in the business once told me, it’s better to lose auto parts than body parts. For wealthy countries fighting far-off insurgencies, replacing vehicles is far better than replacing people. If the vehicle is destroyed but saves enough of its occupants often enough, then it has done its job.

Now, if we assume that all the fatalities in the brigade occurred inside Stryker vehicles—and that is almost certainly not the case—then the fatality rate exceeds 0.068 dead per vehicle per year. One should only get worked up about this number if it is markedly worse than the alternatives. In October 2006, I published an analysis of fatality rates in US armored vehicles in Iraq during the first three years of counterinsurgency work there. I found these figures:

0.024 per Abrams tank
0.033 per Stryker troop carrier
0.044 per Bradley troop carrier

Note that I am not comparing these rates to those in MRAPs, which have rather lower fatality rates amongst the infantry. Still, we do know from this analysis that the Stryker’s annual fatality rate per vehicle, in this one brigade in southern Afghanistan this year, could conceivably be twice that of comparable to rates in several Stryker brigades fighting in northern Iraq over three earlier years. That would assume, however, that all the fatalities were incurred in the brigade’s Strykers, which again, is almost certainly not the case. By itself, this is not news. We need more data to draw any useful conclusions.

In the interest of full disclosure, I will note that I am not currently undertaking any work for General Dynamics, which makes the Stryker, nor for any other company with a significant part in the program. Rather, I am just looking for answers, but specifically, answers to the right questions. The question here is whether the armor, speed, and situational awareness afforded troops by the Stryker are better or worse than those features afforded by other vehicles. This article doesn’t quite tell us otherwise.

MRAPs, of course, are wholly another matter. The MRAPs that did such great duty in Iraq, of course, were largely meant for urban applications, and Afghanistan is clearly a much more rugged place. That is why the USMC is retrofitting its Cougars in Afghanistan with Oshkosh’s TAK-4 offroad suspension, and why the services together have invested so much in the MRAP All Terrain Vehicle (M-ATV) program. They’re seeking roughly the offroad mobility of the Stryker, but with better mine protection.

So, if there is any indication here about what comes after FCS, it’s that any vehicles ordered must feature enough mine protection to seriously complicate road-mining by foot-mobile sappers, who can only carry explosives so large, and enough protection against shaped charges to repeatedly ward off strikes from at least RPG-7s. Separately, I will argue that ballistic protection in excess of 14.5 mm all-around protection at 200 meters in relatively unimportant, since US infantry are not likely to frequently encounter cannon-armed opponents inside missile range. (It’s just that can’t derive that per se from this article.)

At the same time, as much of counterinsurgency is about securing the lines of communications for the populace, the high road speed and low road impact of vehicles like the Stryker or of the MRAP class shouldn’t be discounted. Charging through farmers’ fields with tanks in search of guerrillas does nothing for hearts and little for minds—a point perhaps yet lost on those who spent their careers preparing for the big fight in the Fulda Gap.

An impending train wreck of procurement for the USMC? Read the slides.

This week someone brought to my attention a budgeting briefing from this past July that someone in the Navy Department seems to have inadvertently posted on the Internet. At least, briefings marked like this usually aren’t intended for public consumption:

—————

Pre-Decisional. This briefing contains Planning, Programming, and Budgeting/Execution System (PPBE) data and is not to be disclosed outside the United States Marine Corps and other governmental agencies directly involved in the resource planning and allocation process.

—————

But that was then—it’s in the public domain now, and it says so much about the Pentagon these days. The first item of note is found on page 12, where the text under the headline “QDR” starts simply with

• Bullet (Arial 24 font)

What exactly this says, I’m not sure. Either that’s all that the Marines think that they’re getting out of the QDR, or it simply confirms that format >> content. Either way, it’s just so entertaining.

Also entertaining is page 14, which shows just four programs—EFV, MPC, JLTV, and G/ATOR—eating more than the entire projected Marine Corps procurement budget by FY2018. That this made it onto a slide indicates that some general officer or political appointee has figured out that there’s a real problem here. This, in turn, suggests one of two things:

1. The Marines need a whole lot more money. Perhaps they’re unlikely to get it, but someone in the Corps could start making the case that the Marines should be allowed to steal some share of the US military budget from some set of the other services.
2. If not, then the EFV program is in deeper trouble than its seeming reprieve earlier this year would suggest. The Marines have other ways of getting to shore—hovercraft, helicopters, even displacement landing craft—even if these don’t involve charging through the surf in what’s effectively a tank.

Some of the people involved with this might want to consider Sir Earnest Rutherford’s words: “Gentlemen, we have run out of money. It is time to start thinking.”

And could someone please tell me what (page 10) a “Kusky Process” is? The briefing itself is literally the only link that pops up in a Google search.