Back in the industrial era, the production process was described by interchangeable workers and capital assets measured by plant, equipment, and inventory. Those days have passed. Most job openings require specialized skills achieved by a small fraction of the workforce. Most of today’s asset value is intangible rather than physical, found in software, data, product design, supply chains, training, culture, and so forth. As we march further into the era of intangibles, structural changes will require a re-evaluation of the Department of Defense’s management systems founded on industrial era techniques.
Industrial Administration
Rational management starts with setting objectives and goes on to analyzing alternatives, detailing an execution plan, allocating resources, and measuring performance to plan. In an industrial era dominated by physical commodities, such techniques performed well. The output was generally of a known specification with an existing market price. It was mostly tangible, the bulk of its value coming from raw materials and machining. If the project failed, much of the investment could be recovered in capital stock and intermediate goods. In short, there was less uncertainty.
The output of industrial era corporations depended on physical capital for transportation, fabrication, assembly, and storage. Human labor was largely tactile as it moved, cut, and fitted objects. These facts had implications for management techniques.
If managers wanted to expand output, they’d fully plan what plant and equipment would be procured and at what specifications. All that before allocating resources. If they got behind on opening a new plant, they could expedite it by adding construction labor. And until the plant was at maximum capacity, managers could increase output by adding labor.
At the time, introducing a new worker into construction or onto the assembly line was easy. It took little time to ramp the employee up and just as little coordination from the manager. Indeed, this was the genius of industrial era management. Managers could specialize employee functions to a routine, requiring little oversight and only brief—linear—communication between each-other. To increase output, you simply added more units of “capital” and “labor.”
Into Defense
Rational management really hit its stride after WWII and taking the lead was the Department of Defense. In the 1950s, RAND applied systems analyses to formulating objectives and analyzing alternative solutions. Its analysts looked to use quantitative techniques such as optimizations to pre-determine the best course of action. Once the rational analysis determined the program and its associated funding level, it was up to the project office to procure the required system.
To set the project up for success, a long planning period was initiated. The PERT method broke down the project into significant tasks and inter-related their network dependencies. Then, estimates of time and cost were given. In this way, the project execution can be compared to the plan.
If execution falls behind, additional resources can be put into the relevant tasks. Managers could decide, “should I delay the project, or, should I add more labor to make the project on-time but over-cost?” It also assumes the same for requirements. “Technical performance is coming in at 80% of the requirement, but I could get it back to 100% by doubling the engineering, or by giving them another year.”
Then, with historical cost reports of system development and production, analysts can feed the information back into the rational management machine to make better decisions on future programs. This system of rational management devised over the course of the 1950s and 60s continues to exist in the 21st century defense establishment.
Hints from Software
In 1975, Frederick Brooks released a book of essays about software development projects based on his experience in the Department of Defense. He flat-out rejected the assumptions found in rational planning methods. In “The Mythical Man-Month,” Brooks wrote:
“Cost does indeed vary as the product of the number of men and the number of months. Progress does not. Hence the man-month as a unit for measuring the size of a job is a dangerous and deceptive myth. It implies that men and months are interchangeable.”
Brooks argued that industrial era managers reliably increased output by deploying more labor to partitionable tasks. New hires could learn a routine skill and provide added value without needing to understand the rest of the production process. The desired output could therefore be accomplished in less time if additional manpower was put to it, making men and months interchangeable.
However, Brooks found that the relationship does not hold for software development. Adding more labor wouldn’t increase output. He discussed two major problems: uncertainty and communication.
First, uncertainty was apparent in the continual optimism of planners. The probability that the actual course of work goes according to plan is “vanishingly small.” That is because the software development projects, like all R&D projects, explore new concepts and ideas. The projects are not deploying capital and labor in routine ways.
Second, communication in software is more burdensome because each task is largely inseparable from the whole project. Each worker must be trained not just in some specialized area, but a far larger set of technologies, goals, and strategies. That training is specific to each firm and project. Additionally, the workers in software must personally integrate with one another by communication. That means, as the project’s number of workers grows linearly, the number of inter-personal relations grows exponentially.
Brooks concluded that
“… when schedule slippage is recognized, the natural (and traditional) response is to add manpower. Like dousing a fire with gasoline, this makes matters worse, much worse. More fire requires more gasoline, and thus begins a regenerative cycle which ends in disaster.”
Intangibles
Industrial management techniques reflected in DOD acquisition policies have become not only ineffective, but potentially wasteful, in the age of intangibles. This is why we see tech firms limiting project sizes, such as the two-pizza team rule at Amazon. It is also why small tech startups successfully compete with large corporations, an unthinkable proposition to industrial thinkers like J.K. Galbraith and Alfred Chandler.
Researchers Jonathan Haskel and Stain Westlake found that the vast increase in uncertainty and communication resulting from investment in intangible assets stems from four characteristics: scalability, synergies, spillovers, and sunkedness. All are important but just the first and last will be touched on.
Once an intangible asset is invested in, it can increase output at low or zero marginal cost. Consider Uber’s ride-share platform, Google’s search engine, and McDonald’s brand name. The investment is fixed, and it costs nothing to increase output to another unit. Output can grow to conceivably cover the world very rapidly.
Industrial goods, however, had high marginal costs. It took more metal, capital, and labor to build the additional car. The value of the output was well approximated by the cost of production. But intangible assets can service the marginal customer at zero cost. Pricing intangibles is all about recovering sunk costs.
Implications
One might think weapon systems continue to exist in the tangible world. Despite new programs including autonomy and cloud computing, much of the DoD’s investment remains in Cold War era hardware like fighter jets and submarines. For those systems, roughly 70% total costs goes to operations and support, and 20% to recurring production. That means only 10% of weapon systems costs goes to intangible investments in R&D.
This appears to put to DoD at a distance from the broader economy. Whereas tangible assets and earnings explained to more than 80% of firm value in the 1950s, that figure plummeted to about 20% during the period 2000 to 2013. EverEdge investments finds that intangibles account for 87% of all firm value in 2017, which is higher still in the tech industry.
It appears, then, that we are left with more than 80% of private firm value being derived from intangible assets—if you bought the firm, only 20% of the dollars are buying tangible assets and earnings. We also find perhaps only 10% of weapon system value is intangible—if you bought a defense program, about 90% of the dollars are buying tangible output in recurring production and sustainment. So shouldn’t the DoD continue operating to its industrial era techniques?
No. The value of a weapon system is not in bending metal or laying wires, even if that’s where most of the money outlays go. The value is in product design. R&D decisions tightly constrain almost all of the decisions in production and sustainment. The design determines production methods and costs. The design determines reliability, maintainability, affordability, and all the other “-ilities.”
We often hear about how managers should focus attention on where the dollars go. But even though only 10% of money costs go to weapon system R&D, decisions made in that phase contribute to something more like 90% of the total opportunity costs. Weapon system value is derived from intangibles. However, in the phase where decisions really matter—research and development—the Department of Defense continues to manage itself by industrial era techniques, as though men and months are interchangeable.
Rational management by use of business plans and stage-gate approvals has fallen to the wayside. Baruch Lev argues that financial statements matter far less in firm valuation than qualitative strategies. As a result, we find fewer bankers making debt-based investments in firm growth. That baton is passing to equity-investment from Venture Capital, which depends on talent evaluation and deep technical insight rather than the business acumen of assessing revenue streams and tangible assets.
As more of the value from the defense portfolio is derived from intangible assets, and as additive manufacturing and other processes bring tangible costs down, the Department of Defense will need to start devising management strategies that address the new challenges it faces. There are some hopeful signs in programming and contracting, but there is still a far way to go.
Leave a Reply