Political theory matters. As Harry Jaffa eloquently argued, it mattered greatly that Abraham Lincoln opposed allowing slavery in the territories on the basis that all men are created equal. The structures and preferences that have accumulated over time in America should always be weighed against the principles articulated by our republic’s Founders.
But to effectively bring those principles to bear on today’s issues, we need to understand not only the more obvious corruption that fuels the Beltway—we also need to recognize how overreach is nurtured by seemingly non-ideological processes that have become deeply woven into today’s federal government.
We face a series of cascading failures in sectors like defense and national security that pose serious economic, geopolitical, and related dangers to the U.S.’s stability and international presence.
NASA announced in August that Boeing’s Starliner space capsule would return to Earth without its crew. That marked a dismal milestone in a $4.2 billion program that finally achieved a flawed launch to the International Space Station seven years late, and with a budget overrun of $1.6 billion. Despite the escalating cost, Starliner suffered thruster and other major failures, and NASA finally decided not to put the crew at major risk by returning them in the capsule.
In January NASA announced another nine-month delay in Lockheed Martin’s Artemis moon vehicle, which is now two years behind the schedule set in 2019. In 2012 NASA estimated that a moon mission would cost $500 million per launch—in 2022, the estimate ballooned to $4.1 billion.
Congressman Ken Calvert recently disclosed that the U.S. Navy’s program to develop and build a new Virginia class nuclear-powered SSN-774 attack submarine fleet continues to be two to three years behind schedule. It’s experiencing “extraordinary cost growth” of at least $17 billion over its previously projected $184 billion estimate. The lead contractors for this program, which has been underway since initial prototyping in 2001, are General Dynamics and Northrop Grumman. This mess follows on the multiple delays and problems with the Navy’s littoral combat ship program.
These and other examples illustrate significant problems with major federal procurements. But the problem goes deeper and wider, to the procurement process itself and the Beltway ecosystem it has spawned.
Building the Machine
To see the depth of the problems inherent in the federal procurement process, we need to step back to the 1980s. Digital technology burst out of labs and defense research and development, forever changing how we communicate and do business. Cell phones were built on the Army’s work in packet radios. The Internet was the public instantiation of a communications method originally invented to allow senior national leaders to remain in contact with one another in the event of a major nuclear attack on the U.S. Computing chips were becoming mainstream, rapidly packing increasing power into ever-smaller units, enabling both personal computers and powerful software designs to be implemented for daily use. In 1984 the telephone industry was deregulated, and cell towers were connected to existing land telephone lines, which were rapidly converted to high-speed cables, allowing landline phones and cell phones to interoperate.
As this tech improved and its application began to spread, the federal government created the Software Engineering Institute, a new federally funded research and development center (FFRDC) associated with Carnegie Mellon University, in 1984. The Institute’s focus was (and still is) on management practices associated with computing technology, engineering methods for its development, and cybersecurity. Soon the new discipline of software engineering was defined—a structured way to move from requirements, to design, to implementation, to the testing and validation of software systems.
At the same time, new higher level software programming languages were invented, allowing code to be written in ways that reflected end software functionality rather than the hardware level operations inherent in a chip CPU. Compilers and other tools were created to translate high-level code to a particular operating system and hardware environment, allowing third-party companies to offer the same software for various systems. Lotus 1-2-3, and then Excel, allowed lower- and mid-level corporate employees to save, analyze, and visualize operating data without waiting for central IT organizations to manage their data for them.
What rapidly emerged was a complex ecosystem of interacting software and devices, each with its own functionality but able to communicate with and exchange information with others, locally or at a distance. Centralized, coherently-designed large systems—including military equipment—increasingly functioned within such an ecosystem, incorporating subsystems that were often already in existence or were needed to function as part of a larger technical whole.
Systems Over Mission
Cue the new importance of a recently developed discipline: systems engineering. Wikipedia’s definition is reasonably good:
Systems engineering is an interdisciplinary field of engineering and engineering management that focuses on how to design, integrate, and manage complex systems over their life cycles. At its core, systems engineering utilizes systems thinking principles to organize this body of knowledge. The individual outcome of such efforts, an engineered system, can be defined as a combination of components that work in synergy to collectively perform a useful function.
Systems engineering explicitly provides methods for building and operating complex systems made up of many components, some of which may already exist. Although it is in use in commercial industry to some degree, it found its natural home in defense and national security systems specification, design, and implementation.
Consider, for instance, a new fighter jet. Its ecosystem includes communications, radar, weapons, launch, and other systems, some of which it must host and all of which must be compatible. They must also fit into, or explicitly require modifications to, existing operating procedures for those elements. And they must either fit into, or explicitly modify, existing military tactics, training, and doctrine. Those factors shape the requirements which a new fighter’s design and implementation must meet.
As technology advanced quickly, and in response to a major corruption scandal involving the development of a new Navy vessel, the Federal Acquisition Regulations (FAR) and the highly complex Defense Federal Acquisition Regulations (DFAR) were also greatly expanded. Today the FAR and DFAR specify structured processes by which federal agencies acquire systems and services.
This process is as follows: an agency generally crafts a Request for Proposals (RFP) after internal determination of the purpose, scope, likely budget impact, and other aspects of the system or service to be desired. For most acquisitions, bidders then submit formal proposals that outline what they propose to provide, and on what schedule and at what overall cost. Cost estimates for DOD and similar procurements are shaped by the RFP, which specifies whether the agency wants a time and materials bid, where the agency agrees to pay invoices up to some project total maximum; or a cost plus pricing, in which time, materials, and overhead expenses are reimbursed after labor and overhead rates are approved; or a firm fixed price bid, in which the contractor proposes an overall cost for the acquisition. In all three cases the agency will audit some information to assure itself that the bid is reasonable—that is, that the contractor hasn’t significantly underpriced the bid.
As this simplified description suggests, the FAR/DFAR allow an agency a wide range of bid selection approaches with the intent that the one chosen aligns with the nature of the system or service to be procured. If a major system is being developed or significantly modified, there are many interim deliveries and approvals as the system’s detailed requirements, design, implementation, and testing proceed. Payments are often made incrementally as phases are successfully approved.
The name “systems engineering” and the discipline’s parallels in the FAR/DFAR acquisition regulations carries a strong implication that what is being done is engineering and not investigative innovation. RFPs for acquisitions such as military equipment include requirements that limit acceptable responses. Given this, you might expect that by the 1990s a major acquisition of new, very expensive mission critical systems would, according to systems engineering principles, be a great success. At least in the case of the Future Imagery Architecture (FIA) program, which tried and failed to develop a new constellation of advanced space-based reconnaissance satellites, you’d be wrong.
FIA-related contracts were awarded by the National Reconnaissance Office (NRO) over the course of several years. (The NRO’s existence had only recently been made public after the Washington Post revealed the classified agency’s purpose, name, and office location in the outer D.C. suburbs.) Raytheon developed the ground systems, and Boeing led a team to develop the satellites. NRO assembled a team of experts to augment its own staff in drafting the FIA RFP and selecting awardees. Among their team were experts from the Aerospace Corporation, an FFRDC that specialized in space-related systems.
A key mantra within DOD, the intel community, and related agencies at the time was “Faster, Better, Cheaper.” This reflected the belief that technology and processes had matured sufficiently to allow cost to be a driving factor in procurement decisions for space systems. Lockheed Martin’s proposal would have involved specially designed subsystems and was estimated to cost about $1 billion more than Boeing’s, which proposed to use many commercially available components and subsystems.
The FIA procurement process applied another, new discipline: multiple objective decision analysis (MODA). If engineering starts with specific requirements a new system must meet, how does an agency decide on those requirements? For some procurements that might be straightforward. But many procurements—and many major commercial decisions—involve multiple stakeholders that have different priorities and objectives. In other words, there are objectives for the system that are in tension with one another, and sometimes in outright conflict.
In 2002, prior to the FIA program’s cancellation, the journal Military Operations Research published a paper by the Aerospace team members, “Performance Analysis in the Selection of Imagery Intelligence Satellites,” regarding the use of MODA and detailed performance analytics during the FIA procurement decision. It’s an impressive report that provides in-depth analysis to estimate the likely performance success of each of the two bidders’ approaches. And yet the program was a failure, in large part because Boeing was heavily favored based on underestimated costs and the substantial challenges involved in meeting the necessary space and intel-gathering operational needs using off-the-shelf components.
SpaceX 1, The Rest 0
Fast forward to 2024. Despite extensive prior use of the International Space Station and capsules carrying astronauts (complete with key, detailed engineering requirements that each capsule design must meet in order to dock, unload personnel and supplies, etc.), Boeing’s Starliner suffered major thruster and other problems. Boeing has also suffered massive financial losses along the way.
In contrast, SpaceX racks up success after success with Starlink satellite launches and regularly upgrades their capabilities, as well as conducts successful launches of federal satellites. For its private Polaris Dawn manned high orbit mission, for example, SpaceX designed all the equipment and provided a key crew member.
Why is Space X succeeding where Boeing and Lockheed Martin are stumbling with regard to tech performance, schedule, and cost?
There are two key factors here. First, Elon Musk has a major overriding goal for all his companies’ innovations. He intends to help humans colonize Mars. To get there, the mission will require many innovations: space launch, capsules, communications, power generation and use without fossil fuels, and more. SpaceX meets current needs for paying customers while constantly advancing the capabilities of their systems at their own expense toward this long-term goal. The Mars mission also envisions highly complex interacting systems, many involving innovations that are not yet made or matured.
To put it simply, SpaceX is mission oriented.
A top-down engineering approach can only be usefully applied to individual equipment and systems when some of the technical tradeoffs have been identified and scoped. Instead of starting with systems engineering, SpaceX applies SE within the scope of rapid, iterative prototyping of key subsystems.
For example, their heavy launch vehicle is now on its third refinement of its large booster engines. They’ve also refined the ocean-based robotic landing platforms currently used by their Falcon 9 rockets—which at present are the only regularly reused space launch rockets in existence. To meet its long-term, highly ambitious Mars goals, SpaceX incrementally proceeds toward various capabilities they require. They test assumptions and approaches, analyze failures, incrementally adjust, and try again.
But the federal ecosystem has become massively process oriented. Procurement processes have become institutionalized. Formal certifications of many levels in procurement-related topics can be earned by personnel at classes paid by taxpayers and often taught by third-party organizations, which receive contracts for this purpose. Holding these certifications opens the possibility of job promotions both within agencies and companies. Companies bidding on federal procurements tout the key employees who earned them. Not only do the certifications add credibility to a bid—they can justify a higher labor category (and therefore higher charge) for employees who hold them. It’s a self-reinforcing positive feedback loop.
Along the way a number of professional societies focused on systems engineering and related topics serve as meeting places where contractor personnel and federal employees mingle, and where both can give talks and papers that also enhance resumes, promotions, and contract awards.
These mechanisms and incentives play a key role in the metastasis of the federal bureaucracy and the wider Beltway ecosystem. And yet this is all invisible to political theorists who are trying to understand what’s fueling our march away from constitutional government.
Engineering Defeat
The top-down engineering mindset permeates agencies well beyond DOD and the intelligence community. Increasingly we see actions justified on the basis that they will fix the supposed climate crisis, social inequality, and more. The Beltway ecosystem is built upon the assumption that achieving supposedly just and equitable goals is simply a matter of engineering solutions. And the denizens of the D.C. blob aren’t even sophisticated enough to take a disciplined systems engineering approach.
As a result, the swamp boasts repeated, costly failures across many domains. Lofty goals are assumed to justify authoritarian executive branch actions, and the process far outweighs any evaluation of success at meeting putative goals. Starliner is just one symbol of the much more widespread, dangerous fantasy that has metastasized in the Beltway ecosystem, threatening to collapse our society, economy, and basic national defense.
The Beltway ecosystem is failing us in a deeper way as well. By refusing to acknowledge that many policy domains entail multiple objectives that are often in tension, and for which different stakeholders have differing priorities, our federal agencies have repeatedly overridden the principles of federalism and limited government the Founders very wisely implemented, informed by several centuries of often-unaccountable central power. Advocates of limited government would do well to consider seriously the insights MODA offers, starting with the basic fact that in many major policy areas various stakeholders have differing priorities and objectives.
It is long overdue for the ecosystem that has metastasized in and around the Beltway to be pared back and for a much more modest, incremental approach to be adopted in the pursuit of lofty innovation and social goals. To achieve that we need to take seriously the degree to which that ecosystem has long since been detached from measuring mission success in favor of processes that benefit all the Beltway players.