The $300B/Year Procurement System
That Keeps Breaking

Requirements Creep

The requirements process has a structural incentive toward expansion. Each service, each combatant command, and each functional community that reviews a requirements document has an opportunity to add capabilities. Adding a requirement is costless to the person adding it. The cost is borne by the program office that has to engineer, test, and produce it. The F-35 began as a joint, affordable, multirole strike fighter to replace several legacy aircraft at dramatically lower unit cost. Over two decades of requirements refinement, it accumulated stealth, supersonic performance, STOVL capability, carrier compatibility, nuclear weapon integration, advanced sensor fusion, and electronic warfare capability simultaneously across three variants. Each addition was individually justifiable. Collectively, they produced the most expensive weapon system in human history.

Spiral Development and Lock-In

Programs are frequently approved at lower cost and capability than ultimately delivered, using a concept called "spiral development" — the system is fielded in increments, with each increment adding capability. The theory is sound. The practice is that Spiral 1 establishes an industrial base, a contractor team, a government program office, and a congressional constituency. By the time Spirals 2 and 3 arrive with expanded requirements and higher costs, canceling the program would eliminate jobs in multiple congressional districts, strand significant sunk-cost investment, and leave a capability gap. The political economy of large defense programs makes them nearly impossible to cancel once the industrial base is established — even when the program is failing.

Congressional Pork and Constituency Politics

Defense programs are not just weapons. They are employment programs, and every member of Congress whose district hosts a prime contractor, a subcontractor, or a supplier has a direct financial incentive to keep the program funded. The C-17 Globemaster III cargo aircraft is the clearest example: the Air Force declared it did not need additional C-17s in 2009 and requested program closure. Congress continued funding production for two additional years, adding aircraft the Air Force said it did not want. The Boeing plant in Long Beach, California, employed thousands; the political calculus was straightforward. This is not an aberration. It is the norm. The DoD "unfunded priorities list" — the list of things the military actually wants — regularly differs from what Congress actually funds.

"Too Big to Fail" Prime Contractors

Lockheed Martin, Boeing, Raytheon Technologies (now RTX), Northrop Grumman, and General Dynamics are the five prime contractors that dominate the DoD industrial base. Their workforce, specialized facilities, technical expertise, and cleared personnel are genuinely irreplaceable on a short timeline. DoD knows this. The primes know this. When a major program encounters cost growth, schedule slippage, or technical failure, the honest market response — termination and competition — is rarely available. Walking away from a failing Lockheed or Boeing program risks destroying cleared manufacturing capacity that took decades to build and could take a decade to reconstitute. This structural dependency gives prime contractors negotiating leverage that no competitive market would tolerate.

The Valley of Death

Small, innovative companies — often the source of genuinely disruptive technology — face a structural trap in DoD acquisition. A DARPA or Army Futures Command program manager discovers a promising technology at a startup. The startup wins a $2–5M Other Transaction Authority (OTA) prototype contract. They deliver a working prototype. Then the program must transition to a Program of Record — a formal acquisition program with a budget line, a prime contractor team, ITAR compliance infrastructure, cleared facilities, and the ability to respond to the full weight of Defense Federal Acquisition Regulations (DFAR). Most small companies cannot survive the 5–7 year gap between prototype success and production contract award. The technology either dies, gets acquired by a prime (who may or may not develop it), or gets copied by the prime who then wins the production contract. The Valley of Death is why COTS technology routinely outperforms DoD-developed alternatives at a fraction of the cost.

The most expensive defense program in US history. Mission capable rates have never hit the program target. The F-35B STOVL variant weight crisis in 2004 triggered the first of multiple program restructures. Unit cost grew from ~$81M to ~$162M (flyaway cost of ~$82M). See the full F-35 breakdown at /spending/f35.

The LCS was conceived as a fast, affordable, networked surface combatant to replace three legacy ship classes. It would cost less than $220M per hull. When actual costs emerged, the program split into two competing designs (Freedom-class and Independence-class), which increased overhead without achieving competition savings. The Navy's own testing found the ship could not survive in a "moderate threat environment." Ships are being decommissioned well short of their planned service lives.

The KC-46 was supposed to demonstrate that fixed-price development contracts could discipline prime contractor cost growth. Boeing bid low and won. When technical problems multiplied — most critically, a remote vision system (RVS) that couldn't reliably see the receiver aircraft — Boeing absorbed $7B+ in losses rather than let the government escape. The Air Force received aircraft with known Category 1 deficiencies because they needed tankers. The RVS 2.0 fix was promised for years. As of 2024, the aircraft still has unresolved deficiencies.

DCGS-A was the Army's intelligence fusion system — designed to aggregate signals intelligence, imagery, human intelligence, and other sources into a single analytical platform for brigade and division staff. Soldiers who used it described an unintuitive, crash-prone system that required extensive training to produce results that an analyst with commercial tools could produce faster. Units deployed with unofficial workarounds. After spending over $6B, the Army ultimately moved toward commercial platforms. The system exists in Army unit T/O&E today but is widely regarded as acquisition malpractice.

The Air Force spent roughly a decade and over $1B studying how to replace the aging E-8 JSTARS ground surveillance aircraft. Multiple industry teams competed. Then the program was cancelled — not because the mission went away, but because the analysis concluded that distributed UAS and long-range radar systems could perform the ground moving target indicator (GMTI) mission better, cheaper, and more survivably. The money spent on requirements and concept development bought information, but that information could have been purchased faster. The JSTARS replacement gap remains partially unfilled.

IVAS promised soldiers an augmented reality headset providing digital maps, blue force tracking, thermal sensor display, and squad communication overlay on a single wearable device. The technology is real and the capability is genuinely valuable. What happened instead: soldiers in operational testing reported dizziness, nausea, balance problems, and visual interference in ambient light. For two years, soldier feedback was documented and ignored as the program pressed toward fielding. In 2023, after sustained congressional scrutiny, the Army restructured the program. The $21.9B ceiling contract remains, but the path to that number is now uncertain.

Established by the FY2022 NDAA, the Commission on Planning, Programming, Budgeting, and Execution Reform submitted its final report in March 2023. Key recommendations: separate the budget programming timeline from the acquisition execution timeline to reduce the 18–24 month funding lag; create a new "technology transition" budget account to fund the Valley of Death gap between prototype and program of record; reduce classified budget overclassification that prevents Congress from exercising effective oversight. As of 2024, implementation is partial. The biennial budget recommendation has not been adopted. The technology transition account concept has moved into discussion but not statute.

Section 804 of the FY2016 NDAA created "middle tier" acquisition authorities — rapid prototyping and rapid fielding pathways designed to deliver capability in 5 years or fewer. The rapid prototyping pathway was meant for programs that needed a working prototype within 5 years. The rapid fielding pathway was for mature technology that could be produced and fielded within 5 years using proven manufacturing processes. Both authorities exist and are technically available to every program manager. Both are chronically underused because program managers are evaluated on compliance with traditional acquisition milestones, not speed. The institutional incentives have not changed to match the statutory authorities.

Army Futures Command was established in Austin, Texas in 2018 — deliberately located away from Washington D.C. to reduce bureaucratic inertia. AFC's Cross-Functional Teams (CFTs) are purpose-built groups combining requirements officers, program managers, operational users, and industry representatives to accelerate development. Early results have been mixed but genuine. Long-Range Precision Fires (LRPF) has produced ATACMS follow-on capabilities faster than traditional timelines. Next-Generation Squad Weapon (NGSW) — the XM7 rifle and XM250 automatic rifle — moved from requirements to initial fielding in approximately 5 years, fast by historical standards. The IVAS headset remains a cautionary tale. AFC represents genuine institutional reform; it is not a complete solution.

DIU was established in 2015, elevated to reporting directly to the Secretary of Defense in 2022. It is the U.S. government's primary mechanism for accessing commercial technology at speed using OTA. DIU has fielded programs in commercial satellite imagery (with Planet Labs, Maxar), AI-enabled logistics, cybersecurity, and unmanned systems. Its contracting timelines are genuinely faster than traditional acquisition — months, not years, for prototype agreements. Its limitation: DIU represents less than 1% of DoD procurement spending. It can demonstrate that fast acquisition is possible. It cannot transform a $300B/year procurement system through demonstration alone.

The 2020 Adaptive Acquisition Framework revised DoD Instruction 5000.02 to create six distinct acquisition pathways (Major Capability, Middle Tier, Software, Defense Business Systems, Acquisition of Services, and Urgent Capability Acquisition) rather than forcing all programs through one process. The framework is sensible. The problem is that most program managers still default to the Major Capability Acquisition pathway because it is the most documented, the most familiar, and the one for which career patterns and oversight structures have been built. Structural reform requires changing human incentives, not just creating new pathway options.