Spending Intelligence · Acquisition Accountability
Major Defense Acquisition Failures: What Went Wrong and What It Cost
Congress mandated acquisition reform in the 1970s. The programs below failed decades later. The structural causes — requirements creep, concurrency, optimism bias, sole-source dependency — appear in virtually every major overrun. This page documents the record. Every claim is sourced to a public government report.
Sources: GAO-22-104640 · GAO-23-106154 · GAO-21-157 · DoD IG reports · Congressional Research Service · DoD Comptroller
Total MDAP Cost Growth (GAO)
$628B
Above original baselines; GAO-22-104640
Avg. Schedule Delay Across Portfolio
22 mo.
Average across all major programs; GAO-22-104640
DoD on GAO High-Risk List
1990
Year first listed; never removed
DODI 5000.02 Revisions Since 1971
12+
Acquisition policy rewrites; pattern persists
Section 01
Why Programs Fail: The Structural Causes
GAO has identified the same root causes across 30+ years of acquisition reviews. They are structural to the system, not failures of individual programs or individuals.
Requirements Creep
Adding capabilities during development without adjusting cost or schedule baselines. A program approved at a given capability level and cost is a different program if the requirements expand — but the baseline doesn't officially change until a formal rebaseline event. In the interim, the program absorbs added complexity without updated cost authorization. By the time the overrun becomes visible, production is underway and cancellation is politically difficult.
Concurrency: Producing Before Testing
Starting production before developmental testing is complete saves time — and creates retrofit costs when testing finds defects. DoD's own cost models show that fixing a defect during design costs roughly 1×; fixing it in a manufactured unit costs 10×; retrofitting a deployed system costs 100×. The concurrency decision trades known future retrofit costs for schedule acceleration. GAO has documented concurrency as a cost driver in virtually every major program overrun of the past 30 years.
Optimism Bias in Initial Estimates
Initial cost estimates are systematically understated — not always through dishonesty, but through structural incentives. Programs that estimate realistic costs don't get funded. Programs that estimate achievable costs relative to capability offered get approved. The first realistic cost estimate typically comes after a Nunn-McCurdy breach. By then, production is underway. GAO-22-104640 documents that across 61 major DoD acquisition programs, total cost growth is $628B above original baselines — averaging 22 months of schedule slippage per program.
Sole-Source Dependency
Once a contractor is selected for a complex developmental program, DoD's negotiating leverage essentially disappears. The contractor accumulates institutional knowledge, proprietary technical data, and a trained workforce that cannot be replicated without years of effort. Cost overruns in sole-source programs are absorbed — because the alternative is no program. The revolving door compounds this: officials who understand the program's vulnerabilities retire to the contractor.
The Revolving Door, Program-Level
Program managers cycle through 2–3 year assignments. The institutional knowledge of a 20-year program lives with the contractor, not DoD. When a PM retires to the prime contractor, they carry the government's internal assessment of where the program is vulnerable — and the government has lost its most experienced interlocutor. POGO documented that the most valuable post-government hire a defense contractor can make is the most recent program manager for their biggest program.
Revolutionary vs. Proven Technology
Programs derived from existing, proven platforms consistently perform better than clean-sheet revolutionary designs. The P-8 Poseidon (based on Boeing 737) delivered roughly on cost. The F-35 (clean-sheet STOVL) did not. The JLTV (evolutionary Humvee replacement) succeeded. The Expeditionary Fighting Vehicle (revolutionary amphibious assault) was canceled after $3.3B. The pattern is consistent enough to constitute a structural finding, not a statistical anomaly.
Section 02 — Program Case Studies
Five Programs, Five Different Failure Patterns
All figures from public government sources. No allegations against named individuals. Program history is public record.
KC-46 Pegasus Tanker
Boeing
Contract Year
2011
Contract Type
Fixed-Price Development
Original Estimate
$4.9B (development cap)
Outcome / Current Status
$7B+ in Boeing-absorbed overruns; ~$46B total program
The KC-46 Pegasus replaced the aging KC-135 tanker fleet. The 2011 contract was awarded to Boeing as a fixed-price development contract — specifically structured to shift financial risk to the contractor. Boeing has absorbed over $7B in cost overruns on the fixed-price development portion. The aircraft was delivered with significant deficiencies.
What Went Wrong
Remote Vision System (RVS)
The RVS allows boom operators to see the refueling point during air-to-air refueling. It repeatedly failed operational testing due to resolution limitations, depth-perception problems, and image distortion. Category 1 deficiencies (those that could cause death, severe injury, or mission failure) remained open more than four years after first aircraft delivery. Boeing delivered aircraft without a working boom vision system — then had to retrofit them.
Foreign Object Debris
Multiple delivered KC-46s were found to contain foreign object debris (FOD) — tools, rags, manufacturing materials left inside the aircraft. FOD in an aircraft is a serious safety and quality control failure. The Air Force grounded delivery and required Boeing to clean delivered aircraft. The FOD findings reflected systemic quality control failures in Boeing's manufacturing process during this period.
Fixed-Price Paradox
The fixed-price contract structure was intended to discipline cost — Boeing absorbed overruns rather than passing them to DoD. It worked as designed on cost allocation. It did not prevent capability failures. Boeing delivered aircraft that met the cost cap (from DoD's perspective) but failed to meet performance requirements. The fixed price controlled who paid for overruns, not whether the aircraft worked.
Readiness Impact
The Air Force's KC-135 fleet — which the KC-46 is replacing — averages over 60 years old. Delivery delays and operational restrictions on KC-46 have meant the service continues operating aircraft designed before the Vietnam War. Aerial refueling is the enabler that extends the range of every other aircraft in the force; tanker readiness shortfalls constrain the entire air campaign.
Sources: GAO-23-106154 (KC-46 assessment); DoD IG reports on KC-46 Foreign Object Debris; Congressional testimony, House Armed Services Committee, 2022.
Gerald R. Ford-Class Aircraft Carrier (CVN-78)
Huntington Ingalls Industries
Contract Year
2008
Contract Type
Cost-Plus-Incentive-Fee
Original Estimate
$10.5B (FY2008)
Outcome / Current Status
$13.3B+; program ongoing for CVN-79, CVN-80
CVN-78 Gerald R. Ford was the first of a new carrier class, incorporating revolutionary technologies intended to reduce crew size and increase sortie generation rates. Multiple new systems were mandated for first-in-class installation before they had achieved maturity — a textbook concurrency failure at the platform level.
What Went Wrong
Electromagnetic Aircraft Launch System (EALS)
EALS replaced the traditional steam catapult with a linear induction motor system. It was a revolutionary technology mandated on the first ship before it was proven reliable. EALS suffered repeated reliability failures during developmental and operational testing. The carrier was delivered in 2017 but could not conduct full air operations. Years of post-delivery engineering and fixes followed before EALS achieved acceptable reliability. The reliability requirement was 4,166 mean cycles between operational mission failures; EALS initially achieved far fewer.
Advanced Arresting Gear (AAG)
AAG replaced the traditional hydraulic arresting system with a water-turbine energy-absorber system. It was intended to be gentler on aircraft airframes and more capable of handling the full range of aircraft weights. AAG also suffered serious reliability problems post-delivery. The carrier conducted limited flight operations for years after commissioning, partly because AAG could not be trusted for sustained operations.
Dual-Band Radar Cancellation
The Ford-class was designed around a Dual-Band Radar (DBR) integrating the AN/SPY-3 X-band radar with the Volume Search Radar in a single system. The DBR program was canceled due to cost and technical risk after CVN-78. Subsequent ships reverted to proven radar systems, acknowledging that the revolutionary concept was not achievable at acceptable cost.
Readiness Impact
CVN-78 was delivered to the Navy in 2017 and could not perform its primary mission — launching and recovering aircraft at full operational tempo — for years post-delivery. A carrier that cannot conduct flight operations is an $13B asset not contributing to the fleet's combat capacity. The Ford was assigned to operational service in 2022 and deployed to the Eastern Mediterranean in 2023 — six years after commissioning — where its EALS performed acceptably under operational conditions.
Sources: GAO-21-157 (Ford-class carrier readiness); DoD IG Report DODIG-2022-078; Congressional Budget Office carrier cost analyses.
Littoral Combat Ship (LCS)
Lockheed Martin / Marinette Marine (Freedom); General Dynamics / Austal USA (Independence)
Contract Year
2004
Contract Type
Dual-source competitive
Original Estimate
~$220M per ship
Outcome / Current Status
$500–700M per ship; early decommissioning of multiple ships
The LCS program was designed to produce a cheap, fast, modular surface combatant that could be rapidly reconfigured for different missions — anti-submarine warfare, mine countermeasures, surface warfare — by swapping mission modules. Two competing hull designs were funded simultaneously. The mission modules never worked as designed. The ships were decommissioned young.
What Went Wrong
Mission Module Failure
The core innovation of LCS was modularity — a host ship that could accept swappable mission packages. The surface warfare, mine countermeasures, and ASW modules were all significantly delayed and underperformed requirements. The mine countermeasures module — replacing a critical legacy capability — was years behind schedule and never achieved full operational capability during the program's active life. Ships that cannot perform their defined missions are not contributing to the fleet.
No Variant Was Optimized for Anything
The Freedom and Independence variants differed significantly in hull form, propulsion, and systems. Neither was optimized for the modular missions they were supposed to perform. The design compromise required to accept multiple mission modules meant neither variant was as capable in any specific mission as a purpose-built platform would have been. The Navy ended up with two different small surface combatants, neither dominant in its mission area.
Early Decommissioning
In FY2023, the Navy announced plans to decommission multiple LCS hulls that were only 5–15 years old — well short of their 25-year design life. Decommissioning a ship that cost $500M+ within 10 years of commissioning represents a near-total loss of taxpayer investment. The maintenance and reliability record of both variants contributed to the decommissioning decision.
Readiness Impact
The LCS was supposed to address a mine countermeasures capability gap created by retiring the Avenger-class MCM ships and Osprey-class MCMs. It did not. The US Navy's mine countermeasures capability remains a documented readiness gap. A fleet that cannot clear mines cannot safely maneuver in contested waterways — a direct operational shortfall with no current satisfactory replacement.
Sources: GAO-19-384 (LCS Mission Modules); DoD IG Report DODIG-2020-083; Congressional Research Service LCS reports; Navy FY2023 decommissioning announcements.
Army Future Combat Systems (FCS)
Boeing / Science Applications International Corporation (SAIC)
Contract Year
2003
Contract Type
Lead Systems Integrator (total package)
Original Estimate
$92B (2003 estimate)
Outcome / Current Status
Canceled 2009; $18B spent
Future Combat Systems was the most ambitious Army modernization program since World War II — a networked family of 18 manned and unmanned systems designed to replace the M1 Abrams, Bradley Fighting Vehicle, and multiple other platforms. It was canceled in 2009 after $18B in spending without a single system reaching production.
What Went Wrong
Unphysical Weight Requirements
FCS vehicles were required to be light enough to be transported by C-130 aircraft — the standard Air Force tactical airlift platform — while being survivable against contemporary and near-future threats. These two requirements were incompatible. An armored vehicle that can be carried by C-130 cannot carry the armor needed to survive RPG hits, EFPs, and the full range of threats the vehicle would face. The Army knew this tension existed; the program proceeded anyway.
Network Technology Did Not Exist
The FCS concept required a communications network of unprecedented bandwidth and reliability connecting all 18 systems in real time. In 2003, this network did not exist. The program assumed it would be developed in parallel with the platforms. It was not. By 2009, the network was years from maturity, and the platforms it was supposed to connect were still in early development.
Lead Systems Integrator Structure
Awarding the entire FCS program to a single Lead Systems Integrator (Boeing/SAIC) gave DoD one point of contact but eliminated direct government oversight of the 17+ subcontractors delivering individual systems. The LSI structure concentrated information in the contractor team and reduced the government's ability to independently assess schedule and cost. It was, in retrospect, a structural accountability failure.
Readiness Impact
The Army's ground combat vehicle fleet today consists of systems largely designed in the 1970s and 1980s — the Abrams was first fielded in 1980, the Bradley in 1981. FCS was supposed to replace them. Its cancellation left the Army without a modernization path for its heavy ground combat fleet for over a decade. The Army returned to increment-based modernization — the Bradley replacement program (XM30) — after FCS's failure.
Sources: Congressional Research Service R41129 (Army FCS Program); GAO-09-471T (Army FCS: Challenges in Managing the Transforming Army); DoD Comptroller FCS program documentation.
Expeditionary Fighting Vehicle (EFV)
General Dynamics Land Systems
Contract Year
1996 (full-scale development)
Contract Type
Cost-Plus-Award-Fee
Original Estimate
$8.8B total program
Outcome / Current Status
Canceled January 2011; $3.3B spent over 18 years
The EFV was designed to replace the Marine Corps' AAV7 amphibious assault vehicle — a 50-year-old platform. It promised a step-change in amphibious assault capability: a vehicle that could transit from ship to shore at 25+ knots (the AAV7 does 8 knots in water), then fight like a tank on land. It could do neither reliably.
What Went Wrong
Reliability — Catastrophic Failure Rate
During operational testing, the EFV experienced a mean time between operational mission failures (MTBOMF) of approximately 4.5 hours. An amphibious assault requires vehicles to be reliable for the entire transit from ship to shore — which can take 1–4+ hours — plus the subsequent ground fight. A vehicle that fails every 4.5 hours on average is not operationally useful for amphibious assault. The Marine Corps concluded the reliability would cost billions more to fix.
Weight Exceeded Payload
Like FCS, EFV accumulated weight during development to the point where its operational payload was severely constrained. A vehicle heavy enough to be survivable and fast enough to make 25 knots in water requires engineering compromises that reduce what it can carry — troops, weapons, and supplies.
Operating Cost Per Vehicle
Per-vehicle operating costs escalated to the point where the Marine Corps calculated that EFV's life-cycle cost was unsustainable given post-2008 budget realities. Secretary Gates canceled the program in January 2011 explicitly citing unreliable performance and unsustainable cost as the dual rationale.
Readiness Impact
The Marine Corps still operates AAV7s — a vehicle designed in the 1960s with 1970s technology — as its primary amphibious assault platform. The EFV cancellation, combined with subsequent delays in the Amphibious Combat Vehicle (ACV) program, left the Marines in their legacy platform for decades longer than planned. The ACV has since reached operational capability, but the capability gap created by EFV's failure spanned 20+ years.
Sources: GAO-11-372T (Expeditionary Fighting Vehicle); Secretary of Defense Gates statement on EFV cancellation, January 6, 2011; Congressional Research Service EFV analyses.
Section 03
The Watchdog Infrastructure: Who's Watching
Multiple oversight bodies track acquisition programs. Their reports are public. Most service members have never been told they exist.
GAO — Government Accountability Office
Congressional auditor and program assessor
Annual "High Risk List" — DoD acquisition has appeared on it since 1990 and has never been removed. Program-specific assessments (cited by report number throughout this page). Annual Weapon Systems Acquisitions report.
How to Access
All GAO reports are publicly available at gao.gov. Search by report number (e.g., GAO-22-105198) or by program name.
DoD Inspector General
Independent auditor of DoD programs, contracting, and fraud
Program-specific audits, contract fraud referrals, and semi-annual reports to Congress. DoD IG reports have documented cost irregularities in nearly every major program on this page.
How to Access
All public DoD IG reports available at dodig.mil. Reports are numbered in DODIG-YYYY-NNN format.
POGO — Project on Government Oversight
Independent nonprofit watchdog
F-35 tracker, revolving door database, defense contractor misconduct database, and program-specific investigative reporting. POGO has no government affiliation; its analysis is sometimes contested by contractors and DoD.
How to Access
All reports at pogo.org. POGO's F-35 tracker compiles GAO, DoD, and JPO data in one accessible format.
Congressional Research Service (CRS)
Congressional analytical arm
Program-specific analyses, cost-trend reports, and policy option papers on all major acquisition programs. CRS does not take policy positions; its reports document facts and options.
How to Access
CRS reports were not publicly released until 2018 — they are now available at crsreports.congress.gov. Search by program name or R-number.
Defense Acquisition University (DAU)
DoD's acquisition training institution
Teaches acquisition policy (DODI 5000.02), program management, and lessons learned from historical failures. DAU has published case studies on FCS, EFV, and LCS in its acquisition professional curriculum.
How to Access
Case studies at dau.edu. The Acquisition Community Connection contains historical program documentation.
Section 04
What Has Actually Worked
Accountability requires acknowledging success. Not all acquisition programs fail. The programs below delivered roughly what was promised at roughly what was promised. The pattern that distinguishes them is consistent.
MQ-9 Reaper
DeliveredWhy It Worked
Evolutionary from MQ-1 Predator; leveraged existing ground control architecture
Result
Delivered on cost and schedule; proven in sustained combat operations in multiple theaters; operational from 2007 to present
Joint Light Tactical Vehicle (JLTV)
DeliveredWhy It Worked
Competitive acquisition; clear requirements; evolutionary Humvee replacement with defined survivability baseline
Result
Oshkosh won competitive contract; program delivered roughly on budget; now fielded across Army and Marine Corps
P-8 Poseidon
DeliveredWhy It Worked
Derived from Boeing 737-800ERX commercial platform; adapted for maritime patrol / ASW mission
Result
Significant cost growth contained relative to clean-sheet alternatives; aircraft mature by delivery; fully operational with Navy and allied partners
JDAM (Joint Direct Attack Munition)
DeliveredWhy It Worked
Simple concept: a GPS guidance kit bolted to existing dumb bombs; massive scale; multiple contractors
Result
Unit cost fell from ~$40K to under $25K as production scaled; transformed precision strike; hundreds of thousands delivered; operational in every major US conflict since 1999
Virginia-Class Submarine
DeliveredWhy It Worked
Competitive design between Electric Boat and Newport News; detailed design matured before production; block buy contracting
Result
Costs have remained roughly controlled relative to program baseline; submarine delivered at roughly contracted intervals; considered the most capable attack submarine in the world
The Pattern That Distinguishes Success From Failure
Successful programs share a pattern: they start from something that already works and adapt it. The P-8 is a modified airliner. JDAM is a guidance kit bolted to an existing bomb. The JLTV evolved from Humvee operational experience. Virginia-class built on the Seawolf submarine lessons. The programs that fail most spectacularly — FCS, EFV, LCS — are clean-sheet revolutionary designs where the technology required to meet requirements did not exist at program start. This is not an argument against innovation. It is an argument for separating technology maturation from platform production — mature the technology first, then build the platform.
Frequently Asked Questions
What is the Nunn-McCurdy Act and why does it matter?
The Nunn-McCurdy Act (10 USC 2433) requires notification to Congress when a major defense acquisition program's unit cost grows more than 15% above the current baseline ("significant" breach) or 25% above the original baseline ("critical" breach). A critical breach requires the Defense Secretary to certify the program is essential before it can continue. Nunn-McCurdy was intended to create an accountability checkpoint — but the defense industry has learned that programs rarely get canceled after a breach, making the notification more of a speed bump than a stop sign.
What is GAO-22-104640 and what does it say?
GAO-22-104640, "Weapon Systems Annual Assessment: Programs Are Taking Longer and Costing More," is the GAO's FY2022 annual assessment of major defense acquisition programs. It found that across 61 major defense acquisition programs in the DoD portfolio, total estimated costs had grown by $628B over original baselines, and average program schedules had slipped by 22 months. The report is available free at gao.gov.
What is a Lead Systems Integrator and why did it fail in FCS?
A Lead Systems Integrator (LSI) is a contractor given responsibility for integrating all components of a complex system of systems — acting as the single contractor interface DoD deals with, rather than DoD managing dozens of subcontracts directly. In FCS, Boeing/SAIC were the LSI. The theory was that a single contractor could manage complexity better than government program offices. In practice, the LSI structure reduced government visibility into what subcontractors were actually doing and concentrated program knowledge in the contractor. When the program failed, the government had limited independent basis to assess what had gone wrong.
Why don't fixed-price contracts prevent cost growth?
Fixed-price contracts prevent the contractor from passing cost overruns to the government on the fixed-price line items. They do not prevent: (1) the government absorbing cost growth on changes and modifications that expand scope, (2) delivery of non-compliant products that then require retrofit at government expense, (3) the contractor entering financial distress from absorbing losses — which creates its own program risk, (4) cost growth on separately-contracted sustainment and modification work. The KC-46 absorbed Boeing's overruns on development while the government still incurred costs from delayed deliveries and capability deficiencies.
What reforms have been tried and why haven't they worked?
DODI 5000.02 — the DoD acquisition policy document — has been revised more than 12 times since 1971. Congress has passed multiple acquisition reform acts: Goldwater-Nichols (1986), Weapon Systems Acquisition Reform Act (2009), NDAA provisions almost every year. GAO has documented that despite these reforms, cost growth and schedule delays persist at roughly the same rates across program generations. The structural drivers — requirements instability, concurrency pressure, sole-source dependencies, and optimistic initial estimates — are structural to the acquisition system, not problems that policy language alone can fix.
What is DODI 5000.02?
Department of Defense Instruction 5000.02 is the governing policy document for major defense acquisition programs. It establishes the "Big A" acquisition framework — the sequence of milestones (A, B, C) a program must pass before moving from concept to development to production. Each milestone requires a review that assesses technical maturity, cost estimates, and schedule realism. Milestone reviews are the intended checkpoint where programs with unrealistic estimates get corrected. In practice, milestone reviews have frequently approved programs with known cost risks due to schedule and political pressure to proceed.
Explore More
Sources and methodology: Portfolio-level cost growth data from GAO-22-104640, “Weapon Systems Annual Assessment: Programs Are Taking Longer and Costing More” (GAO, 2022). KC-46 data from GAO-23-106154 and DoD IG reports on foreign object debris and Remote Vision System deficiencies. Gerald R. Ford-class data from GAO-21-157 and Congressional Budget Office carrier analyses. Littoral Combat Ship data from GAO-19-384 and Congressional Research Service RL33741. Army Future Combat Systems data from GAO-09-471T and CRS R41129. Expeditionary Fighting Vehicle data from GAO-11-372T and Secretary of Defense Gates cancellation statement (January 6, 2011). Acquisition reform history from Defense Acquisition University case studies. DODI 5000.02 available at dtic.mil. All named programs are cited from public government record. No non-public information is used. Dollar figures are in then-year dollars unless noted. This page reflects facts documented in public record and does not constitute legal, investment, or policy advice.