Infernos raged aboard Japanese aircraft carriers after U.S. Navy dive-bombers found their marks during the Battle of Midway. Many Japanese pilots were incinerated in ready rooms and in the cockpits of their aircraft while they sat idling on flight decks. Those who were airborne returned to find their carriers aflame. They circled until fuel exhaustion and ditched into the sea, and many were never recovered. For Japan, the loss of its carriers was damaging, but the loss of its naval aviators was catastrophic. Her navy never recovered.
Japan overinvested in individual pilots, overconcentrating capability into an exquisite cadre at the expense of regeneration and pilot training throughput. Japan spent substantially more per pilot than the United States. Resources that might have supported higher training throughput were instead devoted to maximizing individual pilot skills, sacrificing system-level regenerative resilience. Many Japanese pilots never had the opportunity to employ their exquisite skills when it mattered most, and their fate was inseparable from the survivability of the fleet itself.
Importantly, once aviator competency was lost, it did not exist in sufficient numbers to regenerate Japan’s training system. This permanently constrained training throughput, even after Japan tried to expand pilot production, underscoring the brittleness of Japan’s model under attrition. Without regeneration, sustaining combat power increasingly depended on preserving the existing cadre, demanding survivability beyond what piloting skill alone could provide.
This story offers a valuable lesson about investing in gold-plated platforms or manpower models. From the overspecialization of the phalanx to investment decisions in prestige capital warships, history provides ample evidence that overinvestment in a capability can accelerate military decline by diverting scarce resources to protect investments at the expense of system regeneration and resilience against attrition.
At present, U.S. industrial realities create a strong temptation to compensate for numerical deficiencies by concentrating capabilities into fewer, over-specialized, and capital-intensive platforms, such as advanced aircraft and large warships. This logic produces a vicious cycle: scarcity drives concentration, concentration incentivizes survivability, survivability increases costs, and rising costs further constrain force size. The result is a force optimized for rapid, decisive engagements, but increasingly brittle and prone to failure in protracted conflict when attrition rates and survivability assumptions diverge from reality.
Concentrated Capabilities
Entering World War II, Japan’s naval aviation system appeared to be superior to that of the United States in many respects. In fact, in 1941, Japan’s naval aviators were the best in the world. Maneuverable aircraft and a selective training pipeline produced competent aviators who possessed more flight hours than their U.S. adversaries. Despite the initial disadvantage, U.S. pilots soon adapted and introduced tactics that reduced Japanese battle success rates, diminishing the marginal battlefield value of the additional training Japanese pilots received. But most importantly, America focused on generating aviators at scale, while systematically incorporating battlefield feedback. By 1943, the U.S. was producing approximately 20,000 naval aviators annually. In contrast, Japan produced about 5,000 aviators across its army and navy during the same period. Japan’s naval aviation cadre demonstrated selection and completion rates comparable to those of contemporary special forces, such as the Navy SEALs — forces optimized for localized, specialized operations but not for warfare at scale.
As the war continued, America’s position became increasingly dominant, while Japan’s withered as losses outpaced its ability to regenerate well-trained aircrew. Regeneration shortfalls were further exacerbated by a lack of fuel and ammunition for training units. Back at home, Japan lacked the means to expand its flight training pipeline after the losses of veteran pilots, and at sea, the remaining cadre of experienced pilots was condemned to fly against U.S. pilots with ever-longer odds. This stark reality ultimately drove Japan to pursue “kamikaze” tactics with minimally trained pilots.
Throughout history, systems that relied on gilded components, whether exquisite platforms or elite manpower, have often proven decisive in limited engagements but costly and irreplaceable in protracted conflict. The success of the Greek, and later Macedonian, phalanx encouraged continued investment and specialization in localized lethality and survivability. Over time, the formation grew tighter, more rigid, and increasingly dependent on cohesion and favorable conditions. When confronted by the more mobile and regenerative Roman legion, which did not conform to phalanx employment assumptions, the military system became untenable under sustained attrition. Despite its lethality and formation survivability, attritional warfare eroded trained warriors faster than they could be regenerated, and the lack of mobility limited the ability to select favorable terrain for engagements.
A similar dynamic reemerged in early modern Europe, where 17th-century Spanish tercios failed as a military system when employed against mobile French and Dutch forces, violating core assumptions about set-piece battles, time to form, and long personnel training pipelines. Under sustained attritional pressure, the system could not absorb losses at a rate consistent with regeneration. In contrast, Gustavus Adolphus, the Swedish king and military reformer of the early Thirty Years’ War, deliberately traded individual survivability and exquisiteness for system resilience. He democratized heavy artillery by making it lighter, more numerous, and easier to replace, imposing battlefield losses on imperial Habsburg armies that exposed the brittleness of older formations and accelerated doctrinal change across Europe.
During World War II, Germany and Japan built large, specialized battleships with the mission of destroying enemy battleship fleets. However, these ships sapped industrial resources that could have been deployed elsewhere, such as destroyer production and shipyard standardization for increased output. The German battleship Bismarck was sunk in the North Atlantic on its maiden voyage in May 1941. Its sister ship was confined to Norway to serve as a fleet in being by the now more cautious Germany navy. Three years later, it was sunk in a Norwegian fjord. Japan’s Yamato battleship spent long periods husbanded in harbor awaiting opportunities for decisive engagements. Japan also increased investment in anti-aircraft armament to enhance survivability. Despite these measures, it was sunk by U.S. carrier aircraft anyway in 1945. Aversion to routine or discretionary employment steadily eroded the platforms’ value to the war effort. Debates surrounding U.S. aircraft carrier deployments against modern threats today provide a contemporary analogy.
Putting the “Sunk” Into Sunk Costs
Military systems comprise platforms, people, communications, and sustainment. With finite resources, the balance is zero-sum. The greater the investment in one component, the more important it is relative to others. Rational actors compensate by making the component more survivable to protect the investment, creating a self-perpetuating paradox. Greater survivability demands greater technical sophistication; greater sophistication drives higher cost; higher cost reduces available resources; and fewer resources increase the need to further ensure the survivability of remaining concentrated capabilities. This dynamic is most often observed in platforms such as ships and aircraft and results from rational actions to address perceived deficiencies.
Platform-centric thinking and sunk cost fallacy can exacerbate and accelerate this dynamic, as the source of military strength is measured by platforms rather than systems. If the military cannot obtain the quantities it needs, it should husband resources by investing further in survivability and risk-mitigation measures.
For example, Germany’s pursuit of platform excellence emerged as a strategic response to unfavorable industrial realities. Germany sought to “compensate through qualitative superiority and tactics,” believing that “technological superiority” could offset “numerical advantage.” This logic pervaded Germany’s force structure and further exacerbated industrial shortfalls rather than alleviating them, as industrial production was tied up in expensive capital assets and in promised war-winning Wunderwaffe (“wonder weapons”) that never arrived.
Germany’s ill-advised logic echoes with contemporary U.S. force design. During the Cold War, the United States also sought to offset the Soviet numerical advantages with superior platforms. Designs were generally informed by the required exchange ratios that eroded enemy combat effectiveness at a cost below or equal to the friendly replacement rate. The danger lies in repeating Germany’s mistake, where assessments of combat performance skew toward platform-versus-platform comparisons rather than holistic fleet-on-fleet or force-on-force interactions, or, in modern terms, systems-versus-systems competition. In systems warfare, realized combat power emerges from multiple complementary operational systems — such as firepower, command and control, and reconnaissance — integrated to generate desired outcomes greater than the sum of their parts.
Systems warfare concepts are abstract and difficult to quantify because they comprise numerous, disparate components controlled by different agencies, disciplines, and military branches. It is easier to measure performance using familiar, locally controlled metrics, such as platforms. The U.S. Air Force knows how to define a requirement for a bomber, the Navy for a submarine, and the Army for a tank, yet it requires a higher degree of coordination, negotiation, and alignment to design and sustain a complex strike-reconnaissance system that fuses intelligence from satellites and real-time communications with the right weapon at the right location and time. At their most effective, none of these platforms operates independently.
The Survivability Paradox
To meet the required exchange ratios for sustained conflict, an effective force should be optimized across cost, quantity, and individual platform capability. This is where the business side of defense becomes decisive, as aggregate capability across a fleet, force, or system is the true measure of effectiveness. To illustrate the point, we conducted a set of simple analyses to examine the relationships among quantity, threat lethality, and combined fleet effectiveness. We modeled a fleet of simple, one-way attack drones based on the Low-Cost Uncrewed Combat Attack Systems, using two fixed, non-stochastic (deterministic-event) operational scenarios. Each target was assumed to require a single drone, and effectiveness was measured by the total number of targets successfully engaged.
Figure 1: The survivability paradox
The survivability paradox describes a self-reinforcing cycle in which increasing platform capability raises its perceived importance, driving higher survivability requirements, higher costs, smaller fleets, and ultimately a reduced ability to absorb losses.
The drone can be equipped with various enhancements to improve individual capability, each intended to increase survivability against an enemy defense network. We evaluated two scenarios by varying the assumed effectiveness of these enhancements to observe system responses. These enhancements included an optics package, an autonomy software package, low-observability treatments, and an active-countermeasure suite. The autonomy software package enables execution of defensive maneuvers if a threat is detected, but requires the optics package to detect incoming threats. Software is expensive to develop, and we assumed that a significant portion of the notional $400 million fixed budget ($150 million) would be consumed by its development, performance monitoring, and updates. We also credited the effect of mass, where greater quantities saturate defenses and increase realized survivability at the fleet level.
Figure 2: Targets Serviced versus Individual Platform Cost
For individual platforms operating in isolation against an enemy defense network, survivability enhancements appear necessary, with a peak before diminishing returns.
Figure 2: Targets Serviced versus Individual Platform Cost
Figure 3: Capability Regions
Capability enhancements are both price-sensitive and effectiveness dependent. When the efficacy of the individual enhancements was altered, the optimal configuration shifted. On the left, attrition from enemy defenses outpaces the gains in quantity from lower complexity. On the right, excessive cost constrains quantities so severely that targets go unserved.
Figure 3: Capability Regions
Figure 4: The Effect of Unconstrained Mass
When the attacker can employ the entire fleet simultaneously, survivability enhancements become detrimental and the optimal force design consists solely of the basic drone model. The takeaway is that when mass can be deployed at sufficient scale, it can overwhelm a threat system without reliance on individual platform survivability.
Figure 4: The Effect of Unconstrained Mass
Figure 5: Capability Regions with Salvo Size Capped
Build timelines, fielding rates, command-and-control constraints, and the need to husband forces over a prolonged conflict all impose practical limits. Once the salvo size is limited to 300 drones per attack wave, the optimized fleet composition shifts, revealing the core survivability paradox.
Figure 5: Capability Regions with Salvo Size CappedWhat happens if the defense industry cannot produce the required quantities, leaving excess budget? For example, if only 600 drones can be made, should the remaining funds be invested in making the remaining fleet more survivable? The answer is likely no, as the system exists to provide scale, and investing further falls into the survivability paradox and sunk cost fallacy. To enhance the survivability of a limited system, resources of expertise, time, and raw materials are diverted from other relevant combat systems. Instead, the cheap solution becomes a newly minted gold-plated one, where the benefit-to-cost ratio skews unfavorably.
Distributed Concepts: Is it Better?
The U.S. defense manufacturing base and its critical supply chains have diminished since the Cold War, with constraining dependencies that did not exist in previous generations. Cost growth driven by production inefficiencies, supply chain fragility, and lost industrial capacity prevents the necessary scaling required to achieve an optimized force size, as evidenced by several Government Accountability Office assessments and think tank reports. For example, the Navy’s fleet size has collapsed since the end of the Cold War and has stagnated at under 300 vessels. At the same time, orders for advanced capabilities across the services were curtailed, including truncated procurements of the Zumwalt-class destroyer, the B-2 Spirit strategic bomber, and now the Constellation-class frigate program.
These reductions in force size increase the required exchange ratios for the remaining platforms, raising their survivability and effectiveness requirements. Combined with the proliferation of advanced threats that elevated forecasted friendly attrition, U.S. planners were forced to embrace distributed force concepts.
Distributed concepts call for capabilities spread across unmanned systems and less complex combatants to augment existing operational systems, known as hybrid fleets. Internal studies at the Naval Postgraduate School reached consistent conclusions: such fleets are cost-competitive and more survivable than contemporary force design constructs. For example, hybrid fleets modeled against a Russian threat survived across all modeled scenarios and achieved kills on enemy combatants at a 40 percent discount relative to a contemporary fleet. Another study modeled in the Pacific theater reached similar conclusions, reinforcing the generality of these findings. Published analysis drawing on some of those models and studies echoed their analysis – hybrid fleets with a mix of unmanned and manned vessels are more effective.
The recent curtailment of the Constellation-class frigate program and the shift toward a trimmed-capability, repurposed National Security Cutter hull demonstrate a tentative embrace of fleet-level survivability through the planned acquisition of more hulls. However, an institutionalized preference for large, efficient, and high-capability platforms persists. The latest example is the announcement of the proposed Defiant-class battleship. The cost, size, and complexity of such a warship, with estimates ranging into the tens of billions of dollars amid constrained shipyard capacity, may limit the number eventually launched. Moreover, concentrating firepower and national treasure at this scale induces levels of operational risk-aversion comparable to those associated with aircraft carriers and other high-value units. A capability that cannot be risked in combat is, by definition, removed from aggressive offensive action and cannot be fully employed.
Old Habits Die Hard and History Rhymes
The danger the U.S. military now faces is that, during a period of recovery from self-inflicted damage to its defense industrial ecosystem, the temptation to load platforms with additional capabilities is growing and, in some cases, already underway: for example, the new Flight III Arleigh Burke-class destroyers. That potentially begins a cycle that leads to fewer, costlier, and increasingly exquisite platforms. While the United States contemplates purpose-built, large combatants, others have proposed less prestigious and costly options to achieve similar aims. For example, navalists have suggested reconfiguring merchant ships into cheap missile carriers, a role that China has explored, which could achieve greater dispersion, logistical flexibility, and cost efficiency than a small number of specialized warships. Another option would be to supplement the fleet’s magazine with unmanned or optionally manned vessels. If long-range weapons can be employed from outside adversary engagement envelopes, the necessity of a highly survivable, bespoke platform warrants rigorous scrutiny.
Even unmanned systems, such as collaborative combat aircraft drones intended to ease burdens through attritable mass and cost in aerial combat, are showing early signs of cost and requirements creep. This trend is driven in part by the services’ efforts to enhance their survivability to protect increasingly valuable capability investments. These cost creeps have drawn congressional attention, raising concerns about how the services will manage “affordability” to procure “sufficient numbers to execute the concept of operations.”
Exquisite platforms increase the likelihood that capabilities are withheld and preserved rather than employed when losses mount, or that disproportionate numbers of a fleet are removed from offensive operations to defend irreplaceable assets. Or worse, as with the Spanish tercios, warfare can advance in ways that nullify core assumptions and render costly investments irrelevant.
The takeaway is to avoid devoting excessive resources to over-engineering individual platforms at the expense of investing in the resilience of the overall system, and to avoid unrealistic expectations that platform survivability alone mitigates operational or strategic risk. Future force design across the services should balance quality and quantity to ensure sufficient mass. When the required mass is unachievable, force design should avoid the survivability paradox and explicitly accept the gap, then deploy resources elsewhere. The U.S. military should prioritize holistic system resilience and the capacity to regenerate forces over marginal gains in individual-platform survivability, even when numerically disadvantaged.
Trevor Phillips-Levine is a U.S. Navy officer and author.
Andrew Tenbusch is a U.S. Navy officer and author.
Walker D. Mills is a U.S. Marine Corps officer and the co-director of Project Maritime at the Irregular Warfare Initiative.
The views in this article are those of the authors and do not represent the opinions or positions of the U.S. Navy, the Marine Corps, the Department of Defense, or any part of the U.S. government.
**Please note, as a matter of house style, War on the Rocks will not use a different name for the U.S. Department of Defense until and unless the name is changed by statute by the U.S. Congress.
Image: Petty Officer 1st Class Sara Eshleman via Wikimedia Commons
