Industrial Base Coherence

America cannot build what it needs merely by deciding to need it.

The Army aimed to produce 100,000 rounds of 155mm artillery ammunition a month by October 2025. The Inspector General reported March 2026 output of 36,000—more than double the earlier rate of 14,000, but barely a third of the target.

A nation's ability to produce depends on how its capabilities fit together. Contracts assign delivery obligations to individual companies. Fulfilling them requires investment, inputs, and approvals from organizations with separate priorities.

Buyers must commit early enough for suppliers to invest, qualify production, and deliver. Industrial base coherence is the fit between this distributed production system and the demands placed on it.

A factory's capacity is partly made outside the factory

BCG's 2026 fieldwork at more than 50 defense suppliers attributed roughly 70 percent of underperformance in that sample to conditions outside suppliers' direct control.

Buyers ordered at current production rates while expecting suppliers to prepare for much higher rates. The order book supported existing capacity; expansion depended on purchases buyers had yet to commit to.

Ordering this way helps preserve the capacity limit the buyer wants removed. Order timing and terms shape suppliers' hiring, equipment purchases, and inventory. Buyers help create the capacity they later judge suppliers for lacking.

Rational companies can form an incapable industrial base

A supplier that expands against a forecast risks idle equipment and payroll if purchases fail to materialize. Its customer also has reasons to wait. Funding, designs, or demand may change.

Each protects itself. Together, they leave the capacity unbuilt.

Another forecast does not settle who pays if the expansion becomes unnecessary. When cancellation risk blocks investment, a buyer's agreement to share the losses gives the supplier a stronger reason to proceed. The buyer accepts some risk of unused capacity to secure capacity when needed.

Inside an enterprise, executives have authority to reconcile competing budgets and priorities. Across an industrial base, those decisions belong to separate companies and institutions. Government has substantial purchasing and policy powers, but a national target supplies neither a common balance sheet nor a common chain of command.

Surge capacity also costs money while idle. Purchasing that rewards low prices under steady demand discourages manufacturers from maintaining equipment and skills beyond ordinary needs. Rapid expansion requires paying for that readiness.

Coherence depends on the production rate, deadline, resilience, and cost required. An industrial base suited to economical steady production may be poorly suited to urgent expansion.

A commitment has a deadline too

When investment waits for an order, delaying the order delays the earliest possible delivery. Equipment must arrive, workers must become proficient, and processes must qualify.

In March 2026, 155mm ammunition production reached 46,000 metal-parts units a month, while loading, assembly, and packing reached 36,000. Sustaining finished output above 46,000 would require more metal parts. Both stages needed substantial expansion to reach 100,000.

An investment's value depends on how much additional finished output it enables, how soon, and at what cost. Expanding one stage produces little gain if the next cannot absorb its output.

Yet waiting until metal parts become the bottleneck postpones investment until the shortage is already delaying deliveries. Equipment orders and qualification for the higher rate must begin while the current bottleneck is being resolved.

A production plan therefore needs deadlines for equipment funding, design releases, qualification, and supplier commitments. Different organizations control these decisions. By the time a shortage reaches a delivery report, the deadline for preventing it may be months past.

An accepted first article proves that a supplier can produce a conforming item. It does not establish sustained production at the required rate. Counting approval as available capacity hides the remaining preparation.

Competing suppliers can share a bottleneck

CNAS's 2026 analysis of solid rocket motors describes new manufacturers dependent on a narrow base of qualified upstream inputs.

Two vendors buying from the same material producer remain exposed to its failure. Several programs expanding through that producer compete for its output even if it never fails. Each customer's plan may fit individually while their combined requirements exceed the supplier's capacity.

Supplier counts reveal neither independence nor room to expand. Both require tracing shared inputs and the competing claims on them.

Buying the ability to supply

In 2022, the sole supplier of Humvee windshield kits and replacement glass closed while the Army was introducing a tougher durability specification.

The Defense Logistics Agency bought against the earlier specification and worked with the Army to authorize an intermediate requirement. It then paid four suppliers to run qualification tests concurrently. The interim requirement preserved ballistic performance but accepted less durability than the final specification sought.

DLA reported four qualified sources and resumed shipments. It planned to build inventory and confirm multiple sources before adopting the final specification.

The recovery required the Army's specification decision and DLA's qualification funding and purchases to work together. Money for testing alone would have left the changing requirement unresolved.

Follow the cause beyond the company

A delivery report names the company that owes the product. It does not necessarily identify who controls the decisions needed to produce it.

Expediting a supplier cannot recover months lost before its order was released. More equipment cannot settle an acceptance requirement. A second manufacturer cannot remove a bottleneck shared by both.

Some shortages do require more equipment or skilled people. Tracing dependencies identifies where the change must occur, inside the late supplier or elsewhere.

The unit of analysis should be large enough to contain the cause. The intervention should be no larger than necessary to change it.

A collection of plausible company plans is not yet a feasible national production plan. A final assembler's commitment depends on upstream investment, qualified production, and access to capacity other programs may also be counting on. National delivery promises rest on whether those commitments hold together.

Sources

Industrial base coherence

The Talbot West Doctrine A Theory of Enterprise Coherence Systemicity Aerospace and defense