The Denominator Ate the Destroyer: How 32 Zumwalts Became Three
A document-first investigation of the Zumwalt programme through fixed-cost allocation, PAUC, quantity collapse, sunk cost, learning curves, and mission value.
By MyAudioBooks.ai ·
Listen free: The Denominator Ate the Destroyer: How 32 Zumwalts Became Three
On February 1, 2010, the Navy notified Congress that the D D G-1000 destroyer program had breached a critical statutory cost threshold. The Navy's explanation was extraordinary: the breach was a "mathematical consequence" of cutting the class to three ships. Roughly nine point three billion dollars of research and development now had to be divided across three hulls, loading about three point one billion dollars of development cost onto the acquisition cost of each ship.
Our verdict is that the denominator ate the destroyer. The headline unit-cost explosion was partly real technical failure and partly accounting arithmetic created by buying ninety percent fewer ships. Treating the combined number as though each hull simply became six times harder to build hides the decision that created it. Treating the quantity cut as harmless math hides the capabilities, learning, and industrial scale that disappeared with the other twenty-nine ships.
At My Audio Books dot A I, you can create your own audiobooks from prompts, turn your documents into audio, and store everything in your personal library.
Section One: Four Numbers That Seem to Contradict One Another
G A O's 2012 program assessment put the contradiction in one table. Between the January 1998 baseline and the August 2011 estimate, research-and-development cost rose from roughly two point three billion dollars to ten point four billion dollars, an increase of three hundred fifty-five point six percent.
Procurement cost moved the opposite way, falling from roughly thirty-two point five billion dollars to ten point six billion dollars because the Navy would buy far fewer ships. Total program cost fell from thirty-four point eight billion dollars to roughly twenty-one billion dollars.
Yet program acquisition unit cost rose from about one point zero nine billion dollars to seven point zero zero billion dollars, a five-hundred-forty-three-point-two-percent increase. Planned quantity fell from thirty-two ships to three, a reduction of ninety point six percent.
All four statements are true. Total cost fell. Procurement spending fell. Development cost rose. Average acquisition cost per ship exploded. The metric changed because both its numerator and denominator changed.
Section Two: What PAUC Actually Measures
Program acquisition unit cost, or P A U C, adds research, development, procurement, acquisition-related operations, and certain other program costs, then divides by the number of units. Average procurement unit cost, or A P U C, excludes research and development and divides procurement cost by quantity.
That distinction matters because research and development is largely indivisible—it does not fall in proportion when the final order shrinks. The software architecture, hull design, radar integration, power system, testing, and program office still have to exist for the first ship.
Using the Navy's rough nine-point-three-billion-dollar development figure, thirty-two ships would have absorbed about two hundred ninety-one million dollars of development cost each. Three ships absorb three point one billion dollars each. No new cheque for two point eight billion dollars is written merely because the denominator changes; the same development pool is reallocated across fewer units.
C R S reports that the 2010 restructuring letter attributed seventy-nine of the eighty-six percentage points in the P A U C increase to truncation, with the remaining seven points attributed mainly to added development work. That is the rare government document that separates arithmetic from execution.
Section Three: The Quantity Cut Had a History
The original plan contemplated up to thirty-two D D G-1000 destroyers. The planned class fell first to a range of sixteen to twenty-four, then to seven, and finally to three.
In 2008 the Navy reversed its destroyer strategy. It said future operations required greater emphasis on open-ocean antisubmarine warfare, anti-ship cruise-missile defence, and ballistic-missile defence.
The Navy stated that it had reevaluated the future operating environment before announcing that reversal. The Navy concluded that the D D G-51 design could perform those missions adequately and would cost less to procure. Congress approved ending D D G-1000 procurement after a third ship and restarting D D G-51 procurement.
The program that became D D G-1000 was announced on November 1, 2001, as part of a planned family of new surface combatants, and the Navy redesignated the effort D D G-1000 in April 2006. The first two ships were procured in fiscal 2007 and the third in fiscal 2009. The Navy plans no further procurement of D D G-1000s.
This was not a spreadsheet accident. Requirements, threat assumptions, budget pressure, and cost estimates changed together. The denominator was endogenous—generated by decisions inside the acquisition system rather than imposed by nature.
That makes the breach hard to interpret. The quantity cut raised unit cost, and high unit cost made the class look less affordable. The metric did not merely report the decision; it fed the political story surrounding the program.
Section Four: The Strongest Case Against the Denominator Story
The strongest case against our reading is that quantity arithmetic can become an excuse. A troubled program can blame its unit cost on Congress buying fewer units even when technology, schedule, requirements, and production performance also deteriorated.
That objection wins here. G A O found that only three of twelve critical technologies were mature in early 2012, with eight others not expected to be demonstrated in a realistic environment until after installation.
G A O reported at the time that the first ship was approximately 63 percent complete and the second 22 percent complete. The long-range land-attack projectile had rocket-motor delays. Shipbuilders had rework and schedule delays, and the program removed the volume-search radar after the critical breach.
By 2018, G A O said the design had not been stable when lead-ship fabrication began, an approach inconsistent with best practices. Ongoing development and shipboard testing produced design changes, schedule delays, and cost increases. The Navy had identified more than three hundred twenty serious deficiencies when the lead ship's hull, mechanical, and electrical systems were delivered.
Quantity did not cause those problems. It amplified their average cost and reduced the number of follow-on ships over which the program could learn.
At My Audio Books dot A I, you can listen to this investigation and thousands of others that explain the hidden systems behind the headlines.
Section Five: The Gun Without a Round
The original land-attack mission depended on two Advanced Gun Systems and a precision long-range projectile. In November 2016, the Navy canceled planned acquisition of that projectile because of its high cost per round.
C R S records press reporting that the projected unit cost of each projectile had risen to at least eight hundred thousand dollars before the cancellation. G A O reported that the Navy developed seven courses of action after the cancellation before changing the ship's mission.
G A O reported that the Navy evaluated five alternatives, none met the requirements, and it chose not to pursue a near-term replacement. The result, in G A O's words, effectively rendered the gun systems useless for combat operations in the foreseeable future.
The Navy then shifted the class's primary focus from land attack to offensive surface strike and examined replacing the guns with additional launch cells. The current program has moved toward large-diameter tubes for Conventional Prompt Strike hypersonic missiles, repurposing three unusual hulls for a different mission.
The first D D G-1000 was commissioned into service on September 7, 2016, and its delivery date was revised multiple times, reportedly to April 2020.
This is the operational price hidden by an average-cost argument. A program may cut quantity to control total spending, yet lose the fleet scale and original mission that justified the fixed investment. The ships still exist. The original business case does not.
Section Six: The Cost Death Spiral
The familiar death spiral starts with a fixed or sunk cost spread across many planned units. Quantity is cut. Average cost rises. The higher unit cost weakens political and customer support. Quantity is cut again, and average cost rises once more.
But the spiral is not purely an accounting illusion.
G A O's review of Nunn-McCurdy breach history reached a similar conclusion: breaches are often the result of multiple, interrelated factors. Lower quantity can damage supplier economics, reduce learning-curve gains, eliminate competition, increase minimum-order costs, and leave fewer units over which engineering fixes can be amortized. Those effects increase the numerator as the denominator falls.
Zumwalt shows both mechanisms. G A O's 2012 table shows the direct denominator effect. C R S also reports real procurement-cost growth after the three-ship baseline: combined procurement estimates rose from about eight point nine eight billion dollars in the fiscal 2009 budget to about thirteen point eight six billion dollars in the fiscal 2026 budget, a fifty-four-point-four-percent increase.
For scale, the Navy's fiscal 2024 budget submission estimated the first two ships' combined procurement cost at about 9.45 billion dollars and the third ship at about 4.34 billion dollars.
C R S says some early growth came from reallocating class-wide procurement costs from the canceled fourth through seventh ships onto the remaining three. It also describes movement toward a higher independent cost estimate and later spending to modify ships for hypersonic weapons.
The correct conclusion is therefore path-dependent—shaped by the sequence of prior choices. The quantity cut raised measured unit cost immediately. Technical and production problems raised real cost. Mission changes then required new investment. Each decision changed the economics facing the next one.
Section Seven: Do Not Confuse Five Different Costs
Executives and public officials should force every troubled program to report five measures separately.
First is total sunk cost: money already spent that cannot be recovered by canceling future units. Second is remaining incremental cost: the cash still required to finish the next unit or capability. Third is average procurement unit cost: production dollars divided by units. Fourth is program acquisition unit cost: development plus procurement and other acquisition costs divided by units. Fifth is capability cost: dollars per usable mission outcome.
The counterfactual—what would have happened under another quantity—is also essential. Dividing sunk development by the new quantity describes the accounting average, but not the cash saved or spent by canceling one more ship. Marginal decisions require marginal cost.
A programme can therefore have a terrifying P A U C and still be rational to finish if the marginal completion cost is below the value of the capability. It can also have a tolerable procurement unit cost and still be irrational if the intended capability no longer works.
Section Eight: The Boardroom Version
The same denominator trap appears outside defence. A pharmaceutical platform spreads trial infrastructure across an expected drug pipeline. A software company spreads a data-centre build across projected users. An automaker spreads tooling across a model run. A hospital spreads fixed overhead across procedures.
When volume falls, finance teams often raise the fully allocated unit cost. That number is useful for understanding recovery of total investment, but dangerous for shutdown decisions. Unavoidable costs do not disappear when a product line is cut; they migrate to the surviving lines.
G A O's own portfolio accounting makes the same separation. In the 2011 portfolio, roughly 29.6 billion dollars of procurement cost growth was attributable to quantity changes on 36 programs. Another 31.1 billion dollars could not be attributed to quantity and pointed instead to production problems, inefficiencies, or flawed initial estimates. G A O also found that over 60 percent of programs lost buying power as measured by program acquisition unit cost. Only 11 of the 61 programs with unit-cost increases had also cut quantities.
Boards should demand a bridge that labels each cost as sunk, avoidable, incremental, or allocated. They should separate quantity-driven unit-cost change from underlying cost growth. G A O's portfolio work makes the same distinction: quantity changes and production inefficiency are different sources of cost movement.
They should also ask what learning and supplier scale disappear with the cut. A one-unit reduction can save its direct material and labour while increasing the cost of every unit that remains. We'd argue this is the question Zumwalt's own reporting never answered cleanly: what was actually saved, per hull, by canceling ship four?
Section Nine: The Denominator Is a Decision
The Zumwalt class did not become cheap when total programme cost fell, and it did not become seven times harder to fabricate solely because P A U C rose. The quantity cut, development growth, technology immaturity, production challenges, and mission change all moved different parts of the cost system.
The best evidence is the government's own split: seventy-nine of eighty-six percentage points in the 2010 P A U C increase came from truncation, not the seven points attributed to added development work. That does not absolve the program. It tells us which failure the metric was measuring.
At My Audio Books dot A I, you can create fiction, non-fiction, and audio from your own documents, all in one library. Learn more today at My Audio Books dot A I.
The denominator ate the destroyer because the Navy built a development programme for a class and bought a demonstration fleet. Twenty-nine absent ships did not merely vanish from a table; they removed scale, learning, mission capacity, and political support. The price of confusing average cost with marginal cost is that leaders can cancel volume to save money, watch unit cost soar, and then cite the soaring number as proof that the cancellation was inevitable. The denominator is a decision.