Munitions stockpiles

Why can a country hold years of ammunition in peacetime and empty the shelves in a month of war?

A defence ministry says it holds five years of artillery ammunition. Fighting starts, and within weeks the same ministry is scouring the world for shells. Nobody lied and nothing was stolen, so the sentence must have meant something other than what a listener naturally hears — and finding out what explains most of the surprise.

Start with the unit. "Years of ammunition" is not a quantity; it is a quantity divided by a rate. And the rate that was used to make the number is the peacetime one — what the guns fire on ranges and exercises, which is set by training budgets and by how much barrel wear a service will accept.

Write the arithmetic out and it does the work by itself. Let training expend E rounds a year, and let the shelves hold 5*E*. That is the "five years". Now suppose sustained fighting expends E in a single week — a multiple that varies enormously with the weapon, the tempo and the theatre, so I will not pretend to a figure. The same shelves now hold five weeks. Nothing about the stock changed. The denominator did.

That answers the question as asked, and thinly, because the obvious rejoinder is: so start making more. Here the second constraint appears, and it is the one that hurts.

Ask what the production line was sized for. Not for war — for replacement. A line kept alive in peacetime is built to replace roughly what training consumes, which is E a year, which on the numbers above is about one week of combat expenditure. Run that line flat out for a full year and you have replaced a week of fighting. This is the piece that most reporting on ammunition shortages skips: the stockpile and the factory are sized by the same peacetime rate, so they fail together rather than one covering the other.

So why not size the factory for war? Here we leave engineering for institutions, where the real mechanism lives.

A munitions plant sized for a war that does not come is idle capital. No firm builds it against a demand spike that may never arrive, because the spike is precisely the demand pattern that cannot be recovered in a price. What makes the investment rational is not a bigger budget in one year but a multi-year committed order — a contract shape, not a sum of money. Annual appropriations, however generous, buy shells; they do not buy presses, because the press outlives the appropriation.

And beneath the assembly line sits a chain of sub-tiers, each with its own lead time: the energetics — propellant and explosive fill — then the chemical precursors those need, then the specialist forging and machine tools that make the bodies and the plants. The lead times compound in series, and the output is capped by whichever is longest, not by an average. That is why the answer to "how fast can you expand" is measured in years and is stubbornly insensitive to how urgently it is asked.

There is a quieter reason the stock was thin to begin with. Among all the things a defence budget buys, ammunition is the most compressible. Cut people and units disappear; cut aircraft and the gap is visible on a runway. Defer an ammunition buy and nothing observable happens for years, because the stock has no readiness signal of its own — it just sits, and it costs money to store, inspect and eventually dispose of as it ages. Every year of deferral is invisible, and the invisibility accumulates.

Here is the test. If the binding constraint is the deepest sub-tier's lead time plus the contract-shape problem, then releasing money should not raise output for roughly that lead time, and the increase, when it comes, should follow the signing of multi-year commitments rather than the rise in annual budgets. The refuting observation would be output that tracked appropriations with a short lag — which would mean the constraint was really funding, and the industrial story an excuse.

Two refinements connect the steps.

First, "days of supply" is not a property of a stockpile. It is a property of a stockpile and an assumed rate, and the assumed rate is a planning choice made years earlier by people who had to guess what a war would look like. When the guess is wrong — and the modern surprise has been a return to prolonged high-volume indirect fire that many planners had written off in favour of short, precise campaigns — the stock number was never wrong, only its denominator.

Second, the expenditure rate is not a fact of nature either. It is partly chosen, and armies under supply pressure choose it: rationing rounds per gun per day, shifting to precision munitions that need fewer rounds for the same effect but have their own far tighter production ceilings, or accepting a slower tempo. The shortage does not simply arrive; it is met by a change in how the war is fought, which is one reason the shelves rarely empty as completely as the arithmetic threatens.

The honest caveats are large. Actual holdings and consumption rates are classified nearly everywhere, and most public figures are estimates published by parties with a direct interest — ministries arguing for budgets, manufacturers arguing for orders, and analysts on retainer to both — so I have deliberately given no real quantities and used only a worked ratio to show the shape. The ratio itself is illustrative. Whether the true constraint on expansion is energetics, machine tools, skilled labour or contract certainty is genuinely disputed, and different countries have found different answers.

• Why precision munitions invert the trade: far fewer rounds needed, but far lower and less expandable production ceilings.

Every term the collection defines is gathered in the glossary.