Borrowed Urgency: The West trying to run a wartime program on peacetime authority
I. Introduction
The West is re-arming, and at a pace with no obvious precedent since the Cold War it is often said to be repeating. NATO's 2025 Hague summit committed member states to spending five per cent of GDP on defence by 2035 — three and a half per cent on core requirements, a further one and a half on broader security and industrial resilience. That commitment has not been left to erode the way the alliance's earlier two per cent target did for a decade after it was set: in 2025 alone, for the first time in NATO's history, every member state met or exceeded that older floor, and European allies and Canada raised defence spending by twenty per cent in real terms in a single year. Germany amended its own constitution — the Basic Law's debt brake, a fiscal restraint the post-Cold War political class had treated as close to sacred — specifically to fund the increase. Global military spending, on IMF figures, is projected above $2.6 trillion for 2026: the fastest sustained rearmament since the Cold War itself.
That is a demand-side mobilisation, committed through the most binding instruments a modern democracy has available to it: treaty obligation and constitutional amendment. The hardware it funds — missiles, radar, drones, submarines, the vehicles and systems that consume neodymium-iron-boron magnets by the tonne — is exactly the demand this essay's supply-side argument concerns. The instrument actually securing the material inputs for that hardware is not a matching wartime footing but a borrowed one: the Defense Production Act of 1950, modelled on the War Powers Acts of 1941 and 1942, invoked for the current rare earth effort not because a war has been declared but because the statute written for one has never required it to be. The authority is real. It has simply never shed the condition of its own renewal — reauthorised by Congress more than fifty times, left dormant for two decades before its 1980s revival, and, in its present use, contingent on continuing appropriations in a way no wartime mobilisation examined in this essay ever had to survive. The gap between how the West is arming itself and how it is supplying itself is what this essay means by borrowed urgency, and it is worth tracing where that gap comes from before turning to the historical case that gives it its sharpest shape.
For decades before the American Civil War, Great Britain supplied high-quality, inexpensive gunpowder, which made domestic production unprofitable. British powder flowed freely into American ports, and there was no commercial reason to build a domestic industry capable of replacing it — until relations turned, at which point Britain stopped selling. That pattern repeated twice, in the Revolution and again in 1812, each time leaving American forces to scramble for a domestic supply that decades of cheap British imports had given nobody any reason to build.
A friend recently sent me a paper that traces what happened when this pattern played out a third time, on a larger scale, with much higher stakes: Derek Taylor's Creating Southern Thunder: The Evolution of Confederate Gunpowder Production during the American Civil War, published in The Welebaethan in 2019. Taylor's subject is narrow and specific — how a nation with almost no domestic powder-making industry, cut off from its principal supplier at the outbreak of war, built one of the most advanced gunpowder works of the nineteenth century within roughly a year. It is not a paper about critical minerals and does not claim to be.
China's dominance of rare earth processing did to Western investment what British powder mills did to American: not through concealment or sabotage, but simply by being good enough and cheap enough, for long enough, that building a costlier domestic alternative made no business sense until the supply was no longer guaranteed.
This piece is not an argument that the two situations are the same, or that history offers a template to be applied mechanically. It attempts to identify where the cross-references hold, where they break down, and what—if anything—can be usefully learned from a case where the underlying problem was resolved and a case where, yet, it has not.
II. Binary Criticality
That dependency was never a matter of degree. Both gunpowder and the modern magnet are composite materials, inert without their scarce ingredient regardless of how much of the finished product that ingredient physically comprises. Saltpetre made up roughly three-quarters of the Confederate charge by weight; neodymium makes up barely a quarter of a modern magnet's. That difference is beside the point. Gunpowder without saltpetre is not weakened gunpowder — it does not exist. A magnet without neodymium is not a weaker magnet — it is a bar of iron. In both cases, the whole supply chain stands or falls on the one input that cannot be substituted, sourced elsewhere, or done without, irrespective of what share of the bill of materials it occupies.
III. The Trent Affair and the Senkaku Precedent
The parallel sharpens further once you look at how Britain's leverage over that supply was used. In November 1861, a U.S. Navy warship stopped the British mail packet Trent and removed two Confederate envoys, James Mason and John Slidell, bound for Britain to seek support for the Southern war effort. Britain's response was immediate and specific: an embargo on gunpowder materials destined for the United States, backed by the threat of war. Washington released both men within weeks. Britain never recognised the Confederacy as a result, and the underlying dispute over the South's status went nowhere — the embargoes only achieved purpose was securing the release. It was pressure applied through a critical-material chokepoint, deployed to force a specific, narrow outcome, then withdrawn once that outcome was achieved.
A century and a half later, a similar sequence played out between China and Japan. In 2010, Japan detained a Chinese fishing boat captain near the Senkaku Islands. China responded by restricting rare earth exports to Japan—the same instrument, aimed at the same kind of target: not a demand for territorial concession, but leverage applied through a chokepoint to force the release of a detained national. Japan released the captain. Nothing about the underlying dispute moved. The lesson embedded in both episodes is the same: a critical-material dependency is a lever that can be pulled for reasons unrelated to the material itself, and the country holding the dependency does not need to win the argument to win the immediate confrontation.
IV. Davis's Veto — Diagnosing the Bottleneck
In January 1862, the Confederate Congress drafted a bill to subsidise private manufacturers willing to expand gunpowder and saltpetre production: a fifty per cent government advance against a twenty-five per cent investment by the manufacturer, contingent on a good faith promise to deliver. Jefferson Davis, the Confederacy's president, vetoed the first draft. His objection was not fiscal caution — it was a diagnosis. Saltpetre was already scarce, and subsidising more powder mills without first securing more of the raw material feeding them would create more manufacturing capacity competing over the same shortfall. The Confederacy, Davis judged, did not need more mills. It needed more saltpetre. Congress responded not by overriding him but by agreeing with the diagnosis: the Nitre and Mining Bureau was established on 11 April 1862, redirecting government effort toward acquisition specifically, ahead of any further investment in manufacturing.
What makes the veto worth dwelling on is not that Davis was right, though the subsequent record suggests he was — it is that he was faced with two plausible-sounding responses to the same shortage and chose correctly between them. More mills and more raw material both look, on the surface, like reasonable answers to “we don't have enough gunpowder.” Only one of them addressed where the constraint sat.
That is the question worth asking of the modern case, and it is answerable now rather than only in hindsight. Recognition of the rare earth dependency is not in doubt — it has been stated at every level from departmental briefings to a Presidential one. Nor is capital in doubt; multiple financing instruments, across multiple governments, have been committed to multiple projects. What is less settled is whether that capital is landing on the actual constraint. A tonne of mined ore and a tonne of separated, magnet-grade oxide are not the same kind of scarcity, any more than raw nitrous earth and refined saltpetre were the same kind of scarcity in 1862 — and financing the wrong one, however generously, produces the same result Davis was trying to avoid: more capacity built against a bottleneck that was never where the money was aimed.
The veto was not a rejection of private-sector engagement — it redirected where that engagement should be aimed. The mechanism Davis insisted on, once pointed at raw material acquisition rather than manufacturing capacity, worked: Isaac St. John, superintendent of the newly formed Nitre and Mining Bureau, organised the South into nitre districts, brought private producers under coordinated control, and by September 1864 had generated 1,735,531 pounds of domestically-produced saltpetre — insufficient against total demand, but real, measurable private engagement achieved through risk-sharing rather than replaced by government alone.
The U.S. International Development Finance Corporation states the identical theory of action as its own founding purpose: to mobilise private capital by derisking investment in markets the private sector otherwise avoids, explicitly including Africa, where much of the modern critical-minerals resource base sits. Years into that mandate, the result Davis achieved within two years of a standing start has not been matched. Africa's share of global exploration investment has fallen, not risen, over the past two decades. European commercial banks continue withdrawing from African mining rather than being drawn back in. And in the clearest available case, government and sovereign capital — a Qatari fund, a U.S. development finance agency, an export credit bank — are doing the work the mobilisation theory says private capital should be doing instead, not standing behind it.
Here the parallel, which has held closely up to this point, breaks — and the break is worth stating as plainly as the similarities, because it is doing real explanatory work of its own.
Davis's veto was not only a correct diagnosis. It was a risk judgment made at the level of the whole war effort, by someone whose mandate was the outcome itself rather than any single link in the chain leading to it. Josiah Gorgas, chief of the Confederate Ordnance Bureau, exercised the same kind of judgment in creating the Nitre Bureau and handing it to a man he personally trusted; George Rains, superintendent of the Augusta Powder Works, exercised it again at the level of engineering practice, overriding established method the moment his own on-site judgment said the risk lay elsewhere — steam instead of water, his own crystallising wheel instead of straight raking, reverting his finishing-process innovation the moment it underperformed. At every level of Taylor's account, someone with standing looked at where they personally believed the risk actually sat and acted on that belief directly. Nothing in the process validated their judgment before they were permitted to act on it. Their authority was the validation.
No equivalent role exists anywhere in the modern chain this essay has traced. Risk is not ignored — it is assessed exhaustively. But every assessment happens at the granular level, by parties each scoped to a narrow fragment of the whole. A credit committee assesses counterparty risk to a single loan; a compliance function assesses the risk of a single registry filing; a board assesses its own fiduciary exposure; counsel assesses litigation risk clause by clause. Each of these assessments, taken on its own terms, is careful and defensible. None of them is the question Davis was answering. Nobody in the structure is mandated to ask, let alone empowered to act on, whether the cumulative pace of all this careful, granular validation is itself the risk that matters most — whether the sum of locally sensible caution adds up to missing a fixed, external deadline entirely. That question currently has no owner. A system built entirely from node-level risk assessors, however diligent each node is individually, will reliably produce an outcome that is locally prudent and globally slow, because minimising risk at every point in a chain is not the same operation as minimising risk to the chain's actual purpose. Rains's Confederacy never had to reconcile those two things, because Gorgas's mandate collapsed them into one question, answered by one person, at every level examined here. The modern effort has re-fragmented that question into as many pieces as there are institutions in the chain—and, in doing so, has made the deadline itself an orphan: tracked by everyone, owned by no one.
V. Three Tiers: Gorgas, St. John, Rains
If Davis's problem was diagnosing which link in the chain was short, Gorgas, St. John, and Rains together answer a different question: once diagnosed correctly, what does it take to close the gap? Gorgas held the authority to redirect the whole war effort. St. John built the acquisition apparatus beneath him. Rains built the execution capacity beneath that. All three were necessary; none was sufficient alone. The modern case that maps onto this structure most cleanly is not the kind of fragmented, multi-counterparty offtake arrangement common elsewhere in current rare earth financing, but the U.S. Department of Defense's intervention in MP Materials — a single, vertically integrated program that, unlike much of what passes for strategy in this sector, was built by one buyer dealing with one company across the whole chain at once.
The mapping holds at each level. Gorgas's role — the authority to intervene directly, overriding the market's own allocation of capital — belongs to DoD itself, which in July 2025 took an equity stake in MP Materials and restructured the relationship from ordinary customer to guarantor of last resort. St. John's role — acquisition of the specific raw material in short supply, not simply more general capacity — belongs to the $150 million loan earmarked specifically for expanding heavy rare earth separation at Mountain Pass, targeting dysprosium, terbium, and samarium rather than more NdPr mining the country already had. And Rains's role — patient capital and technical execution insulated from ordinary market discipline — belongs to MP Materials itself: Mountain Pass, the existing Texas magnetics operation, and the new 10X facility, financed by a $1.0 billion commitment from JPMorgan Chase and Goldman Sachs and underwritten by a ten-year $110/kg NdPr price floor, a separate $140 million annual minimum-EBITDA guarantee, and a DoD commitment to purchase the whole of 10X's magnet output for a decade. Nothing in this file resembles Rains's mandate as closely as that structure does.
And unlike most of the capital catalogued elsewhere in this essay, this mechanism is demonstrably paying out rather than merely standing ready. MP Materials disclosed $42.3 million in Price Protection Agreement income in the first quarter of 2026 and a further $17.6 million in the second — real, quarter-specific transfers from the floor to the company, narrowing as market price recovered rather than newly triggered by a collapse. Whatever else can be said about the pace of this effort, the price floor is not a paper commitment. It is functioning exactly as designed.
There is a case, on this evidence, that the deal's creation was itself a Davis-shaped act in the narrower sense this essay has used that term: someone at DoD made a personal judgment that NdPr price risk was real enough to warrant absorbing it directly, by guarantee, rather than waiting for the market to validate the concern first. That is a genuine data point against the pattern described in Section IV — proof that the capacity for this kind of authority has not vanished entirely from the modern apparatus, even if it remains rare.
But the case for crediting MP Materials as a full modern Rains has two real limits, and both matter more than the one point in its favour.
The first is the same execution-tier gap found everywhere else in this file. The $150 million loan funds heavy rare earth separation capacity that does not yet exist — it is new construction, not an existing line being scaled. Read plainly, DoD's own structure concedes that the United States had no meaningful domestic capability in exactly the fraction of the chain Davis would have identified as the actual constraint. The light rare earth side of the picture — NdPr mining and separation — is real and operating. The heavy rare earth side, the harder and more strategically sensitive half, is still being built from near zero, on the same timeline as everything else this essay has found wanting.
The second is structural, and it is the sharper of the two. Taylor is explicit that St. John divided the South into fourteen separate nitre districts specifically because geographic diversification reduced the risk of production being interrupted by a single attack. The MP Materials program is the deliberate opposite: one mine, one company, one integrated facility complex carrying the entire weight of the strategy DoD has built around it. That concentration is not, in the usual sense, a choice at all. Mountain Pass is not simply the strongest of several producing options DoD might have backed — it is, at the time of writing, the only operating rare earth mine in North America. Other prospects exist. Bear Lodge in Wyoming, NdPr-enriched and further along than most, has spent years working through permitting and financing while still proving a novel extraction process at demonstration rather than commercial scale. Round Top in Texas holds one of the country's largest heavy rare earth resources by total tonnage — upward of a billion tonnes of rhyolite — but at a grade of only a few hundred parts per million, so the same scale that makes the deposit notable also means mining and processing a correspondingly enormous volume of rock for every kilogram of rare earth recovered; the operator is targeting commercial production only from 2028. Bokan Mountain in Alaska holds a genuine heavy rare earth deposit in a location remote enough to complicate any near-term development. Elk Creek in Nebraska remains an advanced exploration project, not a mine. Each carries its own specific barrier to economic exploitation — capital, chemistry, permitting, grade, or geography — and none is close enough to production to function as an alternative today. St. John distributed his effort across fourteen districts that were each already yielding saltpetre. The modern effort has, for now, nothing comparable to distribute across. The single point of failure this essay has identified is not a preference for concentration over dispersion — it is the absence, at present, of anywhere else to concentrate, at the centre of the one part of this file's evidence that most resembles Rains's mandate.
Va. Process Risk: The Old Method and the New
The pattern here is close to Rains's, with one real difference. Rains adapted an already proven process from Waltham Abbey, the Royal Gunpowder Mills near London, via Major Fraser Baddeley's pamphlet—an industrial method imported wholesale. His “new method” replaced a laborious six-day purification process with continuous raking machinery, cutting refinement to a single day while producing a purer crystal; Rains then improved on it further with his own crystallising wheel, pushing purity to what he claimed was beyond that of the most celebrated powder factories of the age.
ReElement Technologies' chromatographic platform has a similar structure at smaller scale: the company was formed as a subsidiary of American Resources Corporation specifically to commercialise column chromatography developed in Professor Linda Wang's Purdue laboratory, taking an exclusive license to the technology — still owned by Purdue, not ReElement — in January 2021. But where Baddeley's method was already proven at full industrial output before Rains adopted it, Wang's was proven only at laboratory scale. ReElement's own purity results, first achieved in 2022, and its stated commercial capacity represent the company's own scale-up work, not a transplant of something already running elsewhere at the volumes now being claimed.
Here the comparison tilts, briefly, in the modern case's favour. Rains was importing a method already validated at industrial scale; ReElement is proving its method at that scale for the first time, under its own risk, rather than importing something already working elsewhere at the volumes now being claimed. That is a harder position to succeed from. It is too early to credit this as a settled achievement — ReElement's figures remain company-reported, without the kind of independent corroboration Taylor can cite for Rains's saltpetre purity. Rains's own history is not uniformly a success story either: his combined dusting-glazing-drying innovation was abandoned after too much powder was rejected on re-inspection, and he reverted to the established method. But it is worth stating plainly, in a piece otherwise concerned with where the West is behind: this may be one place it is not.
VI. Scale and Patience: The Nitre Beds
Not every part of the Confederate response was manufacturing or acquisition in the ordinary sense. Alongside mining existing saltpetre caves, St. John's Bureau pursued a slower, more deliberate method: growing it. Nitre beds — compost mounds of rotted animal manure, urine, and vegetable matter, built to a standard fifteen-foot-long, eight-foot-wide, five-foot-high specification and watered weekly to keep the decomposition active — were a genuine attempt to manufacture a new domestic resource base rather than extract from what the land already held. It was not a shortcut. Joseph LeConte, the chemistry professor whose pamphlet supplied the Bureau's nitre-bed method, put the ripening period at roughly eighteen months; other accounts describe the beds as needing up to two years before they could be leached of usable saltpetre.
That is the clearest illustration in Taylor's account of a problem distinct from diagnosis or acquisition: timing. St. John's own assessment, recorded by September 1864, was direct about the consequence. Total domestic saltpetre production had reached 1,735,531 pounds — a real, substantial achievement for a standing start. Still, the compost piles' prolonged ripening period meant they remained largely unusable within the actual span of the war. The counterfactual is stated plainly in the same account had the war continued another year or two, the nitre plantations would have produced an estimated three to four million pounds, roughly double to triple what was actually achieved. The investment was not wrong. The diagnosis was not wrong. The organisation was not wrong. The clock ran out first.
That is the precise shape of risk this essay's modern case now faces, and it is worth stating without embellishment because the parallel needs none. The heavy rare earth separation capacity described in the previous section — the $150 million loan behind Mountain Pass's expansion — is new construction, not an existing line being scaled, and new industrial capacity of this kind does not commission on demand. Permitting, equipment procurement, commissioning, and qualification each carry their own minimum duration, largely independent of how much capital or political will stands behind them. Nothing about that timeline reflects poor diagnosis or insufficient resourcing. It reflects the same mechanism Taylor documents: a correctly identified, correctly resourced response to a real shortage, running against a calendar it does not control and cannot accelerate simply by wanting to.
The nitre beds did not fail. They did not finish in time to matter for the war they were grown to help win. Whether the heavy rare earth capacity now under construction across this file's evidence finishes in time to matter for the deadlines closing in during November and January is not yet known — but the Confederacy's own experience is a specific, dated warning against assuming that correct diagnosis and genuine investment are sufficient answers to a question that is, in the end, about arithmetic and a calendar neither side is fully able to move.
VII. Conclusion
This has been a comparison of process, not of outcome, and that distinction is worth restating plainly before drawing any conclusion from it. The Confederacy's story is finished; the rare earth story is not. Augusta never failed operationally—the Confederacy did not lose a single battle for want of powder—and the war was lost regardless, for reasons that had nothing to do with saltpetre. That fact forecloses one kind of argument this essay is not entitled to make: that resolving a recognised material dependency guarantees the larger strategic outcome. It does not, and the Confederate case proves it. What the case can offer instead is narrower and fully earned: a documented instance of triage, and a standard against which triage can be tested.
That triage had three parts, not one. Davis correctly diagnosed which link in the chain was short, redirecting effort from manufacturing capacity to raw material acquisition. St. John then distributed that acquisition across fourteen separate nitre districts specifically to deny a single point of failure. And the mechanism used to engage private producers—a government advance calibrated to reduce their capital risk, not replace their participation—worked, generating over 1.7 million pounds of domestically produced saltpetre by September 1864. Diagnosis, resilience, and mobilisation- all three- inside roughly two years, starting from a nation with almost no industrial base at all.
Measured against that standard, the modern effort passes unevenly, and the unevenness is the finding. On diagnosis, there are real signs of triage working: the DoD's intervention in MP Materials is aimed specifically at the constraint Davis would have recognised — a loan earmarked for heavy rare earth separation, not simply more mining capacity, backed by a price floor and an EBITDA guarantee built to keep capital in place regardless of what the market rewards. On distribution, the same case fails by design: one mine, one company, one facility complex carrying the weight the Confederacy spread deliberately across fourteen districts. And on mobilisation, the picture is not what it was even eighteen months ago. Capital is no longer the missing ingredient, on either side of the market. JPMorgan's Security and Resiliency Initiative direct capital into the supply side — up to $1.5 trillion over ten years, with a dedicated $10 billion Strategic Investment Group already funding companies like Perpetua Resources directly. EXIM's Project Vault addresses the demand side of the same equation — a $10 billion government loan matched by roughly $2 billion in private capital from OEMs and trading houses, structured around guaranteed offtake and fixed pricing rather than production financing. Add Vulcan Elements' $1.4 billion partnership and DoD's own package to MP Materials, and the picture is the same from every angle. Whatever is constraining this effort, it is not the absence of willing money on either side of the transaction.
What remains unresolved is time, and specifically the relationship between the capital now committed and the calendar it is meant to beat. The requirement itself is not new — the restriction traces to the 2018 NDAA, was implemented in regulation the following year, and was strengthened again in 2021. Only its final form is recent: a 2024 rule extending it to the full mine-to-magnet chain, with compliance required from 1 January 2027, and — in July 2026 — an executive order tightening the waiver process that had let the requirement go substantially unmet for years. Whether that order functions as a genuine Davis-style act or as one more layer of the same procedural apparatus this essay has already described is not yet settled. Waivers were not eliminated, only routed through a higher bar — a formal mitigation plan, or direct escalation to the National Security Council — which is either the closing of an escape hatch or the addition of one more validating step, depending on how it is used once the deadline arrives. It is also worth being precise about what the underlying law requires, since coverage often flattens the point: the statute excludes material from four named states — Russia, China, North Korea, and Iran — not non-US material generally. Supply from any allied jurisdiction qualifies. That is the structural opening every non-Chinese rare earth project discussed in this essay sits inside — not a requirement that the West build everything itself, but one that routes demand toward whichever allied suppliers can deliver, wherever they happen to be. China's own suspension of its October 2025 export controls lapses on 10 November 2026, with a second suspension clock lapsing seventeen days later—52 and 35 days, respectively, before the defence ban binds. Whether the current cycle of capital and political will actually close that capacity gap in time is not, in the relevant sense, an open question at all: on the government's own optimistic projections, domestic magnet output reaches barely a tenth of national demand by the end of 2026, and no industrial process compresses the remaining distance into the few months left before the deadline binds. What remains genuinely unresolved is not whether the gap closes — it will not — but what happens after the deadline arrives and finds it still open: whether the tightened waiver process holds under real pressure or becomes one more accommodation in a pattern with a decade of precedent behind it. This essay is being written before that second test, not the first. The first has already been decided by arithmetic. The second remains to be seen — and will be decided by whether anyone is finally willing to be Davis, rather than simply louder about the deadline Davis would have moved to meet.
Acknowledgement
This essay exists because of Paul Rainbow — a retired mining engineer, and a friend with whom I share a longstanding interest in the rare earth sector — who sent me Derek Taylor's Creating Southern Thunder: The Evolution of Confederate Gunpowder Production during the American Civil War, published in The Welebaethan in 2019, correctly judging that I would be fascinated by it. Every fact drawn from the Confederate nitre and gunpowder programme in this piece — Davis's veto, the respective roles of Gorgas, St. John, and Rains, the nitre beds and their ripening period — comes from Taylor's research. I am grateful to both: to Paul, for the introduction, and to Taylor, for the scholarship it introduced me to. Any error in how that history has been applied to the present rare earth case is mine alone, not theirs.