The War Economy and Our Obligations to the Future

When I learned that a new weapons company had called itself Covenant, I found I could not put the name aside. The company makes cruise missiles. Its stated purpose is deterrence: to manufacture enough weapons to persuade an adversary that war would be a mistake. I do not know its founders, and I have no reason to doubt their sincerity. What troubled me was the distance a word could travel while retaining its promise of reassurance.

In Genesis, after the flood, God makes a covenant with Noah and every living creature. He promises never again to destroy them by a flood, and places his bow in the clouds as the sign. The Hebrew word for the rainbow is also the word for an archer’s bow. Beneath the familiar colours, I read a more demanding image: the strongest power imaginable hanging up its weapon. The promise is an act of restraint. The vulnerable are offered something more dependable than the continued goodwill of the powerful. They are given a commitment that power will bind itself.

In finance, where I have spent much of my life, a covenant serves a more earthly purpose. It sets conditions under which money is lent. A borrower may undertake to limit further debts or preserve assets against which the lender has a claim. Breach those conditions, and the creditor may acquire the right to demand repayment. The promise now constrains the person who needs another’s resources. It is a less generous arrangement, though it can make cooperation possible between strangers. What remains is the recognition that power and obligation belong in the same agreement.

Attached to a weapons company, the word takes on a different meaning. Safety is promised through the capacity to inflict destruction. The promise is addressed to one set of people and enforced against another. It may sometimes have to be made: an undefended people can lose everything to an aggressor who feels no obligation to restrain himself. Yet those relying on it must ask what it will cost and what limits its guardians will accept. The promise of protection becomes dangerous when it releases power from every other promise.

The name stayed with me because it brought a moral question into the language of my daily work. I study economies through money and the material conditions that give money value. From that vantage point, our turn towards rearmament raises a problem larger than the cost of any weapons programme. A state can borrow against its people’s future production to finance activities that diminish their capacity to produce. It can call upon the savings of the old and the abilities of the young to sustain this process. For a considerable time, the transactions may register as growth. The loss appears elsewhere, and later.

We need a way of recognising when the pursuit of security begins to consume the society it is meant to secure.

Every morning, civilisation must be supplied again. Yesterday’s meals cannot feed today’s population. Electricity must be generated, water made safe, machinery maintained. Beneath the extraordinary variety of our work is this recurring obligation to the physical world. The sophistication of an economy changes how the obligation is met; it does not abolish it.

In my writing on money, I call those who draw our food and usable energy from nature the primary cooperators. The term describes a dependence that prosperity allows us to overlook. A city can contain universities, banks and orchestras because people elsewhere produce more sustenance than they consume themselves, and because others carry that surplus to where it is needed. Their work gives the rest of us time to do something other than secure the means of surviving another day.

Farmers, in turn, need machinery, medicine, credit and knowledge. An engineer who improves irrigation can contribute more to a harvest than an extra pair of hands in the field. The distinction is one of material dependence, not human worth. However elaborate our cooperation becomes, enough people must remain willing and able to renew the surplus on which everyone draws. What I call the natural order begins with this obligation to renewal. We inherit a world that will sustain us only if the work of sustaining it continues.

Money helps secure that willingness. It allows someone who produces more than he needs today to exchange the excess for a claim he can exercise elsewhere, or later. A farmer need not find a dentist who happens to want a sack of grain. Nor must he consume the whole reward for his labour before winter. A monetary system connects his contribution to the contributions of people he may never meet.

For this arrangement to endure, those who accept money must have reason to trust what it will buy. The trust extends across time. A pension represents years of work already performed, but its purchasing power depends on goods and services that other people have yet to produce. Retirement savings do not contain tomorrow’s bread. They contain an expectation that someone will grow the grain, that the baker will light his oven, and that the money saved will still command a share of what emerges.

Those who create and govern money therefore owe something to the people who accept it in exchange for an irrecoverable portion of their lives. The obligation includes preserving the conditions under which money can continue to reward useful work. A currency can retain its name while becoming less faithful to the labour entrusted to it. The loss is measured in purchasing power, but what has been lost belongs also to time: work already done, patience already exercised, years that cannot be earned again.

Credit can help the productive economy grow. A loan that pays for a more efficient power station may help produce the resources from which the loan will be repaid. The difficulty begins when we cease to distinguish between financing additional productive capacity and issuing additional claims upon capacity that has barely changed. It becomes more acute when we issue the claims while damaging the capacity itself.

In 2020, I called the temptation to confuse financial provision with material provision the money virus. Governments supported incomes and central banks sustained credit on an extraordinary scale. Preventing an interruption of work from becoming a collapse of livelihoods was a legitimate aim. The danger lay in treating the power to authorise spending as though it conferred a corresponding power over what would be available to buy. Crops still had to grow. Ships still had to arrive.

The inflation that followed had several causes. Production and transport were disrupted, spending shifted abruptly, and large public interventions helped sustain demand. Russia’s full-scale invasion of Ukraine then intensified an energy crisis already under way. These forces mattered differently across countries and over time, but their conjunction exposed the same limit. Financial support has to contend with the pace at which the physical economy can respond. When claims outrun available goods, a household can receive more money and become poorer.

As central banks raised interest rates, the familiar debate resumed over how much pain would be required to restore price stability. Alongside it, governments made commitments that would be harder to reverse than an interest-rate decision. Factories would be built, military orders placed and careers begun. The demands of war were acquiring a claim on future production even as households were discovering how much of their purchasing power they had lost.

Two prices help us understand the resulting pressure. One is the cost at which a government borrows for many years. The other is the cost of the energy on which its economy depends. Neither is a flawless instrument. A bond yield reflects expectations about growth and inflation, the supply of debt and the compensation investors require for uncertainty. Energy prices respond to technology, weather and competition as well as to scarcity. But together they tell us something about the difficulty of keeping our promises and renewing the means of keeping them.

The two can move against a government at once. A central bank can reduce the short-term interest rate it controls while lenders demand more to hold a government’s debt for thirty years. Investors may expect stronger growth; they may also fear that future inflation will erode repayment. A rate cut alone cannot tell us which explanation is right. Nor can monetary policy make an unavailable cargo of fuel arrive at a port. A country facing higher financing costs and dearer energy has less room to manoeuvre than one facing either problem alone.

The crises of my lifetime have helped form an expectation of rescue. The crash of 1987, the collapse of the technology bubble, the attacks of September 2001, the financial crisis of 2008, the eurozone emergency and the pandemic differed profoundly in their causes and human consequences. Yet each tested the ability of public institutions to contain a rupture in economic life. Across these episodes, the conviction endured that governments and central banks could mobilise the resources needed to make recovery possible.

That conviction had a material foundation. In the collapse of demand late in 2008, and again in the opening months of the pandemic, oil became dramatically cheaper as central banks eased monetary conditions. Falling energy costs did not make the damage harmless, but they relieved one pressure while policy addressed another. The relationship was never a law: oil had risen during the earlier stages of the 2008 crisis even as the Federal Reserve cut rates. Still, the subsequent rescues encouraged us to expect that financial intervention could buy time in which the productive economy would recover.

Even the world wars reveal the importance of where the means of recovery survive. During the First World War, American exports drew gold across the Atlantic, the United States lent to its allies, and the dollar gained international standing as European finances deteriorated. America experienced inflation of its own; its advantage was relative. Yet a belligerent could still draw upon substantial resources beyond the principal theatres of destruction.

During the Second World War, the United States combined vast productive capacity with an ability to hold down the cost of financing the conflict. The Federal Reserve supported a ceiling of 2.5 per cent on long Treasury yields, purchasing securities to sustain it. That was an exercise of monetary power with consequences for inflation. It operated, however, alongside American farms, factories and oilfields largely beyond the reach of the devastation inflicted overseas. After the war, American resources helped finance European reconstruction. The capacity to promise assistance had a counterpart in the capacity to supply it.

The reservoir that matters in such a crisis is the remaining capacity to repair the damage. It may consist of productive resources beyond the fighting, financial institutions able to sustain credit, or supplies that become more accessible as demand recedes. The oil shocks of the 1970s showed what happens when expensive energy and monetary instability instead reinforce one another. Their precedent should make us attentive to the conditions of recovery, and wary of assuming that a rescue which worked once remains available on the same terms.

War makes that difficulty particularly severe. It consumes fuel and industrial capacity directly. It can also destroy the infrastructure through which civilian production continues, make shipping more expensive, and turn a dependable trading relationship into an instrument of coercion. Russia’s restrictions on gas supplies and Europe’s efforts to reduce its dependence due to the war in Ukraine exposed this vulnerability with particular force. The resulting costs were part of the conflict, even when they arrived as a heating bill far from the front.

Cheap energy purchased at the price of strategic dependence can prove ruinously expensive; diversification and sanctions may be necessary. Their necessity does not make them free. Energy markets continue to discover prices under these conditions, but the price they discover is the cost of obtaining energy through the routes that remain open. A barrel in the ground is of little comfort to a country that cannot safely bring it home.

By 2026, the convergence had become visible again. In its annual economic report, the Bank for International Settlements described how the conflict in Iran had disrupted Gulf energy supplies and damaged infrastructure. Expectations of persistent inflation and concern about the fiscal consequences helped drive up long-term borrowing costs. By September, Britain’s thirty-year borrowing costs had reached levels last seen in 1998, while American long-term yields remained high by the standards of the intervening decades. The pressure was reaching the means by which production continues and the terms on which governments borrow against its future.

Nor were governments simply standing aside. Between March and early September, the United States drew down its Strategic Petroleum Reserve from roughly 415 million barrels to 285 million. Gross federal debt had passed $40 trillion. The Treasury expanded its purchases of older long-term bonds to support market liquidity, and the United States joined Japan in an intervention to support the yen. These interventions may have prevented sharper dislocation. They had not restored the earlier combination of cheap energy and easy long-term finance.

What troubles me is the possibility that relief increasingly consumes the means of providing it again. A petroleum reserve can bridge an interruption; repeated withdrawals leave less protection against the next one. Borrowing can sustain livelihoods while supplies recover; if the disruption persists, the obligations accumulate before the capacity to meet them has recovered. Support for financial markets can preserve their functioning without repairing the conditions that made support necessary. The question is whether the time being bought is being used to restore those conditions.

Military spending, meanwhile, has acquired renewed momentum. The Stockholm International Peace Research Institute estimates that the world spent about $2.9 trillion on its armed forces in 2025, the eleventh consecutive annual increase. European spending rose particularly sharply. These figures do not by themselves tell us whether governments spent too much or too little. They tell us that the claims of security are expanding and must be reconciled with other claims upon societies whose resources are finite.

The strongest case for rearmament deserves to be stated plainly. If an aggressor can destroy a power station or seize a port with impunity, the productive economy is already at risk. An air-defence system can preserve the very surplus on which peaceful life depends. Deterrence can spare resources that an invasion would consume, and military research can yield valuable civilian technologies. Any argument that treats every weapon as economically worthless has failed to account for the cost of being unable to defend oneself.

That is precisely why military expenditure needs judgement. The existence of a threat cannot settle the value of every proposed response. A government must explain how a particular commitment makes its people safer, how it can be sustained, and what conditions would permit it to end. Without those disciplines, necessity becomes a designation that expenditure acquires and never loses.

Eisenhower understood the problem from within. In 1953, the president who had commanded the Allied expeditionary forces in Europe asked Americans to reckon the cost of armaments in homes, schools and food. “This world in arms is not spending money alone,” he said. It was spending human labour and scientific ability, and narrowing the futures available to children. His authority came in part from knowing why weapons were needed. He could recognise their necessity without mistaking it for abundance.

His argument becomes more troubling when we follow it through the workings of money. Suppose a government borrows to expand its arsenal while the conflict it is engaged in damages energy supplies. It must compete for resources that have become more expensive. Households seek relief, businesses seek support, and the government takes on further obligations. If lenders come to doubt its ability to meet them without higher inflation, borrowing costs may rise. More public revenue then goes towards interest, leaving less for the investments that could strengthen the economy’s capacity to bear the burden.

Taxes can rise, spending can be redirected, and new supplies can come on stream. With credible institutions and timely decisions, the process can be interrupted. The temptation is to postpone those decisions by making further promises. The state becomes increasingly dependent on future production while its policies make that production harder to achieve.

When these policies reinforce one another, a society is borrowing against the harvest while burning the field. Its debts presume a future surplus; its conduct puts the sources of that surplus in jeopardy. The financial claim survives the damage to what must eventually redeem it. Somewhere, the two must be reconciled. That reconciliation may come to a pensioner at the shop counter long before it appears as a crisis in the bond market.

A country need not formally mobilise its whole economy for these habits to take hold. Military contracts create factories and constituencies with understandable reasons to want the work to continue. Communities come to depend on orders placed in response to an emergency. A government that contemplates reducing them must then answer for lost jobs, as well as for any reduction in preparedness. The cost of stopping becomes visible in particular towns; the possibilities forgone by continuing remain scattered across society, difficult to count and without an organised voice.

A severe downturn could entrench that dependence. As civilian investment falters, military orders may become among the expenditures a government finds easiest to defend. They offer employment, industrial activity and a purpose that can be presented as national necessity. A factory threatened with closure can be kept open by another contract. To the people whose livelihoods it preserves, that decision may be both reasonable and welcome.

Public spending can bring idle people and machinery back into use. Its value still depends on what they are asked to produce. If energy remains scarce and the conflict continues to damage supply, additional military demand may intensify competition for resources needed elsewhere. Measured activity can recover while the foundations of civilian prosperity remain impaired. War can become politically persuasive as a remedy for distress even when its continuation makes recovery more difficult.

This is the inflection I believe we have reached: the emergency is acquiring an economic constituency that may grow stronger as the civilian economy weakens. The original threat can remain real throughout. A government need never have sought war as a remedy for unemployment. Once livelihoods depend on mobilisation, however, employment can become a reason to prolong it. A market crisis might then be absorbed into the case for continuing the very commitments that should be reconsidered. A market can make a policy more expensive without making it less attractive to those who govern.

The cost is especially difficult to see in the working lives of the young. Consider an engineer whose interest in flight began with a model aeroplane. She can apply her abilities to many ends. If the most compelling offer comes from a missile company, complete with excellent colleagues, demanding problems and an assurance that the work will protect her country, she has reasons to accept it. Blaming her would be an evasion of the choices made by the society that recruited her.

The offer, however, does more than purchase her present labour. It begins to shape what she will know how to do. Over a decade she will acquire specialised judgement, form professional loyalties and help train others. These are among the most valuable things an economy produces. A budget records her salary; it cannot record the alternative body of knowledge she might have helped bring into existence. The allocation of talent is one way the demands of a temporary emergency can enter a nation’s future.

The technology industry makes this process unusually visible. Palmer Luckey’s path from Oculus to Anduril carried the skills of consumer electronics into military technology. Covenant’s founders describe an ambition to assemble an exceptional missile-engineering team. The ability is real. So is the question of how much of a generation’s inventiveness we want organised around the ability to destroy. There need be no conspiracy to arrive at an answer we would never have chosen in the abstract. Thousands of reasonable career decisions can give a society a direction.

Artificial intelligence enters the same world of decisions and constraints. Its apparent weightlessness depends on power stations, transmission lines, chips and cooling systems. The International Energy Agency reports that data-centre electricity demand rose by 17 per cent in 2025. AI may help us use energy more efficiently and accelerate useful discovery; its growing demand for physical resources is already a fact. The return on those resources will depend in part on the purposes to which the technology is put.

I do not know how to rank all the dangers that more capable machines may bring. We already know, however, that a technology’s capacity to serve human beings does not determine who will direct it. Decisions about ownership, military use and public accountability cannot be left inside a general promise of progress. If a new tool reduces the cost of using force, political judgement must still answer for its use. Technical efficiency offers no exemption from that obligation.

We should therefore be wary of waiting for markets to impose restraint. There may be no single moment of reckoning: purchasing power and civilian investment can erode over years. A sudden crash, if it comes, may narrow the choices still further. It can injure people who had little part in choosing the policy and leave them more dependent on the state that pursued it. The work of restraint belongs to political judgement while alternatives remain open.

For governments, that begins with treating the restoration of peace as a continuing responsibility of strategy. Deterrence requires a conception of what it is meant to deter and what a tolerable settlement would look like. Diplomacy, reciprocal limits and arrangements that can be verified deserve sustained effort even when trust is absent. Some negotiations will fail. A people under attack may have to fight. None of this relieves leaders of the duty to seek conditions under which the demand for sacrifice can diminish.

The financial obligation is equally concrete. Long military commitments should be accompanied by an account of how they will be paid for, including what other work will have to wait. Where borrowing is necessary, governments should explain how the productive economy will support it. Inflation should never be treated as a convenient way of collecting a sacrifice that leaders have declined to ask their citizens to make openly.

Nor can restraint mean neglecting the means of resilience. A dependable electricity supply, productive agriculture and infrastructure that can survive disruption all enlarge a country’s freedom of action. Investment in them can reduce the leverage an adversary holds. Governments should bring some of the urgency they apply to armaments to the prosaic work of maintaining these systems. The engineer deciding where to spend her life ought to find that keeping a country alive is also work to which ambition can attach itself.

This will require a different conception of economic success. The purpose of a productive surplus is to release human time: to make room for the household, for study, for work undertaken from curiosity or love. Security has value because it preserves these possibilities. A society that continually mortgages them to sustain an emergency must eventually ask what, beyond the continuation of the emergency, its sacrifices are achieving.

The covenant with Noah extends to living creatures that neither negotiate its terms nor possess the means to enforce them. That is part of its moral reach. The power that makes the promise accepts an obligation to those who cannot exact one. Our decisions, too, bind people who cannot sit at the table. Children will inherit the debts. Strangers will live with the consequences of what we destroy. People yet unborn will begin with the possibilities we have left them.

The promise of protection must remain answerable to the obligation of restraint. Power owes the vulnerable a limit on what it will ask of them, as well as protection from what others might do to them. Money and technology have enlarged our capacity to defer that reckoning. They have given us no means of cancelling it.

The bow in the clouds remains a demanding image. The hand that has the power to draw it also has the power to put it down. Everything we mean by civilisation depends on keeping that second possibility alive.