Before the Deadline

The parts arrive before the machine does. In fact, the machine has not been ordered. The shop’s owner has filled a corner of the building with boxes while leaving the proposed work area empty. Anyone watching only the deliveries might call this a business ready to grow. Anyone watching only the unsigned machine order might call it a business afraid to act.

Both would be describing part of the same week.

The owner needs the imported parts to finish work already booked. A possible tariff increase gives her a reason to buy some of them early. The new machine serves a different purpose. It would let the shop take on a larger kind of job for years to come. That choice depends on customers who have not yet signed contracts, a product price the owner cannot yet promise, and costs that may change again.

An extra box can be used next month. A machine built into a new production line needs years of work to earn back its cost. Putting those purchases in the same column, under the heading “business spending,” hides the difference that drives the decision.

This is one reason a busy warehouse can sit beside an idle building site. Action has a direction as well as a size. Money can move toward keeping today’s work going while moving away from tomorrow’s expansion.

The price and the range

Start with the simpler case. Suppose the owner knows that the delivered price of a part will rise on a fixed date. There is no doubt about the change, and she knows how many parts she will need. Buying sooner might save money. She compares the saving with interest, storage, damage, and the cost of using cash now. If the saving is large enough, an early order makes sense.

No uncertainty is needed to explain that choice. The expected future price is higher. Indeed, knowing the exact date can make early buying easier to plan.

Now change the case. The owner knows an increase is possible, but the rate and start date remain unsettled. The part might become much dearer, or an agreement might leave its price close to today’s level. Buying early could avoid a large bill. It could also leave the shop paying to store goods that turn out to have been no bargain.

The owner faces both a view about the likely price and a range of possible prices. These are separate features of her forecast. Two owners can expect the same average cost while disagreeing greatly about the chance of a very bad outcome. Two policy announcements can leave the range much the same while shifting the whole range upward.

That distinction should survive into the story we tell about the economy. A rush of orders before a tariff date does not, on its own, show that uncertainty caused the rush. A known increase could produce it. To make an uncertainty claim, we need some way to separate the effect of the expected bill from the effect of not knowing it.

There is another wrinkle. A wider range does not always mean “buy more now.” The owner might worry that her customers will cancel if their own costs rise. She could protect herself against a dearer part by ordering early, only to discover that she no longer needs the part. The inventory protects one side of the business while exposing another.

So she asks a more exact question: what work will use these boxes under several plausible outcomes? Parts for firm orders may pass that test. Parts for a hoped-for expansion may not. The legal label on a purchase does not tell us how safe it is. What matters is the chain of uses and payments behind it.

A burst that borrows from later

Federal Reserve researchers François de Soyres, Nils Goernemann, and Chris Machol examined the early-2025 rise in shipments to the United States. Their August 2025 note described frontloading ahead of expected tariffs and warned that a temporary trade boost could hide weak domestic spending in exporting economies.¹

They also found different timing across products and places. Their evidence concerns trade flows and national accounts. It does not show that every firm was stockpiling, or isolate uncertainty from expectations of higher tariffs.¹

Keep that limited result in mind when the shop’s delivery calendar changes. Imagine it brings forward some parts it otherwise would have bought later in the year. This month’s order rises. A later month’s order falls. Over that interval, the owner may purchase the same total number of parts. The first month looks strong only because we have stopped the story before the later one.

The delivery is real. The driver’s work is real. What would be mistaken is treating the full rise as evidence of a lasting increase in the shop’s customers.

A supplier trying to decide whether to add a shift needs to know this. A large order might mean the customer has won new work. It might mean the customer has moved old work onto a new calendar. If the supplier hires on the first interpretation while the buyer intended the second, both firms can make choices that look reasonable from their own desks and fit poorly together.

The buyer cannot always promise what will happen later. But she can say which part of the order is early replacement stock and which part serves new demand. That small piece of information is more helpful than a broad statement that the business is confident. It describes the decision another person must make.

A fall in orders afterward deserves the same care. It could mark a loss of final demand. It could also mark the use of goods already bought. A supplier should not dismiss the fall; lower receipts can create a serious cash problem either way. Yet the remedies differ. A short bridge through a known lull is a different need from rebuilding a business whose customers have gone.

The clock affects even our choice of words. Calling the early rise a boom and the later fall a collapse can make a single act of rescheduling sound like two changes of belief. Before accepting that account, ask what happened to the stock of goods and to sales to the final customer.

Cash gives the deadline its force

The owner cannot fill every empty space with parts. She must meet payroll. She may need cash for a repair before any customer pays her. A lender might refuse a larger credit line or offer one at a price that wipes out the expected saving.

These limits change who can take advantage of notice. A company with cheap credit and spare space can buy ahead of a known increase. A smaller rival may see the same notice, understand it just as well, and have no room to act. Giving both companies the same number of days does not give them the same set of choices.

Even where credit is available, using it has a cost beyond the interest charge. The loan may use borrowing capacity that the shop will need if a customer pays late. The owner has to weigh a possible saving on parts against a thinner cushion elsewhere. A purchase meant to make the business safer can make it brittle if it leaves too little room for the next surprise.

Storage is also more than the price of floor space. Someone must receive the goods, count them, protect them, and find them again. A box in the passage may slow work. A part may lose value if a customer’s design changes. Goods can be perfectly sound and still become the wrong goods.

These costs explain why advance notice does not let everyone escape a coming increase. Notice creates a chance to adjust. It does not supply the cash, space, or accurate demand forecast needed to use that chance.

Nor should an institution judge the success of a transition only by how much buying it draws into the weeks before a deadline. A surge can mean that people found a workable route. It can also mean that they took expensive precautions because no better route was available. We would need to know what they gave up to make those purchases.

The shop owner may postpone maintenance, delay paying herself, or use money once reserved for training. In the accounts, the inventory is an asset. In the working life of the shop, getting that asset may have squeezed several other choices. Both descriptions can be true without any bad accounting.

There is a temptation here to praise the firms that acted fastest. Speed is visible. A choice against an early purchase is less dramatic. Yet a manager who keeps enough cash to pay workers through a quiet month may have made the stronger decision. We should examine the obligation being protected before rewarding the appearance of boldness.

The purchase is not a vote

An early order is also easy to misread as support for the policy that prompted it. If someone buys a generator before an announced power cut, the purchase does not tell us whether the person approves of the cut. In the same way, importing before a tariff increase tells us about a response to expected conditions. It does not reveal a considered judgment about the wider policy.

The reverse mistake is possible too. A delayed machine order does not establish that the proposed rule is bad. The owner could face a needed safety reform that raises the price of an unsafe method. Waiting might be privately attractive even when prompt reform would protect other people. We must assess the purpose of the rule separately from the firm’s attempt to manage its effects.

This separation lets us ask harder questions. Did the institution provide enough detail for a sensible purchase? Could a reasonable transition have reduced costly rushing? Would a longer delay simply allow more harm under the old rule? What help reaches businesses that cannot buy ahead? A pile of new boxes answers none of those questions by itself.

Return to the empty work area. The owner has bought enough parts to serve work she can see. She is still deciding whether to buy the machine. Those choices need not cancel one another out. They protect different periods of her business, and they can be judged against different evidence.

She might review the stock order each week because parts move quickly. She might review the machine after a customer decision or an official notice that settles a key design question. Matching each decision to its own trigger is better than putting the whole business on hold until the news feels calm. Calm is a mood. A signed customer order is information that changes a calculation.

There should be a record of what would change her mind. If demand stays strong but the expected tariff rises, an early parts order may become more attractive. If the chance of a severe cost increase rises while customers stay uncommitted, the machine may become less attractive. If a reliable alternative input becomes available, either choice could change again.

Writing those reasons down has a practical use. When the warehouse fills, nobody has to pretend the firm has approved an expansion. When the machine order remains unsigned, nobody has to pretend the firm has stopped working. Workers and suppliers can ask about the decision that affects them rather than trying to decode one broad signal of confidence.

This is what a good account of economic adjustment should make possible. It should let us see effort without mistaking every busy week for progress. It should let us see caution without treating every delay as failure. Above all, it should keep the calendar open long enough to include the quiet month that an early rush may bring.

The boxes will leave the shop one by one. When they do, the owner will learn whether she protected booked work, guessed well about new work, or bought too much. That evidence belongs in the next decision. The deadline gave her a reason to act sooner. It did not make every early action wise.

Source notes