The Reverse Checkout: Sweden's Pant Machines, Money and a Norwegian Sidequest

The Till Runs Backwards

A customer has finished feeding cans into the machine by the supermarket entrance. There is a little paper voucher in one hand, and a trolley in the other. Somewhere behind the wall, a crushed aluminium can has joined a mountain of its relatives. At the checkout, the voucher comes off the shopping bill. For the customer, the system feels very simple. For the person running the supermarket, it is a business with a machine, a cleaning rota, a storeroom, a transporter, a contract, and a rather unusual flow of money.

This is a story about the reverse checkout. In an ordinary sale, the customer hands money over and takes a product away. With pant, the shop hands money back and takes the empty packaging. The same shop might never have sold that particular can. A tourist could have bought it in Malmö and returned it in Kall. That is not an accounting accident. That is the point of a national system. Every participating shop becomes an entry point into a shared chain of collection, counting and reimbursement.

In Sweden, the name everybody recognizes is Pantamera, but the operator is Returpack. Returpack runs the deposit system for aluminium beverage cans and disposable plastic beverage bottles. The company is owned by parts of the brewing and grocery trades. A different name, Tomra, appears on many of the actual machines. Tomra is the Norwegian machinery company. It does not run the Swedish deposit system. It makes the hardware that lets a human return 37 sticky bottles without asking a cashier to count them by hand.

So there are at least three businesses meeting at that little slot. The producer or importer has to put its drinks packaging into an approved system. Returpack keeps the central financial and collection machinery running. And the shop provides the place where somebody can bring an empty container back. The shopkeeper's basic question is not, 'How much deposit money can I earn?' That money belongs, temporarily, to someone else. The question is, 'Who pays me for operating the return station, and does it cover the cost?'

A National IOU

Start with a bottle of soft drink. Its producer or importer registers the packaging in the approved deposit system. The producer pays the appropriate deposit and fees through the chain, the retailer buys and sells the drink, and the consumer sees the drink price plus the deposit. If the label says two kronor pant, the customer has effectively put two kronor into a nationwide promise: bring this package back to the system and get the amount printed on it returned.

The person bringing it back does not have to find the shop where it was purchased. The receiving store pays out the deposit, usually as a voucher redeemable at the till, and the store gets reimbursed through Returpack. There is a practical gap between a customer's instant return and the supermarket's settlement. That is why the automated records matter. The machine supplier reads the shop's counting data, the system checks the figures, and Returpack pays the store through a process called self-billing, normally weekly. The shop is not supposed to make its margin by keeping the deposit.

This also means a small rural shop may receive an unusual mix of containers. It could sell very few large bottles, yet end up with bags of them after a busy holiday weekend. The refund liability lands locally, while the reimbursement system is national. The return machine's scanner and the shop's cash register must be capable of recognizing whatever deposit value is actually printed on each valid package.

Here is a useful legal distinction. Selling drinks in metal and plastic containers brings obligations about approved packaging and deposit labelling. But a shop is not automatically required by law to offer a public return point merely because it sells drinks. A grocer can choose whether to connect as a return station. Once it signs Returpack's shop agreement, that becomes a different matter. The agreement imposes obligations concerning accepting eligible containers, providing access, keeping the equipment operational, paying out the deposit and storing the returned material properly.

The amounts changed in 2025. Sweden increased the deposit for cans and smaller PET bottles from one krona to two, and for larger PET bottles from two kronor to three. But old labels did not magically disappear from shelves or cupboards. During the transition the exact symbol and amount on the container matter. A machine and the cashier's system have to deal with both generations. This is a small example of why a national promise needs meticulous back-office engineering.

The Machine Is Yours

The equipment is one of the most easily misunderstood parts. In the normal Swedish supermarket model, the shop buys and owns its reverse vending machine. Returpack does not simply arrive with a free, refrigerator-sized robot. The shop buys from an approved supplier, arranges space and electrical and physical installation, takes care of service contracts, and lives with the machine's failures. The Returpack customer registration itself is free, but the machine and the work around it most definitely are not.

There is no sensible single catalogue price for a pant machine. A small through-the-wall unit, a powerful high-volume machine with a sorting and compacting room, and a whole automated return hall are not comparable purchases. The hardware quote is only the beginning. There may be construction, electricity, service, maintenance, software integration, bins, consumables, staff time and valuable floor area. A broken belt or an unreadable scanner is a technical fault. A machine filled to capacity with crushed cans during a Saturday rush is an operations fault. The customer does not care which one it is. They just want the bottle to disappear.

When the shop commits to a return station, it has to keep the receiving area presentable and available in line with the agreement. The contract deals with what happens when the unit is out of order, what a shop can reject temporarily during an exceptional disruption, and what information needs to reach Returpack. It also requires the shop to redeem the return voucher correctly. The agreement specifies a minimum six-month validity for vouchers. That apparently trivial strip of paper is part of the shop's customer obligation, not a souvenir.

And somebody has to manage the back room. Most supermarket returns are compacted and collected as loose material. Returpack describes a system in which more than 85 percent of connected shops use collection with compactor trucks, commonly with separate compartments. But remote stores, islands and difficult delivery locations may use sacks collected through beverage distribution or wholesalers. The point is that the local transport arrangement matters. A shop in a city supermarket complex and a shop in a small mountain village are operating different logistics even if the customer sees the same two-kronor symbol.

The return machine is therefore a small industrial asset hiding behind a customer-service ritual. It sorts, measures and records packaging, reduces its volume, and produces the data that underwrites a money claim. Returpack receives the sorted materials downstream and gets them processed for recycling. The shop is doing a real logistics job, not merely advertising environmental virtue. But whether it is a good deal for an individual shop is an accounting question, and we can put actual numbers on part of it.

Twenty Öre and a Spreadsheet

Returpack pays a separate handling allowance, called hanteringsersättning, for receiving eligible cans and bottles. The model for shops with compacting machines has a fixed annual part and a variable amount per accepted container. Returpack builds the model around five cost categories: the machine, service, labour, premises and materials. It revises the rates using cost indexes and expected return volumes. The fixed and variable portions are paid out weekly, together with the reimbursement of the deposit amounts.

The publicly verified schedule introduced in November 2025 provides a useful concrete example. For shops in the standard loose-material collection arrangement, the variable payment was 20.3 öre for each can, 28.3 öre for each small PET bottle, and 43.9 öre for each large PET bottle, excluding VAT. The fixed payment was 25,000 kronor a year for a machine in its first seven years, and about 15,000 kronor in later years. There is a newer 2026 schedule, so do not mistake these verified November 2025 figures for a guaranteed 2026 quotation. We are using them to understand the machinery of the business.

Now give a hypothetical shop 100,000 returned containers in a year. Seventy thousand are cans, twenty thousand are small plastic bottles, and ten thousand are large plastic bottles. The variable handling payment under that late-2025 schedule would total 24,260 kronor. Add 25,000 kronor in the fixed payment for a qualifying machine, and the shop receives 49,260 kronor in handling compensation, excluding VAT. It might also have reimbursed its customers 210,000 kronor in deposits for those same containers if every one carried the new amounts. But those 210,000 kronor are not the shop's trading revenue. They are money moving through the system.

Forty-nine thousand kronor a year sounds rather different once you pay for the machine, service, floor space, staff intervention, cleaning and occasional equipment downtime. The handling allowance is designed to reflect costs across a sample of different stores. That does not guarantee each particular shop breaks even. A high-volume retailer can spread the fixed equipment cost across huge numbers of returns; a smaller shop may value the service for customer loyalty more than for its direct income. Or it might find the machine hard to justify at all.

There is also the boring but dangerous accounting corner. In April 2026 Sweden temporarily reduced VAT on most food from twelve percent to six percent. That changes the VAT treatment of deposits on those drinks, while some alcoholic drinks still have twenty-five percent VAT. Returpack has its own accounting rules for net deposits and handling compensation. The customer-facing two-kronor deposit can remain two kronor even as its VAT components change. If your bookkeeping system treats every pant line as though it were an ordinary sale, you can manufacture a surprisingly unpleasant reconciliation problem. The joke is that the empty can might be easier to process than its VAT.

A Renault, Two Brothers, and a Robot

The Norwegian sidequest begins in 1972. Two brothers, Petter and Tore Planke, were working on a machine that could accept bottles automatically. The story in Tomra's own historical account is wonderfully physical. The prototype travelled to its first shop in the back of Petter's Renault 16. There was little room to spare. The installation took place on January 2, shortly after the birth of Tore's daughter, and the prototype was soon processing around 2,000 bottles a day. That is how a world-spanning industrial company begins: not with a global strategy presentation, but with a machine stuffed into a car.

TOMRA is a name built from the Norwegian words for empty and return. The founders were solving a practical problem. Glass beverage bottles had to be checked, counted, returned and sorted. It is tempting to imagine that reverse vending was invented in one perfect flash, but simpler deposit-return machines existed before. What made Tomra distinctive was pushing the process toward reliable, automated recognition and higher-throughput handling. By 1977 the company had developed an early microprocessor-controlled machine that could be programmed for different container types.

Sweden supplied an important early customer. In 1974, Systembolaget, the Swedish alcohol monopoly, ordered 100 machines from Tomra. Then the history acquired a distinctly Swedish bureaucratic wrinkle. When an aluminium-can deposit system was being designed, the authorities had concerns about using Systembolaget as the return destination. According to Returpack's historical account, the social authorities were not enthusiastic about sending people into alcohol stores to return their empty cans, because the extra visits might encourage alcohol purchases. A recycling system had become a question about alcohol policy.

Sweden introduced the aluminium-can deposit system on March 21, 1984. There was also a fight about who should own the return machines. Retail groups did not want the central deposit organisation to own everything in their stores. They wanted to choose and own their own equipment. That decision helped create the business we recognize today, where Returpack administers the shared scheme but shops buy reverse vending machines from specialist manufacturers.

The manufacturers were experimenting with what to do to a can after identifying it. Returpack's history describes rival approaches in which cans were crushed into puck-shaped discs or flattened another way. The puck design could fit more cans in a sack, which sounds like an obvious win. But the shape also affected how lacquer could be removed from the aluminium during recycling. In other words, a tiny mechanical decision inside a supermarket machine could affect a metal refinery many miles away. That is the kind of engineering argument I would happily spend an afternoon listening to, and absolutely nobody asked for it when they came in with six empty cans.

When Aluminium Attacked the Business Model

Tomra's later history is a useful reminder that deposit machinery is tied to commodity markets. In the middle of the 1980s the aluminium market was rocked by a dramatic fall in prices. Tomra's historical timeline describes Soviet aluminium flooding the market and a price collapse of more than sixty percent. Suddenly business plans linked to aluminium recovery looked much less attractive. The company suffered a serious financial reversal and pulled back from its expansion in the United States. It was profitable again by 1987, but the path was not a straight line from clever prototype to global giant.

That is the first unexpected lesson from the Tomra detour. A reverse vending machine combines retail equipment, regulation, human behaviour and the economics of scrap material. Its success depends on far more than whether the scanner can identify a Coca-Cola bottle. Somebody has to legislate a deposit. Somebody has to build a collection and payment network. Somebody has to service the machine on a Sunday when it jams. The environmental idea becomes a working service only when all of those pieces cooperate.

There is another lesson. Tomra is not just a maker of supermarket machines. Over the decades it expanded into sensor-based sorting of recycling streams, food products and other materials. The general engineering challenge is similar: identify an object accurately, very quickly, and then send it into the right stream. Once you have become extremely good at seeing and sorting a moving object, a beer can and a piece of mixed recycling stop looking quite so different as industrial problems.

Back in Sweden, the scale is enormous. Returpack's figures for 2025 say that around 3 billion cans and plastic bottles were returned, corresponding to a return rate of 88.4 percent. That is hundreds of containers per resident each year. Most of the work begins in shops, with the machines we barely notice. The labour and maintenance are fragmented across thousands of locations, while the materials, accounting and logistics are coordinated nationally.

The Last Voucher

The newest variation is that even the little paper voucher is becoming optional. By 2026 Pantamera had developed a solution for participating shops to pay customers directly through Swish. It is not free magic either. The store needs compatible equipment, a suitable banking arrangement and sufficient liquidity in its Swish account. The scheme has setup and transaction economics of its own. But it shows the direction: the physical can is still physical, while the accounting claim can move straight into a customer's phone.

The truly clever part of pant is not that a machine can swallow a can. Machines can do much more impressive tricks. The clever part is that a stranger can arrive at a shop with an object bought somewhere else, hand it over, and get a small, reliable amount of money. The shop is then reimbursed by a system large enough to make the transaction feel completely ordinary. That ordinary feeling took laws, contracts, cost models, transport routes, thousands of service visits, a Norwegian Renault, and some very serious arguments about how flat a can should become.

The next time the machine at the supermarket makes an angry noise and spits a perfectly good bottle back at you, remember that you are not arguing with a bin. You are arguing with the front door of a national financial and industrial network. And there is probably a shopkeeper nearby who wishes you would stop trying to feed it a jam jar.