We have been here before in this series. The milkman ran a closed loop on a single route, lifting yesterday's empties onto the same wagon that was delivering today's bottles. Coca-Cola built a deposit system that pulled containers back at rates above 90 percent and made the customer the last leg of its own reverse network. Both were circular systems in the full sense, and both held for decades.
In both of those, the bottle belonged to one company. The dairy's bottle came back to the dairy. Coke's bottle came back to Coke, and its contour shape was designed and legally protected so that nobody could mistake it for anything else or quietly put it to another use. Ownership of the container and ownership of the brand were the same thing, and the loop ran inside the walls of a single business.
Canada and Germany arrived at a different arrangement. Competing companies sat at a table and agreed on one bottle design that all of them would use, which meant an empty could be washed and refilled by whichever of them happened to receive it. The container became a shared asset, pooled across an industry rather than owned within a brand. That version produced the highest recovery rates anyone has recorded in consumer packaging, and it is the version that came apart.
The Canadian version began in 1961, through the trade body then called the Dominion Brewers Association. What it retired was a fleet of heavy, bulky bottles in twelve ounce and twenty two ounce formats that varied from company to company. In their place went one squat brown container holding 341 milliliters. It was cheap to make, durable, and light enough to move in volume. The trade called it the stubby, and it held the Canadian market until 1984.
The agreement was possible because very few people had to agree. Canadian Breweries, Molson and Labatt controlled roughly 95 percent of Canadian beer volume by 1962 and owned nearly all of the country's 61 breweries. Three sets of executives could commit the entire national market to one container, and did.
The bottles already came back, and had for decades. Returnable glass was the ordinary way to sell beer across North America and Europe, and it ran on a deposit. The customer paid a few cents extra at the counter and collected the money on handing the empty back. Glass was costly to make and heavy to ship, so a brewer who could wash and refill a container fifteen times held a large cost advantage over one buying new glass every time. Recovery was a solved problem in 1961, and the deposit had solved it.
The unsolved problem sat one step further along. When every brewer used its own bottle, each empty had to be identified, separated, and sent back to the particular plant able to refill it. A depot taking in a mixed load faced a sorting job and then a set of small, uneconomic shipments to competitors. Sorting was expensive enough that the industry later named it as the reason for adopting a single design. What the brewers gave away in 1961 was packaging as a point of difference. What they bought was a bottle any of them could refill, which made an empty worth the same to whoever received it.
The consequences are easiest to see at the counter where the empties arrived. The Beer Store in Ontario had run a deposit return operation since 1927. By the early 2010s it was taking returns at 447 locations, recovering about 94 percent of all containers and 99 percent of refillable beer bottles. Empties went from the counter onto a conveyor into a sorting area, where a worker separated them by hand. An experienced sorter could estimate a bottle's age from its scuff marks, including the white ring that forms around the base over years of handling. Standard brown bottles were stacked together without regard to which brewer had filled them last, because either Molson Coors or Labatt would collect that stack and wash it.
Everything else took a longer road. Moosehead, Sleeman and Steam Whistle used their own bottles, so those were pulled from the flow and shipped back to the brewery that had filled them. Imported bottles such as Corona and Heineken went to glass recycling, because nobody in the country was set up to refill them. They were crushed into cullet, the broken glass that gets melted down to make new containers. A few seconds of sorting decided whether a bottle rejoined the shared pool, took an individual trip home, or left the loop for good.
For the bottles in that shared stack the trip was short. A trailer collected around 16,000 empties from a single store on a Monday morning and ran them to the Molson Coors plant near Toronto. That plant was moving roughly 3.2 million cans and bottles on an average day. Cases were cut open and the bottles went into a washer that took sixty at a time and held about 49,000 in process. Pressurized jets stripped the labels and flushed out whatever customers had posted through the neck.
They came out identical and unlabeled, and then they were graded. An electronic station photographed each bottle under a strobe, looking for a chipped mouth, worn screw threads or a cap dropped inside. About one percent failed and went down a chute. The rest went to a filler running a thousand bottles a minute, were x-rayed to confirm the fill, then capped, pasteurized to 62 degrees and cooled to 3. Any bottle too fatigued to take that stress broke and removed itself from the fleet. Two and a half hours after the empties were unloaded, filled cases were going back out the door. A bottle returned on Saturday morning was on a shelf again by Monday afternoon.
Grading is what makes reuse safe at that speed. A loop only works if the operator can separate a bottle with twenty cycles left from one that will burst under heat. The industry did it twice, once by camera at line speed and once by pasteurization, which destroyed the marginal units under controlled conditions rather than in a customer's hand.
The saving shows up in one ratio. In 2011, Ontario brewers using refillable bottles sold about 1.15 billion bottles and bought about 92 million new ones. The fleet turned over roughly twelve and a half times for every unit of new glass purchased. That is close to the fifteen refills a bottle averaged before wear retired it. The Beer Store put the energy consumed over a refillable's life at about 90 percent below a single-use container that gets recycled.
The arithmetic behind that ratio is a trade between unit cost and utilization. A refillable bottle is built heavier so it survives handling and washing. European costings from the late 1990s put it at about 2.2 times the price of a one-way glass bottle of the same size. Spread across twenty trips, it delivered packaging at roughly a tenth of the per-fill cost of that one-way bottle, and a twentieth of an aluminum can. The container stopped being a consumable bought on every order and became a fixed asset amortized across fills. Labor replaced material as the dominant cost, running at about half of production cost for refillable glass while packaging fell to a fifth.
None of that came from the deposit. A deposit is a price signal that moves a container from a household to a collection point. It says nothing about what the collection point can do next. The standard is what let a returned bottle go onto a pallet rather than into a brand-specific queue. That is the difference between a two-day cycle and a long shipment home. Every extra format in circulation multiplies the sorting work, splits the volume across more washing lines, and lengthens the return leg for whatever cannot join the common stack.
The design carried a cost that was not accidental. A shared standard raises the price of entry for anyone outside the agreement. Research published this year by Heather Thompson at Carleton University argues the stubby worked as a quiet non-tariff barrier. An American brewer wanting shelf space in Ontario had to run a dedicated line for a Canadian-only container. The refill economics that made the bottle cheap for domestic brewers also disappeared once empties had to cross a border to get home. The Big Three held a majority stake in the retail network that set the packaging requirement in the first place.
The American beverage industry ran the same system and abandoned it first. In 1947, refillables carried 86 percent of US beer volume and 100 percent of soft drinks. By 1960 refillable beer was down to 53 percent, by 1970 to 26 percent, by 1984 to 8 percent, and by 1998 to 3.3 percent. Soft drinks held on longer and then fell further, from 95 percent refillable in 1960 to 31 percent in 1980 and 0.4 percent by 1998. The steel can arrived in 1938 and took 11 percent of beer volume inside a decade, and the plastic bottle finished the job in the 1980s.
The reasons were structural rather than sentimental. Glass was heavier and more expensive to ship than the alternatives. Bottling consolidated away from local plants toward centralized production, which lengthened every return leg. Retailers had their own reason to stop, since roughly half their cost was running the deposit return and half was storing containers, with no margin in either. Then municipal curbside recycling arrived, and the cost of managing containers moved from the beverage companies to the taxpayer.
Distance is where the model actually breaks, and the tolerance is narrow. A cost-benefit study prepared for the European Commission in 2001 found that refillable glass beat one-way glass only while the bottling plant sat within about 125 kilometers of the warehouse, assuming twenty trips against a well-recycled one-way bottle. The threshold moved between roughly 100 and 150 kilometers depending on trippage and the recycling rate of the alternative. Past that radius, hauling the empties back cost more than the reuse saved. Centralization pushed network after network past the distance at which refilling pays.
Canada's standard came apart for a more specific reason. The stubby was dropped in 1984 for the taller American-style longneck, holding the same 341 milliliters. American brands were entering through licensing agreements and then free trade, and the longneck offered more surface for branding. The standard survived as long as the industry structure that produced it, and ended when the number of parties at the table grew.
Germany made the same agreement and offers the cleaner test of which half of the system carried the weight. On August 28, 1969, 142 representatives of the German mineral water trade met in Bonn-Bad Godesberg and voted to give up their own bottles for one shared design. The Perlenflasche, a 0.7 liter clear glass bottle with raised beads across the body, grew into the largest reusable pool in Europe and is still refilled up to fifty times.
Source: https://perlenflasche.de/
There the deposit survived and the shared pool did not follow it. Germany runs the strictest container deposit in Europe: a mandatory charge on one-way bottles and cans since 2003, a lower deposit of 8 or 15 cents on refillables, and automated return machines in every supermarket. Containers come back at rates most countries never reach.
German packaging law also sets a national target for the share of drinks sold in refillable containers, at 70 percent. The figure for 2023 was 34.3 percent, up 0.8 points on the prior year, in the survey the Umweltbundesamt commissions annually. The deposit is doing its job and the loop is still leaking. Much of the volume now sits in brand-specific pools and one-way formats rather than in the kind of shared standard the 1969 vote created.
Ontario is running the experiment on the other variable. The Beer Store has closed 119 locations since 2024, and its agreement with the province lets it close as many more as it chooses from 2026 onward. From January 1, 2026, grocery stores over 4,000 square feet must accept empty containers as a condition of their liquor licence. The deposit is unchanged and the standard bottle is unchanged. What is changing is the node. A concentrated network that sorted, graded and consolidated at scale is giving way to a dispersed one, whose operators handle returns as an obligation attached to a licence. Whether a refill pool holds when the collection point fragments is a question this system is about to answer in public.
Every returns operation runs on the same three parts: an incentive that gets the unit back, a place that receives and grades it, and a downstream that can absorb it. Most operators fund the first and tolerate the second. The third is where recovery value was being decided all along. A returned unit that several channels can accept is worth a different number than one that has to travel home to a single facility for a single brand.
Loop by TerraCycle learned this on a compressed timetable. It launched in 2019 as a direct-to-consumer platform running its own container fleet and its own collection leg, and that model was wound down under the weight of the logistics. The version that survived runs through retail partners, so the container comes back to a store the shopper was visiting anyway and the collection point is shared across brands. Containers in the restructured program average around eleven cycles. The reverse leg has to ride on a movement that is already happening, and the container has to be one that more than one filler can use.
For practitioners: How many distinct formats does your returns node have to sort today, and what would it cost you to cut that number in half? For your highest-volume return category, what share of units can be absorbed by any downstream channel you operate, and what share has to be routed back to one facility or one brand owner? What is the price difference between those two paths, and does anyone in your business currently measure it? Where does grading happen at line speed in your operation, and where is it still a manual exception that a person touches one unit at a time? If your primary collection node were replaced tomorrow by twenty smaller ones with no processing capability, how much of your current recovery value would survive the change? And on the incentive side, how much of your returns budget goes to getting units back, compared with what you spend on the node that decides what those units are worth once they arrive?
Sources
The Coca-Cola Company. "The History of the Coca-Cola Contour Bottle" and "11 Facts About the Coca-Cola Contour Bottle" (contour bottle designed by the Root Glass Company of Terre Haute, Indiana; design patent granted November 16, 1915; the brief called for a bottle recognizable by touch in the dark or when lying broken on the ground; shape granted trademark registration in 1977). https://www.coca-colacompany.com/about-us/history/the-history-of-the-coca-cola-contour-bottle
Genossenschaft Deutscher Brunnen eG. "Chronik" and "Mehrweg: Flaschen und Kästen" (1969 adoption of the standard bottle, GDB membership, up to 50 glass refills, 25 PET refills, 100 crate reuses). https://www.gdb.de/gdb/chronik/ and https://www.gdb.de/mehrweg/flaschen-und-kaesten/
Perlenflasche.de. "50 Jahre Perlenflasche" (August 28, 1969 decision by 142 representatives at Bonn-Bad Godesberg; Coca-Cola's entry into the German market as the trigger; Europe's largest reusable bottle pool). https://perlenflasche.de/
Butler, Colin. "Canada's stubby beer bottle finds new meaning in an age of American bluster." CBC News, 2026 (stubby introduced 1961, standard until 1984; 341 ml; Big Three at ~95 percent of the market and 61 breweries by 1962; Beer Store packaging requirement; Heather Thompson, Carleton University public history research paper, spring 2026). https://www.cbc.ca/news/canada/canada-stubby-beer-bottle-1.7556282
Teotonio, Isabel. "The average beer bottle is refilled 15 times in its environmentally-friendly life cycle." Toronto Star, June 28, 2013 (Beer Store deposit return from 1927; 447 locations; 94 percent of all containers and 99 percent of refillable bottles returned; sorting of standard vs. non-standard vs. imported bottles; 16,000 empties per trailer; 60-bottle washer holding 49,000; electronic bottle inspection with ~1 percent rejection; 1,000 bottles per minute filler; pasteurization at 62°C; 2.5-hour turnaround; 1.15 billion refillable bottles sold against 92 million purchased in 2011; 90 percent energy saving). Reproduced by the Toronto Environmental Alliance. https://www.torontoenvironment.org/average_beer_bottle_refilled_15_times
Golding, Andreas. Reuse of Primary Packaging (Main report). Brussels: European Commission, 1999 (cost table for 500 ml containers: refillable glass €0.103 over 20 trips at €0.005 per fill, one-way glass €0.047, one-way PET €0.069, aluminum can €0.103; labor at roughly 50 percent of production cost for refillable glass against packaging at 20 percent; German employment figures for refillable versus one-way). Cited via the Institute for Local Self-Reliance / GrassRoots Recycling Network. https://refillables.grrn.org/economic-benefits-2/
RDC-Environment and Pira International. Evaluation of Costs and Benefits for the Achievement of Reuse and the Recycling Targets for the Different Packaging Materials in the Frame of the Packaging and Packaging Waste Directive 94/62/EC (Draft). Brussels: European Commission, 2001 (330 ml refillable glass versus one-way glass; refillables cost less below roughly 125 km from plant to warehouse at 20 trips against a 91 percent recycling rate, 150 km at 20 trips against 42 percent, 100 km at 5 trips against 91 percent, 120 km at 5 trips against 42 percent; retailer costs split roughly half deposit-return administration and half storage). Cited via the Institute for Local Self-Reliance / GrassRoots Recycling Network. https://refillables.grrn.org/economic-benefits-2/
Container Recycling Institute. "The Decline of Refillable Beverage Bottles in the U.S." (beer and soft drink container mix tables, 1947–1998; steel can debut 1938; refillable beer at 86 percent in 1947 falling to 3.3 percent in 1998; soft drinks from 100 percent to 0.4 percent). https://www.container-recycling.org/index.php/53-facts-a-statistics/glass/428-the-decline-of-refillable-beverage-bottles-in-the-us
Institute for Local Self-Reliance / GrassRoots Recycling Network. "Canada's Experience with Refillable Beverage Containers" (near-100 percent Canadian beer bottle return rates, 15–20 reuses, Quebec return rates near 98 percent, Ontario levy on non-refillable containers, Brewers Retail recovery figures). https://refillables.grrn.org/canadas-experience-with-refillable-beverage-containers/
Umweltbundesamt. "Mehrwegquote bei Getränken verfehlt gesetzliches Ziel" (German reusable beverage packaging share of 34.3 percent in 2023 against a statutory 70 percent target; GVM survey commissioned by UBA). https://www.umweltbundesamt.de/en/node/119602
CBC News. "Got empties? Finding a place to return them in Ontario is getting harder" (119 Beer Store closures since 2024; 300-location floor to end of 2025; closures at the company's discretion from 2026; grocery stores over 4,000 sq ft required to accept empties from January 1, 2026; 1.6 billion containers collected in 2024). https://www.cbc.ca/news/canada/toronto/ontario-beer-store-deposit-return-program-future-9.7012350
Fast Company. "Why Loop's revolutionary reusable packaging platform failed, except in France" (logistics complexity of the standalone model; restructured retail program; containers averaging around 11 use cycles). https://www.fastcompany.com/91350524/loop-reusable-packaging-platform-failed-except-in-france
Thank you for reading Sunday Supply Chain Stories!