Canadian and Americanized milk containers side by side. Image credit: Werner Antweiler

Canada is a metric country. We measure temperatures in Celsius, lengths and distances in meters and kilometers, weights in grams and kilograms, and volumes in litres and millilitres. We fill our cars' tanks with litres of gasoline and our cars' batteries with kilowatthours of electricity. And we most certainly buy milk by the litre, not the quart.

Then why have some Canadian dairy producers such as Montreal-based Saputo (owner of the Dairyland brand) started selling milk in Americanized sizes? As the picture above shows, regular milk is still sold in metric containers, while specialty products such as micro-filtered or lactose-free milk is sold in americanized volumes. The milk carton shows an americanized size of 946 mL, which is one U.S. quart. Why this American size? They could have easily chosen 950 mL or 900 mL or 750 mL, but they chose the American size of one quart. Why are Canadian milk distributors introducing American measures?

Canada adopted metrication in 1970. I have written about Canada's incomplete metrication in the past, but metric measures is what our kids learn in school and metric measures is what scientists use all around the world. When it comes to volume measurement for fluids, we use the litre and its subunits as defined by the International System of Units (S.I.). We use the base unit, the litre, equivalent to a cube of 10-centimeters in width, height, and length. The millilitre (one-thousandth of a litre) is the most common subunit, but in other countries the decilitre (one-tenth of a litre) and the centilitre (one-hundreth of a litre) are readily found on beverage containers. For example, European beer cans come in sizes that are 33cl or 50cl.

Even though the United States adopted the metric system in 1975 as well, U.S. customary units were allowed to continue and remain the dominant commercial system of measurement today. Most American businesses continue to sell goods with volumes measured in fluid ounces, quarts, and gallons. The official definition of the fluid ounce (fl.oz.) is 29.5735295625 millilitres (mL). To avoid confusion, it should be noted that the U.S. fluid ounce is different from the imperial fluid ounce, which is slightly smaller.

Milk in Canada used to be sold everywhere in half-litre, litre, or two litre cartons, and in one-litre, two-litre, or four-litre jugs. But now it milk distributors such as Saputo are selling milk in pint (473 mL), quart (946 mL), and half-gallon (1.89 L) cartons. This means that these containers contain only 94.6% of the equivalent metric container. The purchase price can thus be kept 5.4% lower, which makes the US-measured product appear slightly cheaper than the metric-measured counterpart. Is it an example of shrinkflation? This is a topic that consumers are increasingly feeling irritated about. Consumers who look at milk containers may be easily confused: the litre and the quart containers look identical. In fact, their exterior is measurably identical. Are some Canadian dairy product distributors trying to put the wool over the eyes of Canadian shoppers? I asked Saputo for a statement. Here is what they replied.

We regularly review our packaging sizes and formats to ensure we are able to meet consumer demands for quality, convenience and value. In this case, a slightly smaller packaging format helps to offset the rising costs of ingredients, transportation and manufacturing, which allows us to continue to deliver the quality and value our consumers expect from us.

Let me parse this nebulous statement. Presumably the quality of milk has not changed, so any reference to quality is completely irrelevant. Convenience is also unaffected on a per-purchase basis, because the carton size has in fact not changed one bit; there is just less inside. If anything, consumers are slightly more inconvenienced because they have to visit their grocery story 5.7% more often to buy fresh microfiltered or lactose-free milk again. As carton size has not changed, transportation could not possibly be affected either. And if efficiencies in transportation was important, carton sizes would all be standardized rather than differentiated.

So what about value? Value is defined by the per-unit price, not the sticker price, and that per-unit price has not gone down magically. Of course, producers pass along higher input costs, but that is hardly a reason to change the product size. Reductio ad absurdum: if prices continue to rise, will carton sizes shrink continuously and eventually become miniscule? So it is fair to conclude that shrinking carton size does not generate better value for consumers.

Saputo's statement is a fine example of vacuous speech. I just hope the responsible marketing director is not one of our Sauder alumni, because we teach better marketing than that at our business school.

There is a difference between deception (illegal) and intimation (legal). Canada's Competition Act prohibits deceptive marketing practices. The content of the quart-size cartons is labelled correctly as 946 mL, so there is no deception or falsehood. However, intimation is the art of making something known indirectly, and that is giving the appearance or subtle suggestion that the quart-size container is just the same as a litre-sized container because it looks alike (because the cartons are not different in physical size or shape). Someone who is not looking closely at the label, or is unable to adjust the per-unit price of the product compared to the litre-sized version of a competitor, may end up assessing the purchase price incorrectly. Intimation plays on consumer inattention or consumer indifference.

I replied to Saputo and asked whether there were other reasonable explanations for the switch to American measures. It could be some supply chain issue with cheaper American packaging, or some such logistics problem. Or perhaps it could be a response to competitors' actions and not wanting to appear more expensive than competitors. There are plausible alternative explanations, but Saputo did not answer my questions and never connected me with one of their decision makers in senior management. I suppose it means that they were not prepared to answer my (uncomfortable?) questions, and Saputo's silence to my inquiry can only mean one of two things: they have either joined the shrinkflation bandwagon, or they are preparing us for the end of the dairy supply management regime by introducing American carton sizes now before the competition arrives. Well, I think it is safe to discount the second option; the demise of supply management is far from imminent.

Shrinkflation is also environmentally irresponsible. The 2L and 1L milk cartons that are filled with half-gallons and quarts of milk, respectively, create extra waste: exactly 5.7% more per unit of milk sold. Despite the 10-cent recycling fee on milk cartons in British Columbia, far from all milk cartons are recycled properly. (According to the Carton Council, the recovery rate is only about 60%.) Shrinkflation creates more waste when container sizes remain unchange, and just the content volume is shrunk. And in general larger containers are less wasteful, so Canadian metric containers have a small edge over their slightly smaller US counterparts.

Let me dive down into the economics of shrinkflation. There are two possible justifications of adjusting packaging size, both of which require rigorous empirical analysis. The first economic justification could be that in response to higher real(!) prices, consumers want to reduce their consumption. For this we would need to see a sufficient price elasticity of demand. Research by the USDA estimates the compensated own-price elasticity of whole milk and reduced-fat milk as -1.65 and -1.26. But most of the recent prices increases are a result of overall inflation, not changes in relative prices. As wages catch up to inflation, there is no reason to think that the real price of milk has gone up. The second justification could be a competitive equilibrium where consumers pay too much attention to sticker prices instead of per-unit prices, and producers migrate towards the smaller quart size as a result. This is some type of leader-follower price dynamic. The rationale for the leader in this pricing game is to gain market share by offering a lower sales price. There is limited competition for milk products in grocery stores: many carry only two brands (and often only to complement product lines, not as substitutes) while smaller stores only carry one.

So what can Canadian consumers do? Given the lack of choice (most grocery stores only carry a single brand of milk), it is not easy to find alternatives. But where you do have a choice, buy metric milk, and please stay strong and true to Canadian's metric system. Write to our Canadian dairy producers and distributors and tell them to stop the shrinkflation nonsense and bring back the litre. Don't americanize our Canadian milk by introducing quart-sized and pint-sized packaging if you are so proud about the rbST-hormone-free quality of Canadian milk, which is frequently put forward as a justification for keeping US competition out of Canadian grocery stores!

My own household consumes a fair bit of different types of milk—my two teenagers have a voracious appetite for milk, and the grown-ups enjoy their capuccinos and latte macchiatos. Canadian dairy producers could not be more happy about my family's level of milk consumption. Yet, we are boycotting Dairyland's non-metric cartons and instead buy Natrel, which has sensibly stuck to metric carton sizes, or Lactantia, which also offers 1.5L containers in addition to 2L cartons and 4L bags. (Natrel is a Canadian dairy co-operative based in Montreal; Lactantia is owned by Lactalis, a French multinational dairy corporation controlled by the Besnier family.)

Shrinkflation, and its even more heinous cousin, skimpflation (lowering the product quality surreptiously), while legal, are playing on consumer ignorance. As consumers, the best way to arm ourselves against these practices is to pay attention and read the labels, and make informed purchase decisions.

And to stay on the topic of the erosion of metrication in Canada, I noticed that many coffee stores and restaurants still sell coffee and wine in fluid ounces. Starbucks sells their coffees in sizes described as short (8 ounces), tall (12), grande (16), and venti (20). It would be easy to sell this as 240, 360, 480, and 600 mL—and giving us Canadians a little extra for our money. And restaurants often sell a glass of wine as either 5 or 8 ounces. Instead, it should be either 150 mL (which makes it exactly five servings per 750 mL bottle) or 250 mL (which makes it exactly three servings per 750 mL bottle). And if you like to drop a zero at the end, use centilitres instead of millilitres.

For restaurant patrons it is often difficult to determine if we are being served correctly: glasses have no mandatory fill lines. To protect consumers, the European Union introduced a Measuring Instruments Directive that requires all glassware used in restaurants to have a marking for the volume of the container, a "fill line" (2014/32/EU and WELMEC) as shown here in the picture. In European Union countries you know when you are being sold short on your beverage. Canada should get more serious about protecting consumers too. What is needed, first, is a legal requirement to require metric measures wherever something is sold ("primacy of metric measures"), with alternative measures shown in parentheses on the side or below—mostly for the benefit of our American visitors as well as Canadians who grew up with imperial measures and never fully shifted to the metric system. What is needed, second, is something akin to the EU Directive that enables consumers to ascertain that what they buy is measured correctly, and not measured by eyesight only.