Design Observer

Job Board



Featured Writers

Michael Bierut
William Drenttel
John Foster
Jessica Helfand
Alexandra Lange
Mark Lamster
Paul Polak
Rick Poynor
John Thackara
Rob Walker


Dear Bonnie
Foster Column
From Our Archive
Partner News
Primary Sources
The Academy
Today Column
Unusual Suspects


Cities / Places
Design History
Design Practice
Disaster Relief
Film / Video
Global / Local
Graphic Design
Health / Safety
Info Design
Interaction Design
Internet / Blogs
Politics / Policy
Popular Culture
Product Design
Public / Private
Public Art
Social Enterprise
TV / Radio

Comments (39) Posted 06.13.07 | PERMALINK | PRINT

Tom Vanderbilt

On the Squareness of Milk Containers

99 Cent
99 Cent, Andreas Gursky, 1999

Do you know, or have you ever wanted to know, why milk containers are square and soft drink containers are round?

The answer is simple: Milk needs to be refrigerated. Rectangular containers occupy less shelf space, and less shelf space is more valuable when that shelf space has to be cooled. Soda pop cans and the like are kept at room temperature on the cheap metal shelving found elsewhere in the store (or arrayed into huge pyramids at the end of aisles). Of course, one might make the argument that every container should thus be square, to maximize space, but there is an added benefit to the consumer, one that seemingly overrides the extra cost in lost shelf space: The ability to comfortably hold the container in one's hand. This benefit, while great, is presumably not greater than the cost of refrigeration — hence no array of round milk jugs.

This small design study of sorts comes from Robert H. Frank's wonderful new book, The Economic Naturalist: In Search of Explanations for Everyday Enigmas. I am, admittedly, a sucker for books by economists that plumb the daily mysteries of life (e.g., Steven Landsburg's seminal Armchair Economist, or Tim Harford's delightfully insightful Undercover Economist. I am also a longtime fan of Frank (I interviewed him for Wired in 1999), so I am doubly glad that he has entered the growing field of those bringing the "dismal science" into bright little realms like the grocery store.

My favorite section is one called "The Economics of Product Design." Its sprinkling of short essays reminded me of a simple truism that I sometimes overlook when thinking about design, gazing upon some item in a store or using some kind of interface and wondering how it got to be the way it is, or indeed lamenting the way it got to be how it is (who signed off on that?). That simple fact I forget is that often, an object's design is only marginally influenced by the hand of a designer. Its design has already been preordained by market (or other) constraints. It is not merely that, as Charles Eames famously said, "design depends largely on constraints," but something more elemental. The constraint is the design.

In the red-in-tooth-and-claw rationalistic world of economics, product design, as Frank writes, "entails a trade-off between features that would be most pleasing to consumers and each seller's needs to keep prices low enough to remain competitive." He uses the example of his first car, a 1955 Pontiac Chieftain. It had a heater, which was then only an "option." Eventually, heaters became cheap enough to produce that anyone buying a car could afford their benefit (that is to say, it was no longer viable for the maker to produce cars without heaters). Or take the car's V-8 engine. The V-8, quite standard then, consumed more fuel than a V-6, but it provided more acceleration and, as Frank notes, "gasoline was still cheap in those days." Cut to the oil embargo of the 1970s and continuing price shocks. Suddenly, six cylinder and even, shockingly, four-cylinder engines were in play. Then, gas prices started to drop again, relative to other goods, paving the way for the emergence of the inflated SUVs and pickup trucks of the 1980s and 1990s. Now, of course, Ford et al. are on the skids, arguably because the benefits of the large vehicles they have focused on producing is being eclipsed by their cost. It raises the question: What portion of car design essentially derives merely from the price of gas? (and yes, I do realize the aforementioned "cost/benefit" is a narrow definition of "cost" and "benefit"; if the market were judicious enough to include the externalities of carbon emissions in the cost, the benefits of large engines to the individual user would start to look pretty slim indeed — and certainly not higher than the long-term costs to everyone else).

Frank takes this cost-benefit accounting through a number of interesting design examples (these are posed as questions that his students in "economic naturalism" originally asked). For example: Why, even though the discs are exactly the same size, do DVDs come in such larger packages than CDs? The answer, as supplied here, is that vinyl records came in tightly shrink-wrapped sleeves, 302 mm square. If one made the CD package a little less than half that size, one could fit two CDs in the racks that countless record stores had installed. The height could be kept the same, which is why you may remember CD jewel cases themselves used to wrapped with all kinds of surplus packaging, strange cousins of vinyl sleeves that were to be ripped away and discarded. DVDs, meanwhile, were riding in the wake of the VHS tape, which came in a 191 mm-high box. Keeping DVD cases the same height meant retailers (and consumers) could simply stack DVDs on their VHS shelves. One does wonder if some brave designer, some Howard Roark of the blister-pack set, stood up in an early meeting and asked if the cost of the excess packaging was indeed greater than the cost of retrofitting shelves. It could also be that Hollywood was leery of people thinking they were getting "less" for their money.

It may have also just been a sort of historical inertia — why change what we've been doing? This is why, as Frank mentions in another piece, women's clothes still button from the left, while men's button from the right. As most people are right-handed, from a pure design point of view right-handed buttoning makes more sense. But as buttons first appeared on the clothes of the wealthy, and women, unlike men, were mostly dressed by servants (most of them right-handed), women's buttons went from the left. This historical norm mostly endures today, even though I imagine few women are dressed by servants.

The cost-benefit principle lurks behind all kind of seeming design oddities. Why do refrigerators have lights but not freezers? People don't open the freezer as often, thus the perceived benefit is lower than the cost of installing it; of course, in pricier units, the freezers do have lights (maybe it will eventually become standard as the extra cost equals the marginal benefit). Why do ATMs at drive-up banks have Braille on their keypads? The answer proposed here is that companies are already making them for walk-up banks, so it's no more difficult (and actually easier) to make them for all banks. No sighted patrons will be inconvenienced, and if any non-sighted patrons come to the drive-up banks, they won't be inconvenienced. Does that answer sound convincing? Sure, but as Frank later learns, Braille at drive-up ATMS's is actually required by law — which is itself another constraint.

Sometimes, the cost-benefit formula gets a little surreal. In some cases, the costs are so low that they give rise to benefits that themselves don't seem worth much — although it depends on how you do the accounting. My friend James Surowiecki wrote recently about "feature creep," whereby electronics and other consumer goods are endowed with so many extra features ("benefits") that they actually confuse customers. This is a puzzle, as Surowiecki notes. On the one hand, it costs very little to add many features, and so Windows, for example, comes with 1500 commands, the vast majority of which will never be used. Is it that the benefit is so high to the person who does use the full range that it makes it worth including, even if it might alienate others? The problem is it that doesn't alienate others until later. When buying a product, Suroweicki notes, people crave features, the more the better. When they get home and are saddled up in front of the TV with a remote that looks like a 64-track studio mixing board, they begin to crave simplicity. Perhaps the trick would be to advertise very complex looking products that when, actually taken out of the box, were simple cubes with a few buttons.

This gives rise to my own question for Professor Frank, or any other "economic naturalists" (design naturalists are welcome as well). Why is it that, in the realm of stereo equipment, for example, one has to pay more for an aesthetic of visual simplicity? I went shopping recently and wanted a simple, but substantial, CD player/tuner sort of thing. I was dismayed by the preponderance of flashing lights and boldly advertised features (e.g, Dolby, graphic equalizer, R.D.S., etc.) splayed across the front of these units. It actually seemed like the more things the stereo looked capable of doing, the cheaper it was. When I finally saw something I liked (the Linn Classik), a very minimal looking unit with a few very small and discrete buttons, painted a kind of Cold War-chic dull bluish-gray, I was dismayed to learn it cost more than twice as much as anything else. I had to pay more for less (which somehow puts me in mind of the old Billy Joel screed, "you can't dress trashy till you spend a lot of money.") Is the issue that companies, in trying to reach the widest possible audience and sell the most units, load up their products with features (and, importantly, the appearance of features), to ensure customers that they are getting the most for their money? (let's call this the "Brookstone" effect, or maybe the SkyMall effect — it's an EasyChair and a vibrating massager and a portable cooled drink container!) And those willing to pay more, meanwhile, are given a design that seems more, well, constrained?
Share This Story

Comments (39)   |   JUMP TO MOST RECENT COMMENT >>

in canada we have milk-to-go, which comes in round (cylinder type) bottles.
06.13.07 at 04:40

Another major factor in round drink bottles is carbonation. It's a real pain in the ass to build a rectangular cardboard box that'll contain a carbonated drink!
Peat Bakke
06.13.07 at 05:21

I believe that the soda cans are round because of strength issues relative to carbonation/pressure as well as tooling requirements for the progression dies that create the can.
Mike Doell
06.13.07 at 06:38

It is possible that the round soda can is the offspring of the round, glass sode bottle? Furthermore, there's an interesting take on Coca-Cola's bottle curve found in William Lidwell's Universal Principles of Design. Could the curve also be why the bottles remained round?

So many questions... so little time.

Thanks for the article. I'll definitely pick up the book.
06.13.07 at 06:54

i did not read the article but based on your headline, i would say no.
06.13.07 at 07:06

Excuse me, first sentence.

06.13.07 at 07:10

Just had a conversation about dvd packaging sizes vs. cd case sizes recently. Ironic!
Allison Haselwander
06.13.07 at 08:50

I think Frank got it all wrong on the soda and milk shape issue. As mentioned by others, a round can or bottle is better for pressure, etc. If square for pressurized was better, you would not see so many round containers for explosive gas and pressurized liquids. I can't think of any rectangular containers with pressure but many round and square for no pressure.
I have seen round milk containers in the U.S. Usually it is organic non pasteurized (or homogenized or both?) milk from regional dairies. Here is one outfit that does square milk containers
Also, remember milk was once delivered in square glass bottles but also round depending on the bottle size, dairy and state. Costco uses a square plastic jug for gallon milk.
It is also important to note that much also depends on what country you are talking about! Tetra Pak square containers are very common for many liquids in Europe and are found in the canned milk shelf area near baking goods in American grocery stores. Why? Because Americans just can't deal with milk that does not need refrigeration until it is opened. I've tried to tell my friendly grocers, they would sell more Tetra Pak milk if he kept some with the refrigerated milk, but nope.
Strangely, Americans do OK with juice boxes for kids. Now slowly appearing are Chocolate milk boxes too. Maybe the shelf space and penny pinching has nothing to do with it but rather history, tradition, marketing, and design?
Joseph Coates
06.13.07 at 08:54

Joseph Coates
06.13.07 at 09:07

I'm glad you mentioned that feature creep article, as I've been thinking and talking about it (the article) quite a bit lately.

To take a stab at your last question, I think the implication of the high-end "simple" item, such as the stereo, is that what you are paying for is the high quality of whatever the item was intended to do in its purest form — in that case, sound. But from a packaging perspective I do think there's more going on. I think that consumer modernism has become associated with elite products (and with "good taste"). The more you pay for a hotel room, the sleeker it will be; high-end packaging often has stripped-down design; cheap kitchen products have tens of attachments, expensive ones are heavy, single-tasked and efficient. I find it truly ironic that the Modernist movement which was meant to bring efficiency in design to all people, has resulted in a surface aesthetic for the elite.

marian bantjes
06.13.07 at 10:48

Here in India, milk is sold in plastic bags (along with cooking oil and many other liquid ingredients). I think the container shapes have more to do with historical custom than anything else.
06.14.07 at 03:57

If only it were that simple. The reason packaging has different shapes is to do with how much of a commodity the contents are... perfume is at the luxury end and aesthetics rule. With milk it is often sold a zero or less profit so the cost of transport dominates; also it needs to be opened more than once hence the closure design trends are to reopenability. Baked Beans need to be sterilised for a long shelf life.. hence the use of cans. but cans can be square section but the rate of filling and sealing of a cylindrical can line dominates the cost of producing. Then there is the need to economically fill the vehicle for transport. Packaging is a very complex subject and economics come into it. Do economists?
06.14.07 at 06:24

.... and another thing.. why are the glass bottles of milk in my fridge cylindrical.. they are cool too, and because they are cylindrical maybe the cold air can circulate and cool them down more efficiently.. or any other logical hypothesis.
06.14.07 at 07:34

The organic milk examples that people have mentioned are interesting. I wonder if that's a way simply to differentiate one's product from the traditional dairies -- then again, Organic Valley et al are square. Interestingly, I've got a few bottles of Ronnybrook organic milk in glass bottles in my refrigerator. They almost look 'round,'and are easily graspable, but they do have squared off edges .

To add further intrigue, if that's the word, to the square versus round question, I just came across this news that gum will now be sold in spherical containers, rather than square, which apparently allows more gum to be stored. Bucky Fuller would be proud. But does it take up more shelf space?

here's the link:
Tom Vanderbilt
06.14.07 at 08:49

In Belgium, most of the milk boxes have a rectangular shape. We only know the typical square-beam-shaped milkboxes with missing persons on them from television. Stores don't refrigorate milk over here (not necessary because of Ultra High Temp. treatment?)

Love that kind of facts though..
06.14.07 at 09:57

If "organic" has anything to do with container shape, in theory, an organic producer would put efficient delivery ahead of subtle consumer choices like square V. round.
But that is not always the case. Although, in the case of the one round plastic organic milk jug, that is special plastic - that choice may over ride better efficiency in terms of carbon output/pollution of making new jugs. My guess is that the energy to mine oil or harvest wood, etc. plus pollution it causes, in non organic containers far exceeds either using recycled materials for organic food packaging or new corn based plastic containters - which may be more efficient to make and can be recycled or composted (in theory, depending on container...) More importantly, the corn based plastic is sustainable and that may override all other considerations for an organic producer.
My choice: glass. Milk tastes better in it, it is recyclable, can be almost any shape, and looks great holding any form of liquid drink. Downside is weight. But if the truck delivering it is powered by bio diesel or is hybrid electric diesel, it is about as low impact on environment/us as you can get.
Joseph Coates
06.14.07 at 10:06

I always thought milk was in square containers because they've been made of paper for a long time, and printing and assembling a square box is easier than a round one. Not to mention storing them flat before they're needed would take up a lot less space.

And when they made the move to primarily plastic containers, it probably made sense to keep the overall shape, to fit into the old palates an shipping containers.

Also, I would think that while groceries are concerned with space economics on their shelves, the companies that create the products would actually want LESS space per shelf, so that their product stands out.

The Oberweis glass milk containers are, to me, the most aesthetically pleasing.
06.14.07 at 10:44

Israel also sells milk in plastic bags, and people have a square-ish plastic container at home (very similar to our American jucie pitchers) that hold the plastic bag once once th corner has been snipped off. I believe this is the result of Israel's limited economy and resource during it's early history.
06.14.07 at 11:43

I've had square carboard containers of apple cider that became round as the cider fermented.

We have the plastic milk bags here too, in Canada. They're just cheaper to make and ship, I imagine.
Russell McGorman
06.14.07 at 12:54

Eastern Canada means most milk in sold in a bag of 3 bags which can be molded into and number of different shapes for form-fitting in ones' fridge. We can show our creativity by putting the bags in a variety of cool holders! :)
Tom Wolfe
06.14.07 at 03:07

I'm confused. If the diameter of a circle measures the same as the width of a square then how can round soft drink cans take up more space? Assuming the milk carton and soft drink were the same width.
06.14.07 at 08:45

Organic milk that is sold in round containers is done so because of nostalgia. Most of us are too young to remember the glass containers of the day when milk was delivered to one's door from a local dairy. Round glass connotes that "fresh from the farm" feeling that they wish to evoke.

Square milk containers were designed to be cheap, seal easily and ship easily with enough "give" that they didn't explode immediately when dropped. The ones that we have today with the little plastic screw top are a result of an addition to the manufacturing line of the old folded cardboard milk cartons—an innovation which allowed the manufacturers to say "Yes! You can re-seal our container and milk will stay fresh and clean," because they had to fight when plastic, square-ish containers were winning the market share.

It is all about the green. And I don't mean environmental.

Unfortunately. At least the plastic jugs are recyclable. Not so for the plastic-coated, folded-cardboard container with a plastic-spout in which my organic milk comes.
Michelle French
06.14.07 at 09:46

Milk is sold in sterile "aseptic" packaging (of which a popular manufacturer is the Swedish Tetrapak -- "tetra" means "four" in greek... four sides) that actually provides several benefits to storing milk. Calling it a "paper carton" is only partially correct; aseptic containers are actually made of a laminate of paper (for rigidity and printability), polyethelene (for sterility and air-tightness), and a thin layer of aluminium (to block light).

Glass bottles may look good, but they cause milk to spoil fast because they expose it to light. Plastic cartons may be cheap, but... same deal.

Milk packaged in aseptic containers actually has a non-refrigerated shelf life of about 6 months, thus needs no refrigeration during transport or during sale. Though most American stores still keep it in the dairy case.

So: why are they not round? Probably has nothing to do with economy of space. Probably has everything to do with the ability of forming a structually-sound cylindar with a paper/plastic/aluminum laminate. It's simply easier and stronger to build brick-shaped containers out of it than cylrindically-shaped.

06.15.07 at 01:50

On the flashy lights of cheaper stereo's and 'less is more' esthetic of high-end/audiophile components I have always thought that these designs are based on the assumed target audience of these products. The cheap stereo's with all the visual bells 'n whistles are targeted at teenagers. The NADs and Rotels at a more refined group that chooses sound quality over blinking lights.

To nitpick a bit: Allison's "conversation on dvd packaging sizes vs. cd case sizes recently" would be ironic if e.g. in that conversation it was remarked how no one ever talks about the subject. Otherwise the term is 'coincidental'.
Joost Dikker Hupkes
06.15.07 at 06:54

"I'm confused. If the diameter of a circle measures the same as the width of a square then how can round soft drink cans take up more space? Assuming the milk carton and soft drink were the same width.

Russell McGorman
06.15.07 at 07:14

Every adjacent cylinder forms a square without corners. Using square packaging, you get greater volume of storage in the same amount of space.
06.15.07 at 03:36

26 comments about milk -- did anyone read past the first paragraph?

Interesting post!
06.15.07 at 06:06

26 comments about milk -- did anyone read past the first paragraph?

I noticed that, too. Though I must say, at this early hour of my Saturday morning, as I drink my coffee and eat this wonderful blueberry muffin, I can't be bothered much with feature creep, but I am terribly interested in milk. :)
amber simmons
06.16.07 at 01:38

Writer doesn't seem to understand that there's a difference between a square and a rectangle. He seems to interchange the two. Kinda blows the whole article. Look at opening:

"Do you know, or have you ever wanted to know, why milk containers are square and soft drink containers are round?

The answer is simple: Milk needs to be refrigerated. Rectangular containers occupy less shelf space, and less shelf space is more valuable when that shelf space has to be cooled."

Tom: Get an editor, man. They help!
JJ Hunsecker
06.17.07 at 09:10

Here in the SF Bay Area, Berkeley Farms sells their milk in round plastic containers.
06.17.07 at 11:13

Interesting article, raises a lot of points about our packaging. I was going to say modern day packaging but then as the article hints at most of the packaging is dictated by whats already out there (vhs/dvd et cetra).

Made me pay attention to everything around me, nicely done.
Tony Goff
06.18.07 at 06:00

With all this discussion of round versus square. Check out this energy drinks glass container.

Bomba Energy from austria i believe.
r hadzor
06.18.07 at 09:25

I know I missed this question on my seventh grade mechanical aptitude standardized test. There was a right answer, most likely choice c. The best container for ice cream would be a round cyinder conainer as it had less surface area, thus the ice cream melts less on the way home from the grocery store, or when mom is dishing it out for the eight member family.

Of course, the rectangular brick cheap brand container was much easier to unpack completely and just slice with a knife. One could serve ice cream as a slab rather than being served from the cylinder containing the dairy delight where an inferior, for the consumer generated, rack and pinion scoop only served doubts between family member if treats were being dished out equally to each member.
06.19.07 at 12:31

I enjoyed reading your thoughts about design and constraints rather than the milk container factor, which seems to have drawn the previous readers.
06.21.07 at 09:24

I would imagine that drinks come in round packages so they can fit in car cup-holders...
06.22.07 at 07:02

Tetrahedral milk containers were used in France in the mid-60s. The milk was pasteurized, but it needed refrigeration. It was a valuable cultural lesson seeing milk sold in something that didn't look like a milk bottle or a milk container. I remember that wine was sold from hoses and squirted into reusable wire frame pushdown sealed bottles. You sometimes see bottle like this for high end vinegars.

Later on our family $5/day trip to Europe we'd go to museums and see ancient bottles, some for men's stuff and some for women's stuff. This sounds confusing, but by then I had been sensitized to the cultural cueing and styling issues.
06.25.07 at 12:52

The simplicity esthetic is partly class based. Anyone who reads Jilly Cooper, the novelist, knows that different classes have different aesthetics. For example, in home decorating, a working class family would make do with all the fancy moldings, but they'd paint them in garish colors with cheap paint. An aspiring middle class family would remove or hide the fancy molding, perhaps putting plywood and veneer over a hand carved door, claiming that this would make it more sanitary and easier to clean. An upper middle class family would carefully restore the molding and woodwork and use a light finish, ideally searching for an "authentic" look. An aristocratic family would leave the original molding and woodwork, but they'd drill a hole in an 18th carved door to put up a stainless steel robe hook because they need a place to hang their bathrobe.

As for feature creep, consider the MS and Apple operating system aesthetics. They both compete on feature lists, but Steve Jobs does a better job of editing and presenting his list to consumers as opposed to IT managers. Does anyone really compare the number of commands? Windows and OSX both have tons of commands, but most of them are used by hidden scripts and processes to keep the system running smoothly. Think of them as vacuum tubes or watch jewels. You don't have to see them. Oops! There goes that line of thought. I remember 7 tube radio tuners and 17 jewel watches. Marketing departments just love numbers!
06.25.07 at 12:44

I was particularly struck by the phrase "The constraint is the design." I hadn't thought of it that way before but it certainly makes sense. It relates to that more recent article about the posters and their ambiguous status between form and function.
07.03.07 at 01:35

On the Squareness of Watermelons

These design issues regarding packaging, shelving, transport, refrigeration, and the end-user are all interestingly contained in the story of the cubic watermelon.

Cubic Watermelons
Andrew Comfort
07.29.07 at 05:54

Share This Story


Tom Vanderbilt writers about architecture, design, technology, science and other topics for a wide range of publications, including The New York Times, Wired, The Financial Times, Smithsonian, Slate and Metropolis..
More Bio >>



Mr. Vignelli's Map
Vignelli Celebration: Massimo Vignelli's 1972 New York City subway map is a beautiful example of information design that was ultimately rejected by its users.

Reflections on The Ephemeral World, Part One: Ink
An elegy to the makeready — those sheets of paper, re-fed into a press to get the ink balances up to speed, leaving a series of often random, palimpsest-like, multiple impressions on a single surface — in the digital age.

Cranbrook Commencement Address
"I come to you, like all commencement speakers, as an emissary from the future." The commencement address delivered by Julie Lasky at the Cranbrook Academy of Art on May 9, 2008.

Greening the Grocery Store
It turns out that the "recycling symbol" at the bottom of my yogurt container had nothing to do with its recyclability. So why was it there? My curiosity led to findings around which I built a design class.

O.H.W. Hadank
Paul Rand held Hadank in the highest esteem because he practiced modernist formal principles even though he did not follow its dogma or style. And most important, as Rand said “Hadank was then and always an original. A profile of O.H.W. Hadank by Steven Heller...