On the Squareness of Milk Containers

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?