Art Carden  

Choice Architecture in the Cafeteria: An Experiment for Colleges and Universities

The Mosquito Bite Analogy... Numeracy and Risk in Air Trave...

Last summer, during one of my first meals at Samford University, I thought I had stumbled upon clever choice architecture at one of the drink dispensers. You can get water by pushing your cup against a lever, but getting soda requires holding your cup under a dispenser and then pushing a button with your other hand (unless you are unusually dexterous). It turns out this was just an idiosyncrasy of one machine: at the others, Coke and Pepsi were just as easy to get as water.

This makes me think, though, that there is a lot of potential for an experiment--and possibly a research paper--that would meet the approval of Richard Thaler and Cass Sunstein as well as (say) Glen Whitman and Mario Rizzo, though I don't want to put words in anyone's mouths. I've seen in a few places claims that the "Freshman Fifteen" has become the "Freshman Forty," and my understanding is that soft drinks are one of the leading sources of empty calories. I wonder if rigging the drink machines at the cafeteria to make it much easier to get water than soda would make much of a difference.

Samford's Nutrition majors have to take Principles of Macroeconomics, which is our introductory economics course. Indeed, some of them are probably enrolled in my online Principles of Macro course right now. In it, they learn (or should learn) the basics of the economic way of thinking. If there is a Samford nutrition major looking for a project and reading this right now, hop to it! I look forward to seeing what you find out.

Comments and Sharing

COMMENTS (9 to date)
Tom West writes:

I remember a ski chalet where due to space restrictions, the soft-drink machine (which was self-service) was up a short flight of maybe 6-7 stairs, when the water dispenser was just beside the cash register.

I figured it halved their soft-drink sales right there.

MG writes:

I think less soda will be dispensed (will flow out of the machines), but not all that reduction will translate into "less soda drunk", as a lot of the previous excess was probably wasted. Good measure, either way.

Sadly, if the "15" has turned to "40", we better be looking for many many more culprits than a few ounces of beheaviorably sub-optimal soda!

Chris H writes:

One thing I notice in most self-serve drink machines I encounter is that water is slightly harder to get than sodas (something I notice a lot given that I pretty much only drink water, and no this isn't meant as a "holier than thou" type thing I just never liked soda).

With soda it's simple to get it one handed, just push your cup up against the right lever and you're good to go. With water, it's typically the case that you have to place the cup down and then pull down a tiny lever to get it going. And what's worse is the water dispenser is a lot less prominent and harder to find than any given soda.

Now, none of these things are big hassles and the fact that in the US water can be had for free from most restaurants should be a bigger incentive (though that incentive doesn't exist when you're on a college meal plan), I would like to see what a nudge in the opposite direction did to consumption.

But it also seems like your potential student researcher has an interesting experiment option. They could do a preliminary study by simply examining how often people get soda vs. water at the weird machine you found against at the other machines. Now of course there could be a bit of selection bias going on with people who tend to drink water anyways congregating on that machine, but it could be important to test as if you don't find any significant difference in soda consumption then scaling up the experiment by replacing all the machines is unlikely to be important. The number of marginal soda drinkers to be impacted would be too small to care about.

Also the more cynical side of me thinks you could just pitch this idea to the school adiministration on the basis that it's cheaper to have students drink water while they are unlikely to get many (and certainly not sustained) complaints about such a small change.

MingoV writes:

I don't believe the 'water easier, soda harder' scheme will work. Most people like soda more than water. Most people like to get the most bang for their buck. If beverages are part of the meal plan, then substituting nearly free water for soda makes the meal plan less valuable. Meal plans (not just sodas) are a main component of weight gain among college student: they can eat and drink as much as they want (yum, two desserts!) without paying more.

Colleges love meal plans because they make much money from them (partly because they buy substandard foods). Colleges are unlikely to give up a profitable system just to help their students maintain a healthy weight. (Yes, I'm cynical, but far too many colleges care far too much about money and far too little about student education and welfare.)

Himanshu Sanguri writes:

I have just finished the "Nudge" and now enjoying to jot down the instances around me where I can witness the Choice Architecture in play. I owe salute to all the marketing pundits for how they think and analyze the choices of buyers as well as non buyers to prepare choice menus for them. One such brilliant example is the shopping mart that works very much on he principles of credit card, i.e. buyers will on his own temptations buy the stuff which was not in his list before he entered the mart. Thus by narrating the choice menu, the buyers buying capacity swells without any force but because of a nudge :-)

Hazel Meade writes:

Why not think about this theoretically?
Suppose there is a product that is (a) bad for you in the long run but (b) addictive and very pleasurable in the short term. In the long run, the product is going to be very harmful, both personally and financially, but because humans discount the future, the short term pleasure outweighs the discounted future costs.

What sort of policies might counter this effect?
I can think of two main strategies:
1) Bring future costs into the present, possibly through some sort of insurance scheme (i.e. allow health insurers to charge people that use the product more).
2) Train people to discount the future less heavily. i.e. less living in the moment, more planning for the future.

JohnC writes:

The freshman 15 (or 40) is a myth. In fact, I recall reading that college freshmen gain less weight than same-aged kids not attending college. (And, yes, yes, weight is a crappy measurement.)

The median gain is about 5 pounds; and, on average, about 20% lose weight. (There's some, but not much, variation based on geography. And I dimly recall that there little difference between the more and the less academically challenging schools.*) The same is true of BMI changes, though students with higher BMI's were the most likely to see pronounced increases in weight. And frankly, one expects 17 and 18 year-olds to gain weight. In fact, for males, increased participation in exercise strongly correlates with weight gain (which actually makes the 2 kg change in weight kinda disappointing).

Having said all that, there are other measures that may be concerning (e.g., changes in fasting insulin levels in the early college years).

* I'm slightly skeptical of that one, as would be anyone who's seen first hand the difference between the girls of Florida and their snowbound counterparts: Bikini weather is a strong incentive.

Vera writes:

A better comparison would be pop vs. *diet* pop. They cost the same and taste about the same, addressing several complaints by other commenters.

(Some will say diet pop isn't so much healthier than regular, so you wouldn't want to actually nudge in that direction, but moot point if the test of the nudge itself is of interest. And I guarantee you college students are more concerned with weight gain than long run very slight alleged chances in cancer risk.)

Jeff Perry writes:

As I recall, the availability of carbonated soft drinks did not drive my freshman weight gain. My university was "sponsored" by Pepsi and I am not a fan. It was more from a perspective of eating new, diffrent, and more food in the dining commons. All meals were buffet style. Add to that the availability of university-made ice cream (the AG school even owned the cows) at no additional cost.

Who am I kidding, the main drivers in my weight gain were the lack of participation in organized sports and my introduction to beer.

Comments for this entry have been closed
Return to top