When people use plastic to pay for food, they make more impulse-based purchases, like ice cream, donuts and chips, compared to those who pay with cash, as I wrote here a few years ago. But the research on this effect has been done with adults. Now, this paper, to be published in the journal Obesity, has found the same effect in schoolchildren. Schools and parents may like the convenience of "no cash" card-based systems for children to pay for their lunches, it says, but those cashless systems are associated with less-healthy eating.
Authors David R. Just and Brian Wansink used data from a fine-grained Department of Agriculture study to compare lunchtime purchases of 2,314 public school children (in grades 1-12) in the United States. Of these, 725 attended schools whose lunchrooms were completely cash-free, while another 1,257 ate in cash-only lunchrooms. The remaining 311 went to schools where kids could pay either with cash or with a debit card. As the USDA study included information on the kids' recollection of what they had eaten, Just and Wansink could correlate food choices with method of payment.
They found that kids in the cash or debit-or-cash systems bought more healthy items like fruits and vegetables, and took in more calories from such items. On the other hand, kids at debit-only schools bought more total calories, and got fewer of their calories from healthy foods. The effect is not entirely consistent for each food category (for instance the 311 kids in the cash-or-debit category took in more calories from cheeseburgers than did kids in the cash-only group), but the overall pattern was that the children, like adults, eat more healthily when they have to pay cash.
In the study I wrote about in 2010, Manoj Thomas and his co-authors speculated that there's a "pain of payment" aspect to cash that hinders junk-food purchases. There is a physical hassle of digging around for coins and counting them out, which slows down the purchase and maybe lends itself to second thoughts. And there's also the emotional impact of watching your hard-earned pennies disappear. There's no such pain involved in showing a plastic card, and that makes it easier to go from impulse to purchase.
Just and Wansink don't speculate about these psychological possibilities, instead pointing to the mechanics of debit payment as a possible explanation for the contrast they found. Very few parents would give an 8-year-old $100 in twenties to spend on lunch between now and Thanksgiving. But in most debit systems, they write, parents pay up front for weeks, or even months, of meals. The familiar childhood constraints of cash (what I can buy depends on what is in my pocket) are removed. And if the money is spent more quickly than expected, it's not easy for parents to figure out that their child could have spent less if she'd stuck to the standard lunch and not the (usually junkier) a la carte food.
So the practical bottom line of the study is that school districts (who like the speediness and accounting efficiency of the cards) should make some effort to measure their effect on kids' diets. Moreover, the cash/debit contrast suggests some possible "nudges" that would help children make healthier food choices (and incidentally save parents money)—a debit-for-fruit-but-cash-for-cookies setup, for instance; or a debit system that lets parents set weekly limits on junk-food spending.
I think, though, there's also a more general take-away for those of us who are interested in how people influence each others' behavior. The questions about possible policies are often posed at a high level of abstraction. (Are debit-card payments for school lunches a Good Thing? Or a Bad Thing?) But the answers usually depend on slight variations in detail. In this case, for example, one kind of debit payment plan (monthly, no review) might encourage kids to eat more candy, while another (weekly, with reports to parents on spending) might do the opposite.
The old Rational Economic Man model taught that people make their choices consciously, with reference to coherent, never-changing preferences. It lacked something in realism but it made up for that with coherence and simplicity. There is no equivalent Post-Rational Economic Man model to predict with certainty what effect a given policy will have on people. So, as Michelle N. Meyer and Christopher Chabris point out in this smart look at "nudge" policies, the only way to find out which behavioral nudges actually work is to test them. School lunchrooms seem like a natural venue for comparing different approaches.
Just DR, & Wansink B (2013). School lunch debit cards are associated with lower nutrition and higher calories. Obesity (Silver Spring, Md.) PMID: 23929600
Thomas, M., Desai, K., & Seenivasan, S. (2010). How Credit Card Payments Increase Unhealthy Food Purchases: Visceral Regulation of Vices Journal of Consumer Research DOI: 10.1086/657331
Follow me on Twitter: @davidberreby