Mixing Memory

Money Is Umm… Food?

A while back, I linked to a paper analogically comparing money to drugs. Judging by the comments, those of you who read the paper weren’t particularly impressed by it. But if you thought the money-drug analogy was odd, I’ve got a better one for you.

If you recall, the money as a drug paper by Stephen E. G. Lea argues against a purely instrumental, “tool” theory of the subjective value of money. From the money as a drug perspective, money doesn’t serve purely practical purposes. Instead, people actually seek out money because it gives them some pleasure, like a drug. Apparently, a similar debate has been raging in the literature on eating. On one view, food is valued and desired simply to maintain energy levels (the instrumental, or tool theory of eating). On another, creatively called “positive-incentive theory,” food is desired for the pleasure of eating. In other words, food too is more like a drug than a tool.

Drawing on this theoretical similarity between money and food, Briers et al.1 hypothesize in a recent paper that the biological basis for the value of money may be the same as the basis for the value of food. As support for this hypothesis, they present the results of three experiments.

The first experiment involved two tasks. In one task, the participants, all of whom had been asked not to eat or drink (except water, tea, or coffee) for four hours prior to the experiment to ensure that they were hungry, were told that the marketing department was conducting its annual donation drive, and they were participating in a study to determine which of ten charities the department should choose to donate to. For each charity, they were asked whether they would give if they were approached and asked to donate money. In the second task, participants were told they were part of a taste test. They each ate a piece of cake, and answered several questions about its taste. Participants were divided into two conditions. In one condition, they completed the donation task first and the taste test second, and in the other, they completed the taste test first. Briers et al. argue that if the value of money and food share a biological basis, then hungry participants should treat money as more valuable, and therefore donate less. Therefore, participants who conduct the donation task prior to the taste test should be less likely to give money to charities than participants who did the taste test first. Consistent with this prediction, participants who did the taste test first said they would donate to the charities 44% of the time, while participants who did the donation task first said they would donate 36% of the time (a statistically significant difference).

Concerned that the results of this first experiment might be due to a sense of reciprocity on the part of the participants (if they got cake first, they might feel obligated to give some money in return), Briers et al. manipulated hunger more indirectly in their second study. This time, participants played a “give some” game, which Briers et al. describe thusly:

Participants were allocated 10 coins, which they could either keep or donate to their opponent, who would simultaneously make the same decision. Each coin kept was added to the participant’s account; each coin donated was doubled by the experimenter and added to the opponent’s account. To make the procedure consequential, the experimenter announced that five randomly selected participants would actually be paid according to the outcome of the game. (p. 6 of the SSRN version, linked above)

Half of the participants played the game while “the scent of baking brownies wafted into the laboratory,” and the other half with no scent. Since, as in the first experiment, participants had not eaten for at least four hours prior to the experimental session, the smell of brownies was supposed to induce hunger. Once again, it was predicted that the hungry participants (those who smelled the brownies) would give less money than those who were (presumably) less hungry. Consistent with this prediction, participants in the scent condition (hungry participants) gave an average of 3.1 conditions, while participants in the no scent condition gave an average of 4.4.

Since the first two studies showed that hungry participants gave less money, and thus seemed to value money more than less hungry participants, Briers et al. sought to demonstrate the reverse relationship (valuing money makes people value food more) in the third study. Participants were first told to imagine winning 25,000 Euros in the lottery (“high-desire-for-money condition”) or only 25 Euros (“low-desire-for-money condition”). Next, participants were told that they would be given a taste test. They were given two bowls, one with regular M&M’s and one with crispy M&M’s, and asked to compare the taste. They were told that they could eat as many of each type as they wanted. The key measure was how many M&M’s they ate. Participants were then given an eating behavior questionnaire designed to determine whether they were restrained eaters (i.e., whether they moderate their food intake with the goal of losing or maintaining their weight). This time, the prediction is that, at least for unrestrained eaters (those who don’t feel the need to eat less in order to lose weight), a high desire for money will lead to a greater desire for food, causing participants to eat more M&M’s. And for the unrestrained eaters, that’s what they found, with those in the high desire-for-money condition eating an average of 40 grams of M&Ms, and those in the low desire-for-money condition eating an average of 29 grams.

Together, Briers et al. argue, these three results provide strong preliminary evidence of a connection between the desire for food and the desire for money. Those who are hungry value money more than those who aren’t, and those who have a high desire for money are hungrier (or at least eat more) than those who don’t. There is, however, a problem with these studies. In the research on goals and needs, there are two established effects of activating a goal or need (e.g., need to eat): the valuation and devaluation effects2. In the devaluation effect, activating a goal (e.g., need to eat) causes goal-related items to be more attractive. For example, if a smoker really needs to smoke, then smoking-related items like lighters and ash trays will be more attractive than if the smoker has just smoked a cigarette. In the devaluation effect, items unrelated to the goal (e.g., for need to smoke, a stereo or a pair of shoes are unrelated to the goal) are seen as less attractive when the goal is active than when it is not. A simple explanation for the results of the first two experiments I described is that money is related to the need to eat, and thus, when need to eat is high, money is more attractive. The third experiment may be a result of the reverse relationship: food being related to the goal of attaining money. While this explanation does a connection between money and food, it may simply be that money is related to many basic goals. Smokers, for example, might be less likely to give to charity if they have a high need to smoke. In order to show that there is a unique relationship between money and food, Briers et al. would first have to show that you don’t get valuation and devaluation effects with money and goals other than hunger or need to eat. Until they do so, their hypothesized common biological basis for the value of both food and money remains unsupported.

1Briers, B., Pandelaere, M., Dewitte, S., & Warlop, L. (2006). Hungry for money: The desire for caloric resources increases the desire for financial resources and vice versa. Psychological Science, 17(11), 939-943.
2Brendl, M.C., Markman, A.B., & Messner, C. (2003). The devaluation effect: Activating a need devalues unrelated objects. Journal of Consumer Research, 29, 463-473.


  1. #1 Measure
    February 7, 2007

    How about an experiment where hungry subjects are given a choice between two rewards. Sometimes one of the rewards would be food, sometimes one would be an amount of money equal to the value of the food. But not a choice between the two.

    The other reward could be drawn from an alternating set of non-food items. These could even be set against each other in some instances, so the testers wouldn’t know what was being tested for, and to catch if one of the random items was too desireable.

    So then you run the experiment with a control group, with a stuffed group, and with a starving group, and see if money is chosen as often as food by any of those groups.

    Just something I came up with while thinking about your article.

  2. #2 sharon
    March 14, 2007

    Check out this money experiment on http://www.shivee.com