Sticky Decisions: Peanut Butter in a Time of Salmonella | CDC EID

EID Journal Home > Volume 16, Number 5–May 2010

Volume 16, Number 5–May 2010
Another Dimension
Sticky Decisions: Peanut Butter in a Time of Salmonella
Gülbanu Kaptan and Baruch Fischhoff
Author affiliations: Carnegie Mellon University, Pittsburgh, Pennsylvania, USA (G. Kaptan, B. Fischhoff); and Wageningen University, Wageningen, the Netherlands (G. Kaptan)

Suggested citation for this article

Abstract
We present a consumer-focused perspective on creating communications regarding potentially contaminated foods. It is illustrated with decisions that might have faced US consumers during the 2009 recalls of peanut and pistachio products. The example shows how knowledge about test results and regulatory processes might be made more useful to consumers.

December 12, 2008
A prudent, informed consumer is about to open a jar of peanut butter. It is one of her favorite foods. Mostly, she loves the taste, but she also knows it as a cheap, healthy food—although she is a little fuzzy on those details (1–3). Like most Americans, she always has peanut butter at home. Unlike most, though, she always considers Salmonella risk, before opening a new jar, then decides whether to eat it, toss it, or wait a month to see if any problems turn up. She has done some research, too. Here is her reasoning.

If she eats the peanut butter and gets salmonellosis, then she has to pay for treatment and lose work time. The US Department of Agriculture (USDA) estimated that an average case in 2007 cost $1,821 in lost wages and medical costs (4). She can imagine her case costing less (if her insurance covers the medical costs and she uses otherwise "wasted" sick-leave days) or more (if the opposite is true), but she decides to use $1,821 in her decision making.

If she throws out the peanut butter, she will lose its $3 cost. Getting a refund is such a hassle that it will still feel like losing $3.

If she waits a month, she will incur the opportunity cost of the money tied up in the peanut butter. She puts that at $0.02 (using 8% annual percentage rate).

Now, she just needs to know the probability of salmonellosis. If it is greater than 1/607, then she should toss the jar, comparing its cost ($3) with that of getting sick ($1,821). If it is greater than 1/91,050, then she should wait a month, comparing its cost ($0.02) with that for getting sick ($1,821), assuming that food inspectors find any problem by then, making the risk zero.

But money isn't everything.

If she throws out the jar, getting another will be a small hassle. She decides that the $3 covers that. She would feel bad about wasting the food but also feel good about her prudence. So, those psychological effects balance out. She vaguely worries that the same logic will lead her to throw away the next jar (and the next). That would make her feel bad.

If she waits a month, then the peanut butter might lose taste or nutritional value, or somehow "go bad." However, she can't find any good information about those possibilities and decides to ignore them. It is just a month.

Therefore, if she eats from the jar, the only important nonmonetary consequence is her getting salmonellosis. She knows that it usually involves an illness of 4–7 days, with diarrhea, fever, and abdominal cramps, and that most persons recover without treatment. However, the diarrhea sometimes calls for hospitalization, and the infection can spread to the blood. People can die, if not treated promptly with antimicrobial drugs (5). Even if USDA has not put a dollar value on suffering, she would pay a lot to avoid it. She would pay even more if she had, or was making the decision for someone with, a weak immune system.

If she tosses the jar, then she will face the risk of driving to the store to get it. She puts that at 1 chance in 100,000 of an accident, and 1 chance in 50 of that crash being fatal (6).

If she eats from the jar and it has any Salmonella bacteria, she will probably consume some of the bacteria, given how peanut butter is made. She realizes that she can put an upper bound on that risk: in 2007, a total of 1 in 6,702 Americans contracted foodborne salmonellosis, from all sources (7). Given that most Americans eat peanut butter, her chances must be smaller—unless there are problems.

That probability is much less than the 1/607 threshold. Therefore, based on purely economic consideration, there is no point in tossing the jar, even if she considers the suffering that USDA ignored. That probability is higher than the 1/91,050 threshold for waiting. But 1/6,702 is such a conservative estimate and there are so many nonmonetary reasons not to wait—and the peanut butter looks so good.

She knows how hard the US Food and Drug Administration (FDA) works on food safety, and she checks its website for recall notices (8). No reported problems! So, she opens the jar and enjoys the great taste of peanut butter.

open here to see the full-text:
Sticky Decisions: Peanut Butter in a Time of Salmonella | CDC EID

Suggested Citation for this Article
Kaptan G, Fischhoff B. Sticky decisions: peanut butter in a time of Salmonella. Emerg Infect Dis [serial on the internet]. 2010 May [date cited]. http://www.cdc.gov/EID/content/16/5/900.htm

DOI: 10.3201/eid1605.090854