Placebo Effects and Price Setting

money_pillsDan Ariely and colleagues have done a couple of very interesting studies that look at the role of price in the effect of placebos.  The first study is near and dear to the pharmacy benefit.  As reported in the Journal of the American Medical Association, Dan and colleagues recruited 82 healthy volunteers.  Each subject received a brochure describing what was allegedly a new pain relief medication, recently approved by the Food and Drug Administration (FDA).  In reality, the medication was a placebo pill.

The subjects were then randomized to one of two groups.  The first group was told that the medication had a regular price of $2.50 per pill; the second group was told the same thing, with the additional bit of information that the price had been discounted to $0.10 per pill (with no explanation of the reason behind the discount).
Subjects were subjected to a series of electrical shocks on their wrists, rating the severity of the pain caused by the shocks.  Each participant then took the placebo, and underwent another series of shocks and rated them.  The difference between the pre- and post-ratings served as a measure of the magnitude of the placebo effect.  Subjects were not told the true nature of the study, and research personnel were blinded to group assignment.

The results are fascinating:

More subjects in the “regular price” group experienced a reduction in pain ratings (pre to post), than did those in the “discounted price” group.  The results held for higher voltages as well: that is, the higher priced drug worked better in reducing relatively significant discomfort than did the discounted drug… even though both had no active chemical ingredients.

The second study was of a similar design but involved measuring the effect of an energy drink on puzzle-solving tasks among college students.  In this study, Ariely et al randomized subjects to four groups, each with a different combination of claims about the effect of the drink on mental functioning (slight improvement vs. significant improvement) and cost (discounted price vs. full price).  The following chart summarizes their results:

The effects were statistically significant: telling subjects that the drink would have a significant effect improved puzzle solving, as did charging the subjects the full price ($1.89) versus the discounted price ($0.89).  In addition, there was an interaction between expectation level and price; that is, price had a bigger effect on puzzle solving among those with higher expectations about the effectiveness of the intervention.

As Dan and his colleagues note, this leads to a bit of a conundrum: believing claims of effectiveness that have no basis in fact may cause an actual effect, and that effect may be enhanced by charging a premium for the product.

What quickly comes to mind in the pharmacy benefit is how we promote generics.  If patients have low expectations about the effectiveness of generics relative to brands, and if we emphasize that generics come at a sizable discount, are we getting everything we can out of these very effective medications?

(Note: this entry originally appeared on

About the author


View all posts