« Brains, minus the vat | Main | Lane Kenworthy's blog »
Wining about Neuroeconomics
Neuroeconomics has promise but many of the early results leave me cold. A forthcoming paper in PNAS, Marketing actions can modulate neural representations of experienced pleasantness (subs. required), has all the usual cute pictures of brain scans (see here, if you care) which are used to make the following conclusion.
Our results show that increasing the price of a wine increases subjective reports of flavor pleasantness as well as blood-oxygen-level-dependent activity in medial orbitofrontal cortex [mOFC], an area that is widely thought to encode for experienced pleasantness during experiential tasks.
In short, a $90 bottle of wine tastes better than a $10 bottle of wine even when it is the same wine. But why not just ask people which wine they like best, as many previous studies have done? How exactly does a picture of the wine-addled brain add to our knowledge? Are we really so concerned that people would lie about their experiences that we need to put them into a 3 million dollar fMRI scanner to read their brains? (I wonder if this paper was NSF funded.)
Moreover, the lessons that people are drawing from this study are absurd. One common response, for example, is "It's a marketing expert's dream; if you want people to like your product more, charge a higher price." Uh huh. And what happens when every winery raises its price, will we all purchase more wine?
Living in a market economy the association in the brain between price and quality is constantly reinforced so it's not surprising that sometimes the brain can "jump the gun" in expectation. But don't imagine that the association can be easily exploited for long. Why do you think these sorts of studies always use wine? Could it possibly be because most people can't tell the difference between a cabernet and a merlot let alone between higher and lower quality wine? But try telling people that a $5,000 car is $45,000 and let's see if the medial orbitofrontal cortex lights up with experienced pleasantness.
Thanks to Ted Frank for the pointer.
Posted by Alex Tabarrok on January 15, 2008 at 07:32 AM in Food and Drink | Permalink
Comments
Hey, I know some guys who discovered this in the 80's, when they didn't have enough money to supply a "wine tasting party" and had to refill bottles. They were poor students, and I heard about it because they were amazed that people loved the refilled bottles so much.
Posted by: odograph at Jan 15, 2008 7:37:24 AM
BTW, more seriously, hasn't recent research into the placebo effect shown that placebos affect not only how we feel, but how our bodies respond. A placebo decongestant can clear air passages,etc. That's weird stuff ... but if it is true for placebo drugs, why not placebo food and wine?
I was joking with co-workers that I was going to try to "placebo" potato chips into health food.
Posted by: odograph at Jan 15, 2008 7:40:18 AM
Um, providing a gateway to measuring subjective experience is a BAD thing in economics?
Funny, I think it's great. The first thing critics of these kinds of studies complain about is that people can lie on simple surveys (see every single critic of happiness research). Now Neuroeconomics gives an imperfect way of grappling with thier complaints. Besides, just because some studies confirm things we already know doesn't mean they are not valuable by lending more weight to those conclusions.
Alex, let's be honest. This is really about politics isn't it? Many fans of Neuroeconomics draw non-libertarian conclusions fom it and that eats you alive inside. ;)
Posted by: Student at Jan 15, 2008 8:13:59 AM
Maybe Jesus was performing a neuroeconomics experiment when he switched the water for the wine.
Of course economics is economics, so what goes for wine, goes for, say, medical expenditures. I'm not changing the subject, I'm pointing out a generality.
People believe that if they spend on medicine they will receive wellness. More spending equals more wellness, right? And the idea that the effect is strongest when the layman can't tell what is good and what is bad. Add to that the illusion of authority and the conflict of interest where the doctor who decides what is wrong, prescribes the solution and you get a pretty predictably non-ideal result. General consumer ignorance undermines the improvement in quality that other products undergo as knowledgeable marginal consumers force the improvement. Consumer ignorance is perpetuated by medical authority, low options and third-party payment systems. It usually just doesn't pay you to try to get a good deal or high quality care. There aren't enough others helping you force the system to improve.
This fallacy results in the US's high expenditures and unimpressive results. The fallacy is cited by those who compare us to other countries. It is an erroneous comparison. All it says is that we aren't spending wisely. It does not follow that a socialized system would be better than our current system. It would just cost less because there would be even less incentive to pay for healthcare. It would be even less pleasant of an experience to do so.
It is a fallacy concluded by socialized healthcare proponents that just spending less or breaking the current system will result in a de facto improvement.
Posted by: Andrew at Jan 15, 2008 8:34:22 AM
"Many fans of Neuroeconomics draw non-libertarian conclusions fom it..."
Please. As if many fans of other economics systems don't already draw non-libertarian conclusions from them. They can be just as wrong no matter what fancy new graphics they use to justify their suppositions. I don't think libertarians need to get worked up over whatever new ways these folks determine to delude themselves.
Whether they want to call it "utils" or count the number of happy neurons, there's nothing in neuroeconomics that contradicts the libertarian philosophy that every individual's happiness is their own and just as important anyone else's, as long as it's arrived at without force or fraud. The objectivists may have trouble with it, but not libertarians.
It's like studies that come to the shocking conclusion "more money makes people happy (sometimes they forget the "on average" qualifier)!" Wow, what a revelation! They make people use a single fungible resource as the generic medium of exchange, and then you are surprised that more of this fungible resource results in increased average happiness. Then they shout from the rooftops something to the effect "We've disproven marginal ordinal utility!" Whatever.
Now if you are referring to the strawman that libertarians believe the market works perfectly, I don't. I've never heard one say that either. All I say on that is that it is just as likely for unscrupulous people to seize consolidated government power as to seize wineries, and all the more damaging.
Posted by: Andrew at Jan 15, 2008 8:52:28 AM
I think experimental economics, including neuroeconomics, puts the kibosh on rational market theories. It doesn't directly refute libertarian philosophies, but I think it does say that libertarian arguments have to be framed for real (and irrational) markets.
Not to say that "irrational" is always bad. An unconscious bias toward cooperation, which some studies show - in addition to our economic nature, may be irrational but nice. And societal mechanisms which reinforce those cooperative biases might be nice as well.
Posted by: odograph at Jan 15, 2008 9:05:18 AM
I agree with Tyler. This type of result makes funding seem wasteful. I would rather see tests done on marginal utilities. Such as, on average, I can tell the difference between 2 packs of splenda and 4 in my coffee. Since 4 packs is too sweet, and I can't tell the difference between 2 and 3 packs, I use 2 packs to be more efficient in my use of splenda. Now I would pay $55 for a neuroeconomics book that talked about this and similar!!
Posted by: Brainwarped at Jan 15, 2008 9:15:49 AM
I have been to a seminar by one of the authors (Plassman). The study has limitations, true. It is also risky to generalize these results to category where quality is easier to infer. Categories where consumers have greater expertise or where product functional attributes are more important than the experiential attributes may show lower effects.
However, the authors go to great extents to show that what they are measuring is different from what the customers say when they are asked. The reported happiness and the experienced happiness do not match one-to-one.
I am still full of skepticism about the potential uses. But the fact that drinking the same thing at different "prices" lights the area of the brain that is associated to pleasure is a pretty cool result.
Posted by: londenio at Jan 15, 2008 9:17:41 AM
Read Montague's recent book "Your Brain is (Almost) Perfect" discusses this exact phenomenon. Basically our brain triggers reward signals to the earliest possible signal of quality (in this case, the price) since using value proxies is a cheap and efficient way to navigate the environment (the earlier we recognize something as valuable, the more able we are to acquire it). Now, of course, these value proxies are defeasible; had the "expensive" glass of wine made the subjects vomit, they wouldn't have liked it as much. But they're less defeasible than we'd like to admit, since value proxies in our evolutionary environment weren't prices and brands, but local environmental features.
The fallacy here is to think they're not genuinely enjoying that glass of wine that was labeled more expensive more than the other one. In fact, they are enjoying it, since they get reward signals from the expensive one and not the cheap one. They're not being duped, but responding as their brain has evolved to respond to reliable signals of quality.
Posted by: Urstoff at Jan 15, 2008 9:25:13 AM
I would take a slightly different message from the description of the research (I read the linked blog post, but not yet the PNAS article). The subjects were not being charged and not even asked to think about paying the different prices.
Instead, they were given a taste of a $10 wine and a taste of "$90 wine" which was actually the $10 wine again. If we consider these tastings as gifts provided by the experimenter to the subject, it may be that the added "experienced pleasantness" found in the tests is the result of the subject believing he is receiving a highly-priced gift and not a response to the flavor modulated by price information.
I assume that neural scanning can help distinguish between these two suggestions, that specific areas of the brain "light up" in response to gifts which are not associated with flavor responses. I see the ability to tackle such distinctions more or less directly one of the values of neuroeconomic research.
Also, contrary to the suggestion in the article abstract and the over-enthusiastic "charge more" advice, nothing in the research suggests that the subjects valued the higher priced item more than the lower priced item by as much as the price difference. So, no get rich quick results here. (People have been making over-enthusiastic claims for science for as long as there has been science; don't reject the science based on over-enthusiastic claims.)
Posted by: Mike/KP at Jan 15, 2008 9:28:11 AM
Great post. In my field of psychiatry, the results of studies like this with their pretty pictures are often implied to be more real and important than a person's subjective account of their suffering. Fortunately, it hasn't made it's way into clinical practice yet, and I predict it will be 25 years before brain imaging has any real utility in psychiatry, if then.
As a professor of mine once said, brain imaging is little more than modern day phrenology.
Posted by: Joe at Jan 15, 2008 9:38:42 AM
I encountered a similar situation in Amsterdam. Among the various snack and pastry shops, there is a shop that loudly boasts "The World's Best Shawarma." As a lover of middle eastern cuisine I couldn't wait to get my hands on a shawarma of that caliber. To tell you the truth, the shawarma did not taste any differently from any other. However, when I went up to pay and found out that my single serving cost 20 Euros, or $28, I instantly realized that I did indeed have the world's best shawarma.
This leads me to my next point that I am in agreement with Prof. Cowen as I will not be having that shawarma ever again.
Posted by: axel molotov at Jan 15, 2008 10:30:08 AM
Tyler didn't post this, it was Alex.
I like neato brain pictures, but I nevertheless feel compelled to link to Beware Neuroscience Stories.
Posted by: TGGP at Jan 15, 2008 10:47:56 AM
"I think experimental economics, including neuroeconomics, puts the kibosh on rational market theories"
Well, I'm a libertarian and I reject the popular interpretation of efficient market theory. However, I do accept that markets are more efficient and rational than bureaucracies.
In short, proving that market participants aren't always successful in fulfilling their rational expectations does not prove that the alternative is better.
The question is, can you profit from the irrationality of others? And does the political system allow enough entrants into the market that seek to profit from the irrationality of others. And do these profits bring in competition? Are these contrarians rewarded and slowly there are enough contrarians where the new approach isn't considered contrarian anymore? Or does the political system discourage learning and trial-and-error?
I only drink wine for the resveratrol. I can't taste it. Gimme the cheapest swill you can find. I put 4 Splendas in my coffee because apparently I obtain the same number of "utils" for 4 as Brainwarped does for 2.
Not many people go broke getting drunk on the most expensive wine. So, it seems some rationality prevails even at the margin.
I'm an engineer. It's my experience that always and everywhere a pretty graphic impresses more than sound reasoning. I wonder what the brain scans of researchers look like while they are looking at these brain scans.
Posted by: Andrew at Jan 15, 2008 10:51:05 AM
When you are just learning to drink and appreciate wine you are often uncertain about what you think about a particular wine. You probably don't challenge that this $50 bottle is better than that $20 bottle. If pushed to give an opinion you lean toward what you think is the more sophisticated opinion.
With more experience, comes more self-knowledge and self-confidence, and you can be more objective about which wine is better.
Posted by: John Kunze at Jan 15, 2008 10:56:45 AM
"Maybe Jesus was performing a neuroeconomics experiment when he switched the water for the wine."
Best. Line. Ev4r. =-)
Posted by: Person at Jan 15, 2008 11:00:50 AM
It's odd that some people think this is only (questionable) brain scans.
Don't those experiments with real money and real people show similar things? You heard about the chocolate and sweaters experiment, right?
I think moderate libertarianism is consistent with our fractional economic nature. The problem with extreme market philosophies is that they assume economics is (or can describe) our entire nature. That's not true for most of us.
Though there is the possibility that extremists are correctly describing their own extreme nature.
Posted by: odograph at Jan 15, 2008 11:06:22 AM
The answer to "Why?" is "to bring in a big grant, so I
can hire more grad students, and look like a bigger shot
in the department/university/field".
I would love to see a Marginal Revolution treatment of the
way the funding of academic research works. Assuming
it wouldn't hit too close to home. ;)
Posted by: Cal at Jan 15, 2008 11:57:07 AM
odograph, read your Mises. All deliberate human action is rational.
Posted by: Noah Yetter at Jan 15, 2008 12:42:38 PM
Thorstein Veblen, anyone?
Anytime "quality" of a good is difficult to measure, define, or agree on, price can become a marker of quality. For goods like wine, on which even "expert" wine tasters can disagree--often quite significantly--about "quality," where "quality" is inherently subjective, this tendency is, in fact, reinforced. Alex's comment about automobiles is, in fact, quite relevant. Ask consumers to evaluate the "quality" of the same car, sometimes with a price sticker of $90,000, sometimes with a price of $10,000, sometimes with something in-between (so we get a distribution of prices--and ex hypothesis, a distribution of quality estimates). Do you really think the answers will be all that highly correlated with price?
Posted by: Donald A. Coffin at Jan 15, 2008 12:48:43 PM
Most of wine's quality *is* the placebo effect. There's no criterion you can try to meet to chemically optimize a wine's
quality. Becoming a wine conosieur (sp) who gives the "right" answer about various wines is more of
snobby signaling game. Sorry, the theory fits the data too well.
Posted by: Person at Jan 15, 2008 1:40:59 PM
But why not just ask people which wine they like best, as many previous studies have done? How exactly does a picture of the wine-addled brain add to our knowledge? Are we really so concerned that people would lie about their experiences that we need to put them into a 3 million dollar fMRI scanner to read their brains?
That would defeat the whole purpose of the experiment. People may very well lie about their experiences, or not even so much "lie" as just let the price guide their evaluation.
But, assuming the brain scans are accurately measuring subjective pleasure (a huge assumption), they can establish a link between perceived cost and actual enjoyment, which is the point they want to make. Otherwise, they only establish a link between cost and reported enjoyment, which isn't the same thing at all.
Posted by: JC at Jan 15, 2008 1:51:58 PM
Odograph,
You might be reading the wrong experimentalists. Here is John A List pitting prospect theory against neoclassical theory, but using actual market participants and not students.
I would recommend you check out List's work before you completely give up hope on 'rational' markets.
Posted by: Dan in Euroland at Jan 15, 2008 2:34:12 PM
Next study: People are thrilled when others spend more on them.
Posted by: Lord at Jan 15, 2008 3:10:56 PM
I would like to see the experiment repeated with $10 hookers and $90 hookers.
Posted by: TomHynes at Jan 15, 2008 5:05:14 PM
