Poor people seem to do much better than rich people at finding optimal strategies when small amounts of money are on the line:
A Brain Image of a Microeconomic Theory, Scientific American: The microeconomic law of diminishing marginal utility states that while accumulating a good—pretzels, pencils, nickels, whatever—each successive unit of that good will be less satisfying to acquire than the one before it. ...[R]esearchers at the University of Cambridge in England ... designed a study to see if the haves catch on more slowly than the have-nots when it comes to reward-based learning. Reporting in the current issue of Neuron, the scientists reveal that when a small sum of money is on the line, poorer people learn quickly how to maximize their profits, leaving their wealthier counterparts in the dust.
In a Pavlovian paradigm, a number of abstract shapes flashed in front of 14 participants. After each shape appeared for three seconds, a picture of either a 20-pence coin (roughly 40 cents) or a scrambled image followed. A card of one particular shape was always followed by the coin, and subjects were told that they could take a 20-pence piece home if they could accurately predict when the money card was the next one up.
The participants had in personal assets an average of about $1,700 in their bank accounts, which ranged from zero to nearly $6,000. The group's average income was just over $20,000, spanning from no income for students to the equivalent of about $60,000 for the most well-off of the bunch.
By measuring response time, the researchers got a sense of how quickly people learned which one of the abstract pictures indicated money would follow. They noticed an inverse correlation between how much money a person had (assets and income) and the swiftness with which they were conditioned. The poorer people tended to figure out which card signaled money ahead within about 12 trials, says neurobiologist Philippe Tobler, the study's lead author, whereas the richer people took about 35 trials.
The team next repeated the experiment while the subject's brains were scanned by an fMRI (functional magnetic resonance imaging) machine. Researchers focused their scans on the midbrain (which contains neurons or nerve cells that produce dopamine, a neurotransmitter central to reward-based learning), and the striatum, another reward-based center located under the cerebral cortex. ...
Once again, an inverse association between wealth and learning appeared, with poor people displaying more increased activity in the midbrain and striatum when compared with the more affluent subjects.
Tobler says the study, which is one of the first to try to measure marginal utility in a laboratory setting, challenges the notion held by many economists that utility comparisons cannot be made between people, because they likely value objects differently. He says, however, "It is possible that these kinds of comparisons are more easily done with money, because money is on an absolute scale." A $20 bill is worth the same no matter who had it, but one person might value it more.
I'm missing something - I don't see how this allows utility to be measured and compared across individuals. They measure learning time or the amount of activity in a defined region of the brain, not utility. Utility must be inferred, i.e. how do you turn these measures of learning or brain activity into a measure of utility? How do we know, for example, that doubling the measured level of brain activity is the same as doubling utility? We don't. In addition, suppose you get the same measure for two different people. Does that mean they experience the same level of utility? There's no reason to presume that they do. I can see how this could be used to construct an ordinal measure of utility for an individual (of course, we can already do that), but not a cardinal measure, and not a measure that is comparable across people.
And even if you can measure utility exactly by looking at brain activity and the measure is universal, i.e. comparable across individuals, if one person has a small negative reaction and another a large positive reaction to the same event, say it's taking away $1,000 from one person and giving it to another, it still doesn't mean is justified to hurt one person a little to help another person a lot. That's remains a value judgment.
Update: I found a bit more here. If this is all there is to their marginal utility measure - asking people how likely it is they'd pick up a coin on the street and finding a correlation with wealth, and then separately finding that wealth correlates with brain activity and learning speed - their claims of progress in measuring marginal utility are overblown:
...While the subjects were learning, and unlearning, to associate the reward-predicting image with the coin, the researchers used functional magnetic resonance imaging to measure the activity in their brains’ reward centers. This brain-scanning technique uses harmless radio waves and magnetic fields to measure blood flow in brain regions, which reflects brain activity.
The researchers found that the richer the subjects were—both in terms of assets and income—the slower they learned or unlearned the association between the conditioning image and the coin. The researchers found the same inverse association between wealth and their neural response in reward areas. In contrast, the subjects’ education or age did not correlate with the speed of learning.
The researchers also measured the marginal utility of money by asking the subjects how often they would be likely to pick up a coin from the street. They also found that the greater a subject’s wealth, the lower the chance the subject would retrieve the coin.
Tobler, Schultz, and colleagues wrote that “the progressively smaller gain with increasing wealth would provide decreasing reward value that could lead to the reduced learning speed. Thus, individuals for whom a financial unit has lower marginal utility would show slower acquisition and extinction than individuals for whom the same unit has higher marginal utility. Or, put differently, ‘The rich are different from you and me.’”