Loading...
In 1990, economists Daniel Kahneman, Jack Knetsch, and Richard Thaler ran a deceptively simple experiment at Cornell University. They handed brand-new coffee mugs — emblazoned with the Cornell logo, worth about $6 at the campus bookstore — to half the students in a classroom. The other half got nothing. Then they opened a market: sellers could sell their mugs, buyers could buy one. Rational economics predicted the price should converge around $3, since mugs were randomly assigned and half the room should value them above average, half below. Instead, something strange happened. Sellers demanded a median price of $5.25. Buyers offered a median of $2.25 — less than half what sellers wanted. The mug hadn't changed. The students were randomly assigned. The only difference was three minutes ...
Popular framing: Sellers were greedy and buyers were stingy; rational traders would have converged on a price.
Structural analysis: Three minutes of ownership creates an endowment effect — losing the mug is felt twice as intensely as gaining the same mug, because loss aversion is asymmetric around the reference point. The bookstore price acts as an anchor sellers use as a floor and buyers as a ceiling, locking the 2:1 spread. The status-quo bias means both groups prefer whatever they already hold; the failure to trade is a structural property of how ownership reframes value, not a failure of either side's rationality.
Treating the endowment effect as individual irrationality misses that it is a rational response to a system where transaction costs, information asymmetries, and reacquisition friction are real. Policy interventions that frame it as a bias to overcome (e.g., 'be rational, just sell') will consistently underperform compared to those that work with the reference-point architecture — redesigning transaction sequences to minimize the open-loop period or making the new state feel like a continuation rather than a substitution.
