Members guard points they would never pay cash for, and they feel a lost tier far more than they enjoyed the upgrade. Both habits trace to one finding. People value what they own more than the same thing they do not own yet. For a loyalty team that shapes how expiry, devaluation, status matches and demotion land.

What is the endowment effect?

The endowment effect is the gap between what people ask to give something up and what they would pay to get it. Richard Thaler named it in 1980 (Thaler, 1980). His first example was a wine buyer who paid about $5 a bottle. He later turned down $100 a bottle, though he had never paid more than $35 for one.

Daniel Kahneman, Jack Knetsch and Thaler put it plainly in 1991. People often demand much more to give up an object than they would pay to acquire it (Journal of Economic Perspectives, 1991). They tied it to loss aversion. Giving up an object hurts more than getting it feels good.

Thaler saw firms use it early. A two-week trial with a money-back guarantee works this way. At first the buyer weighs the small cost of taking the item home. Two weeks later, keeping it feels like the default, and the sale is more likely (Thaler, 1980).

What did the mug experiments show?

The classic test used coffee mugs. In 1990 Kahneman, Knetsch and Thaler gave Cornell mugs, sold for $6.00 at the campus store, to students in every other seat. Then they ran a market. If owning made no difference, about half the mugs should have changed hands. Far fewer did (Kahneman, Knetsch and Thaler, 1990).

One version split 77 students at Simon Fraser University into three groups. Sellers got a mug and named their price to sell it. Buyers named their price to buy one. Choosers picked between a mug and cash. The median seller asked $7.12. Choosers valued the mug at $3.12 and buyers at $2.87. About 12.5 trades were expected, and only three took place (Kahneman, Knetsch and Thaler).

The choosers are the key. They had no mug and spent no money, and they valued it like buyers. So the gap came from the pain of giving the mug up. Reluctance to pay played a small part.

StudyWhat it testedWhat it found
Kahneman, Knetsch and Thaler (1990)Mug sellers, buyers and choosersSellers asked $7.12; buyers paid $2.87; three trades out of 12.5 expected (paper)
Knetsch (1989), as summed up by Ericson and FusterA swap of a mug for a chocolate bar90% of those given candy kept it; 11% of mug owners chose candy (review)
Strahilevitz and Loewenstein (1998), same reviewHow long people had owned the itemThe effect grew the longer they had owned it
Plott and Zeiler (2005), same reviewMug prices after paid practice, with answers kept privateThe gap went away in their pooled data from 74 people
List (2003, 2004), same reviewSwaps at sports card showsStrong in new traders; gone in the most seasoned traders and dealers
Wagner, Hennig-Thurau and Rudolph (2009)Loss of elite statusPeople who lost a tier rated their loyalty below those never promoted (paper)

Why does the endowment effect happen?

The leading account is still loss aversion. Ericson and Fuster reviewed the field in 2013 for the Annual Review of Economics. They concluded that loss aversion remains the leading way to explain the effect (Ericson and Fuster). The open question is what people treat as their starting point.

That starting point is not always what they hold today. The review stresses reference points built from expectations and from past ownership. In a 1998 study the effect grew the longer people had owned the item (Ericson and Fuster). For a program, a member who has held a tier for years has a stronger anchor than one who reached it last month.

Is the endowment effect real, or a lab artifact?

It is real, though smaller in some settings than the early studies suggested. Charles Plott and Kathryn Zeiler gave students paid practice and kept their answers private. In their pooled data from 74 people the gap for a mug went away. They argued the gap reflected confusion about the task (Ericson and Fuster).

John List took the test to sports card shows. New traders were slow to swap, in line with the lab. The reluctance shrank with trading experience and was absent in the most seasoned traders and dealers (Ericson and Fuster). Most loyalty members are not seasoned traders of their own points or status. That makes the effect a fair planning assumption, as long as the team tests it.

How does the endowment effect show up in loyalty programs?

Points and status both feel owned. Expiry turns a dormant balance into a loss, so programs that expire points often let any activity reset the clock. Starbucks Green members keep their Stars alive with a purchase, a redemption or a digital reload of $30 or more in the month before they lapse (Starbucks Rewards Terms). Delta takes the opposite stance and tells members that their miles don’t expire (Delta SkyMiles).

A status match runs on Thaler’s trial logic. Delta gives a member with elite status at another airline the matching Medallion tier for 90 days at no charge. The member keeps it by earning set amounts of Medallion Qualifying Dollars in those 90 days (Delta Status Match Challenge). For three months the tier is the member’s own. At the deadline, losing it is a loss.

How should a program handle status it takes away?

With care. Losing a tier weighs more than gaining it did. Tillmann Wagner, Thorsten Hennig-Thurau and Thomas Rudolph tested this in the Journal of Marketing in 2009. They gave 359 members of frequent flyer programs a story about an airline. Those told they had lost their tier rated their loyalty below those never promoted at all (Wagner et al., 2009).

Their field data came from a European department store chain. For 1,423 customers who lost gold status in 2006, revenue fell 17.9% and transactions 21.5% after the demotion, against the same months a year before (Wagner et al.). The same paper adds a caution. Customers upgraded to gold that year also bought less a year later, which the authors put down to one-off spending that earned the upgrade.

Programs soften the drop in published ways. Delta’s Reclaim My Status serves members who lose a tier after a life event, such as a new child or a new job. It returns the previous tier for 90 days, and the member keeps it by meeting the spending threshold in that window (Delta Reclaim My Status).

For illustration, a program demotes 8,000 members a year who each spend $2,000 with it. If their spend fell 17.9%, as in the Wagner field data, the program would lose about $2.86 million a year. A soft landing that costs $150 a member in kept benefits costs $1.2 million. It pays if it saves more than about 42% of that drop. A random holdout of demoted members settles the question.

How GRAVTY handles the endowment effect

Tier rules, expiry rules and soft-landing rules are all program logic. The GRAVTY technology page says “Patented Visual Rules let business teams change program logic”. Visual Rules is a patented no-code rules engine. A team can change a grace period without waiting on a code release.

Measuring a demoted cohort is a question of segments. GRAVTY Agentic AI Compass lets a team ask why a metric changed and compare results across any segment (Agentic AI Compass). That covers members who lost a tier set against those who kept it. Our guide to tier strategy covers where to set the thresholds.

Frequently asked questions

Who coined the term endowment effect?

Richard Thaler coined it in a 1980 paper in the Journal of Economic Behavior and Organization (Thaler, 1980). He used it for the way people give too little weight to costs they do not pay out of pocket. Goods they hold are worth more to them than goods they do not.

What is the difference between the endowment effect and loss aversion?

Loss aversion is the broader rule, and the endowment effect is one result of it. Kahneman, Knetsch and Thaler describe loss aversion as giving up an object hurting more than getting it helps (Journal of Economic Perspectives, 1991). The endowment effect is that rule applied to things people own.

What is a simple example of the endowment effect?

The coffee mug study is the standard example. Students given a mug asked a median $7.12 to sell it, while students without one would pay a median $2.87 (Kahneman, Knetsch and Thaler, 1990). Only three of an expected 12.5 trades took place.

Does the endowment effect apply to loyalty points?

It applies in principle, and the hard loyalty evidence so far comes from status. A 2009 Journal of Marketing study found that people who lost a tier rated their loyalty below people never promoted (Wagner et al., 2009). Points work on the same logic, and programs act on it through expiry rules. A holdout test shows how large it is in a given program.

Can the endowment effect be reduced?

Yes, experience and task design both shrink it. In studies at sports card shows, the most seasoned traders showed no reluctance to swap. In lab tests with paid practice and private answers, the mug gap went away (Ericson and Fuster). Members trading points every week behave more like seasoned traders.