Owned things are valued above identical unowned ones

What it is

The endowment effect: the same object is valued systematically higher by someone who owns it than by someone who does not. The classic evidence is the mug experiments (Kahneman, Knetsch & Thaler, 1990): "sellers" who received mugs at random quoted minimum selling prices roughly twice the maximum price "buyers" were willing to pay — allocation was random, the two groups were indistinguishable beforehand, and the gap was manufactured entirely by ownership status. The effect is measured as the WTA/WTP gap: the compensation demanded to give up an object exceeds the payment offered to acquire it. It is not bargaining posture, and it is not sentimental attachment to an old possession — it appears within minutes of receiving the object, on average across participants, and its source is ownership itself rather than the object's uniqueness.

Why it happens

Ownership moves the reference point. Once an object is booked under "mine," the reference point shifts from "not having" to "holding": giving it up is coded as a loss, while never acquiring it is only a forgone gain. The value function is steeper on the loss side, so "giving up what I own" and "missing what I never had" — objectively symmetric — become subjectively asymmetric: the seller demands compensation to cover the pain on the loss side, the buyer offers only what the pleasure on the gain side is worth, and the trade stalls between the two valuation curves. The core of the mechanism is that the reference point follows ownership; the object itself has not changed — the same mug, the same money, valued differently because of where they are booked.

Studying it

• Paradigm: random-assignment buying and selling experiments (Kahneman, Knetsch & Thaler, 1990): mugs are given at random to half the participants, real trading opens, and sellers' minimum asking prices are compared against choosers' maximum willingness to pay; valuation is elicited with the BDM mechanism or incentive-compatible auctions so that offers carry real consequences. Field versions use real markets such as collector trade shows (List, 2003), contrasting dealers with ordinary visitors.
• Variables: manipulations are whether ownership was assigned, market experience, and purpose of holding (own use vs resale); outcomes are WTA, WTP, swap acceptance rates, and realized trade rates.
• Uses in interface research: value measurements for features and virtual goods step on it — valuations elicited after free access are inflated by ownership; eliciting from a non-owned control group before launch yields estimates closer to the truth.
• Methodological cautions: the WTA/WTP gap is procedure-sensitive — Plott & Zeiler (2005) shrank it substantially with training and incentive-compatible elicitation, showing part of the gap comes from misconceptions about markets rather than loss aversion, so reported effect sizes must state the elicitation procedure; hypothetical offers inflate systematically, so valuation tasks need real consequences; purpose of holding must be logged as a condition — the effect shrinks or vanishes for goods held for resale (Shogren et al., 1994: market experience erased the gap for exchange goods while it persisted for goods held for use), and pooling the two kinds produces unstable conclusions.

Where it stops holding

• Money shows almost no endowment effect: currency held for exchange does not enter the same "mine" ledger as goods, which delimits the effect's domain to objects held for use.
• Market experience weakens it: traders and collectors, operating in repeated feedback-rich markets, show markedly smaller valuation reversals; the effect is strongest in one-shot, no-feedback decisions.
• The effect requires the object to be experienced as "mine": it weakens for objects that can be replaced at no cost, carry no traces of use, and can be re-obtained anywhere; when substitutes are plentiful and switching is cheap, giving up is not a typical loss.
• It amplifies beyond the standard account for objects with personal meaning and traces of use — there an emotional valuation stacks on top of the reference point, and the two pathways must be measured separately.

Applying it

• On the measurement side: randomize ownership status in willingness-to-pay and feature-value surveys — valuations elicited after free access are naturally inflated, and a non-owned control group's estimates are more usable; to de-bias, re-elicit under resale/exchange framing instructions.
• On the design side: the legitimate use is to let users genuinely own before discussing value — portable data and exportable work make "ownership" real; for settings that benefit the user, keeping their existing configuration by default (preserving what they already hold) is less costly and more defensible than compensating afterwards.
• Audit reversibility on your own side: let users truly give things up (export, delete, take their data) so that the valuation of ownership rests on real use value, not on the difficulty of leaving.
• To validate: the owned-versus-non-owned valuation contrast is the standard check — for the same object, elicit WTA from a randomly endowed group and WTP from a non-endowed group simultaneously; the gap's magnitude is the effect's strength. After launch, compare feature adoption and retention between a "configuration kept by default" arm and a "opt-in required" arm to confirm that preservation reflects value rather than inertia.

Related

• Same group: P2.05.1 Same facts framed as gains or losses reverse decisions · P2.05.2 Framing is itself an ethical decision · P2.05.3 Loss framing that manufactures anxiety approaches manipulation · P2.05.5 Free trials turn the default state into ownership · P2.05.6 Countdowns rewrite opportunity cost into loss · P2.05.7 Reference points are set by the interface, not brought by the user
• Nearby: P2.07 Anchoring effects · P2.08 The line between persuasion and manipulation · P1.14.3 Reversibility beats wording in loss situations
• Search terms: endowment effect · WTA/WTP gap · mere ownership