Every accountant who has valued an owner-managed business has had the conversation. The number is lower than the owner expected. There is a well-known piece of behavioural economics that explains this, and it has been in serious dispute for twenty years.
Key Takeaway
Kahneman, Knetsch and Thaler argued the gap between what owners demand and buyers offer reflects a genuine effect of reference positions on preferences, being a manifestation of loss aversion[1]. Plott and Zeiler replied that observed disparities are symptomatic of subjects' misconceptions about the nature of the experimental task, and that controlling for these makes the valuation gap disappear[2][3]. Isoni, Loomes and Sugden then reported no significant gaps in their version of either procedure, concluding the debunking explanation was also wrong[2]. Three teams, three positions.
Our Grades For These Claims
Applying the scheme from the first article in this series.
That owners and buyers frequently name different numbers is not really a research claim at all. It is observed daily in commerce and needs no laboratory support.
That this is caused by loss aversion is Grade C. It is the original interpretation, seriously contested, and the contest is unresolved.
That the laboratory gap is an artefact of elicitation procedure is also Grade C, contested by a third team from a different direction.
That market experience reduces the effect is Grade C, resting on a single well-known paper we did not obtain.
Our position: this is the most thoroughly disputed literature this series has covered, and the practical section of this article deliberately does not depend on which side is right.
A Note On Method
Everything here is verified to August 2026.
We obtained portions of the original 1990 paper from a university-hosted copy, and quote from its own text[1].
The 2005 challenge and the 2011 comment reach us through a working paper reviewing the literature[2], publisher records[3][4] and the authors' own published reply[3]. We did not obtain any of the three in full.
We state no WTA to WTP ratios and no trading percentages, because our sources give them inconsistently or in fragments.
Several claims reach us through an encyclopedia article hosted as a PDF and through reference lists in working papers, and we flag each.
All arithmetic is ours and illustrative.
This article reviews behavioural economics. It is not valuation, investment or legal advice, and nothing in it is a method for valuing any business.
Where The Term Came From
The origin, from the 1990 paper itself.
"Thaler (1980) labeled the increased value of a good to an individual when the good becomes part of the individual's endowment the 'endowment effect.'"[1]
The paper continues: "This effect is a manifestation of 'loss aversion,' the generalization that losses are weighted substantially more than objectively commensurate gains."[1]
Two observations, ours.
The endowment effect is presented as a consequence of something more general, not as a standalone finding. That matters, because an attack on the endowment effect is not automatically an attack on loss aversion, and vice versa.
And the definition is about ownership changing valuation, which is a strong claim. It says the same object is worth more to you at ten past three, having acquired it at three o'clock, than it was at five to.
The Original Experiments
What was done.
Kahneman, Knetsch and Thaler published Experimental Tests of the Endowment Effect and the Coase Theorem in the Journal of Political Economy, 98(6), 1325–1348, in 1990[4].
Two paradigms run through this literature, and a review paper distinguishes them[2].
The exchange paradigm, associated with Knetsch, gives participants one good and offers a swap for another. The prediction under standard theory concerns what proportion should trade; the review records that Knetsch's results strongly reject this prediction[2].
The valuation paradigm, described by the same review as often referred to by its result, the WTP-WTA gap, notes that non-reference dependent theory predicts that the WTA and WTP for an item should be approximately equal, apart from any income effect, whereas a substantial literature had noted that WTA and WTP were typically far apart[2].
We state no trading percentages. Our sources report them in fragments and inconsistently, and we could not reconcile them against the paper.
The Control That Made It Convincing
The methodological feature that gave the original its force, and it is genuinely elegant.
A review records that a major innovation of Kahneman, Knetsch, and Thaler (1990) was to show that while the WTP-WTA gap appeared for goods, it did not appear for induced value rounds[2].
Explaining that, and this explanation is ours. An induced value round replaces the mug with a token whose cash value the experimenter has told the participant. There is no sentiment, no taste, no ownership feeling available. It is a pure test of whether people can operate the trading mechanism.
Two consequences.
If a gap appeared in the induced value rounds too, it would suggest participants simply did not understand the procedure, and the goods result would mean nothing.
No gap appeared there. So for fifteen years, the natural reading was that the goods result could not be dismissed as confusion, because the same people handled the mechanism correctly when there was nothing to be attached to.
That is a well-designed control, and the challenge to it is the subject of the section after next.
The Consequence The Authors Drew
What the finding was taken to imply, in the authors' own words.
They write that the asymmetry between WTA and WTP, far from being a mistake, reflects a genuine effect of reference positions on preferences[1].
And the consequence: "if a good is evaluated as a loss when it is given up and as a gain when it is acquired, loss aversion will, on average, induce a higher dollar value for owners than for potential buyers, reducing the set of" mutually acceptable trades[1].
Two observations, ours.
The phrase far from being a mistake is doing deliberate work. The authors are arguing against treating the gap as measurement error, which was the prevailing view they cite.
And the stated consequence is not that prices are wrong. It is that the set of trades that can happen at all gets smaller. That is the commercially important claim and we return to it.
The Challenge
The 2005 paper that reopened everything.
Plott and Zeiler published The Willingness to Pay-Willingness to Accept Gap, the "Endowment Effect," Subject Misconceptions, and Experimental Procedures for Eliciting Valuations in the American Economic Review in 2005[3].
Their argument, as an encyclopedia summary puts it, is that observed disparities between WTA and WTP measures are not reflective of human preferences[5], but instead are symptomatic of subjects' misconceptions about the nature of the experimental task[2].
The reported result: they report that the willingness-to-pay and willingness-to-accept disparity is absent for mugs in a particular experimental setting, designed to neutralize misconceptions about the elicitation device[2].
In their own later summary: "We find that gaps observed in Kahneman, Knetsch, and Thaler's (KKT) (1990) mug experiments are not due to a kink at the endowment as posited by" endowment effect theory, and "when the experiment is controlled for subject misconceptions about the elicitation device the valuation gap disappears"[3].
The Part That Is Genuinely Clever
How they answered the induced-value control, which is the strongest move in the dispute.
A review records: "PZ replicate the finding of no gap in the induced value rounds in both their version of the Kahneman et al. procedures and in their preferred procedures. However, since they find a WTP-WTA gap for goods with the former procedures but not the latter, Plott and Zeiler argue that the absence of gaps in induced value rounds does not" establish the absence of misconceptions[2].
Working through the logic, and this is ours.
Under KKT's procedures, PZ got no gap for tokens and a gap for goods, reproducing the original.
Under their own procedures, they got no gap for tokens and no gap for goods.
So the token result was identical under both, while the goods result differed. Which means the token control cannot be detecting whatever changed, because it did not move when the thing that mattered moved.
Two consequences.
The induced-value control turns out to be insensitive to the very problem it was supposed to rule out. Passing it does not demonstrate that participants understood the goods task.
And that is a properly good argument. It does not accuse anybody of anything; it shows that a control everyone trusted lacks the diagnostic power it was credited with.
The Third Team
Where it gets genuinely difficult.
Isoni, Loomes and Sugden published a comment in the American Economic Review, 101(2), 991–1011, in 2011[6].
A review summarises: "Isoni et al. have two main findings regarding the WTP-WTA gap for goods: no significant gaps in their version of the KKT and PZ procedures, and no difference in gaps between the two procedures. They conclude that PZ's procedures did not remove misconceptions, but that some other factor drove the absence of the gap."[2]
Three observations, ours, and the first is the one that matters.
They found no significant gap under KKT's own procedures. That is a failure to reproduce the original result, reported by a team that also rejects the leading explanation for why the original result was wrong.
And they found no difference between the two procedures, which directly undercuts PZ's account, because PZ's whole claim is that the procedures matter.
So the third team is not on either side. They are saying the original does not reproduce and the debunking does not explain it either.
The Reply
What Plott and Zeiler said back.
They wrote that Isoni, Loomes, and Sugden (2011) assert that Plott and Zeiler (2005) reported inaccurate results, and that placing ILS's selective quotes into context demonstrates otherwise[3].
Their three conclusions from re-examining the data: "First, all mug data reject endowment effect theory. Second, lottery gaps are associated with unstable attitudes toward uncertainty... Third, lottery outcome beliefs are influenced by whether WTP or WTA is reported, suggesting that changing beliefs, as opposed to the shape of preferences, produce lottery gaps."[3]
Two observations, ours.
The first point is a neat move: they accept ILS's data and argue it supports their conclusion rather than KKT's. If ILS found no gap for mugs, that is consistent with there being no endowment effect to find.
The third point proposes a different mechanism entirely for the lottery results: that asking someone to state a selling price versus a buying price changes what they believe about the odds, rather than changing their preferences. That is a claim about beliefs, not tastes, and it is a distinct third explanation.
A Scoping Concession Worth Noting
A sentence in the reply that a fair reading requires.
Plott and Zeiler state that their work "clearly indicates that we made no claim about all WTP-WTA gaps in all circumstances"[3].
Two observations, ours.
That is a narrower claim than the popular version of their result. The endowment effect is often described as having been debunked, and the debunkers themselves say they made no such general claim.
Their target is specific: endowment effect theory as an explanation of the gaps in the mug experiments. Whether valuation gaps occur elsewhere, for other goods, in other settings, is left open by their own account.
Where That Leaves It
Our own reading of an unresolved dispute.
Four positions are live in the literature we obtained.
The gap is real and reflects loss aversion operating on reference points.
The gap is an artefact of the elicitation procedure and vanishes when misconceptions are controlled.
The gap does not reliably appear under either procedure, and neither existing explanation accounts for the pattern.
And, for lotteries at least, gaps arise from changing beliefs rather than changing preferences.
Our summary: twenty years of high-quality argument between serious economists has not settled what the laboratory result means. Anyone who tells you the endowment effect is established, or that it has been debunked, is describing one round of a fight that continued.
Why It Still Matters Commercially
The section that deliberately does not depend on who wins. Ours.
Notice what the entire dispute is about: whether a laboratory gap survives particular experimental controls. That is a question about mugs, tokens and elicitation devices.
It is not a question about whether owners and buyers name different numbers for a business. That happens, it is observed constantly, and no experiment is needed to establish it.
Three consequences.
The commercial phenomenon does not require the psychological explanation. An owner may value a business above a buyer because of loss aversion, or because they have private information, or because they know what it cost to build, or because their retirement depends on a number.
Which means the practical advice should not rest on the contested part. Anything an adviser tells a client that depends on loss aversion being the cause is resting on a Grade C claim.
And the useful move is to work with the consequence rather than the cause, which the original authors stated plainly and which nobody in the dispute contests.
The Trades That Never Happen
The consequence, and it is the reason this matters more than a pricing error would. This analysis is ours; the framing is the original authors'.
Recall their own words: the asymmetry induces a higher value for owners than for buyers, reducing the set of mutually acceptable trades[1].
Illustrating with our own arithmetic. If a seller asks 100 and a buyer offers 100, a deal exists. At 90, 80, 65 or 50, no deal exists unless somebody moves.
Two observations.
The harm is not that the business sells at the wrong price. It is that the business does not sell. The failure mode is a deal that never closes rather than a deal that closes badly.
And a trade that never happens leaves no record. There is no transaction to review, no price to regret, no entry in any ledger. This is the eighth article in this series to land on the same structure: the costly outcome is the invisible one, so nobody counts it and nobody is accountable for it.
The Finding About Market Experience
The most practically consequential result in this literature, and our sourcing on it is thin.
A reference list in a review paper cites List, J. A. (2003), Does market experience eliminate market anomalies?, Quarterly Journal of Economics[7].
We did not obtain this paper and report only its title and citation. We include it because the title states a question whose answer is directly relevant, and because a reader following this literature will encounter it.
We also note a separate report that findings suggest the endowment effect is less strong when the relatively artificial sense of scarcity induced in experimental settings is lessened, attributed to Shogren and colleagues (1994)[5]. We did not obtain that either.
Our own observation, offered as reasoning rather than as a finding: if experience with trading a class of good does attenuate the effect, then the size of any such effect in a given negotiation depends on how often each party has done this before, which is an observable and highly asymmetric feature of most business sales.
The One-Time Seller
The application, and it is ours. It rests on the market-experience finding we could not obtain, and should be weighed accordingly.
Consider who is in the room when an owner-managed business changes hands.
The seller has sold a business perhaps once, quite possibly never, and will not do it again.
The buyer, if a trade buyer or an investor, may have done it many times.
Three consequences.
Whatever this literature describes, it would apply asymmetrically, and it would apply to the side with more at stake and less practice.
The seller also cannot learn from repetition, because there is no repetition. Every other skill in running a business improves with iteration; selling it is a single-shot problem.
And that is a reasonable argument for hiring the experience rather than trying to supply it, which is what a corporate finance adviser is for. We note the obvious: this publication belongs to a firm that does this work, and the reader should weigh the recommendation accordingly.
What Buyers And Sellers Remember
A small finding with a clean practical use.
An encyclopedia summary reports that buyers tended to recall reasons to buy the mug, whereas sellers tended to recall reasons to keep their mug before reasons to sell it[5].
We did not obtain the underlying study and report only this description.
Two observations, ours.
If it holds, the two sides are not disagreeing about the same evidence. They are each working from a different subset of it, assembled by whichever role they occupy.
And it suggests a cheap discipline for a seller: write down the reasons to sell before the negotiation begins, since the role you are in may make them harder to retrieve once it starts. That is our own suggestion, not a tested intervention.
Other Explanations On Offer
For completeness, the alternatives this literature contains, cited but not obtained.
Reference lists in the papers we reviewed identify Morewedge and Giblin (2015), Explanations of the endowment effect: an integrative review, in Trends in Cognitive Sciences; Bordalo, Gennaioli and Shleifer (2012), Salience in experimental tests of the endowment effect, in the American Economic Review; Horowitz and McConnell (2002), A review of WTA/WTP studies, and Tunçel and Hammitt (2014), A new meta-analysis on the WTP/WTA disparity, both in the Journal of Environmental Economics and Management[7][8].
We obtained none of these and report no findings from any of them.
Two observations, ours.
The existence of two separate meta-analyses of the WTA to WTP disparity indicates a literature large enough to be worth synthesising twice, which is itself information about how unsettled it is.
And an integrative review of explanations, in the plural, tells you the field has more than one candidate mechanism live. A reader wanting a settled answer will not find one here, and we would rather say that than manufacture one.
What To Do
Do not tell a client the endowment effect explains their valuation. The causal claim is contested by two separate research teams from two different directions.
Do not accept that it has been debunked either. The debunkers state they made no claim about all such gaps in all circumstances.
Work with the consequence, not the cause. The gap reduces the set of trades that can happen at all, and that is agreed by everyone regardless of mechanism.
Count the deals that do not close. The failure mode is a sale that never happens, which leaves no record and appears in no review.
Notice who has done this before. On a finding we could not obtain but which a reader will meet, market experience may attenuate the effect, and in a business sale the experience sits almost entirely on the buyer's side.
Write the reasons to sell down in advance. On one reported finding, sellers recall reasons to keep before reasons to sell, and the fix costs nothing.
Get an external valuation before you form a number. An anchoring point set by your own expectations is the subject of the fifth article in this series, and it is easier to obtain the outside figure first than to revise afterwards.
Separate the number from the reason for it. A seller whose price is driven by what their retirement requires has a real constraint, and it should be named as that rather than defended as a valuation.
The Limits Of This Analysis
Several caveats matter. This article reviews behavioural economics and is not valuation, investment or legal advice; nothing in it is a method for valuing any business. Everything is verified to August 2026. We obtained portions of the 1990 paper only, and did not obtain the 2005 paper, the 2011 comment, or the reply in full, relying on a working paper reviewing the literature, publisher records and quoted passages. We state no WTA to WTP ratios and no trading percentages anywhere in this article, because our sources report them in fragments and inconsistently. We did not obtain the paper on market experience, the study on artificial scarcity, the memory study, the integrative review, the salience paper, or either meta-analysis, and report no findings from any of them; several reach us only as citations in reference lists. Several claims reach us through an encyclopedia article hosted as a PDF, which we identify at each use. All arithmetic is ours and illustrative. The one-time-seller argument rests on a market-experience finding we could not obtain and should be weighed accordingly, and we note that this publication belongs to a firm that provides corporate finance advice. The analysis of why the induced-value control lacks diagnostic power, the observation about invisible failed trades, and the practical section are our own reasoning, not findings. This dispute is live and unresolved, and later rounds may exist that we did not locate.
Frequently Asked Questions
Is the endowment effect real?
So has it been debunked?
What was wrong with the original control?
Does any of this matter if I am selling a business?
What is the actual risk?
Is there anything cheap I can do?
References
- Kahneman, D., Knetsch, J. L., & Thaler, R. H. (1990). Experimental Tests of the Endowment Effect and the Coase Theorem. Journal of Political Economy, 98(6), 1325–1348, university-hosted copy, on the asymmetry between willingness to accept and willingness to pay, far from being a mistake, reflecting a genuine effect of reference positions on preferences; on Thaler (1980) having labelled the increased value of a good to an individual when it becomes part of that individual's endowment the endowment effect; on this effect being a manifestation of loss aversion, the generalization that losses are weighted substantially more than objectively commensurate gains; and on the implication that if a good is evaluated as a loss when given up and as a gain when acquired, loss aversion will on average induce a higher dollar value for owners than for potential buyers, reducing the set of mutually acceptable trades. Note: a university-hosted PDF of the original paper; we obtained portions, principally the introduction. web.mit.edu
- Ericson, K. M. M., working paper reviewing the endowment effect literature, on the exchange paradigm and Knetsch's results strongly rejecting the standard prediction; on the valuation paradigm being often referred to by its result, the WTP-WTA gap, with non-reference dependent theory predicting WTA and WTP should be approximately equal apart from income effects while a substantial literature had noted they were typically far apart; on Plott and Zeiler arguing that disparities are symptomatic of subjects' misconceptions about the nature of the experimental task; on Plott and Zeiler reporting the disparity absent for mugs in a setting designed to neutralize misconceptions; on a major innovation of Kahneman, Knetsch and Thaler having been to show the gap appeared for goods but not for induced value rounds; on Plott and Zeiler replicating no gap in induced value rounds under both their version of the Kahneman procedures and their preferred procedures while finding a goods gap under the former but not the latter, and arguing that the absence of gaps in induced value rounds therefore does not establish the absence of misconceptions; and on Isoni and colleagues finding no significant gaps in their version of either procedure and no difference between them, concluding that Plott and Zeiler's procedures did not remove misconceptions but that some other factor drove the absence of the gap. Note: a working paper reviewing the literature; our principal source for the 2005 and 2011 papers, neither of which we obtained. scispace.com
- Plott, C. R., & Zeiler, K., publisher records for The Willingness to Pay-Willingness to Accept Gap, the "Endowment Effect," Subject Misconceptions, and Experimental Procedures for Eliciting Valuations, American Economic Review (2005), and for their reply to Isoni, Loomes and Sugden, on the authors making no claim about all WTP-WTA gaps in all circumstances; on their finding that gaps observed in Kahneman, Knetsch and Thaler's 1990 mug experiments are not due to a kink at the endowment as posited by endowment effect theory; on the valuation gap disappearing when the experiment is controlled for subject misconceptions about the elicitation device; on Isoni, Loomes and Sugden asserting that Plott and Zeiler reported inaccurate results, which the authors dispute; and on their three conclusions that all mug data reject endowment effect theory, that lottery gaps are associated with unstable attitudes toward uncertainty, and that lottery outcome beliefs are influenced by whether WTP or WTA is reported, suggesting changing beliefs rather than the shape of preferences produce lottery gaps. Note: publisher records reproducing abstracts and quoted passages; we did not obtain either paper in full. researchgate.net
- Semantic Scholar record for Kahneman, Knetsch and Thaler (1990), confirming the citation and carrying indexed descriptions of subsequent work, including that the discrepancy between willingness to accept and willingness to pay is supposed to be a manifestation of the endowment effect, and that Plott and Zeiler (2005) report the disparity absent for mugs in a particular experimental setting designed to neutralize misconceptions. Note: a bibliographic record with third-party indexed summaries. semanticscholar.org
- Encyclopedia article on the endowment effect, hosted as a PDF, on buyers tending to recall reasons to buy the mug whereas sellers tended to recall reasons to keep their mug before reasons to sell it; on findings suggesting the endowment effect is less strong when the relatively artificial sense of scarcity induced in experimental settings is lessened, attributed to Shogren and colleagues (1994); and on Plott and Zeiler having challenged endowment effect theory by arguing that observed disparities between WTA and WTP measures are not reflective of human preferences. Note: an encyclopedia article, not peer-reviewed. We did not obtain any of the underlying studies it describes. cognitionandculture.net
- Reference list in a review chapter, identifying Isoni, A., Loomes, G., & Sugden, R. (2011), The Willingness to Pay-Willingness to Accept Gap, the "Endowment Effect," Subject Misconceptions, and Experimental Procedures for Eliciting Valuations: Comment, American Economic Review, 101(2), 991–1011; and recording the reviewing author's own observation that the subsequent literature has produced evidence both in concert with and refuting the original conclusions, and that even imperfect papers can have deep impact. Note: a reference list and commentary in a working paper; we did not obtain the comment itself. s3.amazonaws.com
- Reference list in the review chapter at reference 6, identifying List, J. A. (2003), Does market experience eliminate market anomalies?, Quarterly Journal of Economics; and Tunçel, T., & Hammitt, J. K. (2014), A new meta-analysis on the WTP/WTA disparity, Journal of Environmental Economics and Management, 68(1), 175–187. Note: citations only. We did not obtain either paper and report no findings from either; the List paper is included because its title states a directly relevant question a reader will encounter. s3.amazonaws.com
- Reference list in a working paper on eliciting the endowment effect, identifying Kahneman, D., Knetsch, J. L., & Thaler, R. H. (1991), Anomalies: The endowment effect, loss aversion, and status quo bias, Journal of Economic Perspectives, 5(1), 193–206; Morewedge, C. K., & Giblin, C. E. (2015), Explanations of the endowment effect: an integrative review, Trends in Cognitive Sciences, 19(6), 339–348; Horowitz, J. K., & McConnell, K. E. (2002), A review of WTA/WTP studies, Journal of Environmental Economics and Management, 44(3), 426–447; and Bordalo, P., Gennaioli, N., & Shleifer, A. (2012), Salience in experimental tests of the endowment effect, American Economic Review, 102(3), 47–52. Note: citations only. We obtained none of these papers and report no findings from any of them. arxiv.org
This article reviews behavioural economics and is not valuation, investment or legal advice. Nothing in it is a method for valuing any business. Only the 1990 paper was obtained, and only in part; the 2005 paper, the 2011 comment and the reply were not obtained in full. No willingness-to-accept to willingness-to-pay ratios or trading percentages are stated anywhere, because sources report them inconsistently. Several cited works were not obtained at all and are identified as citations only. All arithmetic is the authors' own. This publication belongs to a firm that provides corporate finance advice, and the recommendation to obtain external assistance should be weighed accordingly.