IKEA effect
People value what they have built themselves more highly than an identical thing built by someone else, but only when they finish it.
People value what they have built themselves more highly than an identical thing built by someone else, but only when they finish it.
Norton, Mochon and Ariely gave one group of people a flat-packed object and another group the same object already assembled, then asked both what they would pay for it. The builders bid more. The premium is not about ownership, since neither group owned the object while bidding, and not about customisation, since the objects were standardised and identical. It attaches to the act of building. The same paper sets the boundary. Labour raises valuation when the labour succeeds, and stops raising it when the task is abandoned or undone.
Replicated An independent meta-analysis of 55 studies and 5,454 people puts the effect at d = 0.57, and every replication located is conceptual rather than a direct repeat of the original experiments.
Pelled, Demetriades and Walter synthesised 55 studies covering 5,454 participants and report a moderate effect of self-assembly on valuation, d = 0.57, p < 0.005, with matching effects on liking, self-concept and sense of ownership. Their abstract states that the result held across moderators including product customisation and tangibility. The full text is closed at Wiley, which refused the request, and Unpaywall and OpenAlex list no open copy, so those figures are read off the abstract deposited at Crossref, which prints no interval around the pooled estimate.
Sarstedt, Neubert and Barth published a conceptual replication in 2017 whose abstract reports support for the original effect and identifies psychological ownership as the mechanism; that paper is closed at the publisher and Unpaywall lists no open copy, so its sample size and statistics are not quoted. The developmental work is open, and the 2022 study preregistered its design, hypotheses and analysis plan.
Marsh, Gil and Kanngiesser tested 128 British and Indian children aged five and six, who raised the value of a monster toy they had built by 1.06 coins (95% CI 0.60 to 1.52) and lowered the value of an identical one they had not built by 0.27 coins (95% CI -0.77 to 0.23), F(1, 119) = 28.23, p < .001, partial eta squared = .19.
Collaborative building produced the same premium as building alone, and the British and Indian samples did not differ. One conceptual replication reports a null. Raghoebar, van Kleef and de Vet had 86 children aged four to six craft a peacock from vegetables or from non-food objects, and found no effect of self-crafting on how much of the vegetables the children ate or how much they liked them.
No direct repeat of the box, origami or Lego experiments was located. The copy of the FORRT Replication Database downloaded on 17 August 2026 holds no row naming the IKEA effect or Norton and Mochon, and searches of Crossref, PubMed, Unpaywall and the open web turned up only the conceptual work described here. Whether the meta-analysis pooled a direct replication could not be checked, because its list of included studies sits behind the paywall.
The bookcase assembled on a Sundayillustration
Someone who spent an afternoon putting up a flat-pack bookcase would not take for it what they would pay for the same model delivered built. The premium attaches to the assembly, and nothing about the wood has changed.
The shed left at the frameillustration
A shed abandoned once the frame was up carries no premium at all, and its owner talks it down. Labour raises valuation when the labour finishes the job, and stops raising it when the job is left undone.
The deck built overnightillustration
A manager rates the slides he made himself above the identical set an agency supplied, and argues longer for their structure. Neither set is his property and neither is customised, so what separates them is who did the work.
Cases marked as illustrations describe nobody in particular. They are there to make the pattern recognisable, not to report that a particular thing happened.
Four experiments, each with real money at stake through the Becker, DeGroot and Marschak bidding procedure or a sealed auction. In Experiment 1A, 52 students either assembled a plain black IKEA storage box or inspected one already built. Builders bid $0.78 (SD 0.63) against $0.48 (SD 0.40) from non-builders, t(50) = 2.12, p < .05, a premium of 63%.
Experiment 1B moved to origami and 106 participants. Builders bid $0.23 (SD 0.25) for the frog or crane they had folded, while other participants bid $0.05 (SD 0.07) for those same creations and $0.27 (SD 0.26) for origami folded by experts, F(2, 100) = 5.34, p < .01. Builders' bids for their own work and outsiders' bids for expert work did not differ, t < 1, p > .45.
Experiment 2 gave 118 people Lego sets of 10 to 12 pieces and had each bid on their own set and a partner's. Bids ran $0.84 against $0.42 in the build condition, $0.32 against $0.26 when the set came prebuilt, and $0.43 against $0.29 for people who built the set and then took it apart. Only the build condition produced a reliable gap, t(39) = 3.08, p < .01.
Experiment 3 gave 39 people an IKEA box and asked some of them to stop before the last two steps. Those who finished bid $1.46 (SD 1.46) and those who did not bid $0.59 (SD 0.70), t(37) = 2.35, p < .05.
3 primary research4 replication and reanalysis
The paper is titled "The IKEA effect: When labor leads to love", and the name honours the Swedish manufacturer whose products arrive with some assembly required. It appeared online on 9 September 2011 and in print in 2012, at 22(3), 453-460. The underlying idea is older, and the authors take care to separate it from effort justification: in Aronson and Mills's 1959 initiation experiment every participant was admitted to the group, so the success of the effort was never varied, which is the variable this paper claims matters.
The claim, and the replication this entry reads its figures from.
Every figure in the original paper is a bid of under two dollars on a storage box, a piece of folded paper or a dozen Lego bricks. Norton and colleagues say plainly that they used small-ticket items and that whether the effect holds for expensive ones is an open question. The completion boundary rests on two experiments with 118 and 39 participants, and the searches recorded above found no repeat of either.
Marsh, Kanngiesser and Hood report that among five- and six-year-olds the bias was not moderated by how much effort the task demanded, so effort and completion are not the same lever. The mechanism is unsettled. Mochon, Norton and Ariely argue that building signals competence to the self and to others, Sarstedt and colleagues point to psychological ownership, and Marsh and colleagues conclude that ownership alone cannot explain the children's results and propose a link to self-concept instead. The meta-analysis was read only in abstract, so whether it tested completion as a moderator could not be checked.
Each of these is named in this entry's own prose as something it gets taken for. Where the two came out differently, that is worth knowing before you quote either.
The IKEA effect is not the endowment effect, which has its own entry here. Endowment concerns things a person already owns. In Experiment 1A both groups were bidding on a box that was not yet theirs, and in Experiment 2 the people who built a Lego set and then dismantled it had spent longer handling it than the builders and bid less.
Norton and colleagues read that as evidence against ownership and against mere touch. The second misreading is that effort alone buys affection. Half-finished boxes and dismantled Lego produced no premium. The third is scale. The paper's own phrase, that builders valued their origami nearly five times higher, describes bids of 23 cents against 5 cents, and the highest mean bid in the four experiments is $1.46.
Every claim on this page was held against these sources on 2026-08-17. Nothing here is written from memory. Written and checked by Krishna Chagti; how entries are written and corrected. Found a mistake? Report it, with the sentence and a source.