Constructed preferences
Also known as: Construction of preferences, Preference construction
We often don't have settled preferences waiting to be reported—we build them on the spot.
What it means
The constructed-preferences view holds that for many decisions people do not retrieve pre-formed, stable values from memory but instead assemble their preferences in the moment of choice, using whatever cues, comparisons, and strategies the task and context make salient. This directly challenges the standard economic assumption of fixed, coherent, well-ordered preferences that elicitation merely reveals. The strongest evidence is that logically irrelevant features of how a choice is posed—the response mode, the framing, the set of alternatives, the order, even the units—reliably reverse what people 'prefer,' as in preference reversals between choosing and pricing the same gambles. Because construction is shaped by the immediate environment, preferences are malleable and manipulable, which is precisely what makes choice architecture powerful and what makes inferring 'true' welfare from observed choice problematic. The view does not claim people have no values, only that expressed preferences are often jointly produced by the person and the situation. It matters for measurement, for nudging, and for the foundations of welfare economics.
Why it happens
People cannot hold a complete, consistent ranking of options in working memory, so for anything past routine choices they assemble a judgment from whatever the task makes salient. Two mechanisms do much of the work. Scale compatibility means an attribute expressed in the same units as the required response gets extra weight: asked to price a gamble in dollars, the payoff looms large; asked simply to choose, the probability of winning matters more. The prominence effect means the dimension that feels more important dominates choice more than it dominates pricing or matching. Because different elicitation methods recruit different weights, one person can rank a set of options one way and value them another, with nothing about the options themselves having changed.
What the evidence shows
The founding demonstration is the preference reversal: people choose the high-probability bet yet set a higher selling price on the high-payoff bet, contradicting every standard model of risk (Lichtenstein & Slovic, 1971). Economists suspected an artifact of hypothetical questions or weak incentives, so Grether and Plott (1979) reran it with real money and market discipline, expecting the effect to disappear; it did not. Ariely, Loewenstein and Prelec (2003) sharpened the point: an arbitrary number, such as digits from a social security number, shifted absolute willingness to pay for ordinary goods, while relative valuations stayed orderly, coherent arbitrariness that imitates stable preference while resting on an accident. A later replication (Maniadis, Tufano and List, 2014) recovered the anchoring effect in the same direction but at roughly a half to a third of its original size, so the phenomenon holds while being more modest than the first estimate. Reversals replicate across payment schemes and domains, though economists still debate how far incentives and error models shrink them.
Where it shows up
Construction bites hardest where people have no rehearsed preference to retrieve: novel, complex, or emotionally charged goods with no familiar price to anchor on. Contingent-valuation surveys that ask what a wetland or an endangered species is worth are the canonical worry, because the figures move with question order, scope, and framing rather than tracking a settled value. The same fragility shows up in choices between unfamiliar medical treatments, in juror damage awards, and in consumer pricing, where a decoy option or a reframed discount reshuffles what buyers say they want. Joint-versus-separate evaluation is a practical lever: options judged side by side draw on different attributes than the same options judged alone, so how you stage the comparison helps decide the winner.
Why it matters for welfare
If the elicitation helps produce the preference, then revealed preference, reading what is good for someone off what they chose, loses its clean footing. A demand curve can look orderly and still not encode stable underlying wants (Ariely et al., 2003), so market behavior does not by itself certify welfare. This cuts two ways. It is the engine under choice architecture: because defaults, orderings, and framings enter the preference a person ends up with, small design decisions carry real weight, the promise and the ethical hazard of nudging alike. And it is a warning for measurement. Satisfaction scores, willingness-to-pay, and survey preferences are best read as products of a procedure rather than readouts of fixed values, especially for goods people rarely choose.
Examples
Asked which of two gambles they prefer, people often pick the safer one; asked the most they would pay for each, they price the riskier one higher—preferring A yet valuing B more.
Shown three wines, most diners take the middle one. Add a pricier fourth and the middle moves up, so the same drinker now 'prefers' a bottle they passed over minutes earlier.
Asked to rate two job candidates one at a time, a panel favours the warm generalist; asked to choose between them side by side, the same panel picks the technical specialist.
Asked what they would pay to save a few thousand migrating birds, then ten times as many, then a hundred times as many, people name almost the same sum, an environmental value assembled from the question rather than retrieved from memory.
Told a surgery has a 90 percent survival rate, most patients accept; told the same operation carries a 10 percent mortality rate, many refuse. The preference forms around the frame, not the outcome.
First described in Slovic; Lichtenstein & Slovic; Payne, Bettman & Johnson.
Key references
- Ariely, D., Loewenstein, G., & Prelec, D. (2003). "Coherent arbitrariness": Stable demand curves without stable preferences. Quarterly Journal of Economics, 118(1), 73-106. academic.oup.com/qje/article/118/1/73/1917051
- Maniadis, Z., Tufano, F., & List, J. A. (2014). One swallow doesn't make a summer: New evidence on anchoring effects. American Economic Review, 104(1), 277-290. www.aeaweb.org/articles?id=10.1257/aer.104.1.277
- Bettman, J. R., Luce, M. F., & Payne, J. W. (1998). Constructive consumer choice processes. Journal of Consumer Research, 25(3), 187-217. academic.oup.com/jcr/article-abstract/25/3/187/1795625
- Slovic, P. (1995). The construction of preference. American Psychologist, 50(5), 364-371. scholarsbank.uoregon.edu/items/8bfbe1ef-a008-470a-a730-625bfc00c192
- Tversky, A., Sattath, S., & Slovic, P. (1988). Contingent weighting in judgment and choice. Psychological Review, 95(3), 371-384. philpapers.org/rec/TVECWI
- Grether, D. M., & Plott, C. R. (1979). Economic theory of choice and the preference reversal phenomenon. American Economic Review, 69(4), 623-638. ideas.repec.org/a/aea/aecrev/v69y1979i4p623-38.html
- Lichtenstein, S., & Slovic, P. (1971). Reversals of preference between bids and choices in gambling decisions. Journal of Experimental Psychology, 89(1), 46-55. scholarsbank.uoregon.edu/items/0412bdbb-67c9-4646-bd57-d039326c1e98