Value Coherent Arbitrariness

In one of the most famous passages of Mark Twain’s novel The Adventures of Tom Sawyer, Tom is made to paint the fence as punishment for playing hooky from school on Friday and dirtying his clothes in a fight. At first, Tom is disheartened by having to forfeit his day off. However, by applying himself to the paintbrush with gusto, he soon cleverly persuades his friends to trade small treasures wuth him for the privilege of doing his work. He does that by making them believe that painting his aunt’s fence is not a punishment but rather a pleasure and a priviledge they should be willing to pay for. Accordingly to Ariely et al. (2006, p.1), and in Twain’s words, Tom “had discovered a great law of human action, without knowing it – namely, that in order to make a man or a boy covet a thing, it is only necessary to make that thing difficult to attain.”

I was responsible myself for a kind of Tom’s Law phenomena during my childhood. After my mum gave me and my brother a package of wafers, mine would often become too soft due to too much air exposure. Because I enjoyed them much better when they were crispy, I would convince my brother that they tasted “so much better” when they were soft. As a result, my brother would conceive to change some of his crispy wafers for my softer ones. These evidences, although anecdotal, contradict the assumption, adopted by orthodox economics, that preferences are stable and fundamental values, and pose a fundamental challenge to all the economic science enterprise. After all, if people don’t reliably know what they like, it cannot be assumed that voluntary trades will improve well-being or that markets will increase welfare.

Experimental Evidence.

If preferences are not stable, why would we observe stable demand curves in the market place? After all, if consumers' valuations of goods are so malleable, demand curves should reflect that arbitrariness. Several experiments showed that although the valuations of goods and experiences have a large arbitrary component, there is consistency across valuations - after one valuation is made, people provide subsequent valuations that are consistent with that first valuation. By using the last four digits of the social security card in several experiments, it was shown that we can easily increase or decrease the price people are willing to pay for an average bottle of wine – making that valuation significantly arbitrary. Nevertheless, the second offer made by the subjects for a fancier bottle of wine will always be higher than the first one, also making that valuation significantly consistent. These experiments allowed us to understand that although absolute values are surprisingly malleable, relative values are not.

Despite the impact of the arbitrary social security number used to monitor the subjects' willingness to pay higher or lower prices, it could be argued that the effect was due to the subjects' uncertainty about what the goods were worth. Nevertheless, the same impact was found with goods that couldn’t be traded but that could be experienced, and hence fully understood, before the valuation task. Using a poem as the good, Ariely et al. (2006, p.3) found that “people don’t have a pre-existing sense of whether an experience is good or bad,” and that the exact same experience – in this case, a poetry reading – can be desired or avoided, depending on context and presentation. Thus, after being asked if they were willing to pay $2 to listen to their professor recite poetry, a significant higher percentage of students were willing to attend the recital for free, compared with another group of students that were asked if they would be willing to accept $2 to listen to their professor recite poetry. Accordingly, with a simple initial question, researchers were able to manipulate the subjects' perception of the poetry reading experience, making it sound as positive or negative due to the way the initial question was framed. Subsequently, the authors also found that subjects, after perceiving the poetry reading as a positive or negative experience, were consistent in terms of the valuations they gave to different amounts (length of time) of experience, increasing the amount of money they would be willing to pay (positive frame) or to be paid (negative frame) to listen to increasing amounts of time of poetry reading. Most impressively, the same effect was still present even when subjects listen to a one minute sample of the poetry reading before providing any response, and even if it was made explicit that there were two conditions (willing to pay vs. willing to accept), and that the assignment to one or the other was random.

Economic Implications.

The idea that preferences are stable derives from the belief that choices correspond to fundamental values – personal likes and dislikes. In contrast, the subjective and social dimensions of value elicited by Ariely et al. (2003, 2006) studies, suggest that responding to incentives may not be the result of fundamental valuation, but the fact that people try to behave in a sensible manner when it is obvious how to do so. Thus, if previous choices are recalled, decision makers try to behave in ways that don´t violate the explicit rules of consistency.

Accordingly, investors may be willing to pay more for the same stocks today than what they had paid for yesterday because of the announcement of an unexpectedly profitable quarter. But that doesn´t mean that yesterday´s valuation was reasonable. Although responsiveness to incentives is a necessary condition to fundamental valuation, people can show responsiveness even when they do so from arbitrary baseline levels.

Arbitrariness is also expected to be enhanced by ambiguity in a good or experience. Some experiences (e.g., living close to a dump) are unambiguously bad, but the truth is that most of the decisions people make (marriage, education, emigration, jobs, vacations) involve streams of heterogeneous experiences that are arguably even more difficult to assess and hence even more vulnerable to arbitrary influences than a simple poetry reading proposition.

Consistency, in turn, will be influenced by how easy it is to spot behavioral inconsistencies, which depends on whether scales and numerical indices are available, the time interval between choices and the apparent connections between the choices. These conditions apply to several types of decisions, of which financial decisions are a good example. If we look at short term fluctuations in prices, we do indeed see stock prices responding appropriately to good or bad news about individual companies or the economy as a whole. Nevertheless, as previous research illustrated, in the long term markets behave in ways that are completely out of line with historical fluctuations and dividend streams, making it impossible to clearly know if the market is over- or underpriced.

The implications of these findings for economic policies are substantial on two levels. The first level is related to how the `the general equilibrium´ of an economy comes into existence. According with orthodox economics, market and production prices achieve equilibrium through the interaction of exogenous consumer preferences with technologies and initial capital. This analysis becomes very unlikely if we consider that preferences are themselves endogenous to the economy, being influenced by the equilibrium states they presume to create. This way, individual preferences are no longer seen as the determinants of the economy; they are considered to be determined by the economy itself. Other authors (e.g., Robert Frank) already explored similar ideas, arguing that the utility we obtain from our consumption pattern is highly relative on the others (the economy) consumption patterns (e.g., I may prefer to go by bicycle to the work, but I can´t because there is only a road for cars between my house and the university). The second level of implications is related with the prescriptive dimension of economics. Economists derive ´welfare implications´ from alternative policies such as taxation or trade, and welfare is defined as the degree to which a policy leads to the satisfaction of individual preferences. Although economists had identified situations in which free market exchange may not increase welfare, such market failures are related to interactions between people with asymmetric information or from externalities, in which people don´t internalize the costs they impose on each other. The suboptimalities related with the coherent arbitrariness phenomena are, in contrast, related with the individual level. If preferences have a large arbitrary component, even strictly personal consumption choices by fully informed individuals may not maximize welfare. Additionally, these individual effects may be exacerbated by social and market interaction (e.g., choosing a restaurant because it has a lot of clients). These effects can be increased by the degree of consumers' uncertainty about their own preferences. As recent research sas shown such degree can be substantial and poses a fundamental challenge to the those who proclaim markets to be the absolute instruments of welfare.


Ariely, D., Loewenstein, G.F. & Prelec, D. (2003). Coherent Arbitrariness: Stable Demand Curves Without Stable Preferences. Quarterly Journal of Economics. 118, 73-105.

Ariely, D., Loewenstein, G.F. & Prelec, D. (2006). Tow Sawyer and the Construction of Value. Journal of Economic Behavior & Organization. 60, 1-10.

Recent Posts in There Are Free Lunches

Neoliberalism Viewed From the Couch

Insights about economics gathered from a psychoanalyst’s couch

Behavioral Science and Development Economics

From an "Idea Too Soon for It’s Time" to the "Defining Idea of Our Time"

Why is Behavioral Economics so Revolutionary?

And what can it mean for our future

An Indian Economist in Bhutan: The Kingdom of Happiness

An interview with Vatsalya Srivastava about his perception of Bhutan.

Happiness: The "Science" Behind the Free Lunch

Building a well-being economy

5 Bullets in the Heart of Demand Analysis

Preferences influence prices, but are also influenced by them.