Decision Utility

When you’re making a decision – whether it’s a life-changing choice or what you want to eat for lunch – you’re usually guided by the question: What would be the best for me?

In other words, you’re considering what the most satisfying or useful decision would be. Every time we make a decision, we evaluate the potential outcomes and how useful they might be. This concept of “utility”—how beneficial an outcome will be for us—lies at the heart of the behavioural economic and psychological study.

Utility is a key term in economics that describes the benefit an agent receives from the consumption of goods or services. In traditional economics, people are generally expected to act rationally and make decisions based on maximizing an outcome’s utility. In theory, this process makes sense. In practice, utility is often difficult to quantify in real life.

In order to further refine the concept of utility, psychologists and economists have differentiated between two types of utility: our perceptions of utility before we experience it, or decision utility, and the actual experienced utility of a choice, called experienced utility.¹ Decision utility describes the usefulness that we perceive and use to make a decision, while experienced utility describes the lived consequences of the decision in reality. These different types of utility have driven new understandings of utility and its role in decision-making.

Utility is an important concept in economics, psychology, business, and our personal lives – it guides our every choice. If we can understand utility, we can understand why and how people come to their decisions, and even make predictions about how people will behave.

The beginnings

Utility as an economic principle goes back multiple centuries and was first described by 18th-century Swiss mathematician Daniel Bernoulli. Over time, however, economists and eventually psychologists developed more nuanced theories of utility, leading to the multiple conceptions of utility used today. Moreover, over the past century, a new branch in economics has emerged that developed a different understanding of our relationship to utility.

George Stigler, an American economist, and future Nobel laureate wrote a 1950 paper giving a historical survey of utility in economics.² His review of utility theory from 1776 to 1915 in this article served as a basis for many other researchers to build on. He begins with a theory developed by English philosopher Jeremy Bentham. In an influential 1789 paper, Bentham proposed measuring the amount of pleasure and pain in the context of developing a rationalist legal system. He gave four dimensions of these two feelings: intensity, duration, certainty, and propinquity. Bentham also realized that individual differences would change how a given person feels pleasure or pain in a specific situation. In this way, he described our process of evaluating utility as one of optimizing pleasure and minimizing pain. Although Bentham’s theory was justified by its convenience and approximating ability, it was not necessarily effective, since the philosopher did not provide a way to measure the pleasure and pain of a situation.

Utility theory did not become heavily discussed or studied in economics until the 1870s, when economists attempted to advance the idea of utility in different ways, such as studying the relationships between price and utility, and demand and utility. While various mathematical formulations and models were tested to estimate the utility of outcomes using variables like price, the quantity of product, and supply and demand, measurement remained a difficult goal of utility theory.

Expected utility theory

Then, in 1944, John Von Nuemann and Oskar Morgenstern developed the expected utility hypothesis, based on Daniel Bernoulli’s first description of how we make decisions by estimating the probability and utility of an outcome. By multiplying the probability of an outcome by the expected benefit of that outcome, we get the expected utility of that choice. We can then use this to form our decision—we choose what will give us the best-expected utility. By using Bayesian statistics and probability, the theory suggested that we make precise calculations about the optimal outcome and decision even when the outcome is uncertain. While this theory became massively influential, it worked best in scenarios where the expected gains and probabilities are easily calculated. For instance, this framework can be applied to games like poker, but not easily to most life decisions, where we have trouble estimating the outcomes and the likelihood that we’ll get a particular outcome.³

Behavioral economics

In 1969, by the time the expected utility hypothesis was well known among economists, two economists undertook further research into applying the theory to real life situations. Intrigued by a psychologist’s observation that people followed this logical principle in their decisions by estimating basic probability, Daniel Kahneman and Amos Tversky performed studies to test how people actually behaved in comparison to the predictions made by decision analysts based on the expected utility hypothesis. They found that people often did not follow the statistical predictions decision analysts used, instead opting for a more intuitive approach.⁴ The concept of utility in the expected utility hypothesis was then either flawed or failed to take into account certain kinds of utility.

Kahneman and Tversky continued studying utility together, and into the later 20th century, economists distinguished between two different kinds of utility: decision utility and experienced utility. Experienced utility was connected back to the utility consisting of pleasure and pain that Bentham described, and characterized as a hedonic quality, or relating to the pursuit of pleasure. Decision utility, on the other hand, was conceived as the “weight of an outcome in a decision”⁵, or the value we optimize in a decision.

In modern economics, experienced utility was largely ignored because of arguments that it could not be observed or measured, and that choices reveal the utility of the outcomes because rational agents optimize their utility. In their paper, Kahneman, Peter Wakker, and Rakesh Sarin argued that experienced utility could, in fact, be measured and was distinct from decision utility. They suggested that normal human cognition could result in our perceived utility being different from our experienced satisfaction from an outcome. They proposed a utility framework consisting of 4 different types of utility: predicted utility, decision utility, experienced utility and remembered utility. Decision utility is the utility present at the time of the decision, meaning it drives our decision-making.⁵ As a result of these different utilities, we may not always act in a way that actually maximizes the expected utility of our decision —even if we think we are behaving logically at the time—although this is what traditional economics alleges. As a result, behavioural economics developed as a separate branch of economics that accounts for psychological aspects of decision-making that may cause us to act irrationally, or away from the maximum utility.

Biological decision utility

Recently, the biological basis of decision utility has also been studied in neuroscience and connected to dopamine mechanisms in the brain. The importance of dopamine in motivation provides a biological basis for Bentham’s theory of “hedonic qualities” driving our decisions. Particular cues based on memory can also alter the utility of a particular action immediately after we encounter them, thanks to the release of dopamine.⁶ For example, when you get stressed, you might feel an overwhelming desire to smoke a cigarette. In other calm situations, however, we would feel no urge to do so. Our different reactions to these situations demonstrate how utility can change due to different levels of brain chemicals. The way our brains remember pleasurable experiences may drive us towards a desire, even if the experience fails to meet the remembered feeling.⁷

From a rather simplistic view in the late 18th century to a nuanced and humanistic perspective at the end of the 20th century, our understanding of utility has evolved dramatically. Today, our knowledge of utility as a complex, emotional, and changing quality can help us recognize short-sighted decisions and improve our choices.

At the center of all decisions, utility is a core concept that we use every day, whether or not we’re conscious of it. Although it seems logical to assume we automatically resort to maximizing the utility of our actions, sometimes this does not appear to occur. As we experience the consequences of our decisions, big or small, it’s common to look back on ourselves and wonder “What were we thinking?”. It can seem like a different person was responsible for a poor decision, not us.

The distinction between decision utility and experienced utility can often be significant, so understanding the difference is critical to improving our decision-making. For instance, we tend to make faulty judgments of life decisions and their effect on overall satisfaction. One study about the perceived versus lived satisfaction of living in California demonstrated that we often fall prey to a “the grass is greener” mentality.⁸ The authors concluded that when thinking about differences in climate and culture, we overestimate the effect they will have on our satisfaction. In reality, these factors do not significantly impact our enjoyment of where we live—the study showed that we often believe that living in the sunny California weather will make us happier than it actually does.

The difference in decision utility and experienced utility can also be explained by something called a projection bias.⁹ We overestimate how much our future preferences will look like our current preferences. Understanding this tendency, we can recognize it when we make a poor prediction of what we need and account for the bias.

Additionally, the way decisions are framed can affect our perception of their utility. Kahneman and Tversky first demonstrated the influence of framing in a 1986 paper.¹⁰ We can know that the outcome will be the same—like if different discounts result in the same price reduction—yet, we will still be more attracted to the higher discount percentage. The psychological aspect of buying something on sale, even if it is the same price as another product of equal quality that is not on sale, is another utility that traditional economical utility does not consider. As core parts of human decision-making, we have to take our cognitive biases into account if we are going to understand how we form ideas of utility and apply it to evaluate outcomes.

A problem often brought up with traditional economics and expected utility theory is their assumption of rationality. The term “Homo Economicus” describes the agent implied by traditional economics. While real humans – Homo sapiens – are significantly affected by cognitive biases and emotions, Homo economicus is rational and economically-driven. Homo economicus may evaluate utility in a narrow way which disregards the social or emotional utility involved in a decision, for instance.

At the turn of this century, Richard Thaler, an economist inspired by Kahneman and Tversky’s work, wrote a perspective piece on this issue, predicting that economics would pivot to incorporating human behaviour.¹¹ In the latter half of the 20th century, there had been a shift towards accounting for irrational human behaviour, and Thaler was correct in predicting this future in the development of behavioural economics. That said, economists have been careful to specify that the movement away from traditional economics does not mean that we are not rational beings; rather, the existing conceptions of rational behaviour fail to describe the logic humans operate by.

We might expect in making decisions, we automatically maximize utility, or go with the choice that leads to the most useful outcome, considering that earlier economists working on utility theory in the late 19th century consistently arrived at this conclusion.³ In a 2006 paper, however, Kahneman and Thaler refuted this hypothesis.¹² They found that because we do not always know what we like, as demonstrated in the California example, we make errors in predicting the future utility of outcomes. As a result, we do not maximize the utility of our decisions because we make erroneous judgments about what will be useful to us. We make intuitive decisions without really thinking things through. When we go to the grocery store on an empty stomach, we often purchase much more food than we actually need – and more than what was on our grocery list – because of how we feel at that moment.

Intuitive decision-making

This error could also be explained by a process of substitution in intuitive thinking, where we wind up answering a different question than the one we intend to address. For instance, when we’re shopping and hungry, we may be making optimized utility decisions for ourselves in that moment, because we would like to eat the food we’re buying. Although we think we’re dealing with food decisions for the week ahead, we are really just addressing our immediate food desires.

In their paper, Kahneman and Thaler addressed four situations where “hedonic forecasting”, or our ability to know what we want in the future, resulted in errors in decision-making:

• Where the emotional or motivational state of the agent is very different at the time of the decision versus time of consumption
• Where the nature of the decision focuses attention on aspects of the outcome that will not be relevant when it is actually experienced
• When choices are made on the basis of flawed evaluations of past experiences
• When people forecast their future adjustment to new life circumstances.

The first instance, as in the example of shopping while hungry, has been proven to result in different outcomes. A similar case has been observed with the current weather influencing the clothes people buy—on an abnormally cold day, it is less likely that people will buy clothes for warm weather, even for future use by ordering on the phone. “Anchoring” in the present moment can result in us making a different decision that we’re not actually conscious of, but our attention becomes focused on our present needs, so we unknowingly make a decision fit for that problem.

In the second case, the way decisions are presented to us can influence how we evaluate the utility of those choices. This is where biases like naive allocation can result in us making a non-optimized decision in terms of utility. Given different assortments of options, we choose differently.

The third case, where we carry imperfect judgments of past experiences, has to do with the way we remember past pain and pleasure. The peak/end rule suggests that our retrospective evaluation of an incident will be composed of the average of our feelings at the most extreme point and at the end of the experience. In other words, we do not remember the beginning or less extreme aspects of an experience as well as the peak/end when thinking about past incidents. Based on studies on different hedonic experiences, such as measuring pain during medical procedures, people’s evaluations of painful experiences can indeed be altered by manipulating the extreme point of the experience or changing the ending. When a procedure ended more gradually and with a period of lesser pain, patients rated it as less painful than procedures that ended abruptly with pain, even though the only difference was the length of the procedure.

The final case describes what happens when people try to envision themselves living in California—we judge our future lives by metrics that won’t actually end up mattering to us. Kahneman also found that we adapt better to situations than we expect. In fact, we often think something will be worse than it actually is. Kahneman compared how paraplegics felt about their lives after they became paralyzed against asking non-paraplegics to estimate their feelings if they became paraplegic. Interestingly, he found that non-paraplegics greatly overestimated the negative effect of the disability and that paraplegics reported doing much better than people imagined.

Thus, utility can still describe the motivation for our decisions, but past models have failed to account for all that we consider useful.

Homo Economicus

This article goes into detail about the concept of Homo Economicus, or the completely logical species imagined by traditional economics. The difference between homo economicus and homo sapiens – or real human behaviour – is addressed, as well as the history of the term.

Behavioural Economics

This article on behavioural economics also highlights the differences between traditional and behavioural economics. As we saw with people’s failure to choose maximum utility, we need to account for our biases and flaws in rational decision-making in order to realistically understand how we make decisions, not just how we ought to make decisions. Behavioural economics, as opposed to traditional economics, takes our psychology and biases into account to study human decision-making.

Evolution of Decision Making: Current State

This article, the third part of a series, gives an overview of the current state of behavioural economics, including decision utility. The author also provides examples of how recent research can be incorporated in our professional and personal lives to improve our decision-making.

1. Robson, A. & Samuelson, L. (2011). The evolution of decision and experienced utilities. Theoretical Economics, 6(3), 311-339. https://doi.org/10.3982/TE800\
2. Stigler, G. J. (1950). The Development of Utility Theory. I. Journal of Political Economy, 58(4), 307–327. http://www.jstor.org/stable/1828885
3. von Neumann, J., and Morgenstern, O. (1944). Theory of Games and Economic Behavior. Princeton University Press.
4. Kahneman, D., & Tversky, A. (1973). On the psychology of prediction. Psychological Review, 80(4), 237–251. https://doi.org/10.1037/h0034747
5. Kahneman, D., Wakker, P. P., & Sarin, R. (1997). Back to Bentham? Explorations of Experienced Utility. The Quarterly Journal of Economics, 112(2), 375–405. http://www.jstor.org/stable/2951240
6. Berridge, K. C., & Aldridge, J. W. (2008). Decision Utility, The Brain, and Pursuit of Hedonic Goals. Social cognition, 26(5), 621–646. https://doi.org/10.1521/soco.2008.26.5.621
7. Berridge, K. C., & O’Doherty, J. P. (2014). From Experienced Utility to Decision Utility.
8. Neuroeconomics, 2, 335-351, https://doi.org/10.1016/B978-0-12-416008-8.00018-8.
9. Schkade, D. A. & Kahneman, D. (1998). Does Living in California Make People Happy? A Focusing Illusion in Judgments of Life Satisfaction. Psychological Science, 9(5), 340-346. http://www.jstor.org/stable/40063318
10. Tversky, A., & Kahneman, D. (1986). Rational Choice and the Framing of Decisions. The Journal of Business, 59(4), S251–S278. http://www.jstor.org/stable/2352759
11. Loewenstein, G., O’Donoghue, T., & Rabin, M. (2003). Projection Bias in Predicting Future Utility, The Quarterly Journal of Economics, 118(4), 1209–1248. https://doi.org/10.1162/003355303322552784
12. Thaler, Richard, H. (2000). From Homo Economicus to Homo Sapiens. Journal of Economic Perspectives, 14(1), 133-141. https://www.aeaweb.org/articles?id=10.1257/jep.14.1.133
13. Kahneman, D., & Thaler, R. H. (2006). Anomalies: Utility Maximization and Experienced Utility. The Journal of Economic Perspectives, 20(1), 221–234. http://www.jstor.org/stable/30033642