In this paper, we propose a novel theory of choice that is especially relevant when one of the options is the default. We employ the concept of opportunity cost to factor in the foregone value of the alternative not chosen. We have two main objectives: First, to explain experimental evidence related to the “endowment effect”, especially the heterogeneity of behavior among decision makers. While the endowment effect is typically observed on average, there are significant minorities who exhibit the opposite behavior, akin to the phenomenon of “Fear of Missing Out”. Second, to offer an alternative to regret theory (Loomes & Sugden, 1982) that is not based on the utility difference between the options, but on the psychological cost of forgone opportunity.

Opportunity cost is widely considered as one of the fundamental concepts in microeconomic theory. It actually appears to be a misnomer: the opportunity cost is not a cost actually incurred, the concept simply draws attention to the inescapable fact that choosing a particular course of action implies giving up alternative ones. A typical example is a store that sells buttons and thread in one of the main shopping streets of Palma de Mallorca. When its owners pocket their monthly profit of 2000 euros, they better be aware that they could rent out their locale for 3000 euros. The sentimental value of following in the footsteps of countless generations1 may compensate for the foregone gain, but the opportunity cost is to be reckoned with. In this paper, we propose a mechanism that helps to justify their decision (see Section 5 for our resolution of the example).

Empirical evidence suggests that people take into account the alternatives not only at the time of choosing – internalizing the cost of opportunity – but, in fact, already their evaluation of each option is affected by a comparison to the alternatives.2 We propose that the two phenomena are connected: the way relative evaluation modifies an option’s valuation is by incorporating a loss that closely corresponds to the opportunity cost attached to not choosing the alternative option. More precisely, we model a decision maker (DM) who internalizes an actual opportunity cost of his (potential) choices: when evaluating an option, he subtracts from its stand-alone utility – the utility of the option if it were the only one available – (a fraction of) the stand-alone utility of the best alternative. Put in another way: the familiar Homo Oeconomicus first evaluates each option separately and then compares them; while the Homo Consimilis that we propose incorporates the comparators into his evaluation of each option already, accounting for an opportunity cost, before choosing the best one.

We thus propose that the mere fact that an available option has been considered may lead to a psychological loss when it is not chosen. An immediate consequence is that the value of an option is different (lower) when it is evaluated in the context of choice. This is precisely what Bordalo et al. (2012) observed in a set of experimental choices between monetary lotteries. Participants valued lotteries considered in isolation more than the same lotteries considered along alternatives. For an everyday example: a person who chooses to spend the evening at the beach will feel less happy about her choice when she could also go to a concert, even if she prefers the beach.

If the value of an option is diminished as a function of the options foregone, then choosing entails a form of losing.3 Our message goes beyond identifying and (potentially) measuring this utility loss. We wish to explore how the actual choice behavior may be affected by the perceived opportunity costs. In a choice between two options, say A and B, choosing A over B entails the cost of losing B; and symmetrically, choosing B over A entails the cost of losing A. If these losses were the same fraction of the stand-alone utilities, it is straightforward to show4 that our approach would lead to the same choices as those of Homo Oeconomicus, predicted in the absence of relative evaluation.5

However, an asymmetry can emerge between the relative valuations of A and B when one of the options is the default. In many natural situations, DMs decide between the status quo and a new option (Samuelson & Zeckhauser, 1988). For example, whether to keep the old car or to buy a new one, to keep or replace an employee, to work with the current supplier or to try a new one, to do business in the same region or to explore new areas, etc.6 We argue that losing the default option in favor of an alternative may have a different – in size, but not direction – impact on (relative) valuation than giving up the alternative in favor of the default, even if their stand-alone utility is the same.

Importantly, the direction of the asymmetry need not be the same between individuals. Some DMs may show behavior consistent with the “endowment effect” (Kahneman et al., 1990) where they are affected more strongly by the loss of the owned default,7 while others may exhibit the opposite reaction, being particularly reluctant to give up a chance to try something different – a phenomenon similar to the “fear of missing out” or “FoMO” (Przybylski et al., 2013), but without the social context.8 Thus, we can explain the general finding that in experiments documenting the endowment effect – willingness to pay lower than willingness to accept – a significant number of subjects actually display the opposite behavior (Birnbaum et al., 2016, Chapman et al., 2023, Gal, 2006). This could be important, as a frequently used policy design – promoted by the nudge program (Thaler & Sunstein, 2008) – involves setting a socially desirable option as the default, encouraging its choice (for example, as in the case of organ donation, Yan and Frank Yates (2019). The evidence suggests that such policies can work on average (Jachimowicz et al., 2019), but they might backfire for a fraction of citizens. Therefore, a better understanding of how defaults influence choice may help design better policies.

Our key observation is that for every DM the direction of the bias is the net effect of two opposing forces. In addition to the stand-alone utilities, they factor in two potential opportunity costs, each of them pulling in opposite directions. The net effect depends on the relative size of the parameters that capture what fraction of the opportunity cost is internalized.

Note that we propose a form of “loss aversion” that differs from the original formulation (Kahneman & Tversky, 1979). In prospect theory, the potential outcomes of an uncertain option – or the values in different dimensions of a riskless option (Tversky & Kahneman, 1991) – are classified as gains or losses relative to a reference point, and the potential losses within an option receive more weight than the potential gains. Our approach is qualitatively different. Although we do not exclude the possibility of loss aversion in the evaluation of outcomes within a – potentially uncertain – option, we model the ‘loss’ of the entire option. In other words, in case the option is a lottery, we are not concerned with how loss aversion plays out within the lottery, as Kahneman and Tversky do. Instead, we consider the loss of the lottery itself, relative to the reference point of having the lottery. This view of a loss caused by resigning an option not chosen is effectively the opportunity cost of choosing another option. Our view entails another conceptual difference from prospect theory. While the reference point in prospect theory is unique and serves as the benchmark to evaluate all potential outcomes in a choice, our model assumes that the reference point changes: When evaluating A, the loss of B is relative to having B; and when evaluating B, the loss of A is relative to having A.

It is important to single out a key precursor of our model: regret theory. Loomes and Sugden (1982)9 also propose that the utility of a potential outcome of a lottery is affected when an alternative lottery is available. Specifically, according to their formulation, the utility is reduced or increased as a function of the difference between the stand-alone utilities of the chosen and foregone options (in the realized state of nature), and thus the DM may experience regret or rejoice, after the realization of uncertainty. Crucially, regret (or rejoice) disappears when the available options have the same stand-alone utility. This view of foregone options does not correspond to the concept of opportunity cost, which depends only on the value of the alternative. In our view, the negative effect of giving up a worse option is actually largest, when it is valued nearly the same as the chosen option. For example, a person who likes to dine out with friends and to go to concerts equally, and prefers both to going to the opera, would rank her happiness in a dinner out as: best if neither concert nor opera is on, second-best if opera is on, and worst if concert is on, independently of whether there is opera on. This phenomenon is not captured by regret theory, and here we propose an alternative formulation of regret that does (addressing Bourgeois-Gironde’s, (2010), call to refine the original concept). We discuss in detail the relationship between prospect theory, regret theory, and ‘opportunity cost theory’ (OCT) in Section 3. In Section 2 we set up and analyze our model for the choice between the default and an alternative option. In Section 4 we argue that the extension to several alternatives is relatively straightforward. In Section 5 we discuss a crucial detail: How can the choice problem be isolated from a real-world scenario? Here we appeal to the ceteris paribus principle: Features that are common to the options are ‘canceled out’ when the evaluation takes place. In Section 6 we discuss our results, while Section 7 contains some concluding remarks.