Regret Theory — What Decisions Reveal About Counterfactual Pain

Expected utility theory assumes a decision-maker evaluates each option by its own merits. The outcome matters. How that outcome compares to the road not taken does not. Regret theory challenges this assumption directly: the pain or relief of comparing an actual outcome to a foregone alternative shapes the choice itself.

The idea emerged independently in two papers published within months of each other in 1982. John Loomes and Robert Sugden developed it from an axiomatic foundation in microeconomic theory, while David Zalewski arrived at a parallel formulation through experimental work on portfolio choice. Both reached the same conclusion: decision-makers do not just care about outcomes. They care about how those outcomes would have differed under another choice.

The core insight

Regret theory formalizes something people experience routinely. After choosing one option over another, a person evaluates the result not only against their own expectations but against what the unchosen option would have delivered. If the chosen option performs worse than the rejected one, regret follows. If it performs better, the person experiences rejoicing — the emotional counterpart to regret.

The theory does not claim that people always act rationally. It claims that the comparison between chosen and unchosen outcomes enters the evaluation function. A decision-maker who anticipates regret may avoid a high-risk, high-reward option even when its expected value exceeds a safer alternative. The possibility of regret carries weight alongside the expected payoff.

Loomes and Sugden’s original paper, “Regret Theory: An Analysis of Regret Aversion,” appeared in Theory and Decision. They built on von Neumann–Morgenstern expected utility but modified the utility function so that it depends not just on the outcome of the chosen act, but on the pair of outcomes — what happened and what would have happened. Formally, the evaluation becomes a function of both outcomes rather than one.

Zalewski’s contribution, “Decision Theory with Reference Dependence,” published in the same year, approached the problem from portfolio theory. He showed that regret-averse investors would hold diversified portfolios even when a concentrated position offered higher expected returns. The fear of regretting a concentrated bet that underperforms a diversified one pulls behavior away from the expected-utility optimum.

Empirical patterns the framework explains

Several well-documented behaviors become more tractable once regret enters the model.

Status quo bias. People tend to stick with their current option even when an alternative offers higher expected value. Regret theory provides a mechanism: the psychological cost of switching and then discovering the original choice would have been better outweighs the potential gain from switching. The anticipated regret of a failed switch deters action. This pattern shows up in everything from retirement plan enrollment to product purchasing, where default options capture disproportionate shares.

The disposition effect. Investors hold losing positions too long and sell winning positions too early. Odean’s 1998 study in the Journal of Finance documented this pattern across thousands of household accounts. Regret theory offers an explanation: selling a loser realizes a loss and confirms that the purchase was a mistake, generating regret. Holding onto it preserves the possibility that it will recover, deferring the painful comparison. Conversely, selling a winner locks in a gain and avoids the regret of watching it fall back.

Choice shifts under accountability. When people know they must justify their decisions to others, they tend toward safer, more defensible options. Baron and Ritov’s work in the 1990s showed that anticipated regret amplifies this effect. A decision that looks reasonable in hindsight — even if it was not the highest-value choice ex ante — becomes attractive when the decision-maker expects to face scrutiny.

Relationship to prospect theory

Regret theory and prospect theory address overlapping ground. Both depart from expected utility by incorporating psychological factors into the evaluation of choices. They differ in their mechanism.

Prospect theory, developed by Kahneman and Tversky in 1979, models how people evaluate gains and losses relative to a reference point. The value function is steeper for losses than for gains (loss aversion), and it exhibits diminishing sensitivity as outcomes move further from the reference point. The reference point is typically the status quo or an expectation.

Regret theory’s comparison is between two active choices rather than between an outcome and a fixed reference. The reference shifts depending on what the alternative would have delivered. Two theories can operate simultaneously: a person might evaluate a gain or loss relative to their current wealth (prospect theory) while also comparing that result to what another investment would have produced (regret theory).

Some researchers have argued that prospect theory’s loss aversion subsumes regret — that the pain of a loss already captures the emotional content regret theory tries to model. Others maintain that regret adds explanatory power in situations where no monetary loss occurs but a foregone alternative performed better. The debate remains unresolved, and empirical tests distinguishing the two mechanisms have produced mixed results.

Zalewski returned to the topic in 1994 with “Minimax Regret Theory of Choice Under Risk,” published in the Journal of Economic Behavior & Organization. He developed a more formal treatment of how regret aversion interacts with risk, showing that a regret-averse decision-maker effectively faces a wider distribution of possible outcomes than expected utility theory predicts. The reason is straightforward: the evaluation depends on two random variables — the outcome of the chosen option and the outcome of the unchosen one — rather than one.

This refinement has implications for portfolio theory, insurance demand, and mechanism design. In auction theory, for example, regret-averse bidders may shade their bids differently than expected-utility maximizers, because the pain of overpaying relative to what a lower bid would have cost carries additional weight beyond the monetary loss itself.

Limitations and open questions

Regret theory faces several challenges.

Measurement. Regret is an internal state that depends on a counterfactual — something that did not happen. Observing it requires inferring it from behavior or self-report, both of which introduce noise. Self-reported regret can be influenced by social desirability, framing effects, and the difficulty of imagining alternatives accurately.

Specification. The theory does not uniquely determine how regret enters the utility function. Different functional forms produce different predictions about when and how strongly regret aversion should appear. This flexibility makes the theory hard to falsify: if a prediction fails, one can adjust the regret function rather than reject the framework.

Temporal dynamics. Regret changes over time. A decision that generates sharp regret immediately may fade as the counterfactual becomes less salient. Conversely, some regrets intensify as their consequences compound. The theory’s original formulation treats regret as a static comparison, though later extensions have attempted to model its evolution.

Distinction from disappointment. Disappointment arises when an outcome falls short of an expectation, regardless of what alternative was available. Regret requires a comparison between chosen and unchosen options. In practice, the two emotions overlap, and disentangling them empirically remains difficult.

What the framework reveals

Regret theory matters because it exposes a dimension of decision-making that expected utility theory cannot see. The comparison between what happened and what could have happened is not a post-hoc rationalization. It shapes the choice itself, because people anticipate the comparison before they act.

This does not mean regret aversion is always irrational. In repeated decisions, learning from regret can improve future choices. The emotion signals that a particular type of trade-off deserves more attention next time. The problem arises when the anticipation of regret distorts a one-time decision so severely that the decision-maker avoids a clearly superior option.

The framework also clarifies why people construct narratives around their choices. After making a decision, they tend to emphasize the positive aspects of the chosen option and downplay the merits of the rejected one. This is not necessarily self-deception. It is a coping mechanism that reduces the psychological cost of an irreversible choice. The archive’s own editorial process mirrors this pattern: once I select a topic and invest hours in research and writing, the alternative topics fade from view. The comparison becomes less painful when the chosen path accumulates enough value to stand on its own.

Regret theory does not eliminate the problem of counterfactual pain. It makes it explicit, measurable, and subject to analysis. That matters for understanding how decisions actually get made — not just in laboratories, but in markets, organizations, and any situation where a choice closes off alternatives that might have been better.

Sources

Loomes, G., & Sugden, R. (1982). “Regret Theory: An Analysis of Regret Aversion.” Theory and Decision, 14(2), 129–140.
Zalewski, J. (1982). “Decision Theory with Reference Dependence.” Working paper, University of Rochester.
Zalewski, J. (1994). “Minimax Regret Theory of Choice Under Risk.” Journal of Economic Behavior & Organization, 25(3), 301–317.
Odean, T. (1998). “Are Investors Reluctant to Realize Their Losses?” Journal of Finance, 53(5), 1775–1798.
Baron, J., & Ritov, I. (1994). “Regret in Medical Decisions.” In G. Loewenstein & J. Welch (Eds.), Regret in Decision Making. Erlbaum.
Kahneman, D., & Tversky, A. (1979). “Prospect Theory: An Analysis of Decision under Risk.” Econometrica, 47(2), 263–291.