The Blind Spot in Traditional Economic Models
Economics has long treated consumption as a clean transaction: you buy something, you enjoy it, the utility spike dissipates, and the ledger closes. What happens after the transaction—what lingers in your memory and tugs at tomorrow's choices—simply does not appear in the equation.
Stefania Minardi at HEC Paris and Andrei Savochkin at Bocconi University decided that omission was a problem worth formalizing. Their paper, "Time for Memorable Consumption," published in Games and Economic Behavior, builds a framework that forces standard models to reckon with something psychologists have known for decades: the experience lives on after the receipt is shredded.
Their approach introduces a split most econ models ignore. You get moment utility—the satisfaction you extract while eating the meal, riding the rollercoaster, sitting through the wedding reception. But you also accumulate remembered utility: a durable payoff (or cost) that surfaces later, when a recollection nudges your next decision about risk, saving, or whether to book another trip to Portugal.
This distinction sounds small on paper. In practice, it breaks the assumption that economic agents are optimizing over present and future material outcomes alone. Memories aren't material outcomes. They're psychological artifacts—and they change the math.
Predicting Human Decisions with Behavioral Theories: Where Memory Enters the Frame
Any framework that tries to predict human decisions using behavioral theory has to confront the fact that people don't store experiences like security footage. We store them as highlight reels with a weird editing process.
Minardi and Savochkin draw on what the psychology literature calls the peak-end rule. We tend to encode an experience by how we felt at its most intense moment and at its conclusion. A two-week vacation where day four involved a perfect sunset dinner and the final morning brought a cancelled flight? You remember those two points. The mediocre middles evaporate.
The source article illustrates this with examples like weddings and career achievements, episodes whose remembered satisfaction continues to shape well-being long after they occurred. That persistence is what the model formalizes. An experience qualifies as "memorable" when the memory itself carries utility at a future date, independent of whether anything about your material circumstances has changed.
I want to be careful here. The peak-end rule is useful shorthand, not a law of physics. It captures a reliable bias in how people reconstruct past episodes, but it doesn't predict every individual's recall with equal confidence. The authors are transparent about that limit. What they do is incorporate the average pattern into an economic model and then ask: if people behave this way, what happens to predictions about their choices?
How Remembered Experience Tilts Risk-Taking
One payoff of the framework shows up in corporate decision-making. Managers make project choices under uncertainty all day. Standard agency theory says they should evaluate expected returns, adjust for risk aversion, and move on.
Minardi and Savochkin show that when managers carry vivid memories of past project outcomes, their risk posture shifts in non-obvious ways. A manager who remembers a spectacular success from a previous risky venture may become reckless, chasing the remembered thrill rather than recalculating the odds from scratch. Flip the script: a manager whose memory encodes a painful failure at its peak moment can become excessively conservative, forgoing mildly risky projects that carry positive expected value.
The model doesn't judge which response is more common. It maps the conditions under which each emerges, depending on what the memory's peak and end look like relative to actual expected returns. The point is that risk-taking becomes path-dependent in ways traditional models can't explain, because those models have no place for a memory that still burns two years later. This stickiness of emotionally encoded priors echoes what we've explored elsewhere in terms of serotonin's role in reducing belief stickiness: updating in light of new evidence is hard precisely when the old signal carries a strong neurochemical imprint.
Life-Cycle Saving: Your Vacation Remembers You
Here's where it gets personal. The authors apply the model to intertemporal consumption choices, the life-cycle savings decisions that every household eventually faces.
Suppose you took a memorable vacation last year. The peak-end rule tells us you'll retrieve a disproportionately vivid, positive snapshot of that trip. When you sit down to plan next year's finances, that remembered utility raises the anticipated payoff of future similar experiences. Result: you save more, earmarked for another shot at the high.
Now flip it. The same trip cost a fortune, left you stressed at its end, and your brain encodes it as a net negative. Next year's saving decision looks different. You might cut travel budgets entirely or redirect savings toward goals that aren't experience-based at all.
The framework formalizes these responses within a life-cycle model. Income shocks hit, but the sign of the saving response is no longer deterministic, it depends on the emotional architecture of what you remember from prior consumption. That's a level of heterogeneity traditional saving models simply flatten away. Patience itself has a biological substrate here, as we've covered in our look at how serotonin and a confidence threshold shape human patience; remembered utility adds a psychological layer on top of that machinery.
Why This Changes the Explanation, Not Just the Prediction
It would be easy to file this work under "nuance the parameter, keep the model." Minardi and Savochkin's result is more structural than that. Once you accept that consumption generates a second utility stream, remembered utility, several standard conclusions lose their grip. Consumption is no longer just about present welfare and future material payoffs. Well-being becomes partially a function of the memories you're actively constructing through today's purchases and experiences.
This doesn't mean we should abandon rational-choice apparatus. It means the rationality constraint needs an enriched state variable. The agent carries not just wealth and income expectations but a memory set shaped by peak-end encoding. Predictions that ignore that set will systematically mis-forecast who saves, who gambles, and who books the ticket. And memory is not free to maintain: the biological limits on decision-making run deep, a theme our earlier piece on bioenergetic constraints in decision shifts takes seriously.
I find the elegance compelling: one modification, the addition of a memory-augmented utility channel, generates testable predictions about real behavioral patterns that standard models leave unexplained. That's what good interdisciplinary work does. It doesn't replace the parent discipline. It shows where the parent discipline was quietly assuming something inconvenient could be ignored.
It couldn't be ignored here. And that single concession opens an entire research agenda about how the architecture of human recall reshapes the economics of lived experience.
Source
- How Memories Shape Decisions and Future Well-Being, Neuroscience News, summarizing research from Bocconi University and HEC Paris.
