What Is Behavioral Economics? Contrasting Rational Choice Theory and Real-World Decision-Making

Last Wednesday after school, you stood in line at the McDonald’s counter outside your campus, clutching the $12 in your pocket: $8 from your weekly allowance and $4 you earned walking your neighbor’s golden retriever, Buddy, three times while they were out of town. You’d planned to buy the $10 Big Mac Combo—Big Mac, medium fries, medium Coke—when you spotted the bright red sign above the register: “Family Duo Meal: Only $11.99! Includes 2 Big Macs, 2 large fries, 2 medium Cokes, and a free apple pie. Total value: $23.94!” Your brain did a quick calculation: $23.94 worth of food for less than $12? That’s more than 50% off! You handed over your $12, took the paper bag heavy with two burgers, two overflowing fry containers, and two sodas, and walked home feeling like you’d just pulled off a heist. By the time you finished one Big Mac and half an order of fries, you were so full you couldn’t look at the second burger. You threw half the fries in the trash, left the second Coke half-full on your desk, and ate the apple pie the next day when you were already still full. Later, your mom asked you why you bought so much food when you only wanted one combo, and you said, “It was a better deal.” If that sounds familiar, you’ve already encountered the core puzzle this book will unpack: why do people so often make choices that don’t line up with the “rational” decisions traditional economics says we should make?

For decades, the standard model taught in every introductory economics class has relied on a single, unproven assumption: that all human beings are *Homo Economicus*, the hypothetical perfectly rational agent at the core of neoclassical economic theory. This model assumes that every person, when faced with a choice, will automatically gather all available relevant information, weigh the costs and benefits of every possible option without emotional bias, and select the option that delivers the highest possible *utility*—a term economists use to describe the level of satisfaction or happiness a person gets from a good, service, or choice. Under this framework, you buying the Family Duo Meal would only make sense if the total utility you got from eating two Big Macs, two large fries, two Cokes, and an apple pie was higher than the utility you would have gotten from the single Big Mac Combo, minus the $1.99 extra you spent and the negative utility of wasting food. If you only ate one Big Mac and half an order of fries, the model would predict you made an irrational choice—one that will make you worse off in the long run.

The rational actor model has been incredibly useful for economists. It simplifies the messy complexity of human behavior into a predictable, mathematically consistent framework that lets us model everything from how supply and demand set prices to how tax policy changes consumer spending. But it has a fatal flaw: it does not describe how real people actually make decisions. Real people do not have perfect information. We do not have unlimited time to calculate every possible outcome of our choices. We get hungry, we get tired, we get excited about a new video game drop, we want to impress our friends, we hate the feeling of missing out on a “good deal.” These psychological factors are not “noise” that distorts rational decision-making: they are the core of how we make choices, every single day. Traditional neoclassical economics has no way to account for them, which is why it so often fails to predict what real people will actually do.

To see this gap in action, let’s break down the Big Mac Combo choice using the traditional rational model, then compare it to what you actually did. First, the model’s calculation: if you single-item ordered a Big Mac ($6.99), medium fries ($2.99), and medium Coke ($1.99), the total cost would be $11.97, so the $10 combo saves you $1.97. The Family Duo Meal has a listed “total value” of $23.94, so buying it for $11.99 seems to save you $11.95—far more than the $1.97 you save with the single combo. The model would therefore predict you choose the Family Duo Meal, which is exactly what you did. But the model ignores three critical pieces of real-world context: first, you have no way to use all the food in the Family Duo Meal, so the “total value” of $23.94 is meaningless to you personally; second, wasting food creates negative utility—you feel guilty, you waste money, you have to throw away food that could have been eaten by someone else; third, the “saving” of $11.95 is an illusion, because you would never have paid $23.94 for food you can’t eat. When you account for these factors, the actual utility of the Family Duo Meal is lower than the utility of the single Big Mac Combo: you spent $1.99 more, felt full and uncomfortable after eating, and wasted food that made you feel guilty. The model got your choice right, but it got the reason for your choice completely wrong—and it had no way to predict that you would feel regret after eating.

Let’s take a second, even more common teen scenario to test the limits of the rational model. Imagine you get $20 a month in allowance, plus $5 for every week you complete your extra chore of taking out the trash without being reminded. You’ve been eyeing the limited-edition *Splatoon 3* Pro Controller for months: it’s $70 retail, but there’s a 24-hour flash sale next week where it will be $60. You have two clear options: 1. Option A: Spend $15 this week on the new *Splatoon 3* DLC that came out today, save the remaining $5, and repeat this pattern for the next three weeks. After three weeks, you will have spent $45 on DLC, saved $15, and still need $45 more to buy the controller at the sale price. 2. Option B: Skip the DLC this week, save the full $20, and keep saving for the next three weeks. After three weeks, you will have exactly $60 to buy the controller at the flash sale price, plus an extra $10 left over for snacks at the mall.

The rational model would immediately flag Option B as the strictly better choice. You get the high-value controller you want for $10 less than retail, you have extra disposable income left over, and you avoid the “impulse purchase” of the DLC that you might only play for two weeks before getting bored of the new levels. If you are like the vast majority of teens surveyed in a 2023 study of adolescent spending habits, however, you will pick Option A. When asked why, 78% of teens in the study said the DLC “felt more fun right now,” while 62% said they “didn’t want to wait

while 62% said they “didn’t want to wait for something they could have right now,” even when they could clearly articulate that the controller was a higher-value long-term purchase. A separate 41% admitted they would have regretted buying the DLC the next day, but said they “couldn’t help it” in the moment. This gap between what people *say* they want, what they *know* is better for them, and what they *actually do* is not a quirk of teenage immaturity: it is a consistent, replicable pattern of human decision-making that traditional economics has no framework to explain.

For most of the 20th century, the rational actor model was not just a simplifying assumption for economists—it was treated as a factual description of how all humans make choices. Its origins trace back to 19th-century classical economists like John Stuart Mill, who first formalized the idea of *Homo Economicus* as a hypothetical tool to model market behavior, never intending it to be a literal portrait of real people. Over time, however, the tool became the truth: introductory economics courses taught the rational model as the default way humans operate, and deviations from it were dismissed as "irrational" errors, ignorance, or a lack of self-discipline. This framing had a critical, unexamined flaw: it assumed that if people’s choices did not match the model’s predictions, the people were wrong, not the model.

The first major empirical challenge to this assumption came not from economists, b