The Wine Shop That Started a Revolution

It was a Thursday afternoon in late October when everything changed for a quiet professor named Thierry. He had walked into La Maison du Vin, a venerable wine shop tucked into a limestone corner of Bordeaux, seeking nothing more than a bottle of wine for dinner. The shop had been there since 1923, its wooden shelves heavy with bottles from the world's most celebrated vineyards. Thierry wasn't a wine connoisseur—he taught behavioral economics at the university—but he found himself frozen in place, confronted by a wall of prices that seemed to stretch into infinity.

The cheapest bottle cost €8. The most expensive, a Château Margaux from an exceptional vintage, cost €4,200. In between lay a dizzying array of options: €15 bottles, €30 bottles, €50 bottles, €100 bottles, €200 bottles, €500 bottles. Thierry, a man who prided himself on rational decision-making, suddenly felt something he had never experienced before—a complete inability to choose. His eyes kept drifting back to a bottle priced at €85, a well-regarded but hardly legendary Pessac-Léognan. It felt, somehow, like the right choice. The sensible choice. The choice that wouldn't make him look cheap, but wouldn't make him look foolish either.

What Thierry didn't know was that his confusion—and his eventual choice—had been observed, catalogued, and dissected by researchers for decades. The €85 bottle he kept returning to had a name in behavioral economics: it was the "decoy," or more precisely, the item that benefited from something called the "compromise effect." But before we can understand why Thierry and millions like him consistently choose the middle option, we need to understand a much stranger phenomenon that lies at the heart of how humans make decisions.

This is a story about numbers. Not the numbers themselves, but what happens inside our minds when we encounter them. It's a story about why writing down your Social Security number might cause you to pay more for wine you'd never heard of. Why the first price you see for a house can haunt you for decades. Why the opening salary in a negotiation can determine your earnings for years. Why a judge's breakfast might influence whether someone goes to prison.

This is the story of anchoring—and it will change the way you see every number you've ever encountered.

The man who first noticed this phenomenon, at least in its modern scientific form, was a psychologist named Tversky. He was working in the 1970s with his colleague Kahneman, developing what would become the foundations of behavioral economics. Their early experiments yielded a finding so counterintuitive that even the researchers themselves struggled to believe it at first.

They asked students at an Israeli university to spin a wheel—intentionally rigged to land on either 10 or 65—and then write down the number that appeared. After the number was recorded, the students were asked whether the percentage of African nations in the United Nations was higher or lower than their written number. Finally, they were asked to estimate the actual percentage.

The results were staggering. Students who had landed on 10 estimated that African nations made up about 25 percent of the UN. Those who landed on 65 estimated the figure was closer to 45 percent. The wheel had been completely random. It had nothing to do with African nations. Yet the first number encountered—the anchor—had powerfully shaped the subsequent judgment. The higher the anchor, the higher the estimate. Every single time.

This was the first glimpse into what would become one of the most robust and pervasive phenomena in all of human cognition. We are all, it turns out, deeply susceptible to anchoring. The first number we see becomes a reference point that we never fully escape. It pulls our estimates toward itself like a gravitational field, distorting our judgments in ways we cannot detect and would never voluntarily choose.

When Thierry finally bought his €85 bottle at La Maison du Vin, he had no idea that his decision had been influenced by the prices flanking it. The €8 bottle on the left said: "This is what cheap wine looks like. This is what you become if you choose this." The €4,200 bottle on the right said: "This is what excessive looks like. This is what foolish people do." The €85 bottle said: "This is the reasonable choice. This is what smart, sensible people pick."

But of course, none of those messages were written anywhere. The prices simply existed, and Thierry's mind had done the rest.

The experiment that would most dramatically demonstrate anchoring's power came decades later, in a paper that has since become one of the most cited in behavioral economics. Dan Ariely, then at MIT, along with George Loewenstein and Drazen Prelec, wanted to know whether completely arbitrary numbers could anchor people's willingness to pay for ordinary items.

They gathered students and showed them various products: a bottle of wine, a chocolate bar, a wireless keyboard, a book. Before bidding on each item, students wrote down the last two digits of their Social Security numbers—the most arbitrary, meaningless number imaginable. Then they wrote whether they'd be willing to pay amounts starting with those digits.

The results were extraordinary. Students with Social Security numbers ending in the high 90s were willing to pay significantly more for every single item than students with numbers ending in the low teens. The highest-fifth of students, those with SSN endings between 80 and 99, bid an average of $26.20 for the wine. The lowest-fifth, with endings between 00 and 19, bid only $11.80. The arbitrary number had influenced a real economic decision by more than double.

When asked afterward whether their Social Security numbers had affected their willingness to pay, nearly everyone said no. They were certain their bids reflected their true valuations. The anchor had done its work invisibly, without their awareness or consent.

But how? What exactly is happening in the brain when an arbitrary number takes hold and refuses to let go?

The mechanism turns out to be surprisingly intuitive once you understand it. When we encounter a number in any context, our minds automatically use it as a reference point—a starting place from which we then make adjustments. This adjustment process is typically insufficient. We don't move far enough away from the anchor to fully compensate for its influence. Psychologists call this "insufficient adjustment," and it appears to be a fundamental feature of human cognition.

Imagine someone asks you whether the population of France is greater or less than 50 million. You don't know the exact figure, but you have some vague sense that it's around 65 million. So you answer: greater. But now imagine someone asks whether the population is greater or less than 70 million. Your answer might be the same—greater—but the mental image of France's population that now forms in your mind has shifted upward. Your estimate is now anchored to 70 million, and you'll adjust from there.

This adjustment mechanism evolved for a reason. In the ancestral environment, numbers were rare and calculations were expensive in terms of cognitive effort. Using a reference point and then adjusting made evolutionary sense—it was efficient, fast, and good enough most of the time. The problem is that in modern environments filled with numbers, prices, statistics, and metrics, this ancient cognitive shortcut leads us systematically astray.

The wine shop scenario illustrates this perfectly. When Thierry walked in and saw a wall of prices, his brain did what brains do: it grabbed the first number it encountered—or perhaps the most salient numbers, the extremes and the middle—and used them as reference points. Then, when trying to decide what to pay, he adjusted from these anchors. Since the most expensive bottle was €4,200, even €500 seemed moderate. Since the cheapest was €8, even €30 seemed like a serious commitment. The €85 bottle la