Predictably Irrational

Random numbers shape real spending. Ariely asked MBA students to write down the last two digits of their Social Security numbers, then bid on wine and electronics. Students with high-ending digits (80 – 99) bid 216 to 346 percent more than those with low digits (1 – 20). The arbitrary number became a price anchor influencing all subsequent valuations — a phenomenon Ariely calls arbitrary coherence.

Anchors explain how markets are made. Salvador Assael turned worthless Tahitian black pearls into luxury items by placing them in Harry Winston's Fifth Avenue window beside diamonds and rubies. Starbucks escaped Dunkin' Donuts pricing by creating an entirely different experience — different names, sizes, and ambience — so customers formed a new anchor rather than comparing on an old one.

One penny makes all the difference. When a Lindt truffle cost 15¢ and a Hershey's Kiss cost 1¢, 73% chose the truffle. Drop both by one penny — truffle at 14¢, Kiss at free — and 69% grabbed the Kiss. The relative gap was identical, but free flipped the majority toward the inferior chocolate.

The zero price effect appears everywhere. Amazon offered free shipping above a threshold and saw sales spike globally — except in France, where shipping cost one franc (about 20¢). Switching France to truly free triggered the same surge. We overvalue free because it eliminates loss: a free item carries zero risk of regret. Ariely suggests policymakers use this power deliberately — making colonoscopies and mammograms free, not just cheap.

We live in two exchange worlds. Ariely asked participants to drag circles across a screen. Those paid $5 dragged 159; those paid 50¢ dragged 101. Those asked as a social favor, with no payment, dragged 168 — outperforming both paid groups. When AARP asked lawyers to help retirees at $30/hour, they refused. Asked to volunteer free, they agreed.

Once market norms invade, social norms vanish. An Israeli daycare introduced fines for late pickups. Parents arrived later — converting guilt into a fee. When the fine was removed, lateness didn't decrease. Both the social contract and the penalty were gone. Companies that brand themselves as family but impose punitive fees face the same trap: one market-norm violation can destroy years of social goodwill.

Ariely measured the Jekyll-Hyde gap. Berkeley undergraduates answered questions about sexual preferences, moral boundaries, and condom use twice: once calm and once while sexually aroused. In the aroused state, desire for unusual activities jumped 72%, predicted willingness to pressure a date rose 136%, and likelihood of skipping a condom increased 25%.

The gap was consistent and enormous. Even students with extensive experience of arousal couldn't predict their transformation. This has direct implications: "Just say no" campaigns assume we can override passion at will. Ariely argues we must either avoid triggering situations entirely or prepare practical safeguards — like always carrying condoms — while calm, because our emotional self simply won't cooperate.

Dictated deadlines beat freedom. Ariely gave three classes the same assignments with different rules. One got dictated deadlines (weeks 4, 8, 12). Another set their own deadlines with penalties. The third had only an end-of-semester deadline. Dictated deadlines produced the best grades; total freedom produced the worst; self-imposed deadlines fell in the middle.

Precommitment works beyond school. Ford struggled with vehicle maintenance until copying Honda's approach: bundling 18,000 potential service items into three simple intervals. Service bays went from 40% empty to full in three years. Thaler and Benartzi's "save more tomorrow" program let employees commit future raises to retirement savings — boosting rates from 3.5% to 13.5% without anyone feeling the pinch.

Duke basketball tickets tell the story. After a campout lottery, Ariely called winners and losers. Non-owners offered about $170 for a ticket. Owners demanded about $2,400. Not a single trade occurred — a 14-fold emotional chasm created by random chance.

Three quirks drive the endowment effect:
1. We fall in love with what we already have
2. We fixate on what we'll lose, not what we'll gain
3. We assume buyers see the same value we do

Virtual ownership makes it worse: the longer you lead an eBay auction, the more "yours" the item feels. Trial promotions and 30-day guarantees exploit this — once something enters your home, the pain of returning it exceeds the logic. Ariely and Mike Norton call the related labor-to-love phenomenon the Ikea effect.

Ariely served vinegar beer to unsuspecting bar patrons. "MIT Brew" — Budweiser spiked with balsamic vinegar — was preferred by most tasters when they didn't know about the vinegar. Told beforehand, they hated it. Told afterward, they liked it as much as blind tasters. Knowledge before the experience didn't just change opinion; it changed the taste itself.

Price reshapes physiology. Participants given vitamin C capsules branded as "Veladone-Rx" at $2.50 per pill nearly all reported pain relief from electrical shocks. At 10 cents, only half did. Students who paid full price for SoBe energy drinks solved 28% more word puzzles than discount buyers. fMRI scans confirmed Coke's advantage over Pepsi lives in brand associations, not chemistry.

Given a chance to cheat, most people do — but barely. Across experiments at Harvard, MIT, Princeton, and Yale, participants who could inflate test scores claimed about 3.5 extra correct answers. Even when they could shred all evidence, they didn't cheat more. The risk of getting caught barely influenced behavior — something inside held them back.

But a moral nudge eliminated cheating entirely. Participants asked to recall the Ten Commandments before the test — even those who remembered only one or two — didn't cheat at all. Signing a statement about an honor code had the same effect, even at MIT, which has no honor code. It wasn't the content that mattered but contemplating any moral benchmark at the point of temptation.

Ariely left six-packs of Coke in MIT dorm fridges — gone in 72 hours. Plates of dollar bills left in the same fridges? Untouched. When he replaced cash rewards in cheating experiments with tokens exchangeable for money seconds later, cheating more than doubled. Of 150 token-condition participants, 24 cheated "all the way" versus just 4 of 2,000 in cash experiments.

Distance from cash lubricates rationalization. Insurance claimants inflate losses by about 10%. Wardrobing — wearing clothes then returning them — costs retailers $16 billion annually. Expense reports get looser when submitted through assistants. As society shifts from physical money to electronic transactions, Ariely warns this moral buffer will only grow thinner and opportunities for small-scale dishonesty will multiply.