By Richard H. Thaler in the New York Times:
If a business school professor is running short on cash, there is a sure-fire solution: run a dollar auction game in class.
To start, the professor offers to sell the class a $20 bill. Bidding starts at $1 and goes up in $1 increments. The winner pays the professor whatever the high bid was, and gets the $20. Here’s the catch: the second-highest bidder also has to pay, but gets nothing in return.
Typically, a few brave or stupid students — nearly always male — open the bidding but fairly quickly only two bidders remain and they discover they are in a war of attrition. The bidding slows when someone bids $20, but then resumes with neither wanting to “lose.” If the two students are particularly stubborn, prices can go over $50. (The professor typically gives the money to charity, or claims to.)
The dollar auction game was invented by a pioneer of game theory, Martin Shubik of Yale, and it illustrates the concept of “escalation of commitment.” Once people are trapped into playing, they have a hard time stopping. (Consider Vietnam.) The higher the bidding goes, and the more each bidder has invested, the harder it is to say “uncle.” The best advice you can give anyone invited to play this particular game is to decline.
Some games and battles are like that: even when you win, you lose. When you see at the start that such a dynamic is likely, you’re better off just walking away.
These situations crop up regularly in spheres as diverse as politics, romance and business. The 1987 movie “Ishtar,” for example, starring Warren Beatty and Dustin Hoffman, is still being studied for the decision-making process that created a cinematic flop that was five months late and about 80 percent over budget.
But it is rare that the theory of escalation of commitment can be defined as neatly as in Professor Shubik’s auction game, which is why a self-described “entertainment shopping” site, Swoopo.com, is so fascinating. Here, the theory is employed in some diabolically inventive ways.
Swoopo sells new merchandise using unusual auction formats. Let’s concentrate on one of them, the so-called penny auction.
Typically an item — say, a laptop that retails for $1,500, is offered for sale. The bidding starts at a penny, and goes up in one-cent increments, but it costs bidders 60 cents to make a bid. Each auction has a scheduled closing time, but as the deadline nears, that time is extended by 20 seconds whenever someone bids.
The site’s home page displays several attractive objects for sale with closing times fast approaching. It is mesmerizing.
One winning strategy might seem to be this: Bid at the last second, just before an auction is about to end. To “help” you do so, the site offers an automatic bidding program called a Bid Butler that allows you to make bids in the last 10 seconds. Alas, others can also use this automatic program, and you soon discover that just as the clock is ticking down and you’re about to make your big score, a bunch of other Bid Butlers get busy, the price jumps by a few cents, and the clock adds more time. Items can remain “in their final seconds” for days.
What makes this procedure so devilish is that while bidders are looking at what seem to be amazing bargains, the Web site is raking in the money. Because Swoopo collects 60 cents for each penny bid, its revenue is the selling price multiplied by 60. This means that if a computer you covet sells for $100, seemingly a bargain, Swoopo collects $6,000 in revenue, a very juicy profit.
Swoopo has even sold cash using this format — specifically, checks for $1,000. My colleague Emir Kamenica and I looked at 26 such auctions we found in a data set posted on the Swoopo Web site. For each of these, the average revenue to Swoopo was $2,452. Winning bidders also did well: Of the winners, all but two made money even after accounting for the cost of their bids, with an average profit of $658. Still, the important point to remember is that, collectively, bidders are losing money. Only the lucky last bidder is a winner.
It’s no surprise that Swoopo has attracted controversy. Bloggers have called it “evil,” but the company says that its winners save substantially since the winning bids are always well below the retail price of the product and that everyone else gets an exciting experience, which is worth the price of admission.
Frank Han, the North America general manager of Entertainment Shopping, Swoopo’s parent company in Germany, said, “The thrill of winning one of the auctions that ends at a very low price is what keeps customers coming back.”
AND some lucky bidders do get bargains. The site’s list of completed auctions includes a laptop that sold for $23.27, a video game for $5.88 and a microwave oven for $60.96, prices that don’t include bidding fees.
How much danger does Swoopo pose to consumers? Your view of that may depend partly on whether you think people are playing for fun or merchandise. If they are looking for the thrill of chance, it’s hard to argue that Swoopo.com doesn’t have as much right to life as slot machines or state lotteries.
But if people are looking for a good deal, the right comparison isn’t to gambling, but to a discount retailer. The difference between Swoopo and Best Buy is that at Swoopo you end up paying for stuff in the other guy’s shopping cart.
What is Swoopo.com’s future? Of course, it’s possible that competition will eventually drive down its profit margin. Someone may offer to sell bids for only 30 cents, inducing Swoopo to drop its price to a quarter, and so forth. But so far that hasn’t happened.
Swoopo has added a feature allowing you to apply bidding fees from an auction toward buying the product outright, at a retail price listed on the site. If you don’t want to buy that item at that price, you’ve lost your bidding fees.
Market competition, meanwhile, seems to be proliferating the auction strategy rather than undermining it. A start-up called Ibidcondo.com charges $100 for a seat at an auction in which a condo is sold. For a condo worth $280,000, it sells as many as 2,800 seats. Do the math. Caveat emptor.
Richard H. Thaler is a professor of economics and behavioral science at the Booth School of Business at the University of Chicago.