by Robert Litan
Google uses auctions for other parts of its business, such as allocating servers between its business units. Most famously, the company used an auction process when it decided to offer some shares to the public in 2004 rather than going the traditional route of hiring an investment bank to sell shares and set the price. The particular process Google used for its initial public offering (IPO) is called a Dutch auction. The auctioneers (the investment banks that carried out the auction and that were eager for the business) accepted bids for both quantity and price, and then determined the highest price at which all of the shares being offered would be sold. Take, for example, this hypothetical. The company wants to sell 2,500 shares: One bid comes in at 1,000 shares at $104, another for 800 shares at $103, still another at 700 shares at $102, and so on. The clearing price is the price at which the sum of the quantities bid equals the shares put up for sale at the price at which this occurs, which in this hypothetical would be $102 or lower.
Companies using this method do not have to use the exact clearing price, however, in order to help ensure a “pop”—a jump in the stock price above the IPO price, shortly after the IPO—for investors who initially buy the stock. Having a pop reduces the chances that some initial purchasers will welch on their commitments to follow through with the stock purchases (they may be disinclined to do that if the stock price trades below the IPO price shortly thereafter, meaning that if the purchases were honored, the buyers would immediately suffer loss). Google adopted a variation of a pure Dutch auction in large part for this reason, according to Varian (who advised on the structure of the auction, along with other leading economists he brought in as consultants on the matter).
A natural question arises: Why don’t all companies going public use an auction like Google did to manage their IPO, rather than continue to use investment banks to set the IPO price, as is still routinely done? One answer to this question is that few private firms have the kind of public recognition before their IPOs that Google has. As a result, companies going public may feel the need for investment banks to drum up demand for their stock before the offering actually takes place. In other cases, well-known private companies may want an investment bank to set a high price to effectively cash in on the bank’s fame or that of the firm’s founder. A good example is Facebook, which was one of the world’s best-known privately held companies when it went public in 2012 but which did not use an auction to set its IPO price. That may have turned out just fine for Facebook since its stock price fell for some substantial period of time after the IPO, leaving investors upset, but it allowed Facebook’s owners to sell fewer shares to raise the money they were seeking than would have been the case had the initial price been lower.17 (Within a year of the IPO, Facebook’s stock price had more than recovered from its initial post-IPO decline.) Of course, we will never know if Facebook could have fetched an even higher price had it used an auction to set its IPO price.
Another reason why some private companies may prefer an investment bank to set the stock price is that the bank also buys all the shares before they are resold to the public. This gives companies going public certainty about how much money their offering will raise, rather than risk taking in a lower amount if the stock were sold through an auction. Whether investment banks will continue to dominate the IPO process, or whether more firms going public will turn to auctions to set their opening prices, may not be decided for some time.
Getting back to Google, Hal Varian’s life certainly changed after his meeting with Eric Schmidt. Several years after assuring Schmidt about the wisdom of using the Vickery auction, Varian left his tenured job as professor of economics and dean of Berkeley’s School of Information Sciences to become Google’s full-time chief economist. Varian has hired a team of statisticians and econometricians (a unique type of economist I discuss further in Chapter 5) who have applied economic and statistical ideas in a wide range of areas in the company. One well-known success is Google Trends, which allows users to graph over time the number of times particular search terms have been requested. This tool has been adapted by Google to show trends in the flu in different geographic areas, before official government warnings, or can be used by anyone to predict other variables, such as the unemployment rate by looking at trends in search words or phrases like “unemployment claims.”
Varian is modest in claiming parentage for other ideas for which he and his team have been instrumental at Google. Upon some coaxing, he volunteered to me a few examples (I am sure there are more, but Hal is not the kind of person to toot his own horn). Among the various innovations at Google to which he contributed were the occasional use of A/B testing (a subject I discuss in greater detail in Chapter 6), and the introduction of quality assurance techniques to assure that Google’s new products do not have bugs. He also has brought other well-known economists to consult with the company, mostly those with research experience relating to auctions. And he has been a pioneer in developing or overseeing the development of various techniques for discerning patterns in very large bodies of data, or “Big Data,” a subject we explore in several later chapters of this book.18
Other Uses of Auction-Based Pricing
Google has not been the only pioneer of auctions on the Internet. Another leading success story is eBay, which has become famous for its auctions, too. eBay’s history is amazing in a different way from Google’s because the former’s success does not rest on having unique (and secret) software, or in using an alternative to the conventional high-price auction, but rather in using the power of the Internet to make virtual auctions economic in a way that physical auctions never could. I will not discuss eBay further, despite its huge success and importance, because there is no unique economic idea to credit for the company’s success, nor do I have any evidence that any of the sellers on eBay (some of whom have become commercially successful largely by taking advantage of eBay’s auction platform) or buyers use economists to assist with their bidding strategies. Many of the buyers on eBay, for example, have learned through experience that it can pay to wait until the last minute to submit a winning bid, a strategy some economists might recommend but one that anyone can figure out.
However, there are instances where economists have played important roles in advising participants in auctions or those operating an auction platform. For example, in 2008, economists helped design and manage a then-controversial auction platform for setting benchmark prices for wholesale milk. The market has since taken off, and celebrated its fifth-year anniversary in July 2013.19 Likewise, economists also helped Ocean Spray establish a quarterly auction to sell its intermediate product, cranberry concentrate.20
And then there is the remarkable case of Stanford economist Susan Athey, the first woman to win the coveted Clark Medal awarded (now each year) to the economist under 40 doing the most promising work in the field.21 Athey has conducted research on a wide range of microeconomic topics, beginning with timber auctions as an undergraduate at Duke, which led to later work establishing the virtues of sealed bids to help avoid collusion among bidders for government contracts. But what got her attention from the business community—Microsoft’s former and longtime President Steve Ballmer in particular—was her research on the digital advertising market. Ballmer reportedly was especially interested in her view that the market required more than just one competitor (Google), and so he summoned her to his Redmond office, where eventually Athey spent a full year, and has continued ever since as Microsoft’s chief economist while also teaching at Stanford.
Like many economists, Athey has strong mathematical training, supplemented by computer science, both of which she studied in college. She also is a prodigy, having entered college at the age of 16 and earning her PhD at Stanford at the young age of 24. Her doctoral dissertation, which suggested a new way to model uncertainty, helped her become one of the most highly sought after young faculty members in the country at the time. She chose to begin her research and teaching career at MIT, moving to Harvard, and then later
returning to Stanford.
In Chapters 6 and 11, you will meet other economists who have wrestled with the intricacies of auctions in different contexts, where the government has either mandated the use of auctions (for electricity in some states) or is an active seller itself of something the private sector very much desires (like the scarce electromagnetic spectrum used for telecommunications or U.S. Treasury bonds). These auctions can be quite complicated, as can be the strategies bidders use for achieving successful outcomes. It will not surprise you to learn that economists have played important roles in these situations.
Conditional Offer of Purchase (Name Your Price): Priceline
One of the fascinating things about the Internet is that it has created opportunities for not only hundreds of thousands, if not millions, of new businesses, but also new business models. Priceline, the first website allowing customers to name their price to potential suppliers of travel services, is an example of such a new model.
Priceline was born out of a series of brainstorming sessions convened by an already highly successful pre-Internet entrepreneur, Jay Walker. Walker made his initial smaller fortune by founding the Synapse Group, a company that processed magazine subscriptions using the credit card network. He sold a controlling interest in the company in 2001 to Time Warner for over $600 million.22
Walker was not content to sit on his success, however, especially once the Internet became more than an academic curiosity in the early 1990s. Walker could see in broad terms that the Internet would fundamentally transform business, but he wasn’t sure exactly how, so he assembled a small group of his associates, including software engineer Scott Case. The team eventually came up with the business that became Priceline, where Case became its chief technology officer.23
Like many entrepreneurs, Walker was looking not only for something different and scalable, but an idea that could be patented so the company would have intellectual property to protect. His opening was the Supreme Court’s decision in 1988 in State Street Bank v. Signature Financial Group, making clear the patents were available for business methods (although that principle had been established by the Supreme Court as early as 1790, the U.S. Patent Office rejected the theory during the computer age until the Supreme Court resurrected it). Amazon would later get a patent for its one-click button that enables customers to purchase online. Although many academic scholars have since argued that patents should not be issued for business methods because they can inhibit innovation if not easily worked around (as Barnes & Noble and other companies did by adding a second click to get around Amazon’s one click), business method patents are still available, although the courts have made them more difficult to get after the Supreme Court ruled in Kappos v. Bilski that a business method patent was not appropriate for a tax-efficient method of hedging commodity price risks (even where a mathematical formula is involved).
The business method that Walker and his team came up with, and soon patented, was the conditional purchase offer, and they built Priceline around it. Here’s how they came up with the idea.
Any unsold seats on a given flight or unbooked hotel rooms for any given day disappear after the flight takes off or the day is over. As Walker explains it, he and his team brainstormed how to solve three interrelated problems confronting all sellers of perishable merchandise, of which airline travel and the hotel business are prime examples:24
How can sellers discount the seats or the rooms (or any perishable item, for that matter) and attract new buyers without encouraging or allowing other buyers who are willing to pay full price from also taking advantage of the discount?
How can regular, full-price customers be discouraged from delaying their purchases in order to receive last-minute discounted fares or prices?
Because sellers cannot see the demand curve for their products (how many buyers there are at each price), they cannot discover it without lowering their prices in a way that cuts into their profits.
The conditional price offer solves all of these problems by initially offering a more restricted version of airline tickets and later other travel services, and asking customers to submit conditional offers—not bids. The offers had to be binding and Priceline assured they were by requiring those making the offers to give their credit card information at the time of making the offer. The conditional price offers had no effect on full price or regular customers, nor did these customers have any incentive to submit last-minute low-price offers for what, in essence, were somewhat impaired versions of the normal service: flights whose times or departure airports (within a given radius of the customer’s departure location) only Priceline could control.
Walker recounts that he had difficulty at the beginning persuading airlines to participate. Each feared that selling seats at cheaper prices, even at the cost of not selling any at all, would somehow dilute its brand. TWA, which is no longer flying, thought differently and was the single exception. Eventually, others followed suit.
Walker and his cofounders realized that this business model would appeal only to leisure travelers (and perhaps a few business travelers really trying to save money) who cared only about price and were flexible on other terms. In more technical terms, Priceline was targeting the most price-sensitive customers on the demand curve for travel, and ignoring the rest (thus engaging in a variation of Ramsey pricing, which I describe below, but with the twist that customers rather than vendors set the price).
The rest, as they say, is history. Using a series of clever television ads, starring William Shatner from Star Trek, Priceline quickly became a phenomenon in the late 1990s, had some growing pains after the Internet bust, but has gone on to become one of the leading travel booking sites on the web. Walker left the company long before all of this success was achieved, moving on to patent inventions relating to vending machines and inventing and manufacturing casino games.
So where, you may ask, is the economist or the economic idea in the Priceline story? The answer is Walker himself, who was trained in college, at Cornell University, in its well-known industrial and labor relations program, which essentially is all about microeconomics. In other words, Walker studied in school the fundamentals of demand curve analysis that eventually he would one day apply to help build a major Internet-based company resting on a unique business model. Walker’s story proves that one of the best ways of transferring economic ideas from the Ivory Tower to business is through the entrepreneurial efforts of single individuals trained in economics.
Different Prices for Different Folks
It doesn’t take an economist to tell you that different people are willing to pay different prices for the same item or service. That is because people have different preferences. The demand curve you see in standard economics textbooks is downward sloping, meaning that as prices fall, consumers want more units. This reflects the fact that consumers (or businesses needing supplies) at the very top of the curve are willing to pay much more, and are generally less sensitive to variations in prices, than those at the bottom at the curve.
Many different economic ideas have flowed from this very simple insight, and a number of them have found their way into commercial practice. Let’s begin with perhaps the simplest case, a situation where there is only one seller, or a monopolist. In the 1920s, a young math prodigy in the United Kingdom named Frank Ramsey (see following box) formally proved that such a firm would maximize its profits if it could charge higher prices to those least sensitive to price changes (in formal terms, those whose demands are price-inelastic) than to customers who were more price sensitive (in economic jargon, those with price-elastic demand). Ramsey formalized his finding by proving that the optimal strategy for a monopoly firm is to set a markup over variable cost that is inversely proportional to the consumer’s elasticity of demand (sorry, I couldn’t resist, but in plain English it’s the formal equivalent of everything in the paragraph that precedes it).
Now, in the real world, the prices that monopolies like public utilities charge are limited
by regulation precisely because consumers have no other choices. So for practical purposes, Ramsey’s proof is of most direct use to regulators rather than the firms themselves, though firms that earn their monopolies fair and square, through hard work, luck, or foresight, or by owning a lawfully granted patent can maximize their profits following his strategy. The major objection to Ramsey pricing is on grounds of equity, because some consumers who highly value the product of the monopoly, say electricity, may also have modest means, and so many believe it is unfair to charge them more than well-heeled consumers.
Frank Ramsey: The Life of a Genius Cut Short
Frank Ramsey’s life is interesting, but tragic.25 Ramsey was born in 1903 to an accomplished father, Arthur Ramsey, who was a mathematician and president of Magdalene College at Cambridge University in England. His mother, Mary Agnes Stanley, raised four children.
Ramsey demonstrated intellectual prowess at a very young age, showing wide-ranging interest in literature and mathematics. He was said to have suffered from mild depression and developed an intellectual interest in psychoanalysis, even to the point of being psychoanalyzed by a disciple of Sigmund Freud.
Following in his father’s footsteps, Ramsey began his studies at Cambridge at the age 16. While there, he attracted the attention of Keynes and Arthur Pigou, another of Cambridge’s leading economists at the time, both of whom urged Ramsey to turn his mathematical mind to economics.
He did so with relish, producing a number of papers, the most famous of which led to the theorem about monopoly price discrimination developed as a by-product of a seminal paper on taxation.26 The pricing theorem is now widely known as Ramsey pricing in his honor. Keynes called another of Ramsey’s papers about economic growth “one of the most remarkable contributions to mathematical economics ever made.”27
