14

Behavioral Economics and Pricing

The foundation for the standard model of price response is the assumption that consumers are rational utility-maximizing agents. Before making a purchase, a rational consumer gathers information and then systematically and dispassionately evaluates her alternatives. A rational consumer has a willingness to pay (which may be less than zero) for every item on offer to her by every seller. She chooses the set of purchases that maximizes her expected utility (measured in monetary units) from the purchases minus the total amount she must pay. Variation of willingness to pay among customers generates a price-response function as described in Chapter 3. In a similar vein, neoclassical economics assumes that firms choose among alternatives to maximize expected profit; uncertainty on the part of sellers about how different customers value their product offerings relative to the competition gives rise to the bid-response function described in Section 13.3.

The idea of rationality is based on a handful of plausible axioms about how consumers and firms behave, and it provides an intuitive and compelling basis for price optimization. However, it is not difficult to find examples of behaviors that are not consistent with the assumption of consumer rationality. Consider the following two examples.

Coca-Cola makes customers hot. In a 1999 interview with the Brazilian magazine Veja, Douglas Ivester, chairman and chief executive officer of the Coca-Cola Company, announced that his company was testing a new vending machine that would change the price of soft drinks in response to outside temperature. When it was hot outside, a cold can of Coke would cost more than when it was cold outside. As he explained:

Coca-Cola is a product whose utility varies from moment to moment. In a final summer championship, when people meet in a stadium to enjoy themselves, the utility of a chilled Coca-Cola is very high.¹ So it is fair it should be more expensive. The machine will simply make this process automatic. (Martinson 1999)

While Douglas Ivester might have thought this was fair, others emphatically did not agree. The San Francisco Chronicle (1999) called the idea “a cynical ploy to exploit the thirst of faithful customers,” the Honolulu Star-Bulletin labeled it “a lunk-headed idea” (Memminger 1999), and the Miami Herald referred to Coca-Cola management as a bunch of “soda jerks” (Hays 1999, 18).

Coke began to backpedal immediately. “We are not introducing vending machines that raise the price of soft drinks in hot weather,” said spokesperson Ben Deutsch (Reich 1999). Although Coca-Cola is rumored to have tested weather-sensitive machines in Japan, it has never employed them in the United States. And the temperature-sensitive vending machine gaffe is widely considered to be one of the factors contributing to Ivester’s losing his job.

Amazon flubs DVD pricing. One night in August 2000, a customer ordered from Amazon the DVD of Julie Taymor’s movie Titus, paying $24.49. Browsing on Amazon a few nights later, he found that the price had jumped to $26.24. As an experiment, he deleted the cookies on his computer that identified him to Amazon as a regular customer. Revisiting the Amazon website, he was quoted a price of $22.74 for the same DVD (Streitfeld 2000).

This customer happened to be a regular purchaser of DVDs from Amazon, and he came to the reasonable conclusion that Amazon was charging higher prices to its most loyal customers. When he recounted his experience with Amazon on the website DVDTalk.com, the reaction was immediate and overwhelmingly negative. Comments posted on DVDTalk. com over the next weeks included:

• “Amazon is over in my book.”

• “I will never buy another thing from those guys!”

• “I am so offended by what they did that I’ll never buy another DVD from them again.”

• “Amazon is suck [sic].”

The tale spread via email and chat rooms, and Computerworld published a story about Amazon’s pricing policies. The national press picked up the story. Press coverage generally jumped to the conclusion that Amazon was experimenting with “dynamic pricing,” by which they would charge different customers different prices. As one reporter commented, “Dynamic pricing . . . involves putting customers into groups based on how price sensitive they appear to be. If someone accepts the $10-higher price, say, three times in a row, maybe that’s all they will ever get in the future” (Coursey 2000).² While a few reporters noted that dynamic pricing was hardly unknown in the offline world, the response to Amazon’s action was almost universally negative.

In response to the negative publicity, Amazon apologized and claimed that it was merely running a price-sensitivity test: “It was done to determine consumer responses to different discount levels,” said Amazon spokesman Bill Curry. “This was a pure and simple price test. This was not dynamic pricing. We don’t do that and have no plans ever to do that” (Streitfeld 2000, A1).

What happened? In both cases, consumer reaction did not follow the standard model of rational economic decision making. After all, a rational consumer should care only about her price—not the prices charged to other customers or what the company might have charged under different conditions. In the cases of Amazon and Coke, apparently sensible pricing schemes were abandoned as a result of consumer reaction. Clearly, the consumers quoted in these stories are not automatons comparing price to willingness to pay and dispassionately deciding whether or not to buy. Rather, they are human beings reacting emotionally to such seemingly irrelevant issues as the prices other people might be offered or how prices might change with the weather.

The topic of this chapter is how consumers react to prices in ways that deviate from the assumptions of rationality—as in the Amazon and Coca-Cola examples above—and the implications of these deviations for pricing and revenue optimization. Although the basic model of customer price response is quite robust, there are many types of behavior that it cannot represent. In particular, customers do not view buying and selling dispassionately as simple transactions among consenting adults. Rather, they often react to prices in ways that economists consider irrational.³ Customer response to price is not always based only on the price and product being offered. Instead, a number of other, seemingly irrelevant, factors can influence buying behavior, including these:

• How the price is presented and packaged

• How much profit the customer believes the merchant will realize

• How this price compares to the customer’s expectation of what the price should be

• Prices the customer believes are being charged to other customers

All of these factors can influence consumer behavior, but none of them can be easily accommodated within the standard model in which willingness to pay and price alone govern whether or not a customer will purchase.

The study of deviations from rationality and their implications is called behavioral economics. Many behavioral studies have found effects of pricing that might be hard to use in the real world—for example, that customers are less sensitive to prices when the cents match their birthday (Coulter and Grewal 2014) or the final digits are associated with a favorite sports team, such as $.49 for a fan of the San Francisco 49ers (Husemann-Kopetzky and Köcher 2017). While phenomena such as these may have little practical significance, three important and well-established categories of consumer irrationality are important for pricing and we examine them at length.

The first category consists of violations of the law of demand—the principle introduced in Chapter 3 that specifies that, all else being equal, an increase in price should lead to decreased demand, and a decrease in price should increase demand. However, there are situations in which the opposite effect might occur—raising price might lead to increased demand, and lowering price might decrease demand.

The second category consists of presentation issues in which customers react not only to a price but also to the way in which it is communicated. One well-known example of a price presentation issue is the digit-ending phenomenon—buyers in Europe and North America tend to overweight the leading (leftmost) digits when comparing prices so that, in immediate impression, $19.99 seems to be a greater savings from $20.00 than $19.93 from $19.99. Because of this phenomenon, certain penny-endings such as 9 and 5 are far more common than others.⁴

The third category of irrational price response involves the issue of fairness. The belief that one is entitled to fair treatment is deeply ingrained in most people. Perceptions of unfair treatment can trigger strong emotional responses such as those expressed in the DVDtalk.com postings about Amazon.

These three categories all influence consumer decisions, but they manifest themselves in different ways. A company that is doing a poor job at price presentation will simply sell less than it might otherwise, but it is unlikely to hear anything from its customers. After all, customers are unlikely to complain if a merchant’s prices end in $.83 rather than $.99. Similarly, customers are unlikely to complain if they think that an item is priced too low for them. In contrast, customers are likely to complain—and loudly—if their sense of fairness is violated. Poor price presentation will probably not land you on the front page of the newspaper, but pricing considered unfair might—just ask Coca-Cola and Amazon.

14.1 VIOLATIONS OF THE LAW OF DEMAND

The law of demand specifies that demand for a good decreases as its price is increased and that demand increases as price is decreased. This corresponds not only to intuition but also to the fact that the price-response function is derived from the distribution of willingness to pay across a population. If this is the case, then raising prices should always lower demand and vice versa. Therefore, any violations of the law of demand must arise from situations in which the willingness-to-pay assumption does not hold. Three kinds of cases have been identified in which individual purchase behavior is not well described by a maximum willingness to pay, and therefore the law of demand might be violated.

14.1.1 Giffen Goods

Economic theory allows for the possibility of Giffen goods, whose demand rises as their price rises because of substitution effects. An example might be the case of a student on a strict budget of $8.00 per week for dinner. When hamburger costs $1.00 per serving and steak costs $2.00, she eats hamburger six times a week and treats herself to steak once a week. If the price of hamburger rises to $1.10, she stops buying steak and buys hamburger seven times a week to stay within her budget. In this case, a rise in the price of hamburger causes her consumption of hamburger to increase. While this behavior might conceivably occur at an individual level, a Giffen good requires that enough buyers act this way that they overwhelm other buyers who would buy less hamburger as the price rises. Giffen goods are rarely, if ever, encountered in reality—in fact, many economists doubt whether they have ever existed in the real world.⁵

14.1.2 Price as an Indicator of Quality

In some cases, price may be used by some consumers as an indicator of quality: higher prices signal higher quality. In this case, lowering the price for a product may lead consumers to believe that it is of lower quality, and demand could drop as a result. Typically, markets in which this is an issue have a large number of alternatives, and some buyers who do not have the time or resources to research the relative quality of all the alternatives instead use price as a proxy. Wine is a classic example: Faced with a daunting array of labels and varietals, many purchasers are likely to use a rule such as a $10 bottle for dinner with the family, a $15 bottle if the couple next door is dropping by, and a $25 bottle if our wine-snob friends are joining us for dinner.

Figure 14.1 Price as signal of quality.

Using price as an indicator of quality is not necessarily an irrational strategy on the part of a consumer, particularly when quality is correlated with price. Consider a customer who is not highly knowledgeable about wine but has consumed a number of different California Chardonnays at different prices. Each dot in Figure 14.1 represents the utility that she enjoyed (denominated in dollars) from drinking a Chardonnay at that price point. Clearly her enjoyment of Chardonnay is correlated with its price, but the correlation is not perfect. In addition, her enjoyment demonstrates decreasing marginal utility—spending more at higher prices delivers less marginal utility than at lower prices. The solid curve shows her expected utility from purchasing a bottle as a function of price based on her experience. Assume that she is at a wine store shopping for a bottle and is unfamiliar with any of the brands on offer. With only price to guide her, it is perfectly rational for her to purchase at the price point that maximizes her surplus as defined by the difference between her expected utility and the price—which, in this case of the customer in Figure 14.1, occurs at about $20. If enough customers share similar experiences, a Chardonnay might well sell more bottles at $20 than at $12.

The price-as-an-indicator-of-quality effect can be particularly important when a new product enters the market. A medical-product company developed a way of producing a home testing device at a cost 75% below the cost of the prevailing technology. It introduced the new product at a list price 60% lower than the list price of the leading competitors, expecting to dominate the market. When sales were slow, the company repackaged the product and sold it at a price only 20% lower than the leading competitor. This time, sales took off. The company’s belief is that the initial rock-bottom price induced customers to believe its product was inferior and unreliable. The higher price was high enough not to raise quality concerns but low enough to drive high sales.

More recent studies have shown that not only may price be a signal of quality, but the price itself may influence customer enjoyment. In one study, the price of an energy drink influenced the ability of purchasers to solve puzzles after consuming the drink (Shiv, Camon, and Ariely 2005). And in a widely publicized study, researchers performed a brain scan while subjects drank wine. They found that subjects reported enjoying a wine more if they thought that it was priced higher. Perhaps more surprisingly, the brain scan showed that they apparently did enjoy the same wine more if they thought it cost more—regions of the brain associated with pleasure were more highly activated when a subject consumed what they believed was a high-priced wine than when they consumed the same wine but believed it was low-priced (Plassmann et al. 2008).

14.1.3 Conspicuous Consumption

The sociologist Thorstein Veblen coined the term conspicuous consumption for the situation in which a consumer makes a purchase decision in order to advertise his ability to spend large amounts. As he put it, “Conspicuous consumption of valuable goods is a means of reputability to the gentleman of leisure” (Veblen 1899, 57). Thus, it is likely that a large segment of the purchasers of Bentley Continental GT automobiles (advertised as starting at $202,500) and Louis Roederer Cristal Brut Champagne (about $200 –$3,000 per bottle) are not necessarily purchasing entirely based on their willingness to pay for qualities of the car and the taste of the champagne, respectively, but are also making a statement about their ability to afford such luxuries. For a product that has been successfully marketed as an expensive luxury product, there may be those who will purchase it in part because it has a high price—and others know it. For such products, maintaining the air of exclusivity will influence marketing and pricing—for example, many luxury brands will not allow their products to be sold through the usual channels at a discount, even though it might be profitable to do so, because of the potential damage to the brand. Instead, they will sell through alternative channels, such as outlet stores, or ship to other countries to sell or even destroy product to maintain the image of high price and exclusivity.

14.2 PRICE PRESENTATION AND FRAMING

Consider the following three scenarios:

1. You walk up to the box office to buy two tickets to a movie. A sign in the window reads:

General Admission:	$15.00
Senior Citizen (65 and over):	$13.00
Youth (18 and under):	$13.00

2. You walk up to the box office and the sign in the window reads:

General Admission:	$13.00
Middle Aged (over 18 and under 65):	$15.00

3. You walk up to the box office and the sign in the window reads:

General Admission:	$13.00
Middle Aged (over 18 and under 65):	$2.00 surcharge

Chances are that most people would consider the first scenario to be unremarkable. However, the second scenario would likely raise eyebrows. The effect is likely to be even stronger under the third scenario. In fact, you can almost hear the cries of “Why should I be punished because of my age? It’s not fair!” Yet, of course, all three price schemes are exactly the same, at least in the sense that everyone would pay the same amount for admission under each scheme. However, people perceive them and react quite differently. The most widely accepted reason for why this is so is provided by prospect theory.

14.2.1 Prospect Theory

Prospect theory was introduced by Daniel Kahneman and Amos Tversky in 1979 based on their observation that people do not evaluate opportunities based on a strict evaluation of expected costs and benefits. Rather, they use various mental shortcuts and rules of thumb to make decisions. This can lead to outcomes that are inconsistent with economic rationality. For example, prospect theory proposes that people tend to experience losses much more intensely than they value gains: this is called asymmetrical treatment of gains and losses. This idea is illustrated in Figure 14.2. Customers value a gain of Δ much less than they feel the pain of a loss of equal value. Since paying a higher price is viewed as a loss, this means that a surcharge (or price rise) of $5.00 is viewed much more negatively than a discount (or price drop) of $5.00 is viewed positively. In short, price rises are felt more intensely than price drops.

An important implication of prospect theory for price presentation is that the presentation of prices can establish the baseline against which the actual price is compared. As an example, consider the following scenario.

Figure 14.2 Asymmetrical treatment of gains and losses.

• Station A sells gasoline for $2.60 per gallon but gives a discount of $0.10 per gallon if you pay with cash.

• Station B sells gasoline for $2.50 per gallon but charges a surcharge of $0.10 per gallon if you pay with a credit card.

The two approaches are identical in terms of what any particular buyer will pay. Yet Station B’s approach is never seen in the real world. Why? According to prospect theory, the reason is immediate from Figure 14.2. For each station, the list price establishes the baseline price in the buyer’s mind. For Station A the $0.10 discount (+Δ) is viewed as a gain relative to the baseline price of $2.60, while for Station B the $0.10 surcharge (–Δ) is viewed as a loss relative to the baseline price of $2.50. Prospect theory predicts that a buyer using a credit card would be more likely to patronize Station A than accept the surcharge associated with Station B.

For each of the three movie-pricing scenarios, the general admission price establishes the baseline against which the price that a customer will actually be charged is compared. When the actual price includes a discount, it is viewed as a gain, which is viewed favorably. When the actual price includes a surcharge, it is viewed as a loss, which is felt negatively. Asymmetric treatment of gains and losses also explains why the middle-age-surcharge approach to movie pricing seems so jarring—and is never found in the real world.

Asymmetrical treatment of losses and gains may explain some of the negative reaction to Coca-Cola’s announcement of temperature-dependent vending machine pricing. Note the wording of Douglas Ivester’s announcement: “In a final summer championship, when people meet in a stadium to enjoy themselves, the utility of a chilled Coca-Cola is very high. So it is fair that it should be more expensive.” From a prospect theory point of view, he chose the worst price presentation possible—he framed the scheme in terms of a surcharge relative to a baseline price. This framed the higher price as a loss and triggered an immediate negative reaction. Coca-Cola would probably have fared better if Ivester had said something like: “On a cold day, the utility of a chilled Coca-Cola is not as high as on a hot day. So it is only fair that it should cost less—our customers deserve a price break in that situation.” In fact, when faced with the negative reaction to the original suggestion, the company tried to reframe the description in just those terms by insisting that “its machines would be more likely to lower prices during off-peak hours in offices, for instance” (Martinson 1999).

Prospect theory provides an explanation for a nearly universal phenomenon in pricing—discounts are common, and surcharges are rare. As Chapter 12 notes, almost 80% of apparel sold in the United States in 2006 was sold at a discount from list price (Levy et al. 2007). It is safe to assume that only a tiny fraction (if any) of retail apparel sales took place at a surcharge from list price. Clearly, sellers believe that promotions and markdowns stimulate sales more than is explainable simply by the lower price. If the absolute price was all that mattered, sellers would be indifferent between running a sale and lowering list price. Rational consumers would simply compare their willingness to pay against the current price and decide whether to purchase. It would not matter to a rational consumer if the price she was paying was the list price or a sale price.

Figure 14.3 Two Amazon price listings showing discounts from a “List Price” (left) and from a “Was Price” (right). Source: Amazon.com.

But both brick-and-mortar retailers and online retailers like to present prices as discounts from higher standard prices. As an example, Figure 14.3 shows the prices of two different popular toys displayed on Amazon.com. The Pull and Sing Puppy on the right has a price of $9.99 presented with a “Was Price” of $12.97, while the Take Apart Racing Car on the left has a price of $15.99 presented with a “List Price” of $26.65. In both cases, the selling price is presented in combination with the List or Was price and the corresponding savings. The List and Was prices establish a higher benchmark price compared to which the actual price can be evaluated as a gain.

The practice has become sufficiently ubiquitous that some customers have become skeptical about the validity of the list price. A New York Times reporter wrote:

Le Creuset’s iron-handled skillet, 11 3/4 inches wide and cherry in color. Amazon said late last week that it would knock $60 off the $260 list price to sell the skillet for $200. . . . The suggested price for the skillet at Williams-Sonoma.com is $285 but customers can buy it for $200. At AllModern.com, the list price is $250 but its sales price is $200. At Cutleryand-More.com, the list price is $285 and the sales price is $200. An additional 15 or so on-line retailers—some hosted by Amazon, others on Google Shopping—charge $200. On Le Creuset’s own site, it sells the pan at $200. (Streitfeld 2016)

Indeed, the power of perceived discounts is so strong that sellers are tempted to establish fictitious list prices against which they can pose their actual prices. A 1989 lawsuit held that the May Company “deceived customers by artificially inflating its so-called ‘original’ or ‘regular’ (reference) prices and then promoting discounts from these prices as bargains. This practice has become so rife in the U.S. retail environment that many consumers will only buy ‘on sale,’ and then are skeptical about whether they have really bought at a fair price” (Kaufmann, Ortmeyer, and Smith 1991, 117). While displaying fictitious list prices is clearly deceptive (and likely illegal), whenever a seller is legitimately selling an item at a discount, it is extremely wise to make sure that customers are made aware of the fact.

14.2.2 Reference Prices

Kent Monroe (1973) introduced the idea that buyers shop with a reference price in mind that they consider fair. A price close to the reference price will be considered reasonable, and a consumer is likely to purchase if the selling price is close to her reference price. Prices much higher than the reference price may even trigger anger—“$30 for a hamburger! That’s ridiculous!”

While it is good to be priced below a customer’s reference price, a price much lower than the reference price may cause suspicion that the item for sale is inferior or damaged. As an example, a start-up invented a process that could produce paper strips for use by diabetics to measure their blood sugar at a fraction of the cost of previous methods. Historically, these strips had sold for about $2.00 each. With the new process, the company decided to sell its strips (which were of equal quality to those already on the market) for $0.50 each. To its surprise, the strips did not sell well. However, when the company raised the price to $1.75 each, sales skyrocketed. The historical selling price had become the reference price in consumers’ minds and the substantially lower price was interpreted as meaning that the strips were of lower quality—a major consideration to customers whose health depended on accuracy of measurement.

The reference price is not the same as willingness to pay. Willingness to pay is theoretically a property only of the buyer’s utility for the item under consideration, which should be independent of current or past prices. On the other hand, reference price is a function of observed prices, past and present. In the example of the gasoline station discussed earlier, we assumed that the reference price was the list price. The use of “list prices” and “was prices” by retailers is clearly an attempt to influence consumers’ reference prices.

This raises the issue of how reference prices are formed. As noted by Markus Husemann-Kopetzky, “People continuously learn new price information from past purchases, from current offerings of various competitors, from prices of comparable products, from regular or list prices, and so on. When consumers consider this information, they might adjust their internal reference price in either direction” (2018, 111). How different consumers process different types of information to update their reference prices is a topic of ongoing research. A useful survey lists a number of factors that could influence reference price, including price history, promotion history, store-visit history, brand, whether the purchase was planned or unplanned, store type (everyday-low-pricing [EDLP] stores like Walmart versus “hi-lo” stores that cycle between list prices and discount prices), advertised prices, frequency of purchases, durability, and associated services (Mazumdar, Raj, and Sinha 2005). These factors carry different weights for different types of customers. Given this complexity and the heterogeneity of consumers, it is impossible to perfectly engineer the reference price for any single consumer or chosen group.

While perfect reference price engineering may be impossible, there are several practical implications of reference pricing. If a customer becomes used to seeing an item at a low price, she will expect the price to be low in the future and will be less likely to purchase if it is high. This can create a dilemma for the seller. Dropping the price on an item now may be optimal from a tactical point of view (that is, it may maximize short-run contribution), but it may have future repercussions if it lowers the reference prices of potential future customers. Several studies have shown that, if consumers utilize past prices in formulating their reference prices for future purchases, a cyclical hi-lo policy of periodic discounts can be optimal. It is also likely that if sellers ignore the effects of their pricing on reference price formation, then they will tend to underprice.⁶

14.2.3 Mental Budgeting

You are planning to go to a play by yourself one Sunday afternoon. Consider the following two scenarios.

• Scenario 1: You have already bought a $20 ticket for the play. Just after leaving your apartment, you open your wallet, and the ticket slides out and slips down the storm drain, where it is lost.

• Scenario 2: You do not have a ticket but plan to buy one at the box office. Just after leaving your apartment, you open your wallet and a $20 bill slips out of your wallet and is lost down the storm drain.

The question under both scenarios is: Would you go to the play and buy a $20 ticket at the box office? (Assume in both cases that you know that tickets will still be available for $20 when you arrive.) Kahneman and Tversky (1979) found that the percentage of people who would go to the box office and buy a ticket is much higher under the second scenario than the first. Of 250 college students presented with the second scenario, 187 said they would buy a ticket, while only 120 of 250 said they would in the first scenario. This is contrary to economic rationality. The wealth effect of losing a $20 bill and a $20 ticket are exactly the same. Why are people more willing to purchase a ticket in one case and not the other? The explanation, which has been supported by numerous experiments, is that people maintain mental budgets and evaluate expenditures against these budgets. Thus, someone might have a mental budget of $20 (or $30) for the day’s entertainment. The initial purchase of a $20 ticket is counted against this budget, so the purchase of a second ticket would exceed the mental budget. On the other hand, the loss of a $20 bill is not counted against the day’s entertainment budget (perhaps it is counted as a loss against general funds), so there is still sufficient entertainment budget to purchase a ticket.

Mental budgeting can have implications for how (and if) products and services should be bundled. If certain options, say, a top-of-the-line car stereo, are viewed as counting against a separate mental budget than the base product (the car itself), then it will generally be more effective to sell them separately.

14.2.4 Compromise Effects, Attraction Effects, and Decoys

Not only do customers use their prior price experiences in making current purchase decisions; they also use the prices of substitutes and variations in their purchase decisions, sometimes in surprising ways. A particularly striking example of this was described by Dan Ariely (2010). In a study, he offered students one of two different choice sets for subscribing to the Economist:

Choice Set 1	Choice Set 2
Internet subscription: $59	Internet subscription: $59
Print and internet subscription: $125	Print-only subscription: $125
	Print and internet subscription: $125

A total of 68% of the students presented with Choice Set 1 chose the internet subscription, while 32% chose the print-and-internet subscription. However, when presented with Choice Set 2, 84% chose the print-and-internet option, while 16% chose internet only and 0% chose print only. The addition of an irrelevant third option substantially changed customer choice. In this case, it changed the average sales price per transaction from $80.12 to $114.44—a 43% increase resulting from nothing more than the addition of an option that was of no interest to any customer.

In the example above, the print-only subscription in Choice Set 2 is an example of a decoy: an irrelevant product that is added primarily to boost the sales of another product—in this case the print and internet subscription—by making it look like a bargain in comparison. Needless to say, the fact that decoys can work by boosting the sales of other alternatives appears irrational—if consumers were fully rational, the addition of an irrelevant alternative should not influence their choice between two relevant alternatives. Yet, as the Economist example shows, this is exactly what happens.

Decoys can be considered as particularly extreme examples of the attraction effect—the fact that introducing a dominated alternative can increase relative sales of a high-end product. The idea is illustrated in Figure 14.4. Here, we assume two products in the marketplace at two levels of quality and at different prices represented by points A and B. Corresponding to customer expectations, the lower-quality product A is priced less than the higher-quality product B—which means that neither product dominates the other. Note that any product in the shaded region is dominated because a better product is available at a lower price. Product C₁ is a decoy that is dominated by product A, which is the same quality but lower price. Product C₁ is unlikely to generate much demand on its own, but offering it will move some demand from product B to product A. Similarly, product C₂ is a decoy that is dominated by product B. Again, product C₂ is unlikely to attract much demand itself, but it is likely to shift some demand from product A to product B.

Figure 14.4 The attraction effect and the compromise effect.

A related phenomenon is the compromise effect in which adding a superior product to an existing product’s market at a higher price will cause demand to shift to higher-price products. One example of the compromise effect is provided by Starbucks, which initially offered its drip coffee in three sizes: small (8 oz.), tall (12 oz.), and grande (16 oz.). In 1996, Starbucks introduced a fourth size, the 20-ounce venti. According to some sources, in addition to attracting sales itself, introduction of the venti led to a 10–15% increase in grande sales and a corresponding reduction in small and tall sales (Shartsis 2016). The increase in grande sales is an example of the compromise effect—the introduction of the superior alternative led to increased sales for the next best alternative. In Figure 14.4, the venti might be represented by point D—it is actually superior (i.e., larger) than the best existing alternative represented by point B. Because it is not dominated by any of the alternatives, it may attract some customers on its own merits, but it also will cause business to shift from point A to point B.

The compromise effect has been described as the tendency of customers to choose an intermediate option when they are uncertain about the relative quality of different options or their own strength of preference among them.⁷

14.3 FAIRNESS

The reaction to Amazon’s DVD pricing illustrates the emotional response that differential pricing can evoke. Amazon learned the hard way that differential pricing can be the third rail of pricing. However, the Amazon story does not mean that differential pricing has to be entirely ruled out. On the one hand, a reputation for unfair pricing may alienate current and future customers. On the other hand, as Chapter 6 shows, the potential profitability increase from differentiated prices is great enough that companies that can find acceptable ways to segment their customers and charge differential prices have a strong incentive to do so.

This section addresses the issue of fairness as it applies to customer reactions to pricing. Fairness in the sense discussed here—and the way that most people mean it when they cry, “That’s not fair!”—was not part of neoclassical economics. However, behavioral economics has begun to research how consumers form their perceptions of fairness and how these perceptions influence the choices that they make. This is still an active area of research, and many aspects of perceived fairness in pricing are not fully understood. However, it has been shown that customers evaluate the fairness of a price in at least two ways. The first is relative to the perceived profit being earned by the seller. The second is relative to the prices the customer believes other buyers are paying. We consider these two in order.

14.3.1 Dual Entitlement

The first place that fairness can enter into pricing is via the buyer’s perception of the seller’s profit. Most people would agree that “greedy” companies should not generate “obscene” profits by “gouging” customers with “unfairly high” prices. Most people could probably give examples of pricing they believe violated this principle—although they might be harder pressed to give an objective criterion of when profits become obscene or when a high price becomes gouging. Nonetheless, most customers believe that it is unfair for companies to charge too much for their goods and services—even if customers are willing to pay a high price for the item. This idea was formalized by Daniel Kahneman, Jack Knetsch, and Richard Thaler (1986b) as the concept of dual entitlement. Dual entitlement postulates that consumers believe they are entitled to a “reasonable” price and firms are entitled to a “reasonable” profit. Pricing that seems solely aimed to increase profit beyond what is viewed as reasonable is likely to be viewed as unfair. Specific implications of the principle of dual entitlement include the following.

• Raising prices to recoup increased costs is viewed as fair by most consumers.

• Raising prices simply to increase profits is likely to be viewed as unfair—even if people are willing to pay the higher price.

• Deviations from customary practices are often viewed with suspicion as hidden attempts to raise price.

• Providing information about the reason for increasing a price can often help acceptance.

One of the predictions of the theory of dual entitlement is that customers should be more accepting of price rises resulting from cost increases than they would be of price rises resulting from supply shortages. This does, in fact, appear to be the case—for example, charging more for vital supplies immediately before or after the arrival of a hurricane is viewed as very unfair by consumers. In this case, suppliers appear to be raising prices simply to grow profits unrelated to any change in costs in violation of dual entitlement. In the face of widespread consumer outrage after some suppliers hiked prices in the wake of Hurricane Andrew in 1992, Florida passed a law against price gouging. As of 2019, 35 states have passed laws against price gouging during periods of emergency (FindLaw 2020).

Further evidence of dual entitlement was gathered by Kahneman, Knetsch, and Thaler, who posed the following scenario to a sample of consumers: “A hardware store has been selling snow shovels for $15. The morning after a large snowstorm, the store raises the price to $20. Rate this action as: completely fair, acceptable, somewhat unfair, or very unfair” (1986b, 729). Of 107 respondents, 18% rated the pricing action as “acceptable or completely fair,” while 82% rated it as “somewhat unfair or very unfair.” In a similar vein, Kahneman, Knetsch, and Thaler asked what a store should do if it found it had exactly one item left in stock of that year’s most popular toy, assuming that all other stores were sold out of it (in their 1986 paper, it was a Cabbage Patch doll). A vast majority of those surveyed thought that it would be “very unfair” for the store simply to sell the toy to the highest bidder.

One problem created by dual entitlement is due to the fact that customers do not have a very good idea of sellers’ costs or profits. A 2001 survey of price and profit perceptions produced the results shown in Table 14.1. Each of the numbers in Table 14.1 is a gross overestimate of the actual margin. A typical discount store might realize a net margin of 4% and a typical grocery store 2%—not the 30% and 27.5%, respectively, that customers apparently believe. Part of the discrepancy may be explained by reference pricing, since consumers tend to set their reference prices based on the lowest prices they see for an item and they assume that these prices are still profitable for the seller. Since merchants often mark down items below cost, consumers tend to assume that the original list price must have included a vast profit for the seller. This creates a problem. As described in Chapter 12, markdown pricing is often the optimal tactical policy for a merchant selling a stock of perishable goods. However, the short-term gain in profitability from the markdowns must be compared with the longer-term effect of creating the perception in consumers’ minds that the cost of the item is low and that the merchant is reaping exorbitant profits at list price.

TABLE 14.1

Responses to consumer survey questions regarding perceived profit

SOURCE: Bolton, Warlop, and Alba 2003.

The principle of dual entitlement seems to be closely entwined with reference prices in the minds of customers. Consider the following two scenarios, posed by Kahneman, Knetsch, and Thaler (1986a).

• A shortage has developed for a popular model of automobile, and customers must now wait two months for delivery. A dealer has been selling these cars at list price. Now the dealer prices this model at $200 above list price.

• A shortage has developed for a popular model of automobile, and customers must now wait two months for delivery. A dealer has been selling these cars at a discount of $200 below list price. Now the dealer sells this model only at list price.

In the first case, 71% of those surveyed thought the dealer’s action was unfair, while only 42% thought it was unfair in the second case. It appears that respondents are using the (unspecified) list price as the reference price at which the dealer is making a fair return. Pricing at this reference price is viewed as intrinsically more fair than pricing above it, independent of the absolute price being charged or the historical pricing practices of the seller. This is likely one reason why many customers view a dealer pricing a hot new model of car above the manufacturer’s suggested retail price as unfair.

In view of this concern with fairness, music concerts and sporting events sometimes deliberately set prices below market based on the desire of the artist and/or promoter not to be perceived as gouging the fan base. As the late rock-star Tom Petty put it:

My top price is about $65, and I turn a very healthy profit on that; I make millions on the road. I see no reason to bring the price up, even though I have heard many an anxious promoter say, “We could charge 150 bucks for this.” . . . It’s so wrong to say, “OK, we’ve got them on the ticket and we’ve got them on the beer and we’ve got them on everything else, let’s get them on the damn parking.” (Quoted in Wild 2002, 23)

Kid Rock decided to take a “pay cut” on his 2013 tour by selling tickets in every venue for $20, even though the market-clearing price was much higher (Waddell 2013). Bruce Springsteen gave away an estimated $3 million in potential revenue to his fans at a Philadelphia concert by selling tickets at a single price that was well below the price that would have cleared the market. Since the concert grossed about $1.5 million, this is evidence of a serious desire to avoid perceptions of unfair pricing (Perry 1997).

Many economists would argue that these attempts to keep concerts affordable for the average fan are in vain because underpriced tickets are rapidly purchased by scalpers who either resell in person or use online marketplaces such as RazorGator to sell them at the much higher market-clearing price. Indeed, many of the artists and promoters who want to make tickets available more cheaply have resorted to complicated rationing methods to try to ensure that scalpers do not purchase all the cheap tickets to resell on the secondary market. The fact that some artists believe that such strenuous exercises are worthwhile is strong evidence that they want fans to believe that they are being fair in setting ticket prices and not merely chasing massive personal profits.

Dual entitlement also helps explain some of the negative reaction to Coke’s temperature-sensitive pricing scheme. Raising the price of a Coke on a hot day is viewed as similar to raising snow shovel prices after a snowstorm. This perception was reinforced by the fact that Coke’s announcement focused on the increased-price scenario, causing consumers to see the cold-weather price as the baseline price. It is likely that they believed Coke was making a reasonable profit at the cold-weather price and the hot-weather price was an attempt to gain an unreasonable profit, violating the principle of dual entitlement.

14.3.2 Interpersonal Fairness

As the Amazon case shows, price differentiation is an extremely sensitive issue. From the seller’s point of view, one of the most annoying aspects of interpersonal price comparison is that it can turn a happy customer into an unhappy customer for reasons totally unrelated to the quality of the underlying product or the responsiveness of customer service. The disgruntled Amazon customer who complained to DVDtalk.com about the price quoted for the DVD was not dissatisfied with his purchase or with Amazon’s service; he was unhappy because of the idea that someone may have been quoted a lower price. As one of the reporters covering the Amazon story wrote, “I don’t like the idea that someone is paying less than I am for the same product, at the exact same instant” (Coursey 2000). In other words, “I was happy with my purchase until I found out that someone else paid less.” This ex post effect will be familiar to anyone who has ever been responsible for setting salaries: employees who were perfectly happy with their raises become enraged upon hearing that someone considered less worthy received a larger raise. Similarly, learning that someone else received (or was offered) a better price for the same good can instantly convert a happy customer to an unhappy one. This is irrational, at least in the narrow economic sense: my consumer’s surplus should depend only on the price I paid, not on what others paid. Despite this, as the Amazon DVD pricing example shows, pricing that is perceived as unfair can generate a tremendous amount of negative emotion, ill will, and bad publicity.

Many cruise lines have faced this issue over the years. As discussed in Chapter 8, cruise lines have constrained, perishable capacity and most of the larger cruise lines actively practice revenue management to maximize the revenue for different sailings. This means that, if a sailing is not booking up as expected, the cruise line will lower the price for different cabin types to stimulate demand. Thus, different passengers can end up paying different prices for the same cabin type on the same sailing. It is likely that such price differences will not remain secret. People on a cruise dine together every evening. On the second or third night, the conversation will almost inevitably turn to prices. Mr. Smith mentions that his travel agent got a great deal for him and Mrs. Smith—an upper-deck, ocean-view stateroom for only $3,200. Mr. Jones responds that his travel agent got him an upper-deck, ocean-view stateroom for $2,800. Mr. Smith’s face falls, and his cruise is suddenly ruined—at least for the moment. The cruise industry has tried various mechanisms to avoid this scenario. Currently the most popular appears to be a low-price guarantee or best-price guarantee in which a cruise line will provide a rebate—often in the form of an onboard credit—if a lower price is advertised within 48 hours of a purchase.

There is evidence that emotional reactions to unfair treatment are widespread across cultures and that the phenomenon has deep roots. Just how deep these roots might go was demonstrated by researchers at the Yerkes Primate Lab who performed a set of experiments in which capuchin monkeys could “buy” food using stone tokens. The food reward was either a cucumber slice or a grape. The monkeys preferred grapes, but those who received a cucumber slice in return for a token were happy until they saw that another monkey was getting a grape. When this happened, the behavior of the monkeys receiving cucumbers changed abruptly—they refused to eat the cucumber slices and threw them back at the researchers—reactions that never occurred when all of the monkeys received cucumber slices. The researchers concluded: “Capuchin monkeys . . . react negatively to previously acceptable awards if a partner gets a better deal” (Brosnan and de Waal 2003, 299).⁸

By this point, you might be forgiven for believing that any form of price differentiation will be considered unfair by consumers. However, this is not the case. We have seen that dynamic pricing is standard practice in many industries, including both online and brick-and-mortar retailers. Furthermore, price differentiation is widely exercised in many industries through such practices as couponing, sales promotions, markdowns, regional pricing, and peak-load pricing with little or no objection from consumers. In addition, some examples of direct, customer-based price differentiation that are widely practiced with little controversy include these:

• Senior citizen discounts

• Student discounts

• “Kids ride free”

• Family specials

• Lower tuition for in-state students than for out-of-state students at state universities

• College scholarships for low-income students, athletes, or those who score higher on tests

These examples are all pure price differentiation, or group pricing in the terminology of Chapter 6, in the sense that they are based entirely on characteristics of an individual customer and do not reflect any product or cost-of-service differences. Because of this, we would expect them to be controversial. Yet, in most cases, such policies are not particularly controversial, probably because they correspond to an ideal of giving discounts to “deserving” groups. What groups are deemed deserving will vary by country and culture. For many years, the French national railway (SNCF) maintained special discounts (“social tariffs”) for certain classes of workers, schoolchildren, apprentices, the unemployed, the disabled, the families of soldiers killed in World War I, and large families. The large-family (Familles nombreuses) tariff provided discounts of 30% for families with three children, 40% for families with four, 50% for families with five, and 75% for families with six or more children. This particular discount was adopted in 1921 to encourage repopulation of the country after World War I and remained in place until 2008 (De Boras et al. 2008; Chemin 2008).

Perhaps the most controversial form of group pricing that is still widely practiced is gender-based pricing. It is well established that women tend to be charged more than men for certain products and services. A study by the New York City Department of Consumer Affairs in 2015 found that across a sample of identical (or very similar) products marketed separately to men and women, the women’s version was priced higher 42% of the time, while the men’s version was priced higher only 18% of the time (Bessendorf 2015). A survey by researchers at the University of Central Florida showed that, on the average, women tended to be charged $2.17 more than men for a haircut at a low-end salon, $11.71 more at a midrange salon, and $12.24 more at a high-end salon. Furthermore, women tended to be charged an average of $3.95 for dry-cleaning a shirt compared to $2.06 for men, and women’s deodorant averaged $0.29 per ounce more than men’s deodorant. The prices of other types of dry-cleaning services, such as two-piece suits, blazers, and slacks, and other types of personal grooming products, such as razors, body spray, and shave gel, did not show statistically significant differences between women and men (Duesterhaus et al. 2011).

The surcharge that women pay over men for these products and services has been called the “pink tax.” There have been popular press articles on the pink tax (see Sebastian 2016; Ngabirano 2017), and a few jurisdictions, such as California, Massachusetts, Miami/Dade County, and New York City, have outlawed gender differences in pricing. On the other hand, anti–pink tax legislation has failed to pass in a number of locations in the face of opposition by business lobbyists. Even in jurisdictions with laws banning gender discrimination in pricing, enforcement has been sporadic and the practice persists. For example, in California, fewer than five suits were brought in the 20 years following the passage of the Gender Tax Repeal Act in 1995. And, although gender-based pricing for salons was (and still is) widely practiced in California, only five lawsuits were brought in the 10 years following the adoption of the legislation (Jacobsen 2018). This might suggest that much of the public has learned to accept the practice—or is at least not sufficiently outraged by it to pursue legal remedies.⁹

You may realize by now that there is no universally reliable test to determine what types of pricing will be widely accepted by consumers and what will raise complaints. Cultural, social, and educational factors all seem to play a role—what is acceptable to one group (or one consumer) may not be to others. Policies that favor groups perceived as disadvantaged, such as the elderly, seem to be more acceptable than those that appear more arbitrary, despite the fact that senior discounts could often be taking advantage of the tendency of seniors to be more price sensitive than other groups. Surveys have shown that business and economics students tend to accept differentiated and dynamic pricing tactics more readily than the general public. For example, in the snow shovel scenario discussed earlier in the chapter, 76% of University of Chicago economics students said that raising the price of snow shovels after a snowstorm was acceptable, compared to 18% of the general public. This disparity would suggest that the prospect for future pricing debacles is good, since it is the business and economics students who are likely to be formulating the pricing schemes of tomorrow and, as a group, they may be oblivious to presentation and fairness issues.

14.4 IMPLICATIONS FOR PRICING AND REVENUE OPTIMIZATION

The topic of irrational consumer behavior and its influence on real-world pricing is relatively new and not perfectly understood—especially when it comes to judgments of fairness. As one researcher put it, “Fairness, it seems, is a complex and dynamic concept with many varied inputs” (Andreoni, Brown, and Vesterlund 2003, 22). Given this, you might conclude that the current state of knowledge is that consumers act rationally except when they do not. To some extent this is a fair judgment. Nonetheless, enough is known that it is possible to make specific recommendations to increase the effectiveness of pricing and revenue optimization and increase its chances for consumer acceptance. We start by looking at a summary of the factors that make tactical pricing schemes more acceptable. We then consider the implications of these for implementing pricing and revenue optimization. Finally, we return to the concept of fairness and examine two cases in which unfair pricing may have won out over fair pricing, despite consumer reactions.

Key elements that influence the acceptability of pricing and revenue optimization programs are shown in Table 14.2. The impact of these can be summarized as follows.

• Product-based. Product-based pricing differences are much more readily accepted than group pricing. As Chapter 6 shows, there is not a clear distinction between product-based differentiation and customer-based differentiation. To the extent that a pricing difference can be presented as product based rather than customer based, it is much more likely to be readily accepted.

• Openness. Customers certainly claim to prefer pricing schemes that are open. Some of the negative press that Amazon received for its DVD pricing dwelled on the idea that Amazon was doing something in secret. Furthermore, one of the lessons of the Amazon case is that any attempt to implement differential or dynamic pricing on the internet will be discovered and publicized. You should avoid any online pricing scheme that does not meet the New York Times test—that is, you would be averse to having it the topic of an article in the New York Times.

TABLE 14.2

Pricing acceptance factors

• Reasonable. Customers consider certain reasons for raising prices more reasonable than others. They are more accepting of a price raise due to increased costs than they are of one that is just about increasing profit. Customers also have implicit beliefs about a seller’s costs and—via dual entitlement and reference pricing—some idea of what reasonable prices should be. Under the right circumstances, communicating the reasons behind discounts, but especially price increases, can help acceptance.

• Discounts and promotions are much more readily accepted than surcharges. Prospect theory predicts that the same price will be viewed more positively if it is the result of a discount rather than a surcharge. Furthermore, finding that another customer received a discount is much more likely to be received with equanimity than finding that one has been charged a surcharge—even when the net result is the same. As a result, list prices tend to be high, with discounts and promotions commonplace and surcharges rare.

• Rewards. In a similar vein, a reward for loyalty or volume purchases is much more likely to be acceptable than any kind of penalty. Airlines do much better offering frequent-flyer rewards than they would offering infrequent-flyer penalties.

• Easy to understand. Programs viewed as easy to understand are generally more accepted. Programs that are difficult to understand will cause customers to be concerned that they are not getting a good deal simply because they do not know how to play the game. Several studies have shown that, in some cases, customers may actually prefer suppliers who offer them fewer options—even if their resulting price is higher (Iyengar and Lepper 2000). Furthermore, customers often struggle to find the “best” option in cases in which computing the “real” price of an option is difficult—this is a particular issue in the case of financial products such as mortgages in which the total interest paid is a complex function of the interest rate, the term, the down payment, and other factors. In cases with many complex choices, many customers will revert to a simple heuristic such as “We’ll take the mortgage with the shortest term that has a monthly payment within our budget” even though that may not be the best choice.¹⁰

• Available. There seems to be a critical distinction in customers’ minds between a deal that is potentially available to them and one that is not—a distinction closely related to the difference between product-based and customer-based segmentation. Customers are much more likely to be accepting of a discount that was (at least in theory) available to them than one that was not available to them under any circumstance. Thus, the acceptance of early-booking discounts in the airline industry was helped by the fact that any customer could take advantage of them, if only they had been able to make their travel plans earlier. In the same vein, the wide acceptance of discount coupons seems to stem in part from the fact that those who do not take advantage of the coupons realize that they could have done so by taking the time to clip them out of the newspaper. These behavior-based differentiation strategies seem to be much more broadly acceptable to customers than identity-based differentiation, which is predicated on factors such as race, age, and gender.

• Familiar. Customers are naturally conservative—they prefer traditional and customary ways of doing business and distrust innovations. Virtually any change in the way that prices are calculated, billed, or presented is likely to be viewed with suspicion as an attempt to raise prices. For example, there was widespread belief among consumers in the Eurozone that merchants took advantage of the conversion from the native currency to the Euro to raise prices.¹¹ As another example, ride-sharing companies such as Uber and Lyft traditionally based rates on a “surge multiplier” applied to the base fare (see Section 7.6.4 for details on how these multipliers are calculated and used). In 2019, Uber experimented with “additive surge,” in which the amount that drivers received would be a fixed surcharge—for example, $2.10—to the base price of a ride rather than a multiplier (Garg and Nazerzadeh 2019). Many drivers assumed—largely without evidence—that the new practice was designed to reduce their earnings and enrich Uber.

You should anticipate negative reactions to virtually any deviation from traditional pricing practice. The least disruptive way to introduce pricing and revenue optimization in a market is within the context of existing and accepted tactical pricing practices, such as markdowns, promotions, and customized pricing.

One final note: While it is important to be sensitive to customer price perceptions, it is equally important not to allow anticipated customer reaction to squelch all possibility of pricing innovation. Many of the key findings on price perception, prospect theory, and fairness have been based on surveys, and, as any marketing professional knows, survey responses do not necessarily reflect how people will react in real life. For example, a majority of golfers stated that charging different prices depending on when they booked a tee time (e.g., two weeks in advance versus two days in advance) would be unacceptable (Kimes and Wirtz 2003). Nonetheless, several golf courses and online golf booking services have implemented time-of-booking pricing, with few or no customer complaints. While a pricing innovation may initially meet resistance and generate complaints, over time it may become the accepted norm. This is certainly what happened with airline fares—when revenue management was first initiated, customers complained about changing fares and that “the person sitting next to me is paying half of what I paid.” However, over time, dynamic pricing and revenue management became accepted as the norm, and the complaints subsided.

The lesson is that customer complaints and negative press do not necessarily mean that a pricing program is a failure. Consider the case of the First National Bank of Chicago (usually called First Chicago). In the first quarter of 1992, the First Chicago realized that it had a problem with its retail banking operations. It calculated that 67% of its retail customers were not achieving adequate profitability. A primary reason: whereas many customers used tellers rarely if it all, a small number were so-called transaction hounds using tellers more than four times a month while maintaining less than $2,500 in their accounts. Since teller salaries and benefits were a major part of the bank’s variable cost, it seemed only fair to charge customers for the service. As Jerry Jurgensen, First Chicago’s executive vice president in charge of branches, put it, “If nobody changed their behavior, 80% of my customers would never see the $3.00 charge. For the other 20%, I want one of three things to happen: I need for them to change their behavior; I need for them to be willing to pay more, or I need for them to find another bank” (O’Sullivan 1997, 44).

First Chicago held a breakfast press conference to announce the new plan. It was not well received. The headline in the Chicago Tribune read, “First Chicago Loses Touch with Humans.” Competing banks leaped at the chance to tout their free checking services. The story hit the national papers and was even featured in a joke on The Jay Leno Show.

First Chicago was baffled by the reaction to its new pricing scheme. It had tested the plan for 18 months on different focus groups. It had gone out of its way (perhaps too far out of its way) to be open, announcing the new scheme at a press conference. And, after all, the $3.00 teller fee was based on passing a real cost through to the customer, just the type of price change that the principle of dual entitlement predicted would be acceptable.

This would seem like a classic case of a pricing fiasco, and it is often presented as such: A clueless marketing department instituted a bonehead pricing program that ignored customer sensitivities and reaped a whirlwind of criticism. Another disaster, to be put alongside Amazon and Coca-Cola in the annals of pricing infamy. However, there is a twist to the story. From a financial point of view, the First Chicago program was a success. According to an article in the American Banking Association Banking Journal:

Banks are uniform in disassociating their profitability-probing behavior from First Chicago’s, which some consider to have been “PR suicide.” “They fired their marketing people the next day [after the teller-fee story broke],” said a source who preferred not to be named. However, it is widely conceded that First Chicago has had material success. It reportedly lost less than 1% of customers, not the 10% predicted—despite some other Chicago-area banks advertising against it—while branch staff decreased 30% and ATM activity “skyrocketed” (ATM deposits grew 100% in three months). Speaking seven months after the restructuring, [former First Chicago senior vice president Marion Foote] said the bank was making an “adequate return” on 44% of customers, versus 33% beforehand, and customer satisfaction scores were higher than ever. (O’Sullivan 1997, 43)

Was teller surcharging a success or a failure at First Chicago? It depends on whom you ask. The marketing and public relations staff who lost their jobs would probably claim that it was PR suicide. Anecdotal evidence seemed to indicate that some potential new customers were lost because of the negative publicity. But from the cold numbers it would appear that the teller fees were successful at achieving the announced goals: increasing customer profitability. It is suggestive that, despite the overwhelmingly negative publicity, First Chicago retained the fees for more than seven years, until it became part of BankOne. And since 1995, more and more banks have begun to charge teller fees (often as part of bundled pricing packages). For many banks, the purpose of the fee is to motivate customers to move from using human tellers (at an estimated cost of $4 per transaction) to using video tellers (at an estimated cost of $2 per transaction) (Lake 2018). However, the banks who have adopted teller fees since First Chicago have had the good sense not to call a press conference to announce that they were doing so.

Finally, competitive and market forces may not allow a company to simplify its pricing, even if it wants to. Consider the case of American Airlines.

By the spring of 1992, Robert Crandall, chairman and CEO of American Airlines, decided that airline fares had gotten too complex. While American’s revenue management system had originally given it an advantage over its rivals, United and Continental had developed systems of their own and it was not clear that American retained any advantage. Furthermore, customers continued to complain about a system where business travelers paid much more money than leisure travelers for substantially the same service. Simplifying fares seemed to be a logical choice—it would allow all of the carriers to jettison the increasingly complex fare structures, eliminate much discounting, and improve customer satisfaction, all at the same time. American had led the majors into this swamp, and as the world’s largest carrier, perhaps American could be the airline to lead them out.

On April 9, 1992, American Airlines called a press conference to announce its “value pricing” program. Crandall explained the rationale for change:

We have said for some time that our yields are too low, yet the conventional solution, higher and higher full fares and an ever-growing array of discount fares surrounded by an ever-changing plethora of restrictions, simply does not work. . . . In our unsuccessful pursuit of profits, we have made our pricing so complex that our customers neither understand it nor think it is fair. (Quoted in Silk and Michael 1993, 5)

Under value pricing, American sold only four fares in every market: first class, regular coach, 7-day advance purchase booking discount coach, and 21-day advance purchase booking discount coach. This simple structure was to replace the previous situation, in which American often offered 10 or even 20 different fares in a market, with different booking restrictions in different markets. Value pricing would reduce the total number of fares in the network from 500,000 to 70,000 (Silk and Michael 1993). As part of the plan, American also lowered its fares—first-class fares were reduced by 20–50% and regular coach fares were cut by about 40%. However, a myriad of other promotions and discount programs for groups and corporate customers were eliminated. Crandall’s plan was that the new and simpler fares would induce enough additional traffic to outweigh the fare reductions.

American expected other carriers to follow its lead. After all, the complex fares associated with traditional revenue management were just as unpopular at competing carriers, and the cost savings from fare simplification would be significant across the industry. American announced its intentions well ahead of time and did everything it could to signal to the other airlines that its new fare structure was about simplification, not starting a new fare war, and that simplification would be good for the entire industry. And the initial response was promising. Within a week, most major carriers, including United, Delta, Continental, and Northwest, adopted the value pricing fare structure, either in part or in whole. Initially, it looked like American had been able to lead the airlines successfully to a new, simpler structure without touching off a fare war.

Unfortunately, the hope was premature. TWA began a fare war in the second week of April. Initially, American stood by its new pricing scheme. However, in late May, Northwest issued a more serious challenge. It announced a special fare promotion called “Grown-Ups Fly Free” that offered free tickets to an adult when accompanied by a paying passenger between 2 and 17. The other airlines, including American, matched the promotions with 50% discounts of their own. These discounts were certainly effective in stimulating demand, and aircraft were flying with record loads—in many markets load factors approached 100%. However, the demand stimulated by the discounts did not outweigh the cannibalization, and many of the airlines suffered record losses along with their record load factors.

Increasingly, it became clear that value pricing was doomed. With an increasing variety of discounts and promotions being offered, things were reverting to the bad old days. American’s rivals were setting up fare structures that bracketed American’s three coach fares and then using their revenue management systems to feast on American, just as American had used its system to feast on PeopleExpress. By October 1992, it was evident that value pricing was finished. In an interview with the BBC, Crandall said that value pricing was “dying a slow death” and that the plan had “clearly failed” (Silk and Michael 1993, 7). By November, American had jettisoned value pricing completely and was back to setting and manipulating thousands of fares throughout its system.

Value pricing is often presented as a failed attempt at price leadership—which it was. Competitors found that by offering more complex fare structures, they could outcompete American in key markets, and the lure of doing so proved irresistible. Ultimately, complex fare structures backed up by sophisticated revenue management beat out the simplicity of value pricing.

But there is another lesson from the value pricing failure: Customers may prefer simple pricing to complex pricing, but they are not willing to pay for it. Customers praised the simplicity of value pricing, but when the time came to purchase a ticket, they bought from the airline that was offering the cheapest seat—not from the one offering the simplest fare structure. The customers who flocked to take advantage of “Grown-Ups Fly Free” tickets from Northwest did not care that they were destroying value pricing—they were simply looking for a bargain. This indicates that the success of everyday low pricing is primarily due to the fact that it is low, not to the fact that it is every day.

The most successful pricing and revenue optimization tends to be the most invisible. Systems that simply do a better job of setting prices within the context of an existing and accepted tactical pricing structure have the highest chance of success and acceptance. Nonetheless, capturing the full benefits of pricing and revenue optimization may require new pricing practices. You can increase the chance of success by paying close attention to issues of price presentation and fairness in implementation. However, you should also expect that virtually any pricing innovation will be met with skepticism (at best). Part of the art of pricing and revenue optimization is knowing when to change in response to consumer resistance and when to persevere.

14.5 SUMMARY

• Customers do not always respond to prices in a way that is consistent with economic theory. Instead they are influenced by a variety of considerations, including the way the price is framed and the prices offered to others. When this occurs, it is said that customers behave irrationally. The study of the ways in which customers deviate from the predictions of economic rationality is called behavioral economics.

• In general, lower prices lead to higher demand and vice versa; this is the law of demand. However, there are three situations in which demand may actually increase with price:

• With Giffen goods, an inferior good whose price is rising is substituted for a superior good.

• Price is considered a signal of quality by consumers in some cases.

• In conspicuous consumption, the price of a purchase signals to others the presumed wealth of the buyer.

• Customers are sensitive to the way that prices are presented—a price that is presented as a discount will typically attract more demand than the same price presented as list price. In addition, customers tend to be very averse to any price that is presented as a surcharge or a penalty. For this reason, sales promotions and discounts are common, while surcharges are rare. This phenomenon has been explained by prospect theory, which specifies that people are more averse to perceived losses than they are attracted to gains. A discount is perceived as a gain relative to a baseline price, which may be a reference price or a published list or “was” price, while a surcharge is perceived as a loss.

• Consumers believe it is fair that a company makes a reasonable profit and that they are offered a reasonable price—this is the principle of dual entitlement. Consumers are likely to reactive negatively if they believe that a company is violating this principle. This is an explanation for the fact that most consumers find price raises due to cost increases more acceptable than those due to scarcity.

• Consumers are also concerned with interpersonal fairness. In particular, they are likely to get upset if they learn that another customer paid (or was offered) a lower price for the same item at the same time. This makes many forms of group price differentiation controversial. Indeed, some forms of group price differentiation are illegal, depending on the jurisdiction. As a result, although individualized pricing on the internet may be tempting, it raises the very real likelihood of strong consumer pushback.

• Consumers also tend to be conservative when it comes to pricing approaches. They are likely to believe that any change in pricing structure is an excuse for raising prices and violates the principle of dual entitlement.

• Pricing changes are more likely to be accepted if they are:

• Product based rather than customer based

• Open rather than hidden

• Posed as a discount or reward rather than a surcharge or a penalty

• Easy to understand rather than hard to understand

• Available to all customers rather than restricted

• Cost based rather than profit based

• Familiar rather than unfamiliar

• Consumers’ conservatism means that they are likely to object to any change in the way that goods are priced. But this does not mean all pricing innovation should be stifled. For example, customers objected to First Chicago’s introduction of teller fees; however, it turned out to be reasonably successful. On the other hand, while consumers complained about the complexity and volatility of airline fares, American Airlines was unsuccessful in introducing a simplified fare structure because, at the end of the day, consumers preferred low fares to simple fares.

14.6 FURTHER READING

Behavioral economics has roots in the 1950s with the concept of bounded rationality developed by Herbert Simon. During the 1960s and 1970s, Daniel Kahneman and Amos Tversky published a series of papers that essentially established behavioral economics as a field. An entertaining description of the early development of behavioral economics through the collaboration between the very different personalities of Daniel Kahneman and Amos Tversky can be found in Lewis 2016.

The early part of this century saw an explosion not only of academic interest but also of books written for a wider audience, including Nudge, by Richard Thaler and Cass Sunstein (2008); Predictably Irrational, by Dan Ariely (2010); and Thinking, Fast and Slow, by Daniel Kahneman (2011). Behavioral economists study a wide range of topics, but pricing has been an important focus of research: surveys of research on the behavioral economics of pricing can be found in Özer and Zheng 2012, Hinterhuber 2015, and Husemann-Kopetzky 2018.

The use of brain scans to investigate the effect of prices and other marketing actions on cognition and behavior is relatively new but is rapidly expanding. A useful survey can be found in Hubert and Kenning 2008.

14.7 EXERCISES

1. Giffen good. A penurious graduate student has a food budget of $100.00/week. To survive with sufficient energy to attend classes, he knows that he needs to consume 50 protein units per week. The only two foods he can stand to eat on a regular basis are beans and hamburger. He derives twice as much pleasure per protein unit from eating hamburger as he does from beans.

a. Assume that hamburger costs $3.00 per protein unit and beans cost $1.00 per protein unit. Formulate the student’s diet problem as a linear program. (You can assume he wants to maximize his total utility from his diet and that he gets 1 utile from each protein unit of beans he consumes and 2 utiles from each protein unit of hamburger.) What is the optimal consumption of beans and hamburger in this case?

b. Plot the student’s price-response function for beans as the price of beans goes from $0.01 to $2.00 per protein unit, assuming that everything else (including the price of hamburger) stays constant. Note that his individual price-response function is indeed upward sloping. Why? What happens when the price of beans exceeds $2.00 per unit?

c. The student receives a scholarship that enables him to spend $300 a week on food. What is his price-response function for beans now?

2. A particular market was served by two products offered by Acme and a competitor. The prices and quality ratings of the two products are as follows:

	Price	Quality rating
Acme brand	$2.00	50
Competitor brand	$2.50	70

The two brands each have a 50% share of the market. Now Acme introduces a new product with a price of $2.20 and a quality rating of 30. What would you expect to happen in the market?¹²

a. The new product will grab a dominant share of the market.

b. The new product will split the market with the existing products into thirds.

c. The new product will not gain much market share, but Acme will increase market share at the expense of the competitor.

d. The new product will not gain much market share, but the competitor will gain market share at the expense of Acme.

e. The new product will not gain much market share, and the relative shares of Acme and the competitor will be unchanged.

NOTES

1. Since it was an interview with a Brazilian magazine, Ivester was referring to a soccer championship match.

2. In this case, the term “dynamic pricing” was used as a misnomer for “price differentiation” or “price discrimination.”

3. The term “irrational” is not meant to be pejorative in this context—it simply refers to behavior that does not accord with the neoclassical economic model of rational consumer decision making.

4. The research on price-ending effects is quite extensive. For useful surveys, see Özer and Zheng 2012 and Husemann-Kopetzky 2018.

5. Although the idea of Giffen goods dates from the mid-nineteenth century, the first empirical evidence for a Giffen good in the real world—rice markets in Hunan Province in China—was not provided until the twenty-first century (Jensen and Miller 2008), although their claim is controversial. If Giffen goods do exist in the real world, they are quite rare and have little relevance to the practical problem of setting prices.

6. For the optimality of hi-lo pricing in the presence of reference price formation, see Greenleaf 1995 and Kopalle, Rao, and Assunção 1996. Popescu and Wu 2007 shows that cyclical pricing can be optimal and that ignoring reference price effects can lead to underpricing.

7. The idea of a decoy was introduced in Huber, Payne, and Puto 1982. For a discussion of attraction and compromise effects, see Özer and Zheng 2012 and Husemann-Kopetzky 2018.

8. A video showing the outraged reactions of capuchin monkeys to unequal pay can be found in a 2013 TED Talk by Frans de Waal, available at http://bit.ly/1GO05tz.

9. A counterexample to the pink tax is provided by bars that offer “ladies’ nights,” during which women are charged less than men for drinks (a practice that is illegal in California) or dance clubs that charge more for men than women in an attempt to attract more women and achieve gender balance (Lam 2014). In articles on both the pink tax and on ladies’ nights, the lower price is taken as the baseline price, and the higher price is seen as potentially unfair—a likely example of dual entitlement.

10. For a discussion of behavioral issues in mortgage selection and other consumer credit pricing situations, see Phillips 2018, chap. 8.

11. For a survey on consumer perception of price increases and actual increases following adoption of the Euro, see Pufnik 2017, which concludes that actual price increases were far lower than perceived increases for every country that joined the Eurozone up through 2016.

12. The results of this experiment are from Huber, Payne, and Puto 1982.