There's a classic cartoon plot device that represents a struggle with temptation. A tiny angel pops up on the conflicted character's left shoulder, urging him to follow the path of righteousness. A tiny devil sits on his right shoulder, pressing him to give into his desires.

In real life, it turns out that an everyday item has the power to act as both angel and devil every time we go to the grocery store. It lurks in car trunks and pantries all over the world, waiting to guide us simultaneously down paths of virtue and vice. What is this surprising Svengali?

It's a reusable shopping bag.

New experimental research shows that shoppers are more likely to buy virtuous organic items when they bring their own reusable bags to the store than when they opt for paper or plastic bags at the checkout counter. At the same time, those who bring their own bags are more likely to buy indulgent items like ice cream and cookies. Moreover, consumers tend to place a higher value on both organic products and decadent treats when they bring their own bags than when they don't.

Researchers Uma R. Karmarkar and Bryan Bollinger report their preliminary findings in their working paper BYOB: How Bringing Your Own Shopping Bags Leads to Treating Yourself, and the Environment. (The collaborative effort addresses each of their particular interests. Karmarkar, an assistant professor and neuroscientist in the Marketing unit at Harvard Business School, studies factors that affect consumer choice. Bollinger, an assistant professor at NYU's Stern School of Business, studies the marketing of sustainable products.)

"There are all these little things that we're supposed to do to be better to the environment, like turning off the lights when we leave the room or recycling our bottles," Karmarkar says. "Bringing bags is interesting in that it's a difficult thing to remember to do, and actually requires a fairly big behavioral change on the part of the consumer. Our question was, when you succeed at this big behavioral change, does it change other elements of what you're doing as well?"

A SERIES OF EXPERIMENTS

As their working paper explains, the researchers combined empirical and experimental methods to test the purchasing effect of reusable bags.

Using reusable bags increases our tendency to buy both organic and indulgent items.
Photo: iStockPhoto

Looking at loyalty card data from a large grocery chain in California, Karmarkar and Bollinger tracked and analyzed 936,232 purchases by 5,987 households across two years. To assess organic purchases, they looked for transactions in which the consumer could choose either an organic or a nonorganic option—a carton of milk, for example. In monitoring what they called "indulgent" purchases, the researchers looked at sugary items like ice cream and candy bars, as well as salty treats like potato chips.

The data showed a definite correlation: Shoppers who had brought their own bags bought decidedly more indulgences and chose more organic products than those who didn't. But this wasn't necessarily enough information to establish causality—that is, that both effects were specifically due to bringing their own bags. "There are a lot of things going on in a store and a lot of inputs," Karmarkar says.

So she and Bollinger dug deeper with a series of experiments, enlisting participants for a number of online surveys.

In the first experiment, the researchers assigned participants to one of two conditions. The "with bags" participants were asked to imagine approaching a supermarket to do their grocery shopping with their own bags. The "without bags" group received nearly identical instructions, but nothing about bags was mentioned. All the participants looked at a floor map of the grocery store and listed 10 items they would most likely purchase on their hypothetical outing.

Regarding indulgent items, the results depended on whether the participants had children in their households. For those with dependents, there was no significant difference between the with-bags and the without-bags condition. For those without children, the with-bags participants were more likely to imagine buying ice cream and potato chips than the -without-bags- participants.

But the results couldn't speak to organic items; while participants listed items such as milk and vegetables, they generally didn't list whether their hypothetical choices were organic milk and vegetables.

"We could support some of the story but not all of it yet," Karmarkar says.

And so she and Bollinger conducted a second experiment, in which participants reported how much they'd be willing to pay for each of nine specific products. These included both organic and indulgent items, as well as "baseline" items like canned soup. Again, the participants were divided into hypothetical conditions of "with bags" and "without bags."

Consistent with the empirical data, the idea of bringing their own bags increased the likelihood that participants would buy both indulgent and organic items. Moreover, it increased the amount of money they'd be willing to pay for those items.

But the researchers had another question: Does it matter whether a reusable bag is the consumer's choice? "We wanted to examine whether it was important that you made the decision to bring the bags as opposed to a store policy that requires it," says Karmarkar, noting that some stores obligate customers to bring their own bags; others charge customers a fee for single-use carryout bags per a local government mandate.

In the next experiment, all the participants imagined bringing their own bags to the hypothetical grocery store. But while some were told to imagine bringing reusable bags of their own volition, others imagined that they had to bring bags due to a store policy.

Participants then rated their willingness to purchase organic, indulgent, and baseline items. In this case, the results showed no significant difference between the two groups with regard to organic items, which rated highly across the board. However, participants were more likely to buy indulgent foods if they imagined that bringing bags was their own choice.

"A simple way I think about those results is that if you do something good, you reward yourself," Karmarkar says. "You did something good for the environment, so you can have a cookie."

IMPLICATIONS

For retailers, the results suggest that store managers should reconsider where they display their organic items. In short, it may make sense to locate the kale near the Kit Kats.

"The research implies that the area near the checkout counter is a good place to display organic or environmentally positive items," Karmarkar says. "That's the place where shoppers' attention is probably going to be most focused on this element, the bag, which seems to encourage them to buy these things."

For consumers, she recommends that they just think about the findings as they stroll down the grocery store aisles.

"I'm of the mindset that it's useful to know the kinds of things that influence your own behavior," Karmarkar says. "If you're trying to maintain a strict diet, maybe you can recognize the bag's influence, and consciously fill the desire to treat yourself in another way that doesn't interfere with your goals. Maybe you can treat yourself to an extra half hour of sleep."

The "sharing economy" is a burgeoning business model in which people offer their personal belongings and personal services to others, usually through online marketplaces that facilitate the transactions.

It seems absolutely egalitarian at face value: Anyone who owns a sometimes-unused thing—an apartment, a car, a boat—now has an easy way to advertise and share it. And anyone with the time and skills—driving, running errands—can find customers who need these services. But new research shows how online marketplaces can work in ways that are anything but egalitarian: They can inadvertently fuel racial discrimination.

In Digital Discrimination: The Case of Airbnb.com, Benjamin G. Edelman and Michael Luca investigate the possibility of racial discrimination against people who advertise properties on Airbnb, a popular online marketplace that lists temporary rooms and homes in some 34,000 cities across 192 countries.

Analyzing a data set that focuses on New York City (the company's biggest market), the researchers found that black hosts charged approximately 12 percent less for rentals than nonblack hosts—even when the properties were equivalent in terms of location and quality.

How did the researchers find out the race of each host? The same way potential guests do—they looked at the hosts' profile pictures on the Airbnb website. In fact, large profile pictures in online marketplaces were the inspiration for the research.

"In the early days of the Internet, online marketplaces like eBay were relatively anonymous," says Luca, an assistant professor in the Negotiation, Organizations & Markets unit at Harvard Business School. "You just saw a seller's ID with a seller score, or a buyer with a buyer rating."

These days, however, profile pictures are commonplace. It's a phenomenon that Edelman attributes to Facebook. In fact, many sites require customers to use Facebook for a login mechanism—not just marketplaces but media outlets as well. For instance, ESPN.com and USATODAY.com both ask readers to sign in through Facebook in order to post a comment on a news story. "If you log in to a site through Facebook, they get your profile picture, among other things," says Edelman, an associate professor and Marvin Bower Fellow at Harvard Business School. "Now, rather than standing on its own, your comment is juxtaposed with a picture of you."

While pictures have the potential to foster a sense of community in an online marketplace, they also enable racial discrimination. Thus, the researchers set out to find out whether potential discrimination was affecting the online marketplaces of the sharing economy.

A marketplace leader

Edelman and Luca focused their research on Airbnb for a couple of reasons. One, the San Francisco-based company, founded in 2008, is among the most successful sharing-economy marketplaces in the world, with 500,000 listings and counting. (By way of comparison, hotel industry giant Marriott offers around 530,000 rooms worldwide.) Airbnb, which has been valued at roughly $2.5 billion, recently closed a $200 million round of financing, according to a February 6 SEC filing.

Removing profile pictures could prevent racial discrimination in online marketplaces.
Photo: iStockPhoto

Two, the site requires hosts to post large profile photos, which are displayed next to the pictures of the rental properties. "You can't be a host without a profile photo," Edelman says. "Sure, you could put a picture of your dog or the Eiffel Tower instead of a picture of you. But if you have a photo of a dog where your face should be, it might deter people from renting your property."

The researchers' data set consisted of all Airbnb New York rentals available on a randomly selected day in July 2012. For each listing, they recorded information including the asking price of the rental, the characteristics of the apartment or room, and the average rating each host had received. (The site's structured rating system lets guests rate the hosts and properties on location, check-in, communication, cleanliness, and accuracy.)

The researchers also hired a team of workers through the crowdsourcing marketplace Amazon Mechanical Turk (AMT), a popular source of data processing assistance. The workers examined the property photos on each listing and rated them on a seven-point scale from "This is a terrible apartment; I would not stay here at any price" to "This is an extremely nice apartment; I would stay here even if it were a lot more expensive than a nice hotel room." With these ratings, the researchers controlled for apartment/room quality as seen by a user examining an Airbnb listing.

Other AMT workers were hired to look at the public profile pictures of the New York hosts, coding them into one of the following categories: white, black, Hispanic, Asian, unclear but nonwhite, multiple races, unclear/uncertain, or not applicable (because the host had posted a nonhuman picture of, say, a dog or the Eiffel Tower).

Even controlling for factors including perceived quality and location, the researchers found that listings with nonblack hosts earned roughly 12 percent more than those with black hosts for apartments with similar ratings and property photos. "Moreover, black hosts receive a larger price penalty for having a poor location score relative to nonblack hosts," the researchers write in the paper. "These differences highlight the risk of discrimination in online marketplaces, suggesting an important unintended consequence of a seemingly routine mechanism for building trust."

To Airbnb's credit, the company's terms of service forbid content that "promotes discrimination, bigotry, racism, hatred, harassment, or harm against any individual or group." Hosts are responsible for setting their own prices for the properties they advertise on Airbnb, so it's not as if the company is explicitly biased.

But the "pricing" section of the website advises that new hosts set rates with market demand in mind. For example, Airbnb tells new hosts they "may want to charge lower than average rates to attract travelers comparing your place with existing reviews. Once you have a review or two, adjust your rates as needed."

The researchers interpret black hosts' lower prices as evidence of a similar, but permanent, disadvantage. In particular, they maintain that their findings show how black hosts have a harder than average time attracting travelers, and therefore must keep their prices lower than average in order to compete successfully.

"Airbnb's pricing suggestions implicitly acknowledge that price is a market-driven outcome," Luca says. "In a market, prices move until supply meets demand. And the data indicate that discriminatory preferences are factoring into prices."

Advice for sharing-economy companies

Thus far, Airbnb has declined the researchers' offer to visit the company and discuss possible mechanisms to facilitate trust without encouraging unintentional discrimination. But they do have recommendations, not just for Airbnb but also for any online marketplace in the sharing economy.

For starters, Luca and Edelman suggest evaluating whether profile pictures provide any necessary information to the customer. "For instance, ask yourself what useful information you learn from looking at the Airbnb host's face," Edelman says. "The last time I was an Airbnb guest, I never laid eyes on the host except on the website. There was a neighbor facilitator who dropped off the keys and picked them up."

In any case, the researchers recommend putting profile pictures in a less prominent place, perhaps on a separate page from the picture of the product or property. (On Airbnb, the host's profile picture is the second-most prominent element on any listing page, displayed next to the default picture of the property.) "Research has shown that when something is more salient, you respond more to it," Luca says. "This is something a platform should think about in its design decisions. Is a host's physical appearance really the information Airbnb most wants to emphasize?"

More generally, the researchers recommend that sharing-economy companies take website design seriously and to question early on whether their system design encourages racial bias. "These questions are too important to be left to engineers, and too important to be left to graphic designers," Edelman says.

As they continue this line of research, Edelman and Luca are studying some of the online ride-sharing services of the sharing economy. Such services are especially relevant to racial bias studies, they explain. Booking a cab online has the potential to eliminate the possibility that a driver will decline to pick up a black person who is hailing a cab. But racial bias will still be a problem if the customer is required to provide a profile picture when signing up for the service.

"In so many contexts, online information systems can be much better than the real-world alternatives than they replace," Edelman says. "But they can also be much worse. The burden is on us, as system designers, to get it right."

Note to readers: Edelman and Luca are interested in conducting experimental studies with firms that operate online marketplaces.

"Online marketplaces are a great innovation with extraordinary potential, but they also present important challenges, as explored in this paper. We hope companies will design their sites to facilitate transactions and build trust while avoiding unintended consequences such as discrimination," Luca says. "There's a great opportunity for an interested company to take the lead on publishable research in this area. We're willing to help implement changes based on behavioral economics principles, including finding ways to reduce discrimination without hurting the bottom line."

If you think your company might benefit from this research, please write to Ben Edelman at bedelman@hbs.edu or Mike Luca at mluca@hbs.edu.

There are three ways to differentiate in retailing: location, location, and location. The problem is that as markets mature, location becomes less potent as a competitive advantage because the consumer has a growing abundance of convenient choices.

That's one reason why mass retailing in mature markets is a sector notorious for its lack of differentiation between players. Once location has played out its magic, retailers tend to get squeezed in a business characterized by the infernal duo of low margin and high fixed cost. In such businesses, price wars are never far away.

Creating non-price differentiation is difficult in retail as well because development of such advantages takes time and is difficult to execute. All the while, low-price players are constantly looming to pounce.

Can low-margin retailers afford customer loyalty programs?
Photo: iStockPhoto

It is therefore not surprising that many retailers have adopted loyalty programs as a convenient mechanism of meaningful differentiation. Ultimately, loyalty programs should offer incentives for shoppers to reduce their store switching by offering them better value. Loyalty programs have the added advantage of not taking much time to establish, making the threat of price wars from competing retailers less credible.

UNCERTAINTIES, AMBIGUITIES, AND DOUBT

The many advantages of loyalty programs have been well debated in the massive literature available on the subject. Yet, there remain uncertainties, ambiguities, and doubts in the minds of many retailers over whether loyalty programs offer a sustainable and profitable business model.

Retailers such as Sainsbury's in the United Kingdom illustrate this hesitation. They were early adopters of loyalty programs but dropped them, claiming they were too costly without compensating tangible benefits. (Recently they reinstalled a program in a market where all non-Every Day Low Pricing players offered them.)

A lot of this uncertainty is due to a number of myths. In this paper we want to clarify the most prominent of misunderstandings associated with loyalty programs in retailing. The six myths can be classified in three groups: (1) architecture of shopper reward (size and convexity of rewards), (2) architecture of retailer rewards (sustainable profitability and horizontal competition), (3) type of retail sector (frequently versus infrequently purchased goods).

We discuss each of these myths in turn.

SIZE OF THE REWARD

Loyalty programs do not really work in grocery stores because the rewards that grocery retailers can afford to offer are too small. They cannot offer better rewards, according to conventional wisdom, because of their razor-thin margins fueled by extremely competitive market conditions.

That doesn't stop many retailers from trying. Indeed, grocers like Tesco, Sainsbury's, Kroger, Safeway, and Stop & Shop give back 1-2 percent of the total spent to their card-carrying shoppers. Stores in other categories—Staples in office supplies, Barnes and Noble in books, Best Buy in electronics—offer similar rewards. Boots, the UK-based pharmacy and beauty supply retailer, offers a substantial reward of 4 percent, but that rate is more exception than rule.

This type of reward is actually a "thank you" to the consumer from the store for the privilege of collecting loyalty card customer data. As former Tesco CEO Sir Terry Leahy put it in his recent book referring to the 1 percent loyalty discount: "It was a thank you, pure and simple."

Can such small rewards convince consumers to favor one retailer over another? Will consumers bestow their loyalty by spending most of their grocery dollars with one retailer for 1-2 percent payback?

Let's look more closely at the actual rewards. They come in two forms : (1) turbo charged vouchers (doubling, tripling, or even quintupling the value of the voucher at partner companies such as hotels, restaurants, and theme parks and (2) "individual shopper targeted" offers, whereby each shopper regularly receives an extensive set of tailor-made coupons predominantly paid for by supplier brand-owners. The value of tailor-made coupons and their benefit to each individual shopper builds with the retailer's increasing understanding of the contents of the consumer's shopping basket, and relies heavily on the magnitude of dollars spent in the store.

The actual value for the shopper of turbo-charged vouchers and tailor-made coupons are far more substantial than the 1 percent "thank you" discount offered on the total purchases at Tesco. The more a consumer spends in the store, the more information the store is collecting about consumer tastes and shopping habits, allowing it to direct more relevant rewards toward such loyal customers.

Thus an increase in dollars spent in the store not only increases the dollar value of the coupons received by the customer, but also yields higher redemption rates reflecting the massively improved relevance of the tailor-made offers to each individual shopper based on loyalty card data. For example, at Tesco, redemption rates on individual tailor-made coupons vary between 10 percent and 20 percent as compared to the usual 1 percent or less redemption rates.

It is therefore not the 1-2 percent payback to the consumer that makes the loyalty program work for the grocer but rather the turbo-charged vouchers and the total value of tailored coupons that shoppers stand to lose if they go elsewhere. Store-switching costs created this way increase the spending and loyalty of shoppers.

For retailers lacking the capability of using information about their card-holding customers and turning these data into tailor-made individual shopper-specific propositions, loyalty systems are no more than an expensive gadget.

CONVEXITY OF REWARDS

It is necessary but not sufficient to offer substantial rewards to generate more loyalty from shoppers. If competing retailers copy these substantial rewards, shoppers will not be incentivized to stay loyal to only one retailer—promiscuous shoppers would not be worse off than extremely loyal shoppers if rewards were linear and uniform across retailers. In principal, rewards received by shoppers should be non-linear—spending $x at one retailer should offer the shopper a better reward than spending $x/n in different stores.

One approach to achieve this convexity in rewards is to create loyalty systems that are tiered. It is often argued that tiered rewards, such as those offered by airlines, hotels, casinos, and financial service firms, are necessary for loyalty programs to thrive.

Many companies in travel and service businesses reward frequent users by bestowing a status that allows them the privilege of high-value services. Take, for example, British Airways, which allows Executive Club members the privilege of access to airport lounges around the world. Services offered in the lounges—showers, spa, Internet connection, appetizing free food and drinks, along with the opportunity to rest during a layover—is exceptional value to members. There is even a higher class of lounge facilities for an even more frequent traveler in first class. Similar facilities, free upgrades, and better rooms are also offered by hotels.

Casinos designate their guests as silver, gold, and platinum members, rewarding frequent visits with an increasing level of attention and free services available in each class of service. In designing a tiered system at Harrah's, CEO Gary Loveman was very clear that he wanted to ensure that customers who visited another casino would lose something by not accumulating rewards with Harrah's: the opportunity to get to higher levels of rewards and services that have a material impact on the customer's experience.

Furthermore, tiered systems can stimulate customers to increase their purchases to reach the next stage-and therefore more and better rewards-in the hierarchy of loyalty classes.

Even though many believe that loyalty programs that lack this feature won't induce enough loyalty to be sustainable, no retailer offers a tiered loyalty program. Conventional wisdom has it that it is difficult for retailers to offer differentiated services or levels of prestige to their different levels of customer loyalty.

The belief that loyalty programs lacking explicit tiered reward structures are ineffective is a myth. We believe this to be a myth because successful application of loyalty systems by retailers goes beyond tiered rewards—they offer individually tailored rewards.

Tiered reward systems create some form of stratified shopper segmentation—a shopper belongs to the silver category or the platinum category. These predetermined categories are far from homogeneous and substantial variation exists between consumers in the same category. So tailor-made loyalty systems go way beyond these heterogeneous customer groupings by offering propositions at an individual customer level (truly 121 propositions).

Information technology has enabled retailers to do on a mass scale of millions of customers what shopkeepers did in their villages a century ago: offer a differentiated service to each shopper. Shopkeepers knew each one personally and treated them accordingly. If that customer went to another retailer who didn't know the shopper as well, he or she would receive an inferior service. Then as well as now, the more a shopper spends at a particular store, the better that retailer can offer specific value propositions that competitors can't match. Promiscuous shoppers will be disadvantaged by receiving the "vanilla service" designed for the average shopper.

While retailers do not offer tiered loyalty programs, they do offer tailored coupons and other offers valued at par with the total spent in the store. The convexity in the reward structure comes from continuous improvement in making the rewards more relevant and meaningful to the shopper, which is more likely with ever increasing data available on the customer's loyalty card.

SUSTAINABLE PROFITABILITY

As mentioned earlier, retailers are often thought of as being unable to afford substantive loyalty reward programs because of their low margins. Offering an unconvincing 1 percent loyalty reward to customers already represents a reduction of 30 percent to 50 percent of the retailer's net profits. When margins are low, the sustainability of loyalty rewards often depends on two possible sources: abundant excess capacity and exogenous subsidies.

Airlines, casinos, and hotels can offer significant rewards to their loyal customers because they often operate with substantial overcapacity. Rewards based on this overcapacity, such as air miles or stays in hotels and resorts, are highly valued by potential customers.

In regard to subsidies, retailers have two types of "sponsors" for their loyalty reward investments: manufacturers of the branded goods sold by retailers, and other partner companies presumably with higher margins, who find it economically advantageous to pay retailers to attract additional business from their shoppers (Disney, restaurants, theme parks, etc.). The former finance most of the targeted coupons while the latter contribute the lion's share of the financing of turbo-charged vouchers.

Since manufacturers already invest significantly in retailer promotions, targeted promotions are a welcome vehicle for brand owners to improve the effectiveness of these promotional investments. They are particularly attractive to challenger brands, because targeted coupon activities are extremely scalable—they can participate for a varying number of brands for a varying period of time. This scalability of the coupon activities makes them less attractive to leader brands. They prefer, ceteris paribus, heavier, less scalable promotional activities, like in-store demonstrations, big prize sweepstakes, and the like, which are less affordable for challenger brands. However, leader brands are forced to keep a fair share of these targeted coupon promotions as a defensive strategy against their challengers.

Overall, retailer-initiated loyalty systems are truly triple win marketing activities simultaneously benefiting retailers (more loyalty, more sales, better profitability given their high fixed costs); brand-owning suppliers (more effective couponing, and an offensive tool for challengers and a defensive tool for market leaders); and shoppers and consumers (more appropriate product and price propositions). Given these substantial advantages for all parties involved, retailer loyalty programs can be sustainably profitable across the whole value chain.

As the percent of retail sales associated with promotions increases over time, improving the effectiveness and the efficiency of promotional spending will continue to be of utmost importance. It is not that loyalty systems increase the amount of promotional activities in the retail industries; it is redirecting these promotions (coupons, vouchers, etc.) toward more effective and efficient uses.

UNINTENDED CONSEQUENCES

Since loyalty systems are beneficial for suppliers, retailers, and shoppers, these systems should help improve the retailer's image, but a number of unintended consequences can overshadow these positive effects for the retailer.

First, loyalty programs can negatively influence the consumer's price perception of retailers. This conclusion directly follows from the arguments presented above in that retailers are able to charge higher posted prices due to reduced price competition. However, a negative price image is again a myth because it does not distinguish between highly valued loyal consumers and less valued switchers. While the price image may suffer in eyes of the switchers, many will still patronize stores for reasons other than price. On the other hand, the most loyal consumers have a price perception that is significantly affected by the nature and value of the targeted offers they receive in exchange for their business, and therefore have a much more favorable price image relative to the competition.

Loyalty programs end up creating different realities for the most loyal and less loyal shoppers, emphasizing the different reasons for why customers shop at a particular store.

Some unintended consequences are not a myth; retailers need to consider these when they apply loyalty systems.

First, some shoppers who spend less, like older and lower-income consumers, may feel they are getting a rough deal. Individually differentiated prices tend to favor higher spending loyal shoppers. Less favored shoppers might create negative influences against the retail chain's image, especially if their story with issues of discrimination is taken up by the media or regulators.

Second, as there is more room for sales growth with consumers who are not yet top spenders compared with those who are already top spenders, strict application of loyalty system strategies enhancing sales and profit growth might result in the former shoppers receiving more appealing offers than the latter group. This might cause concerns of inequitable treatment from the point of view of your "best" shoppers.

Third, as every shopper encounters a set of prices that are different from prices provided to other shoppers, retailer price transparency might become very murky. Potentially, this can have two implications, one on retailer price image and the other on retailer competition. Since every shopper effectively faces different prices at the same retailer, how does "the market" form a price image of that retailer? Will the generic in-store prices determine the price image or will every consumer have a different price image of the retailer? Furthermore, if price offers, through couponing and vouchering, vary almost continuously over time, will the shopper truly understand the prices at a point in time at each retailer?

This complexity might lead to less price transparency for shoppers and result in more monopolistic behavior by retailers. In this case the regulator might have to step in to ensure shopper interests are guaranteed through proper competitive price offers.

LOYALTY SYSTEMS AND COMPETITION

"I am very store loyal; I hold loyalty cards from five different retailers!"

Shoppers simultaneously hold loyalty cards from competing retailers and often shop regularly at more than one competing store. Hence the belief that loyalty cards are ineffective because shoppers subscribe to multiple programs from competing retailers. This is particularly true of grocery retailing where consumers regularly shop in more than one, and often quite a few, grocery stores.

However, it is a myth to believe that loyalty programs are ineffective when consumers hold loyalty cards from competing stores. First, although consumers shop in a number of stores, the dominant spend is in the favorite store—typically, more than two-thirds of the total grocery spend occurs in the favorite store. For the shopper, the dominant store will have an informational advantage over less patronized stores.

As a result, this favorite store will be able to make far more appealing and targeted propositions to that shopper, reinforcing that customer's spending there. Of course, that shopper will continue to visit other stores for specific reasons, on the way to work, for example. One hundred percent loyalty to one store is rare and would be uneconomical for the retailer to aspire to.

Actually, loyalty systems are unique in the sense that copying of a loyalty system by direct competitors reinforces the strength of the loyalty generating capabilities of all participating retailers. In this case, each retailer will focus on its customers for whom it has an informational advantage rather than trying to attract shoppers who are loyal to another retailer. For the latter, the focal retailer has an informational disadvantage that makes it uneconomical to try to attract those shoppers. Furthermore, such a competitive move would probably trigger retaliation from the attacked retailer, who would subsequently try to steal the focal retailer's loyal shoppers. It can easily be seen that such a strategy would punish all players, except the shopper.

The condition for a favorable equilibrium for all loyalty driven retailers is that all players are reasonably satisfied with their current market share. If some players want to gain share at the cost of profitability, then loyalty systems can escalate into intense price wars from which shoppers win short term and the lowest cost retailer wins long term.

Loyalty systems are most valuable for Hi-Lo market leaders, players who target both ends of the spending spectrum. As loyalty systems are first and foremost effective in retaining customers, the market leader has the most to gain from using these systems. Market leaders, having more shoppers, also have an informational advantage over their followers. (It is more difficult to make loyalty systems effective for smaller followers in the market, because they tend to have fewer customers and therefore an informational disadvantage.)

Tesco's loyalty system in Poland is getting little traction because they lack "shopper critical mass"; they are a distant middle-of-the-pack player in the country. Although they try to use the same loyalty approach as in the UK, the Polish application cannot reproduce the benefits accrued to Tesco's UK market leadership position. The lack of success of Tesco's loyalty system in Poland is also in part due to a very strong market leader, Biedronka, which is a world class Every Day Low Price (EDLP) retail operation.

If one of the competitors is an EDLP player, the final outcome of the competition between the Hi-Lo loyalty retailer and the EDLP retailer is less predictable. In theory, and assuming all retailers want to maximize profits, the Hi-Lo retailer should be able to hold on to its shoppers for which it has an informational advantage and therefore for whom it can offer a more appealing offer. It will be difficult for the loyalty retailer to attract loyal shoppers from the EDLP retailer because it doesn't have the necessary shopper information.

The EDLP player can, however, be a real cause of concern for the loyalty retailer. The shopper population of the loyalty (Hi-Lo) retailer can be classified in three groups: (1) those redeeming their loyalty card offers, (2) those who are price sensitive but for some reason don't redeem their loyalty card offers, and (3) those who are loyal for non-price reasons. It is the second group that is vulnerable to aggressive offers from the EDLP competitor. The latter retailer might offer a price guarantee of lower prices for the total shopping bill of a minimum-sized shopping basket of comparable products. A significant proportion of the second group of shoppers might find this offer appealing and switch to the EDLP player. This is a concern for the loyalty retailer because this is probably a sizeable group. The loyalty retailer then needs to watch carefully its generic prices and in the meantime offer a price matching guarantee with the EDLP player.

Such a situation exists in the UK where grocer Asda, owned by Walmart, follows an EDLP approach and competes with several retailers that are loyalty Hi-Lo players. To attract shoppers from other (loyalty playing) retailers, Asda has offered a 10 percent price advantage guarantee for its shoppers and will reimburse shoppers if Asda is not 10 percent cheaper on the comparable shopper's total shopping basket. This has been effective to the extent that Tesco now offers a price-matching promise vis-à-vis Asda and other retailers, on top of its own loyalty program.

Normally, one would expect this price-matching guarantee to call an end to the price war. But other, more global, issues seem to be at stake and Asda continues to guarantee a 10 percent price advantage. For Asda, short-term profits don't seem to top its list of objectives.

LOYALTY SYSTEMS AND INFREQUENTLY PURCHASED GOODS

Another myth is that loyalty programs do not work in categories that are purchased infrequently or do not affect the behavior of customers who shop infrequently in the category. Consider the impact of frequent flyer programs for families that travel mostly for vacation. It might take them more than a year to accumulate enough miles to get a free airline ticket. Would a frequent flyer program affect the choice of airline when making the next reservation for a family vacation? In other words, it is difficult to make a loyalty program effective if the consumer does not believe that the rewards are achievable.

This myth is partly true because most retailers agree with the observation but have found ways to get around it. They have done so by either making the award more lucrative and therefore more salient in the minds of the consumers, or they help consumers accumulate rewards in other ways that could be redeemed in their stores. To solve the challenge of the infrequent vacation traveler, airlines introduced credit cards so consumers can earn points or rewards for all their spending on the card and advance to levels that make the rewards more easily achievable.

Another way to support a more lucrative program is by finding other partners to pay for the rewards. This is again possible if the data available through the loyalty program can be used in a meaningful way by other partners.

SUMMARY

Retailers seem to have an uneasy relationship with loyalty rewards programs. Many believe the programs are unsustainable in low-margin businesses or industries.

But as we have shown, these myths are just that, and a variety of loyalty programs can be fine-tuned by retailers to produce differentiation and increased profits. When location, location, location is no longer enough to attract customers, a program that rewards their loyalty certainly will be.

As Major League Baseball camps reawaken for spring training over the next few weeks, the same scene will repeat across the country: A pitcher will take the mound. A batter will stare back. And behind him, an umpire will tense in anticipation, ready to call a ball or a strike depending on where the pitcher throws the ball.

How much would you expect the race of the umpire and the pitcher to determine the outcome of the call? That's the question Christopher A. Parsons, Harvard Business School visiting associate professor in the Finance unit, poses in Strike Three: Discrimination, Incentives, and Evaluation, a paper published in 2011.

Parsons has made a habit of asking surprising questions about economic issues, looking at situations where small factors have an unexpected effect on the outcome. In addition to asking about baseball and race, he's explored questions such as how much money motivates ministers, and how much financial journalists affect the stock market.

"I am interested in what motivates people," he says, "and what causes them to make the decisions they do economically and otherwise."

In the cases of umpires and ministers, Parsons's work falls under the broader category of incentive research. "They are about what causes people to behave in predictable ways, and how that changes over time," he explains. Those causes and effects can have broader implications for business and economics as a whole.

EYES ON THE BALL

In order to determine the effect of racial discrimination on baseball games, Parsons and colleagues Johan Sulaeman of Southern Methodist University, Michael C. Yates of Auburn University, and Daniel S. Hamermesh of the University of Texas at Austin waded through pitching data for every Major League Baseball game from 2004 to 2008. When the data was analyzed, they found that indeed, race matters.

When umpires and pitchers are a different race or ethnicity, the umpire is slightly more apt to call a pitch a ball, which favors the batter. Although the effect is small—averaging about one pitch a game—it is consistent and statistically significant, ruling out simple human error.

A pitcher's race may influence how an umpire calls balls and strikes. Photo: iStockPhoto

What was more surprising, however, is how pitchers reacted to this trend. For a select number of games in 2007 and 2008 the researchers also analyzed where the pitchers threw the ball over the plate. They found that when the umpire and pitcher were of a different race or ethnicity, the pitchers were less likely to throw the ball to the edges of the strike zone—"painting the corners" in baseball parlance—and more likely to throw it straight over the plate where it is easier to hit.

In other words, the pitchers seemed to compensate for discrimination by throwing the ball in areas where the outcome was less subjective, even if it meant potentially hurting their own performance.

"That little bit of conscious or unconscious discrimination spills over into the entire game," concludes Parsons. "It might only directly affect a pitch or two a game, but indirectly it affects every pitch through tiny little changes. Ultimately, when a black pitcher is pitching to a black umpire, he is more likely to win the whole game."

Parsons speculates that the discrimination may not be conscious. Since the effects increase as the game goes on, he suggests the animosity that can grow between a pitcher and an umpire over time is magnified by race. A pitcher expressing disgust to an umpire can eventually tick the ump off. "Umpires have a shorter fuse for guys they don't like. If you are a black guy and I don't like black guys, and you shake me off or throw your glove down, now I'm going to get you."

The researchers didn't observe the same effect when umpires were a different race or ethnicity than batters, perhaps because batters are in front of an umpire for only a few pitches a game--pitchers square off face-to-face with umpires for many innings in a row.

The most significant finding in the paper to Parsons is when umpires don't discriminate. During the time the researchers investigated, MLB had installed cameras pointed at the strike zone in a third of ballparks as a way of monitoring umpires' accuracy. They found that when cameras were present, umpires made calls the same no matter the race of the pitcher. "They didn't install these cameras to look for racism, but it turns out those incentives really matter to umpires," Parsons concludes. When an umpire's job was on the line depending on how accurately he called strikes, then he was more likely to work hard to make his calls fair.

Findings from the study have broader implications beyond baseball, says Parsons. A person concerned about being discriminated against may be less likely to take chances, sticking instead to more objectively measurable tasks, in the same way that a pitcher stops taking chances by painting the corners and throwing straight over the plate. That could go for a police officer who chooses to spend time writing traffic tickets rather than conducting investigations; or a worker who chooses to push paper rather than pursue more entrepreneurial activities.

At the same time, it points to the need to monitor supervisors in the workplace for evidence of discrimination. "We know racism and sexism exist. The question is what can we do about it? Does evaluating the evaluators change discrimination? Does the fact that I know I'm being watched change my behavior?"

The answer, at least from this study is yes. "That is interesting," says Parsons, "because it implies that policy can change things."

MONEY IN THE PEWS

The same holds true for Parsons's study into how money motivates ministers, Is a Higher Calling Enough? Incentive Compensation in the Church, published in the July 2010 issue of the Journal of Labor Economics. Perhaps more than any other profession, we hold religious leaders above such base incentives as money. At the same time, it's the rare individual for whom money isn't at least part of a factor in the work they do.

"Just because a person has intrinsic motivation, that doesn't crowd out other incentives," says Parsons. "I think it's a safe call to say that not many ministers get into the ministry to make a buck, but that doesn't mean a monetary incentive for doing a good job doesn't matter."

Parsons and fellow researchers Jay C. Hartzell of the University of Texas at Austin and David L. Yermack (HBS MBA1989/JD1991, PhDBE 1994) of New York University wondered to what degree money was a factor in minister performance, and what it could tell us about pay for performance in general.

To answer those questions, they relied on data from the Methodist Church in the Oklahoma conference, which meticulously collected numbers on all aspects of the ministry, including pay, size of churches, and number of baptisms and conversions over a 43-year period.

They found a direct correlation between the number of parishioners added to a congregation and the amount the minister was paid. In fact, for every member added, a minister was paid an average of $14.71 in inflation-adjusted dollars. And not all new parishioners were valued equally. A conversion from another faith was worth $17.78—but a defection from another Methodist church in the area was worth nearly double that, $32.51.

Objectively, that discrepancy makes little sense. "Those are the easiest people to recruit, since they are already Methodists," says Parsons. From the perspective of members of the congregation's council that is setting minister pay, however, it stands to reason they would be worried about losing members to neighboring churches and weigh those recruits more highly.

For the national Methodist church, on the other hand, it doesn't matter which local congregation a Methodist belongs to. "It's like one division of Google stealing from another," says Parsons. "If one is paid only on the profits of his division, then he has incentive to poach people from another division. You don't want that."

Within the Methodist church, at least, there is a check on this sort of behavior: promotions. The national church rotates ministers between churches every three to five years. When that occurs, Parsons and his colleagues found, ministers who "poached" were less likely to be assigned to the highest-paying congregations.

In a broader context the method of the Methodists spells out the necessity for any organization to structure performance pay in a way that properly aligns incentives.

"Even for groups that you expect would not be influenced by incentives, they are extremely influenced by incentives," says Parsons, "and different parties within an organization can have vastly different incentives."

BEARS AND BULLS IN THE NEWS

While not in the realm of incentives, per se, another of Parsons's papers also shows how extremely open to influence people can be.

It stands to reason that news can influence events as much as it reports on them—but what about the way news is reported? In an analysis of the Wall Street Journal's "Abreast of the Market" column, Parsons found that even savvy investors are susceptible to influence, in a way that can have real effects on the stock market.

"Even the most sophisticated players in the world are not completely immune from framing and interpretation," Parsons says. The column is like a weather report for financial news, reporting on the events in the market on the previous day. It's not always the same weatherman doing the reporting, however; rather, the paper employs a rotating cast of columnists who break down the day's financial highlights.

In Journalists and the Stock Market, published in the 2012 the Review of Financial Studies, Parsons analyzed the effects of these different perspectives along with colleagues Casey Dougal of Drexel University, Diego García of the University of North Carolina at Chapel Hill, and Joseph Engelberg of the University of California, San Diego.

The researchers used computer software to analyze the tone of each columnist's writing, determining whether they wrote with a generally bearish or generally bullish perspective. "Some authors are always Debbie Downers, and some authors always pump up the market," says Parsons.

Then they recorded how the stock market performed on the days following each article's appearance. As it turned out, the market consistently dropped by several points after a bearish columnist wrote, and consistently rose by several points following a bullish columnist. "I describe the weather and you describe the weather and people dress differently for it," says Parsons by way of analogy.

What was most interesting was that the rise and fall seemed to have nothing to do with the content, but everything to do with the tone with which it was presented.

"This is not like something where a journalist would have any information on the overall movement of the market—professionals can't predict the market, let alone some random journalist," says Parsons. "What I found most surprising is that the framing of the day's events mattered as much as the events themselves."

Most of the market results were short-lived, disappearing after a couple of days, so the journalists didn't have any lasting effect on market value. Even so, the study provides a cautionary tale: Investors, as well as the public at large, should be as conscious of how reporters relay the news as they are about the news they relay.

Throughout his research, Parsons forces us to look beyond the generally perceived wisdom—that umpires are impartial, ministers are nonmaterialistic, and journalists are objective—to discover the hidden incentives and influences that change behavior. Such "curve balls" can teach us to look more deeply into these hidden biases and how to change them.

"Goodbye tension, hello pension!"

That used to be the triumphant cry of millions of new retirees. For decades, Americans assumed a good job came with a good pension, guaranteeing them regular monthly payments from their parent company until the day they died. The plans were also known as "defined benefit" plans because they assured recipients of a set monthly amount they could always rely on.

Then, starting in the 1980s, the nest egg started to crack. As firms began competing globally, pension perks began to be seen as very expensive liabilities. Within a few decades, nearly all corporations ceased offering defined benefit plans in favor of "defined contribution" packages such as 401(k) plans—in which employees contribute a set amount from their paychecks that would be individually invested for their retirement.

Although companies often match those contributions, they are under no obligation to continue to do so after retirement, and employees can't rely on a predetermined monthly amount the way they could with earlier benefit plans.

"That world has been disappearing," says Luis M. Viceira, the George E. Bates Professor at Harvard Business School. "In the past few years, there have been zero defined benefit plans created in the United States. The trend, at least in the corporate world, has moved very quickly. "

Companies that traditionally offered defined benefit plans have closed them to new hires, and even frozen them for existing employees. Just look to Seattle, where Boeing workers on January 3 narrowly ratified a contract that will convert their traditional pension plans for newly hired machinists to a 401(k) program. Boeing threatened to move construction of the 777X jetliner to another state without the concession.

FUNDING THE FUTURE

Even so, companies are still on the hook for paying benefits to those employees who have already been promised them. As their workers age, employers face the difficult question, How are we going to make good on those promises?

Pensions are a costly legacy for many companies.
Photo: iStockPhoto

The question is particularly urgent now, says Viceira, who teaches in the area of investment management and capital markets. For starters, the financial crisis depleted many pension plans by dramatically reducing the value of investments, even while companies were still responsible for paying predetermined benefits.

Increasing the pressure are two other factors. Life expectancy has increased, adding to the length of time corporations are required to pay. And interest rates have fallen to historic lows, increasing the funding that companies must set apart to make up for the lower yield on the assets already in place.

"Companies have had to increase their contributions exponentially as interest rates declined," says Viceira. That strain was a major factor in bankruptcies in the steel, airline, and car industries. More and more, companies are looking for a way out of pension plans, while still making good on their obligations.

They have three choices, says Viceira. The first is to do nothing and continue to invest in equities, hoping the numbers will work out. The second is to work a deal with employees for a lump-sum payment covering the value of their pension, walking away without further obligations. That number can be large, however, and few companies can afford to pay out all that money at once.

REDUCING PENSION PLAN RISK

The final option is for companies to "de-risk" their pension plan by putting assets into more predictable investments that generate enough income while still reducing the risk due to market or interest rate volatility. To do that, some companies are turning to the experts in evaluating risk: insurance companies.

In the HBS case study Prudential Financial-General Motors Pension Risk Transfer: Back to the Future?, Viceira, with Emily A. Chien, wrote about the historic de-risking of GM's pension plan for salaried employees, a $25 billion deal negotiated last year. GM transferred its assets to Prudential, which then promised to make good on the benefit payouts in the form of guaranteed annuities.

The deal made sense—after all, who better than insurance companies to estimate life expectancy and long-term risk. And by pooling the GM annuities along with its wider population of beneficiaries, Prudential could manage risk better than the automaker could on its own.

That doesn't mean the deal was without its challenges. The two companies had to decide who was going to assume the assets and the liabilities—was GM going to "buy-in" annuities from Prudential while still maintaining full responsibility for paying out the pensions; or was it going to "buy-out" the annuities, transferring both the assets and the liabilities of the plan to Prudential's own balance sheet? The decision would determine who was ultimately on the hook if either company went under.

Eventually, the companies agreed that GM would buy-out the plan by transferring the assets and liabilities to Prudential. But even that had to be done carefully, since selling the assets in the pension plan all at once to buy annuities could potentially affect the value of what they were worth. Finally, there was the overall price of the deal—requiring months of complex number-crunching to determine the value of the investments and cost of the pay-outs over time.

But both companies had incentive to come to an agreement—which they eventually did for an undisclosed sum. GM removed an uncertain liability from its balance sheet, and Prudential got to take a piece of the pension business away from the asset management companies that have traditionally handled those investments.

"Even though [defined benefit plans] are dying elephants, it's going to take a long time for those elephants to die," says Viceira. "Insurance companies are now back in the game of managing these assets, some of which we might see moving from the BlackRocks of the world to the Prudentials of the world."

Another potentially lucrative target being considered by insurance companies are public pensions, "the huge elephants that are very much alive and kicking," says Viceira.

MORE BUSINESS AHEAD

Time will tell if insurance companies are able to stay in the game long term, but so far other pension-pressed firms have shown interest and followed GM's lead; Verizon, for example, also completed a deal with Prudential. Other companies have been stopped in similar migrations only because their plans are not fully funded—but that could change with a moderate rise in interest rates, sending more elephants stampeding into the waiting arms of insurance companies.

"If interest rates go up and [companies] find themselves fully funded or over funded, they will start going for these deals," says Viceira.

While corporate pensions may ultimately go extinct, these kinds of deals may ensure that currently existing pension liabilities will continue to be paid well into the future.