The Effect of Superstitious Beliefs on Performance Expectations

J. of the Acad. Mark. Sci. (2009) 37:161–169 DOI 10.1007/s11747-008-0116-y

ORIGINAL EMPIRICAL RESEARCH

The effect of superstitious beliefs on performance expectations
Lauren Block & Thomas Kramer

Received: 7 January 2008 / Accepted: 15 August 2008 / Published online: 9 September 2008 # Academy of Marketing Science 2008

Abstract We explore superstitious beliefs as a basis of product performance expectations and their impact on initial purchase likelihood and subsequent satisfaction. In doing so, we demonstrate instances when superstitiondriven expectations cause consumers to make purchase decisions that run counter to economic rationality. In the first set of studies we find that Taiwanese consumers are relatively more likely to purchase a product with positive superstitious associations based on its “lucky” color, and are more likely to purchase and are willing to pay more money for a product with a smaller but “lucky” number of units contained in the package (e.g., eight tennis balls compared to ten). In contrast, consumers who do not hold such superstitious beliefs adhere to the more rational choice paradigm. Next, we show that the differences in purchase likelihood are driven by superstition-based performance expectations. We further generalize these findings to product satisfaction, and find support for expectation disconfirmation sensitivity as a moderator of the effect. Keywords Superstition . Consumer behavior . Irrational beliefs . Performance expectations

Lauren Block and Thomas Kramer contributed equally and are listed in alphabetical order. The manuscript benefited greatly from the insightful comments and suggestions received from the editor and four reviewers. L. Block : T. Kramer (*) Department of Marketing and International Business, Zicklin School of Business, Baruch College/CUNY, Box 12-240, One Bernard Baruch Way, New York, NY 10010, USA e-mail: [e-mail] L. Block e-mail: [e-mail]

Consumer expectations play an important role in marketing because of their impact on initial purchase decisions, satisfaction judgments, and subsequent repurchase behavior (e.g., Kopalle and Lehmann 2006; Oliver 1980; Oliver and Bearden 1985). Research has shown that expectations can be based on variety of factors, including advertising or published quality ratings (Kopalle and Lehmann 1995), trial (Goering 1985), company promises and word-of-mouth (Zeithaml et al. 1988). However, consumers’ superstitious beliefs as drivers of expectations have only received scant attention in the marketing literature. This gap in the literature is even more surprising given how frequently marketers rely on superstitions in their communications, thereby potentially creating or changing consumer expectations. For example, prominently featuring a number perceived to be lucky in Western cultures, the Ritz Carlton Hotel in New York offered a July 07, 2007 wedding package with a reception for 77 guests, a seven-tier wedding cake, seven Tiffany diamonds for the bride, and a seven-night honeymoon at any Ritz Carlton in the world for $77,777. On a smaller scale, Wal-Mart ran a “Lucky in Love Wedding Search” contest, in which seven lucky couples were selected to get married in the Wal-Mart lawn and garden area of their local Supercenter on July 07, 2007 and received a wedding reception for 77 guests. Finally, Icelandair recently ran a “Lucky 7s $7” promotion (similar to Continental Airlines’ “$888 to Beijing” campaign), which allowed customers to add on several excursions in Iceland for $7 each—as long as they were booked by lucky July 07, 2007. The economic boon of lucky numbers is not limited to the US and Western cultures. For example, the number 8 is considered lucky in Chinese cultures; astonishingly, at a government auction of license plates in Guangzhou, China, the most expensive plate (AC6688) went for 80,000 yuan, which is noteworthy not only because it is roughly 11 times

162

J. of the Acad. Mark. Sci. (2009) 37:161–169

the country’s per capita income (Yardley 2006), but also because the auction price itself contains the lucky number 8. Importantly, while these are examples of positive economic influences of superstitious beliefs, note that negative superstitious beliefs can lead to severe economic losses. The long-term cost to the airline industry alone due to superstitious beliefs is substantial, with an estimated 10,000 fewer people in the US flying on Friday the 13th (Shields 2008). The lost business due to the fear of flying and decreased daily business transactions in general on this “unlucky” day is $800–$900 million every time the date occurs (Palazzolo 2005; Shields 2008). Yet despite marketers’ reliance on, and marketplace implications of, superstitious beliefs, there is surprisingly little academic or empirical research that systematically studies how such beliefs impact consumer behavior, or more specifically, if or under which conditions superstitionbased expectations influence purchase intentions and product satisfaction. Yet, such an investigation is not only important theoretically, but also managerially, given that it can provide marketers with key recommendations (such as raising expectations by systematically changing product colors, changing digits in product prices or adjusting the product quantity sold in a set to make them appear “lucky”) to increase choice share of their products and brands or to manage customer satisfaction. Importantly, however, superstitious associations likely represent a double-edged sword for marketers. In particular, if superstitions set up higher performance expectations, consumers may be more likely to purchase the product initially. Yet, these higher performance expectations will be relatively more difficult for marketers to meet or exceed, potentially resulting in lower satisfaction and lower repurchase intentions. We begin in the current research with the impact of superstition-based expectations on purchase likelihood and demonstrate that Taiwanese consumers are more likely to purchase a product in a “lucky” versus neutral color. Next, we find a greater purchase likelihood of a “lucky” product with fewer units contained in the package (e.g., eight tennis balls compared to ten). However, the effect of superstition on purchase likelihood is shown to depend on the degree to which consumers actually hold particular superstitious beliefs. That is, consumers who do not hold superstitious associations with the number 8 adhere to the more rational choice paradigm. Interestingly, we also find that Taiwanese consumers are willing to pay significantly more for a package with fewer units. Importantly, we provide support for our hypothesis that differences in performance expectations drive the effect of superstition on purchase likelihood. We generalize these findings to the impact of superstition-driven expectations on product satisfaction ratings, and furthermore find support for expectation disconfirmation sensitivity as a moderator of the effect. These results

suggest that satisfaction for “lucky” products is independent of positive versus negative expectation disconfirmation— that is, independent of actual product performance. Next, we review relevant previous research and develop our hypotheses, followed by a discussion of a series of studies that tested our predictions. We then discuss the implications of this research for marketers.

Superstitious beliefs Superstitions are beliefs that run counter to rational thought or are inconsistent with known laws of nature (Vyse 1997). Superstitions can be classified as either cultural or personal, and are invoked either to bring good luck or to fend off bad luck. For example, cultural superstitious beliefs likely to impact consumer behavior include the number 8 bringing good luck and the number 4 bringing bad luck in Chinese cultures, whereas the number 7 and 13 are associated with good luck and bad luck, respectively, in the US and other Western cultures. Examples of personal superstitions or rituals relevant to marketers include consumers’ buying and wearing lucky accessories, like charm bracelets, lockets, pens, or cufflinks. Consumption rituals also include using a particular product before an important event that is associated either with high likelihood of failure or a high level of uncertainty (for example, a sports game; Case et al. 2004). Additionally, the degree to which consumers rely on superstitious beliefs in their consumption decisions is likely to depend on the associated level of stress, risk or uncertainty (Keinan 2002; Malinowski 1954). For example, Keinan (2002) found that residents living in areas more likely to be exposed to a missile attack during the Gulf War were more superstitious (that is, were more likely to engage in “magical thinking”) than those living in safer zones. Such an increase in proclivity toward superstitious thought has been found in times of economic uncertainty as well, for example during the great depression (Padgett and Jorgenson 1982). The prevailing view is that resorting to superstitions provides a sense of control, or at least explains why control is not possible (Dudley 1998). Thus, extant literature explains how individuals rely on superstitious thinking and engage in superstitious rituals expecting that performing this ritualistic behavior will bring them luck, or at least ward off bad luck. That is, positive superstitions (taking an exam with a lucky pen) may set up higher performance expectations (receiving a better grade). Conversely, negative superstitions (taking a taxi with an unlucky license plate number on the day of one’s college entrance exam; Yardley 2006) might set up lower performance expectations (failing the exam). Yet, how does this translate to the marketplace? Do products with superstitious associations (such as having a lucky color or price) set up

J. of the Acad. Mark. Sci. (2009) 37:161–169

163

expectations as to how they should perform? Interestingly, in a prior study by the current authors (Kramer and Block 2008), we find that superstitious associations (e.g., lucky numbers and colors) influence consumer behavior to a great extent, and furthermore often do so through an automatic process. Specifically, they found that following product failure, consumers were less (more) satisfied with a product for which they hold positive (negative) superstitious associations. However, while we suggest that superstition’s influence on satisfaction might work through an increase or decrease in product performance expectations, we do not provide any direct evidence or test of this presumption. Thus, a direct test of performance expectations fills in a missing link in the process chain by which superstitious beliefs operate on consumer judgments. Furthermore, we limited our study to the effect of superstitious beliefs on product satisfaction. Although satisfaction with products is, of course, a crucial measure of marketplace success, managers are also fundamentally concerned with the purchase decision itself. However, the effect of superstitious beliefs on purchase has not yet been examined. Therefore, in addition to investigating performance expectations as the underlying process driving the effect of superstitious beliefs on satisfaction, the current paper also seeks to test the impact of superstitious beliefs on the purchase likelihood of products. Specifically, we propose that product attributes with superstitious associations influence performance expectations that in turn determine purchase likelihood. For example, consumers who are deciding between two identical products that only differ in color may choose the one that comes in a “lucky” color over others in a neutral color. Generally, we propose that consumers will be more likely to purchase a product with which they have positive superstitious associations, as compared to a product with which they do not. Furthermore, we propose that differences in purchase likelihood are driven by superstition-based performance expectations. That is, similarly to consumers who engage in superstitious rituals because they are associated with improved performance, such as wearing a lucky outfit to a job interview or taking exams with a lucky pen, we suggest that performance expectations are also transferred to products with attributes that are associated with superstitious beliefs. Based on the above, we hypothesize: H1: Consumers will be more likely to purchase a product for which they hold positive (vs. neutral) superstitious associations. H2: Differences in purchase likelihood are driven by superstition-based performance expectations. We test our hypotheses in a series of studies with Taiwanese consumers. Being of Chinese cultural back-

ground, Taiwanese individuals tend to be more superstitious than their Western counterparts (Simmons and Schindler 2003; Tsang 2004).

Study 1 The objective of study 1 was to test H1 and H2; namely, to test of superstitious associations impact purchase likelihood and to provide evidence for superstition-based performance expectations driving this effect. We chose the color red as the lucky product attribute based on previous findings in the literature demonstrating that consumers perceive this color to be lucky (e.g., Kramer and Block 2008). Method Participants and procedure Forty-four students from a Taiwanese university participated in a study on consumer preferences. Subjects read a scenario in which they were told to assume that they were going to the store to buy a new rice cooker and were considering a model that was described in terms of three features (capacity, timer, keepwarm time) that were kept constant between conditions. The fourth attribute, color, differed between conditions, and was either red (positive superstitious beliefs) or green (neutral), depending on randomly assigned condition. Subjects then rated how likely they were to buy the rice cooker, where 1=not at all likely to buy, and 7=very likely to buy. Next, to assess subjects’ performance expectations, subjects answered the following question: “In general, how do you expect the rice cooker to perform,” where 1=poorly and 7=well. Finally, subjects indicated their agreement (where 1=strongly disagree and 7=strongly agree) on the following scales, “red is a lucky color,” and “green is a lucky color” to ascertain that participants differed in their superstitious beliefs concerning the two colors chosen for the study.

Results Subjects differed in the positive associations they had concerning the two colors [M=5.23 vs. 4.11 for the color red vs. green; t(43)=3.37, p