1 Introduction
“I would never sell my engagement ring or grandma’s clock or…” The sentiment is common; people balk at putting a price on symbols of love. But applying the same prohibition to purchases could be costly. We explore the financial consequences of failing to make cost-benefit trade-offs for purchases associated with loved ones.
Our studies examine how trade-offs between love and money in wedding and funeral contexts lead decision makers to abandon cost-benefit reasoning, focusing their decisions on product quality over monetary savings. Moreover, we show that a reluctance to put a price on symbols of love extends to more common purchase events, such as bringing dessert to a party. The inquiry highlights an important consequence of decisions that demand pricing love: people spend more than is otherwise necessary.
1.1 The case of weddings and funerals
Wedding- and funeral-related decisions occur frequently in the marketplace – and are among the most costly in a consumer’s life. An estimated 2.1 million weddings and 2.4 million funerals occur annually in the U.S. (Reference Tejada-Vera and SuttonTejada-Vera & Sutton, 2010). The average U.S. funeral costs $6,500 (and more than $10,000 with burial costs) and the average U.S. wedding costs over $27,000 – totaling roughly $20 billion and $42 billion spent yearly on funerals and weddings, respectively (AARP, 2000; FTC.gov; Reference GlaserGlaser, 2009; Reference RheaultRheault, 2007; Reference TheThe Wedding Report. 2010).
Wedding and funeral decisions are subject to numerous social and cultural expectations (Reference Bonsu and BelkBonsu & Belk, 2003; Reference LadermanLaderman, 2003; Reference MeadMead, 2007; Reference Ratner and KahnRatner & Kahn, 2002; Reference RichinsRichins, 1994). Consumers seem especially vulnerable to sales and marketing tactics used by the wedding and death care industries (Reference BodenBoden, 2003; Reference Fan and ZickFan & Zick, 2004; Reference Kopp and KempKopp & Kemp, 2007a, 2007b; Reference MeadMead, 2007; Reference MitfordMitford, 1998). Wedding consumers, for instance, receive substantially higher price quotes for products and services (e.g., cakes, photographers) than for identically-described products and services for a birthday (Reference BrowneBrowne, 2009). We highlight another challenge of wedding- and funeral-related purchases (and acquisition decisions for sacred purposes more generally): wedding and funeral consumption contexts exemplify situations in which people put a price on love.
1.2 Sacred items
The sacred is set apart and transcends the mundane particulars of life (Durkheim, 1925/1976). The Constitution is more than a parchment with words, just as the Mona Lisa is more than a painted canvas. Sacred items are not limited to the extraordinary; commonplace items, such as cars or clothes, can also symbolize sacred values related to love, youth, or autonomy (Reference AcquavivaAcquaviva, 1979; Reference Belk, Wallendorf and SherryBelk, Wallendorf & Sherry, 1989).
Distinctive behavioral patterns emerge when people make judgments and decisions about the sacred (see Bartels et al., 2015, Reference Fiske and TetlockFiske & Tetlock, 1997). Decisions involving the sacred elicit judgment errors, such as the omission bias (Reference Ritov and BaronRitov & Baron, 1999), and greater negative emotions (Reference Hanselmann and TannerHanselmann & Tanner, 2008; Reference Baron and RitovBaron & Ritov, 2009). Sacred items tend to elicit large and erratic selling prices (Reference McGraw, Tetlock and KristelMcGraw, Tetlock & Kristel, 2003) and less hedonic adaptation (Reference Yang and GalakYang & Galak, 2015). The central characteristic of sacred values is their exemption from trade-offs with the secular (e.g., money; Reference Baron and SprancaBaron & Spranca, 1997). Trade-offs between the sacred and the secular are treated by decision makers as “taboo”. In economic terms, the marginal rate of substitution of a sacred good for a secular good is infinite; no amount of money can substitute for the Mona Lisa.
1.3 Responses to taboo trade-offs
Taboo trade-offs – such as putting a price on love – trigger identity threats and distress, which people are motivated to avoid (Reference Tetlock, Kristel, Elson, Green and LernerTetlock, Kristel, Elson, Green & Lerner, 2000). People protect the sacred from trade-offs with the secular (e.g., money) in a variety of situations (e.g. selling heirlooms; money for hostages; end-of-life care; Reference Baron and SprancaBaron & Spranca, 1997; Reference Tetlock, Peterson and LernerTetlock, Peterson & Lerner, 1996; Reference TetlockTetlock, 2003). Forgoing cost-benefit reasoning prevents the sacred from being reduced to the status of a commodity (e.g., “no amount of money would make me sell X”; Reference Baron and SprancaBaron & Spranca, 1997; Reference McGraw and TetlockMcGraw & Tetlock, 2005). Blocking an exchange of goods reduces negative emotions and protects against being “caught” undervaluing the sacred (Reference McGraw, Tetlock and KristelMcGraw, Tetlock & Kristel, 2003; Reference Tetlock, Kristel, Elson, Green and LernerTetlock et al., 2000).
Consumer research also identifies how people switch from compensatory to non-compensatory decision strategies when facing distressing (typically taboo) trade-offs (Reference Bettman, Luce and PayneBettman, Luce & Payne, 1998). People will weigh costs (e.g., car price) against attribute quality (e.g., styling) until saving money could endanger lives (e.g., compromising on safety equipment); then people focus primarily on quality over price (Reference Luce, Payne and BettmanLuce, Payne & Bettman, 1999, 2000). For instance, people choosing between apartments will avoid cost-benefit trade-offs by choosing to live in a safe neighborhood with little concern for costs (Reference Luce, Payne and BettmanLuce et al., 1999). Importantly, Luce and colleagues find the effect holds when controlling for attribute importance.
1.4 Inquiry
Money is a prototypical secular value, and monetary trade-offs are particularly threatening when sacred values are in play (Reference Baron and SprancaBaron & Spranca, 1997; Reference TetlockTetlock, 2003). We investigate people’s reluctance to engage in monetary trade-offs for products whose acquisition is symbolic of a sacred value: love. Our studies find that a reluctance to put a price on love causes people to spend more money than the consumption context requires because — all things being equal — they forego lower price options (Study 1), search less for lower prices (Study 2), and avoid negotiating lower prices (Study 3).
2 Study 1A
Study 1 investigates how trade-off avoidance could lead to overspending on engagement rings.
2.1 Method
Twenty-one undergraduates (47% female, M age = 21, SD = 1.4) participated for course credit. Gender as an independent variable or covariate yielded no differences (ps >.250).
Participants engaged in matching-choice tasks for an important wedding-related purchase (Reference Luce, Bettman and PayneLuce, Bettman & Payne, 1997; Reference SlovicSlovic, 1995; Reference Tversky, Sattath and SlovicTversky, Sattath & Slovic, 1988). The study varied three attributes of the engagement rings: price, carat weight, and color (on a 10-point scale from “Faintly tinted, usually yellow” to “Colorless 1”, see Appendix for example stimuliFootnote 1). Larger carat weight and more colorless diamonds are higher quality. We presented each participant twelve pairs of rings by factorially varying low and high levels of two attributes to create three trade-off types (price-carat, price-color, carat-color; see Table 1). Pairs of rings were presented in a fixed random order.
Note: Results from Study 1A matching-choice task displayed. For each pair of rings, the median judgment for the missing attribute value (filled-in by participants) is italicized. The average percentage choice for each pair of rings (with the now filled-in attribute values) is displayed. Higher values for color and carat indicate a higher quality diamond. Aggregate choices by matching-choice task trade-off type are displayed in the last three rows. When price is included in the matching-choice task, the choice of option B represents the higher price option; when price is not included (i.e., color vs. carat), option B represents the larger carat option.
* p < .05.
First, participants matched all pairs of engagement rings by providing a value for a missing attribute to make the two options equally appealing (Table 1). For example, option A was .5 carat and cost $1,275; option B was 1.25 carat with a missing price. The procedure accommodates idiosyncratic differences across participants, thus making latter choices comparable. Then, participants selected the ring they preferred in each of the twelve now-matched pairs (i.e., pairs with the filled-in values). Finally, after selecting rings, participants rated the importance (1 = very low, 7 = very high) of the carat, color, and price attributes when choosing diamond engagement rings (see Reference Luce, Payne and BettmanLuce et al., 1999).
Although matching requires cost-benefit reasoning between high and low levels of an attribute, the choice task allows for trade-off avoidance. If respondents willingly engaged in price-quality trade-offs after matching, they should choose either option with about equal frequency (Reference Carmon and SimonsonCarmon & Simonson, 1998). For trade-offs involving price (price-carat or price-color) we expected participants to avoid trade-off reasoning by choosing the high price, high quality option. As a point of comparison, we examined decisions absent price considerations (i.e., rings that differ only on quality; color-carat).
2.2 Results and discussion
We collapsed pairs of engagement rings into two categories: price-quality trade-off (four price-carat, four price-color) and quality-quality trade-off (four color-carat) pairs of engagement rings. For price-quality trade-offs, on average 89% of participants chose the high cost option (χ2(1,241) = 56.89, p < .001). For quality-quality trade-offs, participants did not choose differently from chance (58% chose colorless diamonds, 42% chose large carat diamonds, χ2(1,79) = 2.09, p = .148). Eleven of twelve trial level comparisons conformed to our hypotheses (Table 1).
In order to control for differences in attribute importance, we conducted a logistic regression with attribute ratings and trade-off type (price-quality vs. quality-quality) as simultaneous predictors. The effect of trade-off type on choice (Wald χ2(1,244) = 27.89, p < .001) remained significant, controlling for carat importance ratings (Wald χ2(1,244) = 5.97, p = .015), price importance ratings (Wald χ2(1,244) = 3.92, p = .07) and color importance ratings (p = .250). These results cannot be accounted for by the prominence hypothesis — wherein the more important attribute looms larger (serves as a tie-breaker) in the choice portion of a matching vs. choice procedure (Reference Tversky, Sattath and SlovicTversky et al., 1988). This is because participants indicate that price and carat are equally important (5.62 vs. 5.24; respectively, p > .40) and color is less important than both (4.14, ps < .05). The results suggest trade-off avoidance causes people to select equally appealing, higher quality options in a wedding context – but only when money is involved in the purchase decision.
3 Study 1B
We examine whether Study 1A’s effects replicate with an experienced population of wedding show attendees (Reference Alba and HutchinsonAlba & Hutchinson, 1987; Reference Bettman and ParkBettman & Park, 1980).
3.1 Method
A researcher approached people as they entered a wedding show. Twenty-two participants (86% female, M age= 31, SD = 9.1) agreed to participate. Eighty-two percent of participants indicated that either they, or their partner, had previously purchased a diamond engagement ring. Gender as an independent variable or covariate yielded no differences in forthcoming analyses (all ps >.200)
Participants completed a matching-choice task for one pair of engagement rings. Half of participants were randomly assigned to the price-quality trade-off condition, where option A cost $1,275 and was .5 carats, and option B cost $3,000 and carat size was missing. The other half of participants were assigned to the quality-quality trade-off condition, where option A was an eight on the color scale and .5 carats, and option B was a four on the color scale and carat size was missing. (Higher numbers on color scale indicate higher quality; see Appendix). Participants matched rings by providing the carat size that made the rings equally appealing (Table 2). Then participants selected their preferred of the now-matched rings.
Note: Results from study 1B’s matching-choice task displayed. For both pair of rings, the median judgment for the missing attribute value (filled-in by participants) is italicized. The average percentage choice for each pair of rings (with the now filled-in attribute values) is displayed. Higher values for color indicate a higher quality diamond.
* p < .05.
3.2 Results and discussion
Results were consistent with study 1A (see Table 2). For price-carat trade-offs, the participants reliably selected higher price diamonds (91% vs. 9%, binomial test p = .012). For color-carat trade-offs, participants did not reliably choose differently from chance (55% chose colorless diamonds, 45% chose large carat diamonds, binomial test p > .250). A logistic regression confirmed that price-quality trade-off type increased probability of choosing the larger carat ring (Wald χ2(1,20) = 4.21; p = .040).
4 Study 2
Study 2 investigates people’s search behavior when purchasing items symbolic of love (see Reference Ehrich and IrwinEhrich & Irwin, 2005). We expected a lower willingness to search for lower prices when making sacred purchases.
4.1 Method
Ninety-five undergraduates participated for course credit. We randomly assigned participants to conditions in a 2 (Purchase type: sacred, funeral vs. secular, non-funeral) x 2 (Search variable: lower price vs. higher quality) between-subjects design. Depending on purchase condition, participants imagined they were purchasing a container either for the cremation of a loved one (sacred purchase) or for the storage of a grandfather clock (secular purchase). Depending on search variable condition, participants saw that the container was high price or low quality. For the high price container, participants saw an $80, higher quality (pine wood reinforced fiberboard construction) container. For the low quality container, participants saw a $49, lower quality (triple-walled corrugated cardboard construction) container. Depending on respective condition, participants indicated their willingness to search for a lower price or higher quality container (1= “Not at all”, 7 = “Extremely”) and the importance of the search variable (1 = “Not at all Important”, 7 = “Extremely Important”).
4.2 Results and discussion
Participants were least willing to search when seeking lower priced alternatives for a sacred purchase (see Figure 1). An ANOVA on willingness to search revealed a main effect where people were more willing to search for higher quality than lower priced alternatives (F(1,91) = 6.54, p = .012), no main effect of sacred versus secular purchase (F(1,91) = .45, p > .250), and a significant purchase type by search variable interaction (F(1,91) = 20.22, p < .001). To test whether participants were least willing to seek out lower priced options for a sacred purchase, we compared the sacred-price condition to the other three conditions (i.e., secular-price, sacred-quality, secular-quality) in a planned contrast; in the sacred-price condition, participants were less willing to search for better (i.e., lower price) alternatives (F(1,91) = 13.69, p <.001).
Results from Study 2. Mean willingness to search for lower price or for higher quality items depending on a sacred (cremation container) or secular (clock storage container) purchase context. Error bars represent 95% confidence intervals of the means.
In an ANOVA on importance of the search attribute, search attributes were rated as more important when people were searching for a secular purchase (F(1,91) = 6.98, p = .010), importance of price and quality attributes did not differ (F(1,91) = 1.54, p = .218), and there was a marginal purchase type by search variable interaction (F(1,91) = 3.78, p = .055). Next, we examined each attribute’s importance depending on purchase type. When searching for a sacred purchase, price was rated as less important (M sacred = 3.21, SD = 1.23, M secular = 4.71, SD = 1.55, t(46) = 3.66, p = .001). The importance of quality did not differ between sacred and secular purchases (M sacred = 4.25, SD = 1.70, M secular = 4.48, SD = 1.81, t(45) = .45, p >.25).
5 Study 3
Study 3 investigates people’s willingness to negotiate. We expected people would be less willing to negotiate for lower prices when making sacred purchases.
5.1 Method
One hundred and six undergraduates (51% female, M age = 20, SD = 1.12) participated in exchange for course credit. Gender as an independent variable or covariate yielded no differences in forthcoming analyses.
Participants were randomly assigned to conditions in a 2 (Relationship type: sacred, loved one vs. secular, acquaintance) x 2 (Negotiated variable: price vs. quantity) design, with the first factor within-subjects and the second factor between-subjects. Presentation order of the within-subjects condition was counterbalanced and had no main or interactive effect.
Participants wrote about their feelings toward someone “they care about deeply or love” [“they know, but not very well”] who did [did not] occupy a big part of the participant’s emotional life (sacred and secular conditions, respectively). Participants also wrote that person’s first name. The name was subsequently inserted into the scenario and questions in order to make the materials more personally relevant.
Participants imagined purchasing cupcakes for a birthday party in honor of the person they wrote about, where cupcakes cost twenty dollars per dozen. Participants indicated how likely they were to negotiate for a lower price [higher quantity], how many dollars off [additional cupcakes] they would ask for, and their comfort negotiating for a lower price [higher quantity]. Finally, participants indicted how much more they would be willing to pay to have the baker write “Happy Birthday [person’s name]” on the cupcakes on a seven-point scale (1 = not much at all, 7 = a lot) and in an open-ended response. We expected people to be least willing to negotiate for lower prices when the purchase was associated with a loved one.
5.2 Results and discussion
We conducted a series of 2 (Relationship type: sacred, secular) x 2 (Negotiated variable: price, quantity) mixed ANOVAs on negotiation likelihood, initial negotiation offers, and negotiation comfort (Table 3). We examined initial offers in two ways: in their original form (additional dollars off or additional cupcakes asked for, depending on experimental condition) and after converting all offers to a dollar equivalent (by multiplying the number of cupcakes by $20/12 cupcake, the price per cupcake specified in the scenario).
Note: Mean values for dependent variables for each of the four cells (price vs. quantity negotiation by sacred vs. secular) are displayed, along with results from 2 (Relationship type: sacred or secular) x 2 (Negotiated variable: price vs. quantity) mixed ANOVAs and planned contrasts between negotiating on price for a sacred purchase versus the other three conditions (i.e. negotiating on price for a secular purchase and negotiating on quantity for a sacred or for a secular purchase). We display initial offer in two ways: in the original form, where half of participants asked for dollars off and half asked for additional cupcakes, and in a dollar equivalent, where cupcake offers were multiplied by the price per cupcake (i.e., $20/12 cupcakes). We do not conduct a planned contrast on the initial offers in the original form because number of cupcakes and number of dollars are not directly comparable.
Next, we assessed whether participants treated negotiation for better prices for a sacred purchase differently from other types of negotiations. In planned contrasts on each dependent variable, we assessed whether the sacred-price condition differed from the other three conditions (i.e., secular-price, sacred-quality, and secular-quality). For two of the three variables, planned contrasts were significant; when negotiating prices for sacred purchases, participants were relatively unlikely to negotiate and made relatively modest initial offers (Table 3).
Finally, we examined participants’ willingness to pay for cupcakes with a personalized birthday message. In planned contrasts using both the seven-point willingness to pay scale and the open-ended willingness to pay question, participants were willing to pay significantly more for customized confections if the purchase was for a loved one (for seven-point scale: M sacred = 4.42 vs. M secular = 2.03, F(1,104) = 207.79, p < .001; for open-ended: M sacred = $6.69 vs. M secular = $2.70; F(1,104) = 122.77, p < .001).
6 General discussion
People around the world purchase items of symbolic importance on a regular basis. Behavioral research offers established theories of sacred values and trade-off avoidance to help understand the challenges of the sacred purchasing process. Our inquiry reveals that, when a purchase is symbolic of love, people are reluctant to seek cost saving options and thus spend more money than is necessary given the availability of lower cost (yet equivalent quality) items in the marketplace. If consumers make many purchases of items symbolic of love over a lifetime (e.g., for weddings, funerals, birthdays and anniversaries), then these monetary costs could add up — especially for consumers who are already having trouble making ends meet.
We explored three aspects of the purchase process: choice, search, and negotiation. Respondents facing taboo trade-offs not only chose higher price items over lower price items they had judged to be equally desirable (Study 1), they also avoided searching for lower priced items (Study 2), and negotiating for lower prices (Study 3). The effects extended to experienced consumers (Study 1B) and positive and negative occasions (Study 2). Trade-off avoidance, however, ceased either when price was removed from consideration (Study 1-3) or when the item was not associated with love (Studies 2 and 3).
Loved ones want to demonstrate their commitment and avoid painful trade-offs when purchasing symbolic items. It is unclear whether the price-insensitivity we captured experimentally — and its real world analog — would survive long if people were fully aware of the psychological processes that underlie their behavior. Loved ones might change their behaviors if they realized third-party providers were pricing symbolic items exploitatively (e.g., funeral homes; Reference MitfordMitford, 1998). Third-party providers, in turn, might alter their pricing strategies if they realized loved ones were aware of their tactics. In this light, how obvious are the influence patterns to each party? How aware is each party of the influence patterns at play at a given point in the evolution of the relationships? We suspect the phenomena we have studied are influenced by a co-evolution of cultural norms and behavioral strategies, in a world with two types of players: consumers who have finite resources but want to believe some things have infinite value and sellers who have financial temptations to exploit this dissonance inside consumers (Reference Akerlof and ShillerAkerlof & Shiller, 2015).
Appendix: Example stimulus (Study 1)
Example stimuli for Study 1 are displayed below. So that higher scores indicated higher quality, the color scale was reverse scored such that values went from 1 = “M = Faintly tinted, usually yellow” (the least colorless option) to 10 = “D = Colorless 1” (the most colorless option).
The instructions read: “Please indicate the color of Option B that would make the two options equally appealing. Assume the two rings are the same for any attributes that are not mentioned:”
Open access
