Buy Now or Buy Later

Buy Now or Buy Later: The Effects of Scarcity and
Discounts on Purchase Decisions
Colin B. Gabler and Kristy E. Reynolds
This research investigates a burgeoning pricing strategy and its effects on purchase behavior. Drawing
from the expected-utility and prospect theories, we test the counteracting variables of scarcity and future
discount across two studies. We first implement a flea market scenario to demonstrate that scarcity creates
emotional value that increases purchase likelihood. Next, we determine the levels of scarcity and discount
that maximize purchase in a department store context. The findings suggest that the level of discount
predicts the purchase of highly visible products; for less visible products, scarcity drives the decision.
These relationships are moderated by involvement with the product class.

In the wake of the most crippling economic downturn
since the Great Depression, Americans have curbed their
shopping habits, and retailers are trying to adapt. Experts
do not expect the financial circumstances and buying practices of the average consumer to bounce back as they have
after past recessions (International Council of Shopping
Centers 2009), which means that value-based pricing—a
staple of retail—may be shortsighted (Anderson, Wouters,
and van Rossum 2010).
A new consumer requires a new strategy. Instead of selling a high volume of the same product at minimal margins
above cost, it may be more beneficial to implement a pricing
strategy that will (1) drive customers to visit the store more
frequently; (2) allow for a smaller inventory, and therefore
decrease warehouse costs; and (3)  maximize profits for
particular items by understanding how scarcity and price
promotions influence the purchase decision.
In an effort to protect fragile margins, some retailers have
begun to lower inventory to avoid offering huge discounts
on overstocked products; however, rising costs continue
to exert pressure on these firms (Holmes 2011). Although
pricing strategy has received a great deal of attention in
the literature (e.g., Samli and Jacobs 1993), the subarea of
price promotions is underdeveloped (Sivakumar 1996). The
current research investigates a pricing strategy that employs

Colin B. Gabler (Ph.D., University of Alabama), Assistant Professor of Marketing, College of Business, Ohio University, Athens,
OH, [email protected].
Kristy E. Reynolds (Ph.D., University of Alabama), Bruno Professor of Marketing, Culverhouse College of Commerce and Business
Administration, University of Alabama, Tuscaloosa, AL, kreynold@
cba.ua.edu.

the counteracting effects of product scarcity and a known
price decrease on the same purchase decision. Essentially,
the store offers some product in limited supply and then
lowers the price incrementally over the coming days or
weeks until all are sold. The customer has full knowledge
of the number of items remaining as well as the timing
and amount of the discount schedule, which is posted on
a sign or the price tag. A typical price tag on a shirt may
read: $100—April 1 through April 7; $75—April 8 through
April 14; $50—April 15 through April 21; $25—April 22 and
after. The strategy combines the techniques of Dutch auctions and price skimming but differs in that there is often
more than one item and the customer has full knowledge
of the price decrease.
The decision is whether to buy the product at the current,
higher price or at the future, lower price, with the implied
risk of the product no longer being available. This strategy
is commonly seen in consignment shops, flea markets, and
garage sales, but we posit that it could be applied to the
broader retail sector. Specifically, traditional retailers often
need to sell seasonal merchandise to make room for new
inventory. This model may help retailers sell their remaining inventory more quickly and at a higher profit margin.
We call it the “steadily increasing discount” (SID) model,
because it shares a foundation with Tsiros and Hardesty’s
(2010) steadily decreasing discounting (SDD) model for
ending a price promotion. Whereas the SDD model suggests that firms can maximize profits by bringing a heavily
discounted product back to its normal price in steps rather
than all at once, the SID predicts that firms can maximize
profits on the front end by gradually arriving at the deep
discount in similar steps. In short, this pricing strategy
could be used to speed up the clearance process while
optimizing profits on those items.
Journal of Marketing Theory and Practice, vol. 21, no. 4 (fall 2013), pp. 441–455.
© 2013 M.E. Sharpe, Inc. All rights reserved. Permissions: www.copyright.com
ISSN 1069–6679 (print) / ISSN 1944–7175 (online)
DOI: 10.2753/MTP1069-6679210407

442  Journal of Marketing Theory and Practice

Figure 1
Conceptual Model

Few scholars have investigated the SID model, and even
fewer retailers have adopted it. Zara International has implemented a strategy of “tantalizing exclusivity” (Ferdows,
Lewis, and Machuca 2004), and Lululemon uses scarcity
to attract customers (Mattioli 2012). But like Tsiros and
Hardesty’s (2010) model, the SID model does not have a
foothold in mainstream retailing. The U.S. consumer wants
fewer—but more specific—products, and this trend is not
likely to change in the near future, which means the SID
could have marketplace potential.
We begin with a review of the literature on pricing,
discounts, expected-utility theory, and scarcity. We then
use prospect theory to form hypotheses that are tested and
analyzed across two studies. Study  1 isolates the effects
of scarcity and tests the mediation of emotional value on
the purchase decision while Study 2 compares the effects
of product scarcity with that of a known future discount.
Involvement with the product class is tested as a moderator in each context (see Figure 1). We conclude the paper
with a discussion of the important findings, implications
for theory and managers, limitations, and future research
directions.

CONCEPTUAL FRAMEWORK
Pricing, Discounts, and Expected-Utility Theory
The pricing literature demonstrates that when people expect
a price to rise, they buy now, and when they expect a price to
drop, they wait for the sale (Jacobson and Obermiller 1990).
But consumers view prices relative to what they paid in the
past and what they expect to pay in the future. In effect,
consumers create their own reference price or standard by
which they judge the price of an item (Monroe 1973), usu-

ally forming a range that they deem acceptable for a given
product (Rao and Sieben 1992). So, when consumers make
purchases, they are analyzing how much they think they
should pay for an item relative to what they paid in the past,
what the price is now, and what it will be later.
Consumers use reference prices to make decisions about
the quality of the product and the monetary sacrifice
required to make the purchase (Monroe 2003). Generally,
individuals view a deviation above their reference price
negatively and a deviation below it positively. But a typical
reference price is dynamic—not just because of the normal
fluctuations of supply and demand but because of price
promotions. Consumer reactions to price promotions are
not always rational; they can even influence consumers to
abandon the cost-benefit analysis that usually governs their
decisions (Frank 2007).
Expected-utility theory (EUT) says that when consumers
face uncertain decisions, they try to maximize the expected
utility of their final assets (von Neumann and Morgenstern
1947). EUT assumes that, presented with a choice where
the outcomes are not known, individuals will compare the
weighted sum of option one versus the weighted sum of
option two, where the weighted sum is the utility of each
outcome multiplied by its probability (Mongin 1997).
All purchase decisions can be seen as a balance between
two types of this utility maximization. On the one hand,
consumers manage the price–quality relationship, wanting
to maximize the value gained for the price incurred. On the
other hand, they view the purchase as a monetary sacrifice,
wanting to maximize the utility (or minimize the disutility) of spending money (Monroe 2003). When a purchase
opportunity is unlimited, individuals place more weight
on the price–sacrifice relationship. Because they have time
to evaluate other indicators of quality (design, craftsman-

442  Journal of Marketing Theory and Practice

Figure 1
Conceptual Model

Few scholars have investigated the SID model, and even
fewer retailers have adopted it. Zara International has implemented a strategy of “tantalizing exclusivity” (Ferdows,
Lewis, and Machuca 2004), and Lululemon uses scarcity
to attract customers (Mattioli 2012). But like Tsiros and
Hardesty’s (2010) model, the SID model does not have a
foothold in mainstream retailing. The U.S. consumer wants
fewer—but more specific—products, and this trend is not
likely to change in the near future, which means the SID
could have marketplace potential.
We begin with a review of the literature on pricing,
discounts, expected-utility theory, and scarcity. We then
use prospect theory to form hypotheses that are tested and
analyzed across two studies. Study  1 isolates the effects
of scarcity and tests the mediation of emotional value on
the purchase decision while Study 2 compares the effects
of product scarcity with that of a known future discount.
Involvement with the product class is tested as a moderator in each context (see Figure 1). We conclude the paper
with a discussion of the important findings, implications
for theory and managers, limitations, and future research
directions.

CONCEPTUAL FRAMEWORK
Pricing, Discounts, and Expected-Utility Theory
The pricing literature demonstrates that when people expect
a price to rise, they buy now, and when they expect a price to
drop, they wait for the sale (Jacobson and Obermiller 1990).
But consumers view prices relative to what they paid in the
past and what they expect to pay in the future. In effect,
consumers create their own reference price or standard by
which they judge the price of an item (Monroe 1973), usu-

ally forming a range that they deem acceptable for a given
product (Rao and Sieben 1992). So, when consumers make
purchases, they are analyzing how much they think they
should pay for an item relative to what they paid in the past,
what the price is now, and what it will be later.
Consumers use reference prices to make decisions about
the quality of the product and the monetary sacrifice
required to make the purchase (Monroe 2003). Generally,
individuals view a deviation above their reference price
negatively and a deviation below it positively. But a typical
reference price is dynamic—not just because of the normal
fluctuations of supply and demand but because of price
promotions. Consumer reactions to price promotions are
not always rational; they can even influence consumers to
abandon the cost-benefit analysis that usually governs their
decisions (Frank 2007).
Expected-utility theory (EUT) says that when consumers
face uncertain decisions, they try to maximize the expected
utility of their final assets (von Neumann and Morgenstern
1947). EUT assumes that, presented with a choice where
the outcomes are not known, individuals will compare the
weighted sum of option one versus the weighted sum of
option two, where the weighted sum is the utility of each
outcome multiplied by its probability (Mongin 1997).
All purchase decisions can be seen as a balance between
two types of this utility maximization. On the one hand,
consumers manage the price–quality relationship, wanting
to maximize the value gained for the price incurred. On the
other hand, they view the purchase as a monetary sacrifice,
wanting to maximize the utility (or minimize the disutility) of spending money (Monroe 2003). When a purchase
opportunity is unlimited, individuals place more weight
on the price–sacrifice relationship. Because they have time
to evaluate other indicators of quality (design, craftsman-

Fall 2013  443
ship, detail, etc.), consumers view price less as an indicator of quality and more for the financial burden it incurs
(Lichtenstein, Bloch, and Black 1988). Utility maximization
of this relationship occurs when the price represents the
smallest sacrifice possible. When a purchase opportunity is
limited, individuals place more weight on maximizing the
utility of the price–quality relationship (Suri and Monroe
2003). Because they cannot process as much information,
the price tag serves more as an indicator of quality than
monetary sacrifice and utility maximization hinges more
on the level of perceived quality (Suri, Kohli, and Monroe
2007).
Given an unlimited purchase opportunity, buying a
discounted product represents a smaller sacrifice, and
therefore higher utility. Using Jacobson and Obermiller’s
(1990) logic, if individuals know that a price will drop in
the future, they would be less likely to buy the product in
the present. Thus, the knowledge of an upcoming discount
should lead to purchase postponement.

Scarcity, Prospect Theory, and Decisions
Under Risk
One way that retailers can limit the purchase opportunity
is by manipulating scarcity. Scarcity can take the form of
a purchase limit (e.g., “Limit X per customer”), a purchase
precondition (e.g., “Product X only available to those who
buy product Y”), a time limit (e.g., “Call within the next
30 seconds!”), or the focus of this research, a product limit
(e.g., “Only X number of products remain!”) (Inman, Peter,
and Raghubir 1997). Brock’s (1968) commodity theory posited that individuals assign values to commodities based on
their availability, viewing scarce products as more attractive than readily available ones. Furthermore, increases in
scarcity also lead to increases in perceived value (Cialdini
1993) and desirability (Lynn 1989).
Consumers normally prefer to gather information
before the purchase decision, but retailers can induce
excess demand to force consumers to decide before they
are comfortable (DeGraba 1995). Apple creates one of these
“buying frenzies” with each iteration of the iPad. Similarly,
Lululemon purposely keeps a low inventory to drive faster
purchases (Mattioli 2012). In such a restricted consumer
setting, consumers often use availability as their main
source of information to make the decision (Inman, Peter,
and Raghubir 1997).
EUT has failed to predict some very common human
behaviors. One reason is that many studies focus on backward-looking consumer choices although most research on

consumption shows that people, in general, are forwardlooking (Hall 1978). Another reason for the theory’s lack
of explanatory power is that money is not always the most
important outcome in consumer decisions (Bell 1982);
consider the people waiting in line for the newest release
at the Apple store. But the main reason that EUT does not
apply to many situations is that people do not always opt
to maximize their utility.
Numerous non–expected-utility models have been developed to address this behavior. For instance, prospect theory
says that people tend to place more weight on outcomes
they consider to be certain than on those they consider to
be probable. Therefore, individuals will more often choose
a modest gain if it is a “sure thing” over some probability
of a larger gain. However, when forced to choose between
a certain modest loss and some probability of a larger loss,
individuals more often risk the larger loss (Kahneman and
Tversky 1979).
Suppose that Apple is selling iPads for $500 today and
$450 tomorrow. This pits the certainty of a modest gain
(full-price iPad) against the probability of a slightly larger
gain (discounted iPad). Given an infinite number of products, the rational individual would wait for the lower price,
as predicted by EUT. But what if there were 100 people who
wanted to buy 10 iPads? Then, an iPad is a scarce commodity, and the probability of obtaining one is 1 in 10, or
10 percent. If there were 50 iPads, that probability would
increase to 50 percent; however, if there was only one, it
would decrease to 1 percent. Scarcity shares a theoretical
root with probability; the literature tells us that individuals
would perceive more value and desirability in the lone iPad
than 1 of the 10, and more in 1 of the 10 iPads than 1 of
the 50 (Cialdini 1993; Lynn 1989). Because people tend to
overweight outcomes based on their certainty (Kahneman
and Tversky 1979), and the probability of obtaining an item
changes with the probability, as scarcity increases, so does
purchase likelihood. Formally,
Hypothesis 1: Scarcity positively influences purchase.
But what if there is a limited number and an upcoming
discount? Prospect theory suggests that the overweighing
of certainty favors the risk averse in the domain of gains
and the risk seeking in the domain of losses (Kahneman and
Tversky 1979). Risk aversion is a personality trait that, like
scarcity and pricing discounts, influences the probability
that an individual will accept uncertainty over certainty
(Burton et al. 1998). In an experiment, individuals preferred
a definite one-week vacation to England to a 50  percent
chance of a three-week vacation to England, France, and

444  Journal of Marketing Theory and Practice
Italy. In the iPad scenario, the $500 product would be the
certain one-week trip and the $450 product the uncertain
three-week trip. In weighing the two prospects, the former
is certain, hence increasing its prospect, but the latter has
greater value, thus increasing its prospect. In general, the
latter prospect increases or decreases with the number available. An individual will be more likely to take the certain
modest gain (product) over the probability of a larger gain
(product plus discount) because the certainty of option 1
outweighs the added value of option 2.

Perceived Emotional Value
Emotions are subjective states of being that involve a
hedonic component and motivate behavior specific to that
emotion (Baron 1992). In a shopping context, emotional
value is the utility derived from the feelings created by
a product (Sweeney and Soutar 2001), and we use this
definition in our study. Cialdini (1993) found that people
assume that something is more valuable if it is less common
and suggested that scarcity clouds the brain and limits the
ability to process information. This response is similar to
that of impulse buying (Rook 1987). Both trigger emotions
that may lead to purchase based on something other than
attributes of the product itself. However, impulsivity is typically a trait of the individual while scarcity is a situational
characteristic. Accordingly, under high scarcity conditions,
the perceived emotional value of a product should be high;
in a low scarcity condition, the perceived emotional value
should be low. Therefore,
Hypothesis 2: Scarcity positively influences perceived
emotional value.
When time and quantity are unlimited, one can pay
more attention to the important information and relevant
cues. The lack of temporal pressure is associated with extra
processing (Bozzolo and Brock 1992). In auctions and other
scenarios where time or quantity is limited, the emotional
arousal compromises a person’s cognitive ability, which
decreases the number of information cues one can grasp
to assess the decision (Dhar and Nowlis 1999). Under such
conditions, this distraction leads to decision making based
on heuristics (Cialdini 1993). Because emotional value
serves as a heuristic in the purchase decision, we posit that
it will act as a mediator between scarcity and purchase.
Therefore,
Hypothesis 3: Perceived emotional value mediates the
relationship between scarcity and purchase.

Involvement with Product Class
In the marketing literature, involvement is defined as the
importance of a product or decision to an individual (Mittal
1995). Involvement creates a state of arousal or motivation
and drives us to process information more thoroughly
(Rothschild 1984). Enduring involvement is ongoing; it is
independent of a specific occasion, and it usually enforces
an individual’s self-concept (Richins and Bloch 1986).
While scarcity limits our ability to process information, it also creates two key antecedents of involvement,
perceived risk and perceived pleasure value (Laurent and
Kapferer 1985), each of which is fundamental in determining the complexity and extensiveness of the cognitive
process (Celsi and Olson 1988). When an item is scarce,
its perceived pleasure value increases as does the perceived
risk of missing out on the opportunity (Bloch and Richins
1983). An individual involved with the product class will
process information more readily because the product—and
consequently the decision—are both more important and
interesting.
If an individual is involved with music, he or she cares
about it and most likely is knowledgeable about it (Brucks
1985). For these individuals, a scarce album would likely
have a higher pleasure value and risk associated with its
purchase. However, scarcity alone would not cause lowinvolvement individuals to suddenly become interested
in music, so the pleasure value and risk will not change
for them. Thus, scarcity will heighten the importance of
a product for high-involvement individuals but not lowinvolvement individuals. Formally,
Hypothesis 4: Involvement with the product class
moderates the positive relationship between scarcity
and purchase such that when involvement is high this
relationship is strong and when involvement is low, this
relationship is weak.

STUDY 1
Sample and Procedure
Based on a pretest, we developed manipulations for high
and low scarcity as well as the choice of product for the
main study. The sample consisted of 247 undergraduate
students at a large southeastern university. The students
completed the survey in class using paper and pencil. When
finished, they were given extra credit and debriefed about
the purpose of the study. Their average age was 21 with a
range from 18 to 35, with 45 percent male and 55 percent

Fall 2013  445
female. The subjects identified their favorite musical artist or band and the dollar amount they would spend on
an album of that band’s first concert. Then we presented
them with a CD-purchasing scenario. Random assignment
to one of two scarcity conditions (high-low) placed the
respondents at a flea market where they found that CD for
$5 higher than the price they identified. The low-scarcity
condition had three CDs while the high-scarcity condition
had one. The scenario read that the flea market offered
any unsold items at half-price the following Saturday, and
asked subjects to decide between purchasing the CD that
day or next week.

Measures
Laurent and Kapferer’s (1985) three-item scale (α = 0.85)
was used to measure involvement with the product class and
Sweeney and Soutar’s (2001) five-item scale (α = 0.91) was
used to measure perceived emotional value. As discussed,
certain individuals are more prone to buy on impulse than
others. For that reason, Rook and Fisher’s (1995) modified
impulse buying scale (α  =  0.91) was included as a covariate so that the emotional response of scarcity could be
isolated. Consumers also differ on their general aversion
to risk. Therefore, Burton et al.’s (1998) four-item risk aversion scale (α = 0.76), which measures the degree to which
an individual avoids taking risks in life, was included as
a covariate. The survey also contained a manipulation
check for scarcity (“The number of CDs at the flea market
is . . .”) using a seven-point scale ranging from (1 = low)
to (7 = high). Appendix A contains the scenario and scale
items after purification. Purchase was measured as a binary
response variable (yes/no).

Analysis and Results
Participants in the high-scarcity scenario described the
CD as significantly (F(1,245) =  27.415; p  <  0.001) scarcer
(M = 5.52) than those in the low-scarcity scenario (M = 4.59).
An interesting finding was that the amount the participants
were willing to spend positively influenced both their likelihood to purchase (p < 0.01) and the perceived emotional
value assigned to the CD (p < 0.001). The only gender difference that emerged was that women placed significantly
more emotional value than men on the CD (p < 0.05) .
Sixty-three of the 120 respondents (53 percent) on the
low-scarcity condition chose to buy the CD while 93 of the
127 respondents (73 percent) on the high-scarcity condition
chose to buy the CD (χ2(1,245) = 11.484, p = 0.001). This

supports H1; despite the knowledge of a future discount,
scarcity positively influences purchase behavior. We implemented Baron and Kenny’s (1986) three-step regression
procedure to determine if perceived emotional value mediated the relationship between scarcity and purchase, adding
the covariates in each step to demonstrate unique variance
(Neubert et al. 2008). While ordinary least squares regression
was appropriate for step 1, steps 2 and 3 required binary
logistic regression. Using a procedure developed by (MacKinnon and Dwyer 1993), the coefficients were multiplied by
the standard deviation of the predictor variable and then
divided by the standard deviation of the outcome variable.
This step allowed the authors to compare continuous and
binary variables in the same analysis (see Table 1).
The results from step 1 support H2. Essentially, the scarcer
the CD, the more emotional value the respondents attached
to owning it. Step 2 replicates H1 while step 3 included both
scarcity and perceived emotional value and tested their
relationship with purchase. The beta for scarcity dropped
with the addition of perceived emotional value; however, it
remained significant (p < 0.05), which exhibits partial mediation and support for H3. This is evidence that scarcity led
consumers to purchase the CD, in part because they placed
more emotional value on the idea of owning the CD.
Finally, it is worth noting that the average amount that
respondents were willing to spend on the CD was $23. The
scenario specifically asked if they would pay $5 more than
that amount. Five dollars represents an increase of 22 percent
to the average self-reported price of $23. Therefore, scarcity
did not just lead people to buy the CD at their upper limit,
the respondents were willing to spend, on average, 22 percent more than their upper limit.
A Sobel (1982) test confirmed that the indirect effect was
significant. The test statistic was 2.01 and was significant on
both the one-tail (p < 0.05) and two-tail (p < 0.05). These
results indicate that the indirect effect of the independent
variable, scarcity, on the dependent variable, purchase,
through the mediator, perceived emotional value, was
significant. Thus, we can conclude that the regression was
significant on all levels.
To test the moderating effect of involvement with the
product class, a median split was performed to create high
and low levels. A t‑test confirmed that a high-involvement
respondent was significantly more involved (M  =  6.54)
than someone with a low level of involvement (M = 4.49,
t(1,244) = 21.03, p < 0.001), which allowed us to treat the
high and low levels as unique variables. Next, we conducted
a mean difference test on both levels of involvement for the
high- and low-scarcity conditions. We found support for H4,

446  Journal of Marketing Theory and Practice
Table 1
Regression Analysis for Study 1
Dependent Variable: Purchase
Step 1: IV → Med
Variable

ß

Control Variables
Risk Aversion
Impulse Buying
Independent Variable
Scarcity
Mediator Variable
Perceived Emotional Value
Adjusted R2 or Nagelkerke R2
F(df) or χ²(df)

–0.185
0.054
0.167**

Step 2: IV → DV

S.E.

VIF

0.060*
0.055

1.000
1.000

0.135

1.000

0.054
5.668 (3, 242)**

ß
–0.060
0.115
0.898***

S.E.

Step 3: IV/Med → DV
ß

S.E.

0.122
0.113

0.030
0.094

0.131
0.117

0.273

0.771*

0.282

0.068
12.499 (2, 244)*

0.505***
0.141
0.141
26.804 (3, 243)***

Notes: IV = independent variable; Med = mediator variable; DV = dependent variable; VIF = variance inflation factor; S.E. = standard error; df = degrees
of freedom. Listwise n = 247. Standardized betas are reported. * p < 0.05; ** p < 0.01; *** p < 0.001.

confirmed by a difference of proportions test (Z = –0.406,
p < 0.001). For high involvement individuals, those in the
high-scarcity condition bought the CD 85 percent of the
time compared to those in the low-scarcity condition who
bought the CD 53 percent of the time. Conversely, for low
involvement individuals, the purchase percentages did not
change significantly as a result of the scarcity condition
(53  percent in low-scarcity condition; 61 percent in the
high-scarcity condition). Interestingly, in the low-scarcity
condition, the same percentage (53 percent) of individuals
chose to buy the CD regardless of their level of involvement
with the product class. This means that for those with low
involvement, the relationship between scarcity and purchase
was not just weak, but actually nonexistent (see Figure 2).

STUDY 2
Study 2 was conducted to assess how the relationships from
Study 1 might change when the level of discount is also
manipulated. A more traditional retail setting (shopping
for jeans in a department store) was chosen to increase
generalizability.
Suppose that the iPad, originally priced at $500, sold
for $375 the next day. This pits the certainty of a modest
gain (just the iPad) against the probability of an even larger
gain (iPad + 25 percent discount). Finally, imagine that the
iPad sold for $250 the next day. This pits the certainty of
a modest gain (just the iPad) against the probability of a
much larger gain (iPad + 50 percent discount). Based on
weighted probabilities, the first prospect remains the same
while the second prospect gains value as the incentive, or

discount, increases (Kahneman and Tversky 1979). Consequently, the value of the first prospect decreases relative to
the second. At some threshold, the value added of prospect
two outweighs the certainty of prospect one.
Consumers use the level of perceived risk as a factor in
product decision making (Bettman 1973), and individuals
use availability to estimate that level of risk (Tversky and
Kahneman 1973). We predict that when there is a lot of the
product in stock and the manager puts that product on a
high future discount, the risk is low, and thus the incentive
to wait is large. If the discount was low, the possible savings
would not be worth the wait; if fewer products were available, the risk that they would be purchased would be high
(Folkes 1988). In both cases, the incentive to wait decreases.
Therefore, the most ideal time to wait for the sale is when the
scarcity is low and the future discount is high. Conversely,
the most ideal time for a consumer to buy the product—or
when the incentive to wait is smallest—is when the product
is scarce and the future discount is low. Formally:
Hypothesis 5: Those in the low-scarcity and high future
discount condition are least likely to purchase while those
in the high-scarcity and low future discount condition
are most likely to purchase.
As discussed, involvement is the importance of a decision based on an individual’s specific needs, values, and
interest in that product (Zaichkowsky 1985). An unimportant purchase decision is then, by definition, uninvolving
(Richins and Bloch 1986). Scarcity implies a perceived risk,
which would heighten the relevance of that product to one’s
values and interests (Bloch and Richins 1983). Similar to

Fall 2013  447
Study 1, we predict scarcity to accentuate the importance
of the decision for high-involvement individuals in the
clothing product category, making the decision more risky,
interesting, and relevant.
Low involvement with a product class translates to less
arousal or motivation to process information (Rothschild
1984). It means that the product is not as essential to one’s
self-concept (Richins and Bloch 1986) and the decision is
not as important. Therefore, for low-involvement individuals, scarcity will not make the weight of the decision more
salient. We predict it will extenuate the interest, risk, and
importance of the decision. In fact, when consumers have
low involvement with a product, price often acts as the
main indicator of quality (Monroe 1973), and consequently,
determines behavior. Studying red wine, Zaichkowsky
(1985) found that low-involvement individuals placed more
emphasis on price than high-involvement individuals. In
essence, the deviation below the reference price was the
predictor of purchase for these consumers (Winer 1986).
In this study, the discount is the price cue, and we predict
it will drive behavior for low-involvement consumers.
Conversely, high-involvement consumers care more about
product attributes and are willing to pay for them (Lichtenstein, Bloch, and Black 1988). Generally, higher prices
are deemed more acceptable by consumers with high
involvement than low involvement (Bloch, Sherrell, and
Ridgway 1986); therefore, the effect of the discount should
be minimal for these individuals. Therefore,
Hypothesis 6: Involvement with the product class moderates the negative relationship between future discount
and purchase such that when involvement is low this
relationship is strong and when involvement is high, this
relationship is weak.

Sample and Procedure
To test H5 and H6, we manipulated both scarcity and
future discount in a retail shopping scenario and analyzed
their effects on the purchase decision. We conducted a
pretest to uncover shopping habits that would adversely
affect the data, such as hiding or reserving the product,
and addressed these issues in the final scenario. For the
main study, we employed Bitner, Booms, and Tetrault’s
(1990) student recruitment method to identify subjects,
first training them and then giving each a URL (Uniform
Resource Locator) to present to their recruit. The students
had the opportunity to recruit two people to participate,
receiving extra course credit for each respondent. To validate the recruited sample, we contacted 10 percent of the

Figure 2
Moderating Effect of Involvement for Study 1

Low Scarcity
High Scarcity

respondents to verify their participation in the study and
uncovered no problems.
The participants accessed the survey through a secure
Web site, which generated a random combination of manipulations each time the URL was clicked. The recruits could
only access the link once. The identification, recruitment,
and data collection period lasted two weeks, resulting in a
total of 423 respondents. The mean age was 30 and ranged
from 18 to 64, with 60 percent female and 40 percent male.
The scenario first asked the subjects to imagine that they
needed a new pair of jeans and to identify their “go-to”
brand. It then asked how much they would be willing to
spend for a pair of jeans from that brand in “the exact style,
fit, color, etc., that you are looking for.” The mean price
was $67 while both the mode and median were $50.
The scenario placed the respondents at a store where they
found that exact pair of jeans at the price they identified.
We manipulated scarcity at high and low conditions, with
ten pairs of jeans representing the low condition and two
pairs representing the high condition. To manipulate future
discount, the scenario contained the line “Above the table
hangs a sign that reads, ‘All jeans 50%-off [25%-off, 10%off] next Saturday,’” which created high, moderate, and
low conditions. The Web site then randomly assigned each
respondent to one of the two scarcity conditions and one
of the three future discount conditions before presenting
them with two choices: “Buy the jeans today” or “Come
back next Saturday to buy the jeans.”

Measures
Laurent and Kapferer’s (1985) three-item scale (α = 0.91)
was again used to measure the involvement with the product

448  Journal of Marketing Theory and Practice
Table 2
Overall Logistic Regression Results for Study 2
Dependent Variable: Purchase
Model 1

Model 2

Model 3

Variable

Coefficient

Wald

Coefficient

Wald

1/exp(ß)

Constant
Impulse Buying
Scarcity
Discount
Scarcity × Discount
Model c2 (df)
Block c2 (df)
Hosmer and Lemeshow
c2 (df)
Percent Correct
Predictions
Nagelkerke R2

–0.439
0.290***

2.78
15.22

–0.492
0.310***
0.289**
–0.304*

3.39
16.66
7.47
5.63

1.36
1.34
0.74

16.01 (1)
9.78 (8)
63.6
0.05

Coefficient

Wald
3.57
16.94
7.29
5.46
0.64

29.21 (3)
13.13 (2)
5.70 (8)

–0.507
0.314***
0.286**
–0.300*
0.103
29.85 (4)
0.64 (1)
14.94 (8)

66.7

66.3

0.09

0.09

Notes: The Wald statistics are distributed chi-square with 1 degree of freedom. *p < 0.05; ** p < 0.01; *** p < 0.001.

class. Rook and Fisher’s (1995) modified impulse buying scale
(α = 0.94) was included as a covariate. The survey also contained a manipulation check for scarcity (“The number of
pairs of jeans on the table is . . .”) and future discount. (“The
discount being offered is . . .”) using a seven-point scale
ranging from (1 = low) to (7 = high). Appendix B contains
the scenario and scales items after purification.

Analysis and Results
Both of the manipulations were successful. The participants in the high-scarcity scenario described the jeans as
significantly (F(1, 425) = 23.03, p < 0.001) scarcer (M = 4.50)
than those in the low-scarcity scenario (M = 3.67). Similarly,
the discount was rated as significantly different at each
level (Mhi = 5.39, Mmod = 4.29, Mlo = 3.42, F(2, 424) = 53.22,
p  <  0.001). To determine whether scarcity and future
discount produced a significant interaction, we tested a
binary logistic model. We entered three blocks separately;
starting with the covariate (impulse buying), followed by
the manipulated scarcity and future discount variables,
and finally the interaction. We assessed the overall model
fit after each addition.
The first model with just impulse buying was significant
and predicted 63.6 percent of the purchase outcomes. The
second model, which contained the manipulated independent variables, showed superior fit as both the block and
the model were significant and the percentage of correct
predictions increased 3.1 points to 66.7 percent. The final

block, which included the interaction term, was not significant and actually had less predictive power (66.3 percent),
allowing interpretation of coefficients from model  2. To
achieve probabilities, we used the formula, P = 1/exp(b),
where the expected beta is the effect of the independent
variable on the odds ratio. The resulting probability is
the change in likelihood that an individual will purchase
given an increase in that variable. As seen in Table 2, those
in the high scarcity condition were 1.36 times more likely
to buy than those in the low-scarcity condition (p < 0.01),
which replicates H1 from Study 1. Those in the high future
discount condition were 0.74 times less likely to buy than
those in the moderate future discount condition, who were
0.74 times less likely to buy than those in the low future
discount condition (p < 0.01).
We used a difference of proportions test to analyze H5.
Each of the six conditions produced a proportion of those
who chose to buy the jeans. Comparing this proportion to
the overall proportion determined if there was a significant
difference in any one condition. The low-scarcity–high
future discount condition yielded the lowest purchase
probability (48.5 percent), which represents a significant
difference (p < 0.05) from the average of 62.6 percent. The
high-scarcity–moderate future discount condition yielded a
slightly larger purchase probability (72.4 percent) than the
high-scarcity–low future discount condition (71.2 percent).
Because these two percentages are not significantly different
from one another and the high-scarcity–low future discount
is significantly different from the low-scarcity–high future

Fall 2013  449
discount condition (p < 0.01), we demonstrate partial support for H5 (see Table 3).
To test H6, as well as replicate the results of H4, we
performed a median split to create high and low levels of
involvement. The t‑test confirmed that the high-involvement
group was significantly more involved (M  =  5.85) than
the low-involvement group (M  =  3.49, t(1,418)  =  27.34,
p < 0.001).
Using binary logistic regression, we attempted to replicate H4 with the new data. In Study 1, scarcity affected those
with high involvement and had no effect on those with low
involvement. In this scenario, we found the opposite effect.
For individuals reporting low involvement with clothing,
scarcity was the driver of purchase (1/exp(b) = 1.56), meaning that regardless of discount, low-involvement individuals
were 1.56 times as likely to buy the jeans when scarcity was
high than when it was low (p < 0.01). Furthermore, highinvolvement individuals were more influenced by the level
of the discount (1/exp(b) = 0.65) as the respondents were
0.65 times less likely to buy a pair of jeans at each discount
interval (p < 0.05) (see Tables 4 and 5).
In essence, low involvement increased the salience of
scarcity while high involvement increased the salience of the
discount, which is opposite of the predicted effects in H4
and H6. When the product class was music, high involvement was pronounced by scarcity. The decision was more
risky and interesting, leading to a greater proportion of the
respondents to purchase the CD. However, with clothing,
scarcity had a significant effect on the low-involvement
individuals while the high-involvement individuals were
affected by the level of discount. These counterintuitive
findings is addressed in the next section.

GENERAL DISCUSSION
Theoretical Implications
While demographics and personal characteristics certainly
influence purchase behavior in a retail setting, this research
asserts that the purchase scenario itself can have a major
impact on the decision. Mazumdar, Raj, and Sinha (2005)
suggest that context plays an important role in the purchase
process, and discount and scarcity certainly fall into that
category. Auction research has examined scarcity and future
price increases (Campbell 1999), but in this research we set
out to investigate the counteracting effect of scarcity and
a future price discount in the same scenario.
Theoretically, we demonstrate another instance in which
EUT fails to predict choice behavior. We find that prospect
theory better explains the way consumers react to the

Table 3
Proportion of Subjects Who Chose to
Purchase for Study 2
Scarcity
Discount

Low
(Percent)

High
(Percent)

Total
(Percent)

Low
Moderate
High
Total

66.2
56.5
48.5+
57.2

71.2**
72.4**
60.3
67.9

68.8
64.5
54.6
62.6

+ p < 0.05: represents a significant difference in relative to the total
average cell. ** p < 0.01: represents a significant difference relative to
the low-scarcity–high future discount cell.

counteracting variables. In Study 1, scarcity led people to
choose the certain modest gain (product at full price) over
some probability of a larger gain (product at discounted
price). One reason for this behavior is that consumers attach
emotional value to a product based on its availability and
use this heuristic in the purchase scenario. Our findings
show that not only will individuals pass up a discount
but they will even purchase a product at a higher price
than their upper limit. In Study 2, when discount was also
manipulated, individuals were more likely to risk a certain
modest gain for the chance of a larger one, depending on
the size of the discount.

Managerial Implications
Promotional discounts are effective because most consumers want to purchase products at the lowest price possible.
However, firms generally want to sell products at the highest price possible and as soon as possible. Our study shows
that by framing the purchase scenario with the dual aspects
of scarcity and future discount, consumers will generally
forsake small discounts to avoid missing the opportunity
to purchase the product. This means that firms can either
manipulate their inventory, their pricing schedule, or both,
in order to maximize profitability. Suppose a retailer has a
large stock of a product it needs to sell to make room for
incoming merchandise. When creating the pricing structure
for this item, our results suggest that the retailer will sell
more of that product at full price by advertising a low future
discount on the product. However, if that product is scarce,
a larger discount will yield a higher full-price purchase
rate, and therefore maximize profit margin on the product.
Furthermore, because individuals attach emotional value
to scarce products, managers can use this knowledge to lay
out the message in the most emotion-provoking way.

450  Journal of Marketing Theory and Practice
Table 4
Logistic Regression Results for Low Involvement for Study 2
Dependent Variable: Purchase
Model 1
Variable
Constant
Impulse Buying
Scarcity
Discount
Scarcity × Discount
Model c2 (df)
Block c2 (df)
Hosmer and Lemeshow c2 (df)
Percent Correct Predictions
Nagelkerke R2

Model 2

Coefficient

Wald

Coefficient

Wald

1/exp(ß)

–0.407
0.228*

1.43
5.11

–0.532
0.274**
0.445**
–0.214

2.28
6.82
9.45
1.47

1.31
1.56

5.26 (1)

16.72 (3)
11.46 (2)
8.12 (8)
64.8
0.10

7.29 (8)
58.9
0.03

Notes: The Wald statistics are distributed chi-square with 1 degree of freedom. * p < 0.05; ** p < 0.01; *** p < 0.001.

Table 5
Logistic Regression Results for High Involvement for Study 2
Dependent Variable: Purchase
Model 1
Variable
Constant
Impulse Buying
Scarcity
Discount
Scarcity × Discount
Model c2 (df)
Block c2 (df)
Hosmer and Lemeshow c2 (df)
Percent Correct Predictions
Nagelkerke R2

Model 2

Coefficient

Wald

Coefficient

Wald

–0.352
0.320*

0.683
7.78

–0.233
0.300**
0.081
–0.194*

0.286
6.63
0.26
5.09

8.35 (1)
4.59 (8)
62.7
0.06

1/exp(ß)
1.35
0.65

13.76 (3)
5.41 (2)
4.23 (8)
67.8
0.10

Notes: The Wald statistics are distributed chi-square with 1 degree of freedom. * p < 0.05; ** p < 0.01; *** p < 0.001.

The SID model differs from traditional clearance pricing
models in two ways. First, while customers generally know
when sales occur either through promotional efforts or
the changing of shopping seasons, the SID model relies on
the fact that customers have full knowledge of the entire
discount schedule. This access to specific information triggers the heuristics studied in this research. Second, for the
SID model to work, the retailer must advertise only within
the store. The discount schedule is public information, but
because the retailer does not promote it, it forces customers to visit the store. Then, it pits customers against one
another, and like casino poker, the house always wins.

This is especially relevant given the current economic
climate. Many marketers are reducing prices to encourage
consumers to buy, but they effectively decrease their profitability as consumers simply wait for these sales. The SID
model would allow firms to advertise a sale and thereby
maintain their concern for the U.S. consumer, but to sell
most of their inventory before the sale occurs. In essence,
they could have their cake and eat it, too.
Involvement with the product class proved an important
customer characteristic in both settings. As predicted, high
involvement magnified the effect of scarcity when the
decision involved music, but counter to our hypothesis,

Fall 2013  451
low involvement magnified the effect of scarcity when the
decision involved clothing. Jeans, and clothing in general,
are a more visible product, and consequently, they imply a
greater social risk (White 1966). Low involvement entails a
low knowledge (Brucks 1985) and low motivation to process
(Rothschild 1984). Brannon and Brock (2001) found that
when motivation and knowledge are low, scarcity results
in consumers paying more attention to task-relevant cues.
Those with a lack of motivation and knowledge about clothing and fashion are more likely to use the availability as a
task relevant cue of its value and desirability, and ultimately,
as a driver of purchase.
High-involvement consumers, by contrast, are not as
reliant on that cue. They may also shop more often, follow
the trends, and know whether or not the product is worth
the price. These individuals tend to be product experts.
They pay more attention to product information, spend
more time processing that information, and put forth more
effort to the search process (Forehand and Deshpandé
2001) to ensure that they make a wise choice (Celsi and
Olson 1988). A wise choice, in this case, means evaluating
all the alternatives, perhaps looking online, and making
sure they get the best price for the product. After all, even
high-involvement individuals seek low prices (Lichtenstein,
Bloch, and Black 1988), which could be why the discount
had a greater effect on this group.
The consumer’s level of involvement is therefore a
vital consideration for retailers looking to implement the
SID pricing schedule. While involvement is an individual
characteristic, some product categories—like clothing—are
intrinsically more involving than others. Lichtenstein,
Bloch, and Black point out: “A buyer may feel quite differently about purchasing a higher-priced brand of peas than
about purchasing a higher-priced television” (1988, p. 245).
These highly visible product categories are more likely to
have high-involvement shoppers who are less susceptible
to scarcity but are very conscious of the pricing structure.
For less visible products, involvement may not be as high,
meaning individuals will place more weight on availability
and use scarcity as a decision cue. Managers should use
this information to cater a SID schedule to their specific
product.
Finally, the successful implementation of the SID pricing
schedule requires a change in all aspects of the business.
Zara’s success stems from the integration of its pricing
model to each function of the supply chain (Ferdows, Lewis,
and Machuca 2004). Their distribution centers have to be
responsive, their databases accurate and current, their staff
well-trained, and their designers cutting-edge and prolific.

It is easier to carry a large volume of the same product than
it is to carry a large assortment of products; but to truly
work in the retail sector, an extensive product breadth
must balance the lack of product depth. That way, when a
rack with three dresses runs out, customers will not leave
the store; they will find another rack with three different
dresses. Being out of stock is no longer a liability, but may
actually become an asset. Pricing promotions also typically
increase store traffic (Pan and Zinkhan 2006), and this has
been true in the case of Zara. The constant turnover of new
and unique products attracts more frequent visits than a
store with low product turnover.
One word of caution. Often when consumers miss a
bargain, they decide not to purchase that product at the
normal price because it would serve as a reminder of the
missed opportunity (Sevdalis, Harvey, and Yip 2006). While
Tsiros and Mittal (2000) found that regret positively influences repurchase intentions, a fine line separates positively
influencing repurchase behavior and negatively affecting
attitudes toward the company. A key component of the SID
model is transparency and truthfulness. Managers must be
up front about the discount as well as the stock of products.
For instance, customers may feel manipulated if a retailer
carried dozens of a style of jeans but only placed a few on
the sales floor at one time. The unique cues of this pricing
schedule make it relevant only for products that will not be
replenished in the near future. Finally, the SID model shares
a foundation with Tsiros and Hardesty’s (2010) SDD model,
and could effectively dovetail to this strategy to create a
comprehensive promotional pricing schedule.

Limitations and Future Research
While academic journals have validated the student
recruitment method and student data in general, it poses
a limitation in this study. Future research might test the
SID model in an actual retail setting so that real choices
could be observed. In this study, we isolated the effects of
scarcity and future discount into one purchase decision;
respondents had a choice of buying the product now or at
the later date. Thus, this study is a snapshot of one interval
in the SID schedule. To fully assess the applicability of the
SID model, purchase decisions would have to be assessed
and measured at different intervals throughout the course
of the promotion.
Because the effect of involvement differed based on
the product type, this construct and how it relates to scarcity and future discount warrants further investigation.
Lichtenstein, Bloch, and Black (1988) note that involvement

452  Journal of Marketing Theory and Practice
is product specific and varies across products and situations, and clearly this was the case in our study. A study
could be developed where the effects of scarcity and future
discount are measured across several industries/product
types to determine if involvement is industry or product
type specific.
Other moderators may be especially pertinent to the SID
model. For instance, the same discount can yield different
responses when presented as either dollar-off or percentageoff (Estelami 2003). In both the EUT and prospect theory,
how the message is framed and interpreted largely determines the behavioral response. Future research should
consider message framing and conflict of decision, as well as
deal proneness, need for uniqueness, regret, and repurchase
intentions. Finally, while the SID model may prove effective,
by its very nature of promoting sales that may never come
to fruition, it warrants ethical consideration.

REFERENCES
Anderson, James C., Mark Wouters, and Wouter van  Rossum
(2010), “Why the Highest Price Isn’t the Best Price,” Sloan
Management Review, 51 (2), 69–76.
Baron, Jonathon (1992), “The Effect of Normative Beliefs on Anticipated Emotions,” Journal of Personality and Social Psychology,
63 (2), 320–330.
Baron, Reuben M., and David A. Kenny (1986), “The ModeratorMediator Variable Distinction in Social Psychological
Research: Conceptual, Strategic, and Statistical Considerations,” Journal of Personality and Social Psychology, 51 (6),
1173–1182.
Bell, David E. (1982), “Regret in Decision Making Under Uncertainty,” Operations Research, 20 (5), 961–981.
Bettman, James (1973), “Perceived Risk and Its Components: A
Model and Empirical Test,” Journal of Marketing Research,
10 (2), 184–190.
Bitner, Mary Jo, Bernard H. Booms, and Mary Stanfield Tetrault
(1990), “The Service Encounter: Diagnosing Favorable and
Unfavorable Incidents,” Journal of Marketing, 54 (1), 71–84.
Bloch, Peter H., and Marsha L. Richins (1983). “A Theoretical
Model for the Study of Product Importance Perceptions,”
Journal of Consumer Research, 47 (3), 69–81.
———, Daniel L. Sherrell, and Nancy M. Ridgway (1986), “Consumer
Research: An Extended Framework,” Journal of Consumer
Research, 13 (1), 119–126.
Bozzolo, Anita, and Timothy C. Brock (1992), “Unavailability
Effects on Message Processing: A Theoretical Analysis and
Empirical Test,” Basic and Applied Social Psychology, 13 (1),
93–101.
Brannon, Laura A., and Timothy C. Brock (2001), “Limiting Time
for Responding Enhances Behavior Corresponding to the
Merits of Compliance Appeals: Refutations of HeuristicCue Theory in Service and Consumer Settings,” Journal of
Consumer Psychology, 10 (3), 133–146.
Brock, Timothy C. (1968), “Implications of Commodity Theory
for Value Change,” in Psychological Foundations of Attitudes,

Anthony G. Greenwald, Timothy C. Brock, and Thomas M.
Ostrum, eds., New York: Academic Press, 243–275.
Brucks, Merrie (1985), “The Effects of Product Class Knowledge on
Information Search Behavior,” Journal of Consumer Research,
12 (1), 1–16.
Burton, Scot, Donald Lichtenstein, Richard Netemeyer, and Judith
Garretson (1998), “A Scale for Measuring Attitude Toward Private Label Products and an Examination of Its Psychological
and Behavioral Correlates,” Journal of the Academy of Marketing
Science, 26 (4), 293–306.
Campbell, Margaret C. (1999), “Perceptions of Price Unfairness:
Antecedents and Consequences,” Journal of Marketing Research,
36 (2), 187–199.
Celsi, Richard L., and Jerry C. Olson (1988), “The Role of Involvement in Attention and Comprehension Processes,” Journal of
Consumer Research, 15 (2), 210–224.
Cialdini, Robert B. (1993), Influence: Science and Practice, New York:
HarperCollins.
DeGraba, Patrick (1995), “Buying Frenzies and Seller-Induced Excess
Demand,” Journal of Economics, 26 (2), 331–342.
Dhar, Ravi, and Stephen M. Nowlis (1999), “The Effects of Time
Pressure on Consumer Choice Deferral,” Journal of Consumer
Research, 6 (4), 389–405.
Estelami, Hooman (2003), “The Effect of Price Presentation Tactics
on Consumer Evaluation Effort of Multi-Dimensional Prices,”
Journal of Marketing Theory and Practice, 11 (1), 1–15.
Ferdows, Kasra, Michael A. Lewis, and Jose A.D. Machuca (2004),
“Rapid-Fire Fulfillment,” Harvard Business Review, 82 (11),
104–110.
Folkes, Valerie S. (1988), “The Availability Heuristic and Perceived
Risk,” Journal of Consumer Research, 15 (1), 13–23.
Forehand, Mark R., and Rohit Deshpandé (2001), “What We
See Makes Us Who We Are: Priming Self-Awareness and
Advertising Response,” Journal of Marketing Research, 38 (3),
336–348.
Frank, Robert H. (2007), The Natural Economist: In Search of Explanations for Everyday Enigmas, New York: Basic Books.
Hall, Robert E. (1978), “Stochastic Implications of the Life CyclePermanent Income Hypothesis: Theory and Evidence,” Journal
of Political Economy, 86 (6), 971–988.
Holmes, Elizabeth (2011), “Retailers to Rein in Discounts,” Wall
Street Journal, July 8 (available at http://online.wsj.com/article/
SB10001424052702303544604576431572289761998.html).
Inman, J. Jeffrey, Anil C. Peter, and Priya Raghubir (1997), “Framing
the Deal: The Role of Restrictions in Accentuating Deal Value,”
Journal of Consumer Research, 24 (1), 68–79.
International Council of Shopping Centers (2009), “How the
Recession Has Impacted Consumer Shopping Habits,” 2009
Shopping Habits Report, New York.
Jacobson, Robert, and Carl Obermiller (1990),“The Formation of
Expected Future Price: Reference Price for Forward-Looking
Customers,” Journal of Consumer Research, 16 (4), 420–432.
Kahneman, Daniel, and Amos Tversky (1979), “Prospect Theory:
An Analysis of Decision Under Risk,” Econometrica, 47 (2),
263–291.
Laurent, Gilles, and Jean Noël Kapferer (1985), “Measuring Consumer Involvement Profiles,” Journal of Marketing Research,
22 (1), 41–53.
Lichtenstein, Donald R., Peter H. Bloch, and William C. Black
(1988), “Correlates of Price Acceptability,” Journal of Consumer
Research, 15 (2), 243–252.

Fall 2013  453
Lynn, Michael (1989), “Scarcity Effects on Desirability: Mediated
by Assumed Responsiveness?” Journal of Economic Psychology,
10 (2), 257–274.
MacKinnon, David P., and James H. Dwyer (1993), “Estimating
Mediated Effects in Prevention Studies,” Education Review,
36 (4), 144–158.
Mattioli, Dana (2012), “Lululemon’s Secret Sauce” Wall Street
Journal, March 22 (available at http://online.wsj.com/article/
SB10001424052702303812904577295882632723066.html).
Mazumdar, Tridib, S.P. Raj, and Indrajit Sinha (2005), “Reference
Price Research: Review and Propositions,” Journal of Marketing, 69 (4), 84–102.
Mittal, Banwari (1995), “A Comparative Analysis of Four Scales
of Consumer Involvement,” Psychology & Marketing, 12 (7),
663–682.
Mongin, Philippe (1997), “Expected Utility Theory,” in Handbook
of Economic Methodology, John Bryan Davis, D. Wade Hands,
and Uskali Maki, eds., London: Edward Elgar, 342–350.
Monroe, Kent B. (1973), “Buyers’ Subjective Perceptions of Price,”
Journal of Marketing Research, 10 (1), 70–80.
——— (2003), Pricing: Making Profitable Decisions, 3d ed., Burr
Ridge, IL: Irwin.
Neubert, Mitchell J., K. Michelle Kacmar, Dawn S. Carlson, Lawrence B. Chonko, and James A. Roberts (2008), “Regulatory
Focus as a Mediator of the Influence of Initiating Structure
and Servant Leadership on Employee Behavior,” Journal of
Applied Psychology, 93 (6), 1220–1233.
Pan, Yue, and George Zinkhan (2006), “Determinants of Retail
Patronage: A Meta-Analytic Perspective,” Journal of Retailing,
82 (3), 229–243.
Rao, Akshay, and Wanda A. Sieben (1992), “The Effect of Prior
Knowledge on Price Acceptability and the Type of Information Examined,” Journal of Consumer Research, 19 (2),
256–270.
Richins, Marsha L., and Peter H. Bloch (1986), “After the New
Wears Off: The Temporal Context of Product Involvement,”
Journal of Consumer Research, 13 (2), 280–285.
Rook, Dennis W. (1987), “The Buying Impulse,” Journal of Consumer Research, 14 (2), 189–199.
———, and Robert J. Fisher (1995), “Normative Influences on
Impulsive Buying Behavior,” Journal of Consumer Research,
22 (3), 305–313.
Rothschild, Michael L. (1984), “Perspectives on Involvement:
Current Problems and Future Directions,” in Advances in
Consumer Research, vol. 11, Thomas C. Kinnear, ed., Provo,
UT: Association for Consumer Research, 216–217.

Samli, A. Coskun, and Laurence W. Jacobs (1993), “International
Pricing Decisions: A Diagnostic Approach,” Journal of Marketing Theory and Practice, 1 (4), 29–41.
Sevdalis, Nick, Nigel Harvey, and Michelle Yip (2006), “Regret
Triggers Inaction Inertia—But Which Regret and How?” British Journal of Social Psychology, 45 (4), 839–853.
Sivakumar, K. (1996), “Tradeoff Between Frequency and Depth of
Price Promotions,” Journal of Marketing Theory and Practice,
4 (1), 1–8.
Sobel, Michael E. (1982), “Asymptotic Confidence Intervals for
Indirect Effects in Structural Equations Models,” in Sociological Methodology, Samuel Leinhart, ed., San Francisco:
Jossey-Bass, 290–312.
Suri, Rajneesh, and Kent B. Monroe (2003), “The Effects of Time
Constraints on Consumers’ Judgments of Prices and Products,” Journal of Consumer Research, 30 (1), 92–104.
———, Chiranjeev Kohli, and Kent B. Monroe (2007), “The Effects
of Perceived Scarcity on Consumers’ Processing of Price
Information,” Journal of the Academy of Marketing Science,
35 (1), 89–100.
Sweeney, Jillian C., and Geoffrey N. Soutar (2001), “Consumer
Perceived Value: The Development of a Multiple Item Scale,”
Journal of Retailing, 77 (2), 203–220.
Tsiros, Michael, and David M. Hardesty (2010), “Ending a Price
Promotion: Retracting It in One Step or Phasing It out Gradually,” Journal of Marketing, 74 (1), 49–64.
———, and Vikas Mittal (2000), “Regret: A Model of its Antecedents
and Consequences in Consumer Decision Making,” Journal
of Consumer Research, 26 (4), 401–417.
Tversky, Amos, and Daniel Kahneman (1973), “Availability: A
Heuristic for Judging Frequency and Probability,” Cognitive
Psychology, 5 (2), 201–232.
von Neumann, John, and Oskar Morgenstern (1947), Theory of
Games and Economic Behavior, 2d ed., Princeton: Princeton
University Press.
White, Irving S. (1966), “The Perception of Value in Products,”
in On Knowing the Consumer, J.W. Newman, ed., New York:
Wiley, 90–106.
Winer, Russel S. (1986), “A Reference Price Model of Brand Choice
for Frequently Purchased Products,” Journal of Consumer
Research, 13 (2), 250–256.
Zaichkowsky, Judith Lynn (1985), “Measuring the Involvement
Construct,” Journal of Consumer Research, 12 (3), 341–352.

454  Journal of Marketing Theory and Practice

Appendix A
Scenario and Items for Study 1
A lot of people have a favorite musical band or artist. Think about your favorite band or artist and write it here
__________________.
About how much would you be willing to spend for a live CD of this band or artist’s first-ever performance? Please
write the amount here __________________.
Now imagine that this Saturday morning you go your local flea market and find that CD [three of them in fact] except
it is priced $5 higher than what you wrote above. You know that the flea market has a policy of selling any unsold items
at half off the following Saturday, but you have no idea if [any of] the CD[s] will still be there next week. You cannot hide
a CD, place one on hold, or do anything of that nature. Would you purchase a CD that day or come back the following
Saturday with the hope that at least one was left and you could buy it for half off?
Table A1
Scale Items with Factor Loadings and Reliabilities for Study 1
α
Involvement with Product Class (Laurent and Kapferer 1985)
The clothes I buy are very important to me.
For me, it matters what clothes I own.
Clothes are an important part of my life.
Perceived Emotional Value (Sweeney and Soutar 2001)
This CD is one that I would enjoy.
Buying this CD would make me want to listen to it.
Owning this CD would make me feel good.
Owning this CD would give me pleasure.
I would feel relaxed about owning this CD.
Risk Aversion (Burton et al. 1998)
I don’t like to take risks.
Compared to most people I know, I like to gamble on things.
Compared to most people I know, I like to live on the edge.
I have no desire to take unnecessary chances on things.
Impulse Buying (Rook and Fisher 1995)
I often buy things spontaneously.
“Just do it” describes the way I buy things.
I often buy things without thinking.
“I see it, I buy it” describes me.
“Buy now, think about it later” describes me.
Sometimes I feel like buying things spur of the moment.
I buy things according to how I feel at the moment.
I carefully plan most of my purchases.
Sometimes I am a bit reckless about what I buy.
Note: Items anchored by 1 = “strongly disagree” to 7 = “strongly agree.”

λ

0.85
0.91
0.88
0.85
0.91
0.85
0.86
0.89
0.90
0.81
0.76
0.74
0.82
0.85
0.60
0.91
0.80
0.85
0.86
0.86
0.80
0.67
0.72
0.56
0.72

Fall 2013  455

Appendix B
Scenario and Items for Study 2
Imagine that you need a new pair of jeans, and that this Saturday you go to the mall in search of a pair. A lot of people
have a favorite or “go-to” brand when it comes to jeans. Think about your favorite brand or a brand you like and write
it here ________________________.
Imagine that the brand you wrote above sells one pair of jeans that has the exact style, fit, color, etc., that you are
looking for. About how much would you be willing to spend for a pair of these jeans? Please write the amount
here ________________________. (If you cannot think of a specific brand, write how much you’d pay for a really nice pair
of jeans with the exact style, fit, color, etc. you like)
Now imagine that you enter a store that is running a promotion on the brand of jeans that you wrote above. You soon
find a table display with 10 [2] pairs of the exact style, fit, color, etc., that you are looking for in your size. Above the table
hangs a sign that reads, “All jeans 50% off [25% off, 10% off] next Saturday.” Keeping in mind that this pair of jeans will
not be available in other stores or online, you begin to debate purchasing the jeans today or coming back next Saturday for
the discount.
Table B1
Scale Items with Factor Loadings and Reliabilities for Study 2
α
Involvement with Product Class (Laurent and Kapferer 1985)
The clothes I buy are very important to me.
For me, it matters what clothes I own.
Clothes are an important part of my life.
Impulse Buying (Rook and Fisher 1995)
I often buy things spontaneously.
“Just do it” describes the way I buy things.
I often buy things without thinking.
“I see it, I buy it” describes me.
“Buy now, think about it later” describes me.
Sometimes I feel like buying things spur of the moment.
I buy things according to how I feel at the moment.
Sometimes I am a bit reckless about what I buy.
Note: Items anchored by 1 = “strongly disagree” to 7 = “strongly agree.”

λ

0.91
0.85
0.85
0.82
0.94
0.79
0.86
0.85
0.86
0.83
0.80
0.82
0.78