Yes, We Have No Bananas: Consumer Responses To Restoration of Freedom

Research Report
Yes, we have no bananas: Consumer responses to restoration of freedom
Sarah G. Moore
a,
, Gavan J. Fitzsimons
b
a
Alberta School of Business, University of Alberta, Edmonton, AB T6G 2R6, Canada
b
Fuqua School of Business, Duke University, Durham, NC 27708, United States
Received 16 April 2013; received in revised form 27 March 2014; accepted 3 April 2014
Available online xxxx
Abstract
When stockouts restrict consumers' freedoms, two independent responses can occur: product desirability, or a reactance-based increase in the
desire for the unavailable option, and source negativity, or general frustration with the source of the restriction. In four studies, we provide a novel
investigation of consumer responses to stockout-restoration and examine how these two forces combine to affect consumer responses after
freedoms are restored. To do so, we investigate two moderators that inuence the activation and strength of product desirability and source
negativity, respectively: trait reactance and attributions. While all consumers experience source negativity in response to stockouts, only consumers
high in reactance experience product desirability, leading to differential responses to stockout-restoration. Compared to an in-stock condition, high
reactance consumers respond positively to stockout-restoration, while low reactance consumers respond negatively to stockout-restoration, in
terms of store and product evaluations and store choice. However, when high reactants attribute a stockout to the store, thereby increasing source
negativity relative to product desirability, they respond negatively to stockout-restoration.
2014 Society for Consumer Psychology. Published by Elsevier Inc. All rights reserved.
Keywords: Reactance; Restrictions; Restoration; Freedom; Stockout
Consumers respond to restrictions of freedom such as space
constraints (Levav & Zhu, 2009; Xu, Shen, & Wyer, 2012),
unsolicited recommendations (Fitzsimons & Lehmann, 2004),
and product unavailability (Fitzsimons, 2000; Verhallen &
Robben, 2005) with reactance motivation (Brehm, 1966). This
motivation can elicit positive responses, including an increased
desire for restricted versus available items (Worchel & Brehm,
1971). Some firms have successfully capitalized on such
positive responses. For example, Nintendo's resurrection of
the video game industry in the 1980s has been credited to a
controlled dearth of game cartridges (Berry, 1989; Wolpin,
1989, p. 38), and consumers consistently line up for new and
scarce iPhones (Petrecca, 2012). Some firms have even
advertised that their products are disappearing at a store near
you (Elliott, 1993). Independent of reactance, however,
restrictions also elicit negative responses, including negativity
toward the source of the restriction (Clee & Wicklund, 1980;
Fitzsimons, 2000). Negative responses were widespread when
the launches of Microsoft's Xbox and Nintendo's Wii were
accompanied by significant shortages (Huang, 2007; Robischon,
2005). Thus, prior work has shown product desirability (positive)
and source negativity (negative) responses to restrictions of
freedom.
We investigate positive and negative consumer responses to
restoration of freedom. We use the context of stockouts, which are
common, costly marketplace restrictions (Anderson, Fitzsimons,
& Simester, 2006; Jing & Lewis, 2011; Schary & Christopher,
1979), and explore responses to stockout-restoration, when
formerly unavailable products become available. We consider
how product desirability and source negativity responses to
The authors are grateful to Jim Bettman, Don Lehmann, Abhijit Guha, Yael
Zemack-Rugar, Virginia Weber, Jennifer Argo, and the editor, associate editor,
and review team for their constructive comments on this manuscript. The
authors gratefully acknowledge the support of a Walmart Seed Grant from the
School of Retailing at the Alberta School of Business.
Corresponding author. Tel.: +1 780 492 3664.

E-mail addresses: sarah.g.moore@ualberta.ca (S.G. Moore),
gavan@duke.edu (G.J. Fitzsimons).
http://dx.doi.org/10.1016/j.jcps.2014.04.001
1057-7408/ 2014 Society for Consumer Psychology. Published by Elsevier Inc. All rights reserved.
Please cite this article as: Moore, S.G., & Fitzsimons, G.J., Yes, we have no bananas: Consumer responses to restoration of freedom, Journal of Consumer Psychology
(2014), http://dx.doi.org/10.1016/j.jcps.2014.04.001
Available online at www.sciencedirect.com
ScienceDirect
Journal of Consumer Psychology xx, x (2014) xxxxxx
JCPS-00410; No. of pages: 8; 4C:
stockouts impact product and store evaluations post-restoration. In
other words, what happens when Xbox or Wii consoles become
available following a shortage? Do consumers forgive or punish
the firm? Do they like or dislike the product?
Responses to restoration
Prior work argues that reactance will be reduced by the
restoration of freedom (Worchel & Brehm, 1971, p. 294).
Supporting this prediction, some studies find decreases in
product desirability post-restoration (Worchel & Brehm, 1971),
though others do not (Schwarz, 1984). However, some of these
studies may not have induced reactance (Schwarz, 1984), and
others lack control conditions (Pallak & Sullivan, 1979),
complicating interpretation. Further, prior work has focused
solely on product desirability after restoration. Thus, it is
unclear if the product desirability and source negativity
instantiated by stockouts are eliminated or persist after
restoration of freedom.
We extend this work by arguing that product desirability and
source negativity might 1) persist and 2) jointly influence
responses to restoration, and provide a framework to predict
when and how they do so. First, since chronic goalssuch as
the pursuit of freedomcan influence behavior even after they
are (temporarily) satisfied (Fishbach, Eyal, & Finkelstein,
2010; Moore, Ferguson, & Chartrand, 2011), product desir-
ability and source negativity might persist and influence
product and store evaluations post-restoration. Second, halo
effects demonstrate spill-over between and among overall
attitudes and attribute-specific attitudes (Kardes, Posavac, &
Cronley, 2004; Wirtz & Bateson, 1995), suggesting that the
product desirability and source negativity elicited by stockouts
might jointly influence both product and store evaluations
post-restoration. Such joint influence is particularly likely in
our studies, as the stockout and restoration share an origin (a
local store).
Specifically, we suggest that responses to restoration will be
jointly driven by the relative strengths of consumers' product
desirability and source negativity responses to stockouts, and that
the stronger of these two forces will dominate (Chowdhury,
Olsen, & Pracejus, 2008; Fiske, 1980; Taylor & Thompson,
1982). If product desirability outweighs source negativity in
response to a stockout, responses to restoration should be positive;
if source negativity outweighs product desirability in response to a
stockout, responses to restoration should be negative. Thus, either
positive or negative responses to restorationin terms of product
and store evaluationsare possible. To investigate when these
occur, we identify two moderators that should influence the
activation and strengths of source negativity and product
desirability, respectively: trait reactance and attributions
regarding the stockout (Fig. 1).
Trait reactance should determine whether a stockout activates
product desirability; high reactants experience product desirabil-
ity when restricted, while low reactants do not (Hong & Faedda,
1996). Further, while product desirability is a motivational
response uniquely linked to reactance and the goal to restore
freedom, source negativity is a generalized affective response to
restriction, independent of reactance (Brehm, 1966; Wicklund,
1974, 2001). Thus, all individuals should respond to stockouts
with source negativity, but only high reactants should simulta-
neously be motivated by product desirability. Relative to an
in-stock condition, then, high reactants should respond to
stockouts with negative store evaluations and positive product
evaluations, while low reactants should respond to stockouts
with negative store evaluations and neutral or negative product
evaluations. Further, since source negativitybut not product
desirabilityis activated for low reactants in response to
stockouts, relative to an in-stock condition, low reactants should
respond negatively to stockout-restoration. However, relative
to an in-stock condition, high reactants should respond positively
to stockout-restoration; although both product desirability and
source negativity are activated for high reactants in response to
stockouts, we argue that product desirability will play a greater
role in determining responses to restoration, due to its motivational
nature (Brehm, 1966).
In addition, reactance and attributions should interact to
predict responses to restoration. While reactance should
determine the activation of source negativity and product
desirability, their relative strengths should be influenced by
attributions. We examine product popularity and store failure
attributions for stockouts, which should strengthen product
desirability (Verhallen & Robben, 1994; Ge, Messinger, & Li,
2009) and source negativity (Bitner, 1990), respectively. Since
high reactants experience both product desirability and source
negativity in response to stockouts, the stronger force should
drive responses to restoration. Thus, high reactants should
respond positively to restoration following popularity attribu-
tions and negatively to restoration following store failure
attributions. In contrast, since low reactants do not experience
product desirability in response to stockouts, there is no
motivational force to strengthen; thus, popularity attributions
should not affect their responses to restoration. Rather, source
Store evaluation
Product evaluation
Store choice
Trait reactance
Attributions
Restoration of
freedom
Restriction of
freedom
Source negativity
Product desirability
Attributions
Fig. 1. Responses to restriction and restoration of freedom.
2 S.G. Moore, G.J. Fitzsimons / Journal of Consumer Psychology xx, x (2014) xxxxxx
negativity should drive low reactants' (negative) responses,
regardless of attributions.
We test these predictions in four studies. Studies 1 and 2a/b
examine how reactance moderates responses to restriction and
restoration, while study 3 tests the joint impact of reactance and
attributions on responses to restoration.
Study 1
Study 1 examines high and low reactants' store and product
evaluations after in-stock, stockout, and stockout-restoration
shopping trips.
Methods
Individuals from Mechanical Turk (N = 134) were paid to
complete a 3 (shopping trip: in-stock, stockout, stockout-
restoration) by measured reactance between-subjects design.
Participants thought of a favorite snack food, typed in the type
and flavor of this snack food (e.g., BBQ chips), rated it on a 0
100 scale, and imagined going to their local grocery store to
purchase it. In-stock participants imagined retrieving their item,
paying, and leaving the store. Stockout participants imagined
that their item was not available, and left the store with no item.
Stockout-restoration participants imagined that their item was
not available, then imagined shopping for several other
products; when they returned to the front of the store through
the snack aisle, their item had been re-stocked, so they
purchased it prior to leaving.
Participants then evaluated the store (satisfaction, likelihood
of returning) and their item (satisfaction, likelihood of
repurchasing) on 7-point scales, and completed some demo-
graphic measures. Finally, participants completed the reactance
scale (Hong & Faedda, 1996; 11 items, e.g., I find contradicting
others stimulating; 1 = strongly disagree, 5 = strongly agree).
Results
We created overall store evaluation (r = .81, p b .001),
product evaluation (r = .65, p b .001), and reactance measures
( = .88, M = 2.88, SD = 0.75); reactance was not predicted
by shopping trip (p N .33).
Next, we regressed product evaluations, F
(6, 127)
= 2.59,
p = .02, and store evaluations, F
(6, 127)
= 6.89, p b .0001, on
shopping trip (dummy-coded; dummy1: stockout = 1, 0 else;
dummy2: stockout-restoration = 1, 0 else), reactance (cen-
tered), their interactions, and initial product ratings (centered).
Initial ratings were a positive and significant covariate
(b
product
= 0.0095, t
product
(127) = 2.36, p = .04; b
store
= 0.02,
t
store
(127) = 3.54, p b .001). Below, we discuss simple inter-
action tests by shopping trip for store and product evaluations.
In-stock vs. stockout
Product evaluations showed a shopping trip (dummy1) by
reactance interaction, b = 0.28, t(127) = 2.19, p = .03. High
reactants (+1 SD) rated the product marginally more positively
after the stockout than the in-stock trip, b = 0.31, t(127) =
1.89, p = .06, while low reactants (1 SD) rated the product
marginally more negatively, b = 0.31, t(127) = 1.79, p =
.07.
In contrast, store evaluations showed a simple effect of
shopping trip, b = 0.64, t(127) = 4.81, p b .001, but no
interaction (p N .81). All individuals rated the store more
negatively after the stockout (M = 4.30) than the in-stock trip
(M = 4.94), confirming that, in response to a stockout, both
high and low reactants experience source negativity, but only
high reactants experience product desirability.
In-stock vs. stockout-restoration
Product evaluations revealed a shopping trip (dummy2)
by reactance interaction, b = 0.29, t(127) = 2.59, p = .01.
High reactants rated their product more positively, b = 0.30,
t(127) = 2.00, p = .05, but low reactants rated their product
more negatively, b = 0.34, t(127) = 2.25, p = .03, after the
stockout-restoration than the in-stock trip.
Store evaluations also showed only a shopping trip by
reactance interaction, b = 0.46, t(127) = 2.80, p b .01. High
reactants evaluated the store directionally more positively, b =
0.31, t(127) = 1.42, p = .15, but low reactants evaluated the
store significantly more negatively, b = 0.71, t(127) = 3.16,
p b .01, after the stockout-restoration than the in-stock trip. Thus,
compared to an in-stock condition, high reactants responded
positively, but low reactants responded negatively, to stockout-
restoration.
Stockout vs. stockout-restoration
For this comparison, dummy1 was re-coded (in-stock = 1, 0
else); dummy2 then tests stockout against stockout-restoration.
Product evaluations did not differ by reactance, shopping
trip (dummy2), or their interaction across the stockout and
stockout-restoration trips (ps N .91), suggesting that stockout-
restoration has no independent effect on product desirability
beyond the stockout.
In contrast, store evaluations showed a simple effect of
shopping trip (M
stockout
= 4.30, M
restoration
= 4.73), b = 0.44,
t(127) = 3.37, p b .001, and a trip by reactance interaction,
b = 0.42, t(127) = 2.30, p = .02. High reactants evaluated the
store more positively after the stockout-restoration than the
stockout trip, b = 0.90, t(127) = 3.69, p b .001, but low
reactants' evaluations did not differ (p N .93), confirming
high reactants' positivity and low reactants' negativity in
response to restoration.
Discussion
Study 1 provides initial evidence that high and low reactants
respond differently to restriction and restoration of freedom.
Compared to an in-stock scenario, all individuals responded to
a stockout with negative store evaluations, but only high
reactants responded with positive product evaluations. Similar-
ly, compared to an in-stock scenario, high reactants responded
more positively to the store and product after a stockout-
restoration. Low reactants showed the opposite effects. Thus,
relative to a regular, in-stock transaction, high reactants are
happier, and low reactants are unhappier, after their freedoms
have been restricted and restored.
Studies 2a & 2b
We next replicated study 1 using actual choice. Across
several within-subject shopping trips, undergraduates selected
and received different flavors of jellybeans. Shopping trip
outcome was manipulated. On in-stock trips, jellybeans were
received without incident; on stockout trips, no jellybeans were
received; on stockout-restoration trips, jellybeans were initially
unavailable but then were received.
Study 2a
Methods
Undergraduates (N = 53) received cash and jellybeans for
completing a 3 (within, shopping trip: in-stock, stockout,
stockout-restoration) by 6 (between: trip order) mixed design
with reactance as a measured variable. Participants were not
told how many shopping trips they would make, to minimize
stockpiling (Blattberg, Eppen, & Lieberman, 1981) and variety
seeking (Kahn, 1995). Participants were informed that the
flavors available would vary across trips. Shopping trips were
described on-screen in the second-person, where participants
shopped for jellybeans and interacted with a local shopkeeper.
On each of ten trips, participants chose one flavor from a list of
five available flavors. There were ten flavors in total. On the
first and second trips, participants were exposed to all ten
flavors in two random sets of five. On the remaining eight trips,
participants saw two other sets of five flavors. Set 1 flavors
appeared on the in-stock, stockout, and stockout-restoration
trips. Set 2 flavors appeared on the filler trips.
After choosing, participants raised their hands, and the
experimenter observed their jellybean choice; on the filler and
in-stock trips, they received a scoop of that flavor. On the
stockout trip, participants chose a flavor but were informed by
the shopkeeper on-screen (before raising their hands) that
those jellybeans were out-of-stock: participants received no
jellybeans. On the stockout-restoration trip, participants select-
ed a flavor, but were informed by the shopkeeper that those
jellybeans were out-of-stock. Participants were asked to
re-choose; however, after they made a second selection, the
shopkeeper discovered some of their original jellybeans, and
participants received these from the experimenter.
After each shopping trip, participants evaluated the store
(satisfaction, return), their chosen jellybeans (satisfaction,
repurchase, attractiveness), and the other jellybean flavors
available on that trip (attractiveness), on 7-point scales. The
next shopping trip then commenced. Finally, participants
completed demographic questions and the reactance scale.
Results
First, we created overall reactance ( = .83, M = 2.94,
SD = 0.70), store evaluation (rs N .70, ps b .001), and product
evaluation measures (s N .69). Order showed no main effects
or interactions with shopping trip or reactance on evaluations
(ps N .19) and was excluded from subsequent analyses.
We used a repeated measures ANOVA with shopping trip
and reactance to predict evaluations, and found significant
interactions for store, F
(2, 102)
= 4.54, p = .01, and product
evaluations, F
(2, 102)
= 3.52, p = .04. Below, we report paired
comparisons by shopping trip.
In-stock vs. stockout
Product evaluations revealed a simple effect of shopping trip
(M
in-stock
= 5.22, M
stockout
= 4.00), F
(1, 51)
= 9.56, p = .003,
and a trip by reactance interaction, F
(1, 51)
= 4.53, p = .04.
High reactants (+1 SD) evaluated the jellybeans more positively
after the stockout than the in-stock trip, F
(1, 51)
= 8.55, p = .005,
while low reactants (1 SD) showed the opposite pattern,
F
(1, 51)
= 11.19, p b .001.
For store evaluations, we also found a simple effect of
shopping trip (M
in-stock
= 5.10, M
stockout
= 2.36), F
(1, 51)
= 22.39,
p b .001, and a trip by reactance interaction, F
(1, 51)
= 5.78, p =
.02. While all individuals evaluated the store more negatively after
the stockout, low reactants, F
(1, 51)
= 28.84, p b .001, responded
more negatively than high reactants, F
(1, 51)
= 18.61, p b .001.
Consistent with study 1, this pattern demonstrates high reactants'
product desirability and all individuals' source negativity in
response to a stockout.
In-stock vs. stockout-restoration
Product evaluations showed a simple effect of shopping trip
(M
in-stock
= 5.22, M
restoration
= 5.51), F
(1, 51)
= 7.42, p = .009,
and a trip by reactance interaction, F
(1, 51)
= 10.26, p = .002.
High reactants evaluated their jellybeans more positively, F
(1, 51)
=
7.98, p = .006, while lowreactants evaluated their jellybeans more
negatively, F
(1, 51)
= 6.53, p = .01, after the stockout-restoration
than the in-stock trip.
Similarly, store evaluations showed a simple effect of
shopping trip (M
in-stock
= 5.10, M
restoration
= 4.89), F
(1, 51)
=
7.72, p = .008, and a trip by reactance interaction, F
(1, 51)
=
6.92, p = .01. High reactants rated the store more positively,
F
(1, 51)
= 7.63, p = .008, but low reactants rated the store
more negatively, F
(1, 51)
= 7.81, p = .007, after the stockout-
restoration than the in-stock trip. Thus, compared to an in-stock
condition, high reactants respond more positively, while low
reactants respond more negatively, to the store and product
following a restriction and restoration.
Stockout vs. stockout-restoration
Product evaluations showed only a simple effect of shopping
trip (M
stockout
= 4.00, M
restoration
= 5.17), F
(1, 51)
= 4.60, p =
.04, which became non-significant (p N .79) when in-stock
trip product evaluations were used to covary out ownership
differences. As with study 1, this suggests that restoration has
no independent effect on product evaluations beyond the
product desirability instantiated by the stockout.
Again controlling for in-stock trip product evaluations,
store evaluations showed a simple effect of shopping trip
(M
stockout
= 2.68, M
restoration
= 4.91), F
(1, 50)
= 11.07, p = .001,
and a directional trip by reactance interaction, F
(1, 50)
= 2.39, p =
.12. All individuals rated the store more positively after the
stockout-restoration than the stockout trip, though high reactants
responded more strongly, F
(1, 50)
= 34.04, p b .001, than low,
F
(1, 50)
= 11.09, p b .001. Lowreactants' positive store responses
to a stockout-restoration compared to a stockoutin contrast
with their neutral responses in study 1may be because this
study uses real choice.
Study 2b
Study 2b examined behavioral responses to restoration. The
study was identical to study 2a, but had only five shopping
trips: three filler trips, an in-stock trip, and a stockout-
restoration trip. In addition to reporting store (7-points:
satisfaction, return) and product evaluations (7-points: satisfac-
tion, repurchase; 100-points: attractiveness), undergraduates
(N = 120) chose between two stores before each trip.
Using a repeated measures ANOVA for store (rs N .76,
ps b .001) and product (standardized; s N .72) evaluations,
we found only reactance ( = .80, M = 2.90, SD = 0.64) by
shopping trip interactions (F
store
(1, 116) = 4.10, p = .05;
F
product(1, 116)
= 3.84, p = .05). High reactants (+1 SD)
evaluated the store and product more positively after the
stockout-restoration than the in-stock trip (F
store(1, 116)
= 4.43,
p = .04, F
product(1, 116)
= 3.71, p = .05), while low reactants
(1 SD) evaluated the store and product more negatively
(F
store(1, 116)
= 4.54, p = .04, F
product(1, 116)
= 3.54, p = .06
(Fig. 2).
A repeated measures logistic regression revealed a similar
shopping trip by reactance interaction for store choice,
2
(1,
119) = 3.59, p = .05. High reactants were more likely,
2
(1,
119) = 3.90, p = .05, while low reactants were less likely,
2
(1, 119) = 4.02, p = .04, to return to the same store after the
stockout-restoration than the in-stock trip (Fig. 3).
Finally, we conducted additional analyses to corroborate our
product desirability findings and rule out generalized affect
as an alternative explanation. If our results stem from a
reactance-based desire for a restricted item, product desirability
responses should be specific to the formerly unavailable
jellybean. Supporting this idea, non-chosen flavor ratings did
not differ by shopping trip, reactance, or their interaction in
study 2a or 2b (ps N .20).
Discussion
These studies provide further evidence that high and low
reactants respond differently to stockout-restoration, and confirm
that source negativity and product desirability carry over
post-stockout. Compared to an in-stock condition, after a
stockout, all individuals evaluated the store more negatively, but
only high reactants evaluated the product more positively.
However, compared to an in-stock condition, after a stockout-
restoration, high reactants evaluated the store and product more
positively, and lowreactants evaluated the store and product more
negatively. These evaluations were mirrored in store choice: high
reactants were more likely, and low reactants were less likely,
to stay with a store after a stockout-restoration. In short, high
reactants reward firms, but low reactants punish firms, for
restricting and then restoring their freedom.
Note: In study 2b, jellybean attractiveness was measured on a 0-100 scale and satisfaction and likelihood of repurchase
were measured on 1-7 scales; thus, product evaluation was standardized for analysis.
1.00
2.00
3.00
4.00
5.00
6.00
7.00
S
t
o
r
e

e
v
a
l
u
a
t
i
o
n

Reactance
STORE EVALUATION BY
SHOPPING TRIP AND REACTANCE
In-stock Stockout-restoration
-1.00
-0.75
-0.50
-0.25
0.00
0.25
0.50
0.75
1.00
1 (Low) 2 3 4 5 (High) 1 (Low) 2 3 4 5 (High)
P
r
o
d
u
c
t

e
v
a
l
u
a
t
i
o
n
(
s
t
a
n
d
a
r
d
i
z
e
d
)
Reactance
PRODUCT EVALUATION BY
SHOPPING TRIP AND REACTANCE
Fig. 2. Store and product evaluations by shopping trip and reactance, study 2b.
0%
10%
20%
30%
40%
50%
60%
70%
1 (Low) 2 3 4 5 (High)
S
t
o
r
e

s
t
a
y
i
n
g
Reactance
STORE STAYING BY SHOPPING TRIP
AND REACTANCE
Fig. 3. Store staying by shopping trip and reactance, study 2b.
Study 3
In study 3, we manipulate whether a stockout is due to
product popularity or to the store's failure to order products,
and investigate how attributions and reactance interact to
predict responses to stockout-restoration. For low reactants, we
expect consistently negative responses to restoration. Since low
reactants do not experience product desirability in response to
stockouts, even popularity attributions will not alter their
responses to restoration, and source negativity should elicit
negative responses across attribution conditions. For high
reactants, popularity attributions should strengthen product
desirability and replicate the positive responses to restoration
seen previously; however, store failure attributions should
strengthen source negativity relative to product desirability and
attenuate or even reverse these positive responses.
Methods
Individuals (N = 291, 55% female, M
age
= 40) in a national
online panel were paid for completing a between-subjects study
with measured reactance and shopping trip (3: in-stock, stockout-
restoration with popularity attribution, stockout-restoration with
store failure attribution). Participants went on one hypothetical
shopping trip and chose between five flavors of jellybeans. In
the in-stock condition, individuals requested and received
their jellybeans with no incident. In the stockout-restoration
conditions, individuals were exposed to the same manipulation
as in studies 2a/b, where their jellybeans were allegedly
out-of-stock. In the popularity attribution condition, the shop-
keeper stated that the store was out-of-stock because their
jellybeans were popular. In the store failure attribution condition,
he stated the store was out-of-stock because he had not ordered
their jellybeans last week. After this, in both stockout-restoration
conditions, the shopkeeper discovered that their jellybeans were
in-stock.
Following the scenario, participants reported store and
product evaluations (7-points: satisfaction, return, repurchase)
and their feelings toward the store and their jellybeans
(5-points: happy, excited, upset, angry). These items were
reversed as appropriate, standardized, and combined into
overall store ( = .80) and product ( = .85) evaluations, with
higher numbers indicating positivity. After answering some
demographic questions, participants completed the reactance
scale ( = .89); reactance was not predicted by shopping trip
(p N .60).
Results
We used reactance (centered), shopping trip (dummy-coded;
dummy1: popularity attribution = 1, 0 else; dummy2: store
failure attribution = 1; 0 else), and their interactions to predict
store, F
(5, 285)
= 18.37, p b .001, and product evaluations,
F
(5, 285)
= 26.05, p b .001. Below, we discuss simple interaction
tests by shopping trip for store and product evaluations.
In-stock vs. stockout-restoration with popularity attribution
For product evaluations, we found only a shopping trip
(dummy1) by reactance interaction, b = 0.37, t(285) = 2.07,
p = .04. Compared to the in-stock trip, high reactants (+1 SD)
evaluated their jellybeans directionally more positively, b = 0.44,
t(285) = 1.61, p = .10, while low reactants (1 SD) evaluated
their jellybeans more negatively, b = 0.59, t(285) = 2.12,
p = .03, after the stockout-restoration with popularity attribution
trip.
Store evaluations also showed only a shopping trip by
reactance interaction, b = 0.54, t(285) = 2.86, p b .01.
Compared to the in-stock trip, high reactants evaluated the
store more positively, b = 0.68, t(285) = 2.34, p = .02, while
low reactants evaluated the store more negatively, b = 0.82,
t(285) = 2.82, p b .01, after the stockout-restoration with
popularity attribution trip.
In-stock vs. stockout-restoration with store failure attribution
For product evaluations, we found only a shopping trip
(dummy2) by reactance interaction, b = 0.36, t(285) = 2.29,
p = .02. Compared to the in-stock trip, low reactants rated their
jellybeans more negatively after the stockout-restoration with
store failure attribution trip, b = 0.91, t(285) = 3.61, p b
.001, though high reactants' product evaluations did not differ
across trips (p N .70).
Store evaluations revealed only a simple effect of shopping
trip, b = 0.28, t(285) = 2.24, p = .03, and no trip by
reactance interaction (p N .17). Compared to the in-stock trip
(M = 0.15), all individuals evaluated the store more negatively
after the stockout-restoration with store failure attribution trip
(M = 0.17),
Discussion
Study 3 examined the impact of stockout attributions on the
relative strengths of source negativity and product desirability,
and thus on responses to restoration. We replicated our
previous findings for stockout-restorations with popularity
attributions: relative to an in-stock condition, high reactants
responded positively to restoration, while low reactants
responded negatively. In contrast, high reactants' positive
product responses were attenuated, and their positive store
responses were reversed, for stockout-restorations with store
failure attributions. We argue that this occurred because
popularity attributions reinforced high reactants' product
desirability relative to source negativity, while stockout-
restorations with store failure attributions strengthened their
source negativity relative to product desirability. Consistent
with previous studies, low reactants responded negatively to
restoration, regardless of attributions; we argue that their lack of
product desirability allows source negativity to dominate their
responses to restrictions.
General discussion
This paper examined consumers' responses to restoration of
freedom in the marketplace, extending past work that has focused
on responses to restrictions. We argued that responses to
restoration are driven by the combination of product desirability
and source negativity elicited by restrictions, and examined two
moderators that affect the activation and strength of these two
forces, respectively: trait reactance and attributions.
We predicted that stockouts would elicit source negativity
for all individuals, but product desirability only for high
reactants, leading to different responses to stockout-restoration.
Supporting our framework, in studies 1 and 2a/b, high reactants
responded positively to stockout-restoration, while low reac-
tants responded negatively, in terms of store and product
evaluations and store choice. In study 3, we examined how
stockout attributions moderated the relative strengths of
product desirability and source negativity for high and low
reactants. High reactants responded positively to stockout-
restoration after popularity attributions, but negatively after
store failure attributions. In contrast, low reactants consistently
responded negatively to stockout-restoration, regardless of
attributions.
This examination of consumer responses to restoration
opens up several pathways for future research. While our
studies demonstrate how restriction-specific moderators such as
reactance and attributions influence responses to restoration
through product desirability and source negativity, restoration-
specific moderators of these responses could also be examined.
For example, responses to restoration likely depend on the
restoration source, which may be the store that incurred the
stockout (as in our studies), or another store. In this latter case,
responses to restoration should not be contaminated by source
negativity, so restoration from a different store may elicit more
positivity than restoration from the same store. Alternately,
consumer- versus firm-restoration could be examined; con-
sumers may feel more source negativity if they locate an
out-of-stock product than if an employee does so. Further,
rather than identifying traits that moderate responses to
restriction (e.g., reactance), traits that moderate responses to
restoration could be identified (e.g., affect intensity; Litt, Khan,
& Shiv, 2010). Finally, in our studies, restoration follows soon
after the stockout. While immediate restoration can occur in the
marketplace, future research could explore responses to
restoration following longer stockouts, with and without signals
of restoration (e.g., rain checks); perhapsat least for high
reactantsthere is an optimal time course for stockouts and
restorations.
Appendix A. Methodological Details
Methodological details for this article can be found online at
http://dx.doi.org/10.1016/j.jcps.2014.04.001.
References
Anderson, E., Fitzsimons, G. J., & Simester, D. (2006). Measuring and
mitigating the costs of stockouts. Management Science, 52(11), 17511763.
Berry, J. (1989). Marketer of the year. Adweeks Marketing Week, 30.
Bitner, M. J. (1990). Evaluating service encounters: the effects of physical
surroundings and employee responses. Journal of Marketing, 54(2), 6982.
Blattberg, R. C., Eppen, G. D., & Lieberman, J. (1981). A theoretical and
empirical evaluation of price deals for consumer nondurables. Journal of
Marketing, 45(1), 116129.
Brehm, J. W. (1966). A theory of psychological reactance. New York:
Academic Press.
Chowdhury, R. M., Olsen, G. D., & Pracejus, J. W. (2008). Affective responses
to images in print advertising: Affect integration in a simultaneous
presentation context. Journal of Advertising, 37(3), 718.
Clee, M. A., & Wicklund, R. A. (1980). Consumer behavior and psychological
reactance. Journal of Consumer Research, 6(4), 389405.
Elliott, S. (1993, April 29). Advertising. The New York Times.
Fishbach, A., Eyal, T., & Finkelstein, S. R. (2010). How positive and negative
feedback motivate goal pursuit. Social and Personality Psychology
Compass, 4(8), 517530.
Fiske, S. (1980). Attention and weight in person perception: The impact of
negative and extreme behavior. Journal of Personality and Social
Psychology, 38(6), 889906.
Fitzsimons, G. J. (2000). Consumer response to stockouts. Journal of Consumer
Research, 27(2), 249266.
Fitzsimons, G. J., & Lehmann, D. R. (2004). Reactance to recommendations:
When unsolicited advice yields contrary responses. Marketing Science,
23(1), 8294.
Ge, X., Messinger, P. R., & Li, J. (2009). Influence of soldout products on
consumer choice. Journal of Retailing, 85(3), 274287.
Hong, S. -M., & Faedda, S. (1996). Refinement of the Hong Psychological
Reactance Scale. Educational and Psychological Measurement, 56(1),
173182.
Huang, E. (2007, March 28). GameStop calls Wii shortage intentional. http://
www.gamepro.com/news.cfm?article_id=107465.
Jing, X., & Lewis, M. (2011). Stockouts in online retailing. Journal of
Marketing Research, 48(2), 342354.
Kahn, B. E. (1995). Consumer variety-seeking among goods and services: An
integrative review. Journal of Retailing and Consumer Services, 2(3),
139148.
Kardes, F. R., Posavac, S. S., & Cronley, M. L. (2004). Consumer inference: A
review of processes, bases, and judgment contexts. Journal of Consumer
Psychology, 14(3), 230256.
Levav, J., & Zhu, R. (2009). Seeking freedom through variety. Journal of
Consumer Research, 36(4), 600610.
Litt, A., Khan, U., & Shiv, B. (2010). Lusting while loathing: Parallel
counterdriving of wanting and liking. Psychological Science, 21(1),
118125.
Moore, S. G., Ferguson, M. J., & Chartrand, T. L. (2011). Affect in the
aftermath: How goal pursuit influences implicit evaluations. Cognition and
Emotion, 25(3), 453465.
Pallak, M. S., & Sullivan, J. J. (1979). The effect of commitment, threat and
restoration of freedom on attitude change and action-taking. Personality and
Social Psychology Bulletin, 5(3), 307310.
Petrecca, L. (2012, September 18). Apple iPhone fans line up for days. http://
usatoday30.usatoday.com/money/business/story/2012/09/18/iphone-fans-
line-up-at-the-5th-ave-apple-store/57803532/1.
Robischon, N. (2005, October 31). Xbox 360: Retail Bully. http://gizmodo.
com/gadgets/home-entertainment/xbox-360-retail-bully-134320.php.
Schary, P. B., & Christopher, M. (1979). The anatomy of a stockout. Journal of
Retailing, 55(2), 5970.
Schwarz, N. (1984). When reactance effects persist despite restoration of
freedom: Investigations of time delay and vicarious control. European
Journal of Social Psychology, 14(4), 405419.
Taylor, S. E., & Thompson, S. C. (1982). Stalking the elusive vividness effect.
Psychological Review, 89(2), 155181.
Verhallen, T. M., & Robben, H. S. (1994). Scarcity and preference: An
experiment on unavailability and product evaluation. Journal of Economic
Psychology, 15(2), 315331.
Verhallen, T. M. M., & Robben, H. S. J. (2005). Unavailability and the
evaluation of goods. KYKLOS, 47(3), 369387.
Wicklund, R. A. (1974). Freedom and reactance. Lawrence Erlbaum.
Wicklund, R. A. (2001). Reactance, psychology. In N. J. Smelser, & P. B.
Baltes (Eds.), International encyclopedia of the social & behavioral
sciences (pp. 1278612789). Oxford: Pergamon.
Wirtz, J., & Bateson, J. E. G. (1995). An experimental investigation of halo
effects in satisfaction measures of service attributes. International Journal
of Service Industry Management, 6(3), 84102.
Wolpin, S. (1989). How Nintendo revived a dying industry. Marketing
Communications, 14, 3640.
Worchel, S., & Brehm, J. W. (1971). Direct and implied social restoration of
freedom. Journal of Personality and Social Psychology, 18(3), 294304.
Xu, J., Shen, H., & Wyer, R. S. (2012). Does the distance between us matter?
Influences of physical proximity to others on consumer choice. Journal of
Consumer Psychology, 22(3), 418423.

Yes, We Have No Bananas: Consumer Responses To Restoration of Freedom

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Yes, We Have No Bananas: Consumer Responses To Restoration of Freedom

Uploaded by

Copyright:

Available Formats

Research Report

Yes, we have no bananas: Consumer responses to restoration of freedom

Corresponding author. Tel.: +1 780 492 3664.

You might also like