Content uploaded by Linda J Skitka
Author content
All content in this area was uploaded by Linda J Skitka on Dec 13, 2017
Content may be subject to copyright.
Patriotism or Nationalism? Understanding
Post-September
11,2001,
Flag-Display Behavior'
LINDA
J.
SKITKA~
Universiry
of
Illinois at Chicago
People reacted to
the
September 11, 2001, terrorist attacks in a number of different ways.
One reaction was to display the American flag on one's home, car, or person. The goal of
this research was to understand the underlying motivations that led to this widespread
behavior. Specifically, to what extent was post-9ilI flag-display behavior motivated by
patriotism
(love
of country and in-group solidarity), nationalism (uncritical acceptance
of
national, state, and political authorities and out-group antipathy), or a combination of
both? Results of a national survey
(N
=
605) provided much stronger support for the
hypothesis that post-9/11 flag-display behavior was
an
expression of patriotism, not
nationalism. Other results supported the notion that patriotism can exist without national-
ism, even in the context of people's reactions to a terrorist attack.
Our country! In her intercourse with foreign nations,
may she always be
in
the right; but our country, right
or wrong.
-Stephen Decatur
(1
8
16)
What do we mean by patriotism in the codtext
of
our
times?
I
venture to suggest that what we mean is a
sense
of
national responsibility
.
.
.
a patriotism which
is not shout, frenzied outbursts
of
emotion, but the
tranquil and steady dedication
of
a lifetime.
-Adlai Stevenson
(1952)
One
of
several reactions Americans had to the terrorist attacks on the World
Trade Center and the Pentagon on September
11,2001,
was an impulse to display
'1
would like to thank Ronnie Janoff-Bulman for prompting me to account for why people flew
the American flag post-9/11, Knowledge Networks for their professional assistance with sampling
and data collection, and Chris Bauman and Liz Mullen
for
their valuable comments on a previous
draft of this manuscript. Data collection and preparation
of
this article were supported by grants from
the National Science Foundation, SBR-0210053 and SBR-OI11612.
*Correspondence concerning this article should be addressed
to
Linda
J.
Skitka, Department of
Psychology (dc 285), University of Illinois at Chicago, 1007
W.
Harrison Street, Chicago,
1L
60607-
71
37.
E-mail: lskitka@uic.edu
1995
Journal
of
Applied
Social Psychology,
2005,35,
10,
pp.
1995-201
1.
Copyright
0
2005
by
V.
H.
Winston
&
Son,
Inc.
All
rights
reserved.
1996
LINDA
J.
SKITKA
the American flag. National surveys found that between
74%
and 82% Americans
reacted to the attacks with flag displays on their homes, cars, or person
(e.g.,
Moore, 2003; Roberts, 2002). What drove the impulse to display the American
flag? Did people fly the flag to express their solidarity with the victims
of
the
attack, the victims’ families, and their fellow citizens? Or did they display the flag
as
a battle standard
to
symbolize dominance and hostility toward out-groups?
Anecdotal evidence suggests that some flew the flag for love of nation.
For
example, Todd Gitlin,
a
former president of Students for Democratic Society (an
activist organization known for burning the American flag to protest American
imperialism abroad) wrote
“I
took inspiration from the patriotic activists who
seem to have brought down Flight 93 over Pennsylvania and probably saved the
White House.
.
,
.
It dawned on me that patriotism was the sum of such acts” (as
cited in Horowitz, 2003). Gitlin then did something he previously felt was
unthinkable: He draped an American flag from his window.
Reactions to the pervasive display
of
the American flag were not, however,
unilaterally positive. For example, one commenter noted “My office
.
.
.
seems
swept into a surreal state of ‘flag waving,’ and ‘kill a coinmie for mommie’ patri-
otism” (AlterNet,
2001).
Barbara Kingsolver
(2001),
like others whose political
consciousness was shaped largely by the Vietnam War, wrote “Patriotism threat-
ens free speech with death” and “the American flag stands for intimidation, cen-
sorship, violence, bigotry, sexism, homophobia, and shoving the Constitution
through a paper shredder.” A number of university administrations apparently
agreed with Kingsolver’s view, or for other reasons also felt squeamish about dis-
plays
of
the American flag. Arizona State, Central Michigan, Lehigh, Marquette,
Texas A
&
M, and a number of other universities banned flag displays following
September
11,
2001, as potentially offensive and insensitive to diversity on cam-
pus (Chow, 2001).
In
sum, people differed in the degree that they felt post-9/11 flag displays
primarily symbolized love of country, uncritical conformity, derogation
of
other
nations, or all of these. The goal
of
the research presented here was to explore
empirically the degree
to
which post-9/11 flag-display behavior is connected to
love of country and affirmation of cultural values versus hostility toward out-
groups and conformity to authority.
Patriotism and Nationalism
The concern about what it means to fly the American flag in response to the
terrorist attacks
is
rooted
in
assumptions about whether national pride necessarily
implies ethnocentricity or xenophobic regard for others. In short, some may be
skeptical that
it
is possible for people to express
patriotism,
defined as love
of
country and attachment to national values (e.g., Adorno, Frenkel-Burnswik,
Levinson,
&
Sanford, 1950; Bar-Tal,
1997;
Kelman, 1997; Kosterman
&
PATRIOTISM
OR
NATIONALISM?
1997
Feshbach, 1989) without also expressing
nationalism,
which is uncritical
acceptance of national, state, and political authorities combined with a belief in
the superiority and dominant status of one’s nation (e.g., Adorno et al., 1950;
Hechter, 2000; Kosterman
&
Feshbach, 1989; Meloen, 1999; Schatz
&
Staub,
1997; Schatz, Staub,
&
Lavine, 1999). Theoretically, patriotism is an affective
attachment to the in-group independent of one’s feelings about the out-group or
authorities. Nationalism, however, is explicitly connected to out-group antipathy
(Blank
&
Schmidt, 1993,2003).
Factor-analytic studies have indicated that patriotism and nationalism are dis-
tinguishable constructs. Different clusters of attitudes correspond to a positive
but critical appreciation of one’s country and its symbols on the one hand, and an
orientation best characterized by “my country, right or wrong” on the other
(Blank
&
Schmidt, 1993, 2003; Kosterman
&
Feshbach, 1989; Schatz
&
Staub,
1997; see also Sullivan, Fried,
&
Dietz, 1992, for a five-factor alternative). Not
surprisingly, however, these constructs are nonetheless somewhat correlated
because they both include a component of positive in-group evaluation
(Kosterman
&
Feshbach, 1989).
In a similar vein, other research taking a social-identity perspective has found
that in-group enhancement and out-group derogation are not like conjoined
twins, such that wherever there is one, there necessarily is the other (Brewer,
1979, 1999; de Figueiredo
&
Elkins, 2003; Peiia
&
Sidanius, 2002).
A
number of
studies have found that increased in-group identification leads to negative out-
group attitudes only when the out-group poses a threat to the in-group (e.g.,
Brewer, 1999; Brown, 1995). Therefore, flag-display behavior may have been a
consequence
of
patriotism without nationalism for those less threatened by the
terrorist attacks. However, among those who were more threatened by the terror-
ist attacks, flag-display behavior may have reflected more nationalism, or a com-
plex blend of nationalism and patriotism.
Consistent with the proposition that flag-display behavior might be a complex
blend of both patriotism and nationalism, value-protection theorists have argued
that people are intuitive prosecutors who respond to moral transgressions (e.g., a
terrorist attack) with a strong sense of motivated arousal and distress (Tetlock,
2002; Tetlock, Kirstel, Elson, Green,
&
Lerner, 2000). This motivated arousal
leads people to respond with both moral outrage (a reaction that includes cogni-
tive, affective, and behavioral components, including negative attributions and
vilification of the transgressor, rage, and punitive behavior) and value affirmation
(attempts to morally cleanse by reaffirming one’s commitment to important cul-
tural and moral values or by doing good deeds to reassure oneself of one’s own
comparative moral commitment and worth).
If people viewed the terrorist attacks as
a
moral breach that violated, for
example, their perception of what constitutes the tenets of just war (Walzer,
2000), the value-protection model (VPM) predicts that they would respond with
1998
LINDA
J.
SKITKA
both moral outrage and value affirmation. The first would be a more interper-
sonal and nationalistic response geared toward shoring up the moral perimeter
and guarding against future threat, whereas the latter would be a more intrapsy-
chic and patriotic response designed to reassure oneself and other in-group mem-
bers of one’s commitment to in-group ideals. According to this perspective, flag
displays may well be the result of increased expression of both nationalism and
patriotism, given that people tend to respond to moral threats with redundancy
and overkill, rather than using one or another strategy (Tetlock et al..
2000).
In sum, drawing on previous research and theory, we derived three hypothe-
ses for why there was such a dramatic display of the American flag following the
September
11,
200
1,
terrorist attacks. The patriotism hypothesis posits that flag-
display behavior in this context was a reflection of love of nation and a need to
express and defend core American values, and was not a consequence of a need
to rally around authority and to defend against out-groups. If the patriotism
hypothesis is true, then flag-display behavior should relate more strongly to mea-
sures of in-group enhancement and the degree that people engaged in other
efforts to support cultural standards of value as a consequence of the attacks (e.g.,
donated blood, gave money to charity) than to measures of out-group derogation
or antipathy.
The nationalism hypothesis, in contrast, posits that flag-display behavior in
the context of
911
1
was a symbolic expression of threat-induced needs to uncriti-
cally support American leaders and to defend against hostile out-groups.
If
the
nationalism hypothesis is true, then we would expect measures of perceived
threat, out-group derogation, moral outrage, and uncritical support for leaders to
emerge as stronger predictors of flag-display behavior than measures of in-group
enhancement or the tendency to engage in value-affirming behaviors.
Finally, the overkill hypothesis predicts that because the threat posed by the
terrorist attacks was
so
severe, people may have felt compelled to respond with
seemingly redundant reactions, and therefore that flag-display behavior was a
result of increased nationalism and patriotism. If the overkill hypothesis is true,
then we would expect to see measures of threat, out-group derogation and hostil-
ity, increased support for authority, in-group enhancement, and value-affirming
behaviors to each explain unique variance in the tendency to display the
American flag.
Method
Participants
The study sample was drawn from a panel of respondents maintained by
Knowledge Networks
(KN).
KN
recruits panel members using random-digit-
dialing telephone-selection methods, and therefore maintains a true probability
PATRIOTISM
OR
NATIONALISM?
1999
sample. As expected, given the random selection of participants, the characteris-
tics of the panel therefore closely match those of the U.S. Census.3 Once a panel
member agrees to participate, they are given a free interactive device to access
the World Wide Web (e.g., a Web TV), and free Internet access in exchange for
participation in regular surveys. About 50% of the panelists had no prior access
to the Web before becoming
KN
members,
so
the
KN
panel is the only Web-
enabled household panel that is truly representative of the American public.
As part of a larger data-collection effort (Skitka, Bauman,
&
Mullen, 2004), a
random sample
of
605
panel members (an
88%
within-panel cooperation rate)
responded between December 28,2001, and January 14,2002, to a password-
protected e-mail that alerted them to the survey. The e-mail had a clickable link
that allowed them to initiate the survey. Participants could access the survey only
once, and the survey was protected from nonpanel member access. There were no
significant differences in demographic profiles between those who did versus did
not respond to the invitation to participate. A summary of sample characteristics
is provided in Table
1.
The survey assessed people’s degree of nationalism, perceived threat of
future terrorist attacks, and retrospective reports of behavioral reactions to the
terrorist attacks (including whether people displayed the American flag). Finally,
the survey assessed the degree that they perceived that their feelings about in-
groups and out-groups were more positive, remained the same, or more negative
since September
1
1,2001.
Measures
Nationalism and support
for
authority.
Theorists have organized a variety of
overlapping personality and attitudinal variables into ideological/affective/
cognitive stylistic resonances (Alker
&
Poppen, 1973; Carroll, Perkowitz,
Lurigio,
&
Weaver, 1987). For example, one resonance,
cognitive
conservatism,
combines support for traditional power structures and opposition to egalitarian-
ism with personality measures of dogmatism, authoritarianism, and intolerance
of
ambiguity (a resonance reminiscent
of
the classic work on authoritarianism;
Adorno et al.,
1950).
Focusing on operational definitions, one could argue on
both conceptual and psychometric grounds that an item such as “We need strong
national leaders” could as easily be part
of
an ideology, authoritarianism, or atti-
tude scale.
Therefore, we used four items from Altemeyer’s (1996) right-wing authori-
tarianism (RWA) scale as an attitudinal (rather than a personality) measure of
nationalism and support for authority. People were asked to indicate on 5-point
3See
http://www.knowledgenetworks.com/ganp/
for comparisons of the panel with current
US.
Census figures.
2000 LINDA J.
SKITKA
Table
1
Unweighted Demographic Characteristics
of
the Sample
Variable Percentage
Gender
Male 48
Female 52
18-29 16
30-44 30
45-59 33
60
or
older 20
Less than high school 14
Some college 26
Bachelor’s degree or higher
25
Age (in years)
Highest level
of
education
High school 35
Household annual income
Less than $14,999
9
$15,000-$29,999
15
$30,000-$49,999 33
$75,000-$99,999
10
$100,000-$124,999
5
$125,000
or
more 4
White
79
Black 8
Hispanic 9
Other
4
Urban 88
Rural 12
Northeast 20
Midwest 23
South 36
West
18
Note.
N
=
605.
Urban
areas were classified
as
those locations that had a census Metro-
politan Statistical Area (MSA) code. For brevity’s sake, ranges are reported, but some
variables were measured at more fine-grained levels (e.g., age, income). Because the
study sample was a true probability sample, departures
of
sample characteristics from
what would be expected based
on
current census estimates could be corrected by apply-
ing
sample
weights.
Ail
other descriptive statistics and substantive analyses, therefore,
are based on weighted data. Hypothesis testing with and without weights yielded the
same results, however.
$50,000-$74,999 24
Race/ethnicity
Context
Region
PATRIOTISM
OR
NATIONALISM?
2001
radio-button scales4
(strongly agree, agree, neutral, disagree,
and
strongly dis-
agree):
“Our country desperately needs a mighty leader who will do what has to
be done to destroy the radical new ways and sinfulness that are ruining us”; “Our
country will be great
if
we honor the way of our forefathers, do what authorities
tell us, and get rid of the ‘rotten apples’ who are ruining everything”; “Our coun-
try will be destroyed someday if we do not smash the perversions eating away at
our moral fiber and traditional beliefs”; and “The way our country can get
through future crises is to get back to our traditional values, put tough leaders in
power, and silence troublemakers spreading bad ideas.” The items were reverse
scored
so
that higher scores on the measure reflect higher levels
of
nationalism.
The measure had a Cronbach’s alpha of
.89
within our sample.
Perceivedpersonal threat.
Perceived personal threat was measured with eight
items that tapped the degree of worry people felt about future terrorist attacks,
flying in commercial aircraft, getting infected with anthrax, other kinds of bioter-
rorism, retaliation for the war in Afghanistan, the personal safety of themselves
and their family, being in tall buildings, and large public gatherings on 5-point
radio button scales that ranged from
not at all
to
very much
(a
=
.92).
Group differentiation: In-group enhancement and out-group derogation.
Par-
ticipants were asked how much their feelings about a number of groups had
changed since September
11,
2001,
on 5-point radio-button scales
(much more
negative, more negative, stayed
the
same,
more positive,
and
much more posi-
tive).
A
principal-components analysis of the group-differentiation items with an
oblique rotation revealed a two-component solution, Feelings about in-group tar-
gets-Americans as a whole, American political leaders, firefighters and
police-loaded on one component (eigenvalue
=
2.42);
whereas feelings about
out-group targets-new immigrants, Arab American citizens, Palestinians, and
those who live in Islamic or Middle Eastern countries-loaded on another com-
ponent (eigenvalue
=
3.18).
Given that (a) the expected in-group and out-group
components emerged; (b) the components were uncorrelated
(r
=
-.04,
ns),
despite using a data-reduction technique that allowed correlated components; and
(c) attitudes toward in-groups and out-groups were on average more positive and
negative, respectively, there was a sound foundation for separate measures of in-
group enhancement and out-group derogation or distancing
(a
=
.77
and
.86,
respectively). The finding that in-group and out-group reactions factored out sep-
arately and were subsequently relatively uncorrelated was interesting in its own
4Radio-button scales present scale points as “bubbles”
or
“buttons.” When a participant selects a
given response option with a
mouse
click, the button becomes colored in, and the response is
recorded in a computer file. These scales had verbal point labels (e.g.,
strongly
agree)
but no numeric
labels. The computer recorded all responses numerically (using
the
values
1-5
for
5-point scales, and
1-7
for
7-point scales), with higher scale values reflecting greater degree
of
the variable measured
(e.g., higher values on the threat scale reflected greater threat).
2002 LINDA
J.
SKITKA
right. In-group and out-group differentiation were clearly distinguishable reac-
tions in how people reacted after
911
1.
Patriotism.
A single-item measure was included to tap the degree that people
felt an increased level of patriotism after the terrorist attacks. Specifically, partic-
ipants were asked to indicate
“To
what extent did you feel a surge of patriotism
following the attacks?” on a 5-point radio-button scale
(not at all, slightly, mod-
erately, much,
and
very much).
Behavioral checklists: Measures of moral outrage, value affirmation, and
flag-display behavior.
Participants were presented with a behavioral checklist of
different things they may have done in reaction to the September 11,200
1,
terror-
ist attacks. This list included three categories of behavior: (a) behaviors that
reflected moral outrage (e.g., said something like “We should just nuke them,”
engaged
in
some behavior in an attempt to blow off steam); (b) behaviors that
reflected non-flag-related forms of value affirmation (e.g., donating blood,
increased attempts to do nice things for family and friends, donating money to
charity); and (c) the separate category
of
flag-display behavior (displaying the
flag at one’s home; displaying the flag on one’s car, or wearing clothing
or
jewelry that depicts the American flag). Counts of how many of each
of
these
categories of behavior were checked off, therefore, served as our behavioral
index of moral outrage (with a range from
0
to
4)
and value affirmation (with a
range from 0 to
7).
Two measures of flag-display behavior were used: whether people displayed
the flag, and how many different ways people displayed the flag (on their homes,
their cars, their person, or some combination thereof). This measure was scored
on a 4-point scale ranging from
0
to
3.
Profile information.
In addition to the measures included on our survey, we
also had profile information about each respondent’s age, gender, education,
income, region of the United States, and political orientation.
Results
National-Level Analyses
Of our sample, 74% indicated that they had engaged in at least one form of
flag-display behavior as a result of the September
1
1,
200
1,
terrorist attacks, a
finding consistent with the results of other national surveys (Moore,
2003).
Moreover, 25%
of
our sample displayed the flag on their home, their car, and
their person after the attacks; similar percentages displayed the flag in only one
or
two
of these contexts. There was not a significant difference in whether or how
much Americans flew the flag as a function of region of the United States (i.e.,
whether the respondent lived in states in the South, Northeast, East, or Midwest),
F(
1,
60
1
)
=
2.19,
ns,
and
F(
1,
601)
=
1
SO,
ns,
respectively. African Americans
Table
2
Means, Standard Deviations, and Intercorrelations
of
Background Variables, Predictors, and Post-September
11,
2001,
Flag-Display Behavior
MSD2 3 4 5 6 7
8
9 1011 121314
1.
Extentofflagdisplay 1.47 1.12 .78**
.08
.lo*
-.lo*
.15**
.20** .11** .36** .12** .27** .40** .52**
.I]**
2. Flag display 1.26 0.44
-
.08 .14** -.lo* .14** .13** .05 .30** .03 .24** .35** .47** .05
3. Gender 1.52
0.50
-
.05
.01
-.09*
.14** .04
.08*
-.05
-.04 .16** .lo*
-.03
4. Age 45.67 16.29
-
-.04 .11**
.OO
.08
.lo* .02 .01
.08
.19** .12**
5.
Education 4.01 1.63
-
.34** -.21**
-.28**
-.lo* -.14**
-.09*
.01
-.03
-.O6
6. Household income 10.15 3.60
__
-.09* -.12** -.09*
.05
.03 .14** .12**
.08
7. Threat 2.56 0.96
-
.22** .15** .14** .17** .22** .16** -.12**
8.
Nationalism 2.63 1.08
-
.27** .13**
.19**
.03 .18** .20**
enhancement 4.78 1.26
-
.03 .19** .34** .40** .06
10. Out-group derogation 3.69 1.06 .25** .07 .19**
.OO
9. In-group
-0
.27** .35**
.05
-
1
1.
Moral outrage 1.25 1.27
-
D
-i
-
;
14. Political orientation 4.13 1.60
-
2
12. Non-flag-related
2
9
5
8
value-affirming
-
2
behavioral reactions 1.84 1.54 .36**
-.05
13. Patriotism 3.83 1.16
-
.15**
D
Note.
N=
605.
Extent of flag display ranged from
0 (no
display
of
the American Flag)
to
3
(the person displayed the American Flag
on
his or herperson.
reflecting some college
(but
no
degree). Annual household income was measured with
17
intervals; a mean of
10
equates to
$35,000
to
$39,000.
Moral out-
rage behaviors, range
=
0
to
4.
Nan-flag-related value-affirming behaviors, range
=
0
to
7.
All
other measures were rated on 7-point scales (ranging from
1
to
7).
Higher values
of
political orientation reflect greater conservatism.
*p
<
.05. **p
<
.01.
car: or home).
Flag display was coded
1
(Yes)
and
0
(No). Gender:
1
=
male,
2
=
female. Education was measured in nine categories, with a score
of
4
cn
9
0
0
2004
LINDA
J.
SKITKA
were less likely to display the flag
(M
=
0.46,
SD
=
0.50)
and less likely to dis-
play the flag in as many places (M=
0.75,
SD
=
0.98)
than were other ethnic
groups (i.e., non-Hispanic Whites, Hispanic, or other;
M=
0.81,
SD
=
0.39,
and
M
=
1.62,
SD
=
1.02,
respectively), an effect that was not significant when all
other demographic and predictor variables (e.g., income, education) were
included as covariates.
As can be seen in Table
2,
people who displayed the American flag
were somewhat older and less educated, had higher incomes, and were more
threatened by the terrorist attacks than were people who did not display the flag.
More strongly predictive of displaying the American flag, however, was the
degree that people engaged in in-group enhancement, expressed moral outrage
about the attacks, engaged in non-flag-related value-affirming behaviors like
donating blood, and expressed a higher degree of patriotism as a result of the
attacks. Measures of nationalism and out-group derogation were uncorrelated
with whether people displayed the American flag.
Although displaying the American flag was, not surprisingly, strongly corre-
lated with how many places people displayed the American flag, there were some
differences nonetheless in the pattern
of
results across these two measures. Age,
education, and income were weakly predictive
of
how many places people dis-
played the flag. Threat, nationalism, and out-group derogation each emerged as
stronger predictors of breadth of flag display than whether people displayed the
flag at all. That said, the strongest predictors
of
extent of flag display were vari-
ables more closely aligned with patriotism than nationalism; specifically, in-
group enhancement, engaging in non-flag-related value-affirming behaviors,
higher levels of patriotism, but also moral outrage.
To better understand the forces that best explained people’s tendency to dis-
play the
flag
following September
I
I,
200
1,
we explored the degree that our
predictors explained unique variance in flag display and extent of flag display by
using standardized regression.
As
can be seen in Table
3,
the results were
more consistent with a patriotic than a nationalistic or overkill explanation
for
post-9il1 flag displays. Lower levels of education and higher incomes explained
unique variance in both whether people displayed the flag and the number
of
ways they displayed it. Threat emerged as a weak but significant predictor
of
breadth of flag displays, but not whether people displayed the flag. Nationalism
was a weak predictor of whether people displayed the flag, but did not predict
how broadly they did
so.
The strongest predictors
of
both whether and how many
ways people displayed the flag were in-group enhancement, engaging in non-
flag-related value-affirming behaviors, and self-reported patriotism. As each
of
these increased, so too did whether they displayed the American flag and how
widely they did
so.
Including interaction terms of threat with out-group deroga-
tion, moral outrage, nationalism, in-group enhancement, non-flag-related value-
affirming behaviors, and patriotism in the regression equations did not yield any
PATRIOTISM
OR
NATIONALISM?
2005
Table 3
Standardized Regression Weights Associated With Background Variables,
Predictors, and Post-September
lI,
2001,
Flag Display Behavior in the Full
National Sample
Extent of flag
Predictor display Flag display
Gender
Age
Education
Household income
Threat
Nationalism
In-group enhancement
Out-group derogation
Moral outrage
Non-flag-related value-affirming behaviors
Self-reported patriotism
Political orientation
R
.04
-.01
-.12**
.13**
.ox*
.05
.15**
.oo
.05
.18**
.34**
.06
.60
.02
.04
-.14**
.14**
.02
.08*
.11**
.ox*
.ox*
.15**
.33**
.04
.54
Note.
N
=
605.
Extent
of
flag display ranged from
0
(no
display
of
the American Flag)
to
3
(the person displayed the American Flag
on
his
or
her person,
car,
or
home):
Flag
display was coded
1
(Yes)
and 0
(No).
*p
<
.05.
**p
<
.01.
significant effects, a result indicating that the degree that people felt threatened
by the terrorist attacks did not moderate the effects of the variables that predicted
whether and how widely people displayed the American flag.
Regional Analyses
Although there were not significant differences in extent of flag display or
whether people flew the flag across different regions
of
the United States, one
still might ask whether the reasons people flew the flag varied as a function of
region.
To
explore this question, we re-ran our regression analyses separately by
region (see Table 4 for more detail; regional categories are those defined by the
U.S.
Census as states
in
the Northeast, Midwest, South, and West). These anal-
yses revealed that although there were some differences in predictors of flag
Table
4
h)
8
cf,
r_
z
0
D
c
Standardized Regression Weights Associated With Background Variables, Predictors, and Post-September
11,
2001,
Flag-Display Behavior Broken Down by Region
of
the United States
Northeast Midwest South West
Extent of Flag Extent
of
Flag Extent
of
Flag Extent
of
Flag
(N
=
123)
(N=
134) (N=216)
(N=
129)
v)
5
Variable display display display display display display display display
D
x
Gender
.06
-.01
-.
17 -.06
.08
.10
.08 .03
Education
-.I6
-.19*
-.22** -.I2 -.01
-.06
-.15
-.I0
Household income
.08
.10 .I3 .02 .04
.12
.16*
.14
Threat
-.04 .02 .06 -.12 .20** .13* -.03
.03
Age
.II
.I4
.05 .03 -.01
.08
-.17* -. 12
Nationalism
.02 .01 .06
.09
.04 .07 -.04 -.04
In-group enhancement
.16
.12 .17* .18* .06 .07 -.04 -.09
Out-group derogation
.ll .14
-.05
.14 .01 .04 -.02
.09
Moral outrage
.10 -.02 .04
.09
.o
1
.02 .09 .04
Non-flag-related value-affirming
behaviors
.19* .I2 .17*
.08
.21** .14* .19*
.21**
Self-reported patriotism
.23** .30** .I4 .25** .41** .35** .39** .37**
Political orientation
-.01
.oo
.18* .07
.o
1
-.02
.03
-.01
R
.51** .49** .51** .45** .61** .55** .64** .59**
~
Note.=ent of flag display ranged
from
0
(no
display
of the
American
Flag)
to
3
(the person
displayed
the
American Flag
on
his
or
her
person,
car,
or
home).
Flag
display was coded
1
(Yes)
and
0
(No).
The distribution
of
participants across different regions
of
the United
States is consistent with what would be expected given
census
statistics.
*p
<
.05.
**p
<
.01.
PATRIOTISM OR NATIONALISM?
2007
display as a function of region, a regular pattern emerged nonetheless. Variables
consistent with the patriotism hypothesis consistently predicted flag-display
behavior across all regions, whereas variables consistent with the nationalism
hypothesis did not.
For example, extent of flag display was most strongly predicted by non-flag-
related value-affirming behaviors and self-reported patriotism for people in the
northeastern United States. Turning to whether people in the Northeast displayed
the flag, we found that only self-reported patriotism (the strongest predictor) and
education explained unique variance in whether people flew the flag. Education
was negatively related, whereas self-reported patriotism was positively related to
whether people flew the flag.
Education was also a negative indicator of the extensiveness of flag display in
the Midwest. Breadth of flag display was higher among those who responded to
9/11 with stronger feelings of in-group enhancement and who engaged in higher
levels of non-flag-related value-affirming behaviors in response to 9/11, Two
findings were unique to the Midwest: (a) self-reported patriotism was not reliably
associated with breadth of flag display; but (b) political orientation was.
As
polit-
ical conservativism increased,
so
too did breadth of flag display. That said, polit-
ical orientation was not correlated with whether people flew the flag in the
Midwest. Instead, whether people in this region displayed the flag was predicted
by the degree that they responded to 9/11 with increased in-group enhancement
and self-reported patriotism.
Unique
to
the South was a reliable association between perceived threat with
both whether and how extensively people displayed the American flag. That said,
nationalistic variables did not predict flag-display behavior in the South; instead,
patriotic variables did. Specifically, besides threat, only the degree that people in
the South responded to 9/
1
1
by engaging in non-flag-related value-affirming
behaviors and with increased self-reported patriotism predicted whether and how
extensively they displayed the flag.
Finally, the extent that and whether people flew the flag in the West was most
strongly predicted by non-flag-related value-affirming behaviors and self-
reported patriotism. Extent of flag display was also associated with lower levels
of education and higher levels of income for this portion of the sample.
In sum, the constellation of variables that predicted both flag display and
extent of flag display supported the patriotism hypothesis more strongly than the
nationalism or the overkill hypothesis at both national and regional levels. More-
over, even though there were some differences in what led to flag-display behav-
ior across different regions of the United States, the most consistent predictors
were variables that reflected positive in-group regard, not out-group antipathy.
Other results also supported the notion that people’s symbolic expression of
patriotism through display of the American flag was independent
of
nationalism.
Specifically, nationalism and patriotism were only weakly correlated
(r
=
.18,
2008
LINDA
J.
SKITKA
p
<
.Ol), and in-group enhancement as a reaction to the terrorist attacks was
uncorrelated with out-group derogation
(r
=
.03,
nsj.
Taken together, these results
indicated that feeling good about one’s own group does not necessitate feeling
bad about other groups, even in the context of people’s reactions to terrorist
attacks.
Discussion
There were some good reasons to be concerned that 9/11 might create a
nationalistic upsurge. The passage of the Patriot Act, thought by some to pose
serious challenges to civil liberties, in addition to widespread willingness on the
part of Americans to sacrifice some civil liberties in an effort to fight terrorism
(Huddy, Khatid,
&
Capelos,
2002)
together suggested that security concerns
might lead people toward a form of nationalism that could undercut many of the
cornerstones of a functioning liberal democracy.
Even though Americans reported high levels
of
moral outrage and perceived
threat after the 911
1
terrorist attacks, our results indicated that these reactions
were associated more strongly with increased in-group consideration and
enhancement than with rampant ethnocentrism or out-group hostility. Moreover,
our results indicated that Americans’ post-9/11 display of the American flag was
a phenomenon that was connected more closely to the same impulses that led
people to donate blood and to give millions to charity, rather than to a national-
istic desire to rally around in-group political authorities and institutions, or to
express out-group derogation or hostility. In short, our results supported the prop-
osition that patriotism and in-group regard do not necessarily go hand in hand
with nationalism and out-group derogation, even under conditions of threat, such
as a terrorist attack.
The conclusion that displays of the American flag in the immediate aftermath
of 9/11 reflected patriotism and a desire to show solidarity with fellow citizens,
rather than a desire to express out-group hostility, does not mean that displays of
the flag are always or even often expressions
of
in-group enhancement without
out-group implications. The flag and other symbols
of
group identity can clearly
shift in meaning as a function of the particular social context in which they are
used. National polls, for example, have indicated that the number of people who
still display the flag since the Iraq War began
(56%)
was significantly lower than
the
74%
to
82%
of those who displayed the flag in the months immediately fol-
lowing 9/11 (Moore,
2003
j.
One can speculate that what it means to display the
flag since the Iraq War began may have shifted more toward the nationalistic end
of the spectrum, a sentiment that fewer Americans may be prepared to endorse
unequivocally. In short, displaying the flag appears to have considerable meaning
to those who engage in it. However, what this meaning
is
in any given context
seems likely to vary.
PATRIOTISM
OR
NATIONALISM?
2009
Of course, this study, like all field studies, is open to alternative explanations
and potential criticism. For example, perhaps we would have found stronger
effects for nationalism if we had used measures besides items from Altemeyer’s
(1 996) RWA scale or our out-group derogation measure to represent different
aspects of this construct, Nonetheless, the variance observed with these measures
was sufficiently high to allow the detection of correlations with other variables,
so
the null findings for nationalistic variables cannot be said to be a result of
restriction of range. In addition, it would have been ideal to have a more in-depth
measure of patriotism. Nonetheless, the high correlation of our patriotism item
with value-affirming behaviors (e.g., donating blood, charitable giving) and with
in-group enhancement bolsters the conclusion that this item was tapping the
intended construct.
Although this research, like all research, is open to methodological criticism,
it also has some particular methodological strengths. For example, this study
tested the more general social identity theory hypothesis that in-group enhance-
ment need not go hand in hand with out-group derogation in a painfully real,
important, and high-impact context that involved people’s real and deep feelings
about their country under a condition of an unexpected threat. Although some
research has supported the independence of out-group derogation and in-group
enhancement in the comparatively sterile confines of the laboratory (for a review,
see Brewer, 1979), this research provided an important test of the generalizability
of these findings to a more high-impact context. Moreover, because hypotheses
were tested using a national probability sample
of
the American public, the find-
ing that flag-display behavior was rooted more in patriotism than in nationalism
is
one that we can be confident represented the thoughts, feelings, and behavior
of the mass public; moreover, one that generalized across different regions of the
United States.
In closing, the results
of
the study presented here provided insight into the
psychological factors that shaped one of the many reactions people had to the
September 11, 2001, terrorist attacks; specifically, why there was such a wide-
spread display of the American flag. Although other factors such as conformity
or concern about criticism from others for being nonpatriotic also may have
played roles in why people displayed the flag, our results indicated that a sense of
increased patriotism, positive in-group identity, and a desire
to
affirm cultural
standards of value were each more strongly related to flag-display behavior than
were feelings of nationalism or out-group derogation. These findings also pro-
vided real-world support for the notion that engaging in in-group enhancement
does not always mean that people will also derogate out-groups. Even under con-
ditions that inspired considerable moral outrage and serious concerns about
safety, Americans nonetheless flew the flag to symbolize their commitment and
connections to their fellow citizens, not to declare that the United States was
superior and dominant, or that out-groups should beware.
2010
LINDA
J.
SKITKA
References
Adorno, T.
W.,
Frenkel-Brunswik, E., Levinson, D.
J.,
&
Sanford,
R.
N. (1950).
The authoritarian personality.
New York, NY: Norton.
Alker,
H.
A.,
&
Poppen, P.
J.
(1973). Personality and ideology in university
students.
Journal
of
Personality,
41,652-671.
Altemeyer, B. (1 996).
The authoritarian specter.
Cambridge, MA: Harvard Uni-
versity Press.
AlterNet. (2001).
What do the stars andstripes mean
now?
Retrieved October 8,
2003, from
http://www.alternet.org/story.htm1?StoryID=ll559
Bar-Tal, D. (1997). The monopolization of patriotism. In D. Bar-Tal
&
E. Staub
(Eds.),
Patriotism in the lives
of
individuals and nations
(pp. 246-270).
Chicago, IL: Nelson-Hall.
Blank, T.,
&
Schmidt, P. (1993). Verkletze oder verletzende Nation? Empiricsche
Befunde zum Stolz auf Deustchland
[Hurt
or hurting nation? Empirical
results on pride in Germany].
Joumal.fiir SoziaEforschung,
33,391-41
5.
Blank,
T.,
&
Schmidt, P. (2003). National identity in
a
united Germany: National-
ism or patriotism? An empirical test with representative data.
Political Psy-
Brewer, M. B. (1979). In-group bias in the minimal intergroup situation: A
Brewer, M. B. (1999). The psychology
of
prejudice: In-group love or out-group
Brown, R.
(
1995).
Prejudice: Its socialpsychology.
Oxford,
UK:
Blackwell.
Carroll,
J.,
Perkowitz, W., Lurigio, A.,
&
Weaver, K. (1987). Sentencing goals,
causal attributions, and personality.
Journal
of
Personality and Social Psy-
Chow, C. (2001).
American flag banned
on
campuses across the nation.
Retrieved October 8, 2003, from
http://www.academia.org/campus-reports/
200
l/nov-200
1-1 .html
de Figueiredo, R.
J.
P., Jr.,
&
Elkins,
Z.
(2003). Are patriots bigots? An inquiry
into the vices of in-group pride.
American Journal
of
Political Science,
47,
Decatur,
S.
(1816, April). US. naval commander. Toast proposed at a banquet in
Norfolk, Virginia, to celebrate Decatur
’s
victory over Algerian “Barbary
pirates.” Retrieved October 10, 2005, from
http://www.bartleby.com/66/46/
15946.html
Hechter,
M.
(2000).
Containing nationalism.
Oxford,
UK:
Oxford University Press.
Horowitz, D. (2003).
9/11
and the “anti-war” left.
Retrieved October 8, 2003,
Huddy,
L.,
Khatid, M.,
&
Capelos, T. (2002). Trends: Reactions to the terrorism
chology,
24,289-3 12.
cognitive-motivational analysis.
Psychological Bulletin,
86,307-324.
hate?
Journal
of
Social Issues,
55,429-444.
chol~g~,
52,
107-118.
171-188.
from
http//www.gopusa.com/davidhorowitz/dh~O9
15p.shtml
attacks of September 11, 2001.
Public Opinion Quarterly,
66,418-450.
PATRIOTISM
OR
NATIONALISM?
201
1
Kelman,
H.
C. (1997). Nationalism, patriotism, and national identity: Social-
psychological dimensions. In D. Bar-Tal
&
E. Staub (Eds.), Patriotism in the
lives of individuals and nations (pp. 165-1 89). Chicago,
IL:
Nelson-Hall.
Kingsolver, B. (2001). And our$ag was still there (Sun Francisco Chronicle).
Retrieved October
8,
2003, from
http://www.organiconsumers.org/corp/
kingsolver09260
1
.cfm
Kosterman,
R.,
&
Feshbach,
S.
(1989). Toward a measure of patriotic and nation-
alistic attitudes. Political Psychology,
10,
247-273.
Meloen,
J.
D. (1999). The political culture of state authoritarianism. In
S.
Renshon
&
J.
Duckitt (Eds.), Political psychology: Cultural and cross-
cultural perspectives (pp. 108-127). London, UK: Macmillan.
Moore, D.
W.
(2003). War makes Americans confident, sad: Personal lives less
affected than during the first
Guy
War and 9/11 (Gallup News Service).
Retrieved October
6,
2003, from
http://www.gallup.com/poll/releases/
pr030326.asp?Version=p
Peiia,
Y.,
&
Sidanius,
J.
(2002).
U.S.
patriotism and ideologies of group domi-
nance:
A
tale
of
asymmetry. Journal
of
Social Psychology, 142, 782-790.
Roberts,
J.
T. (2002). WhoJlies theflag? Retrieved October
8,
2003, from http://
faculty.wm.edu/jtrobe/whofliestheflag .htm
Schatz, R. T.,
&
Staub,
E.
(1 997). Manifestations
of
blind and constructive patri-
otism: Personality correlates and individual-group relations. In D. Bar-Tal
&
E.
Staub (Eds.), Patriotism in the lives
of
individuals and nations (pp. 229-
245). Chicago, IL: Nelson-Hall.
Schatz,
R.
T.,
Staub, E.,
&
Lavine,
H.
(1999). On the varieties
of
national attach-
ment: Blind versus constructive patriotism. Political PsychoZogy, 20, 15 1-1 74.
Skitka, L.
J.,
Bauman, C.
W.,
&
Mullen, E. (2004). Political tolerance and com-
ing to psychological closure following September 11, 2001
:
An integrative
approach. Personality and Social Psychology Bulletin, 30, 743-756.
Stevenson, A. (1952, August 27).
U.S.
Democratic politician. Speech to
American Legion Convention, New York City, “The Nature of Patriotism.”
Retrieved October 10,2005, from
http://www.bartleby.com/66/59/56459.html
Sullivan,
J.
L., Fried,
A.,
&
Dietz, M. G. (1992). Patriotism, politics, and the
Presidential election
of
1988. American Journal
of
Political Science,
36,
Tetlock,
P.
E. (2002). Social functionalist frameworks for judgment and choice:
Intuitive politicians, theologians, and prosecutors. Psychological Review,
Tetlock, P.
E.,
Kirstel,
0.
V.,
Elson,
S.
B., Green,
M.
C.,
&
Lemer,
J.
S.
(2000).
The psychology
of
the unthinkable: Taboo trade-offs, forbidden base rates,
and heretical counterfactuals. Journal
of
Personality and Social Psychology,
Walzer,
M.
(2000). Just and unjust wars: A moral argument with historical illus-
200-234.
109,451-471.
78,
853-870.
trations (3rd ed.) New York, NY Basic Books.
