Social Psychological and Personality Science

http://spp.sagepub.com/

Living Dangerously: Culture of Honor, Risk-Taking, and the Nonrandomness of ''Accidental'' Deaths
Collin D. Barnes, Ryan P. Brown and Michael Tamborski
Social Psychological and Personality Science 2012 3: 100 originally published online 8 June 2011
DOI: 10.1177/1948550611410440

The online version of this article can be found at:

http://spp.sagepub.com/content/3/1/100

Published by:

http://www.sagepublications.com

On behalf of:
Society for Personality and Social Psychology

Association for Research in Personality

European Association of Social Psychology

Society of Experimental and Social Psychology

Additional services and information for Social Psychological and Personality Science can be found at:

Email Alerts: http://spp.sagepub.com/cgi/alerts

Subscriptions: http://spp.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

Citations: http://spp.sagepub.com/content/3/1/100.refs.html

>> Version of Record - Dec 2, 2011

OnlineFirst Version of Record - Jun 8, 2011

What is This?

Downloaded from spp.sagepub.com at Universidad del Norte on August 7, 2013

 Social Psychological and
Personality Science
Living Dangerously: Culture of Honor, 3(1) 100-107
ª The Author(s) 2012
Reprints and permission:
Risk-Taking, and the Nonrandomness of sagepub.com/journalsPermissions.nav
DOI: 10.1177/1948550611410440
‘‘Accidental’’ Deaths http://spps.sagepub.com

Collin D. Barnes1, Ryan P. Brown2, and Michael Tamborski2

Abstract
Two studies examined the hypothesis that the culture of honor would be associated with heightened risk taking, presumably
because risky behaviors provide social proof of strength and fearlessness. As hypothesized, Study 1 showed that honor states
in the United States exhibited higher rates of accidental deaths among Whites (but not non-Whites) than did nonhonor
states, particularly in nonmetropolitan areas. Elevated accidental deaths in honor states appeared for both men and women
and remained when the authors controlled for a host of statewide covariates (e.g., economic deprivation, cancer deaths,
temperature) and for non-White deaths. Study 2, likewise, showed that people who endorsed honor-related beliefs reported
greater risk taking tendencies, independent of age, sex, self-esteem, and the big five.

Keywords
culture of honor, manhood, masculinity, risk taking, accidental deaths

The circumstances that require men to prove their manhood are To recapitulate what has been more fully explained
numerous and varied, and this fact renders manhood a volatile elsewhere (e.g., Cash, 1941; Fischer, 1989; Nisbett, 1993;
social status that must be constantly upheld with demonstra- Nisbett & Cohen, 1996; Wyatt-Brown, 1982), economic fac-
tions of toughness and bravery (e.g., Bosson, Vandello, tors and pervasive lawlessness in the southern and western
Burnaford, Weaver, & Wasti, 2009; Vandello, Bosson, Cohen, United States produced an insecure social environment that
Burnaford, & Weaver, 2008; Weaver, Vandello, Bosson, & helped perpetuate the cultural ideology of honor of the Ulster
Burnaford, 2010; see also Archer, 1994). Acts of physical Scots (or ‘‘Scotch-Irish’’), who immigrated to the United States
aggression constitute one means by which men prove their in great numbers during the 18th century (Fischer, 1989). This
masculine status because such behaviors convey to the self ideology of honor emphasized the relentless, and sometimes
and others that a man is strong, fearless, and willing to act violent, defense of masculine reputation, which is presumably
despite risks to personal safety (e.g., Bosson et al., 2009). Dan- a social adaptation to an environment characterized by scarce
gerous behaviors, such as motorcycle riding without a helmet resources, frequent intergroup aggression (e.g., raiding), and
and mountain climbing without a partner, represent another the absence of the rule of law. Social institutions (Cohen &
class of masculinity-confirming behaviors because their per- Nisbett, 1997), gender identity schemas (Brown & Osterman,
formance, like overt acts of aggression, demonstrates one’s in press), and beliefs about society’s support for honor-
strength and fearlessness. restorative aggression (Vandello, Cohen, & Ransom, 2008)
The problem with engaging in excessive risk taking, of have been cited as a few of the mechanisms by which this pre-
course, is that it can sometimes be deadly. In this article, our occupation with the defense of honor persists over time, and
primary interest is in how deaths associated with accidental studies have demonstrated the tendency for people (White
causes (automobile accidents, falls, electrocutions, etc.) might males, in particular) from honor states to respond to reputation
be facilitated by the social dynamics underlying the culture threats with higher levels of hostility and violence compared to
of honor, a characteristic of societies that place special
emphasis on the aggressive defense of reputation (Cash,
1941; Fischer, 1989; Nisbett, 1993; Nisbett & Cohen, 1996; 1
Institute for U.S.–China Issues, The University of Oklahoma, Norman, OK,
Wyatt-Brown, 1982). We reasoned that insofar as these USA
2
‘‘accidental’’ deaths are more frequent in so-called honor Department of Psychology, The University of Oklahoma, Norman, OK, USA
states and not attributable to regional differences in other
Corresponding Author:
variables, they might reflect the influence of a cultural ideol- Collin D. Barnes, Institute for U.S.–China Issues, The University of Oklahoma,
ogy that places a premium on proving that one is strong and 729 Elm Avenue, Hester Hall, Room 118, Norman, OK 73019, USA
fearless. Email: [email protected]

Downloaded from spp.sagepub.com at Universidad del Norte on August 7, 2013

Barnes et al. 101

people from nonhonor states (e.g., Brown, Osterman, & Study 1

Barnes, 2009; Cohen & Nisbett, 1994; Cohen, Nisbett, Bowdle,
In Study 1, we examined deaths classified as ‘‘accidental’’ by
& Schwarz, 1996; Nisbett, Polly, & Lang, 1995).
the U.S. Centers for Disease Control and Prevention (CDC)
Although prior culture-of-honor research has predomi-
as a function of the honor or nonhonor classification of each
nantly focused on retaliatory aggression (e.g., Brown
U.S. state. Although our primary analyses focus on statewide
et al., 2009; Cohen et al., 1996), and although such aggres-
accidental death rates across genders as a function of culture-
sion might contribute to some of the accidental deaths we
of-honor status, we also examine death rates separately for
examine below (e.g., instances of retaliatory road rage), it
males and females. In addition, because theory and prior
could be that a preoccupation with proving personal honor
research suggests that the honor classification of a state should
compels men to act riskily in other contexts as well,
only be relevant for outcomes among Whites, we hypothesized
whether these actions are preceded by a challenge—e.g.,
that honor states would exhibit increased accidental deaths
‘‘What’s wrong McFly? Chicken?’’ (Gale, Canton, &
among Whites, but not among non-Whites.1
Zemeckis, 1989)—or not. Exposing oneself to potentially
deadly situations provides social proof that one is strong and
fearless, and because this proof is such a salient concern Method
(especially for men) in cultures of honor, people living in
such cultures could suffer accidental fatalities at higher rates
Data
than people living elsewhere. Consistent with this hypoth- Our analyses relied on cause-of-death data compiled by the
esis, Roebuck and Murty (1996) found that sociocultural CDC, which were coded according to the most recent revision
scripts concerning the defense of masculine reputation of the International Classification of Diseases (i.e., ICD-10
played a key role in the recidivism of White male drunk codes), and which include the years 1999 through 2006. Only
drivers. Other studies have considered regional differences deaths associated with unintended, external causes (i.e.,
in safety and health-related risk taking, but these studies ‘‘Transport accidents’’ [code V01-V99] and ‘‘Other external
focused on the role of religious based fatalistic beliefs and causes of accidental injury’’ [code W00-X59]) were exam-
have reported inconsistent results (Cohen & Nisbett, 1998; ined.2 Transport accidents include those involving pedestrians,
Sims & Baumann, 1972). motorcyclists, automobiles, water vehicles, and others. Non-
Finally, we should note that although our discussion of transport accidents include accidental deaths associated with
cultural norms associated with risk taking has been focused falls, exposure to inanimate and animate forces, drowning,
on issues of greatest relevance to the defense of masculine burning, overexertion, and the like. A full list of ICD-10 codes
identity, these norms have implications for both men and is available from the CDC’s website. The death rates we report
women. First, some research using non-American samples were calculated per 100,000 persons and adjusted for age
suggests that conformity to the masculine stereotype predicts based on the U.S. standard population in the year 2000.
risk taking independent of biological sex (e.g., Granié, 2009; We used Cohen’s (1998) approach to distinguishing honor
Özkan & Lajunen, 2006; Raithel, 2003). Insofar as confor- states from nonhonor states. The states located in the southern
mity to stereotypes can occur among those to whom they and western census regions of the country (i.e., South Atlantic,
do not directly apply (e.g., Bargh, Chen, & Burrows, 1996; East South Central, West South Central, Mountain, and
see also Bem, 1975), and to the extent that environmental Pacific), except for Alaska and Hawaii, were categorized as
primes and general social pressures to exhibit ‘‘manly’’ beha- honor states; states located in the northern part of the country
viors (e.g., strength and fearlessness) are more pervasive (i.e., New England, Middle Atlantic, East North Central, and
within honor states than within nonhonor states, it might not West North Central), plus Alaska and Hawaii, were identified
be surprising for the predicted regional pattern in accidental as nonhonor states. To measure honor status in a more contin-
deaths to emerge among women as well as men. Second, uous fashion, some researchers have used Gastil’s (1971)
because the outcomes we are examining in Study 1 are death southern cultural influence index (e.g., Nisbett & Cohen,
rates that do not discriminate between people who suffered 1996). Although Gastil’s index has been criticized (e.g., Loftin
accidental fatalities because of their own risk taking or & Hill, 1974), it is, at present, the only continuous indicator of
because of the risk taking of others, we cannot be sure who the culture of honor available for U.S. states. By using Cohen’s
the precise cause of these deaths is. Thus, women in honor (1998) dichotomous classification and Gastil’s continuous
states might exhibit elevated accidental deaths because they index, we hoped to provide converging evidence for our
are (wittingly or unwittingly) adhering to behavioral scripts predictions.
associated with ‘‘manly’’ virtues, or because the risky beha- Prior research has shown that honor states differ from non-
viors of men cause women to die as victims at higher rates honor states on a number of variables (e.g., Brown et al., 2009;
in honor states. The women-as-risk-takers and women-as- Cohen, 1998), each of which could serve as an alternative
victims interpretations cannot be conclusively tested with the explanation for any regional differences we might observe.
kinds of data available to us at the statewide level, which is We attempted to eliminate these alternative explanations by
our focus in Study 1. Because of this limitation, we address taking into account a host of state-level controls, including the
this issue further at the individual level in Study 2. proportion of individuals living in rural areas (U.S. Census

Downloaded from spp.sagepub.com at Universidad del Norte on August 7, 2013

102 Social Psychological and Personality Science 3(1)

Table 1. Intercorrelations Among and Descriptive Statistics for Study 1 Variables

1 2 3 4 5 6 7 8 9 10 11 12 13

1. CHC –
2. SCI .69** –
3. Rurality .01 .09 –
4. Temperature .47** .74** .21 –
5. Economic .47** .63** .31* .39** –
6. Cancer .01 .26y .35* .10 .32** –
7. HPSA .37** .33* .20 .16 .55** .07 –
8. Speed .41** .24y .13 .06 .28y .24y .61** –
9. Auto-travel .17 .06 .56** .06 .11 .00 .41** .33* –
10. Police .17 .24y .48** .33* .11 .08 .31* .23 .20 –
11. Accidents .58** .68** .51** .34* .69** .20 .64** .42** .43** .02 –
12. Transport .59** .66** .55** .34** .68** .20 .67** .51** .54** .06 .92** –
13. Nontransport .43** .55** .32* .26y .53** .15 .44** .21 .19 .04 .86** .60** –
M – 17.80 0.28 52.24 0.00 196.74 0.14 65.06 0.00 32.38 39.66 18.13 21.53
SD – 9.04 0.15 8.15 .79 13.73 0.08 4.50 0.84 6.54 8.88 5.62 4.27
Correlations are based on age-adjusted state-level data for Whites.
For CHC ¼ culture-of-honor classification (0 ¼ nonhonor state; 1 ¼ honor state); SCI ¼ Gastil’s (1971) southern cultural influence index; Rurality ¼ proportion of
population living in rural areas; Temperature ¼ mean annual temperature; Economic ¼ economic deprivation; Cancer ¼ number of cancer-related deaths per
100,000 persons; HPSA ¼ proportion of population living in a health professional shortage area; Speed ¼ Average speed limit across road types; Auto-Travel ¼
composite auto-travel variable; Police ¼ number of police officers per capita; Accidents ¼ total fatal accidents per 100,000 persons; Transport ¼ transportation
accidents per 100,000; NonTransport ¼ nontransportation accidents per 100,000.
y
< .10. * p < .05. ** p < .01.

Bureau, 2000), as rurality might be associated with treacherous Results and Discussion
roadways and diminished access to medical care; mean annual
We tested the association between the culture of honor and
temperatures (National Oceanic and Atmospheric Administra-
accidental deaths using two indicators: (a) Cohen’s (1998)
tion, 2000), as hotter temperatures could be linked with aggres-
dichotomous, culture-of-honor classification (CHC) and
sive driving; economic deprivation (a composite of poverty
rates [National Center for Education Statistics for 2000 and (b) Gastil’s (1971) continuous, state-level measure of southern
average of 2004–2006], unemployment rates [U.S. Bureau of cultural influence (SCI), along with the covariates described
Labor Statistics, 2000, 2004], median household incomes above. Intercorrelations among and descriptive statistics for all
[U.S. Census Bureau, 2000, 2004], and proportion of individu- study variables appear in Table 1. For the sake of brevity, only
als with at least high school degrees [National Center for results related to our focal predictors (i.e., CHC and SCI) are
Education Statistics, 2000, 2004]; a ¼ .88), as economically presented in the text; statistics for all study variables are
deprived states might have fewer resources for promoting their displayed in Tables 2 and 3.
residents’ safety; age-adjusted cancer death rates for 1999 Accidental death rates for Whites varied widely across
through 2006 (obtained from the CDC’s Compressed Mortality states (from 22.80 per 100,000 to 60.20 per 100,000); thus,
File; Center for Disease Control and Prevention, 1999–2006), there was substantial interstate variability to be explained.
as higher accidental deaths might simply reflect higher mortal- We found that CHC was a significant predictor of these deaths,
ity rates that have nothing to do with cultural ideology of such that Whites in honor states (M ¼ 42.03) experienced more
honor;3 and the proportion of each state’s population living accidental deaths than Whites in nonhonor states (M ¼ 36.89),
in a primary health care professional shortage area (HPSA; b ¼ .29, t ¼ 3.17, p ¼ .003, d ¼ 0.97, and this pattern was
current data obtained from the U.S. Department of Health and corroborated by the SCI index, b ¼ .54, t ¼ 4.35, p < .001,
Human Services, Health Resources and Services Administra- d ¼ 1.33. Furthermore, this effect obtained among men and
tion, 2011), as accidental deaths might be more prevalent in women, whether the culture of honor was operationalized using
areas with fewer healthcare resources. Finally, for analyses of the CHC or SCI (ps  .008). In contrast, and consistent with
transport-related accidents, three additional covariates were prior research, neither CHC, b ¼ .06, t ¼ 0.53, ns, nor SCI,
included: (a) average state speed limits (current data obtained b ¼ .03, t ¼ 0.17, ns, significantly predicted accidental
from the Insurance Institute for Highway Safety, 2011), (b) a deaths among non-Whites, even when non-White males and
composite auto-travel variable (obtained from the U.S. Depart- females were examined separately.4
ment of Transportation’s Federal Highway Administration; an The nature of these data precludes our making firm causal
average of the number of vehicles owned [2006] and miles conclusions about the role of the culture of honor in accidental
traveled per capita [2005]), and (c) number of police officers deaths, as we cannot hope to account for all possible confound-
per capita (data for 1996 based on U.S. Census Bureau esti- ing variables. However, to the extent that any statewide con-
mates obtained from www.Allcountries.org). founds would affect persons of all demographic groups

Downloaded from spp.sagepub.com at Universidad del Norte on August 7, 2013

Barnes et al. 103

Table 2. Standardized Regression Results for Accidental Deaths Using Cohen’s (1998) Culture-of-Honor Classification (CHC)

Accidents Rurality Temperature Economic Cancer HPSA Speed Auto-Travel Police Culture of Honor

Total (White) .42** .18y .17 .00 .32** – – – .29**

Male .42** .20* .17 .03 .34** – – – .25**
Female .43** .09 .18 .00 .27* – – – .36**
Total (Non-White) .20y .26* .11 .11 .72** – – – .06
Male .20y .19 .10 .13 .73** – – – .04
Female .21y .36** .13 .08 .67** – – – .11
Transport
Total (White) .43** .18* .18y .02 .23* .08 .15 .09 .26**
Male .44** .18* .18y .02 .25* .04 .15y .10 .26**
Female .40** .16y .20y .05 .15 .20* .14 .05 .24**
Nontransport
Total (White) .24 .10 .22 .00 .18 – – – .21
Male .23 .14 .23 .03 .21 – – – .14
Female .25y .02 .21 .01 .11 – – – .31*
Rurality ¼ proportion of population living in rural areas; Temperature ¼ mean annual temperature; Economic ¼ economic deprivation; Cancer ¼ number
of cancer-related deaths per 100,000 persons; HPSA ¼ proportion of population living in a health professional shortage area; Speed ¼ Average speed limit
across road types; Auto-Travel ¼ composite auto-travel variable; Police ¼ number of police officers per capita; CHC ¼ culture-of-honor classification
(0 ¼ nonhonor state; 1 ¼ honor state).
y
.10 > p  .05; *p < .05; **p < .01.

Table 3. Standardized Regression Results for Accidental Deaths Using Gastil’s (1971) Measure of Southern Cultural Influence (SCI)

Accidents Rurality Temperature Economic Cancer HPSA Speed Auto-Travel Police Culture of Honor

Total (White) .37** .06 .09 .07 .34** – – – .54**

Male .37** .02 .09 .03 .36** – – – .49**
Female .38** .15 .09 .08 .30** – – – .59**
Total (Non-White) .20y .22 .09 .09 .73** – – – .03
Male .20y .17 .09 .12 .74** – – – .00
Female .22y .27y .07 .05 .70** – – – .07
Transport
Total (White) .34** .00 .14 .03 .24* .10 .19* .05 .42**
Male .34** .00 .13 .02 .26* .05 .19* .05 .43**
Female .32** .02 .17 .01 .15 .23* .18* .02 .35**
Nontransport
Total (White) .19 .15 .12 .06 .19 – – – .52*
Male .19 .07 .14 .01 .21 – – – .42*
Female .20 .31 .10 .09 .13 – – – .64**

Rurality ¼ proportion of population living in rural areas; Temperature ¼ mean annual temperature; Economic ¼ economic deprivation; Cancer ¼ number
of cancer-related deaths per 100,000 persons; HPSA ¼ proportion of population living in a health professional shortage area; Speed ¼ Average speed limit
across road types; Auto-travel ¼ composite auto-travel variable; Police ¼ number of police officers per capita; CHC ¼ culture-of-honor classification
(0 ¼ nonhonor state; 1 ¼ honor state).
y
.10 > p  .05; *p < .05; **p < .01.

similarly, controlling for accidental deaths among non-Whites p < .001, d ¼ 1.56. Thus, although we cannot measure all
should strengthen our confidence that we are observing a possible confounds directly, controlling for rates of non-
culture-of-honor phenomenon among Whites, not simply a White accidental deaths should strengthen our inference about
spurious association. To test this idea, we regressed White the role of culture of honor in accidental deaths among Whites.
accidental deaths on CHC and (separately) SCI, controlling for Keeping our focus on deaths among Whites, we next exam-
non-White accidental deaths. Although non-White deaths ined whether culture-of-honor status was a significant predictor
were predictive of White deaths, b ¼ .40, t ¼ 3.83, p < .001, of both transport and nontransport accidents. For our analysis
d ¼ 1.12, so too was CHC, b ¼ .52, t ¼ 4.91, p < .001, of transport accidents, we included three additional covariates:
d ¼ 1.43. Likewise, SCI was a significant predictor of White (a) state average speed limits, (b) the composite vehicle-travel
accidental deaths, b ¼ .66, t ¼ 7.80, p < .001, d ¼ 2.28, con- variable, and (c) number of police officers per capita. CHC was
trolling for non-White accidental deaths, b ¼ .45, t ¼ 5.35, a significant predictor in this analysis of transport accidents,

Downloaded from spp.sagepub.com at Universidad del Norte on August 7, 2013

104 Social Psychological and Personality Science 3(1)

association with accidental deaths was significant within

60
Metro metro, F(1, 42) ¼ 6.11, p ¼ .018, d ¼ 0.90, and nonmetro
50 Non-Metro regions, F(1, 42) ¼ 15.22, p < .001, d ¼ 1.43, though the inter-
Accidental Death Rates

action indicates that the latter association is significantly

40 larger. Thus, from all of these analyses, it appears that the
culture of honor has implications for fatalities that result
30
from unintentional causes, a unique finding in the culture-of-
20 honor literature.

10
Study 2
0 The statewide analyses of Study 1 were predicated on two
Non-Honor States Honor States
assumptions: (a) the accidental deaths cataloged by the CDC
Culture of Honor Classification are, at least partially, reflective of underlying risk-taking
behaviors and (b) the United States can be divided into regions
Figure 1. Age-adjusted accidental death rates per 100,000 persons for that differ on the degree to which they are characterized by an
Whites as a function of culture-of-honor classification and metropolitan/ ideology of honor. Although we have built a case for the
nonmetropolitan status.
tenability of these assumptions, more convincing evidence for
a link between the culture of honor and risk taking might
with Whites in honor states (M ¼ 19.46) having significantly emerge if we assessed these variables directly, at the individual
higher transport-related accidental deaths than Whites in level. To this end, we conducted Study 2 to examine whether
nonhonor states (M ¼ 16.57), b ¼ .26, t ¼ 3.19, p ¼ .003, higher levels of risk taking might be found among individuals
d ¼ 1.01. Importantly, this finding was corroborated by the who endorse the ideology of honor, using a self-report honor
SCI index, b ¼ .42, t ¼ 3.91, p < .001, d ¼ 1.24, and analogous ideology scale recently devised and validated by Barnes,
patterns obtained when White males and females were Brown, and Osterman (2011).
analyzed separately (ps  .006). Eliminating the transport-
related covariates from the model and focusing only on
nontransport accidents, we found that CHC was a significant Method
predictor of death rates, but only among White females, Participants
b ¼ .31, t ¼ 2.07, p ¼ .044, d ¼ 0.63. SCI, on the other hand,
significantly predicted nontransport accidents whether Participants were 103 undergraduates (79 females) from the
White males and females were analyzed together or separately University of Oklahoma who received course credit for taking
(ps < .05). part; their mean age was 18.6 (SD ¼ 0.8) years, and the major-
ity identified themselves as White/Caucasian (73.5%), Asian/
Pacific Islander (9.8%), Hispanic/Latino (6.9%), Black/African
Accidental Deaths in Metro Versus Nonmetro Areas American (4.9%), Native American (2.9%), and Other (2%).
As in Study 1, our primary analyses are limited to self-
Nisbett and Cohen (1996) found that homicide rates were identified White/Caucasians or Hispanic/Latinos. One partici-
especially pronounced in small towns in honor states, where, pant did not complete all of the measures and was excluded,
among other reasons, inhabitants have greater reason to be con- leaving a final sample 81 participants (62 females).
cerned with their reputations being known throughout their
communities. To test for a similar moderator, we analyzed
accidental deaths according to whether they occurred in areas
Measures and Procedure
classified by the CDC as metro (counties in metro areas with All participants completed the Honor Ideology for Manhood
populations greater than or equal to 50,000) or nonmetro Scale (HIM; Barnes et al., 2011) and the Rosenberg Self-
(counties with urban populations less than 50,000), in a two- Esteem Scale (RSES; Rosenberg, 1965) during an online
way, repeated measures analysis of covariance (ANCOVA).5 testing session at least 2 weeks prior to the laboratory portion
Cohen’s (1998) CHC variable was the between-states factor, of the study. The HIM contains 16 statements (a ¼ .91) derived
and whether the deaths occurred in metro or nonmetro areas from prior research on the U.S. culture of honor (e.g., Cohen &
was the within-states factor. Annual temperature, economic Nisbett, 1994). Eight items reference beliefs about the nature of
deprivation, cancer death rates, and HPSA were entered as a ‘‘real man’’ (e.g., ‘‘A real man doesn’t let other people push
covariates. Consistent with prior research on homicide, the him around’’), and the other eight items reference beliefs about
between-within interaction was significant, F(1, 42) ¼ 5.07, conditions under which men are justified in engaging in retalia-
p ¼ .030, with accidental death rates among Whites being tory violence (e.g., ‘‘A man has the right to act with physical
highest in nonmetro regions of honor states (see Figure 1). Sig- aggression toward another man who calls him an insulting
nificant covariates in the model included economic deprivation name’’). The items of the HIM are not self-descriptive, but
and HPSA (ps  .024). Importantly, the culture-of-honor ideological. Thus, both men and women can readily endorse

Downloaded from spp.sagepub.com at Universidad del Norte on August 7, 2013

Barnes et al. 105

(or not endorse) these items. Respondents indicate their level of Considering the likelihood that men’s risky behaviors and
agreement with these items on scales ranging from strongly dis- masculine risk-taking schemas also affect women, we tested
agree (¼ 1) to strongly agree (¼ 9). The RSE is a 10-item scale our culture-of-honor hypothesis by examining accidental
(a ¼ .89) containing statements regarding self-worth, to which deaths across genders. In addition, we controlled for a host of
respondents indicate their level of agreement on scales anchored statewide variables that might account for the expected differ-
with strongly disagree (¼ 1) and strongly agree (¼ 4). ences in accidental deaths (e.g., rurality, temperature, eco-
In the laboratory, participants were told that the study was nomic deprivation). As hypothesized, we found that state
designed to explore certain types of decision making. They honor status (assessed in two different ways) significantly
completed several questionnaires (randomized for each predicted accidental deaths among Whites from 1999 to
participant), including the Domain-Specific Risk-Taking 2006, and that this regional difference was especially pro-
Scale (DOSPERT; Weber, Blais, & Betz, 2002) and the Big nounced among Whites living in nonmetropolitan areas.
Five Inventory (BFI; John, Donahue, & Kentle, 1991). The Consistent with previous research on regional differences in
DOSPERT consists of 40 risky behaviors (e.g., bungee jump- violence, culture of honor was not associated with accidental
ing off a bridge, gambling away a week’s income) that span deaths among non-Whites, and its association with death rates
five domains: ethical, financial, health/safety, recreational, and among Whites even persisted when we controlled for non-
social (a ¼ .79). Participants estimated the likelihood they White death rates.
would perform each behavior if given the opportunity. The BFI In terms of absolute magnitude, culture of honor appears to
is a 44-item index of the five personality dimensions of have a stronger association with White male risk taking than
extraversion (a ¼ .87), agreeableness (a ¼ .79), conscientious- with White female risk taking—specifically, the difference
ness (a ¼ .83), neuroticism (a ¼ .84), and openness (a ¼ .84) between covariate-adjusted accidental death rates among
that has shown strong consistency and predictive validity in White males in honor states (M = 57.68) and non-honor states
previous studies. After completing these measures and a task (M = 51.56) was about 1.5 times as large as the same difference
related to another study, participants were debriefed and among White females in honor states (M = 27.42) and
excused. non-honor states (M = 23.26). Even so, the association between
culture of honor and female accidental deaths was statistically
significant in Study 1. This association was further supported at
Results and Discussion the individual level in Study 2, which examined self-reported
To examine whether endorsement of honor-related beliefs risk taking and behavioral inhibition within a sample of White
predicts enhanced risk taking at the individual level, we college students, most of whom were females. Although our
regressed DOSPERT scores on the HIM, controlling for age, hypotheses about the relation between culture of honor and risk
sex, self-esteem, and the BFI subscales. Consistent with our taking concern social dynamics surrounding the promotion and
predictions, the HIM was positively associated with risk preservation of masculine identities, our evidence regarding
taking, b ¼ .26, t ¼ 2.37, p ¼ .02. Only agreeableness excessive risk taking among women can be understood in at
(b ¼ .32, t ¼ 2.58, p ¼ .01) and extraversion (b ¼ .32, t ¼ least two ways. First, some evidence indicates that the degree
2.70, p < .01) were significant covariates. These individual- of individuals’ conformity to the masculine stereotype predicts
level analyses thus support our interpretation of the accidental risk taking independent of biological sex (Granié, 2009; Özkan
death rates at the statewide level in Study 1. Also, because most & Lajunen, 2006; Raithel, 2003), and insofar as conformity to
of the participants in Study 2 were women, these data suggest the masculine gender role can be heightened by situational
that the significant association between culture of honor and primes (Bargh et al., 1996) and social pressures (e.g., Bem,
accidental death rates among women in Study 1 was not simply 1975) that are, conceivably, more pronounced in honor states
a function of women being passive victims of male risk taking. than nonhonor states, it seems plausible that the accidental
deaths of women result from their mimicking, wittingly or
unwittingly, the risky behaviors of men. Second, even if
General Discussion women do engage in excessive risk taking themselves, it is also
Despite the argument that men from every cultural background possible that they die as a consequence of men’s dangerous
are motivated to prove their masculine identity (e.g., Bosson actions, as would be the case if a husband crashed his car as
et al., 2009; Vandello et al., 2008), research suggests that men a result of driving recklessly through traffic with his wife and
(White men, in particular) in the southern and western United children as passengers, or if a male maintenance worker elec-
States who have been influenced by an ideology of honor are trocuted himself and his female coworker while working with
especially driven to achieve this goal (e.g., Cohen et al., live wires in a puddle of water.
1996; Cohen & Nisbett, 1994; Nisbett et al., 1995). Conse- Given these findings, what might be done to reduce acciden-
quently, men from culture-of-honor regions might be more tal deaths arising from excessive risk taking by people socia-
prone to engage in risky behaviors that sometimes lead to lized by an honor ideology? Although large-scale attempts at
death, relative to men from nonculture-of-honor regions, cultural engineering are unlikely to succeed, it might be
because such behaviors signify that one possesses the ‘‘manly’’ possible to use the force of the culture of honor itself to pro-
attributes of strength and courage (Bosson et al., 2009). mote the public welfare (see Apiah, 2010). Specifically,

Downloaded from spp.sagepub.com at Universidad del Norte on August 7, 2013

106 Social Psychological and Personality Science 3(1)

perhaps interventions that shame people into safer behaviors on action. Journal of Personality and Social Psychology, 71,
(e.g., ‘‘Don’t be a sissy, buckle up!’’), or rely on strong, high 230-244.
status figures as models of responsible conduct would be effec- Barnes, C. D., Brown, R. P., & Osterman, L. L. (2011). Don’t tread on
tive at modifying social schemas and scripts about the meaning me: The cultural ideology of honor and militant responses to
of strength and courage. Whether such an approach would terrorism. Manuscript submitted for publication.
effectively diminish accidental deaths in cultures of honor is Bem, S. L. (1975). Sex-role adaptability: One consequence of
uncertain, but it seems reasonable to infer from the present psychological androgyny. Journal of Personality and Social
findings that merely including seatbelts in motor vehicles or Psychology, 31, 634-643.
warning labels on firecrackers might not be enough to prevent Bosson, J. K., Vandello, J. A., Burnaford, R. M., Weaver, J. R., &
accidents in cultures that confer social status on those who are Wasti, S. A. (2009). Precarious manhood and displays of physical
willing to throw caution to the wind and live dangerously. aggression. Personality and Social Psychology Bulletin, 35,
Thus, the cultural dynamic that we have captured in these 623-634.
studies reveals a threat to public health that goes beyond the Brown, R. P., & Osterman, L. L. (in press). Culture of honor, violence,
risk of interpersonal violence documented in previous and homicide. In T. Shackelford & V. W. Shackelfrod (Eds.),
research and points to the possibility that life in honor- Oxford handbook of evolutionary perspectives on violence, homi-
oriented societies is more treacherous than previously realized. cide, and war. Oxford University Press.
Brown, R. P., Osterman, L. L., & Barnes, C. D. (2009). School
Acknowledgment violence and the culture of honor. Psychological Science, 20,
The authors thank Mauricio Carvallo and Peter H. Gries for their 1400-1405.
helpful comments on an earlier draft of this manuscript. Cash, W. J. (1941). The mind of the south. New York, NY: Vintage.
Center for Disease Control and Prevention. (1999–2006). Compressed
Declaration of Conflicting Interests mortality file: ICD 10 coded mortality rates for 1999-2006. Retrieved
The author(s) declared no potential conflicts of interest with respect October 12, 2010, from http://wonder.cdc.gov/mortsql.html
to the research, authorship, and/or publication of this article. Cohen, D. (1998). Culture, social organization, and patterns of
violence. Journal of Personality and Social Psychology, 75,
Funding 408-419.
The author(s) received no financial support for the research, authorship, Cohen, D., & Nisbett, R. E. (1994). Self-protection and the culture of
and/or publication of this article. honor: Explaining southern violence. Personality and Social
Psychology Bulletin, 20, 551-567.
Notes Cohen, D., & Nisbett, R. E. (1997). Field experiments examining the
1. Some culture-of-honor studies have distinguished between Whites culture of honor: The role of institutions in perpetuating norms
of Hispanic and non-Hispanic origins, but others have not (e.g., about violence. Personality and Social Psychology Bulletin, 23,
Cohen et al., 1996). In our analyses, we include Whites of both 1188-1199.
Hispanic and non-Hispanic origins. Analyses that included only Cohen, D., & Nisbett, R. E. (1998). Are there differences in fatalism
Whites of non-Hispanic origins did not differ appreciably from between rural southerners and midwesterners? Journal of Applied
those reported here. Social Psychology, 28, 2181-2195.
2. Having little to do with risk taking, deaths caused by ‘‘exposure to Cohen, D., Nisbett, R. E., Bowdle, B. F., & Schwarz, N. (1996). Insult,
forces of nature’’ (e.g., hurricanes, tornadoes, earthquakes) were aggression, and the southern culture of honor: An ‘‘experimental
excluded from the data. ethnography.’’ Journal of Personality and Social Psychology, 70,
3. Cancer death rates were obtained separately for Whites and 945-960.
non-White for use in White and non-White analyses, respectively. Fischer, D. H. (1989). Albion’s seed: Four British folkways in Amer-
4. Taking the natural log of accidental deaths among non-Whites ica. New York, NY: Oxford University Press.
reduced the skewness of this variable and appeared to improve the Gale, B. (Producer/Writer), Canton, N. (Producer), & Zemeckis, R.
degree to which the regression assumption of homoscedasticity (Director/Writer). (1989). Back to the future part II [Motion
was satisfied. Even with this modification, the results remained picture]. United States: Amblin Entertainment.
nonsignificant, both at the aggregate level, and when non-White Gastil, R. D. (1971). Homicide and a regional culture of violence.
males and females were analyzed separately. American Sociological Review, 36, 412-427.
5. Because of the absence of nonmetro regions in New Jersey and Granié, M (2009). Effects of gender, sex-stereotype conformity, age
Rhode Island, these states were excluded from the analysis. and internalization on risk-taking among adolescent pedestrians.
Safety Science, 47, 1277-1283.
References Insurance Institute for Highway Safety. (2011). Maximum posted
Apiah, K. A. (2010). The honor code. New York, NY: W.W. Norton. speed limits, March 2011. Retrieved March 25, 2011, from
Archer, J. (1994). Violence between men. In J. Archer (Ed.), Male www.iihs.org/laws/speedlimits.aspx
violence (pp. 121-140). London: Routledge. John, O. P., Donahue, E. M., & Kentle, R. L. (1991). The big five
Bargh, J. A., Chen, M., & Burrows, L. (1996). Automaticity of social inventory: Versions 4a and 54. Berkeley: University of California,
behavior: Direct effects of trait construct and stereotype activation Berkeley, Institute of Personality and Social Research.

Downloaded from spp.sagepub.com at Universidad del Norte on August 7, 2013

Barnes et al. 107

Loftin, C., & Hill, R. (1974). Regional subculture and homicide: An U.S. Census Bureau. (2000). Census 2000 summary file 3 (SF3)—
examination of the Gastil-Hackney thesis. American Sociological Sample data. Retrieved October 16, 2008, and February 24,
Review, 39, 714-724. 2009, from http://factfinder.census.gov/servlet/DatasetMainPage
National Center for Education Statistics. (2000, 2004). Table 11. Servlet?_program¼DEC&
Educational attainment of persons 18 years old and over, by state: U.S. Census Bureau. (2000, 2004). Table H-8: Median household
Selected years, 1994 to 2004. Retrieved March 24, 2011, from income by state: 1984 to 2009. Retrieved October 21, 2010, from
http://nces.ed.gov/programs/digest/d04/tables/dt04_011.asp http://www.census.gov/hhes/www/income/data/historical/
National Center for Education Statistics. (2000, 2004-2006). Table 20. household/h08.html
Household income and poverty rates, by state: 1990, 2000, and U.S. Department of Health and Human Services, Health Resources
2004-06. Retrieved March 25, 2011, from http://nces.ed.gov/ and Services Administration. (2011). State population and health
programs/digest/d07/tables/dt07_020.asp professional shortage areas designation population statistics.
National Oceanic and Atmospheric Administration. (2000). Average Retrieved March 25, 2011, from http://bhpr.hrsa.gov/shortage/
mean temperature index by month: 1971–2000. Retrieved October primarycare.htm
16, 2008, from http://www.cdc.noaa.gov/USclimate/tmp.state. U.S. Department of Transportation, Federal Highway Administration.
19712000.climo (2005). Table PS-1. Selected measures for identifying peer states.
Nisbett, R. E. (1993). Violence and U.S. regional culture. American Retrieved March 25, 2011, from http://www.fhwa.dot.gov/policy/
Psychologist, 48, 441-449. ohim/hs05/htm/ps1.htm
Nisbett, R. E., & Cohen, D. (1996). Culture of honor: The psychology U.S. Department of Transportation, Federal Highway Administration.
of violence in the South. Boulder, CO: Westview. (2006). Table MV-1. State motor-vehicle registrations. Retrieved
Nisbett, R. E., Polly, G., & Lang, S. (1995). Homicide and U. S. March 25, 2011, from http://www.fhwa.dot.gov/policy/ohim/
regional culture. In B. Ruback & N. Weiner (Eds.), Interpersonal hs06/htm/mv1.htm
violent behavior: Social and cultural aspects (pp. 135-151). New Vandello, J. A., Bosson, J. K., Cohen, D., Burnaford, R. M., &
York, NY: Springer. Weaver, J. R. (2008). Precarious manhood. Journal of Personality
Özkan, T., & Lajunen, T. (2006). What causes the differences in and Social Psychology, 95, 1325-1339.
driving between young men and women? The effects of gender Vandello, J. A., Cohen, D., & Ransom, S. (2008). U.S. southern and
roles and sex on young drivers’ driving behavior and self- northern differences in perceptions of norms about aggression:
assessment of skills. Transportation Research Part F: Traffic Mechanisms for the perpetuation of a culture of honor. Journal
Psychology and Behaviour, 9, 269-277. of Cross-Cultural Psychology, 39, 162-177.
Raithel, J. (2003). Risikobezogenes verhalten und geschlechtsrolle- Weaver, J. R., Vandello, J. A., Bosson, J. K., & Burnaford, R. M.
norientierung im jugendalter [Risk-taking behavior and gender (2010). The proof is in the punch: Gender differences in percep-
role orientation in adolescents]. Zeitschrift fu¨r Gesundheitspsy- tions of action and aggression as components of manhood. Sex
chologie, 11, 21-28. Roles, 62, 241-251.
Roebuck, J. B., & Murty, K. S. (1996). The southern subculture of Weber, E. U., Blais, A., & Betz, N. E. (2002). A domain-specific
drinking and driving: A generalized deviance model for the south- risk-attitude scale: Measuring risk perceptions and risk behaviors.
ern White male. New York, NY: Garland. Journal of Behavioral Decision Making, 15, 263-290.
Rosenberg, M. (1965). Society and the adolescent self-image. Wyatt-Brown, B. (1982). Southern honor: Ethics and behavior in the
Princeton, NJ: Princeton University Press. Old South. New York, NY: Oxford University Press.
Sims, J. H., & Baumann, D. D. (1972). The tornado threat: Coping
styles of the North and South. Science, 176, 1386-1392.
Bios
U.S. Bureau of Labor Statistics. (2000, 2004). Unemployment rates for
states. Retrieved March 25, 2011, from http://www.bls.gov/lau/ Collin D. Barnes is a postdoctoral research fellow with the Institute
lastrk04.htm, and http://www.bls.gov/lau/lastrk04.htm for U.S.-China Issues at the University of Oklahoma.
U.S. Census 2000 on www.Allcountries.org. (1996). 353. Number and
Ryan P. Brown is an associate professor of psychology at the University
rate of full-time sworn police officers in state and local law
of Oklahoma.
enforcement agencies. Retrieved March 25, 2011, from http://
www.allcountries.org/uscensus/353_number_and_rate_of_full_ Michael Tamborski is a doctoral candidate in social psychology at
time.html the University of Oklahoma.

Downloaded from spp.sagepub.com at Universidad del Norte on August 7, 2013