Download as pdf or txt
Download as pdf or txt
You are on page 1of 15

7

- . _ f Política! Behavior
The Mu/tiple Bases o

. ,, mpasses a wide array of phenomena,


The term "political behav10r efco ple interacting with, and responding
O
most of which convey .ª hsense penod them People follow contemporary
1· · 1 · I t at surrou
to, the po 1t1ca stimu 1d.
·
d 's newspaper or watching this evening's
. · I events by rea mg d"to pprove
po l1t1ca ay ·
of the job the pres1dent currently 1s
·
TV news; they approve or isa . . ' . 1· h f hº
· d h f candidates 1n th1s year s election. In 1g t o t 1s
domg· an t ey vote or 1· . l b h .
· d' · · · derstandable that most accounts of po 1t1ca e av1or
1mme 1acy, 1t 1s un .
contemplare how people may be influenced by environm~ntal_ forces, ~nd
especially by forces in operation just prior to the behav1~rs 1n quest1on.
Thus, a great deal of research examines matters such as the 1mpact of news
media on political attitudes, of partisan mobilization on voter turnout, and
of campaign content on candidate choice. Or, reaching back to earlier influ-
ences, studies consider the effects of childhood political socialization, edu-
cation, and the person's exposure to others within various social contexts.
To the extent that such research calls attention to one key set of deter-
minants of political behavior, we should have no qualms about a focus
on the influence of environmental forces. However, diligence is needed
to ensure that ~~ch a focu~ does not produce analytical myopia. Any
ahccount of P?htical behav1or positing, either explicitly or implicitly
t at only environmental forces matter necessanly • assumes that people'
fi rst encounter the w0 ld 0 f ¡1· 1•
assumption is wro ~ f
recognize the impa:g~f th u '!ºh
11
po ~ c~ as political blank slates. And that
st1c theory ~f political behavior must
tors, innate biological or b~e~ ~oa~ s~ts of variables: environmental fac-
10
between environmental d _ogica Y influenced factors, and interactions
b an 1nnate fo M
een ~o help bring the second d . r~es. Y purpose in this book has
the .~1x by highlighting the . an thrrd of these sets of variables into

:;:t t:t ~:sonality is significant :n~t~::1ve, I ha~e sought to empha-


poht1cal b h ·
.
importance of pe ¡· .
· e av1or. In pursuing thi b. . rsona 1ty as an 1nfluence on

Y ect of personality 1·mpl· . l as ª pers1stent force but also


ic1t y signa ls t he underlying
. ' role of
The Mu/tiple Bases of Political B e h aviar
.
. Jogy. Empirically,
ºb this study . has shown th at f un damental ps h 1 . 1
bJO
·fferences contn ute to vanance in political beh av1or
. both di y_c o ¡ogicad
• h
d1 ough interacuon •
w1t environmental infl
t1 uences. ence b rect Y'd
H an
.1gr rhe role of personahty, a much richer and
.
d, y c_o~s1 er-
1n f .. lb h . more nuance depicuo f
rhe bases o po 1iuca e av10r becomes possible. no
Although. the theory
f h . fl I have advanced facilitates de l f
ve opment o a more
intricare v1ew o . t e m uences on human behav1or, · such a view . may
bring consternatton
. . h b to sorne. The present discussiºon is ar y t e fi rst
· h dl h
time· rh1s 1ssue as
, ¡· Ph'll' een confronted. In her biography o f th e wnter
· Al'ice
Sheldon, Ju te 1 1ps recounts th~ nervous skepticism sorne of Sheldon's
sychological. research. had faced m the 196os, comment'mg t h at (2006
P ¡ "Ther_e 1s a s1de to behaviorism that's very liberating: if the huma~
204
psyche begms , h as a blank
h. slate,
· then we can change anything . Bu t w h at
if you can ~ e a~ge t ~ngs m yourself?" I have posited that personality is
a stable, b1olog1cally mfluenced psychological structure. It changes ve
little, if at all, during on~'s adul~ years. This implies, as one example, rh17t
introverts cannot refash1on their psychological compositions into those
of e~rav~rts. I_ntrovert~ may endeav~r to become more outgoing and
talkattve m their behav1or, but they w1ll not do so by transforming their
basic psychological tendencies; instead, they can only strive to adjust
their behavior despite the inertial force of their core personality traits.
Two facets of this discussion require further elaboration. First, my the-
sis is not one of biological or psychological determinism. Under my con-
ceptualization, personalit)'. _trai_~ ~ e\ psychological structures that give
rise to certain tendencies in oehavior, but traits do not predetermine that
such behavior must and will occur. The trait literature abounds with evi-
dence consistent with my perspective. For each of us, personality traits
establish central tendencies, but actual behavior can and <loes vary mark-
edly around these cenrral tend~_ric~. Recall, for example that i11isc~el
(1979) , a chief critic o trait approaches, found an average correlat1on
of 0.30 in trans-situational behavior. Although the precise estímate can
be debated, I ag ree with Misch el's hroa der po int that traits are not abso-
lute determinant~ uf behavior. As apphed to poli.tics, my argument is not
that people are born to be well informed, conservative, or politically
engaged, but instead that biologically influenced psychological structures
exert meaningful and systematic impact on the likelihood of the~e and
other forms of political behavior. Tendencies matter, but tendencies are
not requirements.
th
1 Alice Sheldon gained fame as a science fiction writer, writing primarily uocle~ e
name James Tiptree Jr. Earlier in her life, she had been an African explorer, p~i~te~,
' . d • h · d th t ave rise to Ph11hps s
WAC, poultry farmer, CIA operattve, an , m t e peno a g . 1 6
comment, a graduare student who earned a doctorare in psychology m 9 7·

183
. and t he Foundations of Política!
Personaltty . Behavior
d · · al pomt . 1s. that what I advocate 1s much more th an
A secon ~rmc f new variables to our models. lmplicit in n,
the mere add11 n °
10
/º¿'1; students of political behavior require a ne y
approach 1s the be1h~f t ª d way of thinking about the bases of politi wl,
. nd mu t1 acere h 1 . 1 d ºff ca
expans1ve,
. w, a st recogmze . that basic psyc o og1ca . fl1 erences exist· "" we
act10n. we mu h
·ze t at t ese h di'fferences often exert
h fCm uence on politic a¡
must recogm nner that is causally prior to_ t e e ,ects of envirorunen.
behav10r
If mLastly:
ª ma we mus t recognize that env1ronmental
. . b . factors will not
taI orces. :d
s produce I ent1ca . l effects on all ind1v1duals,
' d ut mstead
d that these
.
a way ·¡¡ as a function of people s eep-roote psychological
effects ~ftenOw1
tendenc1es. nce var:onsider
we the possible impact of 11 personality,
. f 1our the-
ones . of po11t1ca
. · 1 behavior expand from one conste at1on o exp anatory
factors to three. . e ·d bl · · •
The thesis advanced here imphes the ~eed 1or cons1 era e llllttative
m
• terms of theory building ' methodolog1cal advances, . and substantive
inquiry. The present effort has centered on the dommant contem~orary
d I of personality trait structure, the five-factor approach. lnd1cators
mothee Big Five have been developed, an d t h e m~p
of · 1·1cat10ns
· of these trait
dimensions for a large number of dependent variables have been consid-
ered. The tests have explored both direct effects and numerous forms of
indirect, nonmonotonic, and conditional relationships. At virtually every
turn, we have seen that personality <loes indeed constitute a key founda-
tion of política! behavior. Although the substantive ground covered in this
book has been expansive, it should be dear that anything accomplished
here represents only the very first step in pursuit of what promises to be
an extremely rich stream of scholarly research. My hope is that this book
will have persuaded at least sorne readers that there is value in studying
the political effects of fundamental psychological differences, and will
have sparked interest in further research on the multiple bases of political
behavior. In the remainder of this chapter, I discuss a few of the lingering
questions pages.
preceding and future directions suggested by the analyses reported in the

THE ROLE OF PERSONALITY IN THEORIES OF POLITICAL


BEHAVIOR
st
The mo immediate need in further contemplating the possible impact
of personality on political behavior is for theory building and the careful
derivation
espec1 11 dof additional
· b hypotheses. For direct effects' this matter is not
l
. ª Y auntmg ecause past research in trait psychology offers amp e
gru_dance regarding the likely consequences of variance in the Big Five
btra1t dnnens1ons. In the preceding chapters numerous hypotheses have
een tested offerin 11·d 1.e ' f d· t
' g ª so oundation for subsequent study o 1: ec
The Mu/tiple Bases of Political Behavior
f{ ts. However, in m~ving beyond these direct effects, a great <leal of
e ec 'zing will be reqmred before the full significance of enduring psy-
rheorl
chological .
d1fferences w1·11 be ~ea 1·1zed . In part1cu!ar,
. c_rea~ive and diligent
effort will be needed to _spec1fy the many poss1ble mdirect and condi-
. ¡ effects of persona 11ty.
uona . h
In the three pnor e apter~, exam~les were offered of the chief types
of personality_ effects f?r wh1ch I_ beheve a_dditional hypotheses must be
crafted. First, in many mstances, lt _se~ms hkely that the influence of per-
nality will operate through med1atmg factors. For political attitudes
::e plausible pat~, as _t~sted in Chapter 5? includes personality effect~
on political pred1spos1t10ns such as part1sanship and ideology, with
those predispositions, in turn, shaping political attitudes. This same line
of reasoning must be applied in theory building pertinent to other sorts
of dependent variables. For instance, in Chapters 4 and 6, openness
to experience and extraversion were found to exert positive effects on
numerous forros of political engagement, but it is not yet known what
intermediary factors, if any, must be added for the causal path to be
complete.
The second possibility that requires further theorizing concerns non-
monotonic effects of personality. In most of the tests reported in the prior
three chapters, indicators of the Big Five were modeled in simple additive
form. It is conceivable, however, as was demonstrated by the case of con-
scientiousness and preference for legislator action in Chapter 5, that many
personality effects will be nonmonotonic. As one example, extraversion
influences the inclination to engage in social forms of political partici-
pation, but the impact of this trait dimension does not necessarily rise
steadily across the introversion-extraversion scale. One obvious alternate
is that there may be sorne sort of threshold effect, either in which extra-
version must be very low before social forms of participation become
unlikely or in which only the very most outgoing individuals gravitate
toward these behaviors. Positing the existence and direction of a relation-
ship are much easier matters than specifying its form, but our theoretical
and empirical models will remain imprecise absent the latter.
Additional work also is required to construct hypotheses regarding
~ossible conditional effects of personality. One type of conditional rela-
ti?nship involves interactions between two or more of the Big Five trait
dimensions. · We saw examples of this in Chapter 4 with t e mteraction
h . .
between openness to experience and conscientiousness in two models
. po1·Itical
of · knowledge. Research in trait psychology often hypotheSizes .
mteractions between Big Five traits, but applications of the five-factor
framework in the study of political behavior must advance beyond ~he
present fledgling state for such conditional effects to become routme
co~ponents of theory-driven inquiry.
. d the Foundations of Politica/ Behavior
Personaltty an . . .
·t · onal relat1onsh1p 1nvolves the inte·
d of con d1 1 f racti
A secon . cype ·r and environmenta 1 actors. My thesi·s h 0 n
licy tra1 s d . old
between persona_ l . lly influenced forces an envuonmental f s,
h
first, t at both b10 og1ca d h
. . 1 b havior. and secon , t at spec1 e element f"fi actors
matter or f P ohttca
d eri·es of 'variables l"k1 e1y operate 1n• mteract
. ·
s rom .
b a catego . 10 n W tth
these two ro . . 1 put variance in persona11ty· often leads peo l
other. S1mp Y ' • ¡· · P e to
one an . .; 1 to environmenta1 st1mu 1. 1t 1s one thing ho
d d1frerent Y . d . ' Wever
respon h 1·ntuition that personahty an env1ronment inte ,
embrace t e .fi h ract
to . ther to delineare and test spec1 e ypotheses regard· '
and qmte ano . h. F d d tng
form of such relat1ons 1ps. or eca es, scholars in th
the nature an d . d "bl • . e
field of trait psychology _have menttone poss1 e 1:11reract10?~ hetween
. vana
tra1t • bl es and situatwnal forces, . · yet
1 theory-dnvenbemp1ncal studY
f 8 h effects remains disappo1nt1ng y rare. It may e that scholars
~00 Jicting applied research will need to take the lead _role in. identify-
ing these conditional effect~. Two e~amples of person~hty-env1r~nment
·nteractions were reported 1n the pnor chapters, that 1s, concermng the
:ize of discussion networks and exposure to cross-cutting discourse, and
negative campaign tone and projected vo~er turno~t. Countless addi-
tional relationships such as these are poss1ble. In v1rtually any labora-
tory experiment in political psychology, for example, we should at least
pause to consider whether our manipulations may produce effects that
vary as a function of participants' personality traits. Absent attention
to personality-environment interactions, our explanations of political
behavior tend to have a simplistic "one-size-fits-all" character. Moving
beyond that point will require rigorous effort, but the pay off promises
to be more elaborare and powerful theories of how citizens engage the
political world.
In addition to theorizing regarding the f orm of personality effects, fur-
ther attention must be directed to how personality fits within broader
a~counts of political behavior. One matter is whether research on person-
ahty and politics should be wedded to the Big Five. The emergence of five-
factor models in the late 1980s genuinely revolutionized the study of trait
psyc~olo?y, and it is an extreme understatement to note that research 00
th
e ~ig Five in the past two decades has been impressive in its scope ao<l
qualhity. ~ch~Iars in many fields have conducted applied work drawing
on t e B1g F1ve In rh· 1· • . h ame.
.
As d1scussed . · is sense, po 1t1cal sc1ence has come late to t e g
1n Cha t 2 fi f ctor
tax · d P er , most of the lingering dispute about ve-ª
onom1es oes not h 11 . r· ous-
n . e ª enge the centrality of openness conscien 1
ess, extravers1on a bl ' d me
psychologists d. ' greea eness, and emotional stability. Instea , s~l"tY
in either redu . isagrhee r~gat<ling whether there might be greater utl l r ·
c1ng t e Btg p- . . . · ns o
splitting some or U0 f ~ve ~o two h1gher-order tra1t d1men sto ' rs,
th
Given that cons ª e Big Five traits into multiple subsidiary fact~at
ensus strongly supports the Big Five, my belief is t
The Mu/tiple Bases of Política/ Behavior
. oach presently offers the most useful and well-grounded start·
h s appr . . . d. . mg
t \nt for political s~1_ ent1sts mtereste m conductmg applied research on
Pº lity and pohtics.
persona f h B' F
Although I advocat_e obcul_s don t Teh ig iv~, we should not invoke this
while wearmg m ers. ree pomts warrant mention F' t
approac h . b d . irs ,
rn e purposes 1t may e a equate, and perhaps even desirable to
for sont only for t h e two h.1gh er-or d er tra1t · d.1mens1ons
· '
that apparently
accournpass the Big . p·1ve. o 1gman
· (1997 ) suggests that a h1erarchical
. view
enco d . . h'
connects openn~ss ~n extravers10n m one 1gher-order trait dimen-
. and consc1ent10usness, agreeableness, and emotional stability in
s1on, .
ther. Digman further argues that these h1gher-order factors are easily
ano with severa1 tra d'
linked ·
. 1t1ona1 t h eones
· o f personality. In man y of the
tests reported in the pnor three chapters, and especially those concern-
ing informati?n in_ Chapter 4 and ~~rticipation in_ Chapter 6, openness
and extravers10n y1elded strong pos1t1ve effects wh1le the remaining trait
dimensions produced a great number of modest negative effects. This
pattern is at least loosely consistent with Digman's empirical observa-
tions. It would be extremely premature to conclude that research in polit-
ical science should forego the Big Five in favor of a "Big Two" approach.
Still, the pattern emerging from present empirical tests is suggestive, and
thus, at a mínimum, we would be wise to be receptive to the possibility
that sorne of the Big Five trait dimensions bring reinforcing influences,
yielding general patterns consistent with Digman's claimed higher-order
trait dimensions.
A second possibility, one with the opposite logic of a move to two
global factors, is that applied research on personality and political behav-
ior will be most fruitful when the Big Five dimensions are parsed into
subsidiary traits. In psychology, a typical progression begins with iden-
tification of effects of one or more of the Big Five dimensions, followed
by subsequent research that seeks to pinpoint the specific elements of
the ~imension that are most important for the phenomenon in question.
For mstance, if initial tests show that conscientiousness matters for sorne
aspect of job performance, follow-up work might explore which _sub-
c_omponents of conscientiousness drive this relationship. I see it as highly
hkely that political scientists will find value in pursuing a similar course.
B~t, fir st things first more work is needed to posit and test effects of th e
Big p· ' . .
b ive. Down the road, refinements centered on subsidiary traits are to
e welcomed.
A final point to reiterate concerning the Big Five is that even the frame-
Work's
t m0st ardent advocates recognize that the taxonomy does no t cap-
the eOfrrety o f variance in psychological di ff erences (e.g.,.Sa ucier
Gureldb . . and
pi . erg 1 998). In most research on politics, the starting pomt m mcor-
rating personality in our theories should be the Big Five. However, we

187
. nd the Foundations of Política[ Beh .
Persona ltty ª avior
tive to the possibility that particular trait f
should remainbrecehp five-factor framework also may be e actors not
sed y t e . . onseq
enco_mpas S cier and Goldberg (r 99 8) 1dent1fied a handfut UentiaI.
For mstance,f l~u captured by the Big Five, including at least of traits
that are
. not du Y 1·udice - that may . b f 1
e o re evance in . man. two _ re¡·_
giou~ness an pre y studies of1
olitical behav10r. . fi . f .
P . b ond the immed1ate con nes o the Big Five 0
Movmg ey . . . , ne top·
'd closely involves the mterre 1at1onsh1p among gen . te to
1
cons er more k h . etics
l.
sona 1ty, an d politics At
· . present, we now
. t at
d vanance
.. in g 'Per-
enes e .
ds with variance m both persona1ity an poht1cal behav· Or-
respon . d .h 1or, and
h d 'fferences in personahty a1so correspon w1t variancei. .
t at 1 d . f n Polu
. behavior. A reasonable inference to enve rom this array of fi -
1ca1 ¡· . l b h . nd-
.mgs 1s· that biological effects on po ltlca . e hav1or . operate
. at least Panly
through personality. Ge~es sh~~e baste p~yc o1og1cal d1fferences that, in
t rn influence patterns m poht1cal behav1or.
u The evidence regarding these three relationships is strong. Nonetheless
the implications of extant work should ~ot be exagg~rated. First, the da~
that personality functions as a mechamsm connectmg genes and politics
has been established only by inf~r~nce. A~d~tional wor~ is needed to pro-
vide direct evidence of personahty s med1atmg role. Th1s presumably will
require analysis of twin or molecular data from a dataset that also includes
indicators of personality and political engagement. Second, although bio-
logical influences on politics may operate partly through personality, it
should not be assumed that personality is the only intervening mechanism.
Thus far, students of biology and politics have been much more active
in identifying genetic influences on political behavior than in explaining
how and why these influences operate. I believe personality to be one key
linkage mechanism, but we should not permit the study of personality
to foreclose consideration of other alternates. Third, if genetic effects do
operate through personality, we should not assume that the causal chain
includes only genes, traits, and the political variables under consideration,
nor should we assume a simple and direct causal chain. As noted previ-
ously, for instance, personality may influence long-term political predispo-
sitions that then shape more immediate attitudes and behaviors. Furt~e~,
even granting a strong influence of biology on both personality and pohti-
cal behavior, the precise expression of these effects may hinge on complex
·
mteractions · ·
mvolving environmental factors. Lastly, one conceiva · blYalso
could extend the causal chain in the opposite direction, specifying those
factors that give rise to biological differences. h
This last pomt · ra1ses
· a question about 1· ust how far back we must reacbl
t 0 devise· satis · factory accounts of political behavior. Th1s • is
· a rea sona del
query. On the one hand, it would not be especially revealing to :t
atteo<lance at the February 17th PTA meeting solely on atte n<lan

188
The Mu/tiple Bases of Political Behavior
F bruary 3rd meeting. On the other hand students f r .
1
rhe Ied not be. forced to ánchor our causal the~ries. in adº. po 1~1cs surfehy
shou . h h h . 1scuss1on o t e
. . B ng Some m1g t argue t át t ere 1s no need for our .
B1g a · . l . . 1. exp1anat1ons to
. rporate 610 ogy, or even persona tty, and that focus on what h
¡neo
•orto the behav1or
· m· quest1on
• 1s
. adequate I obvio
1 d'
appens
¡~st p1~s1 that we must reach back far enough to i~clude tusndy tsagrele.b~y
v1eW d d'ff b arnenta, 10-
. Ily influence 1 erences ecause only by doing· so ca k
Iog1ca f h h · n we ac now1-
he effects o t e t ree central bases of political beh · •
edge t . f d . . av1or: mnate
forces, env1ronmenta1 orces, an t~e1r mteractions. It follows that pre-
. ly how far back we must reach 1s not yet known. If we were t fi d
c1se h b. I . . fl o n ,
for instance, t at 10 og1ca1m uences on politics operate exclusively via
personality, then I ~ould contend t~at students of politics could fare well
without incor~ora~mg proc~sses pnor ~o pers?nality effects in our ernpir-
ical accounts; m th1s scenano, personahty variables would be sufficient to
capture the influence of deep-rooted forces. Conversely, if, as seerns Iikely,
personality ~s. found to be only one _of a set of rnechanisrns linking biol-
ogy and poht1cs, then a cornprehens1ve treatrnent of biological influences
dearly would require data on personality's fellow rnechanisrns.
A related issue concerns the appropriate division of labor arnong
scholarly disciplines. For instance, the irnpact of biology on personality
2

dearly is relevant for political scientists, yet ernpirical exarnination of the


biology-personality link arguably should be left to behavioral geneticists
and trait psychologists. Sorne line has to be drawn. After all, absent sorne
reasonable division of labor, we fall prey to an infinite regress. Genes
matter for personality, but what factors lead to variance in genetics?
Prenatal exposure to certain chemicals may influence genes, but what
factors cause prenatal exposure to those chernicals? The cornposition of
local soil partly determines environmental exposure to certain chernicals,
but what affected the composition of the soil? Continuing on this path,
it would not be long before we found ourselves asking what caused indi-
gestion arnong dinosaurs. But the way to escape this infinite regress is not
to draw an arbitrary line. Instead, where we draw the line must be guided
by the tenets of our theories. If we theorize a biological basis to political
behavior, then our causal chain necessarily must include indicators- such
as measures of personality traits - that capture biological forces and that
differentiate those forces from environmental influences. But we reason-
ably can leave to other disciplines the task of identifying the processes
that gave rise to our predictors. .
The impact of personality on political attitudes suggests a need to rethmk
other_ long-standing theories of political behavior. For instanc~, wh~t
<loes it imply for rational-actor perspectives if, due to differences m therr

I thank Paul Quirk for suggesting this point.


2

189
·ty and the Foundations of Political Behavior
Persona lt
.. 'dentically situated individuals form disparate .
onahties two l f 1· . 1 po1icy
pers ' h ose different courses o po 1t1ca engagement? 1f var·
P references or e o . f •anee
.in persona . .. t its influences behav1ors or pre erences, can we reco .
11ty ra f .. • • . nc11e
. 'th d't'onal conceptions o utl 1
1ty max1m1zat1on? Suppose f
this w1 tra 1 1 bl' • , or
o neighbors, one a trem 1ng 1ntrovert and the oth
examp1e, t hat tw . k . . . era
. extravert hold ident1cal sta es 1n an upcom1ng c1ty council di·
b01sterous , fi d' · h s-
cmmooo . fa zoni'ng ordinance. Based on n 1ngs 1n C apter 6 (and she
common sense) , we surely would expect. that .the extraverted neighbo r
ld be more likely to attend the counc11meeting, yet an exclusive focus
::;conomic self-interest seemingly would be silent on this matter.
Partisanship offers a related test case. The "running tally" view of par-
tisanship (e.g., Downs 1957; Fiorina 1981} casts partisan preference as
the fluid outcome of a person's ongoing assessment of party performance.
This perspective admits no enduring basis to partisan attachment. Under
such a framework, why would partisan affiliation vary as a function of
stable personality traits such as openness to experience, conscientious-
ness, and agreeableness, as was found in Chapter 5? Following these
models, people with identical utility functions should not differ in par-
tisan preference merely because one person is more conscientious than
another, yet we have seen personality effects of precisely this sort.
Personality traits logically serve as sources of stability for political
judgments. For instance, a high level of conscientiousness should act in a
consistent, persistent manner to nudge a person toward ideological con-
servatism. Personality similarly may constitute a stabilizing influence on
assessments of policy. Since at least the time of Converse (19 64 ), prefer-
ence stabil!ty has been championed as a desirable property in citizens. 3
But adulat~on of attitudinal stability presupposes that that stability arises
from ~ell-mformed deliberation, not from a biologically influenced psy-
cholog.ical te~dency. Personality exerts strong influence on political prefer-
~:ces, mcludmg, as was reported in Chapter 5 a noteworthy influence 00
1
heology. ~f t~is effect contributes to attitudi;al stability as seems likely,
t en the . significance convent1ona · 1ly attnbuted
. ' · · must be
to such stab1hty
recons1dered A p ·h · ce
· erson wit a very high level of openness to experien
an d a very low level f · • · be a
staunch l'b 1 1 b . o consc1ent1ousness may persistently cla1m to
gest that sterb~l? ~t it ~u r ely would require a definitional stretch to sug-
1
a 1ty m th1s · 'fi
To sorne ext h case sigm es sophistication. .0
which exclu . elnt, t ~se matters merely represent concrete instances t
s1ve y enviran 1 ounts
of political beh . R menta perspectives offer incomplete acc e
av1or. un · ll í ce ar
purely environme . ~mg ta Y models of partisan pre1eren r¡
nta1' treatmg actors as blank slates who calculate par
3 F
or a recent example of resea . . aod
Snyder (2008). rch m this tradition, see Ansolabehere, Rodden,
The Mu/tiple Bases of Política/ Behavior
ts after compiling and tabulating evidence of ,
assesfs~luernes. Adding the impact of personality to the nux·a rPeartysult s _successes
and ai h nsive and rea ¡·IStic · d epiction
· · o f partisan
. choice As s mha more
mpre e d b e avior, any unpact of personality. . torw. .at .con-
h . . -
co 5 lf-intereste
sritut~s p\es that there is nota single right answer for all
ment 1mic definitions o f rationa
O
in:
~od ItilcaUJuddg-
. 1·Ity, we may struggle to expJa,...h
VI . ua s n er
·
1
·

econom · d" ·d l h ld fu d u i w Y two


. ly similar m 1vI ua s o n amentally different polit" .
seemmg . l. dh . 1ca VIews
h our hypothetica mtrovert staye orne while the extravert ·d1 .gh'
or w Y . . . B e nei -
bor attended the c1~ counc11meetmg: ut onc~ we factor in personality, a
d~a ent interpretation of what const1tutes rat1onal behavior may eme ,
mer . h e e . . . rge.
WT. may recogmze t at a prererence 1or 1 Imtted, caut1ous policymakin •
. ht given one person 's personar1ty, w h"l
we 1 e a preference for more aggressive g IS
ng emmental action is right for the othei; justas a quiet evening at home
::kes more sense for ~he introvert than for the extravert.
These various quest1ons of theory clearly are not all answerable as yet.
1 have called attention to them to provide a sense of what a personality
approach implies for alternate perspectives, and to suggest areas in which
refinements in theory may be needed. For what it is worth, my own ten-
dency is to see different scholarly points of view as potential complements,
not as competitors, and such is the case here. Although I believe research
00
political behavior must account for fundamental psychological differ-
ences, Ido not advocate a focus on personality as a replacement for prior
approaches. Instead, as I have emphasized throughout this chapter, I believe
that simultaneous attention to stable psychological factors, environmental
factors, and interactions between the two is necessary if we are to develop
a comprehensive treatment of the multiple bases of political behavior.

THE EMPIRICAL STUDY OF PERSONALITY AND POLITICS

Issues in measurement are perhaps more mundane than questions of the-


ory, but issues in measurement still must be addressed. I strongly advocate
that research on political behavior include a focus on personality traits,
and I believe that the Big Five presently affords our best opportunity to
pursue this objective. However, I am less wedded to particular indicators
0
f_the Big Five. The good news on this front is that data gather~d via
umvocal, bipolar, and questionnaire formats consistently yield evidenc_e
?f five-factor trait structure. In an important sense, it follows that tt
is difficult to go wrong when selecting Big Five measures from _among
th0
se already tested by trait psychologists, particularly if it is posSible for
respondents to complete large, multi-item batteries. 4

For a g°0 d, recent d1scuss1on


lohn . . of the propert1es
. of ~anous
• B"1g Five measures, see
'Naumann, and Soto (2008, 130-8).

191
Persona lt·ty and the Foundations ol Political Behavíor
. . . applications of the Big Five proceed, two methodolo .
As po1inca1 . . p·· . I h .. h · •. g1c,
. .11 . •re further attent1on. irst, a t oug 1t 1s true rhat muI . 1

1 fu . . 1 . . ttp1e
l

issues w1 reqm
. be used to deve op ncttona measures, 1t 18 not . .
item formats can . p·. h b· 'd ifi e1ea
'd 1b • f measures of the B1g 1v_e ave een 1 ent ed. As w h
that 1 ea ne . . h b . b! .. e av(
. proach on th1s pomt as een to use 1polar ttems a d
seen my own ap . f 1 ' n 'a1
. ' h
a simp1e c ec
k on the robustness
.
o resu
Thts, to vary
. sorne of the part·
1cu1a1
'
. . f . m one survey to the next. e two-1tem scales developed h
item pa1rs . ro b 'd f h b . f . ere
. d well, and the same can e sa1 o t e ne B1g Five meas ures
funct1one
sling Rentfrow, and Swann (2003), Woods and Ham·p
reported by Go , . h . son
. nd Rammstedt and John (2007). St1 11 , t ere 1s a clear differe
(2005 )'a 'bl "W' h f h nce
between "good enough" and "best poss1 e. 1t. urt er research on mea-
s rement and especially with direct tests of vanous competing measures it
º be' possible to identify the umque
should · attn'b utes of our best contenders, '
and ultimately to propose a preferred battery for use on future surveys.
Validity should be the chief concern, but we also must be attentive to how
much survey time is required to obtain Big Five data and whether inclusion
of trait items adversely affects interview rapport. Ideally, our long-term
objective should be for a common trait battery to gain regular use on a
wide array of surveys and experiments. Once this goal is met, it will be
possible for scholars to incorporate basic psychological differences in their
empirical research as a matter of routine, facilitating ongoing attention to
the multiple bases of political behavior. As this research takes place, use of
a well-established common trait battery will help the work to be cumula-
tive and mutually reinforcing, thus hopefully avoiding a return to the idio-
syncratic, piecemeal approach rightly criticized by Sniderman (1975).
A second, and more complicated, challenge will be in the development
of high-quality Big Five indicators for use in cross-national research. As
noted in Chapter 2, psychologists have reported a great <leal of evidence
regarding the universality of the five-factor structure. Big Five batteries
have been translated into multiple languages, and the scales have per-
formed w~ll. Hence, working within any given country, it should not be
p~oblemat1c for political scientists to devise good indicators of the Big
~~ve. ~o~ever, it is a considerably more difficult matter to represent the
ig Five m a comparable manner across nations and especially across
f
language~. The development of good indicators or use in comparative
~esearc~ is an age-old problem, and there is no need to belabor the point
elreb. Iht 18 ~uf~cient to note that, as in all cross-national research on políti-
ca
. e av1or:' 1t will be important
· for analysts to be particularly cautIO · us
m matters concerning measurement.

PERSONALITY ANO POLITICS


A deepening of our und d. d as
the ultimate goal of erSran mg of citizens and politics sta~ s A
research on personality and political behav10r. s
192
The Multiple Bases of Political Behaviar

. d throughout this book, I believe that a m h . h
s1ze . . f h d. . . uc ne er and
JllPba preciatton o t e urt erpmmngs of pol'f 1 b . more
eflaaoced ap we incorporate both innate and environ l ica l fehav1or will
e once . . . . menta orces 1
efllerg . interact10ns, m our accounts. As th1s research ' a ong
with thetrf specific substantive questions can be pursupdrocAeeds,.a great
ber O f lose research avenues no· e · t th1s ea r1Y
11uJll I ne1•rher wish to oree . · r to suggest h
stage, . s are more pressmg than others. Hence 1·n 11. . t at
ropic d . ' ca tng attenf
soJlle ossible issues, I o so w1th the intent of providing· ion
feW P . · f ·examp1es of
to a . of substantive quest1ons uture research on personal'ty . d
he sorts . l h h . i an pol-
~. s rnight pursue, not to imp Y t at t ese particular questions should be
it1C d d any privileged status.
awar e .d . h h . fl uence of personali'ty
one 1·ssue to cons1 er 1s w. et er any .
m . · on pat-
. civic engagement y1elds an madvertent partisan or ideol . · . 1
terns in . . d' I h . og1ca
unbalance in pohttc~1 1sc?urse. n t e pnor th~ee chapters, we observed
rous instances m wh1ch openness to expenence produced a •positive
nurne .h . . d d .
t and in wh1c consc1ent1ousness pro uce a negat1ve effect in mod-
effec f f .. .
els regarding many .ª~ets º. c1v1~ engag~~ent. Th1s pattern was seen in
ts concerning pohttcal d1scuss1on, poht1cal knowledge, opinionation
tes
. terest in politics, effi cacy, an d work'mg m · campaigns. In several addi-'
:onal cases, either a significa~t ~ositive effect of openness or a signifi-
cant negative effect of consC1ent1ousness was observed. In themselves,
these results establish that personality matters for civic engagement, but
the findings carry no overt political importance. However, these findings
are placed in a new light when we recall from Chapter 5 that openness
predicts ideological liberalism and a Democratic partisan affiliation, and
that conscientiousness predicts the opposite. Putting these results together,
the possibility emerges that liberal Democratic voices speak more loudly
in the American political scene than do conservative Republican voices,
purely dueto what may be nothing more than a coincidental correspon-
dence in the traits that influence political predispositions and those that
influence engagement.s
Future research might examine the magnitude of any such personality-
based effect in an effort to gauge its aggregate strength. 6 For any given
individual, the effect may well be modest, but even marginal individual-
level influences can take on considerable systematic importance when
aggregated across tens of millions of citizens. Also, further research could
find that the effect suggested here is transient, linked somehow to politi-
cal developments at the time this book's data were gathered rather th ªn
to more lasting features of the American political system. Lastly, it shoul~
be clear that if there is a personality-based tilting of the playing field, th15

: ~y t~anks to John Zaller for suggesting this possibility.


van~nt of the simulation techniques used by Althaus (2003) and Bartels (I99 6 ) to
study info .
rmat1on effects could be employed.

1 93
Personalt·ty and the Foundations of Political Behav ior
·
11 be offset by differences in mobilization efforts 1'h
effecr_mayfinwde h t any advantage for liberals and Democrats w · bat is,
we rrught t a . l . . h e o se
.•cy is studied in 1so at1on vams es, or even rever
hen persona l1
l'Ve
. h . ses, wh
w . 1factors are added to t e m1x. en
env1ronmenta h fu
substantive concern t at may warrant rther att .
A secon d .
. . h ·nterplay between t h e persona¡·1t1es
· o f c1tizens
.. _ a dent1on
mvo1ves t e 1 _ A • · d n thos
·¿ tes for elected office.7 s JUSt re1terate , we have s
0 f can d1 a
e
. . d een tha
. to experience and consc1ent10usness pro uce opposite eff .t
openness . f M ._ . ects 10
ideology and parttsan pre erence. y assumpt1on In inte
terms of h . d' . rpret-
. th se findings has been that t ese tra1t 1mens1ons correspond . h
mg e · h · b l'b wu
characteristic views regardmg w at it means_to e I . eral or conserva.
t1·ve and what it means to be a Democrat ora Repubhcan. For instan
, . . d d ce,
people who are high in _consc1ent10usness ten towar _ca~t~on and per-
sonal responsibility, attnbutes that should attract these md1v1duals to th
principies of small government. But an alternare possibility is that peopl:
value the traits they themselves possess, seek out those traits in others
and find them in fellow partisans. 8 '

I have long been interested in the influence of candidate character


on candidate evaluation (e.g., McCurley and Mondak 1995; Mondak
199 5a). Previously, I have focused on candidate traits such as competence
and integrity, traits that all voters presumably value. But if, as posited
here, there is also a personality connection, it might be that voters differ
in the personality traits they find desirable in candidates. All voters should
prefer candidates who are honest, but all voters will not necessarily pre-
fer candidates who are extraverts. The person who is open to experience
may be averse to dogmatic politicians, and the person who is careless or
dishevelled may be unimpressed by a candidate's conscientiousness.
If there is such a personality-based affinity between voters and candi-
dates - and at this point, definitive demonstration of the existence of this
link requires further research - it would suggest two noteworthy implica-
tions. First, an impact of personality on candidate preference generally
would have the effect of reinforcing partisan and ideological preferences,
but for a reason unrelated in any direct manner to policy. In the extreme
case, if a conscientious individual votes solely on the basis of which can-
didate is the most responsible, then the fact that both the voter and the
candidate are Republicans would be a coincidence. But second, because

lmportant first steps toward studying this matter have been taken by Caprara and
7

colleagues (see especially Caprara, Barbaranelli, and Zimbardo 2002; Caprara et al.
2003) .
8
r
A elated, but somewhat different, possibility is that voters derive ideological infer-
e~ce~ from assessments of candidates' traits. For instance if the candidate appears
bit dis~evelled, might a voter equate this with a low mark on conscientiousness, an
th
us with the assumption that the candidate is not a conservative?

1 94
The Mu/tiple Bases of Political Behavior

.
.s no t set in stone that. all. Democrats
'd are open to experience or thata ll
1t 1 bl'cans are consc1ent10us, ev1 ence that voters seek d'd
Repu I I . 1· . Id out can 1 ates
·milar to themse ves m p~rson~ tty _wou ~uggest a campaign strategy for
si d'dates from the poht1cal mmonty. For mstance Democraf .
can I bl' d' · · h . .. ' 1c nommees
. stro ngly Repu b1can 1stnctsh . . m1gh t. 1mprove their prospects, even
. I'f on1y
10 margins, y emp as1zmg t e1r work ethic and sense f
on the b . . f h . o persona1
ponsibility, and y om1ttmg rom t e1r personal histories any menti
d'their surling accolades, skill at cooking Tibetan cuisine, and experien;e~
street performer. In short, cand1dates may find that the person l'ty
as a II . d' . a1
artributes that go over "':e m one istnct may not in another, and that
by highlighting those tra1ts the voters find to be most desirable, the candi-
dates can make i~roads even where partis~n forces are unfavorable.
The third poss1ble research stream I w1ll note concerns cross-national
work on political behavior, and especially on politics and culture. Students
of comparative political behavior have long been intrigued by the sig-
nificance of political culture, with Ronald Inglehart's research achieving
particular prominence (e.~., lnglehart 1988, 1~89)._Recent contributions
in chis area suggest a poss1ble role for personahty. F1rst, sorne works have
commented on a possible interplay between genetic and cultural influ-
ences (e.g., lnglehart and Klingemann 2003). Second, in a very interesting
discussion, Hofstede and McCrae (2004) report and analyze correlations
between the Big Five and Hofstede's cultural dimensions. Hofstede and
McCrae posit numerous possible accounts for these relationships, includ-
ing that the distribution of genetically influenced personality traits may
vary by nation, and thus that "there are innate temperamental differences
between ethnic populations that give rise to cultural differences" (Hofstede
and McCrae 2004, 77). Hofstede and McCrae rightly note that extant
data do not support conclusive statements regarding the causal connec-
tion between personality and culture, but the authors also emphasize the
importance of future research on this matter. As this work proceeds, it
may be of particular value for students of comparative political behavior
if it provides new insight on the origins, meaning, and significance of cul-
ture. Hence, simultaneous attention to personality and political culture
exemplifies well the promise I see in incorporating fundamental psycho- 9
logical differences in our substantive examinations of political behavior.
st
The examples of possible research streams outlined here sug?~
two general lessons about personality and politics. First, o~p?rt~mties
10
abound for dramatic breakthroughs in the depth and sophi 5ticat n. of
60th our theories and our substantive explanations of political behavwr.

9
As th is research proceeds, it will be important for scholars to be attentive t_o issues in
cross-level inference as work moves between an individual-level and ª natwnal-level
focus (e.g., Seligson 2002).

1 95
Personality and the Foundations of Political B e havzor
.
Second, achieving these breakthroughs . d will require con s1•d erabl
l
however. Research on persona 1ty an mass politics _ ¡ d e effort
on ali innate m · fl uences on po 1·1t1ca
. l beh av1or
· - remainn .eed ' research'
state. Not~worthy inr~ads in the study of genetics and po~i;:s~ fledgling
accumulatmg at a rap1d pace, and the present effort to call . av~ been
the significance of personality hopefully has succeeded in ~~t~n~ion to
the corresponding importance of psychological differences Nig hghting
the combmed· bo d y o f extant researc h m· t h ese areas 1s
· minute
· onethei
1 . ess,
what can and should be achieved with further effort. For thi·s research re ative to
realize its full potential, scholars must embrace a new perspecti to
spective that fully acknowledges the significance of the multipleveb, ª Per-
.. . ases of
poht1cal behav1or.

You might also like