Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

HIGLEY, John BURTON, Michael FIELD, G. in Defense of Elite Theory PDF

Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

In Defense of Elite Theory: A Reply to Cammack

Author(s): John Higley, Michael G. Burton and G. Lowell Field


Source: American Sociological Review, Vol. 55, No. 3 (Jun., 1990), pp. 421-426
Published by: American Sociological Association
Stable URL: http://www.jstor.org/stable/2095766 .
Accessed: 19/02/2015 14:09
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

American Sociological Association is collaborating with JSTOR to digitize, preserve and extend access to
American Sociological Review.

http://www.jstor.org

This content downloaded from 200.130.19.138 on Thu, 19 Feb 2015 14:09:30 PM


All use subject to JSTOR Terms and Conditions

IN DEFENSE OF ELITE THEORY:


A REPLY TO CAMMACK

JOHN HIGLEY

MICHAEL G. BURTON

Universityof Texas at Austin

Loyola College in Maryland

G. LOWELLFIELD
Universityof Connecticut

Cammack's (1990) attackon "thenew elite


paradigm"provides a welcome opportunity to engage his criticisms and clarify our
versionof elite theory.At bottom,we aretrying
to pin down the elite structures, outlooks,
choices, and actionsthatmay accountfor gross
disparities among the political records of societies that in economic, class, cultural, and
otherrespectsappearremarkablysimilar.Looking firstat Europeansocieties afterthe sixteenth
century, we ask why the political records of
Britain and its English-speaking offshoots,
Sweden, the Netherlands,and Switzerland,are
from early dates marked by more gradual,
peaceful change thanotherEuropeansocieties,
and why politics in most of the other societies
become benign only after World War II. Second, we ask why the political recordsof Latin
Americansocieties areuniformlycharacterized
by instabilityandviolence duringthe nineteenth
century, and why a handful of them establish
stable, increasinglydemocraticregimes in this
century,while othersdo not. Third,among the
many developingcountriesthatachieved independenceafterWorldWarII, we ask why countries so similarlysituatedas IndiaandPakistan,
Malaysia and Indonesia,Senegal and Nigeria,
Botswanaand Ghana,have strikinglydifferent
politicalrecords.Fourth,we ask why the major
revolutionsandcounterrevolutionsof the modern periodhad such diverse political outcomes
even thoughtheircauses and processes in most
cases had a strongfamily resemblance.Finally,
we ask why politics in the prosperous,globally-dominant,post-industrialsocietiesof northwesternEurope,NorthAmerica,and Australasia swing from a placidphase duringthe 1950s
and early 1960s to a much more fragmented
and turbulentphase duringthe late 1960s and
1970s, only to reach what today looks like
immobility in the face of profoundsocial and
economic problems.

We believe that cogent and parsimonious


answers to these questions can be found in a
modified version of elite theory that ties the
diverse patternsof regimes and their broader
political context to continuities and transformationsof nationalelites. We do not claim that
elites alone runthe political show, or thatsome
simple,readilyfalsifiableexplanationthatholds
for political change in all places and times can
be derivedfrom elite (or any other)theory.We
argueonly thatelite theoryilluminatesthe flow
of modem political history and contemporary
events betterthancompetingtheories, and that
by synthesizingelementsof elite andclass theories, it may be possible to build a more powerful theory. In his attack, Cammack gives the
impressionthathe is strikingat the heartof our
position. But his criticisms are based on a partial readingand caricatureof our work. He ignores our attemptedsynthesis of elite and class
theory,andhe says nothingaboutthe closeness
or remoteness of our theoretical stance as a
whole to modem political history. Instead,he
concentrateson methodological issues in the
nexus between elites and regimes, and his only
notice of historical or contemporarypolitical
change comes in a few remarksaboutcompeting interpretationsof Englishpolitics in the late
seventeenthcentury.
Cammackfirst takes issue with our definition of elites, calling it too permissivebecause
it ostensibly includes, in his words, "leadersof
movements that may lack any permanentorganizational structure,"thereby making any
"fully consensual elite . . . unstable by definition."This is incorrect.Startingwith an extended discussion of the subject 20 years ago
(Field and Higley 1973), we have consistently
followed Weber and Michels in conceiving of
elites as rooted in bureaucraticorganization.
Movement leaders are elites only to the extent
that the movements are bureaucraticallystruc-

AmericanSociological Review, 1990,Vol. 55 (June:421-426)

This content downloaded from 200.130.19.138 on Thu, 19 Feb 2015 14:09:30 PM


All use subject to JSTOR Terms and Conditions

421

AMERICANSOCIOLOGICALREVIEW

422
tured and thus powerful on a sustainedbasis.
This alone enables such leadersto affect political outcomesregularlyand substantially,as our
definitionof elites stipulates.We do not deny
thatunstructuredand more or less spontaneous
popularprotestsanduprisingsoccasionallyhave
majorpolitical effects, but unless they take on
bureaucratictrappingsand articulatea distinct
leadershipcomponentthey usually dissipateor
are promptly suppressed. Whether relations
between movement leaders, in our sense, and
otherelites are consensual is thereforean empiricalquestion.
Threeadditionalpoints aboutdefining elites
need mentioning.First, in our empiricalinvestigations of national elites, none of which is
mentionedby Cammack,we have, with various collaborators,applied the concept restrictively, not permissively, in order to focus on
the few thousanduppermostleaders of major
organizationsand movements in societies like
theU.S., Australia,Norway,andWest Germany
(Higley, Field, and Groholt1976; Higley, Deacon, and Smart 1979; Higley and Moore 1981;
Higley, Hoffmann-Lange, Kadushin, and
Moore, unpublished). Second, problems in
defining and identifying elites exist, but they
are minor comparedwith the problems of defining and applying competing concepts like
"rulingclass,""powernetworks,"or "thestate."
As Marcus(1983, p. 25) pointsout, elite theory
has the great advantageof being the only macrotheoryin political sociology thatoperateson
a "small-group,personallevel of conceptualization,"focusing on small numbersof people
who can plausiblybe viewed as sources of politicalcontinuityandchange,who possess some
importantdegree of internalorganization,and
who can often be described in considerable
detail. Third,by adoptinga Weberian-Michelsian conception of elites, we draw attentionto
the organizationalconstraintson elite behavior. The need to maintaincontrolof the organizations that give them power shapes their actions and outlooks in many ways, frequently
overridingthe substantiveinterests and vague
ideological principles on which Cammack
places so much emphasis, and often impeding
their ability to enter into the settlements and
othertransformationsfrom disunity that Cammack thinks should be routine if elites generally gain from such maneuvers.
Cammacknext takes us to task for positing
allegedly circularrelationshipsbetween types
of elites and patternsof regime stability/insta-

bility because our indicatorsof these patterns


are not independentof the causally prior elite
types (see Field and Higley 1985; Higley and
Burton 1989). This objection is difficult to
grasp. Cammack does not quarrel with our
specificationof regime instabilityas indicated
by irregular seizures, attempted seizures, or
widely expectedseizuresof governmentexecutive power by force. He adduces no cases for
which this specificationis inadequate,nor does
he suggest any other specification. Instead,he
assertsthatin discussingthreepotentiallyproblematic cases out of a hundredor more - the
FrenchThirdRepublic,andUruguayandChile
duringthe presentcentury- we fail to separate elite and regime characteristicsenough to
avoid circularity.In fact, we use those cases to
show how importantit is to make such a separation. We recall that contemporaryobservers
of the three regimes almost unanimously regarded them as stable, whereas by one of our
indicatorsthey were clearly unstable:Expectations of power seizures were widespreadand
were actuallyplottedon a numberof occasions.
Our larger point is that to avoid mistaken
assessments of such regimes, analysts should
first apply our indicators of instability rigorously, and if they are still in doubt,they should
look more closely thanhas been customaryfor
evidence thatelites thatwere clearly disunified
prior to the regime and period in question French elite struggles during the 1850s/1860s
andearlier,the Uruguayancivil warof 1903-4,
the Ibanez dictatorship in Chile prior to its
overthrow in 1932 - had achieved unity.
Analysts should (1) scrutinizea country's political recordfor one of the transformationsto
unity that we specify and that in theory would
lay the basis for regime stability;and (2) look
for the attitudesof elites towardbasic rules of
the game and the elite network configurations
that we specify as characteristicof a unified
nationalelite. Lackingevidence of these events
and patterns,analysts should presumethat the
earliermanifestdisunityof elites continues,and
thatthe regimein questionis basicallyunstable.
More generally,Cammackwants us to abandon our distinction between disunified and
consensuallyunified nationalelites because, in
his view, elite structureand functioning is so
elastic that it defies typological treatmentand,
therefore,any relatedspecificationof transformations in elite types. The unity of national
elites ebbs and flows along a continuum, he
tells us, with elite groups sometimes becoming

This content downloaded from 200.130.19.138 on Thu, 19 Feb 2015 14:09:30 PM


All use subject to JSTOR Terms and Conditions

423

ELITETHEORYDEFENDED
"imperceptiblycloser"only laterto "driftapart."
Thus, our distinction between disunified and
consensually unified elites is excessively "polarized"and "disproportionate"
to the regime
patternswe seek to explain. We have several
responses to this criticism.
First,in allegingthatwe "polarize"elites into
consensually unified and disunified types,
Cammackforgetsthatwe also distinguisha third
elite, the ideologicallyunifiedtype.If one wants
to thinkin termsof a continuum,ideologically
unified elites would clusterat one end, consensually unifiedelites would be arrayedalong the
middle, and disunified elites would cluster at
the other end. In a recent paper, we invoked
Sartori's (1976) discussion of typologies to
point out that our distinctionsconstitute ideal
or pure types that "representthe standards,
parameters,or models against which ... concrete instances can be compared in terms of
greateror lesser proximity"(p. 145). We wrote:
"Inreality,nationalelites appearto vary along
a continuum while nevertheless clustering
around three fundamentally distinct forms"
(Field, Higley, and Burton 1989, pp. 13-14),
andwe referredto the clustersthata numberof
contemporarynationalelites form.
Second, it is easy to assert that the relative
unityof nationalelites is a continuousvariable,
but since neitherCammacknor anyoneelse has
produceda continuousmeasure of elite unity/
disunity,the assertionleads nowhere theoretically; indeed, without typologies and the classifications they engenderthere can be no theory. Third,our distinctionscorrespondclosely
to the typological work done by other students
of elites (Burton and Higley 1987a). Finally,
we (along with various collaborators) have
conducted several empirical investigations of
national elites that appearto be consensually
unified, andmuch of this work is aimed at providing the criteriaand referentsby which such
elites can be identified. But, again, Cammack
ignores these investigations.
Today it is nearlyaxiomaticin political sociology that a key featureof stable democracies
is substantialconsensus and accommodation
amongelites on rules of the political game and
the worth of political institutions.In Sartori's
(1987) recent formulation, stable democracy
requires that elites perceive politics as "bargaining" ratherthan "war"and that they see
outcomes as positive- not zero-sum (pp. 2246). Whatdivides the field is how this elite conditioncomes aboutandhow it is sustained.Is it

producedby a democraticpolitical culture,by


the workings of democraticinstitutions,or by
class hegemony resulting from a "re-ordering
of class forces," as Cammackphrases it? Our
answeris "noneof the above."We contendthat
stable regimes that operate according to principles of representative politics, which may
evolve into "democracies"in the modem sense
if other conditions are favorable, originate in
consensual, and thus voluntary,unificationsof
the most powerfulelite factions aroundrules of
the game and new or existing institutions.Often such unifications occur in the process of
gaining national independence.But in a number of crucial cases, they involve sudden or
otherwise clear-cut elite transformationsfrom
endemic, long-standingdisunity to consensual
unity. Much of our recent work is aimed at
specifying two kinds of transformations:elite
settlements and elite convergences (or "twostep"transformations).
Cammacktries hard to explode the idea of
elite settlements,claiming that if they occur at
all theyplay a smallrole in largerclass struggles.
Mainly, he caricaturesour analysis. Thus, we
do not say that settlements include all elites;
they are"comprehensive"in the sense thatthey
occur among major factions that previously
were at each other's throats,but they may well
exclude or destroy factions that stand in the
way (e.g., English and Swedish royalists in the
late seventeenthand early nineteenthcenturies
respectively, allies of the deposed Colombian
and Venezuelan dictators in the late 1950s).
Nor do we say that the settlementsin England,
Sweden, and Colombia involved only political
elites, and that the settlement in Venezuela
involved only political and business elites. In
highly condensed treatmentsof the four cases,
we concentrateon the most visible actors, although, depending on the country, military,
church,tradeunion,media,andotherelites were
partiesto the settlements.Similarly,we do not
claim thatelite autonomyfrom mass pressures
is the sole or main cause of elite settlements.
Instead we emphasize legacies of costly but
inconclusive conflicts, new crises which
threaten the renewal of those conflicts, and
exceptionalnegotiatingcircumstancesthatseem
necessary in fashioning settlements.However,
we do not presume to "explain"elite settlements,which Cammackrepeatedlysays we fail
to do. Settlementsare highly contingentevents
that depend on such factors as the skills and
choices of elite persons who happen to be in

This content downloaded from 200.130.19.138 on Thu, 19 Feb 2015 14:09:30 PM


All use subject to JSTOR Terms and Conditions

424

AMERICANSOCIOLOGICALREVIEW

place, and they cannotbe fully "explained"in


the usual social scientific sense. Finally, we do
not say thatelites arewithoutsubstantiveinterests. Together with organizationalconstraints
on elite behavior, the existence of strong and
irreconcilableinterests among elites accounts
for the persistence of elite disunity in many
countries over long periods. Nevertheless, in
extraordinarycircumstances,elites sometimes
find ways to compromiseconflicting interests
while maintainingorganizationalsupport,and
the settlement concept capturesthese turning
points in the politics of certaincountries.
England's Glorious Revolution of 1688-89
is the only empiricalcase to which Cammack
refers in his attackon the settlementconcept,
and a brief restatementof how we view elite
change in thatevent is warranted.Long before
and after 1688-89, a limited suffrageconfined
political influence in England to the upper,
largely landed,class. But before 1688-89, that
class was deeply divided between "Tory"supporters of royal power and the established
Anglican Church, and "Whig" advocates of
what amounted to republicanismand greater
religious tolerance. There was no systematic
differencein the economic interestsof the two
factions. During the voluntary and deliberate
Tory-Whig conspiracy that deposed James II
and made England into a "crownedrepublic"
in 1688-89, Tory leaders droppedtheir insistence that kings could not be resisted (much
less deposed),while Whig leadersacceptedthe
privileged status of the Anglican Churchover
otherProtestantsects. Together,these repudiations insulatedeach faction againstattemptsby
the otherto take a principledposition and seek
exclusive dominancein civil conflicts. Neither
factionneededto fearthatthe otherwould push
any political advantageto intolerablelengths.
After 1688-89, in short,the majorelite factions
were consensuallyunified, andthey dominated
subsequent English politics in long stretches
withoutseeking to go muchbeyond the limited
representationof landedwealth. Thus, all subsequentdissidentmovementswere deprivedof
major elite support.A prudentand restrained
politics very different from the divisive, violent politics of earlier periods of English history became the norm, and peaceful transfers
of governmentexecutive power via elections,
which eventually became democratic, took
place periodically.
Like other scholars to whom we referredin
our discussion of this case (see Burton and

Higley 1987b), Cammackbelieves that 168889 was merely the political coda to what had
been a class-conflict symphony, i.e., the English Civil Warsandupheavalsof 1640-60. This
fairlyorthodoxMarxistinterpretation
has drawn
devastatingcriticism(see Goldstone 1983 for a
compelling summary).Our thesis of elite settlement is in step with a growing body of historical scholarshipthat stresses the watershed
characterof the events of 1688-89 (e.g., Jones
1972; Schwoerer 1981; Speck 1988). Full appreciation of our thesis is possible, however,
only when it is placed in comparativecontext,
where it offers a much more plausibleinterpretationthanMarxismfor why it took so long for
most Europeansocieties to achieve the kind of
peaceful conflict managementthat has characterized English politics since 1689, and why
many societies in the world still have not
achieved it.
But let us also dispel the idea that our aim is
to refute class analysis. In some of our work
(e.g., Field and Higley 1980, esp. pp. 21-32;
Field, Higley, and Burton 1989), we hold that
the interdependenceof elites and mass publics
can best be capturedby combining elite and
class theory.In slogan form,ourposition might
be renderedas "Classespush, elites effect" (cf.
Dahrendorf,1988, pp. 53-4). The difficulty is
that classes presumably"push"in ratheruniform ways among similar societies, yet political outcomes are radically different. It is thus
necessary to consider how elites "effect"such
differentoutcomes, and thereinlies the significance of the elite transformationswe examine.
In our investigation of elite convergences
("two-step"transformations),we do not reverse
our method, as Cammackcharges.Analyses of
both settlementsandconvergencesaretriggered
by observations that previously unstable regimes achieved stabilityat some fairly specific
point (e.g., England after 1689, Sweden after
1809, Italy and France duringthe 1970s). According to our theory, such stabilizationof regimes (which also operate according to principles of representative politics) means that
elites achieved substantialconsensus and unity
prior to, or perhapscoterminouswith, stabilization. In concrete cases, the research questions are:Is there evidence of such consensual
unification and, if so, how did it occur? Elite
settlement is one possibility. Another is elite
convergence, in which some of the powerful
but previously warringfactions in a disunified
elite form a winning electoral coalition and,

This content downloaded from 200.130.19.138 on Thu, 19 Feb 2015 14:09:30 PM


All use subject to JSTOR Terms and Conditions

ELITETHEORYDEFENDED
throughrepeatedelectoralvictories,eventually
leave dissidentelites no alternativebut to compete for governmentexecutive powerunderthe
same rules of the game and other tacit understandingsas the winning coalition. This seems
to capturethe processby which French,Italian,
and other nationalelites became consensually
unifiedin recentdecades,andit is, for example,
the process throughwhich South Koreanelites
are apparentlymoving at present. Though we
thinkthatour thesis of elite convergenceneeds
furtherinvestigation,Cammacktakes no issue
with its accuracy in the cases we have examined.
Cammackis right that we have not as yet
spelled out, let alone documented,the complex
ways in which elite consensus and unity,
whether created through settlements, convergences, strugglesfor nationalindependence,or
otherways, are sustainedover time. He is right
to observethatnew elites emerge andold elites
arepushedaside, andthatthereis, in his words,
"a continual process of negotiation among
changingelite groups"withconsensusandunity
waxing and waning. But he is wrong to assert
thatthe rules of the game change in any fundamentalway. The basic rules of the game, summarized as "restrainedpartisanship,"do not
change.Cammackalso goes astrayin imputing
to us the contentionthat elite consensualunity
is an all-or-nothingcondition, so thateither an
entire national elite agrees publicly on everything importantor it is disunified. Our basic
distinctionbetween ideologically unified elites
(which are what Cammack actually seems to
have in mind)andthe consensuallyunifiedtype
avoids just this confusion. The hallmarkof the
former is uniform public utteranceson major
policy questions in conformancewith a single
ideology. Lacking ideological unity, members
of a consensually unified elite regularly and
publiclydisagreeon suchquestions,thoughthey
just as regularlypull their punches shortof the
point where violence and regime instability
would begin. Finally, while much needs to be
done to specify why and how consensually
unified elites persist in various circumstances,
we cannotidentify a single case in the modern
historicalrecord in which a consensually unified elite, once created,failed to make the adjustments necessary to its continuance. Cammack seems equallyunableto find such a case,
while the history of the last 300 years in his
own country, Britain, testifies forcefully not
only to the lasting effects of elite settlements

425
and the permanenceof this kind of elite, but to
the importanceof elites for comprehendingthe
broadsequencesof politicsin themodemworld.
is Professor of Sociology and
Governmentand Director of The EdwardA. Clark
Centerfor Australian Studies at The Universityof
Texas at Austin.

JOHN HIGLEY

MICHAELG. BURTON is Professor of Sociology at

Loyola College in Maryland.


G. LOWELLFIELD is EmeritusProfessor of Political
Science at The Universityof Connecticut.
The authors are writing a book tentativelytitled A
New Elite Frameworkfor Political Analysis.
REFERENCES
Burton,Michael G. and John Higley. 1987a. "Invitationto Elite Theory."Pp. 133-43 in Power Elites
and Organizations,edited by G. William Domhoff and Thomas R. Dye. Newbury Park, CA:
Sage.
Burton,Michael G. and John Higley. 1987b. "Elite
Settlements." American Sociological Review
52:295-307.
Dahrendorf,Ralf. 1988. The Modern Social Conflict. New York:Wiedenfeld & Nicholson.
Field, G. Lowell and John Higley. 1973. Elites and
Non-Elites: The Possibilities and their Side Effects. Andover, MA: Warner Modular Publications.
. 1980. Elitism. Boston: Routledge

&

Kegan Paul.
. 1985. "National Elites and Political Sta-

bility."Pp. 1-44 in Research in Politics and Society: Studies of the-Structure of National Elite
Groups, Vol. 1, edited by Gwen Moore.
Greenwich,CT: JAI Press.
Field, G. Lowell, John Higley, and Michael G.
Burton. 1989. "A New Elite Frameworkfor Political Sociology." Paper presented to the
Roundtableon "TheNew Elite Paradigm:Opposing Views," American Sociological Association
annualmeetings, San Francisco.
Goldstone, Jack A. 1983. "CapitalistOrigins of the
English Revolution:Chasinga Chimera."Theory
and Society 12:123-42.
Higley, John and Michael G. Burton. 1989. "The
Elite Variable in Democratic Transitions and
Breakdowns." American Sociological Review
54:17-32.
Higley, John,Desley Deacon, and Don Smart.1979.
Elites In Australia. London:Routledge & Kegan
Paul.
Higley, John, G. Lowell Field, and Knut Groholt.
1976. Elite Structureand Ideology:A Theorywith
Applications to Norway. New York: Columbia

This content downloaded from 200.130.19.138 on Thu, 19 Feb 2015 14:09:30 PM


All use subject to JSTOR Terms and Conditions

426

AMERICANSOCIOLOGICALREVIEW

UniversityPress.
Higley, John, Ursula Hoffmann-Lange, Charles
Kadushin,andGwen Moore. Unpublished."Elite
Integrationin Stable Democracies:A Reconsideration."
Higley, JohnandGwen Moore. 1981. "EliteIntegration in the UnitedStatesandAustralia."American
Political Science Review 75:581-97.
Jones, J.R. 1972. The Revolution of 1688 in England. New York:W.W. Norton.
Marcus, George E. 1983. "'Elite' as a Concept,
Theory, and Research Tradition." Pp. 3-27 in
Elites: EthnographicIssues, edited by George E.

Marcus. Albuquerque:University of New Mexico Press.


Sartori,Giovanni. 1976. Parties and Party Systems.
New York: CambridgeUniversityPress.
. 1987. The Theory of Democracy Revis-

ited: TheContemporaryDebate, Vol. 1. Chatham,


NJ: ChathamHouse.
Schwoerer,Lois G. 1981. TheDeclaration of Rights,
1689. Baltimore:JohnsHopkinsUniversityPress.
Speck, W.A. 1988. ReluctantRevolutionaries:Englishmen and the Revolutionof 1688. New York:
OxfordUniversityPress.

MANUSCRIPTSFOR THE
ASA ROSE SOCIOLOGYSERIES
Manuscripts(100 to 300 typed pages) are solicited for publica-

tion in the ASA Arnoldand CarolineRose MonographSeries.


The Series welcomes a variety of types of sociological workqualitative or quantitative empirical studies, and theoretical or
methodological treatises. An author should submit three copies
of a manuscriptfor consideration to the Series Editor, Professor
Teresa A. Sullivan, Universityof Texas, PopulationResearch Center, 1800 Main,Austin,TX 78712.

This content downloaded from 200.130.19.138 on Thu, 19 Feb 2015 14:09:30 PM


All use subject to JSTOR Terms and Conditions

You might also like