Recombinant Property in East European Capitalism

Author(s): David Stark

Source: The American Journal of Sociology, Vol. 101, No. 4 (Jan., 1996), pp. 993-1027
Published by: The University of Chicago Press
Stable URL: http://www.jstor.org/stable/2782236 .
Accessed: 12/01/2011 16:13

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .
http://www.jstor.org/action/showPublisher?publisherCode=ucpress. .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

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.

The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to The
American Journal of Sociology.

http://www.jstor.org
Recombinant Property in East European
Capitalism1
David Stark
CornellUniversity

Recombinantproperty is a formoforganizationalhedgingin which

actorsrespondto uncertainty assets,redefining
by diversifying and
recombining resources.It is an attemptto hold resourcesthat can
be justifiedbymorethanone legitimating principle.Propertytrans-
formation in postsocialistHungaryinvolvesthe decentralizedreor-
ganizationof assets and the centralizedmanagementof liabilities.
Togethertheyblurtheboundariesofpublicand private,thebound-
ariesofenterprises, and theboundednessofjustificatory principles.
Enterprise-level fieldresearch,data on the ownershipstructureof
Hungary's220 largestenterprises and banks, and an examination
ofthegovernment's recentdebtconsolidationprogramssuggestthe
emergenceof a distinctively East European capitalismthat will
differas muchfromWestEuropeancapitalismsas do contemporary
East Asian variants.

INTRODUCTION
Sociologybegan as a scienceof transition,foundedat our century'sturn
on studies of the epochal shiftsfromtraditionto modernity,rural to
urbansociety,gemeinschaft to gesellschaft,
feudalismto capitalism,and
mechanicalto organicsolidarity.For thefoundersof sociology,the crisis

1
Researchforthispaperwas conductedwhiletheauthorwas a visitingfellowat the
Institute
forAdvancedStudy/Collegium Budapestand was supportedbygrantsfrom
theNationalScienceFoundation, Reform
IRIS (Institutional andtheInformal Sector),
and theProjecton CorporateGovernance oftheWorldBank/Central EuropeanUni-
versity.My thanksto Luc Boltanski,Ronald Breiger,RogersBrubaker,Laszl6
Bruszt,EllenComisso,Paul DiMaggio,Neil Fligstein,GeoffFougere,Istvin Gabor,
GernotGrabher,Szabolcs Kem6ny,Janos Kornai,JanosLukacs, PeterMurrell,
Laszl6 Neumann,Claus Offe,KentRedding,AkosR6na-Tas,Ivan Szelenyi,Andrew
Walder,PamelaWalters,and especiallyMoniqueDjokic Starkfortheircriticisms of
an earlierdraft.Correspondencemaybe addressedto David Stark,Departmentof
Sociology,UrisHall, CornellUniversity,
Ithaca,New York 14853.
? 1996 by The Universityof Chicago. All rightsreserved.
0002-9602/96/10104-0006$01.50

AJS Volume 101 Number4 (January1996): 993-1027 993

AmericanJournalof Sociology

besettingEuropeansocietiesat theend ofthe 19thcenturywas diagnosed

as a normativeand institutionalvacuum. The old order regulatedby
traditionhad passed, buta new moralorderhad notyetbeen established.
Duringourown finde siecle,notthecrumbling oftraditionalstructures
but the collapse of communismgives new lifeto the transitionproblem-
atic (Nee 1989; Lipset 1990; and see Alexander[1994] foran extended
criticaldiscussion).Withinthatproblematic,thepresentis studiedas an
approximationof a designatedfuture(Blanchard, Froot, and Sachs
1994), riskingan underlyingteleologyin which conceptsare drivenby
hypostasizedend-states.In the frameworkof transitology, the transi-
tionalpresentis a periodof dislocationas societyundergoesthe passage
througha liminalstatesuspendedbetweenone social orderand another
(Bunce and Csanadi 1993),each conceivedas a stable equilibriumorga-
nized arounda coherentand moreor less unitarylogic.
But is ours stillthe centuryof transition?And is thatmodel of social
change, so formativein the launchingof sociology,still adequate for
understanding themomentouschangesin contemporary EasternEurope?
Difficult to assimilatewithinthetransition problematicare thenumer-
ous studiesfromEasternEurope documenting paralleland contradictory
logicsin whichordinarycitizenswerealreadyexperiencing, fora decade
priorto 1989, a social world in which various domains were not inte-
gratedcoherently(Gaibor1979, 1986; Szelenyi 1988; Stark 1986, 1989;
Rona-Tas 1994).2 Through surveyresearchand ethnographicstudies,
researchershave identified a multiplicityof social relationsthat did not
conformto officially prescribedhierarchicalpatterns.These relationsof
reciprocity and marketliketransactionswere widespreadinside the so-
cialistsectoras well as in the "second economy"and stemmedfromthe
contradictions of attempting to "scientifically
manage" an entirenational
economy.At the shop-floor level, shortagesand supplybottlenecksled
to bargainingbetweensupervisorsand informalgroups;at the manage-
rial level, the task of meetingplan targetsrequireda dense networkof
informalties thatcut across enterprises and local organizations;and the
allocativedistortions of centralplanningproducedthe conditionsforthe
predominantly part-timeentrepreneurship of the second economiesthat
differed in scope, densityofnetworkconnections,and conditionsoflegal-
ityacross the region(Gaibor1979; Kornai 1980; Sabel and Stark 1982;
Szelenyi1988).
The existenceof parallel structures(howevercontradictory and frag-
mentary)in these informaland interfirm networksthat "got the job

2 East European scholars have long argued that social change is a transformational
reshapingof enduringstructuresexhibitingmultiplicityratherthan uniformity(Kon-
rad and Szelenyi 1979; Szuics 1985; Staniszkis 1993; Szelenyi 1994).

994
 Symposium:Stark

done" means that the collapse of the formalstructuresof the socialist

regimedoes not resultin an institutional vacuum. Instead, we findthe
persistenceofroutinesand practices,organizationalformsand social ties,
thatcan becomeassets,resources,and thebasis forcrediblecommitments
and coordinatedactionsin the postsocialistperiod(Bourdieu 1990; Nel-
son and Winter1982). In short,in place of disorientation, we findthe
metamorphosisof sub-rosaorganizationalformsand the activationof
preexisting networksof affiliation.
If, by the 1980s, the societiesof Eastern Europe were decidedlynot
sys-tems organizedaround a singlelogic, theyare not likelyin the post-
socialistepoch to become,any moreor less than our own, societieswith
a singlesystemidentity.Change,even fundamentalchange,ofthe social
world is not the passage fromone orderto anotherbut rearrangements
in the patternsof how multipleordersare interwoven.Organizational
innovationin this view is not replacementbut recombination(Schum-
peter1934).
Thus, we examinehow actorsin thepostsocialistcontextare rebuilding
organizationsand institutionsnot on the ruins but with the ruins of
communismas they redeployavailable resourcesin responseto their
immediatepracticaldilemmas. Such a conceptionof path dependence
does not condemnactorsto repetitionor retrogression,3 forit is through
adjustingto new uncertainties by improvising on practicedroutinesthat
new organizationalformsemerge(Nelson and Winter1982; White 1993;
Kogut and Zander 1992; Sabel and Zeitlin 1996). The analysisthat fol-
lows emphasizestheorganizationalreflexivity thatis possiblewhenactors
maneuveracrossa multiplicity oflegitimatingprinciplesand strategically
exploitambiguitiesin the polyphonyof accounts of work, value, and
justice that compose modern society(Boltanski and Thevenot 1991;
White 1992; Stark 1990; Padgettand Ansell 1993; Breiger1995).

A New Type of Mixed Economy?

This articleexaminesthe recombinatory logic of organizationalinnova-
tion in the restructuring
of propertyrelationsin Hungary.It asks, Are
recombinantprocessesresultingin a new typeof mixed economyas a
distinctivelyEast European capitalism?
For more than 30 years, policy analystsin Eastern Europe debated
the "correctmix of plan and market"(Stark and Nee 1989). By the

3 See, by contrast,Burawoyand Krotov'saccountof changeas retrogression:

"Our
case studysuggests
thatwiththewitheringawayofthepartystatetheSovieteconomy,
farfromcollapsingor transforming
itself,has assumedan exaggerated
versionofits
formerself' (1992, p. 34).

995
AmericanJournalof Sociology

mid-1980sin Hungary,thedebatehad shiftedto thecorrectmixof"pub-

lic and privateproperty"as the earliersacrosanctstatus of collective
propertyerodedwiththe growthof the secondeconomy.It was thus,in
thewaningyearsofstatesocialism,thatGaibor(1986) and Szelenyi(1988)
coined the term"socialistmixed economy"to designatethe new eco-
nomicconfiguration.4 Meanwhile, Stark(1989, p. 168), amplifyingGa-
bor's call to acknowledgea mixed economy"as a viable hybridform
and not as inherently unstableand necessarilytransitional,"questioned
nonethelesswhethertheconceptofmixedeconomywas adequate to grasp
the emergentphenomenaof late socialism.On the basis of fieldresearch
on "intrapreneurial" subcontractingunitsin Hungarianfirms,I argued
thataspectsofemergentprivateproperty werenotrespectingthebound-
aries of thesecondeconomybut werebeingfusedwithpublic ownership
insidethesocialistfirmresultingin a "diversification
ofpropertyforms."
Identifying "hybridmixturesof public ownershipand privateinitiative"
(Stark 1989,pp. 167-68) I arguedthat,insteadofa mixedeconomywith
well-boundedpublicand privatesectors,analysisshouldbeginto address
the growingpluralityof "mixedpropertyforms"that transgressedand
blurredtraditionalpropertyboundaries.
Scholarsofeconomicreforms in China subsequentlydevelopedrelated
conceptsto analyze the fiscalreformsreshapingincentivesamong local
governmentsgiving rise to "township and village enterprises."Oi's
(1992) conceptof "local corporatism,"Nee's (1992) "hybridproperty,"
and Cui's (in press) notionof "moebius-stripownership"each illumi-
nated a particularfacetof Chinesepropertyreformsthat supportedthe
generalconclusionthat China's is not a simple mixed economybut a
kaleidoscopeof mixedpublic and privatepropertyforms.
Of special relevanceto my concernsis Walder's (1994) insightthat
property reformshouldnotbe equated withprivatization.Walder argues
that "clarificationof propertyrights"in the Chinese fiscalreformscan
yieldperformance enhancingincentiveseven while maintaining"public
ownership"withoutprivatization.Our analysisof the Hungarian case
also demonstrates that propertytransformation can occur withoutcon-
ventionalprivatization.5 The difference,
however,is thatpropertytrans-

4 Szel6nyi (1978) argued that "mixture"characterizedboth East and West: whereas

a redistributivewelfare state mitigatesinequalities produced by markets under ad-
vanced capitalism, in state socialism subordinated marketlike institutionsmitigate
inequalities produced by the dominant redistributivemechanism. Elsewhere (Stark
1986) I labeled this analytic method "mirroredopposition" and used it to analyze
differencesbetween capitalist and socialist internallabor markets.
5 In her analysis of "political capitalism" in Poland, Staniszkis (1991) similarlyidenti-
fies "hybrid forms" of "undefined dual status" in a variety of leasing forms and
cost-shiftingarrangementsthroughwhich nomenclaturacompanies enjoy the benefits
of propertytransformation withoutprivatization.

996
 Symposium:Stark

formationin Hungarydoes not necessarilyclarifypropertyrights.As

we shall see, the emergingnew propertyformsin Hungaryblur (1) the
boundariesof public and private,(2) the organizationalboundariesof
enterprises,and (3) theboundednessofjustificatory principles.To denote
theseprocessesoftripleboundaryblurringI adopt thetermrecombinant
property.
Recombinantproperty is a formoforganizationalhedging,or portfolio
management,in which actors respondto uncertainty in the organiza-
tionalenvironment by diversifyingtheirassets,redefining and recombin-
ing resources.It is an attemptto hold resourcesthat can be justifiedor
assessed by morethan one standardof measure.
The distinctivevariantof organizationalhedgingthatis recombinant
property in Hungaryis producedin two simultaneousprocesses:Parallel
to the decentralizedreorganizationof assets is the centralizedmanage-
mentof liabilities. On the one hand, decentralizedreorganizationpro-
duces the crisscrossing lines of interenterpriseownershipnetworks;on
theother,debtconsolidationtransforms privatedebtintopublicliability.
Althoughthesetwo dimensionsare discussedseparately,theirsimultane-
itygivesdistinctive shape to Hungarianproperty.The clash ofcompeting
orderingprinciplesproduces organizationaldiversitythat can forma
basis forgreateradaptabilitybut, at the same time,createsacute prob-
lems of accountability.
Data.-My argumentsare based on data collectedduringan 11-month
stayin Budapest in 1993-94. That researchincludes(1) fieldresearchin
six Hungarianenterprises,6 (2) compilationofa data seton theownership
structureof Hungary's200 largestcorporationsand top 25 banks,7and
(3) interviewswithleading actorsin banks, propertyagencies,political
parties,and government ministries.'

are amongthe20 largestfirmsin Hungaryand are at thecore

6 Threeofthesefirms
ofHungarianmanufacturing in metallurgy,electronics,
and rubberproducts.Three
are smalland medium-size firmsin plastics,machining, and industrialengineering.
Thisfieldresearch was conducted incollaborationwithLaszl6Neumannand involved
longitudinal analysisof thesamefirmsin whichwe had earlierstudiedan organiza-
tionalinnovation ofinternal subcontractinginsidethesocialistenterprise
(Stark1986,
1989,1990;Neumann1989).
7 These data wereaugmented by ownership data drawnfromthefilesof some 800
firmsundertheportfolio management oftheStateProperty Agency.
8
A partiallistof interviewees includestheformerpresident of the NationalBank;
theformer deputy-minister of theMinistry of Finance;executivesofthefourlargest
commercial banks and two leadinginvestment banks;the formerpresidentof the
,StateHoldingCorporation; directors,advisors,and officialsof the State Property
Agency;seniorofficials of theWorldBank'sHungarianMission;thechiefeconomic
advisorsofthetwomajorliberalparties;thepresident oftheFederationofHungarian
Trade Unions;and leadingofficials of theHungarianSocialistParty(who lateras-
cendedto high-level positionsin thenew Socialist-Liberalcoalitiongovernment).

997
AmericanJournalof Sociology

PROPERTY TRANSFORMATIONIN HUNGARY:

THE POLICY DEBATE
My pointof departureis a questioncentralto contemporary debates in
EasternEurope and theformerSovietUnion:By whatmeanscan private
property becomethetypicalformofproperty relationsin economiesover-
whelmingly dominatedby stateownershipof productiveassets?
Much of thatdebate can be organizedaroundtwo fundamentalpolicy
strategies.First,the institutionalization of privatepropertycan best be
establishedby transferring assets frompublic to privatehands. Despite
differences in thespecificmethodsdesignatedforsuchprivatization(e.g.,
sale vs. freedistribution,etc.), the variousproposalswithinthisradical
perspectiveshare the assumptionthat the creationof a private sector
beginswiththeexistingstate-ownedenterprises, thatis, thebasic organi-
zational unitsof the emergentmarketeconomywill be the preexisting
but newlyprivatizedenterprises.
The secondpolicystrategyarguesfromtheperspectiveof institutional
evolutionary)
(and specifically, economicsthat,althoughslower,themore
reliableroad to institutionalizing privatepropertyrestsin the develop-
mentof a class of privateproprietors. Instead of transferringthe assets
ofa givenorganizationalunitfromone ownershipformto another,public
policyshouldlowerbarriersto entryforsmall and mediumscale, genu-
inelyprivateventures.This perspectivetypicallylooks to the existing
secondeconomyentrepreneurs as thebasic organizationalbuildingblock
of an emergentmarketeconomy.
RecentevidencesuggeststhatHungaryis adoptingneithera big bang
approach nor the policyprescriptions of evolutionaryeconomics.Con-
traryto theoptimistic scenariosofdomesticpoliticiansand Westernecon-
omistswho foresawa rapid transferof assets fromstate-ownedenter-
prises to privateownership,the overwhelmingbulk of the Hungarian
economyremainsstateproperty.Two yearsafterPrimeMinisterJozsef
Antallconfidently announcedthathis new government would privatize
morethan50% ofstateproperty by 1995,thedirectorofthePrivatization
ResearchInstitutefunctioning alongsidetheStatePropertyAgency(SPA)
estimatedthatonlyabout 3% of the state-ownedproductivecapital has
been privatized(Mellar 1992).Accordingto a recentstudycommissioned
by the World Bank (Pistorand Turkewitz1994), by mid-1994the SPA
had onlysold about 11% of the value of its originalportfolio.
Contraryas wellto thehopesofevolutionary economics,a considerable
bodyof evidencenow suggeststhatthe secondeconomyhas not become
a dynamic,legitimateprivatesector:Althoughthe numberof registered
privateventureshas skyrocketed, manyare "dummyfirms,"tax evasion
is pervasive,and manyentrepreneurs (a majorityin some categories)still

998
 Symposium:Stark

engagein privateventuresonlyas a secondjob (Laky 1992; Gaibor1994,

1996). And althoughemployment is slowlyincreasingin the sector,most
researchers agreethattheproportion work(forwhichthe
ofunregistered
statereceivesno social securitypaymentsand the employeereceivesno
benefits)is increasingfaster(Kornai 1992, p. 13).
These tendenciestogetherwith new formsof corruption,extortion,
and exploitationhave promptedone researcherto label the transitionas
one "fromsecondeconomyto informaleconomy"arguingthatit is now,
underthesenew conditions,thatLatin Americancomparisonsare more
applicable to the Hungariansetting(Sik 1992). When privateentrepre-
neurs look to governmentpolicy,theysee only burdensometaxation,
lack ofcredits,virtuallyno programsto encourageregionalor local devel-
opment,and inordinatedelaysin paymentsforordersdeliveredto public-
sectorfirms(Webster 1992; Kornai 1992). Through violations of tax
codes, off-the-books paymentsto workers,and reluctanceto engage in
capital investment,much of the private sector is respondingin kind
(Gaibor1996). Such government policiesand private-sectorresponsesare
clearlynot a recipeforthe developmentof a legitimateprivatesectoras
a dynamicengineof economicgrowth.

THE DECENTRALIZED REORGANIZATIONOF ASSETS

Althoughtheyfailto correspondto the policyprescriptionsof eitherbig
bang or evolutionary
economics,significant
propertytransformations are
takingplace in Hungary.Since 1989,therehas been an explosionof new
economicunits.In table 1, we see that
1. thenumberof stateenterprises declinedby about 60% fromthe
end of 1988 to the middleof 1994;
2. thenumberofincorporated shareholding companies(reszventar-
sasag or RT) increasedby morethan20-fold(from116 to 2,679);
and
3. the numberof limitedliabilitycompanies(korlatolt felelossegii
tarsasagor KFT) increasedmost dramaticallyfromonly 450
unitsin 1988 to over 79,000 by the middleof 1994.

Table 1 clearlyindicatesthe sudden proliferation of new units in the

Hungarianeconomy.But does thetable providea reliablemap of prop-
ertyrelationsin contemporary Hungary?No, at least not if the data are
forcedintothe dichotomouspublic/private categoriesthatstructurethe
discussionabout propertytransformation in the postsocialistcountries.
As we shall see, actorswithinthe large formerly state firmsare trans-
formingpropertyrelationsat the enterpriselevel. The results,however,

999
 00 )c-
t t- ON

000
~~~~~~~~~~c9
cn

,'0-
0' elOC-

o e *O-o

ON

z~~~~~~z
ON c9^ \t S
-e oo OO\ =

O tC

00~~~~~~
-o oo 00tO 0-CO o0

CIS
PC ~~00 C--~~~CLf) 0
Z '

00 cu~~~
C) ~ ~ ~ ~ )

z 0C--
x

ON ~
CO,

C EE 0 0E

0CO
4 00C C 3 C

00-
-)00)

COd~ 0
 Symposium:Stark

are not well-definedrightsof privateproperty,yet neitherare theya

continuationor reproductionof old formsof stateownership.

New Formsof State Ownership

Take firsttheshareholdingcompanies(RTs) on line 2 of thetable. Some
of these corporationsare privateventuresnewly establishedafterthe
"systemchange." But manyare the legal successorsof the state-owned
enterprisesthat would have been enumeratedin the previousyear on
line 1 of the table. Througha mandatoryprocessof "corporatization,"
theformerstate-ownedenterprise transforms itslegal organizationalform
intoa shareholdingcompany.The question,of course,is who is holding
theshares?In mostofthesecorporatizedfirmsthemajorityof sharesare
held by the State PropertyAgencyor the newlycreatedState Holding
Corporation(AV-Rt). That is, as "public" and "private"actors copar-
ticipatein the new recombinantpropertyforms,the natureand instru-
ments of the "public" dimensionchange: Whereas "state ownership"
in socialismmeantunmediatedand indivisibleownershipby a statemin-
istry(e.g., Ministryof Industry),corporatization in postcommunism en-
tails share ownershipby one or anothergovernment agencyresponsible
forstateproperty.
Such corporatization mandatedbya privatization agencyin thecurrent
contexthas some distinctivefeaturesof renationalization. In the 1980s,
managersin Hungary(and workersin Poland) exercisedde factoproperty
rights.Althoughtheyenjoyedno rightsover disposal of property,they
did exerciserightsof residualcontrolas well as rightsover residualin-
comestreams.In the 1990s,corporatization paradoxicallyinvolvesefforts
by the stateto reclaimtheactual exerciseof thepropertyrightsthathad
devolved to enterprise-levelactors. Ironically,the agencies responsible
forprivatizationare actingas agentsof etatization(Voszka 1992).
The "trapof centralization"alreadywell knownin the region(Bruszt
1988) stands as a warning,however,that the effectiveexerciseof such
centralizedcontrolvariesinverselywiththescope and thedegreeofdirect
intervention.One encounters,therefore,proposals for privatizingthe
asset managementfunction.In such programs,thestateretainstheright
to dispose of propertybut delegatesits rightsas shareholderto private
consultingfirmsand portfoliomanagementteamswho overseedailyoper-
ations and strategicdecisionson a subcontracting or commissionbasis.

Interenterprise
Ownership
The state is seldom,however,the sole shareholderof the corporatized
firms.Who are theothershareholdersoftheRTs enumeratedon line 2 of

1001
AmericanJournalof Sociology

table 1? To answerthisquestion,I compileda data set on theownership

structureof the largest200 Hungariancorporations(rankedby sales).9
These firmscompose the "Top 200" of the 1993 listingof Figyeli, a
leadingHungarianbusinessweekly.Like theirFortune500 counterparts
in the United States, the Figyelo 200 firmsare major players in the
Hungarianeconomyemployingan estimated21% of the labor forceand
accountingfor37% oftotalnetsales and 42% ofexportrevenues(Figyelo
1993). The data also include the top 25 Hungarian banks (ranked by
assets). Ownershipdata were obtained directlyfromthe Hungarian
Courtsof Registrywherecorporatefilescontainnot onlyinformation on
the company'sofficers and board of directorsbut also a completelist of
the company'sownersas of the 1993 annual shareholders'meeting.The
data analyzedhereare limitedto thetop 20 shareholdersofeach corpora-
tion.10In the Budapest Court of Registryand the 19 countyregistries,
we were able to locate ownershipfiles for 195 of the 2-00corpora-
tions and forall of the 25 banks, referredto below as the "Top 220"
firms.
Who holds the shares of these 220 largestenterprisesand banks? I
foundsome formof state ownership-with shares held by the AV-Rt,
the SPA, or the institutionsof local government (who had typicallyex-
changedtheirreal estateholdingsforenterpriseshares)-present in the
overwhelming majority(71%) of theseenterprises and banks. More sur-
prisingly,given the relativelyshorttime since the "systemchange" in
1989-90, we found36 companies(i.e., morethan 16% ofthispopulation)
in majorityforeignownership.Hungarianprivateindividuals(summed
down the top 20 owners)hold at least 25% of the shares of only 12 of
theselargestenterprises and banks.
Most interestingfromthe perspectiveof this articleis the findingof
87 cases in whichanotherHungariancompanyis among the 20 largest
shareholders.In 42 of these cases the otherHungarian companies to-

is nota simplematterwherecapitalmarketsare poorlydevel-

9 Such data collection
oped. Thereis no HungarianMoody'sand certainly no corporatedirectoryequivalent
to IndustrialGroupingsin Japan or Keiretsuno Kenkyu(see, e.g., Gerlachand
Lincoln1992).The labor-intensive from
solutionhas beento gatherthatdata directly
the HungarianCourtsof Registry.My thanksto Lajos Vekas, professorof law,
ELTE, and RectoroftheInstitute forAdvancedStudy,CollegiumBudapest,forhis
interventionsto secureaccess to thesedata and to Szabolcs Kemenyand Jonathan
Uphoffforassistancein data collection.
10 This 20-owner limitation is a conventionadoptedin researchon intercorporate
ownershipin East Asia (Gerlachand Lincoln1992;Hoshi 1994).In theHungarian
economywhereonly37 firms are tradedon theBudapeststockexchangeand where
corporateshareholding is not widelydispersedamonghundredsof small investors,
the20-ownerrestriction allowsus to accountforat least 90% of thesharesheld in
virtuallyeverycompany.

1002
 Symposium:Stark

getherhold a clear majority(50% plus one share). Thus, by the most

restrictive
definition,almost20% ofourTop 220 companiesare unambig-
uous cases of interenterprise ownership;and we find some degree of
interenterpriseownershipin almost40% of theselarge companies.
Figure 1 presentstwo discretenetworksformedthroughsuch interen-
terpriseownership.Arrowsindicatedirectionality in whicha givenfirm
holds sharesin anotherlarge enterprise.Weak ties (shareholdingswith
otherfirmsthatdo not have at least one othertie, whetheras owneror
owned, to any otherfirmin the network)are not displayed.1"The rela-
tionsdepictedin the figure,we emphasize,are the directhorizontalties
amongthe verylargestenterprises-thesuperhighways, so to speak, of
Hungariancorporatenetworks.The diagramspresentedin figure1 indi-
way of mappingthe social space of propertytransforma-
cate a different
tion than that suggestedin table 1. Whereas table 1 grouped entities
accordingto theirlegal corporatestatus,here we tracenot the distribu-
tionof attributesbut the patternsof social ties.
In analyzingtherelationaldynamicsof recombinant property,we now
shiftour focusfromthe corporatethoroughfares linkingthe large enter-
prisesto examinethelocal bywayslinkingspin-off propertieswithinthe
gravitationalfieldof large enterprises.

CorporateSatellites
We turnthus to the formwith the most dramaticgrowthduringthe
postsocialistperiod, the newly establishedlimitedliabilitycompanies
(KFT), enumeratedon line 3 of table 1. Some of theseKFTs are genu-
inelyprivateentrepreneurial ventures.But manyoftheselimitedliability
companiesare not entirelydistinctfromthe transformed shareholding
companiesexaminedabove. In fact, the formerlysocialist enterprises
have been active foundersand continueas currentownersof the newly
incorporatedunits.
The basic processof thispropertytransformation is one of decentral-
ized reorganization:Under the pressureof enormousdebt, declining
sales, and threatsof bankruptcy(or, in cases of moreprosperousenter-
prises,to forestalltakeoversas well as to increaseautonomyfromstate
ministries)directorsofmanylargeenterprises are breakingup theirfirms
(along divisional,factory,departmental,or even workshoplines) into
numerousjointstockand limitedliabilitycompanies.It is notuncommon

11The totalpatternof strongand weak tieswill be examinedin a laterstudythat

uses block-modelanalysis,testsforbank centrality,and assesses the relationship
betweenownership tiesand director
interlocks.
The purposeofthatstudywillbe to
identifythemajorcorporategroupingsin theHungarianeconomy.

1003
 Network I

Network II

E* E

B Financial Institution (Bank or Insurance) E Enterprise

FIG. 1.-Two interenterprise ownershipnetworksamong large Hungarian

firms(based on data gatheredfromcorporatefilesof 200 largestenterprisesand
top 25 banks in HungarianCourtsof Registry).
 Symposium:Stark

50.1
12.7E (i)0.1

12.8 98. 50
<Kf

(8X 10
i -0
95 9&
-2(
51S2
15.0 10f

15.9 56.3 100 metal ) 81.3 6 49.0

25.0 (8 /E)S6.feq400

25.9 69.0

26.7 ) 8 (833.3
30.8

Rt = Shareholding Company
Kft = Limited Liability Company
Bt = Partnerships
Numerals in italics indicate Heavy Metal's ownership stake in a given
satellite.

FIG. 2.-Corporatesatellites
at HeavyMetal(basedon data frominternal
company documents).

to findvirtuallyall oftheactivitiesofa largepublicenterprise distributed

among 15-20 such satellitesorbitingaroundthecorporateheadquarters.
As newlyincorporatedentitieswith legal identities,these new units
are nominallyindependent-registered separately,withtheirown direc-
torsand separatebalance sheets.But on closerinspection,theirstatus
in practiceis semiautonomous.An examinationof the computerizedrec-
ords of the Budapest Court of Registryindicates,forexample,that the
controllingshares of these corporatesatellitesare typicallyheld by the
public enterprises themselves.This patternis exemplified by the case of
one of Hungary'slargestmetallurgy firmsrepresented in figure2. As we
see in thatfigure,"Heavy Metal," an enormousshareholdingcompany

1005
AmericanJournalof Sociology

in the portfolioof the State HoldingCorporation,is the majorityshare-

holderof 26 of its 40 corporatesatellites.
Like Saturn'srings,Heavy Metal's satellitesrevolvearoundthe giant
corporateplanetin concentricorbits.Near thecenterare the core metal-
lurgyunits,hot-rolling mills,energy,maintenance,and strategicplan-
ningunitsheld in a kind of geosynchronous orbitby 100% ownership.
In the nextring,wherethe corporateheadquartersholds roughly50%-
99% of the shares,are the cold-rolling mills,wireand cable production,
the oxygenfacility,galvanizingand otherfinishing treatments,special-
ized castings,qualitycontrol,and marketingunits. As this listingsug-
gests,these satellitesare linkedto each otherand to the core units by
tiesoftechnological dependence.Relationsbetweenthemiddle-ring satel-
litesand thecompanycenterare markedby thecenter'srecurrent efforts
to introducestricter accountingproceduresand tighterfinancialcontrols.
These attemptsare counteredbytheunits'efforts to increasetheirauton-
omy-coordinatedthroughpersonalties and formalizedin the biweekly
meetingsof the "Club of KFT Managing Directors."
The satellitesof the outerringare even more heterogeneousin their
productionprofiles(construction, industrialservices,computing,ceram-
ics, machining)and are usuallyoflowerlevelsof capitalization.Unitsof
this outerringare less fixedin Heavy Metal's gravitationalfield:some
have recentlyenteredand some seem about to leave. Among the new
entrantsare someofHeavy Metal's domesticcustomers.Unable to collect
receivables,Heavy Metal exchangedinterenterprise debtforequityin its
clients,preferring thatthesemeteorsbe sweptinto an orbitratherthan
be lostin liquidation.Amongthosesatelliteslaunchedfromthe old state
enterprise are someforwhichHeavy Metal augmentsitsless thanmajor-
ity ownershipwith leasing arrangementsto keep centrifugalforcesin
check.
The corporatesatellitesamongthelimitedliabilitycompaniesenumer-
ated on line 3 of table 1 are, thus, far fromunambiguously"private"
ventures;yet neitherare they unmistakably"statist"residues of the
socialistpast. Propertysharesin mostcorporatesatellitesare notlimited
to the foundingenterprise.Top- and mid-levelmanagers,professionals,
and otherstaffcan be foundon thelistsoffoundingpartnersand current
owners.Such privatepersonsrarelyacquire completeownershipof the
corporatesatellite,preferring to use theirinsiderknowledgeto exploit
the ambiguitiesof institutional coownership.The corporatesatellitesare
thuspartiallya resultof the hedgingand risk-sharing strategiesof indi-
vidual managers.We mightask why a given managerwould not want
to acquire 100% ownershipin orderto obtain 100% of the profit,but
fromthe perspectiveof a given managerthe calculus instead is, "Why
acquire 100% of the riskif some can be sharedwiththe corporatecen-

1006
 Symposium:Stark

ter?"With ambiguousinterestsand dividedloyalties,theserisk-sharing

(or risk-shedding) owner/managers are organizationallyhedging(Sabel
1990).12
Not uncommonly, theseindividualsare joined in mixedownershipby
otherjoint stockcompaniesand limitedliabilitycompanies-sometimes
byindependentcompanies,oftenbyotherKFTs in a similarorbitaround
thesame enterprise, and frequently by shareholdingcompaniesor KFTs
spinningaround some otherenterprisewithlines of purchaseor supply
to the corporateunit(Voszka 1991). Banks also participatein thisform
ofrecombinantproperty.In manycases, theestablishment of KFTs and
othernew corporateformsis triggered byenterprise debt.In thereorgani-
zationoftheinsolventfirms,thecommercialbanks(whosesharesas joint
stockcompaniesare stillpredominantly stateowned) becomesharehold-
ers of the corporatesatellitesby exchangingdebt forequity.
We have used the term"corporatesatellite"to designatethisinstance
ofrecombinant property.An exact(butcumbersome)terminology reflects
the complex,intertwined characterof propertyrelationsin Hungary:a
limitedliabilitycompanyownedbyprivatepersons,by privateventures,
and by otherlimitedliabilitycompaniesownedbyjoint stockcompanies,
banks,and largepublicenterprises ownedbythestate.The new property
formsthusfindhorizontaltiesof cross-ownership intertwined withverti-
cal ties of nestedholdings.

MetamorphicNetworks
The recombinantcharacterofHungarianpropertyis a functionnot only
of the direct(horizontal)ownershipties among the largestfirmsand of
theirdirect(vertical)ties to theircorporatesatellitesbut also of the net-
work propertiesof the fullensembleof directand indirectties linking
entities,irrespectiveof theirattributes(large, small, or of various legal
forms)in a given configuration. The available data do not allow us to
presenta comprehensive map of thesecomplexrelations.Recordsin the
CourtsofRegistryincludedocumentson theownersofa particularfirm,

12 Manyofthesemid-level managershad experiences in the 1980swithan organiza-

tional precursorof the presentrecombinant forms-the intraenterprise partner-
ships-in whichsemiautonomous subcontracting unitsused enterpriseequipmentto
producegoodsor servicesduringthe "offhours"(Stark1986, 1989). Like "second
economy"producers who continued to holda job in stateenterprises,
theseintrapre-
neurialunitswerea widespreadresultof hedgingstrategies in theHungarianecon-
omy.Some of thesepartnerships were-scarcelydisguisedrent-seeking schemesthat
privatizedprofitstreamsand leftexpenseswiththe state-owned enterprise.
Others
creativelyredeployedresourcesfromdiversepartsof theshopfloorand regrouped,
as well,theinformal normsofreciprocitywiththetechnicalnormsofprofessionals.

1007
AmericanJournalof Sociology

AV-Rt ~ -BankSP

0 Kft~~~~~~~~~~~~~~f

FIG. 3. A metamorphic network(basedon datafromHeavyMetalinternal

documents, files,andtheBudapestCourtofRegistry).
SPA files,corporate

but enterprisesare not requiredto reportthe companiesin which they

hold a stake. However, on the basis of enterprise-level fieldresearch,
examinationof public recordsat the SPA, and interviewswithbankers
and executivesof consultingfirmswe have been able to reconstruct par-
tiallysuch networksas represented in figure3.
For orientationin this graphicspace, we positionfigure3 in relation
to figures1 and 2. Figure1 presentedinterenterpriseownershipnetworks
formedthroughhorizontalties directlylinkinglarge enterprises.Figure
2 zoomed in on the corporatesatellitesof a singlelarge enterprise.With
figure3 we pull back to examinea fragment of a broaderinterenterprise
ownershipnetworkbringingintofocusthe ties thatlinkcorporatesatel-
litesto each otherand thatformthe indirectties among heterogeneous
unitsin a morelooselycoupled network. 13

13 The metamorphic network is nota simplesummation of thesetof horizontaland

verticalties:to categoricallylabel thetiesbetweena givenKFT and a givenRT as
"vertical"wouldbe to ignorethewaystheKFTs are recombining To the
properties.
extentthatnetwork qualities(network properties,in thedoublesenseoftheterm)are
emergent in themetamorphic network, thelanguageofhorizontal and verticalshould
giveplace to moreappropriate suchas extensivity,
descriptors density,tightor loose
coupling,strongor weak ties,structural holes,and the like (Breigerand Pattison
1986;Burt1992).

1008
 Symposium:Stark

I label this emergentformof recombinantpropertya metamorphic

network.Here we see thatthe limitedliabilitycompaniesthat began as
corporatespin-offsare orientedthroughownershipties eitherto more
thanone shareholdingcompanyand/orto otherlimitedliabilitycompa-
nies. In themetamorphic network,actorsrecognizethe networkproper-
tiesoftheirinterdependent assetsand regroupthemacrossformalorgani-
zational boundaries. These creative regroupingsfail to respect the
organizationalboundariesbetweenenterprises as well as the boundaries
betweenpublic and private.
Withfewexceptions(Sabel and Prokop1994),theliteratureon postso-
cialist propertytransformation (most of it confinedto "privatization")
assumesthattheeconomicunitto be restructured is theindividualenter-
prise. But the identificationof interfirmnetworkssuggeststhat policies
and practicesaimed at restructuring should targetnot the isolated firm
butnetworks offirms.Such an alternativestrategy ofrestructuringrecog-
nizes thatassets and liabilitieshave distinctivenetworkproperties.
The industrialstructureof the socialisteconomycommonlygrouped,
withina singleenterprise,assets thatwere incompatible(exceptwithin
the logic of centralplanning).Merelyseparatingor simplyregrouping
such assets withinexistingenterprisesalone (on a firm-by-firm basis)
cannotequal the morefruitful recombinations of complementary assets
acrossa setoffirms.Restructuring via recombinantnetworksthusopens
the possibilitiesof increasingthe value of existingassets throughtheir
recombination. This regrouping does notnecessarilyimplybringinginter-
dependentassets underthe commonownershipumbrellaof a hierarchi-
cally organizedenterprise.As such, Hungarian recombinantproperty
providesexamplesofintercorporate networksas alternativesto a dichoto-
mouslyforcedchoicebetweenmarketsand hierarchies.

THE CENTRALIZED MANAGEMENTOF LIABILITIES

In theprevioussection,we examinedthe decentralizedreorganization of
assets. Propertytransformation, however,involvesnot only assets and
rightsbut also liabilitiesand obligations.In thissection,we analyzewhat
happensin a postsocialisteconomywhen actorsare called to accountfor
enterprisedebt.

Taking the Last Small Steps

The liabilitiesmanagementstorybegins in 1991 when the Hungarian
government modifiedthreeimportantlaws regulatingthe
fundamentally

1009
AmericanJournalof Sociology

accountingof assets and liabilitiesin an attemptto maintainits lead in

regionalcompetitionforforeigninvestmentsand internationalcredits.
Hungary's comparativeadvantage, it appeared, was its gradualism,
which,acrossthedecades ofthe 1970sand 1980s,had yieldeda fullrange
of marketlikeinstitutions.Admittedly, thesewere not the institutions
of
a market economy, but theywere close; and so, thegovernment reasoned,
whynottakethelast smallsteps?As thepioneerattemptto bringpostso-
cialistpracticein fullconformity withWesternaccountingand banking
standards,thenew measurescould be cast as a bold move when appeal-
ing to internationallendingagencies. But because they were not big
steps,thenew measurescouldgainexternallegitimation withoutcreating
a domesticshock.
Thus, thenew AccountingLaw of 1991 (whichtookeffecton January
1, 1992)requiredenterprisesto switchto Western-style accountingprinci-
ples. The simultaneously enacted,toughnew Western-style Bankruptcy
Act similarlycontainedstiffpersonalpenaltiesfordirectorsofenterprises
that failed to fileforbankruptcyafterthe accountants(using the new
accountingprinciples)soundedthealarm. At thesame time,thenew Act
on Financial Institutions introducedin December 1991 was designedto
put Hungary'scommercialbanks on a Westernfooting.In particular,
the reserverequirementsfor measuringcapital-adequacyratios were
modifiedand the securitiesand otherfinancialinstruments for provi-
sioningagainstqualifiedloans were respecified.
The last smallstepsprovedto be a leap intotheabyss.Alreadyreeling
fromthe collapse of the CMEA (Council forMutual Economic Assis-
tance) markets,enterprisedirectorsnow learnedfromtheiraccountants
that the new accountingpracticeswere coloringthe companies' books
even redderthanexpected.By theend of 1992,over 10,000bankruptcies
and liquidationproceedingshad been initiated-a figure10 timeshigher
thanduringthepreviousyearwhenenterprises had experiencedtheworst
shockof thecollapsedEasternmarkets(Bokros 1994). Withone-thirdto
one-halfof enterprisesin the red, the loss-makingfirmsbegan to stop
paymenton theirbank credits.By the end of 1992, the overdue loan
stockof the bankingsystemwas 127 billionforints(1.5 billionin U.S.
dollars)up 90% fromthepreviousyear(NationalBank ofHungary1992,
p. 109).
Withthousandsof firmsfilingforbankruptcy,the banks were forced
by the new bankinglaw to reclassifyloans. The subsequentdramatic
increasein the new legallyrequiredprovisioningsagainst poorlyper-
formingloans cut deeply into bank profits,slashed dividends and tax
revenuesfromthe bankingsectorto the state treasury,and turnedthe
banks' capital-adequacyratiosfrompositiveto negative. The banking

1010
 Symposium:Stark

systemwas in crisis-firstannounced,no less, in theFinancial Timesof

London (Denton 1993).

From Small Steps to Big Bailouts

The same government thathad launchedan unintendedfinancialshock
now initiateda bold plan to save thebanks.In its 1992loan consolidation
program,the governmentboughtFt 104.9 billion(about $1 billion)of
qualifieddebt (almostall in the "bad" debt classification) involving14
banks and 1,885companies.In a relatedmovein early1993,thegovern-
mentalso purchasedthe bank debt of 11 giantenterprises (theso-called
dirtydozen) forroughly$300 million.But the loan consolidationand
enterpriserecapitalizationprogramsdid notrestorestabilityin thebank-
ing sector.By September1993,onlyninemonthslater,financialexperts
were estimatingthat loans in arrearshad once again soared to 20% of
totalloan portfolios.And the 10 largestbankswere again hoveringat or
belowthe0% capital-adequacyratio(a conditionoftechnicalinsolvency).
For the government, the new bankingrules did not exclude bailing
out banks and enterprises again and again. But the big bailoutof 1993
had a new twist.Instead of buyingthe debt fromthe banks, thistime
the governmentadopted a two-stagestrategyof firstrecapitalizingthe
banks and then using the banks to work out the enterprisedebt. By
injectingenormoussumsof freshcapital intothe banks, the Ministryof
Finance became the dominant shareholderof the large commercial
banks. The firststage of the strategy,then,could be summarizedin a
phrase:Do not acquire the debt,acquire the banks.
The secondstageof thestrategywas designedto harnesstheexpertise
of the banks to the serviceof the state. Because it was the banks, and
not the state, that would be leftholdingthe qualifieddebt, the banks
would have an incentiveto collectthat debt, or at least the part they
had notalreadywrittenofftheirbooks. And theywould do so, thistime,
not withthe stateas theirsometimepartnerbut withthe state as their
majorityowner. But as Hungary'sconservative-nationalist government
shouldhave learnedfromtheearlierexperienceofstatesocialism,efforts
to exercisecontrolthroughdirectownershipdo not equal moreeffective
statecapacity.Banks have shownalmostno willingness to use theconsol-
idationfundsforactivelyrestructuring firms;and, despitetheassumption
that the Ministryof Finance's ownershipwould yield controlof the
banks,thegovernment has been almostentirely ineffective in monitoring
how the banks use the recapitalization funds.
The massivebailoutprogramswerenot,of course,withouteffects:At
Ft 300 billion($3 billion)-amountingto 10% of Hungarian GDP and

1011
AmericanJournalof Sociology

18.3% of the 1994 national budget(i.e., proportionally more than the

U.S. savings and loan bailout)-the bailouts created a long queue of
banks and firmswith theirhands out, reachingfor the state's pock-
etbook.
Thus, at the same timethatthe corporatenetworkswere engaged in
thedecentralizedreorganization ofassets,theHungarianstateattempted
of
the centralizedmanagement liabilities.That centralizationhas not
leftthe decentralizedprocessesuntouched.From the perspectiveof the
enterprises,"debt consolidation"triggersthe organizationalseparation
ofdebtsfromassets.The Hungariangovernment's attemptat thecentral-
ized managementof liabilitiesstimulatesthe networksto complement
theirstrategiesof risk spreadingwith new strategiesof risk shedding.
Two typesof strategiescan be identified,each based on the organiza-
tionalseparationof assets and liabilities.In one type,assets are distrib-
utedto thesatellitesand debtsare centralized,increasingtheenterprises'
chancesofinclusionin thegovernment-funded debt consolidation.In the
other,assets are closelyheld by the enterprisecenterand liabilitiesare
distributedto the satelliteswherenetworkties and politicalconnections
manipulate proceedingsin a Hungarian version of "bankruptcyfor
profit"(Akerlofand Romer[1993] coin the termin theirstudyof state-
managedliabilitiesin the U.S. savingsand loan bailout).
We thus see a new paternalismin Hungary: Whereas in the state
socialisteconomypaternalismwas based on the state'sattemptsto cen-
trallymanage assets(Kornai 1993a), in thefirstyearsof the postsocialist
economypaternalismis based on thestate'sattemptsto centrallymanage
liabilities.Centralizedmanagementof liabilitieswill not continueindefi-
nitely,but the organizationaldynamicsof enterprises formedunderthe
new paternalisticconditionsare likelyto have strongpath-dependent
effects.

THE MULTIPLE ACCOUNTS OF RECOMBINANT PROPERTY

In the highlyuncertainorganizationalenvironmentthat is the postso-
cialisteconomy,relativelyfewactors(apart frominstitutional designers)
set out withthe aim to createa marketeconomy.Many, indeed would
welcome such an outcome. But theirimmediategoals are more prag-
matic: at best to thrive,at least to survive.And so theystriveto use
whateverresourcesare available. That task is not so simplebecause one
mustfirstidentify the relevantsystemof accountingin whichsomething
can existas a resource.At the extreme,it is sometimeseven difficult to
a
distinguish liabilityfrom an asset. If the liabilities
of yourorganization
(enterpriseor bank) are big enough,perhapstheycan be translatedinto
qualificationsfor more resources.And what could be more worthless

1012
 Symposium:Stark

than a bankruptedlimitedliabilitycompany-except, of course, if you

have shed theriskto thebanks (and thento thestate)and put the assets
in anotherform.Assetsand liabilitieshave value not in themselvesbut
in relationto legitimatingprinciples.
To examinehow economicactorsin thepostsocialistsettingmaneuver
notonlythroughan ecologyof organizationsbut also througha complex
ecologyof orderingprincipleswe need to understandthedoublyassocia-
tive characterof assets. There are no free-floating
resources.To existas
an asset a potentialresourcemustbe mobilizablethroughtiesof associa-
tion among persons(Granovetter1985). And to be of value a potential
resourcemustalso have relativeworthaccordingto a standardof mea-
sure.To be able to circulatethroughthetiesthatbind (and thuscontrib-
uteto thatbinding)an assetmustbe justifiedwithina relativelystabilized
networkof categoriesthat make up a legitimating principle(Thevenot
1985;Boltanskiand Thevenot1991;Latour 1988;White1992). Regroup-
ing assetsthusinvolvesmakingnew associations-not onlyby rearrang-
ing social tiesamongpersonsand thingsbut also by drawingon diverse
repertoires ofjustificatory
principles.
To emphasizethe patternedand the performative aspectsof thispro-
cess,I exploita notionofaccounts.Etymologically rich,thetermsimulta-
neously connotesbookkeepingand narration.Both dimensionsentail
evaluativejudgments,and each impliesthe other:Accountantsprepare
storylines accordingto establishedformulae,and in the accountingsof
a good storyteller we know what counts. In everydaylife, we are all
bookkeepersand storytellers. We keep accountsand we give accounts,
and most important,we can all be called to account for our actions.
When we make such an accounting,we draw on and reproducesocial
orders.We can competently producejustificationsonlyin termsofestab-
lished and recognizedorderingprinciples,standards,and measures of
evaluation.Because we do notsimplygive reasonsbut also have reasons
fordoing things,accountsare not simplyretrospective; the imperative
ofjustification (Boltanskiand Thevenot1991)structures what we do and
not simplyhow we explain. We can never simply"calculate" because
we mustdo so withunitsand instruments ofmeasurement thatare deeply
structured by accountsof what can be of value. We reproducetheseunits
of measurementand we recalibratethe measuringinstruments when we
assert our worthiness,when we deferto the "more worthy,"or when
we denouncetheirstatusaccordingto someotherstandardofevaluation.
When we give an account,we affirmor challengethe orderingcriteria
accordingto whichour actions(and/orthoseof others)have been or will
be evaluated. And it is always withinaccounts that we "size up the
situation,"fornot everyformof worthcan be made to apply and not
everyasset is in a formmobilizableforthe situation.We evaluate the

1013
AmericanJournalof Sociology

situationby maneuveringto use scales thatmeasuresometypesof worth

and not others,therebyactingto validate some accountsand discredit
others.
The multipleaccountsof recombinant propertyrespondto and exploit
the fundamental,thoughdiffused,uncertainty about the organizational
environment. In transformative economies,firmshave to worrynot sim-
ply about whetherthereis demandfortheirproducts,or about the rate
of returnon theirinvestment, or about the level of profitability
but also
about theveryprincipleof selectionitself.Thus, thequestionis not only
"Will I survivethe markettest?"-but also, under what conditionsis
proofof worthon marketprinciplesneithersufficient nor necessaryto
survive?Because thereare multiplyoperative,mutuallycoexistentprin-
ciples of justificationaccordingto which you can be called on to give
accounts of your actions, you cannot be sure what counts. By what
proofand accordingto whichprinciplesof justificationare you worthy
to stewardsuch and such resources?Because of thisuncertainty, actors
will seek to diversifytheirassets,to hold resourcesin multipleaccounts.
This abilityto glideamongprinciplesand to producemultipleaccount-
ings is an organizationalhedging.It differs,however,fromthe kind of
hedgingto minimizeriskexposurethat we would findwithina purely
marketlogic-as, forexample,whentheshopkeeperwho sellsswimwear
and suntanlotionalso devotessome floorspace to umbrellas.Instead of
actingwithina singleregimeof evaluation,these actors use organiza-
tionalhedgingthatcrossesand combinesdisparateevaluativeprinciples.
Recombinantpropertyis a particularkind of portfoliomanagement.It
is an attemptto have a resourcethatcan be justifiedor assessed by more
thanone standardofmeasure(as, e.g., therabbitbreederwhoseroadside
stand advertises"pets and meat" in the documentaryfilm,Roger and
Me). In managingone's portfolioof justifications,one startsfromthe
dictum:diversify youraccounts.
The adroitrecombinant agentin thetransformative economiesof East
CentralEurope diversifies in
holdings response to fundamental uncer-
taintiesabout what can constitutea resource.Underconditionsnot sim-
plyofmarketuncertainty but oforganizationaluncertainty, therecan be
multiple(and intertwined) strategiesforsurvival-based in some cases
on profitabilitybutin otherson eligibility.Whereyoursuccessis judged,
and theresourcesplaced at yourdisposaldetermined, sometimesby your
marketshareand sometimesby the numberof workersyou employin a
region;sometimesby yourprice-earnings ratio and sometimesby your
"strategicimportance";and, when even the absolute size of yourlosses
can be transformed into an asset yieldingan incomestream,you might
be wise to diversify yourportfolio,to be able to shiftyouraccounts,to
be equally skilledin applyingforloans as in applyingforjob creation

1014
 Symposium:Stark

subsidies,to have a multilingualcommandof the grammarof credit

worthinessand the syntaxof debt forgiveness.To hold recombinant
propertyis to have such a diversified
portfolio.
To gain roomformaneuver,actorscourtand even createambiguity.
They measurein multipleunits,theyspeak in manytongues.They will
be less controlledby othersif theycan be accountable (able to make
credibleaccounts)to many.14In so doing,theyproducethe polyphonic
discourseof worththatis postsocialism.
We can hear that polyphonicchorusin the diverseways that firms
justifytheirclaimsforparticipationin the debt-reliefprogram.The fol-
lowinglitanyofjustificationsare stylizedversionsof claimsencountered
in discussionswith bankers,propertyagency officials,and enterprise
directors:Our firmshouldbe includedin the debt reliefprogram
becausewewillforgive ourdebtors(i.e.,ourfirmoccupiesa strategicplace
in a networkofinterenterprise
debt)
becausewe aretrulycreditworthy(i.e.,ifourliabilities
areseparated from
ourassets,we willagainbe eligibleformorebankfinancing. Similar
couldbe provided
translations foreachofthefollowing justifications)
becausewe employthousands
becauseoursuppliers dependon us fora market
becausewe arein yourelectiondistrict
becauseourcustomers dependon ourproduct inputs
becausewe can thenbe privatized
becausewe can neverbe privatized
becausewe tookbigrisks
becausewe wereprudent anddid nottakerisks
becausewe wereplannedin thepast
becausewe havea planforthefuture
becausewe exportto theWest
becausewe exportto theEast
becauseourproduct hasbeenawardedan International Standards Quality
ControlCertificate
becauseourproduct is partoftheHungarian nationalheritage
becausewe arean employee buy-out
becausewe area management buy-in
becausewe arepartlystate-owned
becausewe arepartlyprivatelyheld
becauseourcreditorsdroveus intobankruptcy whentheyloanedto us at
higherthanmarketratesto artificially raisebank profits in orderto
paydividends intoa statetreasury whosecoffers had dwindledwhen
likeourselves
corporations effectivelystoppedpayingtaxes.
And so we mustask, intowhoseaccountand by whichaccountwill debt
flow?Or, in such a situation,is anyoneaccountable?
forgiveness

14 See Padgettand Ansell(1993) foran analysisof such multivocality

in another
historical
setting.

1015
AmericanJournalof Sociology

AN EAST EUROPEAN CAPITALISM?

How are we to understandtheseunorthodoxforms,theseorganizational
"monsters"regroupingthe seeminglyincongruous?In this concluding
section,we reconsiderthethreeaspectsofrecombinant property(blurring
of public and private,blurringof enterpriseboundaries,and blurring
the boundednessof legitimatingprinciples)in termsof threeunderlying
concepts-mixture,diversity,and complexity.

Mixture
Imaginetwo economies,each of equal partspublic and private.In one,
halfthefirmsare fullyprivate,halfare fullypublic. In the other,every
firmis half public, half private. Each is a "mixed economy."Yet is it
likelythattheirdynamicswill be the same?'5No two economiesclosely
approximatethethoughtexperiment's ideal types;but it nonethelessputs
in sharp reliefthe question What is the mix of the postsocialistmixed
economy?
My findingsof corporatespin-offs and recombinantreorganizationat
theenterprise level, and of widespreadpublicownershipcombinedwith
interenterpriseownershipnetworksamong the verylargestenterprises,
challengesthe assumption,widelyheld on all sides of the privatization
debate, thatpostsocialisteconomiescan be adequatelyrepresentedin a
two-sectormodel. That analyticshortcomingcannot be remediedby
moreprecisespecification of the boundarybetweenpublic and private:
the old propertydivide has been so eroded thatwhat mightonce have
been a distinctboundaryline is now a recombinantzone. Hungaryis a
postsocialistmixed economynot because of a simple dualism of well-
bounded state-ownedfirmsin one sectorand privatelyowned firmsin
anotherbut because manyfirmsthemselvesexploitaspectsofpublic and
privatepropertyrelations.16 What we findare new formsof propertyin
whichthequalitiesof privateand public are dissolved,interwoven,and
recombined.Propertyin East European capitalismis recombinantprop-
erty,and itsanalysissuggeststheemergenceofa distinctively East Euro-
pean capitalismthatwill differas muchfromWestEuropean capitalisms
as do contemporary East Asian variants.
15 In a relatedpath dependent thoughtexperiment: Imaginetwo mixedeconomies
each withhalfthefirms fullypublicand halfthefirmsfullyprivate.The firstarrived
at thatsectoralmix froma starting pointof onlypublicfirms.The other,froma
starting pointofonlyprivatefirms.Aretheirdynamicslikelyto be thesame?
16 It was nottheaim ofthisarticleto producea definitive
testoftherelativeweights
ofpublic,private,and recombinant zonesin thepostsocialist
mixedeconomy.To do
so, we willneedorganizationalsurveys,conductedin closeconjunctionwithdetailed
to yieldmorerefined
fieldinvestigations
enterprise-level and nuancedcategoriesand
measures.

1016
 Symposium:Stark

The conceptof a postsocialistmixedeconomyis a usefulfirstapproxi-

mationof an East European capitalism.But its essentialistcategoriesof
"public" and "private" (and the related dualisms of "market" and
"redistribution")-evenwhen opened up to the possibilityof being
mixedtogetherin thesame organizationalsetting-may be morelimiting
thanilluminating.
For decades, capitalismwas definedvis-a-vissocialismand vice versa.
Their systematiccomparisonenrichedour understandingof both, but
the"methodsofmirroredopposition"and similarconstructs(Stark 1986;
Szelenyi1978, 1988)thatworkedwiththesedualismsare no longerfruit-
ful. The demise of socialismchallengesthat analyticallyforcedchoice,
and itoffersan opportunity forenrichingcomparativeinstitutionalanaly-
sis. When we stop definingcapitalismin termsof socialism,we see that,
in our epoch, capitalismas a constructis only analyticallyinteresting
in the plural: capitalismsmustbe definedand comparedvis-a-viseach
other.

Diversity
Our firstanalyticshift,therefore,must be fromthe conceptual tools
aroundtheconceptofmixtureto thosearoundthatof diversity.Capital-
isms are diverse,and that diversityis manifestedin formsthat cannot
be adequatelyconceptualizedas mixturesof capitalismand socialism.17
By analyzingrecombinant property notonlyas thedissolutionand inter-
weavingofelementsofpublicand privatebutalso as a blurringoforgani-
zational boundariesin networksof interlocking ownership,we can es-
cape, forexample,the termsof the debate about whetherthe "lessons
of East Asia forEastern Europe" are the virtuesof neoliberalismor of
neostatism(WorldBank 1993; Amsden1994). Instead we join economic
sociologistswho are studyingthe East Asian economiesfroma network-
centeredapproachin whichnot markets,nor states,nor isolatedfirms,
but social networksare the basic unitsof analysis(Gereffi1994; Hamil-
ton,Zeile, and Kim 1990;Hamiltonand Feenstra1995). In thisperspec-
tive,theabilityof the East Asian economiesto adapt flexiblyto changes
in worldmarketsrestsin theinterlocking ties characteristic
of corporate

17 My argument, thus,bearsno resemblance to "thirdroad"solutions(i.e., themis-

takennotionthattherecouldbe somecombination ofthebestfeaturesof capitalism
withthebestfeatures ofsocialism),and itfollowsthatI am notarguingthatrecombi-
nantproperty is a "bestway."As peoplelivingin East CentralEuropehave known
fordecadesifnotcenturies, all thebestroadsto capitalismstartedsomewhereelse.
I am remindedof thejoke in whichan Irishmanin the farcountryside is asked,
"What'sthe best way to get to Dublin?" He thinksfora minute,and responds,
"Don't startfromhere."

1017
AmericanJournalof Sociology

groups(Orru,Biggart,and Hamilton1991; Granovetter1995), whether

thesebe thepatternsofmutualshareholding withintheJapanesekeiretsu
(Gerlach and Lincoln 1992; Hoshi 1994); the ties of familyownership
withinthe moreverticallyintegratedSouth Korean chaebol (Kim 1991;
Hamilton and Feenstra 1995); the social ties of the more horizontally
integrated Taiwanese quanxiqiye "related enterprises"(Numazaki
1991); or the dense ties thattransgressorganizationalboundariesin the
"buyer-driven" and "producer-driven" networksin Hong Kong, Singa-
pore, and elsewherein SoutheastAsia (Gereffi1994).
These recentstudiesof the social embeddednessand local organiza-
tionalinnovationcharacteristic ofEast Asian corporatenetworkssuggest
thatthestrategicchoiceis notplans or markets,or even clans or markets,
but clansfor markets.Market orientationmust be distinguishedfrom
marketcoordination:a broadvarietyofinstitutions ofnonmarketcoordi-
nationare compatiblewithhighperformance marketorientation (Schmit-
ter1988; Boyer1991; BresserPereira1993). Many of themostsuccessful
formsofnetworkcoordinationin East Asia, moreover,appeared to early
observersas highlyimprobableformswhose atavisticfeaturescould not
possiblysurvivebeyondtheperiodofpostwarreconstruction fromwhich
theyarose.18 Our pointof departure,it shouldbe clear, however,is not
to look to EasternEurope to findHungariankeiretsuor Czech chaebol.
Instead of searchingfordirectcounterparts, East Asian/EastEuropean
comparisonswillyieldnew conceptswhenwe graspthespecificity of the
regionalvariantsby explainingthe differences among the various coun-
trieswithina region.'9
Futureresearchmustexaminewhetherthe East European corporate
networksare becomingsuccessfully orientedto the worldmarket.But it
is not too earlyto pose analyticdimensionsalong whichwe could assess
the potentialforrecombinantpropertyto contributeto economicdevel-
opment.
18 Incongruity,
in itself,neither
insuressurvivalnorcondemns an organizational
form
to an earlydeath.Kim's(1991)discusionof thecombinatory logicof the formation
of the chaebolin Korea immediately followingWorldWar II invitescomparison
withtheformation ofrecombinant structuresduringthecontemporary periodofEast
Europeanreconstruction.
19 Starkand Bruszt(1995),e.g., comparecorporate networksin Hungaryand the
Czech Republic. They findthat Hungariannetworksare formedpredominantly
through enterpriseto enterpriselinks,sometimes involvingbanksyetabsenttiesbe-
tweenbanksand intermediate-level institutions
suchas investment companies.In the
CzechRepublic,bycontrast, ownership networks are formed
predominantly through
tiesat themesolevel in the crossownershipof banksand largeinvestment funds,
but directownershipconnections amongenterprises themselvesare rare. Whereas
Hungariannetworks coupledat thelevelofenterprises
are tightly butlooselycoupled
at themesolevel,Czechnetworks are looselycoupledat thelevelofenterprisesand
tightlycoupledat themesolevel.

1018
 Symposium:Stark

One startingpoint,ready-at-hand fromthe burgeoningliteratureon

the "transitional"economies,would be to ask, Do theycontributeto
creativedestruction? That litmustestis based on a widelyheld assump-
tionthateconomicdevelopmentwill be best promotedby "allowingthe
selectionmechanismto work"throughbankruptciesof underperforming
enterprises.Recombinantpropertywould notreceivean unambiguously
positivescoremeasuredbythisstandard.Indeed,thekindsofinterenter-
prise ownershipdescribedabove are classic risk-spreadingand risk-
sharingdevicesthatmitigatedifferences acrossfirms.By dampeningthe
performance the survivalof the weaker
of the strongerand facilitating
firmsin theinterfirm networks,theymighteven impedecreativedestruc-
tionin the conventionalsense.
But thereis some questionthata tidal wave of mass bankruptciesis
a long-termcure forthe postsocialisteconomies.With the catastrophic
loss of marketsto the East and withthe stagnationof the economiesof
potentially new tradingpartnersto theWest,thedepthand lengthofthe
transformational crisisin East CentralEurope now exceeds that of the
GreatDepressionoftheinterwarperiod(Kornai 1993b).In such circum-
stances, an absolute hardeningof firms'budget constraintsnot only
drivespoorlyperforming firmsintobankruptcybut also destroysenter-
prisesthat would otherwisebe quite capable of makinga high perfor-
manceadjustment(see esp. Cui 1994).Wantondestruction is notcreative
destruction,goes this reasoning,and recombinantpropertymightsave
someofthesestruggling butcapable firmsthroughrisk-sharing networks.
Alongthisline of reasoning,we would want to assess whetherthe sacri-
ficein allocativeefficiencyby retardingbankruptcy is beingoffsetby the
preservation ofassetswithreal potentialforhighperformance in a situa-
tionof economicrecovery.
A related,but analyticallyseparate,point is that riskspreadingcan
be a basis forrisktaking.Extraordinarily highuncertainties of the kind
we see now in thepostsocialisteconomiescan lead to low levelsofinvest-
mentwithnegativestrategiccomplementarities (as when firmsforgoin-
vestmentsbecause theyexpect a sluggisheconomybased on the lack
of investmentsby others).By mitigatingdisinclinationsto invest,risk
spreadingmightbe one meansto breakoutofotherwiselow-levelequilib-
riumtraps.20Firms in the postsocialisttransformational crisisare like
mountainclimbersassaultinga treacherousface, and the networksof
interenterprise ownershipare the safetyropes lashing them together.
Neoliberalswho bemoana retardedbankruptcy ratefailto acknowledge

20
On strategic in thepostsocialist
complementarities economies
see especially
Litwack
(1994).Hirschman(1958)providestheclassicstatement
onlow-levelequilibriumtraps
and theimportanceofriskspreadingforeconomicdevelopment.

1019
AmericanJournalof Sociology

thattheremightbe circumstances whenthismutualbindingis a precon-

ditionforattempting a difficult
ascent. Along thisline of reasoning,we
would want to assess whetherthe opportunities forrisksheddingin the
Hungarian settingcan be offsetwhen networks(ratherthan "develop-
mentalstates"[Evans 1992])performdisequilibrating functionsthatfa-
cilitateand stimulateentrepreneurial risktaking.
Economicdevelopmentin East CentralEurope does requiremoreexit
(some,indeedmany,firmsmustperish)and moreentryas well. But for
destructionto be creative,thesedeaths mustbe accompaniedby births
notsimplyofnew organizationsbutofnew organizationalforms. Organi-
zationalformsare specificbundlesof routines,and thereductionoftheir
diversitymeanstheloss of organizedinformation thatmightbe of value
when the environment changes(Hannan 1986; Boyer 1991; Stark 1989,
1992). Fromthisperspective,an economythatmaximizedallocativeeffi-
ciency(byputtingall resourcesin themostefficient form)would sacrifice
adaptive efficiency. Socialism, in this view, failed not only because it
lacked a selectionmechanismto eliminateorganizationsthatperformed
poorlybutalso becauseit putall itseconomicresourcesin a singleorgani-
zationalform-the stateenterprise.Socialismdrasticallyreducedorgani-
zational diversityand in so doingprohibiteda broad repertoireof orga-
nized solutions to problems of collective action. Along this line of
reasoning,an assessmentof formsof recombinantpropertyin an East
European capitalismshouldstartnot by testingwhethertheyreproduce
statesocialismor harborreal privatepropertybut whethertheycontrib-
ute to adaptive efficiency.
For the propertyrightsschool,it is not destruction(bankruptcy)nor
diversitybut the clarityof propertyrightsthat will yield the rightset
of incentivesto make restructuring in the postsocialisttransformation
performanceenhancing. Instead of reassigningpropertyrightsto an
owner(an ironiclegacyof an essentiallyMarxistnotionof property), this
school argues that propertycan be productively"dis-integrated" (Grey
1980) such thatdifferent actorscan legitimatelyclaim rightsto different
aspectsand capacitiesof thesame thing(Hart 1988; Comisso 1991). But
howeverdisaggregated, propertyrightsmustbe clarifiedifaccountability
is to be insured.Walder(1994), theleadingproponentofthisperspective
in thepostsocialistdebate,forexample,showsthatit is nottheprivatiza-
tionof assetsbut theclarification of propertyrightsthathas contributed
to the dynamismof townshipand village enterprisesin Chinese light
industry.
Along this line of reasoning,we should assess whetherrecombinant
propertyis leadingto well-defined propertyrights.The initialevidence
presentedin this articlesuggeststhat recombinantpropertywould fail
such a test. But fromanotherperspectivein the debate over property

1020
 Symposium:Stark

rights,the blurringof enterpriseboundariesmightbe a viable strategy

to promoteorganizationalflexibility,On thebasis ofresearchin advanced
manufacturing fieldsin Germanyand theUnitedStates,Sabel (1990) and
Kogut, Shan, and Walker (1992) demonstratethat under conditionsof
extreme market volatilityor of extraordinarilyrapid technological
change,economicactorsengagein hedgingstrategiesvis-a-visotheror-
ganizations(partnersor competitors) in theirorganizationalfield.When
the futureis highlyuncertain,it is far fromclear at Tl whetheryour
assets will be interdependent with mine at T2. In such situations,in
additionto the dualism make or buy (hierarchyor market)thereis an
alternative-cooperate.Kogutobservesthatone manifestation of such a
hedgingstrategyis cross-ownership (not simplyamong purchasersand
suppliersbut also among competitors),and he findsdense patternsof
cross-ownership amongcompetitors in thefieldofmicroprocessing where
firmscannotbe certainwhose standardswill be the industrystandards
in the next round. Sabel goes even further,arguingthat, in cases of
extremelycomplex asset interdependence, it is not clear-cutproperty
claims(howeverdensethecross-ownership) but an ambiguityofproperty
claimsthatprovidesflexibleadaptationto the market.Sabel's argument
departs radicallyfromthe propertyrightsschool: he is claimingthat
actorsare not assigneddifferent rightsover different aspects of an asset
butare makingoverlappingclaimson thesame aspect.This is ambiguous
property,not disaggregatedproperty.
The hedgingstrategiesand boundaryblurringin postsocialistrecon-
struction,it seems,findcounterparts in some of thetechnologically most
highlysophisticatedsectorsofNorthAmericanand West European capi-
talism.Along the dimensionof thisline of reasoning,we should assess
whetherrecombinantpropertyis, in fact,contributing to flexibility
and
whetherany gainsthatmightso accrue are enoughto offsetthe possible
sacrificeof accountability.We reencounter thistrade-off of adaptability
and accountability as we turnfromthe issuesof organizationaldiversity
and propertyrightsto the problemof heterogeneouslegitimatingprin-
ciples.

Complexity
In restructuringassets, we mightsay that actorsare "identifying"
new
resources,but thiswould suggestthattheresourcewas simplyhiddenor
underutilizedand onlyneededto be uncovered.In fact,beforerecombin-
ing resources,theymustfirstredefinethem.We call this abilityto "re-
cognize"the propertiesof personsand thingsorganizationalreflexivity.
It cannot be derived fromthe ambiguityof propertyclaims but is a
functioninstead of the ambiguityof organizingprinciples.The key to

1021
AmericanJournalof Sociology

adaptabilityin thisview is not simplythe diversityof typesof organiza-

tionsbutthepossibilities forcross-fertilization
insideand acrossorganiza-
tionswheremultiplyoperativelegitimating principlescollide-or in Har-
risonWhite's(1993) phrase, "values mate to change."'21
Some mightargue,of course,thatmultipleordersare fine-provided
that each occupies a distinctlybounded domain. Such is the model of
modernity in "modernization" theory:throughdifferentiation, each do-
mainof societywould developas a separateautonomoussubsystemwith
its own distinctivelogic. Complexityin thisview requiresdiversitybut
only as the juxtapositionof clearlybounded rationalities.Marxism,of
course,has its own conceptionof complexity:the temporaryoverlap of
mutuallycontradictory principles.Both modernization theoryand Marx-
ism are deeplygroundedin the transitionproblematic.The noisyclash
of ordersis onlytemporary: the revolutionary momentforone, the pas-
sage to differentiateddomainsin the other.
If we break with this transitionproblem,we can escape fromthe
impoverished conceptionsofcomplexity in bothMarxismand moderniza-
tiontheory.In the alternativeconceptionofferedhere,complexityis the
interweavingof multiplejustificatory principleson the same domain
space. That view, ofcourse,shareswithmodernization theorythenotion
of distinctivedomains-relativelyautonomousfieldsof action(Bourdieu
1990). And it shareswithMarxismthenotionofthe collisionof ordering
principles.But unlike modernizationtheory,each domain is a site of
heterogeneity; and unlikeMarxism,thattensionis not consolidatedand
thenreleasedin an all-encompassing revolutionary moment.The noisy
clash of ordersoccursthroughout thesocial world,and it is not transient
but ongoing-punctuatedby relative,localized stabilizationsbut never
equilibrium(Latour 1988).
Postsocialistsocietiesare enteringthisdiscordantworld. To stillthat
noisyclash by theascendencyofone accounting,withprofitability as the
sole metricand marketsas the onlycoordinatingmechanism,would be
to duplicatetheattemptof Communism,withits impositionof a unitary
justificatory principle,a stricthierarchyof propertyforms,and a single
coordinatingmechanism.To replicatethe monochromewitha different
coloringwould be to destroythe heterogeneity of organizingprinciples
thatis the basis of adaptability.
As this account of recombinantpropertyhas demonstrated,postso-
cialist societiesare not lackingin heterogeneousorganizingprinciples.
The problemtherefore is nota simplelack ofaccountability but an over-

21 See esp. Grabher (1994) fora discussion of how rivalryof coexistingorganizational

formscontributesto reflexivityand adaptability. For related views on adaptability
and complexity,see Landau (1969), Morin (1974), and Conrad (1983).

1022
 Symposium:Stark

abundance of accountability:An actor who, withinthe same domain

space, is accountable to every principleis accountable to none. The
adaptabilityof moderncapitalismsrestsnot simplyin the diversityof
organizationsbut in the organizationof diversity:enough overlap of
legitimatingprinciplesacross domainsto fosterrivalryof competingac-
countswithindomainsand enoughboundednessof rationalitiesto foster
accountability.It is notin findingtherightmixofpublic and privatebut
in findingtherightorganizationofdiversity toyieldbothadaptabilityand
accountabilitythat postsocialistsocietiesface theirgreatestchallenge.

REFERENCES
Akerlof, GeorgeA., and Paul M. Romer.1993."Looting:The EconomicUnderworld
ofBankruptcy forProfit."Brookings Paperson EconomicActivity2:1-73.
Alexander,Jeffrey C. 1994. "Modern,Anti,Post, and Neo: How Social Theories
Have Triedto Understand the'New World'of 'Our Time."' Zeitschriftfiur Soci-
ologie23:165-97.
Amsden,Alice. 1994. "Can EasternEuropeCompeteby Gettingthe PricesRight?
ContrastwithEast AsianStructural Reforms." Pp. 81-107 in RebuildingCapital-
ism:Alternative Roads afterSocialismand Dirigisme,editedby AndresSoimano,
OsvaldoSunkel,and MarioI. Blejer.AnnArbor:University ofMichiganPress.
Blanchard,OliverJean,KennethA. Froot,and Jeffrey D. Sachs. 1994.Introduction
to The Transitionin EasternEurope,vol. 2. Editedby OliverJean Blanchard,
KennethA. Foot,and Jeffrey D. Sachs. Chicago:University ofChicagoPress.
Bokros,Lajos. 1994. "Privatization and the BankingSystemin Hungary."Pp.
305-20 in Privatizationin theTransition Process.RecentExperiencesin Eastern
Europe,editedby Laszl6 Samuely.Geneva:UnitedNationsConference on Trade
and Development and KOPINT-DATORG.
Boltanski,Luc, and LaurentThevenot.1991.De la justification: Les economiesde
la grandeur.Paris:Gallimard.
Bourdieu,Pierre.1990.TheLogicofPractice,translated by RichardNice. Stanford,
Calif.:Stanford UniversityPress.
Boyer,Robert.1991. "MarketswithinAlternative Coordinating Mechanisms:His-
tory,Theory,and Policyin the Lightof the Nineties."Paper presentedto the
Conference on theComparativeGovernanceofSectors,Bigorio,Switzerland.
Breiger,RonaldL. 1995."Social Structure and thePhenomenology ofAttainment."
AnnualReviewofSociology21:115-36.
Breiger,RonaldL., and PhilippaE. Pattison.1986. "CumulatedSocial Roles: The
Dualityof Personsand TheirAlgebras."Social Networks 8 (3): 215-56.
BresserPereira,Luiz Carlos. 1993. "The Crisis of the State Approachto Latin
America."Discussionpaperno. 1., Instituto Sul-Norte,November.
Bruszt,Laszlo. 1988."A centralizaci6
csapdajaes a politikairendszer reformalternati-
vai" (The trap of centralizationand the alternatives of reforming the political
system).Medvetanc,no. 1, pp. 171-97.
Bunce, Valerie,and Maria Csanadi 1993. "Uncertainty in the Transition:Post-
Communism in Hungary."East EuropeanPoliticsand Societies7 (2): 240-75.
Burawoy,Michael,and Pavel Krotov.1992."The SovietTransitionfromSocialism
to Capitalism:WorkerControland EconomicBargainingin theWood Industry."
AmericanSociologicalReview57:16-38.
Burt,Ronald. 1992.Structural Holes. Cambridge, Mass.: HarvardUniversity Press.

1023
AmericanJournalof Sociology

Comisso, Ellen. 1991. "PropertyRights,Liberalism,and the Transitionfrom

'ActuallyExisting'Socialism."East EuropeanPoliticsand Societies5 (1): 162-
88.
Conrad,Michael.1983.Adaptability. New York: PlenumPress.
Cui, Zhiyuan. 1994. "Epilogue:A Schumpeterian Perspectiveand Beyond." In
China: A Reformable Socialism?editedby Yang Gan and ZhiyuanCui. Oxford:
OxfordUniversity Press.
. In press. "Moebius-Strip Ownershipand Its Prototypein Chinese Rural
Industry." Economyand Society.
Denton,Nicholas.1993."Two HungarianBanksSaid to Be TechnicallyInsolvent."
FinancialTimes,May 20.
Evans, Peter. 1992. "The State as Problemand as Solution."Pp. 139-81 in The
PoliticsofEconomicAdjustment, editedbyStephanHaggardand RobertKaufman.
Princeton, N.J.: Princeton Press.
University
Figyelo.1993.Top 200: a legnagyobb (The top 200 largestenterprises).
vdllalkozdsok
Figyelo,specialissue.
Gabor,Istvan. 1979. "The Second(Secondary)Economy."Acta Oeconomica22 (3-
4): 91-311.
. 1986."Reformokmasodikgazdasag,allamszocializmus. A 80-as6vektapasz-
6s 6sszehasonlft6
talatainakfelj6d6stani gazdasagtanitanulsagair6l"(Reforms,sec-
ond economy,statesocialism:Speculationon the evolutionary and comparative
economiclessonsoftheHungarianeighties).Val6sdg,no. 6, pp. 32-48.
. 1994."Modernity or a New Type ofDuality?The SecondEconomyof To-
day." Pp. 3-21in TheLegacyofCommunism in EasternEurope,editedby Janos
MatyasKovacs. New Brunswick, N.J.: TransactionBooks.
. 1996."Too Many,Too Small:SmallBusinessesin Postsocialist Hungary."In
Restructuring Networks:Legacies,Linkages,andLocalitiesin Postsocialist
Eastern
Europe,editedbyGernotGrabherand David Stark.New York:OxfordUniversity
Press.
Gereffi, Gary.1994."The Organization ofBuyer-Driven GlobalCommodity Chains:
How U.S. RetailersShape OverseasProduction Networks."Pp. 95-122 in Com-
modityChainsand GlobalCapitalism,editedby GaryGereffi and MiguelKornze-
niewicz.Westport, Conn.: Praeger.
Gerlach,MichaelL., and JamesR. Lincoln.1992. "The Organizationof Business
Networks intheUnitedStatesandJapan." Pp. 491-520inNetworks and Organiza-
tions,editedby NitinNohriaand RobertG. Eccles. Cambridge,Mass.: Harvard
BusinessSchoolPress.
Grabher,Gernot.1994.In Praise ofWaste:Redundancyin RegionalDevelopment.
Berlin:EditionSigma.
Granovetter, Mark. 1985."EconomicAction,Social Structure, and Embeddedness."
American JournalofSociology91:481-510.
. 1995. "Coase Revisited:BusinessGroupsin theModernEconomy."Indus-
trialand CorporateChange4 (1): 93-130.
Grey,Thomas.1980."The Disintegration ofProperty." Pp. 69-85 inProperty,edited
byJ. RollandPennockand JohnW. Chapman.New York:New York University
Press.
Hamilton,GaryG., and RobertC. Feenstra.1995. "Varietiesof Hierarchiesand
Markets:An Introduction." Industrialand CorporateChange4 (1): 51-91.
Hamilton,GaryG., WilliamZeile, and Wan-JinKim. 1990. "The NetworkStruc-
turesof East AsianEconomies."In Capitalismin Contrasting Cultures,editedby
S. R. Cleggand S. G. Redding.Berlin:Walterde Gruyter.
Hannan, Michael T. 1986. "Uncertainty, Diversity,and OrganizationalChange."
Pp. 73-94 inBehavioraland Social Science:FiftyYearsofDiscovery.Washington,
D.C.: NationalAcademyPress.

1024
 Symposium:Stark

Hart,Oliver.1988. "Incomplete Contractsand theTheoryof theFirm."Journalof

Law, Economicsand Organization 4 (1): 119-39.
Hirschman,Albert.1958. The Strategyof EconomicDevelopment.New Haven,
Conn.:Yale University Press.
Hoshi,Takeo. 1994."The EconomicRoleofCorporateGroupingand theMain Bank
System."Pp. 285-309in TheJapaneseFirm:TheSourcesofCompetitive Strength,
editedbyMasahikoAokiand RonaldDore. London:OxfordUniversity Press.
Kim, Eun Mee. 1991. "The IndustrialOrganizationand Growthof the Korean
Chaebol:Integrating Developmentand Organizational Theories."Pp. 272-99 in
BusinessNetworks and EconomicDevelopment in East and SoutheastAsia, edited
by Gary Hamilton.Hong Kong: Universityof Hong Kong, Centreof Asian
Studies.
Kogut,Bruce,and Udo Zander.1992."KnowledgeoftheFirm,CombinativeCapa-
bilities,and theReplicationofTechnology." Organization Science 3 (3): 383-97.
Kogut,Bruce,Weijan Shan, and GordonWalker. 1992. "The Make-or-Cooperate
Decisionin the Contextof an IndustryNetwork."Pp. 348-65 in Networksand
Organizations, editedby NitinNohriaand RobertG. Eccles. Cambridge,Mass.:
HarvardBusinessSchoolPress.
Konrad,George,and Ivan Szel6nyi.1979.Intellectuals on theRoad to Class Power.
New York:HarcourtBrace & Jovanovich.
Kornai,Janos.1980. The EconomicsofShortage.Amsterdam: North-Holland Pub-
lishing.
. 1992. "The Post-Socialist
Transitionand theState:Reflections in theLight
ofHungarianFiscal Problems."AmericanEconomicReview82 (2): 1-21.
. 1993a. "The Evolutionof FinancialDisciplineunderthePostsocialistSys-
tem."Kyklos46:315-36.
. 1993b."TransitionalRecession."Discussionpaper. InstituteforAdvanced
Study/Collegium Budapest.
Laky, Ter6z. 1992. "Small and Medium-SizeEnterprises in Hungary."Reportfor
theEuropeanCommission. InstituteforLabourStudies,Budapest.
Landau, Martin.1969. "Redundancy, Rationality, and the Problemof Duplication
and Overlap."PublicAdministration Review29 (4): 346-58.
Latour,Bruno. 1988. The Pasteurizationof France. Cambridge,Mass.: Harvard
University Press.
Lipset,SeymourMartin.1990. "The Death of the ThirdWay." NationalInterest,
no. 20 (Summer):25-37.
Litwack,John.1994."Strategic Complementarities and EconomicTransition."Dis-
cussionpaper.Institute forAdvancedStudy/Collegium Budapest.
Mellar,Tamas 1992. "Two Years of Privatization." Nepszabadsag,May 22.
Morin, Edgar. 1974. "Complexity." InternationalSocial Science Journal26 (4):
555-82.
National Bank of Hungary.1992. AnnualReport.Budapest: National Bank of
Hungary.
Nee, Victor.1989."A TheoryofMarketTransition." AmericanSociologicalReview
4 (5): 663-81.
. 1992."OrganizationalDynamicsofMarketTransition: HybridForms,Prop-
ertyRights,and Mixed Economyin China." Administrative Science Quarterly
37:1-27.
Nelson,RichardR., and SidneyG. Winter.1982.An Evolutionary TheoryofEco-
nomicChange.Cambridge:CambridgeUniversity Press.
Neumann,Laszl6. 1989. "MarketRelationsin Intra-Enterprise Wage Bargaining?"
Acta Oeconomica40:319-38.
Numazaki,Ichiro.1991."The RoleofPersonalNetworksin theMakingofTaiwan's
Guanxigiye (RelatedEnterprises)."Pp. 77-93 in BusinessNetworks and Economic

1025
AmericanJournalof Sociology

Development in East and SoutheastAsia, editedby GaryHamilton.Hong Kong:

University ofHong Kong,CentreofAsianStudies.
Oi, Jean C. 1992. "Fiscal Reformand the EconomicFoundationsof Local State
Corporatism in China." WorldPolitics45:99-126.
Orru,Marco,NicoleWoolseyBiggart,and GaryG. Hamilton.1991."Organizational
Isomorphism in East Asia." Pp. 361-89 in The New Institutionalism in Organiza-
tionalAnalysis,editedbyWalterW. Powelland Paul J. DiMaggio.Chicago:Uni-
versity ofChicagoPress.
Padgett,JohnF., and Christopher K. Ansell.1993."RobustActionand theRise of
theMedici,1400-1434."American JournalofSociology98:1259-1319.
Pistor,Katharina,and JoelTurkewitz.1994."CopingwithHydraStateOwnership
afterPrivatization: A ComparativeStudyofHungary,Russia,and theCzech Re-
public."Presented to theConference on CorporateGovernancein CentralEurope
and Russia,theWorldBank, December.
R6na-Tas,Akos. 1994."The FirstShall Be Last? Entrepreneurship and Communist
Cadres in the TransitionfromSocialism."AmericanJournalof Sociology100:
40-69.
Sabel, Charles.1990."M6bius-Strip Organizations and Open Labor Markets:Some
Consequences oftheReintegration ofConception and Executionin a VolatileEcon-
omy."Pp. 23-54. in Social Theory fora ChangingSociety,editedby PierreBour-
dieu and JamesColeman.Boulder,Colo., and New York: WestviewPress and
RussellSage Foundation.
Sabel, Charles,and JaneE. Prokop.1994. "Stabilization throughReorganization?
SomePreliminary Implications ofRussia'sEntryintoWorldMarketsin theAge of
DiscursiveQualityStandards."Conference on CorporateGovernancein Central
Europeand Russia,WorldBank, December.
Sabel, Charles,and David Stark.1982."Planning,Politics,and Shop-FloorPower:
HiddenFormsof Bargainingin Soviet-Imposed State-SocialistSocieties."Politics
and Society11:439-75.
Sabel, Charles,and Jonathan Zeitlin.1996."Stories,Strategies, Rethink-
Structures:
ingHistoricalAlternatives to Mass Production."In WorldsofPossibility:Flexibil-
ityand Mass Productionin Western Industrialization,
editedby C. Sabel and J.
Zeitlin.Cambridge:CambridgeUniversity Press.
Schmitter, Philippe.1988."ModesofGovernance ofEconomicSectors."Manuscript.
Stanford University.
Schumpeter, JosephA. 1934. The Theoryof EconomicDevelopment.Cambridge,
Mass.: HarvardUniversity Press.
Sik, Endre. 1992."FromSecondEconomyto Informal Economy."JournalofPublic
Policy 12 (2): 153-75.
Staniszkis, Jadwiga.1991."'PoliticalCapitalism'in Poland."East EuropeanPolitics
and Societies5:127-41.
. 1993. "Ontology,Contextand Chance: Three Exit RoutesfromCommu-
nism."WorkingPaperson Centraland EasternEurope,no. 31. HarvardUniver-
sity,CenterforEuropeanStudies.
Stark,David. 1986."Rethinking InternalLaborMarkets:New Insightsfroma Com-
parativePerspective." AmericanSociologicalReview51:492-504.
. 1989. "CoexistingOrganizationalForms in Hungary'sEmergingMixed
Economy."Pp. 137-68inRemaking theEconomicInstitutions ofSocialism:China
andEasternEurope,editedbyVictorNee and David Stark.Stanford, Calif.:Stan-
fordUniversity Press.
. 1990."Work,Worth,andJustice." Centraland East EuropeWorkingPaper,
no. 5. HarvardUniversity, CenterforEuropeanStudies.
. 1992. "Path Dependenceand Privatization Strategiesin East CentralEu-
rope."East EuropeanPoliticsand Societies6:17-51.

1026
 Symposium:Stark

Stark,David, and Laszl6 Bruszt.1995."Restructuring Networksin thePostsocialist

EconomicTransformation." CornellWorkingPapers on TransitionsfromState
Socialism,no. 95.4. CornellUniversity, EinaudiCenterforInternational Studies.
Stark,David, and VictorNee. 1989. "Toward an Institutional Analysisof State
Socialism."Pp. 1-31 in RemakingtheEconomicInstitutions ofSocialism:China
andEasternEurope,editedbyVictorNee and David Stark.Stanford, Calif.:Stan-
fordUniversity Press.
Szelenyi,Ivan. 1978. "Social Inequalitiesin State SocialistRedistributive Econo-
mies."Theoryand Society1-2:63-87.
1988.SocialistEntrepreneurs. Madison:University ofWisconsinPress.
1994. "SocialistEntrepreneurs-Revisited." WorkingPaper no. 4. Univer-
sityofMichigan,International Institute.
Sziucs,Jen6.1985.Les troisEuropes.Paris:Harmattan.
Thevenot,Laurent. 1985. "Rules and Implements: Investmentin Forms."Social
ScienceInformation 23 (1): 1-45.
Voszka, Eva 1991. "Homalybolhomalyba.A tulajdonosiszerkezeta nagyiparban"
(Fromtwilight to twilight:Property changesin largeindustry).
Tdrsadalmi Szemle,
no. 5, pp. 3-12.
. 1992."EscapingfromtheState,Escapingto theState."Paperpresentedat
the Arne Ryde Symposiumon the "TransitionProblem."Rungsted,Denmark,
June.
Walder,Andrew.1994. "CorporateOrganizationand Local Government Property
Rightsin China." Pp. 53-66 in ChangingPoliticalEconomies:Privatizationin
Post-Communist andReforming Communist States,editedbyVedatMilor.Boulder,
Colo.: LynneRienner.
Webster,Leila. 1992. "PrivateSectorManufacturing in Hungary:A Surveyof
Firms."Manuscript.WorldBank, Industry Development Division.
White,HarrisonC. 1992.Identityand Control:A Structural TheoryofSocial Action.
Princeton, N.J.: Princeton University Press.
. 1993. "ValuesCome in Styles,WhichMate to Change."Pp. 63-91 in The
OriginsofValues,editedbyMichaelHechter,LynnNadel,and RichardE. Michod.
New York:Aldinede Gruyter.
WorldBank. 1993. The East Asian Miracle:EconomicGrowthand Public Policy.
Oxford:OxfordUniversity Press.

1027