Content uploaded by Maria Rost Rublee
Author content
All content in this area was uploaded by Maria Rost Rublee on Jan 09, 2015
Content may be subject to copyright.
COMPARATIVEPOLITICAL STUDIES / February 2000Dickson,Rublee / PARTY MEMBERSHIP IN CHINA
This article compares the relative importance of political capital (in the form of membership in
the Chinese Communist Party) and human capital (in the form of higher education) in urban
China. Survey data from urban China strongly support twokey elements of the intellectualNew
Class theory: intellectuals will have privileged access into the party, and the importance of
educationrelativetopoliticalreliabilitywillincreaseovertime.Thedataalsoshowhowpolitical
capital and human capital are converted into high paying and prestigious jobs. There is also
evidence of a separate path of career mobility: for the most socially prestigious jobs, a college
degree—not a party card—is the key. Economic reforms of the post-Mao era are creating a gap
betweenpoliticalpower,ononehand,andsocialprestigeandeconomicpower,ontheother.This
gap can be expected to grow as the reforms continue.
MEMBERSHIP HAS ITS PRIVILEGES
The Socioeconomic Characteristics of
Communist Party Members in Urban China
BRUCE J. DICKSON
MARIA ROST RUBLEE
George Washington University
I
naneraofeconomicreform,whyshouldpeoplejoinarulingCommunist
Party?During theirStalinist or Maoist phases, thesepartiesenjoyed anear
monopoly on the distribution of goods and resources, such as jobs, housing,
access to education, health care, and scarce consumer goods. Under these
conditions, joining the party had clear material benefits. In addition, mem
-
bership in a Communist Party had a moral appeal, at least for some: Party
propaganda promoted the concept of a better life to come, and some desired
to join the party to contribute to the party’s ideological goals. But once a
Communist Party embarks on economic reform, the picture changes. The
party typically liberalizesits control over society, and traditional communist
87
AUTHORS’NOTE:Fortheir helpful comments on earlier drafts of this article, wewould like to
thank Steven Balla, Melanie Manion, Lee Sigelman, Paul Wahlbeck, Andrew Walder, Martin
King Whyte, Langche Zeng, and several anonymous reviewers.
COMPARATIVE POLITICAL STUDIES, Vol. 33 No. 1, February 2000 87-112
© 2000 Sage Publications, Inc.
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
ideology is downplayedto promote the transition from a command economy
to a market economy. With jobs, housing, and consumer goods available
through the market, dependence on the party for distribution of those valued
commodities is dramatically reduced. The prestige of the party also suffers,
as reformist leaders criticize the excessesof past policies and the mistakes of
past leaders and as current policies create economic and political problems,
such as corruption and rising expectations for further reforms. Under these
conditions, do the benefits of party membership also diminish?
In this article, we examine the incentives to join the Chinese Communist
Party (CCP) during the reform era. During the Maoist era, gaining access to
theCCP wasthesole meansof careermobility,butpost-Mao reformscreated
newopportunitiesforpursuingcareergoals.Individualscouldseekadvanced
degrees from Chinese or foreign universities and thereby pursue careers on
the basis oftheir technical expertise, or they could engage in entrepreneurial
activities to accrue higher incomes. The CCP therefore had to compete with
theseothercareerpaths:Peopleengagedinstudy orentrepreneurialactivities
were reluctant to join the party, which would place additional demands on
their time and limit the range of acceptable behavior.
Inthiscontext,whatvaluedoesmembershipintheCCPstillhold?Wewill
focus on two related questions. First, what is the social composition of the
party? According to the “New Class” theories of communist parties (Djilas,
1957; Gouldner, 1979; Konrad & Szelenyi, 1979; Szelenyi, 1986-1987),
partymembersweredistinctfromtherestofsocietynotonlyintheiraccumu-
lationofpoliticalpowerbutintheirsocialcomposition.Despitetheirclaimto
representtheproletariat,communistpartiesdidnotreflectsocietyatlargebut
increasingly were composed of intellectual and technical elites. Previous
researchers have noted this trend in post-Mao China and have analyzed the
changing composition of party elites, in particular the rise of technocrats
(Hendrischke, 1994; Li & White, 1990), but information on the composition
of the party as a whole has been lacking. Are New Class theories generaliz
-
able to China? This is our first question.
Thesecondquestionisacorollaryofthefirst:IfNewClasstheorypredicts
who the party tries to recruit, does party membership then offer tangible
benefits not availableto the general public, such as higher paying and more
prestigiousjobs?Ifpartymembership isa formof politicalcapital, howis it
converted into economic and social dividends? In an era in which technical
expertiseandprofessionalmeritareemphasizedoverideologicalfervor,we
compare the value of party membership and educational credentials as
alternativestrategiesforcareermobility.Thiswillrevealtherelativeimpor
-
tanceof political capital andhuman capitalin contributingtostratificationin
88 COMPARATIVE POLITICAL STUDIES / February 2000
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
Chinaand allowusto compareChina’s experiencewiththat ofother socialist
countries.
THE THEORETICAL BASIS FOR PARTY HEGEMONY
What is the basis for inequality in communist political systems? Djilas’s
(1957) The New Class was an early and influential analysis of continued
political inequality in communist societies. According to Djilas, communist
parties had become a new ruling class, as despotic as the one they had over
-
thrown.Especiallyduringthetransformation andconsolidationphases ofthe
postrevolutionary communist regimes,the parties were able to dominate and
persecute their enemies, real or imagined, and were able to self-select their
own members and exclude all others. Those without the proper class back
-
groundsorpolitical connectionsfacedpolitical andeconomic discrimination
at a minimum, and in many cases overt persecution.
In China, this was the motivation for several of Mao’s campaigns against
hisownparty-state.Invariousways,the100FlowersCampaign,theSocialist
Education Movement of the early 1960s, and especially the Cultural Revolu-
tiontargeted rulingelitesat the central and local levels. Mao’s complaint that
the bureaucracy had become slow to implement new policies, unresponsive
to his own wishes, and out of touch with the wants and needs of society was
generallytrue.This complaintwasnot particulartoChina—it isthe priceofa
bureaucratized polity (Whyte, 1989). It is especially true for revolutionary
regimes that employ exclusionary recruitment policies that discriminate
against “counterrevolutionaries” and those with bad class backgrounds.
Bureaucracies that are not accountable tend not to be responsive.
As a consequence of these policies, during the Maoist years, party mem
-
bership was the primary—perhaps the only—avenue toward good jobs,
accesstoscarceresources(suchasbettereducation,travel,andvariousgoods
andservices),and socialprestige.Bythetime apersonreached adulthood,he
orshehadtodecidewhethertobecomeapoliticalactivistasapreludetojoin
-
ing the party or to remain among “the masses” (Oksenberg, 1970). Party
membershipwasthe sole path to career mobility,but itwasa mixedblessing.
Party members were vulnerable to the shifting political winds of the Maoist
era;manyofthosewhojoinedthepartywith sinceremotivessufferedterribly
during periodic leftist witch-hunts. Becoming an activist was itself a risky
strategy:Itrequired showingone’sloyalty tothe partyatthe expenseofone’s
friends and coworkers. Deciding not to become a party member was a low-
risk strategy, but it also had a low expected utility.
Djilas’s New Class theory focuses on the Communist Party as its unit of
analysis and does not differentiate the social composition of groups within
Dickson, Rublee / PARTY MEMBERSHIP IN CHINA 89
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
the party. Later works refined his argument, maintaining his focus on party
hegemonybutaddingananalysisofgroupscompetingforpower.Ascommu
-
nistpartiesmovedintothepost-Stalinistphaseofadaptation,theemphasison
utopian goals and social mobilization was replaced with economic moderni
-
zation and a reconciliation of state and society (Jowitt, 1992; Lowenthal,
1970). As a consequence, communist parties abandoned their exclusionary
recruitment policies and began to recruit large numbers of intellectuals and
technical experts. Konrad and Szelenyi argued that the circle of those who
dominatedstate-socialistsocietiesbegantochangeinthe1960s.Intellectuals
joined the ruling circle, and qualified personnel with academic credentials
replaced the “reds” in the bureaucracy. They believed that this transition was
changing the natureof the bureaucracy itself: “It was becoming less archaic,
more rationalistic; it was being intellectualized. At the same time, the intel
-
lectuals were becoming, to some extent, bureaucratized. Thus, bureaucrats
and intellectuals together were forming one new, unified dominant class”
(Szelenyi, 1986-1987, p. 106).
Thus, the New Class theory underwent an important modification:
Whereas Djilas focused on the party’s monopoly of political power as its
basis for hegemony, Konrad and Szelenyi emphasized the educational and
technical credentials of the intellectuals being brought into the party. In a
more general sense, the human capital of education was rising in value rela-
tive to the political capital of party membership. In post-Stalinist parties,
experts were gaining the upper hand over reds.
A similar transition occurred in post-Mao China. Whereas political reli-
ability, or “redness,” was emphasized for much of the Maoist era, in the
post-Mao period, the emphasis has been primarily on expertise to promote
economic modernization. The preference for redness had created a “vir
-
tuocracy,” in which people were recruited into the party and appointed into
official positions not on the basis of theirprofessional ortechnical qualifica
-
tions but on the basis of their correct political virtues (Shirk, 1982). But the
change in regime goals—from class struggle to economic modernization—
required an infusion of expertiseinto the bureaucracy, and the co-optation of
the types of intellectuals who had been persecuted during the leftist waves of
the Maoist era. What had been a virtuocracywas transformed into a meritoc
-
racy, andthe“socialistrevolutionaries” werereplacedby“partytechnocrats”
(Lee, 1991).
In the post-Mao period, the party no longer had a monopoly on career
opportunities,andtheappealofpartymembershipprobablydeclinedaccord
-
ingly (Harding, 1987). Indeed, there were three paths to career success dur
-
ing the 1980s: the “black” path (representing a cap and gown,i.e., education,
90 COMPARATIVE POLITICAL STUDIES / February 2000
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
especially at a foreign university), the “gold” path (going into business for
oneself), and the “red” path (joining the party). Most people who joined the
party during the post-Mao years did so not for ideological reasons but for
material ones (Rosen, 1990). Those who joined during the Maoist era
undoubtedlyalsohad mixedmotives,butmaterial motivesbecamemorepro
-
nounced in the post-Mao era, when ideology and political reliability were
de-emphasized as criteria for membership.
Konrad and Szelenyi’s intelligentsia New Class hypothesis was widely
influential. Its philosophical rationale and evidence derived from their own
experiences corresponded nicely with wider trends of the time. However, it
was the discovery of survey data after the collapse of communism that
allowed a more rigorous test of their hypothesis. For instance, Wong (1996)
found important differences between the Hungarian and Czechoslovak par
-
tiesthat werenotentirely consistentwith the intelligentsiaNewClass theory.
Although a university education greatly improved a person’s chances of
being a party member in Hungary, in Czechoslovakia it improved a person’s
chancesofbecomingapartyfunctionary(although muchlesssothaninHun-
gary)but not of a rank-and-fileparty member. In both countries, the relation-
ship between education and party membership (both for functionaries and
rank-and-filemembers)did notvaryacross agecohorts. FromWong’sanaly-
sis, it appears that Hungarian intellectuals of all generations had a smoother
road to traveltowardclass power than did their counterparts in Czechoslova-
kia. This indicates that the intelligentsia NewClass theory may not be gener-
alizable to all communist parties. The evidence was strongest for Hungary,
whose ruling party was exemplary in its inclusion of intellectuals (and was
the basis for Konrad and Szelenyi’s argument), but much weaker in Czecho
-
slovakia, where more orthodox Leninist norms preserved the value of politi
-
cal loyalty over expertise virtually until the demise of the Czech party. Bian
(1994) found that education did not improve a person’s chances of attaining
partymembershipintheMaoistera,butinthepost-Maoerathosewithhigher
education have a better chance at gaining party membership. This suggests
that “market reforms did alter the basis of political mobility....Theparty
now is interested in those with greater human capital (education) and less
focused on criteria of ideological commitment” (p. 139).
The availability of survey data also allows a complementary analysis:
How is the political capital of party membership and the human capital of
educationconverted into career dividends?Doparty membership and educa
-
tion create social and economic inequalities? More to the point, in the post-
Stalinist (or forChina, post-Maoist) phase, is the value of party membership
devalued relative to education and expertise, givingpeople more meaningful
Dickson, Rublee / PARTY MEMBERSHIP IN CHINA 91
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
options in their career choices? By now, it is an established fact that even in
communistsocietiesall people are not created equal. Despite theabolitionof
private property and the promotion of egalitarian policies and norms, social
and economic inequality continued to exist in all communist countries. For
our purposes, it is not enough to find that political and human capital created
inequalityin jobsand income. Ifthe intelligentsiaNewClass theoryis gener
-
alizable to China, we would expect that education would have a greater
impact than party membership alone, and that its impact would increase for
younger cohorts. Using the same Hungarian data as Wong (1996), Borocz
and Southworth (1996) found that the value of a university degree was
roughlythreetimeshigherthanthatofpartymembership.Therewassubstan
-
tial overlap between these two variables, suggesting that intellectuals had
been recruited into the Hungarian party and lending credence to Konrad and
Szelenyi’stheory.SurveysfromChinaalsofoundthateducationhadagreater
impact on a person’s income than did party membership (Walder, 1990,
1995a; Xie & Hannum, 1996).
In the third section of this article, we will test the intellectual New Class
theorywithdatafromanationwidesurveyofurbanChina.Then,inthefourth
section, we will examine the relationship between political and human capi-
tal, on one hand, and income distribution and job prestige, on the other.
Before turning to that analysis, we will begin with a brief introduction to
changes in CCP recruitment policies and their implications.
A SHORT COURSE ON PARTY BUILDING IN CHINA
WhentheCCPcametopowerin1949,itsmembersweregenerallyyoung,
poorly educated, of peasant background, and more experienced in fighting a
guerilla war than in administering government programs. Therefore, the
partyundertooktwo missions. One wasto expandits membership to staffthe
party and government bureaucracies in Beijing and nationwide. The other
was to upgrade the technical and professional qualifications of its members
to handle the new task of governance. In practice, these two missions proved
tobeincompatibleandwerepursuedsequentiallyratherthanintandem.Dur
-
ing periods of Maoist radicalism, the party recruited large numbers of politi
-
calactivists withmass mobilization skills, and the required qualificationsfor
party membership were sacrificed for the sake of meeting quantitative tar
-
gets. In contrast, during periods of normalcy, when the party attempted to
restore social order and revive economic production disrupted by the previ
-
ous radical upsurge, it froze new recruitment, weeded out unqualified mem
-
bers, and retrained other members with the requisite skills. For instance,
92 COMPARATIVE POLITICAL STUDIES / February 2000
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
during the Great Leap Forward (1958-1960), the CCP grew by almost 6.5
million members, an increase of more than 50% in just 3 years. During the
period of recovery (1960-1964), the party grew by only 600,000 and also
expelled large numbers of local cadres (i.e., party and government officials
with decision-making authority) and members recruited in recent years.
As the Cultural Revolution got under way in 1966, the party once again
began recruiting large numbers of new members, most of them for their
ideological and political reliability rather than their professional skills. As a
result, the party doubled in size during the Cultural Revolution decade
(1966-1976), growing from 18.7 million to 35.1 million members (Dickson,
1997, pp. 78-103).
At the time of Mao’s death in September 1976, the CCP wasbeset by sev
-
eralorganizationalandpersonnelproblems.Notonlywasitbedeviledbyfac
-
tional strife at all levels, but its members were generally too old, too poorly
educated, and too politically radical to be able or willing to implement the
package of reforms identified with Deng Xiaoping. Consequently, the CCP
implemented new criteria for the recruitment of members and the appoint-
ment and promotion of cadres, emphasizing youth and education. Between
1982 and 1990, the average age of party and government officials at the pro-
vincial, prefecture, and county levels declined from 62 to 55. Thepercentage
of officials at these three levels with a college education rose to 73.6, 66.5,
and 64.2, respectively, more than tripling the 1982 levels (Renmin ribao,
1991).Amongallpartymembers,thepercentagewithahighschoolorhigher
educationrosefrom12.8in1978to 30.4in1988 andto39.9 in1994(Xinhua,
1989, 1992, 1995).
In1988, there were 48 millionparty members in a totalpopulation of 1.11
billion, about 4.3% of the total population. The Chinese Household Income
Project(CHIP)surveydata (describedbelow)providea moredetailedlook at
the composition of the CCP in urban China in the late 1980s (see Table 1).
TheCCP isnot arepresentativecross sectionof Chinesesociety.Membersof
the CCP are generally older, better educated, and more typically male than
nonmembers. To be precise, the typical (but hypothetical) party member in
1988 was a 46-year-old male with 11.5 years of education (i.e., senior high
school) who was an office worker. In contrast, the typical (but equally hypo
-
thetical) nonmember was a 39-year-old female with 9.7 years of education
who held the less-prestigious job of a laborer. These two typical people
shared several attributes: They both were Han, the dominant ethnic group in
China, and had permanent jobs in the state sector. The probability of being a
party member increases steadily with education, so that an urban resident
with a university education has an almost 50% chance of belonging to the
Dickson, Rublee / PARTY MEMBERSHIP IN CHINA 93
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
CCP. Urban men are also 3 times more likely to be party members than are
women.
94 COMPARATIVE POLITICAL STUDIES / February 2000
Table 1
Profile of Communist Party Members in Urban China (numbers in cells are percentages)
Within-Group Comparison Between-Group Comparison
(columns sum to 100) (rows sum to 100)
CCP Non-CCP CCP Non-CCP
Gender
Male 75.11 43.38 33.75 66.25
Female 24.89 56.62 11.45 88.55
Age
18-30 4.60 30.34 4.27 95.72
31-40 28.43 30.59 21.50 78.50
41-50 32.32 19.79 32.46 67.54
51-60 27.49 12.18 39.90 60.10
> 60 7.10 7.09 22.76 77.24
Education
< 3 years primary 0.89 4.57 5.38 94.62
3 years primary 1.20 2.35 13.04 86.96
Primary graduate 10.03 14.39 16.98 83.02
Junior high 29.57 38.19 18.52 81.48
Senior high 18.64 23.97 18.58 81.42
Professional school 15.15 8.52 34.30 65.70
Community college 12.07 4.23 45.58 54.52
University 12.45 3.77 49.25 50.75
Job type
Laborer 15.75 65.47 7.03 92.97
Office worker 38.96 19.19 38.97 61.03
Factory manager
or director 5.81 0.55 76.88 23.13
Official 16.65 1.01 83.83 16.17
Professional or technical 22.83 13.79 34.24 65.76
Sector
Private 0.09 1.45 2.00 98.00
Collective 7.93 24.28 9.30 90.70
State 91.76 73.76 28.04 71.96
Nature of employment
Temporary 0.80 2.47 9.21 90.79
Contract 0.14 1.55 2.78 97.22
Permanent 99.05 95.98 24.33 75.67
Ethnicity
Han 95.54 96.51 22.26 77.44
Minority 4.46 3.49 27.35 72.65
Source: China Household Income Project data.
Note: CCP = Chinese Communist Party.
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
It is generally believed that party members are concentrated in the most
influential and most prestigious jobs. As seen in the profile of urban party
members, the share of party members is lowest in the least prestigious jobs
and rises as job prestige increases, except for the most prestigious category,
where the percentage of party members drops sharply. (This drop provides a
clue to the issue of alternative career opportunities, which we will turn to
shortly.) This supports the conventional wisdom, but the strength of the rela
-
tionship between party membership and certain jobs is surprising. Although
itwasgenerallybelievedthatthepartycontrolledthemostpowerfulandpres
-
tigiouspositions,the CHIPsurveydatareveal thatalmost 84%ofresponsible
officials and77% of factorymanagers anddirectors are partymembers.Con
-
versely, party members are dramatically underrepresented among laborers;
whereas almost one quarter of the survey respondents were party members,
fewer than 7% of laborers are party members. This latter point validates a
change in the party’s recruitment strategy. Around the time that this survey
wasconducted,theCCPbeganpublicizingtheconcernthattoo muchempha-
sis was being given to professional and technical criteria, especially level of
education, at the expense of political criteria; in particular, concern was
expressed that farmers and workers on the “front line of production”—the
party’s traditional base—werebeing ignored (DongWanmin, 1991; LiYanxi,
1989; Renmin ribao, 1990; Zhou Peng, 1989). The survey data bear out this
concern. Although communist parties typically co-opt intellectuals during
their postmobilization phase, this shift in recruitment policy encountered
resistance by the reds, whose political virtue and ideological commitment
were prized in the past but were less valued following the change in party
goals.
1
A “NEW CLASS” IN CHINA?
To assess the generalizability of the intellectual New Class theory to
China, we use data from the 1988 CHIP survey conducted in 1989 bya team
of American, British, and Chinese scholars. Interviewees were members of
9,009urbanhouseholdsrepresenting31,827individualsin10provinces.The
survey was conducted by using a multistage probability sample to be repre
-
sentative of urban conditions in different regions of China and in cities and
Dickson, Rublee / PARTY MEMBERSHIP IN CHINA 95
1. In his reflection on his Intellectuals on the Road to Class Power, Szelenyi (1986-1987)
acknowledges that he did not anticipate the counteroffensive of party functionaries against
newly co-opted intellectuals, and that the “intelligentsia New Class project” was not as inevita
-
ble nor as enduring as he originally imagined. Jowitt (1992) correctly noted the tension between
a Leninist party’s goalsof mobilization and inclusion, but theimplication of his argument—that
inclusion would lead to the liberalization of the regime—was not realized in practice.
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
townsof varyingsize.
2
Forour analysis, we restrict the sample to those older
than 18 years, the minimum age of eligibility for joining the party, and
exclude those who did not report whether they were party members. This
leaves 4,758 party members in an urban adult population of 20,927—that is,
22.7% of the adults covered by the urban survey are party members.
This survey represents a snapshot of the party at one point in time. We do
not have time-series data that would allow us to show trends over time. The
main advantageof the CHIP survey data is that this was the firsttruly nation
-
widesurveyofits kind.Walder’s(1990,1995a)researchbased ona1986 sur
-
veyfromthecityofTianjinaddressessimilarquestionstotheonesposedhere
and will be an important reference point for our analysis and conclusions.
Other than the works already cited, there is little comparable research from
China. After a wave of survey fever in the mid- and late 1980s, the CCP lim
-
itedthe types oftopicsthat could beaddressed and thetypesof questions that
could be asked. Political attitudes were generally off-limits and questions
about the party in particular were not allowed (Rosen, 1991). We know of at
least one survey that was confiscated after it was conducted; it was returned
only after responses identifying party members had been erased. Conse-
quently,wecannotbesurethatourfindingsarestillvalidtoday.Nevertheless,
we are sure that our findings offer new insights into the composition of the
party on one hand and individualcareer choices on the other, and will remain
the richest source of such insights until other surveys become available.
TheintellectualNewClasstheoryassertsthatintellectualshaveprivileged
access to the party and especially to positions of authority. Therefore, we
would expect that those with higher levels of education would have a better
chance of being party members and that within the party, cadres would have
higher levels of education than rank-and-file members. We construct our
dependent variable, party membership, by coding party members who work
as laborers, office workers, and professional and technical workers as rank-
and-file members, and party members who work as managers or officials as
cadres. Our main independent variables concern education, which the CHIP
data set codes on an 8-point scale, ranging from less than 3 years of primary
school up to university education. We created dummy variables for high
school(including the categories of senior high school and professional train
-
ing school) and college (both community college and university). As control
variables, we use age and gender (a dummy variable, with female=0and
male = 1). The CHIP data set lacks important social-background variables,
suchasfather’seducation,employment,andpartymembership,aswellasthe
96 COMPARATIVE POLITICAL STUDIES / February 2000
2. Formore detailson the survey’s methodology, see Eichen and Zhang (1993). The data set
and codebooks are available through the University of Michigan’s Inter-University Consortium
for Political and Social Research.
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
family’s class status. Particularly during the Maoist era, family background
was a major consideration in an individual’s life opportunities. A true test of
the intellectual NewClass theory would include these variables,but unfortu
-
nately they are not available to us.
Theresults ofour testare reportedin Table2. Thesurveydatashowstrong
support for the intellectual New Class theory: The coefficients for college
and high school are positive and significant,those for collegeare higher than
for high school, and those for cadres are higher than for rank-and-file mem
-
bers. This is exactly what we would expect to find,and we can reject the null
hypothesis that education is unrelated to party membership. In addition, the
control variables of age and gender are also positive and significant. Older
people are more likely to be party members than are younger people and are
more likely to be cadres than rank-and-file members. In a society known to
veneratetheold,itisnosurprisethatseniorityisanadvantageingainingaccess
to the party and to cadre posts in particular. The survey data also reveal the
advantages of being male: The coefficients for male are large and significant,
andtheoneforcadresistwicethatforrank-and-filemembers.Indeed,thecoef-
ficient for male is even higher than that for college education among cadres.
This initial test of the intellectual New Class theory indicates that its
expectations are borne out in China. But has this always been the case? The
theory also predicts that the importance of education increases over time,
especially after the party abandons class struggle and utopian policies and
concentrates on economic modernization. Because we do not have time-
series data, we use a cohort analysis to simulate the effects of time. As
reported in Table 3, this expectation is also generally met. Coefficients for
high school are small and insignificant for the youngest cohort, highest for
the next youngest, and then decline. This finding is consistent with both the
New Class theory and the priorities of post-Mao recruitment. Education is
now emphasized, and for urban youths a high school education is no longer
sufficient for joining the political elite. The coefficients for college offer an
even sharpercontrast: largeandsignificantfortheyounger cohortsand much
smaller (but still significant) for older cohorts. In short, the cohort analysis
follows the expectations of the New Class theory.
One anomaly emerges from this analysis: The age-group 31-40 gets the
biggest boost from education, and yet this is the cohort that represents the
“Cultural Revolution generation.”
3
This cohort did not have the same access
to higher education as did others, but it appears that those who did receive a
Dickson, Rublee / PARTY MEMBERSHIP IN CHINA 97
3. Cohort groupings were selected so that any possible effects of the Cultural Revolution
(1966-1976) could be noted. The 31-40 age-group represents the generation most likely to be
affected by school closings and loss of career opportunities, because they were between 10 and
19 years old at the beginning of the Cultural Revolution, and 20 to 29 by its end.
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
higher education (perhaps only during the post-Mao era) greatly improved
theirchancesofbecoming party members. As we will see in the nextsection,
however,it isthepolitical capitalof partymembership,not thehuman capital
of education, that most accounts for this cohort’s ability to obtain high-
paying and prestigious jobs.
CONVERTING POLITICAL AND
HUMAN CAPITAL INTO MATERIAL BENEFITS
To promote economic modernization, the CCP loosened its control over
boththeeconomyandsociety,andde-emphasizedthesalienceofparty ideol
-
ogy. As the reform era progressed, the party’s control over scarce resources
and its prestige diminished (Rosen, 1990; Tong, 1997; Walder,1995b). And,
yet, the party continued to grow steadily. Does party membership confer
material privileges not similarly available to nonmembers? Has the extent of
these privileges changed over time? Using the CHIP survey data, we will
examine the changing nature of rewards offered to Communist Party mem
-
bers as well as how political capital (party membership) and human capital
(education) are converted into career advantages. Because the data are not
longitudinal,butonlyfromonepointintime,cohortgroupingswillbeusedto
98 COMPARATIVE POLITICAL STUDIES / February 2000
Table 2
Multinomial Logit Regression of Determinants of Party Membership
(numbers in parentheses are standard errors)
Rank and File Cadre
Age 0.036 0.074
(.012) (.02)
Male 1.097 2.113
(.074) (.155)
College 0.915 1.368
(.099) (.137)
High 0.441 0.618
(.099) (.158)
Constant –3.43 –7.258
(.568) (.925)
χ
2
697.5
R
2
.09
N 4,717
Note: All coefficients are significant at the .001 level (one-tailed tests).
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
99
Table 3
Multinomial Logit Regression of Determinants of Party Membership by Cohort (numbers in parentheses are standard errors)
18-30 31-40 41-50 51-60
Rank and File Cadres Rank and File Cadres Rank and File Cadres Rank and File Cadres
Age 0.244*** 0.650*** 0.080*** 0.148*** 0.036** 0.074*** –0.0354** 0.045*
(.025) (.169) (.013) (.028) (.074) (.02) (.015) (.022)
Male 1.023*** 2.001** 1.152*** 1.667*** 1.097*** 2.113*** 1.326*** 2.865***
(.161) (.703) (.073) (.187) (.074) (.155) (.09) (.231)
College 1.431*** 2.097** 1.632*** 2.298*** 0.915*** 1.368*** 0.936*** 0.914***
(.191) (.703) (.095) (.173) (.099) (.137) (.122) (.17)
High 0.095 0.363 0.570*** 0.725*** 0.441*** 0.618*** 0.331** 0.685***
(.177) (.734) (.087) (.206) (.099) (.158) (.13) (.171)
Constant –10.385** –25.582** –5.322*** –10.177 –3.430*** –7.258*** .273 –1.657
(.699) (4.994) (.467) (1.037) (.568) (.925) (.807) (1.22)
χ
2
291.05 866.59 697.5 611.22
R
2
.1523 .1133 .0912 .1043
N 5,107 6,278 4,717 3,251
*p < .05. **p < .01. ***p < .001 (one-tailed tests).
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
estimate the effect of different policies over time. Using cohorts to substitute
for time-series data,however,is somewhat problematic, because differences
between cohorts could be due to age effects instead of or as well as cohort
effects. For example, the data could indicate that younger cohorts receive
fewerbenefitsfrom party membership than do older cohorts. This difference
could be due to cohort effects: The value of party membership has declined
over time. It could also be due to age effects: As people age, the benefits of
party membership rise, thus making party membership more valuable for
older cohorts. Although discriminating between cohort and age effects is not
alwayseasy,withouttime-series dataitis theclosestapproximation possible.
In addition, we will make note of any possible confounding age effects as
they arise.
INCOME
One motive for joining the CCP may be that party members enjoy higher
standards of living than the rest of the population. Indeed, an enduring com-
plaint about the CCP (going back to the pre-1949 days in Yanan and continu-
ingtothepresent)isthatitsmembers,andhigh-rankingofficialsinparticular,
leadcomfortablelives,whereastherest ofsocietyexperienceshardship(Dai,
1994; Meaney, 1991; Oi, 1991). The CHIP survey data support this general
perception. If we look at the main components of a person’s total monthly
income (regular wages, bonuses, and various subsidies, and not including
pensions or supplemental income for retirees), the average party member
earned 26.5% more than the average nonparty member in 1988 (191 yuan
compared to 151 yuan, a difference of 40 yuan [t = –8.34, p <.0001, df =
17, 316]).
4
But this simple comparison does not take account of other factors
that influence income. Party membership itself is dependent on such factors
asage,education,andgender,allofwhichhelpdeterminebothincomedistri
-
bution and the probability of being a party member.
Followingpreviousresearchon incomedistributioninChina,we examine
regularwages and total income separately to determine whether the explana
-
tory variables have similar impacts on the different components of income.
Regularwages account for roughly half of the total income, with the remain
-
der coming from bonuses and subsidies. We use the natural logs of regular
wages and total income to discount the outlying cases of extremely high
income.
100 COMPARATIVE POLITICAL STUDIES / February 2000
4.Our discussion ofhigher incomefor partymembers does notinclude incomederivedfrom
bribesorother corruptbehavior, awidespreadbutdifficult-to-measurephenomenon. The poten
-
tialto engage in corruption may bean incentivefor becoming a party or governmentofficial, but
not for rank-and-file members.
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
Table4 summarizesthe impactsof personaland job-relatedattributesona
person’sregularwage.Ourfirstmodelincludespartymembershipasaninde
-
pendentvariable.Inthis case,acollegeeducationhasagreaterpositiveinflu
-
ence on regular wages than does party membership. To more precisely test
the relative importance of political and human capital, we break party mem
-
bership into two groups: cadres and rank and file. As Model 2 shows, a col
-
lege education is more important to regular wages than rank-and-file mem
-
bership—but not as important as a cadre position.
Partymembershipdoesindeedhavemonetaryrewards,especiallyforcad
-
res.Ordinaryleastsquares(OLS)coefficientsrevealthatafterage
5
andwork
-
ing in a joint venture, being a cadre had the next largest positive impact on
regular salary. As noted, rank-and-file members received less of a benefit.
Given that its coefficient is less than that of college, this indicates that a col
-
lege diploma may add more to one’s income than simple party membership.
This is not the case with a high school diploma: The coefficient for high
school was small and not statistically significant, indicating that it does not
affect one’s income. Nonetheless, the independent effect of a party card,
whether a cadre or regular member, is to increase one’s regular wages. Men
earn more than women, as shown by the positive and statistically significant
coefficient for male.
6
The value of CCP membership to total income follows a similar pattern.
Aswithregularwages,partymembership doesprovidemonetarybenefitsfor
a person’s total income, which includes bonuses and subsidies as well as sal-
ary. As opposed to regular wages, however, the coefficients for cadre and
rank and file are almost equal, suggesting no difference in total income
between cadre status and rank-and-file membership (we will return to this
anomaly below). Also in contrast to regular wages, the college coefficient is
larger than that for either cadres or rank-and-file members, indicating that a
college degree is more important to accessing bonuses and subsidies than is
partymembership. TheOLS analysisof totalincome revealsanotherpointof
interest: The independent variables explain more of the variation in regular
salary than total income. Our regression model explains almost 46% of the
Dickson, Rublee / PARTY MEMBERSHIP IN CHINA 101
5. Unlike the otherdichotomous variables in our model (e.g., either you are a party member
oryouarenot),ageis measured continuously.To see theimpactofaperson’sageonincome,you
must multiply the coefficient by the person’s age. For any adult, seniority is the most important
determinant of income.
6. The large and negative coefficient for those working in the private sector may seem sur
-
prising, but at the time of the survey, private entrepreneurs in urban areas were generally inde
-
pendent vendors (Young, 1994). In the years after the survey, the category of private entrepre
-
neurs has grown to include the owners and managers of larger enterprises, especially in the
industrializing rural areas.
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
variation for regular wages, but only 25% of the variation for total income.
This finding is consistent with previous research (Griffin & Zhao, 1993;
Walder, 1990,1995a;Xie &Hannum, 1996), andfor thesamereason. Wages
inChinaare distributed onamore or less rational basis (those with more sen
-
iority and education earn more), but bonuses and subsidies are distributed
lesssystematically.Bonuses andsubsidies arelessdependentonanindividu
-
al’s performance or personal attributes than on the profitability of the firm
itself, which, in turn, is less dependent on performance than on irrational
prices and soft budget constraints, for example, easy access to government
subsidies and loans that typically do not get repaid.
Tounpackthe effectofeconomicreform on thebenefitsof party member
-
ship, we ran identical regressions by cohort groupings. All cohorts receive
102 COMPARATIVE POLITICAL STUDIES / February 2000
Table 4
Ordinary Least Squares (OLS) Regression of Monthly Income
(numbers in parentheses are standard errors)
Regular Wage Total Income
Model 1 Model 2 Model 1 Model 2
Membership .077*** .078***
(.006) (.008)
Cadre .109*** .079***
(.010) (.014)
Rank and file .069*** .078***
(.006) (.008)
Male .064*** .064*** .067*** .067***
(.004) (.004) (.006) (.006)
Age .021*** .021*** .017*** .017***
(.000) (.000) (.000) (.000)
College .096*** .096*** .091*** .091***
(.007) (.007) (.009) (.009)
High .008 .007 .022*** .021***
(.005) (.005) (.007) (.007)
Joint .365*** .364*** .282*** .281***
(.036) (.036) (.050) (.050)
Collective –.111*** –.111** –.144*** –.144***
(.005) (.005) (.008) (.008)
Private –.378*** –.378*** –.480*** –.480***
(.067) (.067) (.093) (.093)
Constant 3.541*** 3.545*** 4.288*** 4.288***
(.009) (.009) (.013) (.013)
Adjusted R
2
.457 .458 .250 .251
*p < .05. **p < .01. ***p < .001 (one-tailed tests).
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
higher wages and total income due to CCP membership—both cadre and
rank and file (see Figures 1 and 2). Whereas the difference between cadres
andrank-and-filememberswasnotapparent whenall elsewasheld constant,
thecohort analysisshowsthat cadrestatus does provideanobviousboost toa
person’s total income. What is unexpected is the large benefit of the cadre
position for the youngest cohort. If the reform period has de-emphasized
ideological purity and has stressed technical expertise and education, why is
it that the group being educated and finding their first jobs during this period
receives the largest salary boost from the cadre position? Is political capital
becoming more important? We surmise that the answer is less startling.
Because so few people in the youngest cohort are cadres (only 28 of 4,854
[0.6%]), these luckyfewreceive a comparatively large boost to their income.
As their peers age and more attain cadre status, their comparative benefit
decreases. The age-group 31-40 receives the next largest benefit due to a
cadre position. This may be ascribable to age effects as just described. An
additional factor may be due to cohort effects: During the Cultural Revolu
-
tion (during which this group was educated and first entered the labor mar
-
ket), political purity was valued most, and those in political positions were
rewarded the most.
Therelativeimportance of political capital and human capital over time is
also revealed in Figures 1 and 2. Cadre status has a larger impact on regular
Dickson, Rublee / PARTY MEMBERSHIP IN CHINA 103
Figure 1. Cohort analysis of key determinants of regular wage.
Note: Numbers on Y axis refer to ordinary least squares (OLS) coefficients.
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
wages and total income than a college education for the youngest two
cohorts, buta smaller impact for the older cohorts. The full impact of college
educationislikelymutedbyageeffects:Manyintheyoungestcohort(18-30)
likely have not entered or finished college, and many of those who have may
nothavebeenabletousetheirdiploma tosecurea job.Forthiscohort, thefull
valueofa collegeeducation hasnot yet been realized.Collegeeducation also
has an equal or greater impact on wages and total income than does rank-
and-file party membership, except for the Cultural Revolution cohort. This
finding suggests again that political, rather than human, capital is most
importantto thegroup that waseducatedand firstgainedemploymentduring
the Cultural Revolution.
Our analysis in this section offers two key insights into the question of
whether party membership continues to provide tangible benefits. First,
political capital does still translate into monetary benefits in the People’s
Republic of China (PRC), and those with more capital (cadres) benefit more
thanregularmembers. However,humancapitalisalsoimportant,providinga
greaterbenefitthan regularmembership butless thanthat of acadre position.
Second, cohort effects are visible in our analysis, showing that the Cultural
Revolution cohort is more dependent on political than human capital.
104 COMPARATIVE POLITICAL STUDIES / February 2000
Figure 2. Cohort analysis of key determinants of total income.
Note: Numbers on Y axis refer to ordinary least squares (OLS) coefficients.
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
PRESTIGE
Asecondmaterialincentivefor joiningtheCCP maybeaccess tothemost
powerful and prestigious jobs. For this issue, we ran a multinominal regres
-
sion of job type for the entire sample and by cohort to determine to what
extent party membership and personal characteristics determine job type. In
this case, jobtype has five possible values, ranked in terms of socialprestige
as measured by a 1983 survey in Beijing (Lin & Xie, 1988). Whereas that
study ranked 50 occupations, the CHIP survey has only five categories. We
consider the professional and technical category to be the most prestigious
becausethe10mostprestigiousoccupationsin theLin andXie studyfallinto
thiscategory.Themultinominallogit resultsofjobprestige (Table5),however,
indicate that although the coefficient for membership in the professional/
technicalcategoryis positive, itis smallerthan for every otherjob type.From
Table1,weseethatonly34%ofthepeopleintheprofessional/technicalcate
-
gory are party members, compared with nearly 84% of responsible officials
and 77% of managers. If conventional wisdom suggests that party members
have access to more prestigious jobs, how do we explain the data suggesting
thatjustmorethanonethirdofthepeopleinthemostprestigiousjobcategory
do not belong to the party?
Earlier, we mentioned that the economic reforms of the post-Mao era
opened up newavenues for career mobility besides joining the party. What is
the relative worth of collegeeducation (the “black” path) in the attainment of
prestigious jobs? Although the coefficient for college is positive and signifi-
cantforallfourjobtypes(withlaborersas thecomparisongroup),a compari-
son with party membership is telling. For both responsible officials and fac-
torymanagers,partymembership is roughly equal in importance to a college
education. For an office worker, however, a college education is 1.5 times
more important than party membership. For the most prestigious group, pro
-
fessional/technical,acollege degreeis more thanthreetimes more important
than a party card. Thus, the regression results confirm the existence of two
paths: (a) party membership for party jobs and (b) a college education for
technical and professional positions.
Is the existence of the two paths a new occurrence, perhaps due to the
reforms instituted by Deng? The cohort analysis of Table 6 indicates that
this may not be so. In this table, we estimate how college education and
party membership influence a person’s likelihood of holding a particular
job. A college degree makes a person roughly 7 times more likely to be in
the professional/technical category than is a person with “benchmark” char
-
acteristics (a male, nonparty member, with less than college education), and
this advantage is true for all cohorts. In contrast, party membership alone
Dickson, Rublee / PARTY MEMBERSHIP IN CHINA 105
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
roughly doubles the chances of being in this category for the youngest
cohortsbuthasamuchsmallerimpactforoldercohorts.Partymemberswith-
out a college education in the 41-50 age-group are 28% more likely to be in
this job category than are nonmembers, and those in the 51-60 age-group are
only 13% more likely. Having both a college education and a party card
makes a person much more likely to be in this category than a person with
only a party card, butless likely than a person with only a collegedegree. For
the professional/technical category, human capital is clearly the most impor
-
tant consideration, and for all cohorts. In contrast, the combination of both
human and political capital is the key to becoming a responsible official. A
college degree increases a person’s chances of becoming an official, but not
asmuchas partymembership,and thepeoplemost likelytofallintothiscate
-
goryarethosewithbothacollegedegreeandapartycard.Forthosewhohave
both a party card and a college diploma, the advantage is highest for the two
youngest cohorts and much less for the older cohorts, reflecting the reform
era’s requirement that cadres must be both red and expert. Although we will
need time-series data to provide a more definitive test, our cohort analysis
suggeststhattheexistenceoftheredandblackpathspredatesthereformera.
106 COMPARATIVE POLITICAL STUDIES / February 2000
Table 5
Multinomial Logit Coefficients for Determinants of Job Type
(numbers in parentheses are standard errors)
Professional/
Office Manager Official Technician
Member 1.877*** 3.020*** 3.413*** 1.417***
(.055) (.145) (.108) (.006)
Male –0.008* 1.113*** 0.632*** –0.492***
(.042) (.179) (.108) (.053)
Age 0.034*** 0.069*** 0.096*** 0.064***
(.002) (.006) (.005) (.003)
College 2.78*** 3.024*** 3.665*** 4.331***
(.110) (.180) (.136) (.109)
High 0.637*** 0.592*** 0.708*** 0.332***
(.048) (.151) (.109) (.067)
Constant –2.730*** –8.383*** –8.754*** –4.290***
(.088) (.319) (.230) (.114)
Pseudo R
2
.201
Number of observations 17,589
*p < .05. **p < .01. ***p < .001 (one-tailed tests).
Note: “Laborer” is the comparison group.
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
The multinominal regression also highlights the effect of gender on job
prestige. The coefficients for male are positive and large for managers and
responsible officials, indicating that men are more likely to hold these posi-
tions. Conversely, for office worker and professional/technical, the coeffi-
cients for male are negative, indicating that women are more likely to hold
thosepositions.Howisitthatwomenaremorelikelytoholdthemostprestig-
ious types of jobs in China—those in the professional/technical category?
This category includes scientists, doctors, nurses, teachers, lawyers, engi
-
neers, and artists. The breadth of this category does not allow us to distin
-
guishbetween university professorsand primaryschool teachers,or between
physicistsandpoets.
7
Becauseofthelargernumberofwomeninthisjobcate
-
gory, the coefficient for male is negative. However, analysis of wages and
education of people within this job type indicates that men hold the higher
Dickson, Rublee / PARTY MEMBERSHIP IN CHINA 107
Table 6
Estimated Probabilities of Job Type, by Cohort (numbers in cells are percentages)
18-30 31-40 41-50 51-60
Estimated probability of holding
professional and technical jobs
Benchmark 5.8 7.0 10.5 9.9
College only 39.1 49.5 70.9 76.3
Chinese Communist Party (CCP) only 10.3 12.6 13.5 11.1
College and CCP 26.9 36.2 45.7 54.2
Estimated probability of being a
responsible official
Benchmark 0.1 0.1 1.8 3.5
College only 0.3 4.0 6.2 6.2
CCP only 1.3 6.9 17.1 25.2
College and CCP 2.9 15.7 29.7 28.2
Note: Benchmark probabilities are calculated using coefficients from the multinomial logit
model of Table 5,with “male” setto 1 and “member” and“college” set to 0 (i.e., the benchmark
referstoanurbanmalewhoisnotapartymember and has less thanacollegeeducation).Wethen
estimatethechanged probability ofbeingeither aprofessional/technicalora responsibleofficial
by setting “member” and/or “college” to 1.
7. Using a 1986 survey from Tianjin, Walder (1995a) distinguished between “high” profes
-
sionals (scientists, researchers, professors, engineers, lawyers, and doctors) and “low” profes
-
sionals (primary and secondary schoolteachers, nurses, and technicians). His analysis showed
thatgender,seniority, and educationhadsimilar impacts on“high”and “low”professionals. The
variables with dissimilar impacts (especially party membership) were not statistically signifi
-
cant. Although the Chinese Household Income Project (CHIP) data do not allow us to distin
-
guishbetween high and low professionals, Walder’s analysis indicatesthat this distinctionis not
essentialfor the questionat hand:howpartymembership andcollegeeducation affecta person’s
attainment of prestigious jobs.
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
payingpositionsandhavemoreeducationthanwomeninthiscategory.Thus,
what seems likean indicator of women’s liberation (women holding more of
the best jobs) is not so, because hierarchical discrimination within the group
still exists.
CONCLUSION
The presence of intellectuals in a communist party is a deliberate choice.
The party attracts the best educated into the party to relegitimize its rule
among key social sectors and to co-opt skills needed for new policy initia
-
tives. Theseascriptivepoliciesplace greater emphasison humancapitalrela
-
tive to political capital and thereby diminish the value of party membership
per se and increase the value of higher education. This may create a conflict
withintheparty(andamong recruitmentgoals) betweenmaintaining aprole
-
tarian identity in line with party traditions and attracting the technical and
administrative specialists needed for modernization. In his reflection on his
New Class theory, Szelenyi (1986-1987) noted that intellectuals and techni-
cians did not continue to increase their power and influence as he had origi-
nally expected. Party cadres, more concerned with preserving their hegem-
onythan withthe bestinterests ofthe socialistsystem, usedvariousmeasures
to nullify, to discredit, and to exilethe intellectuals. In recent years, there has
been a similar debate in China regarding the propriety of havingintellectuals
dominate the party at the expense of workers and peasants, the traditional
core of the party. Because China’s top leaders have been more committed to
economic reform and the promotion of market mechanisms than were Czech
or East German leaders, the counteroffensive against intellectuals in China
has been less effective. However, we must await further survey data of the
type analyzed in this article to demonstrate this empirically.
InChina,intellectualsaregenerallydefinedasanyonewithadvancededu
-
cation. Our analysis shows that party members are more likely to have high
schoolandcollegeeducationsthanthegeneralpopulation,andthatcadresare
more likely to be better educated than rank-and-file members. The cohort
analysis lends further credence to the expectations of the New Class theory.
Younger cohorts are generally better educated than older cohorts, and in the
case of cadres the difference is particularly sharp. These results indicate that
the key elements of the intellectual New Class theory are generalizable to
China, but we were not able to perform a definitive test. For that we would
need family background, time-series data, and direct measures of rank and
108 COMPARATIVE POLITICAL STUDIES / February 2000
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
file and cadre status. But the findings of this initial test are consistent with
theory. The NewClass theory may havelimited generalization to communist
partiesthat retain exclusiverecruitment policies,butforparties—asin China
and Hungary—that became more inclusive, the insights of the theory are
moreappropriate. As in Hungary, the Communist Partyin China has actively
sought out and recruited largenumbers of intellectuals to promote economic
modernization.We have alsoshownhowpoliticaland humancapital arecon
-
verted into material benefits. Party members do enjoy material advantages
compared with the rest of the population; specifically, they earn more money
and have easier access to the most prestigious jobs. Although we found
empirical evidence for the existence of alternative career paths, our cohort
analysis suggests that the post-Mao reforms did not create these paths.
Wecanonlyguesshowtheviolentrepressionofthedemocracymovement
of 1989 and the explosive growth of the private sector changed the cost-
benefit calculus of prospective party members. As the party’s control over
society and the economy continues to diminish as a consequence of the
reform and opening policies of the post-Mao (and now the post-Deng) era,
theprestige oftheparty, andpresumably ofparty membership,hascontinued
to decline. Party membership is still a valued commodity for those seeking
careersinthebureaucracy.ThisfactaloneallowstheCCPtoattractenoughof
China’sbestandbrightesttocontinueto grow.Butforthosewho seethe CCP
as the obstacle to change rather than its agent, or as a corrupt machine more
interested in rent seeking than collective welfare, alternativecareer paths are
readily available. More recent data, when they become available, will likely
show that the black path of higher education and the gold path of private
entrepreneurship have increased in value relative to the red path of party
membership in the years since the CHIP survey was conducted. Since 1992,
CCP policies have encouraged more and more people to “jump into the sea”
(in Chinese, xia hai) of private entrepreneurship. The result has been an
explosion of private sector activity that is transforming China. Indeed, the
three paths are merging in this new environment, with engineers and techni
-
cians finding new ways of marketing their expertise in commercial ventures
and with local business and political elites cooperating in the pursuit of eco
-
nomic development. The private sector is being legitimized by its involve
-
mentwith theseprestigious occupations andpowerfullocal elites,and itcon
-
sequently is losing its reputation as the domain of scoundrels and drifters. To
accurately measure these trends, we must await morerecent time-series data
of the type analyzed in this article. In the meantime, we are left with single
snapshots, such as this, with all their richness and limitations.
Dickson, Rublee / PARTY MEMBERSHIP IN CHINA 109
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
REFERENCES
Bian,Yanjie.(1994).Workandinequalityinurban China.Albany:StateUniversityofNewYork
Press.
Borocz,Jozsef, &Southworth, Caleb. (1996). Decomposingthe intellectuals’classpower:Con
-
version of cultural capital to income, Hungary, 1986. Social Forces, 74(3), 797-821.
Dai, Qing. 1994. Wang Shiwei and “wild lilies”: Rectification and purges in the Chinese Com
-
munist Party, 1942-1944. Armonk, NY: M. E. Sharpe.
Dickson, Bruce J. (1997). Democratization in China and Taiwan: The adaptability of Leninist
parties. Oxford, UK, and New York: Oxford University Press.
Djilas, Milovan. (1957). The New Class: An analysis of the communist system. New York:
Praeger.
Dong, Wanmin. 1991. Recruit party members strictly according to standards (in Chinese).
Henanribao,January7,p. 3, inFuyinbaokanziliao,D2: ZhongguoGongchandang(hereaf
-
ter FBZ) (Collections of Reprinted Reports: D2—The Chinese Communist Party). Beijing:
Renmin daxue chubanshe, January 1991, pp. 127-128.
Eichen, Marc, & Zhang, Ming. (1993). Annex: The 1998 Household Sample Survey—Data
description and availability. In Keith Griffin & Zhao Renwei (Eds.), The distribution of
income in China (pp. 331-347). New York: St. Martin’s.
Gouldner, Alvin. (1979). The future of the intellectuals and the tise of the New Class. London:
Macmillan.
Griffin, Keith, & Zhao, Renwei. (Eds.). (1993).The distribution of incomein China. New York:
St. Martin’s.
Harding,Harry.(1987). China’ssecondrevolution: ReformafterMao. Washington, DC: Brook-
ings Institution.
Hendrischke,Hans.(1994).Expertocracyandprofessionalism. In DavidS. G. Goodman&Bev-
erly Hooper (Eds.), China’s quiet revolution: New interactions between state and society
(pp. 144-161). New York: St. Martin’s.
Jowitt,Ken.1992.Newworlddisorder:TheLeninistextinction.Berkeley:Universityof Califor-
nia Press.
Konrad,George,& Szelenyi,Ivan.(1979).The intellectualson theroadto classpower: Asocio
-
logical study of the role of the intelligentsia in socialism. New York: Harcourt, Brace,
Jovanovich.
Lee, Hong Yung. 1991.From revolutionary cadresto party technocrats in socialist China. Ber
-
keley: University of California Press.
Li, Cheng and White, Lynn. (1990). Elite transformation and modern change in mainland China
and Taiwan: Empirical data and the theory of technocracy. China Quarterly, 121, 1-35.
Li, Yanxi. (1989). Limit the number of CCP members (in Chinese), Xuexiyu shixian, January,
pp. 49-51, in FBZ, February 1989, pp. 136-139.
Lin, Nan, & Wen Xie. (1988). Occupational prestige in urban China. AmericanJournal of Soci
-
ology, 93(3), 793-832.
Lowenthal, Richard. (1970). Development versus utopia in communist policy. In Chalmers
Johnson (Ed.), Change in communist systems (pp. 33-116).Stanford, CA: Stanford Univer
-
sity Press.
Meaney, Constance Squires.1991. Market reform and disintegrative corruptionin urban China.
In Richard Baum (Ed.), Reaction and reform in post-Mao China: The road to Tiananmen
(pp. 124-142). New York: Routledge.
110 COMPARATIVE POLITICAL STUDIES / February 2000
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
Oi, Jean C. (1991). Partial market reform and corruption inrural China. In RichardBaum (Ed.),
Reaction and reform in post-Mao China: The road to Tiananmen (pp. 143-161). New York:
Routledge.
Oksenberg, Michel. (1970). Getting ahead and going along in communist China: The ladder of
success on the eve of the CulturalRevolution.In John Wilson Lewis (Ed.), Party leadership
and revolutionary power in China (pp. 304-437). London: Cambridge University Press.
Renmin ribao. (1990, March 11). In Foreign Broadcast Information Service: Daily
report—China. April 5, 1990, p. 29.
Renminribao.(1991). Gongchandang dezuzhigongzuo chengji juda(Tremendousaccomplish
-
ments in the CCP’s organizational work), June 6, p. 1, in FBZ, June 1991, pp. 121-122.
Rosen, Stanley. (1990). The Chinese Communist Party and Chinese society: Popular attitudes
toward party membership and the party’s image. Australian Journal of Chinese Affairs, 24,
51-92.
Rosen, Stanley. (1991). The rise (and fall) of public opinion in post-Mao China. In Richard
Baum (Ed.), Reform and reaction in post-Mao China: The road to Tiananmen (pp. 60-83).
New York: Routledge.
Shirk, Susan. (1982). Competitive comrades. Berkeley: University of California Press.
Szelenyi, Ivan. (1986-1987). The prospects and limits of the East European New Class project:
Anauto-criticalreflectionontheintellectualsontheroadtoclasspower.PoliticsandSociety,
15(1), 103-143.
Tong, Yanqi. (1997). Transitions from state socialism: Economic and political change in Hun-
gary and China. Lanham, MD: Rowman and Littlefield.
Walder,Andrew G. (1990).Economic reform and income distributionin Tianjin,1976-1986. In
DeborahDavis&Ezra Vogel(Eds.), Chinesesociety on theeveof Tiananmen (pp. 135-156).
Cambridge, MA: Harvard University Press, Council on East Asian Studies.
Walder, Andrew G. (1995a). Career mobility and the communist political order. American
Sociological Review, 60, 309-328.
Walder, Andrew G. (1995b). The quiet revolution from within: Economic reform as a sourceof
political decline. In Andrew G. Walder (Ed.), The waning of thecommunist state: Economic
origins of political decline in China and Hungary (pp. 1-24). Berkeley: University of Cali
-
fornia Press.
Whyte,Martin King.(1989). Whohates bureaucracy?A Chinese puzzle. In VictorNee &David
Mozingo (Eds.), Remaking the economic institutions of socialism: China and Eastern
Europe (pp. 233-54). Stanford, CA: Stanford University Press.
Wong, Raymond Sin-kwok.(1996).The social compositionofthe Czechoslovak andHungarian
parties in the 1980s. Social Forces, 75(1), 61-90.
Xie, Yu, & Hannum, Emily. (1996). Regional variation in earnings inequality in reform-era
Urban China. American Journal of Sociology, 101(4), 950-992.
Xinhua.(1989,September19).InForeignBroadcastInformationService:Dailyreport—China,
September 28, 1989, p. 52.
Xinhua. (1992, October 6). In Foreign Broadcast Information Service: Daily report—China,
October 6, 1992, p. 29.
Xinhua. (1995, June 29).In Foreign Broadcast Information Service: Dailyreport—China, June
30, 1995, pp. 12-13.
Young, Susan. (1994). Privateentrepreneurs andevolutionarychange. In DavidS. G. Goodman&
BeverlyHooper(Eds.),China’squietrevolution:Newinteractionsbetweenstateandsociety
(pp. 105-25). New York: St. Martin’s.
Zhou Peng. (1989). Review of the past ten years of party recruitment (in Chinese). Dangzheng
wenhui, January, p. 12, in FBZ, March 1989, p. 133.
Dickson, Rublee / PARTY MEMBERSHIP IN CHINA 111
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
Bruce J. Dickson is an associate professor of political science and international affairs
and directorof theSigur Centerfor AsianStudies atGeorge Washington University. He
is the author of Democratization in China and Taiwan: The Adaptability of Leninist Par
-
ties (1997).
Maria Rost Rublee is a doctoral candidate in the Department of Political Science at
GeorgeWashington University. Her researchconcerns China’s integrationinto interna
-
tional regimes.
112 COMPARATIVE POLITICAL STUDIES / February 2000
at Australian National University on January 9, 2015cps.sagepub.comDownloaded from
