Chinese Entrepreneurs, the Party-State, and Gender: Women Succeed in Business without the CCP

Abstract In this article, we take a gender perspective to explore the relationship between private entrepreneurs’ business success and their Chinese Communist Party (CCP) membership. We theorize that, due to male-dominated CCP-business networks, female entrepreneurs are not able to leverage CCP membership into greater business success. Observable implications of this are that women will experience little or no business advantage from Party membership, while men will experience a significant advantage. The independence of women's business success from Party connections allows the possibility that women have carved out an area of autonomy from the party-state. Using a series of national entrepreneur surveys collected between 2002 and 2012, we assess the degree of empirical support for our expectations. We show that the CCP does appear to function as an exclusive “boys’ club” in terms of profitable patronage or networks, while the most successful women tend not to be Party members. The surveys also provide some evidence consistent with a CCP effort to recruit successful female entrepreneurs in order to curtail their autonomy. Our findings suggest non-CCP female entrepreneurs are a significant but generally overlooked socio-economic group with considerable potential autonomy.

developmental state. In recent decades it has mixed economic performance and nationalism into the legitimizing ideology of the Chinese Communist Party (CCP), while assiduously avoiding any hint of a democratic opening. It has a growing private sector and entrepreneurial class, but there remains significant state and Party involvement in business. Since 2001, private entrepreneurs have been formally allowed to join the CCP (Jiang 2001), although many de facto businesspeople were already members at that time (Wolf 2001).
As many scholars have noted, this raises the question of whether contemporary China's economic liberalization has created an independent business class, autonomous from the CCP. China's private sector entrepreneurs care about property rights (Huang 2008) and might have less interest in or need for Party engagement. Classic theories of democracy (Schumpeter 1942;Moore 1966) suggest that this would have fundamental implications for the development of the Chinese polity. While the authoritarian state retains coercive controls, used if entrepreneurs start to organize around political issues (such as strengthening property rights, rule of law, or reducing corruption), the state is also dependent on private-sector business owners for one of its key claims to legitimacy: economic success.
This raises a further question, which is less frequently asked: who needs whom? Are Party links crucial for private-sector success in the PRC? If not, this suggests a potential for the development of an entrepreneurial class with considerable autonomy from the CCP. As women are systematically under-represented in the CCP and likely to be excluded from key networks, it is possible that their paths to business success involve more self-reliance, creating the potential for greater autonomy from the party-state.
We therefore take a gender-based approach to help shed light on this question, by using four waves of a large-sample survey of entrepreneurs, male and female, collected between 2002 and 2012. Overall, women comprise 14 percent of the entrepreneurs in our sample, a small, but not insignificant group. Their numbers also appear to be growing: between 2002 and 2012, the percentage of female entrepreneurs in our sample women increased from 11 to 16. Our nuanced and divergent expectations based on gender and temporal distinctions in the data increase our inferential leverage and enhance confidence that our findings are not subject to endogeneity or individual-level confounders such as intelligence or skill.
Our findings strongly suggest that there is a gender gap in the ability of entrepreneurs to exploit Party membership for business success. A "boys' club" of Party insiders appears to exist, allowing men who begin entrepreneurial activities while already CCP members to use the Party to support their business success. In 2012, this group of insiders represented more than a fifth of all male entrepreneurs and over 25 percent of the most successful quintile.
We also find evidence corroborating the idea that an autonomous business class is developing outside the Party. CCP membership is not necessary for private-sector success in modern China; male entrepreneurs can achieve success without joining the Party, although they are on balance not as successful as those male CCP insiders who become entrepreneurs only after joining the Party. For women, however, there is no evidence that joining the CCP enhances business success. In fact, the most successful female entrepreneurs are outsiders. As of 2012, nearly 70 percent of the top quintile of female Chinese entrepreneurs never joined the Party; and less than 15 percent of women in the top quintile founded their business while already Party members.
Existing scholarship has acknowledged the possibility that entrepreneurial autonomy exists, opening the door (even if only slightly) to weakening autocratic rule over time, and that this may have implications for understanding General Secretary Xi Jinping's emphasis on centralizing power and focus on nationalism (over economic growth) as a source of legitimacy (Huang and Chen 2020;Heberer and Schubert 2019;Ji 2018;Schubert and Heberer 2017). But systematic evidence to demonstrate the foundations of this potential for autonomy has been lacking. Our analysis takes an important step forward by presenting such evidence and pointing to starkly heterogeneous patterns of Party engagement and business success based on gender. We show that the potential for autonomy is gendered, and women hold the advantage.

Relevant literature
While abundant empirical studies have shed light on the newly emergent entrepreneurial class's self-reported political attitudes (Tsai 2007;Dickson 2008;Chen and Dickson 2010;Huang and Chen 2016;Chen and Lu 2011), it is uncertain how these relate to their past or future political choices. Most studies have found little evidence that China's new middle class (however defined) is an agent of, or even desires, greater political openness (Dickson 2007;Tsai 2007;Chen and Dickson 2010;Chen and Goodman 2013;Chen 2014;Chen and Lu 2011). Further, correlation of political attitudes with socio-economic circumstance gives little insight into causation or motivation. Political views may adjust to suit past choices, and future decisions may be made for reasons independent of political attitudes, such as potential business success.
Classic theories suggest that economic interests, rather than liberal political ideology, motivate the entrepreneurial class to seek rule of law, property rights, and other liberalizing goals. Theoretically, the expectation that entrepreneurs would express pro-democracy or liberal political views in opinion surveys is not wellfounded. Our approach, therefore, is to focus on potential autonomy of interests, rather than self-reported political attitudes.
Despite the increasing importance of markets, the Chinese party-state retains significant control over economic activity through the legal system and state control over loans, information, business networks and resources, including the ability to exercise these economic powers arbitrarily or for political ends. As a result, it is often claimed that members of the entrepreneurial class need to be embedded within the orbit of the CCP to achieve and maintain economic success (Wank 1999;Dickson 2008;Wright 2010;Chen 2014). Dickson (2007, 852) notes that the CCP is engaged in "co-opting entrepreneurs into the Party and encouraging current Party members to go into business." In this way, the Party appears to thwart the development of entrepreneurial autonomy. 1 However, the conclusion that China's entrepreneurs do not comprise an autonomous group with interests distinct from the Party is typically based on the assumption that political embeddedness succeeds because it brings business benefits. Few studies produce systematic evidence of this, or effectively contend with the possibility that an observed correlation may be due either to 1) the importance of Party connections for success, or 2) the fact that success itself often brings embeddedness for the purpose of co-optation (Tsai 2007;Dickson 2008;Chen and Dickson 2010). We make headway in exploring this distinction by focusing on the sequencing of Party membership and entrepreneurial activity-those "insiders" who found their businesses while already possessing CCP membership are likely to try to leverage the available networks and patronage opportunities membership provides, while those entrepreneurs who join the CCP after founding their businesses are potentially targets of successful co-optation efforts. They may also attempt to leverage the advantages of embeddedness, although would do so as "late-joiners," potentially from a weaker position.
Crucially for our analysis, despite the growing numbers of female entrepreneurs, studies do not take into consideration well-known gender differences in the PRC for access to political capital and networks, due to cultural, social, and institutional factors. Patriarchal attitudes and Confucian norms may obstruct women's pathways to political power (Howell 2002(Howell , 2006Edwards 2007). The leash of domestic burdens and cultural attitudes can prevent them from building effective political networks (Chen 2011;Osburg 2013). Gender disparities in political participation in general and political party membership in particular are widely observed worldwide (Bjarnegård 2013). But the one-party dominance makes the Chinese case distinct, as people do not enter formal politics for their political preferences. And despite the CCP positioning itself as a guarantor of women's rights, it is not held to this commitment by inter-party competition or the need to appeal to constituencies. We exploit this gender-based heterogeneity for theoretical insight and analytical leverage.
Male dominance in the party-state is a continuing feature of the PRC. As of 2019, men comprised 72.1 percent of the CCP's membership (CCP Central Committee Organization Department 2020), and in 2017, they held 73.5 percent of leadership positions in the party-state system (State Council Information Office of the People's Republic of China 2019). This means women may be excluded from the CCP's business networks and co-optation efforts, and/or respond differently to them. In this context, the development of entrepreneurial autonomy is plausible to the extent that network exclusion occurs, and co-optation is incomplete. In this sense, we agree with Tsai (2005Tsai ( , 1130) that China's "[e]ntrepreneurs should … be examined at a lower level of abstraction rather than lumped into a catchall capitalist 'middle class.'" But how can we assess the degree of exclusion or co-optation of the entrepreneurial class and its possible gender patterns? This is a difficult research challenge, and one scholar has suggested the study of such issues is "best left to economists and business management specialists" (Dickson 2008, 28). But this implies that there are not important political dynamics at play in these economic processes, a logic we consider implausible. For example, a recent in-depth study of one Chinese business association found that it was able to "maintain its autonomy vis-a-vis the state," "strengthen its own organizational cohesion," "organize … collective actions," and "even advance … universal values, such as equal rights to justice" (Ji 2018, 463).
Most scholars addressing these issues have used qualitative data and methods. Although statistical analysis has advantages for generalization about fundamental socio-economic processes, the closed nature of the Chinese system means that reliable and revealing statistical or survey data are rarely available. We are aware of only one quantitative study examining the possible causal relationship between party-state connections and entrepreneurial success. It finds that CCP connections may contribute to the success of private businesses (Li at al. 2008). 2 While this is a valuable and pioneering study, Li et al. (2008, 284) acknowledge that they do not fully account for potential confounding factors associated with membership in the CCP, nor do they consider the possibility that successful businesspeople are recruited into the Party in order to co-opt them and limit their autonomy, reversing their assumed causal arrow. In addition, we are not aware of any study focusing on the possible gender differences in private entrepreneurs' access to political capital and networks. Our focus on gender not only provides an important nuance in understanding, but aids inference by allowing theoretically informed comparison of male and female entrepreneurs' patterns of business success and relations with the party-state.
Filling these significant gaps, our analysis using a high-quality dataset presents new evidence supporting the contention that there is an autonomous entrepreneurial class in China, and that female entrepreneurs are likely to play a more important role in this group than has previously been understood.

Theory and hypotheses
The debate over the political implications of China's economic transformation focuses on one of the oldest of political theories. From Aristotle's Politics to contemporary scholarship, the connection between the middle class and political democracy has been a staple of theory and empirical analysis. An economically satisfied middle class possessing private wealth is conducive to the establishment of democracy because their interests tend towards social stability and individual autonomy. They do not threaten the existing social structure, but value the status quo that serves their economic interests, eschewing factionalism or radical wealth redistribution. However, the owners of private assets will also be centrally concerned with the protection of individual property rights, rule of law, and individual autonomy in terms of meaningful rights to political self-determination and participation (Lipset 1959;Moore 1966).
Although the bourgeoisie was seen in an entirely different light by Marx, its reinforcement of capitalist order is also central to Marxist thinking. Putting theoretical irony and the ideological contradictions of the modern PRC aside, some see a fundamental conundrum for Chinese policymakers pursuing rapid growth on a capitalist model (Harvey 2007). Can China's leaders harness the economic energy of entrepreneurs while channeling any political activity into reinforcing the power of the Communist Party, rather than establishing an autonomous political role for the entrepreneurial class through which property rights, rule of law, and meaningful political participation might be pursued? Some argue that this is possible (Chen and Dickson 2010;Pearson 1998). But few would disagree the existence of a politically autonomous, economically powerful group, with strong interests in the rule of law on economic issues, would represent a major risk to the current Chinese political order.
China's transition to a market economy has led to an increasingly private-sector-friendly institutional environment, including changes to the workings of the CCP and to the PRC Constitution. This has created the possibility of alternative paths to economic success, outside the party-state. Indeed, economic reform is the very process that unleashes the economic power of individuals, households, villages, and enterprises (Gold 1989;Blecher and Shue 1996;Pearson 1997;Dickson 2003).
Wary of the potential transition of power from (state-controlled) distributors to (increasingly private) producers through market reform, as witnessed in Eastern European socialist systems (Szelenyi 1978;Nee 1989), the Chinese state has attempted to undermine any independent political role of private sector entrepreneurs through corporatization, co-option, and clientelism (Chen and Dickson 2010;Dickson 2003;Dickson 2008;Goodman 1995;Unger and Chan 1996;Wank 1999). The party-state could be expected to actively recruit leading members of the insipient capitalist class into the CCP, converting potential rivals into allies and providing them with additional (political as well as economic) interest in line with the authoritarian rule of the party-state. However, existing studies provide little systematic empirical evidence (beyond self-reported political attitudes) to help us understand whether this has occurred.
More specifically, key analytical challenges are to establish a) whether Party connections are crucial for business success in China, b) whether the CCP seeks to co-opt entrepreneurs who autonomously achieve success, c) whether gender differences exist in these processes; and d) how these processes may have changed over time. Logically, one or the other of the first two dynamics might dominate, but existing scholarship struggles to distinguish them based on the available evidence. While either process might lead to a close relationship between the party-state and the private economic sector in China, it is hard to move beyond this general observation with existing studies' (largely qualitative) methods and data.
In this article we take important steps towards better understanding these processes, theorizing heterogeneous expectations by gender. These expected gender differences, in turn, enhance our analytical leverage. If we can confirm nuanced expectations of gendered differences based on deductive argument, we can increase confidence in causal assumptions beyond those based on a single logic for a monolithic block of entrepreneurs. This analytical leverage is further strengthened against endogeneity concerns by our distinct expectations based on the timing of Party membership relative to the beginning of business activity, as discussed further below. Because we have distinct, often opposite, expectations for the relationship between Party membership, its sequencing with founding a business, and business success for men and women, we are able to rule out unmeasured confounding factors that might be related to both Party membership and business success, such as personal skill or intelligence. We additionally control for age, education, and time in business.
In our analysis, we examine connections between the CCP and male and female private entrepreneurs, and we explore the possible impact on these two groups' business success. We focus on Communist Party membership. This is "the most common political credential" in the PRC (Walder 1995, 312), considered to provide privileged access to valuable economic opportunities, information, and social networks (Saich 2015). Although more recent scholarship indicates that the rewards of CCP membership are increasingly concentrated in state rather than non-state sectors (Dickson 2014;Appleton et al. 2009), we are aware of no empirical research on whether market reform has created an independent pathway to wealth for private business owners who do not join the CCP.
Female entrepreneurs are of theoretical interest because they are potentially more autonomous from the party-state (on average) than male entrepreneurs, as they tend to remain outside of formal structures of political power (Chen 2011). 3 They are less likely to be members of party-state organizations (Edwards 2004); and selection effects may mean that only the more determined, talented, and self-reliant women succeeding in establishing and maintaining a business without the benefit of Party-based networks.
Because of this political disadvantage relative to men (Judd 2002;Fang, Granrose, and Kong 2005;Howell 2002;Gallagher 2001), Chinese women may be especially inclined to self-organization and self-help to overcome challenges, further enabling their potential autonomy. This points to a dilemma for political leaders seeking to thwart the development of autonomous power centers while also maintaining the foundation for capitalist economic development. A group that is excluded from political power may seek autonomy to guard its success in other areas of activity, like the economy. Members of the existing political networks may see their interests based on exclusive access to patronage threatened by the newly successful group, creating resistance to inclusion.
To address this dilemma, the party-state might seek to co-opt particularly successful female entrepreneurs due to the potential threat posed by their effective creation of alternative social or economic networks. However, it is questionable whether female entrepreneurs will respond positively to such invitations. Earlier research (Chen 2011) reveals that female entrepreneurs do not always accept invitations to join the Party when these are extended, due to fear that it may erode their freedom and create additional burdens. Others express intentions not to join for a range of reasons, including incompetence of officials, irrelevance, and corruption.
A complicating factor is the role of family networks in Chinese businesses. Fei (2006) describes the Chinese family as a medium through which various political, economic, and religious activities are organized, involving cooperation of relatives according to their complementary roles and relationships. The relationship between siblings is one of the most important. A survey of Chinese people's social support networks showed that siblings play a more significant role than parents and children in providing financial and psychological support (Zhang and Ruan 1999).
It is therefore possible that women entrepreneurs may rely on their male family members-especially their brothers-to "grease the wheels" of the party-state machine and promote their business success (Chen 2011). This would mean that there would be less need for women entrepreneurs to join party-state organizations themselves in order to gain the benefits of these connections. While we acknowledge that we cannot examine this possibility in the survey data we use, Chen's detailed study of female entrepreneurs provides qualitative data suggesting that many female entrepreneurs do not tend to rely on such family networks. The extent to which female entrepreneurs might leverage family connections to the CCP for business success remains a topic for future investigation, but the existing evidence suggests it is far from a universal pattern.
The above observations lead us to expect that CCP membership will not bring greater success to women operating private-sector business, but it will enhance success for male entrepreneurs. But the existing literature has not adequately addressed the process leading to an observed association between success and CCP membership. The Chinese state might attempt to undermine the potential independent political role of private-sector entrepreneurs through co-optation of those who are more successful, converting potential rivals into allies.
With the available survey data, we can distinguish between CCP members who began their entrepreneurial activity after joining the Party (insiders), and those who joined the Party only after founding their businesses (late-joiners). While not a precise indicator of co-optation, this can serve as a reasonable proxy, since co-optation is only possible for the latter group. Our expectation is that male and female entrepreneurs who join the Party after founding their businesses will tend to be more successful than those who do not join (outsiders), given that it is their business success that provides the foundation for autonomy, and therefore co-optation. To the extent that the support for this hypothesis diverges by gender, we will gain potential insight into observable implications of the CCP's perceptions of gendered entrepreneurial autonomy and rates of co-optation success.
Based on this discussion, we present the following hypotheses addressing the correlations between female and male entrepreneurs' CCP membership and business success.
Hypothesis 1: Female entrepreneurs who join the CCP before founding their businesses are not likely to have greater business success than other female entrepreneurs. [patriarchal exclusion] Hypothesis 2: Male entrepreneurs who join the CCP before founding their businesses are likely to have greater business success than other male entrepreneurs.
Hypothesis 3: Male and female entrepreneurs who join the CCP after starting their businesses are likely to have greater business success than those who do not join the CCP. [co-optation]

Research Design
We test these hypotheses using four waves of the Chinese Private Enterprises Survey (CPES), national surveys of entrepreneurs conducted by the Privately Owned Enterprises Research Project team of the Chinese Academy of Social Sciences. These data were collected in 2002, 2004, 2008 and 2012, and extensive efforts were made to achieve nationally representative samples of entrepreneurs, although we acknowledge the difficulty of sampling such a small sub-set of the population. 4 The surveys include 16,020 respondents, 5,120 (36%) of whom were members of the Chinese Communist Party (CCP). Women comprised 2,311 (14%) of our respondents. Of women, 23 percent were Party members, compared with 38 percent of men. In addition to CCP membership, these surveys asked respondents about the length of time they have been in business, when they joined the Party, 5 and personal income from their private business, which we use as a measure of business success. 6 After presenting a descriptive analysis, we fit two sets of models to these data to examine the role that Party membership plays in business success (higher private income). We take advantage of the design of the surveys to do this. Each survey includes questions on the date respondents joined the CCP and the date they started their business. We fit linear regression models to measure the interaction between a parameter for respondents' time in business and their Party status (non-member, joined the Party first, founded their business first) to examine the relationships between gender, Party membership, and business success. This allows us to get at not only whether business success is associated with Party membership, but whether the CCP recruits already successful businesspeople, or creates them. We fit these models to both a pooled set of data from all four years for which we have surveys (2002, 2004, 2008, and 2012) and each year separately, to examine change in the relationship between business success and Party membership over time.
We fit linear models to a measure of log private income, as our initial examination of the data showed a right-skew of the distribution of our income data, which would otherwise result in heteroskedasticity and a poorer model fit. We fit these models separately for men and women, instead of using complex models with multiple interactions between gender, CCP membership and the length of time in business, as this approach is simpler, easier to interpret (rather than picking apart three or four level interactions) and substantively equivalent to the more complex full-factorial interactions (see for instance Kam and Franzese 2007 for a discussion). Pooled models with third and fourth-order interaction would also strain the statistical power of these surveys.

Descriptive Analysis
We begin by examining the difference in mean personal income by gender and membership of the CCP, shown in Figure 1. We find that for respondents in all four surveys, the individual incomes of male CCP members are higher than non-members, but for women this was not the case in three of the four surveys (with 2012 the exception). This lends some support to hypotheses 1 and 2, regarding patriarchal exclusion of women and the existence of a boys' club within the CCP, although without the nuance of examining the sequencing of joining the Party and starting in business, which we examine below.
Of course, CCP membership may not have a causal effect on personal income. These patterns could be the result of the Party acting to support its hold on power by recruiting talented and already successful entrepreneurs (Hypothesis 3). We obtain a better understanding of the mechanisms involved in the role of success in private enterprise and CCP membership by testing for the existence of a significant difference between those who start their businesses after they become Party members (Hypotheses 1 and 2), and those entrepreneurs who started their businesses prior to joining the CCP (Hypothesis 3). To do this, we use two additional questions: one asking respondents the year their business was registered as a private enterprise, and another asking what year they joined the CCP. Using these data, we can ascertain whether Party membership preceded or followed business formation.
If Party membership has a causal effect on business success, we expect those who join the Party first to have higher incomes. If the CCP is a recruiter of already established successful business owners-to strengthen its hold on the Chinese state and society-those joining after they founded their firms should have the higher incomes than those who do not join. Assessing temporal sequencing helps us more closely approximate our causal expectations (but we acknowledge the limits placed on causal inference by assessing correlations in observational data).
We find a changing pattern over time for men, but a consistent one for women. Shown in Figure 2, the results for men in the early years of our data (2002,2004) are consistent with hypothesis 2. Male Party insiders who founded businesses after joining the CCP earned higher incomes than those recruited after already starting their firm. However, in 2008 and 2012 already established male entrepreneurs earned more. These late-joiners also earned more than non-member male entrepreneurs in 2004, 2008, and 2012, providing some support for hypothesis 3. For women the results are less ambiguous. Those female entrepreneurs who founded their business before joining the Party had higher incomes in all four surveys (although the difference and sample size were both small in 2002). This supports Hypothesis 1: women who are Party members first have no evident business advantage. Women who never join the Party are also more successful than those who founded their business and Of course, Party membership and gender are not the only possible drivers of financial success in business. It may also be the product of a range of factors including age, education, how long a business has operated, and location. We test our hypotheses more robustly, considering these possible confounding factors, by fitting several models to these data.

Regression Models
We begin by further testing for evidence that there is a relationship between CCP membership and business success. Here we are particularly interested in what these data say about whether CCP membership comes before or after business success. We do this by fitting linear regression models separately to the data for male and female entrepreneurs across a combined file comprising the four surveys that contain the required variables (2002, 2004, 2008, and 2012). This allows us to borrow sample size and power across years and allows us to compare CCP members who joined the Party before founding their enterprise, those who started in business first, and entrepreneurs who never joined.
We use these models to predict the personal income of our sample of entrepreneurs on a log scale (to account for the right-skew of the income distribution). In these models we interact a linear parameter for the length of time a respondent's firm has been operating, whether they were a Party member (Figure 3a), and whether they joined the CCP before or after they founded their business (Figure 3b). We also include controls for age, education, the province in which they are located and the year of the survey. Coefficients and standard errors for these models are included in Table 1. Figure 3 shows estimated earnings from the models fit to the survey data from men and women. Each point in these plots is the estimated average income (in ¥ 10,000) of entrepreneurs as a function of their CCP membership and time in business. The  model smooths the parameter for the time a respondent's firm has been in business, but as we log income as the dependent variable (converted back to the original scale for Figure 3), it does not require the relationship to be a completely linear trend. The top row of plots shows the estimated business incomes of men and women entrepreneurs who were Party members (darker points) and non-members (lighter points). The bottom row shows those who either founded their firm before joining the CCP (lighter) or joined the Party before beginning their business (darker; not shown are those who never joined). These plots highlight the interaction between the parameter for time in business and whether a respondent is a CCP member (Figure 3a), and whether they joined before or after starting their business (Figure 3b).
These results indicate that for men there is no general relationship between CCP membership and income (Figure 3a), but that this may in part be because this association is conditional on when a male entrepreneur joined the Party. While for men in general the amount of time they have been in business is positively associated with higher private incomes (even after controlling for potentially confounding factors), the greatest advantage accrued by firm longevity was for respondents who joined the Party first ( Figure 3b). The interaction between time in business for male entrepreneurs who joined the CCP before founding their firm is positive and more than two standard errors from zero, indicating there may be a benefit that accrues over time for male Party members who enter the private sector after establishing themselves within the Party, providing support for Hypothesis 2. It also suggests some nuance to this explanation when considering the trajectory of business success over time. Those who founded their business before joining the Party appear to have earned more in earlier years, but gained less over time. Outsiders and late-joiners are both estimated to earn considerably less than Party insiders after more than a decade in business; with this gap growing increasingly large over time. Male business owners who never join the Party (up to the time of the survey), are estimated to eventually earn more than those who joined the Party after founding their firm.
These results contradict Hypothesis 3 regarding co-optation of successful male entrepreneurs. Instead, our results suggest those who join the Party before going into business gain the greatest benefit, with late-joiners trying to benefit from CCP membership only after starting a business failing to reap the same rewards. Those who do not join appear to do on-average better than late-joiners, perhaps eschewing the Party because it is not needed. But this overall pattern only applies to men.
For women, there is a key difference. Party membership predicts lower incomes for women entrepreneurs in China, with the gap growing in favor of non-members over time (Figure 3a). Unlike for men, disaggregating CCP members into late-joiners and insiders does not reveal a disadvantage for outsiders-they remain the most successful category of female entrepreneur in China (Figure 3b; although the difference is not significant). Specifically regarding our first hypothesis, female entrepreneurs who joined the CCP prior to founding their private firm were estimated to earn no more than others at the beginning of their career, and less over time (Figure 3b). This clearly supports Hypothesis 1, based on a pattern of patriarchal exclusion, since CCP membership is not associated with business success for female "insiders." Those who joined the Party after founding their firm made more than others shortly after starting their business, but later in their career saw slower growth than those who never joined. This perhaps supports Hypothesis 3, suggesting successful female entrepreneurs (early on) may be recruited by the Party, but that like their male counterparts theses late-joiners obtain few long-term advantages from CCP membership. Non-members did slightly better early on, and they earned increasingly larger incomes than Party members the longer their firms were in business, strongly suggesting that business success for female entrepreneurs is not dependent on Party membership.
Overall, we do not find evidence that Party membership offers financial advantages for women in private sector business, regardless of whether they join before or after becoming entrepreneurs, since both groups underperform women who never join. This contrasts sharply with the apparent advantages enjoyed by male "insiders."

Discussion
Does an autonomous entrepreneurial class exist independent of the CCP, or do Party links remain an essential qualification for private-sector success in the modern PRC? If private-sector success is initially possible without the Party, does the CCP co-opt already successful business people to sustain its hold on power? Given the male-dominated nature of the Party and PRC society, are women more likely find pathways to success autonomous from the CCP?
Our findings suggest that there is a distinct gender difference in the dynamics of party-state connections and entrepreneurial success in China. We find evidence of a "boys club" in modern China. Comprised of men who are Party members before becoming entrepreneurs, this Party-business network does not appear to be particularly welcoming to established male business owners who seek membership after founding their businesses, nor to private-sector businesswomen. For male entrepreneurs who are already Party insiders when they found their businesses, CCP membership appears to be a powerful foundation for business success, conveying a distinct advantage over Party outsiders and those who join the Party only after starting their business.
However, we also found support for the conjecture that a successful business class is developing outside the Party. Membership of the CCP is not necessary for private sector success in modern China, and this is especially true for female entrepreneurs. Male entrepreneurs who remain independent of the CCP achieve moderate business success, relative to male Party insiders who found businesses before joining the CCP (late-joiners). But for women, who comprise 14 percent of the entrepreneurial class according to our data, we find evidence of a substantially different pattern, suggesting a different set of mechanisms at work.
Lack of Party credentials does not appear to hinder female entrepreneurs in their business success, while those who do found their businesses from within the Party do not achieve the same success as those who remain autonomous. Late-joiner women, who join the CCP only after starting their businesses, are on average more successful than those who found their businesses as insiders. This suggests a pattern of co-optation that is not evident for male late-joiners.
Our findings are consistent with a process in which women understand that the party-state "insider" system is not a reliable vehicle for supporting their entrepreneurial success. Women who choose to go into business achieve success based on their business skills and/or autonomous networks. This very independence may then make successful businesswomen important recruits for the Party, seeking to co-opt them to prevent the growth of autonomous networks of socio-economic success. Female entrepreneurs do not appear to need the CCP to achieve business success, while there is at least suggestive evidence that the Party perceives a need to co-opt them by recruiting them after they have started their business activity. Also in stark contrast to men, women who use Party membership as a base from which to begin entrepreneurial activity, achieve little success, providing strong evidence for the existence of a boys' club of Party insiders.
Given theories about the entrepreneurial class, economic interests, and the roots of democracy, our findings have potentially significant political implications. This article not only points to self-interested preferences of business elites as opposed to selfreported political attitudes, but shows clearly the stark contrast in the relationship between CCP insider status and business success in the PRC. In the contemporary Peoples' Republic, there is evidence that the business class has carved out an area of autonomy from the CCP, and women appear to be positioned to play a leading role.
Sources of Support. This research is funded by the China Studies Centre (CSC) at the University of Sydney under a CSC Research Project grant.
Competing interests. The authors declare none.
4. The CASS researchers describe the sampling procedure as follows: "multistage sampling is conducted across China in accordance with a certain percentage (about 0.05%-0.03%, the specific percentage of each time depends on survey funding). First, the total number of samples and that in provincial regions, including autonomous regions, and municipalities directly under the central government, should be determined. Second, in each provincial region, theoretically, we should choose the capital city, a prefecture-level city, and a county-level city, and three counties: one is developed, one is underdeveloped, and the third is in somewhere between the two. Third, we determine the number of urban and rural survey households based on the urban-rural population structure. Fourth, the number of survey firms in each industry is decided based on the distribution of industries in the urban and rural areas. Lastly, we select the specific firm according to the equidistance principle. The survey is implemented mainly by local Federations of and Administrations for Industry and Commerce (some earlier surveys were also conducted by institutions such as the Rural Work Department and the Policy Research Office of the CCP Central Committee)." See Chen et al. 2019. 5. These data were not collected in the 2006 survey. 6. We distinguish business success from business survival here, although it can be argued that longevity is a form of success. In this article, we are interested in success in financial measures, since this implies strong interests in property rights and thus potential autonomy. The correlation coefficients between time in business and personal income are quite low. For men they are, .03, .06*, .05 *, and .09* in 2002, 2004, 2008, and 2012 respectively (* indicates p-value <.05). For women the correlation between time in business and income is only somewhat higher: .13*, .13*, .19*, and .15* for these years. We further note that other measures of wealth available in the survey were not appropriate as indicators of business success. Specifically, the survey also asked about respondents' family income, and family wealth. If other family members (e.g., a husband or wife) also had significant income or assets, this would distort our measure for the entrepreneur's own income or wealth. Unfortunately there was no measure of the entrepreneur's own total wealth or the value of the business. We believe that self-reported total wealth or business value would also be especially prone to error.