Institutional Survival under Extreme State Repression and Subsequent Revival

,

S TATE forces are among the most critical agents of institutional change, and state pressures on isomorphic change and conformity constitute prevailing forces not only for institutionalization (Meyer and Rowan 1977;DiMaggio and Powell 1983;Scott 1983Scott , 1987;;Zucker 1987) but also for the deinstitutionalization of prior organizational traditions and practices (Tolbert and Zucker 1983;Hinings and Greenwood 1988;Oliver 1992;Hiatt, Sine, and Tolbert 2009).Although there is ample evidence about the effectiveness of low to moderate state repression in deinstitutionalizing traditional practices and customs, the fate of an institution under extreme state repression remains understudied.When a state uses extreme repression against an institution, the physical and organizational infrastructure of that institution can be erased, and the constituent individuals or groups within the institution can suffer violent attacks.Can the values of an institution survive when a state exercises extreme repression, and how?Institutional values are the traditions, customs, and beliefs that underpin an institution.The institutional values justify the take-for-granted character of an institution by defining the "way things are" and/or the "way things to be done" (Scott 1987: 496), and they make an institution resistant to change and to be self-reinforcing over time (Tushman and Romanelli 1985;Scott 1987;Zucker 1987).
A growing body of literature has recently documented that "custodians" can help revive a delegitimized institution by assembling financial, human, and physical resources, generating identity reproduction, and/or regenerating institutional values or logics (Dacin and Dacin 2008;Howard-Grenville, Metzger, and Meyer 2013;Dacin, Dacin, and Kent 2019;Kroezen and Heugens 2019;Raffaelli 2019).The "custodians" can be either former institutional participants or newly arrived individuals or groups with interests in reviving an institution.
In line with this research, we advance the study of institutional survival and revival in three aspects.First, we move beyond moderately delegitimized cases to study cases that have endured severe state repression.From Stalin's purges to Mao's Cultural Revolution and from state crackdowns on communist rebellions in Southeast Asia to state-sponsored ethnic massacres in Africa, extreme forms of state violence have been an important part of contemporary history and politics (Zemskov 1991;Cribb 2002;Blattman 2009;Walder 2014;Lupu and Peisakhin 2017).The far-reaching impacts of this state violence have opened new frontiers for research on institutional survival and revival.Second, in terms of theoretical mechanisms, the literature has mixed values and material resources.Understandably, it is difficult to disentangle the two in situations of less extreme repression.Nonetheless, extricating the values mechanism from the material resources mechanism would help make more rigorous both the theorization and empirical testing of institutional survival and revival.This separation can be done more easily in the cases of extreme state repression, where the material bases of a delegitimized institution are often destroyed or even obliterated and therefore "controlled", whereas values stand out as a more fundamental mechanism of institutional survival and revival.Third, to the best of our knowledge, previous studies in the institutional revival literature have been qualitative.Although qualitative case studies have the advantage of presenting delicate processes of institutional revival, they lack empirical rigor in terms of examining theoretical mechanisms.Our study uses systematic, high-quality quantitative data to test the impact of state repression and the role of values as a moderating mechanism for institutional revival.

Extreme State Repression, Dormant Values, and Institutional Survival
Institutional theorists have long recognized the special coercive power of the nationstate, which exercises authority over other organizations and their environments (Lindblom 1977;DiMaggio and Powell 1983;Streeck and Schmitter 1985: 20;Krasner 1993;Scott 2014).Previous studies have demonstrated the effectiveness of the state in deinstitutionalizing organizational traditions and customs.For instance, Tolbert and Zucker (1983) showed that during the civil service reforms of U.S. municipal governments, the states' mandate rapidly deinstitutionalized a range of traditional practices and procedures.In their study of the British local governments, Hinings and Greenwood (1988) demonstrated how state pressures led to the delegitimation of professional bureaucratic design among local government offices.Hiatt, Sine, and Tolbert (2009) showed that during the time of the American Temperance movement, states' jurisdiction over and mandatory adoption of scientific temperance in the curriculum of the public-school system tended to deinstitutionalize the practice of alcohol consumption by altering attitudes and beliefs about drinking.
Although the state's ability to exert legitimate coercion makes it a quite distinct type of actor, the literature on institutions has largely focused on relatively moderate levels of state coercion or repression.However, the state may exercise its coercive power to the extreme, whereby it delegitimizes an institution by destroying its physical and organizational infrastructure and inflicting violent attacks on the institution's constituent individuals or groups.States regularly perpetrate violence against civilians on their own territory.The number of victims of Stalinist repressions has been estimated to be 3.8 million (Zemskov 1991), and an estimated 1.1 to 1.6 million people died during China's Cultural Revolution (Walder 2014).Politically motivated violence has been endemic in the developing world: more than half of all nations there have experienced an armed conflict since 1960, with twenty percent suffering ten or more years of war (Blattman and Miguel 2010).Thus, political scientists have asserted that "state-sponsored and politically motivated violence against minority groups remains a defining feature of contemporary politics" (Lupu and Peisakhin 2017: 836).This conclusion raises an important question about whether and how institutions survive severe state repression.
A burgeoning literature on institutional revival has demonstrated the importance of key "custodians" in preserving and reviving a delegitimized institution (Dacin and Dacin 2008;Howard-Grenville, Metzger, and Meyer 2013;Dacin, Dacin, and Kent 2019;Kreozen and Heugens 2019;Raffaelli 2019).The "custodians" in this literature can be either individuals or groups that previously took part in the institution or new institutional participants who arrived at the time of revival.These custodians help revive the once delegitimized institution by assembling material resources, including financial, physical and human resources, and maintaining identity and values.The focus on "custodians" points to the agentic role of individuals and groups as carriers of the delegitimized institution.We build on this line of research but highlight that the agentic role of "custodians" is severely constrained under extreme state repression to the extent that it may not be possible for "custodians" to assemble material resources for institutional survival and revival.A good case in point is the fate of China's pre-1949 entrepreneurial families featured in our study.From the 1950s through the Cultural Revolution (1966)(1967)(1968)(1969)(1970)(1971)(1972)(1973)(1974)(1975)(1976), there was no way for these families to run family businesses and build up assets.The revival of these families in the post-1978 reform period therefore cannot be explained by referencing the mechanism of cumulative material resources.
Instead, we draw attention to persistent values as the fundamental mechanism for institutional revival.Rather than assuming that the "custodians" act as the carriers of values delegitimized by the state, we emphasize that the "custodians" have to adapt and recalibrate in an extremely hostile environment.Unlike moderate state repression that discourages or discriminates against an institution, extreme state repression can eradicate the infrastructure of a particular institution and subject its "custodians"-the constituent individuals or groups who make up the institution-to violent attacks.By institutional infrastructure, we mean the basic physical and organizational structures and facilities needed for the operation of an institution.With legal bans on the infrastructure of an institution, the prospect of institutional survival and revival hinges on the coping strategies of the "custodians".Previous studies have examined the survival of institutions experiencing moderate delegitimating attacks.For instance, Taylor (1989) demonstrated that abeyance women's movement organizations helped maintain continuity between the women's movement of the 1920s and that of the 1960s in the United States.Schneiberg (2007) found that cooperative organizations coevolved with the dominant, "liberal market" logic and corporate hierarchies in the United States through the mid-twentieth century and beyond.He interpreted that cooperative organizations, as "constituted platforms and building blocks" (Schneiberg 2007: 71), carried on a "cooperative" legacy when it was not in vogue and facilitated institutional revival, reassembly, and reemployment.The theoretical mechanisms used in these studies included a mixture of ideational and material resources.
We argue that, when facing extreme state repression, the "custodians" have a much narrower set of options than the abeyance women's movement organizations or the cooperative organizations described in the literature had.They are more likely to preserve the values of a suppressed institution, sometimes in a dormant manner, out of small "safe" spaces such as families or closely knit social groups, instead of larger organizations or communities where it is harder to build cohesive and high-trust relationships.As repression becomes extreme and violent, Scott (1985: 331) maintained that overt actions should be rare and that hidden resistance is more likely to occur at the level of values or ideology than at the behavioral level, where suppressed groups are more constrained by the threat of power and violence.Moreover, he emphasized that the social sites of hidden resistance (or "hidden transcripts") are often small "safe" spaces such as individuals' circles of closest family members and friends or small public spaces like a tavern or a chapel (Scott 1990: 120-124).In the analysis of religious survival under communist repression in Central and Eastern European countries, Tomka (1995) specifically indicated that after the state paralyzed coordinating organizations, including the churches, the real carriers for expressing and transmitting religiosity were the religious families and small communities.In a nutshell, under conditions of extreme state repression, when physical and organizational infrastructure is obliterated and the constituent groups suffer state violent attacks, the maintenance of institutional values by constituent groups should be the dominant mechanism by which to engage in survival and revival.Extreme state repression allows us to center our theoretical mechanisms on values instead of material resources.
In the last decade or so, political scientists have begun to examine the historical legacies of political violence on individual political attitudes and behaviors.Their findings have shown that victims of political violence become more resilient (Blattman 2009;Wood 2003) and demonstrate an in-group bias in political participation and altruist behaviors (Bauer et al. 2016).Balcells (2012) revealed that individual experiences of violence during the Spanish Civil War led to the rejection of the perpetrators' political identity.In a multigenerational survey of Crimean Tatars in 2014, Lupu and Peisakhin (2017) found that descendants of families that suffered more violence during the 1944 deportation have higher levels of in-group attachment, more strongly support the Crimean Tatar political leadership and are more hostile toward Russia.In a study of the long-term impact of Stalin's repression in Ukraine, Rozenas, Schutte, and Zhukov (2017) demonstrated that violence creates strong political preferences in opposition to the perpetrators of violence.
Previous research has shown that social cohesion within constituent groups or organizations facilitates the maintenance of an institution in decline.For instance, Taylor (1989) revealed that strong interpersonal ties and emotional bonds among exclusive and elite National Woman's Party (NWP) members helped maintain the ethos of the U.S. women's movement from 1945 to the 1960s, during which time the movement confronted an inhospitable political and social environment.Scholars have shown that in the face of the decline of religion, churches with stricter doctrines and stronger demand for loyalty tend to better maintain their denominational direction and experience stronger growth, partly through more extensive social interactions between church members and the repudiation of the outside world (Kelley 1986;Iannaccone 1994).Based on this evaluation, the more severe the political violence, the stronger the in-group attachments and the more hostile the attitudes against the perpetrators of violence; consequently, the suppressed groups should be able to more effectively maintain their "deviant" or opposing institutional values, which the perpetrators have attacked.
Hence, we further argue that as state violence escalates, suppressed groups in small "safe" spaces can paradoxically become more resilient in upholding their opposing institutional values by producing in-group solidarity and hostile attitudes toward the perpetrators.Consequently, in localities where the state has inflicted more severe political violence, suppressed groups should be able to mitigate the negative long-term impact of state violence more effectively on institutional revival.
As shown above, we theorize institutional survival under extreme state repression by borrowing different understandings emanating from the foundational social science disciplines.Although these different understandings are not new, our synthesis provides new insights into the understanding of institutional survival under extreme state repression.We believe this approach has the potential to strengthen both the intellectual foundations and rigor of the theoretical development of research on institutional survival and revival.production grew rapidly at eight to nine percent annually between 1912 and 1936 (Chang 1969).By the late 1930s, a substantial modern industrial sector encompassing producer and consumer goods had been created, assimilating not only modern technology but also modern business practices (Richardson 1999).Therefore, before 1949, China had a burgeoning yet unevenly developed capitalist economy (Richman 1969), and many Chinese capitalists were private entrepreneurs (Lu 1994).After the communists took power in 1949, the new regime viewed private businesses as politically unacceptable because they lived on the surplus value created by workers; a successful communist revolution was supposed to eliminate such exploitation entirely.Private businesses were also believed to be economically inefficient.Public ownership of the means of production was seen as superior for accomplishing and coordinating large-scale production (Yang 2007).The Communist Party "called for the transformation of the entire economy from a basically private and market-oriented economy into a centrally planned one.In the transformed economy, all industrial enterprises and all agricultural farms should be owned by the state.Most of the agricultural labor would be organized into collective farms.No place was reserved for capitalist firms, not even for small private peddlers" (Lu 1994: 62).
The private sector was suppressed step by step during the socialist transformation.In early 1952, the state launched the Five-Anti (Wu-fan) campaign to attack the bourgeoisie and private enterprises.Officially, the campaign was proclaimed as a movement to eradicate the "five evils": the bribery of officials, tax evasion, the fraudulent fulfillment of state contracts, the theft of state assets, and the theft of official secrets for private benefit; however, one of its real chief aims was to speed up the transformation of private enterprises into joint private-state partnership or pure state ownership (Richman 1969).Business owners were rigorously interrogated and heavily fined.To pay off the alleged back taxes and criminal fines, many "voluntarily" turned their firms over to the government (Walder 2015: 76).
By 1956, all private enterprises were transformed into joint state-private partnerships or pure state ownership, and private owners became state employees.It was officially announced that the socialist revolution in the ownership of the means of production was largely completed.The private industry's share of total industrial production fell from 39 percent in 1952 to 16 percent in 1955 and then to almost zero in 1956 (Liu and Yeh 1965).Subsequently, the decade-long Cultural Revolution (1966)(1967)(1968)(1969)(1970)(1971)(1972)(1973)(1974)(1975)(1976) attempted to eliminate bourgeois ideology and traditional Chinese culture.Many former capitalists and members of the petty bourgeoisie were targeted and physically abused during the Cultural Revolution on a much larger scale than the Five-Anti campaign had done.Private entrepreneurship was suppressed for three decades between 1949 and 1978.We focus on the impact of political violence that took place during the Cultural Revolution on the post-1978 revival of private entrepreneurship because the Cultural Revolution was a decade-long (1966)(1967)(1968)(1969)(1970)(1971)(1972)(1973)(1974)(1975)(1976) catastrophe that came immediately before the post-1978 reform era.
In December 1978, two years after Mao's death, pragmatic elites led by Deng Xiaoping launched China's economic reform and "open-door" program.The top leadership had realized that private business had to be tolerated-even encouraged-as "a means to curb the acute and potentially destabilizing problem of unemployment caused by the sudden and massive return of urban youth who had been sent to the countryside in the Mao era" (Peng 2004(Peng : 1054)).
During the reform era, private entrepreneurs in China were initially met with a low level of social acceptance.For instance, the China Youth Daily interviewed Liu Guixian, founder of the Happy Guest Restaurant, which was established in 1980 as the first private restaurant in Beijing.She described that, in 1980, "People in my neighborhood hadn't changed their mindset much yet. . . .Most people thought that I was practicing capitalism.That doing private business was going against the Communist Party.They saw many foreigners come to my restaurant and thought that I had become a special agent.The neighbors would tell me: 'Aren't you afraid of being a special agent?' 'You are a special agent now!' 'Just look, there will be a day for you to be sorry sooner or later.The country will give you what you deserve.'. . . .I couldn't sleep well at night" (China Youth Daily, April 10, 2008).
Government policies and formal legal regulations toward private entrepreneurship went through stages of tolerance, accommodation, and encouragement in the post-1978 reform era.In the decade before 1988, the leadership tried to strike a balance between the lingering powerful delegitimizing forces and the pressing need for jobs.The government officially only recognized "self-employed individuals" (getihu) who employed fewer than eight people.For those who hired a greater number, for years, the government purposefully remained ambiguous-neither supporting nor banishing them-until 1988, when the growing number of getihu who hired eight or more employees were officially recognized by the state as "private enterprises" (siying qiye).
During Deng Xiaoping's 1992 South China tour, private entrepreneurship was decoupled from capitalism, and he further defended private entrepreneurship: "it promotes the growth of the productive forces in a socialist society, increases the overall strength of the socialist state, and raises living standards" (Deng 1993: 372).These three criteria-which replaced the notion of exploitation-became the new basis on which to evaluate the legitimacy of private entrepreneurship.In 1997, the state fully recognized private businesses by officially proclaiming private enterprises as "an important component" of China's "socialist market economy" (they had only been a "supplementary" component prior to 1997).
Our observation period starts from the time in which the state loosened its restrictions on private entrepreneurship in 1978; it ends in 1996, the year just before private entrepreneurship was fully recognized by the state.Thus, private enterprise was not fully legitimate over this period even though its level of legitimacy had been increasing.This period is ideal for observing the effect of dormant values of private entrepreneurship on institutional revival.We analyze the extent to which pre-1949 entrepreneurial families served as institutional carriers for private entrepreneurship between 1949 and 1978, especially in the face of the political violence of the Cultural Revolution (1966)(1967)(1968)(1969)(1970)(1971)(1972)(1973)(1974)(1975)(1976) The Cultural Revolution The Cultural Revolution was a decade-long (1966)(1967)(1968)(1969)(1970)(1971)(1972)(1973)(1974)(1975)(1976) political and social catastrophe initiated and led by the Chinese Communist Party's (CCP) Chairman Mao Zedong and "was responsible for the most severe setback and the heaviest losses suffered by the Party, the state, and the people since the founding of the People's Republic" (Resolution on CCP History (1949-1981)).
Endorsed by Mao Zedong and adopted by the CCP Central Committee on August 8th, 1966, the "Decision of the CCP Central Committee Concerning the Great Proletarian Cultural Revolution", also known as the Sixteen Points, was Mao's blueprint for the Cultural Revolution (MacFarquhar and Schoenhals 2006).The document was published in the People's Daily the next day and clearly stated the political values and objectives of the Cultural Revolution.
At present, our objective is to struggle against and crush those persons in authority who are taking the capitalist road, to criticize and repudiate the reactionary bourgeois academic 'authorities' and the ideology of the bourgeoisie and all other exploiting classes and to transform education, literature, and art and all other parts of the superstructure that do not correspond with the socialist economic base, so as to facilitate the consolidation and development of the socialist system (The Sixteen Points; English translation from Schoenhals 1996: 33).
In addition, the document called on the masses to rise up and strike against capitalist roaders, the reactionary bourgeoisie, and other exploiting classes.Responding to this call, students in middle schools and universities across the country formed revolutionary groups known as the Red Guards to smash the "four olds" (old ideas, old culture, old customs, and old habits).Red guards searched people's homes and confiscated or destroyed the properties of people with a "bad" class background.Individuals labeled with a "bad" class background were those from capitalist, rightist, or counterrevolutionary families, as well as rich peasants, bad elements, and landlords (Treiman and Walder 2019).In urban areas, the homes of capitalists were especially alluring sites for house ransacking (Perry and Xun 1997: 12).In Shanghai, 84,222 homes of "bourgeois" families were looted between August 23 andSeptember 8, 1966 (MacFarquhar andSchoenhals 2006: 117).The most frightening aspects of the Red Guards movement were the beating, torture, and killing of innocent people.Many of these people committed suicides due to intolerable physical and mental abuse (Thurston 1987;Wang 2001)."In August and September (1966), altogether, 1,772 were murdered in Beijing.In Shanghai in September there were 704 suicides and 534 deaths related to the Cultural Revolution.In Wuhan during this period there were 32 murders and 62 successful attempts at suicide" (MacFarquhar and Schoenhals 2006: 124).
Workers' participation was also a critical ingredient of the Cultural Revolution (Perry and Xun 1997).Student and worker rebels quickly spiraled out of control.At the end of January 1967, the military was ordered to stabilize public order.Subsequently, the military became an important force serving on the newly formed revolutionary committees (MacFarquhar and Schoenhals 2006), which replaced local parties and governments throughout China.
The revolutionary committees turned out to be an even more violent actor.Walder (2014: 21) demonstrated that state-led acts of terrorism, conducted by military and civilian authorities after the establishment of revolutionary committees, were "far more ruthless and efficiently destructive" than the actions of student and worker rebels.For instance, "Cleansing the class ranks" was the first major campaign led by the revolutionary committees.It started gradually in places such as Shanghai in late 1967 and was well underway in most regions of China by the summer of 1968 (MacFarquhar and Schoenhals 2006).The targets of the campaign included all who were suspect due to their class backgrounds, political histories, or ties to foreign entities.Similar to those who had been victims in the earlier days of the Cultural Revolution, a large proportion of the victims of this "cleansing the ranks" campaign were individuals with overseas and nationalist affiliations or who came from landlord and capitalist backgrounds (Walder 2015: 274).In Beijing, the "cleansing of the class ranks" generated the deaths of 3,731 people between January 1968 and May 1969.In eastern Hebei, more than 84,000 people were persecuted; 2,955 of them died; and 763 suffered permanent disabilities.In Jilin Province, in northeast China, this campaign led to the "death by unnatural causes" of 2,127 people and permanent injured 3,459.In Zhejiang, an estimated 100,000 people were arrested and struggled, and a total of 9,198 deaths occurred.In Yunan, 448,000 people were targeted; some 15,000 became victims of "cleansing", and 6,979 died (MacFarquhar and Schoenhals 2006: 257-258).
Two other major national campaigns were called "One Strike, Three Anti" (yi da san fan) and the "May 16 Conspiracy", which also led to the persecution of millions of victims (MacFarquhar and Schoenhals 2006).The Cultural Revolution ended with Mao's death on September 9, 1976.It has been recognized that political violence is psychologically traumatic and generates fears of future threats from the state (Hobfoll, Cannetti-Nisim, and Johnson 2006;Johnson and Thompson 2008).Given that a dominant and overarching theme of the Cultural Revolution was to ideologically and politically assault "capitalist roaders" and their supporters, other things being equal, the more severe the political violence in a location during the Cultural Revolution, the more individuals became fearful of future state violence, and consequently, private entrepreneurship would revive slowly in these locations in the post-1978 era.
Hypothesis 1: Individuals from locations where political violence was more severe during the Cultural Revolution were less likely to become private entrepreneurs between 1978 and 1996.

Pre-1949 Entrepreneurial Families as Institutional Carriers
During the Mao era , China was governed by a totalitarian regime (Linz and Stepan 1996), and the public spheres were monopolized by the state (Zhou 1993).The state controlled all aspects of life, including the economy, education, art, and science.Walder (1986) elaborated on how highly dependent on the state individuals were in their lives.In this context, "deviant" values are most likely to become By heightening fears of future threats from perpetrators, traumatic experiences of political violence often engender stronger in-group attachments and hostile attitudes toward perpetrators, and families can transmit these feelings across generations (Lupu and Peisakhin 2017).We argue that as political violence escalates in a locality, suppressed groups would become more resilient in upholding their "deviant" or opposing institutionalized values, and consequently, these groups would better mitigate the deterrent effect of political violence on subsequent institutional revival.
Sociologists and psychologists have demonstrated that families become more close-knit when a common enemy is increasingly threatening.In a study of familial interpersonal relations, Stouffer and Lazarsfeld (1937: 68-69) noted that "in some types of social groups, a crisis may be a means of solidifying the group, especially in the face of a common enemy."In a study of Nazi Germany, Allport, Brunner, and Jandorf (1941: 14) wrote that in the context of a social catastrophe, familial attachments strengthened: "Most dramatic are the many instances of return to healing intimacy of the family after bitter experience of persecution on the street, in the office, or in prison."Based on the Harvard Project on the Soviet Social System, Geiger (1955) documented that respondents from families whose immediate members suffered state repression were more likely to feel that family solidarity was enhanced.In addition, among the respondents who claimed to have been "'at one time in favor of the Soviet regime', but to have turned against it subsequently, while still living under Soviet jurisdiction," (67) 39 percent specifically named terror as a reason for their change of political attitude.
That former capitalists and members of the petty bourgeoisie were beaten up or killed during China's Cultural Revolution has been well documented (Wang 2004: 192, 317, 631, 728).Whyte (1995Whyte ( : 1011) ) maintained that although the Cultural Revolution has usually been seen as tearing families apart through political strife, "what is less often appreciated is how for most families-those not affected by pressures on one family member to denounce another-the effect was the opposite.What was involved was a particular instance of a more general phenomenon of any society: in times of crisis, when the social environment became more dangerous and unpredictable, people are thrown back to their closest bonds in order to survive, and particularly on their families." Even when Red Guards denounced a parent, the rupture in the family was often followed by eventual reconciliation and resulted in closer bonds (Yuan, 1987).In an early account, Parish (1975) explicitly indicated that escalating political violence led to strengthened family bonds during the Cultural Revolution.He observed that "In the initial stages of the Cultural Revolution, youth turned against old customs and on their rampages destroyed objects such as ancestral plaques which supported the traditional family.However, as the Cultural Revolution became progressively more violent, many youths became frightened and returned to the safety of their homes for a period of intensive togetherness.The result may have been that the Cultural Revolution eventually did as much to strengthen the family as it did to weaken it" (Parish 1975: 629).Based on interviews with victims of the Cultural Revolution, Thurston (1987: 226) articulated that "one unintended consequence of the Cultural Revolution, an unexpected result of the inclusion of entire families in the web of attack, was a strengthening of familial bonds. . .In these cases, familial bonds served as protection against a hostile state."In a fascinating study of one family through six hundred years of Chinese history, Esherick (2011: 295) maintained that "For all the iconoclastic attacks on the 'feudal' family system and the evils of the Confucian patriarchy, and for all the pressure to distance oneself from miscreant family members, the actual result of the Cultural Revolution may have been a strengthening of family ties.At a time when even best friends could be forced to turn on each other and reveal secrets, one's immediate family members could be the most trusted to stick together."Although the family was not a completely "safe" space due to possible internal political strife, it was probably the most protective social group one could rely on for trust and support during the Cultural Revolution.Li (2006) asserted that it was inevitable for suppressed families to generate hostile attitudes toward the communist regime.In 1968, a 20-year-old student, Xiao Ruiyi, wrote a well-known "constructive" letter to Mao Zedong in which she stated that "We beat up and humiliate landlords and rich peasants, regardless of whether they have changed.We think that some of them became rebellious precisely because we have implemented persistent and cruel repression.If we continue acting like this, who would not want Chiang Kai-Shek (Chinese nationalist leader and opponent of the communists) to come back?" (Song 2018: Chapter 24).Whyte (1991) articulated that members of families tended to share the fate of an individual member who suffered from attacks, which produced "a shared sense of resentment" (727) during the Cultural Revolution.In his study of the Ye family through several, mostly tumultuous periods of Chinese history, Esherick (2011:292) documented that within a family that suffered violent attacks during the Cultural Revolution, although the parents-to avoid further political trouble-may have still verbally encouraged their children to stay loyal to Chairman Mao and the Communist Party, the children tended to show a strong oppositional attitude, and "the dining table became a forum of increasingly intense debate" (292).
In a recent memorial book, Liu Wenzhong (2016) documented in detail how his family members remained resilient under the violent attacks of the Cultural Revolution.His father was politically labeled a "counterrevolutionary" and tortured during the Cultural Revolution because he had been an elite banker before 1949 and worked for the U.S. Citibank and other major banks in the era of the Republic.His elder brother, Liu Wenhui, was well-known as a "counterrevolutionary" who was shot dead for gainsaying the Sixteen Points-Mao's Blueprint for the Cultural Revolution.In the book, Liu Wenzhong described how the members of the Liu family suffered elements of post-traumatic stress disorder such as shock, anger, nervousness, fear, and even guilt; how his parents, and other brothers and sisters comforted and gave each other strength; how his younger brother secretly kept the blood letter of Liu Wenhui after their parents found it inside a quilt returned by the prison; and how he and his sister persisted in fighting for the rehabilitation (ping fan) of Liu Wenhui after the Cultural Revolution.The siblings believed that their father and elder brother Liu Wenhui were innocent, and their hatred toward Mao and the communist regime was evident throughout the book.
A shared sense of solidarity against and resentment toward the Communist Party and state tended to alienate family members from the regime and help revive traditional family values that the state had attacked.In their classic work on China's politics, Townsend and Womack (1986: 183) maintained that the violence and suffering inherent in class struggles caused alienation from the communist regime and triggered the growth of local spheres of individual or group power, where the "deviant" values that the regime had attacked were paradoxically revived.Liang and Wang (2007) asserted that during the Cultural Revolution, the family had persistently been the most powerful entity to resist the "revolutionary storms" (23), and value restoration first took place at the family level once the revolutionary attacks were reduced.In his study of the Ye family, Esherick (2011: 305-307) revealed that after their fathers were attacked and/or jailed during the Cultural Revolution, the Ye family's younger generation was determined to avoid the politics of the party-state, and many of them chose academic or scientific research careers in the post-Mao reform era.This pattern was stronger if the father himself had been a scientific scholar and had been assaulted as an advocate of "bourgeois science" or as an "American spy" during the Cultural Revolution.The Ye family is an extended family consisting of several smaller families, and Esherick (2011:295) emphasized that "It was, however, the nuclear family that proved most resilient."Even when family members were sent to the countryside or different parts of China, the nuclear families grew tighter through letters and New Year visits at home.Therefore, even during the disastrous Cultural Revolution, the values of private entrepreneurship could be maintained at the family level.We hypothesize the following: t Hypothesis 2: Individuals belonging to families that owned private businesses before 1949 were more likely to become private entrepreneurs between 1978 and 1996 than individuals belonging to families that did not own private businesses before 1949.
The stronger the local political violence during the Cultural Revolution, the more likely the members of pre-1949 entrepreneurial families either were tortured or felt more imminent threats from the state.As previous research has shown, it is not necessary for an individual to have directly experienced violence to feel its effects (Novick 2000).Consequently, attachment within these families increased, and more resentment and hostility could be directed against the communist regime and the persons, institutions, and symbols associated with it (Geiger 1955).Therefore, these families could become more resilient in preserving their "deviant" values of private entrepreneurship, which the regime attacked, and exert a stronger buffering effect against the deterrent effect of local political violence during the Cultural Revolution on institutional revival.Therefore, we hypothesize the following: ter insulated from the deterrent effect of political violence on individuals' entry into private entrepreneurship between 1978 and 1996.
In 1950, China's new communist government assigned to each family a label describing its political class (family origin, or jiating chushen).Designed to consolidate political control, the basic principle of this political categorization was to differentiate "proletarian" elements from "exploiters" and "class enemies".Although they were derived from Marxist class analysis, these categories were in fact political statuses, with those in the "proletarian" categories being assumed to be loyal to the regime and those in the "exploiters" categories being assumed to be hostile to the regime.This political classification system was rife with contradictions (see Treiman and Walder 2019, for detailed explanations).For instance, those who joined the Party or the Red Army before their victory were designated as "revolutionary", even if they had come from prosperous "exploiter" households.Conversely, despite one's humble origins, having joined the nationalist party or army would have eliminated one's "proletarian" origins and made one a class enemy.However, regardless of the widespread contradictions, those in the "exploiters" categories were designated as "enemies of the people" and periodically abused in political campaigns (Treiman and Walder 2019).These political categories were attached to entire families and passed down through generations (through the male line).They were enforced for three decades, and it was not until 1979 that the political stigma attached to "bad" class labels was formally removed (Treiman and Walder 2019).
As mentioned above, during the Cultural Revolution, families with a "bad" class background were more targeted in political campaigns and suffered more from political violence (MacFarquhar and Schnoenhals 2006;Walder 2015).Therefore, we suspect that, as local political violence escalated during the Cultural Revolution, politically stigmatized pre-1949 entrepreneurial families should have stronger ingroup attachments and more hostile attitudes against the communist regime than families that were engaged in entrepreneurship before 1949 and were not politically stigmatized.The "deviant" values of private entrepreneurship should then be better preserved among the pre-1949 entrepreneurial families that were stigmatized than among those that were not.Therefore, we further hypothesize the following: Hypothesis 4: When compared to families who were engaged in entrepreneurship before 1949 and were not politically stigmatized for it, those families that were engaged in entrepreneurship before 1949 and were politically stigmatized have a stronger effect on mitigating the deterrent effect of local political violence that took place during the Cultural Revolution on individuals' entry into private entrepreneurship between 1978 and 1996.from all regions of China except Tibet (Treiman 1998), and the other source is the China Political Events Dataset, 1966-1971(Walder 2014), which contains information about 14,405 political events and contextual data from approximately 2,319 county and city-level jurisdictions during the period from June 1966 to December 1971 during the Cultural Revolution.

Data
"Life Histories and Social Change in Contemporary China" Dataset The Life Histories and Social Change in Contemporary China survey gathered extensive information on respondents' life histories and job activities.Samples about rural and urban individuals were drawn separately, yielding 3,087 urban cases and 3,003 rural cases.The survey used the 1990 census data to organize more than 2,500 county-level jurisdictions into 25 strata, based on the proportion of the population that received at least a junior high school education.From each stratum, two countylevel jurisdictions were selected with probability proportionate to the size of the population aged 20-69.Based on the same criterion, a township/town/street committee was selected from each county-level jurisdiction, and two villages or neighborhood committees were selected from each township/town/street committee.
Respondents were randomly sampled from each selected village or neighborhood committee.Details on the sample design are provided in Treiman (1998).
We test our hypotheses using the urban sample of 3,087 cases because most capitalists and members of the petty bourgeoisie lived in urban areas, and the legitimacy of private entrepreneurship was a much more salient issue in urban China at the time of the post-1978 reform era.Within the urban sample, a municipality is identified as a sample municipality if one of its county-level districts is selected into this national multistage stratified design.The urban sample of 3,087 cases comes from 44 sampled municipalities, which themselves emanate from 24 provinces across China.Dataset, 1966Dataset, -1971 The "China Political Events Dataset, 1966-1971" resulted from a two-decade effort by Andrew Walder and his research team to collect data on events that happened during the Cultural Revolution from annals (difang zhi) published by local governments; they also included background data from these annals or statistical yearbooks (Walder 2014;Walder and Lu 2017).The dataset contains the dates for 14,405 Cultural Revolution events and the corresponding number of deaths; for example, the dataset includes events such as "factional fighting between rebel groups" (wudou), "cleansing the class ranks", "One Strike, Three Anti", "May 16 Conspiracy", et cetera, that took place from June 1966 to December 1971 in 2,319 county-and city-level jurisdictions.The repression that characterized the Cultural Revolution significantly weakened after Lin Biao, the CCP's vice chairman and Mao's designated heir and successor, died in a plane crash on September 13, 1971 (Walder 2015: 286).The dataset is unique and rare because it includes data related to death occurring during specific events of the Cultural Revolution at the county and city levels in China.It allows us to measure the severity of state political violence during the Cultural Revolution at the municipality level.The changes in the boundaries between jurisdictions were reconciled to ensure that there were no gaps in coverage or double counting."Some of the jurisdictions in existence in 1966 were merged with others, split in two, or renamed.These boundary changes were reconciled by examining materials in the annals and tracing the history of boundary changes in the national register of jurisdictions (Ministry of Civil Affairs 1998)" (Walder 2014: 516).The data on the total population and urban population in each jurisdiction were gathered from other sections of the local annals or from statistical yearbooks published during the mid-to-late 1960s.

Measurement
We define private entrepreneurship broadly as nonagricultural undertakings in which laborers may or may not be hired.A private entrepreneur earns income mainly from "running a business" -in contrast to a wage earner (who derives income mainly from a salaried job) or a farmer (whose main source of income is farming and agricultural sidelines).Our broad definition of private entrepreneurs therefore includes "self-employed individuals" (getihu) who have fewer than eight employees and "owners of private enterprises" (siying qiyezhu) who hire eight or more people.
An individual's family is defined as an entrepreneurial family before 1949 if his or her father or paternal grandfather owned a business in 1948, as shown in the Life Histories and Social Change in Contemporary China survey.Individuals tend to live with paternal grandparents because the Chinese family is patrilineal (Whyte, 1995).Following Treiman and Walder (2019), we derive the political class category of a respondent's family based on his or her response to a question: "Your family origin (jiating chushen) is. . ." in the Life Histories and Social Change in Contemporary China survey; furthermore, we define as "bad" class backgrounds the family origins of "capitalist", "rightest", "rich peasants", "bad element", "landlords", or "counterrevolutionary" (1158).A family became a class enemy once it had been designated as part of a "bad" class, and the class label was passed down to future generations.We define the "stigmatized pre-1949 entrepreneurial family" as those pre-1949 entrepreneurial families that were labeled as belonging to a "bad" class and the "nonstigmatized pre-1949 entrepreneurial families" those pre-1949 entrepreneurial families that were not labeled as being part of a "bad" class.
A dominant and overarching theme of the Cultural Revolution was the assault on capitalist ideology and the attack on the supporters of that ideology.Because death represents the most terrifying form of violence, we construct a variable we call "number of deaths (per 10000 persons) at the municipal level during the Cultural Revolution" based on the "China Political Events Dataset, 1966-1971" data set, aiming to capture the severity of state political violence during the Cultural Revolution at the municipal level.We exclude from these calculations the number of deaths resulting from "factional fighting between rebel groups" (wudou), "rebel groups attacking the military", or "rebel groups attacking government offices".These are not state repression events.
We control for respondents' cadre status and educational level during the 1978-1996 period, as well as their age and gender, because these factors could influence one's entry into private entrepreneurship during the post-1978 reform era.In line with the literature on the market transition (Walder 1996;Walder, Li, and Treiman 2000;Li and Walder 2001), the variable "cadre" is coded 1 if a respondent was a "group or team leader" or a higher-level cadre and 0 otherwise.We differentiate between the respondents by using four levels of education: college, senior high school, junior high school; primary school or no formal education is the reference category.
We also control for what respondents' job status was before they entered private entrepreneurship.The variable "state-sector job" is a dummy variable, which is coded 1 if a respondent was employed in either the government, public institutions (Shiye Danwei), or state-owned enterprises, and this dummy variable is 0 otherwise.The state-sector job status could affect one's entry into private entrepreneurship because state-sector jobs were "iron rice bowls" with guaranteed life-time job security, medical benefits, housing, education, and other elements of social welfare.We also control for the proportion of state-owned enterprise (SOE) employment at the municipal level at the early stage of the economic reform.The variable "ratio of SOE employment at the municipal level in 1984" is calculated based on data from the 1985 China City Statistical Yearbook-the first official nationwide statistical yearbook at the municipal level in the reform era.This variable captures both the availability of SOE job opportunities at the municipal level and the extent to which the state was organized at the municipal level.
We control for the number of private businesses per capita at the municipal level in 1948 to take into account the prevalence of private businesses in a municipality before 1949.Because the Life Histories and Social Change in Contemporary China survey is a nationally representative survey dataset, we construct the variable "number of local private businesses (per 1000 persons) at the municipal level in 1948" from this dataset, using the percentage of respondents who came from families that owned private businesses in a municipality in 1948.This variable captures the inherent entrepreneurial advantage of a municipality, including geographical advantage and local entrepreneurial culture.We also control for the state's strong relegitimation of private enterprises after 1988.The variable "state's strong relegitimation of private enterprise from 1988 on" is coded 0 for the years before 1988 and 1 for 1988 and the years thereafter.
Regarding the "China Political Events Dataset, 1966-1971", "rules for counting casualties were conservative.Coders were instructed to record a value of zero unless a specific number was mentioned in the text, even if there were repeated references to widespread casualties without mention of specific numbers" (Walder 2014: 516).Due to the restrictive coding rules and the lack of detail in many accounts, the reported number of deaths in the dataset is not a complete count of all events and casualties; instead, this number should be considered as an estimate of the actual numbers, which cannot be directly observed.The annals publication was strongly influenced by province level editorial policy (Walder and Su 2003;Walder 2014)."Their publication was coordinated at the province level: annals followed a relatively standard format within provinces (subject headings, order of presentation, and coverage devoted to different topics)" (Walder 2014: 515).In addition, more urbanized regions may have been able to devote more resources to the compilation of annals (Walder 2014).Therefore, to correct possible biases in the reported estimates in the numbers of deaths, we control for province level fixed effects and the size of the total urban population at the municipal level in the statistical models.Controlling the fixed effects by provinces also helps us control for any time-invariant provincial effect.Enduring regional policy or ideological variations, if any, should be captured by controlling province level fixed effects.
Therefore, concerns about unobserved factors that could affect both death rates during the Cultural Revolution and the development of local entrepreneurship in the reform era can be reduced or dissipated by controlling fixed effects at the provincial level.Wang (2021) examined the effects of a series of historical and geographical variables on local political violence during the Cultural Revolution and found that none of these factors is statistically significant in explaining local variations in political violence during the Cultural Revolution.He concluded that idiosyncratic factors (rather than systematic factors) explain the violence.There has been an almost complete turnover of government personnel since the Cultural Revolution.The radical bureaucrats in post during the Cultural Revolution were gradually forced to retire and leave office in the course of Deng Xiaoping's personnel reform in the 1980s, and they were replaced with young professional bureaucrats (Manion 1993).
Table 1 and Table 2 show the descriptive statistics and pairwise correlations of the major independent and control variables.Overall, the magnitude of the correlations between variables is small.The only exception is that "pre-1949 entrepreneurial family" and "nonstigmatized pre-1949 entrepreneurial family" have a high correlation of .947.This high correlation is due to the fact that the majority of pre-1949 entrepreneurial families were not stigmatized with the label of "bad" class origin.We estimate the effects of these two variables in separate statistical models.

Statistical Model
We use an event history framework to estimate the effects of local political violence that occurred during the Cultural Revolution, as well as the effects of stigmatized and nonstigmatized pre-1949 entrepreneurial families on individuals' entry into private entrepreneurship in the post-1978 era.Event history analyses model the effect of explanatory variables on the rate at which units-in this case, individuals-experience an event or a transition in qualitative states over time (Allison 1984;Tuma and Hannan 1984), yielding the instantaneous rate at which a unit experiences an event during an interval, given that the event has not occurred by time t: where r(t)yields an individual's hazard rate, or "risk", of entering private entrepreneurship.
We use a nonparametric specification of duration dependence, the "proportional hazards" model by Cox (1972).This model specifies the hazard rate to be: where X s are exogenous variables of interest, b sare coefficients estimating the effects of these variables, and h(t) is the unspecified baseline hazard function.The Cox model is appealing because one need not assume a function form for the baseline hazard rate.Incorrect parametric assumptions may lead to biased estimates for the effects of covariates on the hazard rate (Blossfeld and Rohwer 1995).The proportional hazard assumptions for Cox regression models were tested using Schoenfeld residuals and found not to be violated.The Cox model has previously been used by both organizational sociologists (Carroll and Mosakowski 1987) and labor economists (Taylor 1999) to model the transition into or exit from self-employment based on life-history data.
We model an individual's first entry into private entrepreneurship from 1978 to 1996.An individual becomes at risk of becoming a private entrepreneur after either the year 1978 or the year he or she turns 18 years old, whichever comes later.Those who did not become private entrepreneurs by 1996 are right-censored.All time-varying covariates are lagged one year and updated on an annual basis.We analyze 2769 cases after deleting cases with missing values on key variables.A total of 338 out of the 2769 cases transitioned into private entrepreneurship during the 1978-1996 period.

Findings
The main results of our statistical analyses are presented in Table 3. Model 1 is the baseline model testing the main effects of the local deaths that occurred during the Cultural Revolution and the pre-1949 entrepreneurial family as an institutional carrier on post-1978 entrepreneurial entries.Model 1 shows that the "number of deaths (per 10000 persons) at the municipality level during the Cultural Revolu- This result supports Hypothesis 1, which posits that local political violence has a long-term deterrent effect on the revival of an institution under attack.All else being equal, the increase by one death (per 10000 persons) in a location during the Cultural Revolution decreases the likelihood of individuals entering private entrepreneurship by 3.5 percent during the 1978-1996 period.In model 1, the main effect of pre-1949 entrepreneurial families is statistically nonsignificant.This result contradicts Hypothesis 2, which predicts that members of families that owned private businesses before 1949 are more likely to become private entrepreneurs in the post-1978 reform era than the members of families that did not own private businesses before 1949.This finding shows that pre-1949 entrepreneurial families were not able to successfully carry private entrepreneurship as an institution throughout the three decades  of communist delegitimating attacks and have an overall positive effect on institutional revival in the post-1978 reform era.
The entrepreneurial values of these families may have faded under the totalitarian regime.
In model 2 in Table 3, we add the interaction effects of the "number of deaths (per 10000 persons) at the municipal level during the Cultural Revolution" and the variable of "pre-1949 entrepreneurial family" to test whether pre-1949 entrepreneurial families mitigate the deterrent effect of the deaths that occurred locally during the Cultural Revolution on individuals' post-1978 likelihood of becoming entrepreneurs.A likelihood ratio test shows that model 2 fits the data significantly better than model 1 (the baseline model) (p <.05, two-tailed).The interaction between the "number of deaths (per 10000 persons) at the municipal level during the Cultural Revolution" and the variable "pre-1949 entrepreneurial family" is positively significant (p <.05).This result supports Hypothesis 3, which predicts that as local political violence escalated during the Cultural Revolution, pre-1949 entrepreneurial families were better insulated from the deterrent effect of local political violence on post-1978 entrepreneurial entry.
Figure 1 shows the mitigating effect that the pre-1949 entrepreneurial family had on the deterrent effect that deaths occurring locally during the Cultural Revolution had.In localities where the number of local deaths per 10,000 people was close to zero, the history of family business has not affected the likelihood of an individual to enter the private sector in the post-Mao period.This finding captures the overall effect of three decades of communist delegitimating attacks (including the Cultural Revolution) on pre-1949 entrepreneurial families.The value of private entrepreneurship diminished among the pre-1949 entrepreneurial families in general and became too weak to facilitate institutional revival in the reform era.Whatever advantages these families had in terms of business skills, know-how, or "cultural capital", their entrepreneurial values were too deteriorated for them to become effective carriers in the reviving of private entrepreneurship in the first two decades of China's market transition.Only in localities where the number of local deaths was higher during the Cultural Revolution were individuals coming from families that owned private businesses prior to 1949 more likely to establish private   businesses in the post-Mao period, suggesting that these families had a mitigating effect when they had been faced with more severe local political violence.This finding is in line with our core argument that as state violence escalates, suppressed groups paradoxically become more resilient in upholding "deviant" institutionalized values by generating in-group attachment and hostile attitudes toward the state.Consequently, in localities where state violence was more severe, suppressed groups were able to more effectively mitigate the negative long-term impact of state violence on institutional revival.In short, Figure 1 shows both the overall effect of the communists' delegitimating attacks on pre-1949 entrepreneurial families and the mitigating effect of pre-1949 entrepreneurial families in localities where the Cultural Revolution caused more deaths.Model 3 in Table 3 is the baseline model testing the main effects of the number of local deaths caused by the Cultural Revolution on post-1978 entrepreneurial entry and of the role of pre-1949 entrepreneurial families, both stigmatized and nonstigmatized, as institutional carriers.Similar to model 1, the "number of deaths (per 10000 persons) at the municipal level during the Cultural Revolution" has a   of communist delegitimating attacks.
In model 4 in Table 3, we add the interaction effects of the "number of deaths (per 10000 persons) at the municipal level during the Cultural Revolution" and the role of stigmatized and nonstigmatized pre-1949 entrepreneurial families to test the extent to which stigmatized and nonstigmatized pre-1949 entrepreneurial families mitigated the deterrent effect of local political violence on individuals' post-1978 entrepreneurship entry.A likelihood ratio test shows that model 4 fits the data significantly better than model 3 (the baseline model) (p <.05, two-tailed).The interaction between the "number of deaths (per 10000 persons) at the municipal level during the Cultural Revolution" and the role of stigmatized pre-1949 entrepreneurial families is positively significant (p <.001).In contrast, the interaction between the "number of deaths (per 10000 persons) at the municipal level during the Cultural Revolution" and the role of nonstigmatized pre-1949 entrepreneurial families is statistically insignificant.The magnitude of the former coefficient is 2.89 times larger than the latter, and a Wald test shows that the difference between these two coefficient estimates is statistically significant at p <. 05 (two-tailed).This result supports Hypothesis 4, which posits that as local political violence escalated, the fact that a pre-1949 entrepreneurial family had been stigmatized had a stronger effect on mitigating the deterrent effect of local political violence on institutional revival in comparison to such families that had not been stigmatized.Families with "bad" class backgrounds were targeted more acutely in the political campaigns of the Cultural Revolution and suffered more from political violence.Therefore, as local political violence escalated, stigmatized pre-1949 entrepreneurial families generated stronger in-group attachment and more hostile attitudes against the state than nonstigmatized pre-1949 entrepreneurial families did.Consequently, compared to the latter, the former were more resilient in preserving their "deviant" entrepreneurial values and were better insulated from the deterrent effect of local political violence on institutional revival.Figure 2 and Figure 3 clearly show that as local political violence escalated, stigmatized pre-1949 entrepreneurial families and nonstigmatized pre-1949 entrepreneurial families distinctively mitigated the deterrent effect of local deaths during the Cultural Revolution.
The control variables exhibit the expected effects.Those who were younger or male were more likely to become private entrepreneurs over the 1978-1996 period than those who were (respectively) older or female.Individuals with senior high school or college education were less likely to become private entrepreneurs, with the magnitude of the coefficients decreasing from the former to the latter.This result is reasonable given that the state allocated more desirable jobs to the better educated and that private entrepreneurship was still in the process of legitimation.Those with state-sector jobs were less likely to become private entrepreneurs than those with jobs in nonstate sectors.Both the cadre variable and the variable "ratio of SOE employment at the municipal level in 1984" show a statistically nonsignificant effect.The variable called "state's strong legitimation of private enterprise from 1988 on" shows a statistically significant and positive effect, which demonstrates that the influence of the state's clear and powerful post-1988 relegitimation triggered private entrepreneurship."The logged total urban population at the municipal level during the Cultural Revolution" variable shows a negative and significant effect.This result may reflect that the campaigns of the Cultural Revolution were more impactful in more urbanized areas.

Alternative Explanations
The results we have presented so far demonstrate consistent mitigating effects of pre-1949 entrepreneurial families the long-term negative impact of local political violence that occurred during the Cultural Revolution on post-1978 entrepreneurial revival; the results also show that stigmatized pre-1949 entrepreneurial families exhibit a stronger mitigating effect than nonstigmatized pre-1949 entrepreneurial families.There are alternative explanations for this result.We propose an explanation based on the existing scholarship on political violence and families in times; our explanation is that severe political violence engenders stronger in-group attachments and more hostile attitudes against the state.Consequently, as political violence escalates, suppressed families become more resilient in upholding their "deviant" institutional values and better mitigate the deterrent effect of political violence on institutional revival.However, it is possible to argue that in locations where more severe political violence took place during the Cultural Revolution, individuals belonging to pre-1949 entrepreneurial families faced stronger state discrimination and had fewer local opportunities for state-sector jobs or "iron rice bowls".Therefore, these individuals could have been forced into private entrepreneurship during the reform era due to the lack of opportunities to work in the public sector.
To address this alternative explanation, in all the models included in Table 3, we control for whether respondents held a state-sector job and the proportion of SOE employment at the municipal level during the early years of the reform era.As expected, the "state-sector job" variable is negatively significant, which suggests that sociological science | www.sociologicalscience.com individuals with state-sector jobs were less likely to become private entrepreneurs in the first two decades of China's economic reforms.The "ratio of employment at the municipal level in 1984" variable is statistically nonsignificant, indicating that the proportion of local SOE employment did not influence individuals' post-1978 entry into private entrepreneurship.Table 3 shows that our key findings are robust when these two variables are controlled for in the models.
We further address this alternative explanation in Table 4.In model 5 in Table 4, we add the interaction effects of the "stigmatized pre-1949 entrepreneurial family" and the "state-sector job" variable to test whether individuals belonging to stigmatized pre-1949 entrepreneurial families were more likely to become private entrepreneurs when they did not have state-sector jobs.The interaction between these two variables is statistically nonsignificant, suggesting that individuals belonging to stigmatized pre-1949 entrepreneurial families were not more likely to enter private entrepreneurship when they did not have state-sector jobs.Our key findings are robust when we control for this interaction effect.
In model 6 in Table 4, we add the interaction effects of "stigmatized pre-1949 entrepreneurial families" and the "ratio of SOE employment at the municipal level in 1984" to test whether individuals belonging to stigmatized pre-1949 entrepreneurial families were more likely to become private entrepreneurs when the proportion of local SOE employment was smaller.This interaction effect is also nonsignificant, indicating that individuals belonging to stigmatized pre-1949 entrepreneurial families were not more likely to enter private entrepreneurship when the local prevalence of SOE jobs was lower.Our key findings are also robust when controlling this interaction effect.Model 7 in Table 4 is a full model that includes both of the above two interaction effects, and the statistical patterns of our key findings remain.
Therefore, state-sector job opportunities at either the individual or municipal level are unlikely to drive our key findings.

Discussion
Whereas previous studies have shown that low to moderate state delegitimation can deinstitutionalize prior organizational traditions and customs, we theorize and test whether and how the values of an institution can survive under conditions of extreme state repression.We argue that these values become dormant within small "safe" social spaces, such as families and other small-scale close-knit social groups, as opposed to larger communities.Additionally, although state violence can deter the revival of an institution under attack by generating traumatic experiences, which in turn generate the fear of future threats from the state, suppressed groups-as institutional carriers-can paradoxically become more resilient in preserving "deviant" values by generating in-group solidarity and hostile attitudes toward the state.Consequently, when state violence became more severe in a locality, suppressed groups were able to more effectively mitigate the negative long-term impact of state violence on institutional revival later on.By combining two unique nationally representative datasets pertaining to contemporary China, we study the extent to which pre-1949 entrepreneurial families served as institutional carriers for private entrepreneurship during the Mao era , especially in the face sociological science | www.sociologicalscience.com  (1966)(1967)(1968)(1969)(1970)(1971)(1972)(1973)(1974)(1975)(1976), and subsequently shaped individuals' likelihood to enter into private entrepreneurship during the post-1978 reform Contrary to our expectation, we find that overall entrepreneurial transmission at the family level was discontinued by communist repression during the Mao era.This finding corroborates our core argument that extreme and long-lasting state repression can threaten and disrupt institutional survival and revival.Consistent with our hypotheses, we find that in localities that endured more severe political violence during the Cultural Revolution, pre-1949 entrepreneurial families were better insulated from the deterrent effect of the deaths that occurred locally during the Cultural Revolution on institutional revival.Moreover, we find that stigmatized pre-1949 entrepreneurial families-those labeled as having a "bad" class background-exhibited a greater mitigating effect than nonstigmatized pre-1949 entrepreneurial families.Tests using controls for state-sector job opportunities at both the individual and municipal levels suggest that state-sector job opportunities are unlikely to drive our results.The findings reported in our study advance the literature on state and institutional change, institutional survival and revival, and the development of the Chinese economy and entrepreneurship.Below, we discuss these contributions and outline directions for future research.

Contributions to the Literature on State and Institutional Change
Our study contributes to institutional theory by examining institutional survival under conditions of extreme state repression in which an institution is delegitimized, its physical and organizational infrastructure is destroyed, and the individuals or groups constituting the institution suffer violent attacks.Previous studies have demonstrated that low to moderate state delegitimating can deinstitutionalize organizational traditions and customs (e.g., Tolbert and Zucker 1983;Hinings and Greenwood 1988;Hiatt, Sine, and Tolbert 2009).However, we have a limited understanding of whether and how the values of an institution can survive when the state delegitimizes it, eliminates its physical and organizational infrastructure and violently attacks former institutional participants (individuals or groups).
We find clear evidence showing the disruptive effects of extreme state repression on institutional revival.We find that the overall institutional continuity of private entrepreneurship was eliminated at the local community level by communist repression.The variable called "number of local private businesses (per 1000 persons) at the municipal level in 1948" is not statistically significant, and we find that individuals from locations where political violence was more severe during the Cultural Revolution were less likely to become private entrepreneurs between 1978 and 1996.
Surprisingly, the disruptive effect of extreme state repression is even stronger than we expected.The overall institutional continuity in private entrepreneurship also stopped at the family level.This result reflects that the values of private entrepreneurship could not effectively withstand the three decades of extreme communist repression during the Mao era.Nonetheless, we find that in localities that experienced more severe political during the Cultural Revolution, pre-1949 entrepreneurial families were better insulated from the deterrent effect of local deaths during the Cultural Revolution on institutional revival in the post-1978 reform era.In addition, stigmatized pre-1949 entrepreneurial families-those labeled as belonging to the "bad" class-exhibit a greater mitigating effect than nonstigmatized pre-1949 entrepreneurial families.
Together, these findings demonstrate that under conditions of extreme state repression, the values of an institution tend to become dormant within small "safe" social spaces, such as families and other small close-knit social groups, as opposed to larger community networks.Moreover, these values tend to be paradoxically better preserved in small "safe" social spaces in locations in which state repression is more violent.Taken together, these findings point to future directions in the study of state and institutional change.First, future studies should further examine the special coercive power of the state on other institutions and organizations (Lindblom 1977;Streeck and Schmitter 1985: 20;Kraser 1993: Scott 2014) and analyze the consequences of varying levels of state repression on institutional change.Second, depending on the severity level of state repression in a particular context, future studies must consider the varying strengths of different types of institutional carriers (e.g., community networks, organizations, small close-knit social groups) in carrying the values of a delegitimized institution and facilitating institutional revival later on.Third, future studies also must examine how the severity of state repression diminishes or amplifies the maintenance of values among different types of institutional carriers.Indeed, the effects can be drastically different across different types of institutional carriers.
Fourth, we focus on the family as the institutional carrier that suffered extreme state repression.Future studies could test the effects of dormant values on other small social groups.Presumably, the more cohesive a social group is, the more effective it would be at maintaining its institutional values under conditions of extreme state repression, and the stronger its effect would be on institutional revival later on.For both theoretical and comparison purposes, examining institutional survival among different types of institutional carriers with varying levels of social cohesion could further enhance our understanding of institutional survival and revival under conditions of extreme state repression.

Contributions to the Literature on Institutional Survival and Revival
Existing studies on institutional survival (Taylor 1989;Schneiberg 2007) and revival (Dacin and Dacin 2008;Howard-Grenville, Metzger, and Meyer 2013;Dacin, Dacin, and Kent 2019;Kreozen and Heugens 2019;Raffaelli 2019) have mixed values and material resources as theoretical mechanisms.This approach has been appropriate because the institutions in these studies suffered moderate delegitimation, and it is difficult to disentangle the effects of values from those of material resources in less extreme situations.
Our research contributes to the study of institutional survival and revival by extending the research frontier to conditions of extreme state repression.findings shed new light on institutional survival and revival.First, whereas a delegitimized institution under relatively moderate repression can be revived through multiple pathways, in situations of extreme state repression, the persisting values are far more important than material resources in sustaining the delegitimized institution and may even be the only viable mechanism if the institutional infrastructure has been eliminated altogether.Second, the "custodians" of the delegitimized institution facing extreme state repression must retreat into small "safe" places such as families and closely knit social groups for protection and solidarity.These custodians have a much smaller social space and set of options than "custodians" finding themselves in situations of low to moderate repression, where larger communities and more diffuse networks can act as a base for material and nonmaterial support.Third, we use systematic data rather than individual cases to show the sweeping and long-lasting disruptive effects of state delegitimation/repression and the paradox of fragility and resilience among "custodians"-pre-1949 entrepreneurial families in our study-in the process of institutional survival and revival.Our approach allows for a more rigorous testing of the alternative mechanisms of institutional revival illustrated in the literature.In light of our findings, future studies on institutional survival could further examine how different types of institutional carriers (e.g., community networks, organizations, and small close-knit social groups) effectively carry the values of a delegitimized institution under different severity levels of delegitimating attacks and how these carriers contribute differently to institutional revival later on.

Contributions to the Study of China's Economy and Entrepreneurship
Our study also contributes to the literature on the development of the Chinese economy and entrepreneurship.Private entrepreneurship has been a key driver of China's recent economic growth.The social roots of private entrepreneurship in post-1978 China are an intriguing topic, especially given that private entrepreneurship was delegitimized for three decades between 1949 and 1978.The linkage between post-1978 and pre-1949 private entrepreneurship has remained understudied.The literature on the Cultural Revolution has mainly focused on what happened during the Cultural Revolution.Our findings show that private entrepreneurship was deinstitutionalized even at the family level through three decades  of communist delegitimating attacks.The deaths that occurred locally during the Cultural Revolution have had a long-term deterrent effect on entrepreneurial revival in the post-1978 reform era.However, in localities in which the number of deaths was high, pre-1949 entrepreneurial families were able to better mitigate that deterrent effect, whereas the stigmatized pre-1949 entrepreneurial families exhibited a greater mitigating effect.
Although the Cultural Revolution is believed to have destroyed traditional Chinese culture, its long-term impact on China's subsequent development remains sociological science | www.sociologicalscience.com unclear.The first dataset capturing the number of casualties occurring in localities across China during the Cultural was recently released by Walder (2014) and makes it possible to analyze some of the most important yet understudied questions regarding the long-term impacts of the Cultural Revolution.Notably, future research can study the Cultural Revolution not as a single event having an aggregate effect nationwide but as a movement consisting of wide variations across localities in the level of mass mobilization and the severity of outcomes.More systematic insights into the impact of the Cultural Revolution on the Chinese economy and entrepreneurship can be generated.

Hypothesis 3 :
As local political violence escalated during the Cultural Revolution, families engaged in entrepreneurship before 1949 were betsociological science | www.sociologicalscience.com 705 October 2023 | Volume 10 We test our hypotheses by combining two unique nationally representative data sets about contemporary China.One source is the dataset collected in 1996 from the survey of Life Histories and Social Change in Contemporary China, a multistage stratified nationally representative probability sample of 6,090 adults aged 20-69 sociological science | www.sociologicalscience.com 706 October 2023 | Volume 10 sociological science | www.sociologicalscience.com 707 October 2023 | Volume 10 sociological science | www.sociologicalscience.com 708 October 2023 | Volume 10 sociological science | www.sociologicalscience.com 709 October 2023 | Volume 10 sociological science | www.sociologicalscience.com 710 October 2023 | Volume 10 sociological science | www.sociologicalscience.com 714 October 2023 | Volume 10
organizations with strong emotional bonds and high levels of trust.Hence, we focus on the precommunist (pre-1949) entrepreneurial families as institutional carriers of private entrepreneurship.As mentioned above, by 1956, all private enterprises had been transformed into state-private partnerships or pure state ownership.Private entrepreneurship was delegitimized for three decades between 1949 and 1978 (Mao era), and many former capitalists and members of the petty bourgeoisie were targeted and physically abused during the Cultural Revolution.

Table 1 :
Descriptive Statistics Major Independent and Control Variables.

Table 2 :
Correlation Coefficients of Major Independent and Control Variables continued.
sociological science | www.sociologicalscience.com " has a statistically significant (p <.05) and negative effect on individuals' entry into private entrepreneurship after 1978.Therefore, local deaths occurring during the Cultural Revolution deter the post-1978 revival of private entrepreneurship. tion

Table 3 :
Cox models predicting entry into private entrepreneurship in urban China from 1978 to 1996.
Notes: † p < 0.01; * p < 0.05 (two-tailed tests) The reference category for education is primary school or no formal education.The reference category for the pre-1949 entrepreneurial family is a family that was not engaged in business in 1948.The reference category for stigmatized or nonstigmatized pre-1949 entrepreneurial family is a family that was not engaged in business in 1948.All the time-varying covariates are lagged one year and updated on an annual basis.sociological science | www.sociologicalscience.com 715 October 2023 | Volume 10 statistically significant (p <.05) and negative effect on individuals' entry into private entrepreneurship after 1978.The main effects of stigmatized and nonstigmatized pre-1949 entrepreneurial families are not statistically significant.Again, these results may reflect that entrepreneurial values eroded among these carriers in general due to the three decades (

Table 4 :
Cox models predicting entry into private entrepreneurship in urban China from 1978 to 1996.Notes: † p < 0.01; * p < 0.05 (two-tailed tests).The reference category for education is primary school or no formal education.The reference category for the pre-1949 entrepreneurial family is a family that was not engaged in business in 1948.The reference category for stigmatized or nonstigmatized pre-1949 entrepreneurial family is a family that was not engaged in business in 1948.All the time-varying covariates are lagged one year and updated on an annual basis.
sociological science | www.sociologicalscience.com of the political violence of the Cultural Revolution