Bargaining in Socialist Factories

A socialist model of industrial labor relations is characterized by state or collective ownership, a government-sanctioned wage and bonus system, lifetime employment, and benefits for workers, including free medical insurance, full pension, free housing, and other fringe subsidies. The system was believed to prevent capitalist exploitation and mobilize workers effectively. In 1949, in the aftermath of decades of civil wars, foreign invasions, including the Japanese occupation, and natural disasters, the Chinese Communist Party-State started implementing this industrial model nationally, hoping to achieve both political stability and economic development for the new regime. Along with structural revamping, the Party-State also promoted Marxist ideology that proclaimed the working class to be a leading class, the true owner of all the factories. Managers were government employees under the new system, and workers were told they were building a strong nation of their own and thus working for themselves.

In addition to Marxist ideology, a very different type of discourse, derived from Confucian doctrines, was used to mobilize workers. It referred to factories as big families and analogized work relationships with family relationships. For example, workers were supposed to consider each other as brothers and sisters. In addition, it utilized the concept of “reciprocal obligations” (bao-en), an old exchange principle deeply rooted in Chinese culture. The new definition of the old concept bao-en was that, since Chairman Mao and his Party-State “liberated” people from foreign and domestic oppression and exploitation and ended their long suffering, the people, including workers, should be grateful and reciprocate through their loyalty and hard work. For workers who had very little education at the time, these values based on traditional ethics seemed much easier to understand and accept.

Under these political, ideological, and organizational circumstances, what were labor relations like between managers and workers on the shop floor in factories, especially in the SOEs? One argument, made by Andrew Walder, is that the state and factory management were able to have full control over workers through building an “organizational dependence” based on the unique work unit (danwei) system, which provided its employees lifelong economic security (salary, housing, etc.) and performed some social services, such as issuing permits for marriage or travel and lodging.Footnote ⁴ Walder also argued that total control was derived from a powerful party system that penetrated the grassroots level, that is, the production team level in a factory.Footnote ⁵ Both the political control and material benefits offered by a company manager generated clientelistic labor relations in industrial organizations in socialist China. The counterargument is that factory managers lacked control over workers under the danwei system since managers had no right to fire unruly workers and no power to decide workers' salaries and bonuses.Footnote ⁶ The problem here is that both arguments treat socialist labor relations as static and see neither their development in a historical view nor the interactive dynamics between managers and workers in the actual production process.

The socialist regime-building lasted for thirty years. Contradictory forces influencing labor relations changed constantly throughout this period. Both Taylorist “scientific management” and workers' participation in management were implemented in most of the SOEs at about the same time. Sometimes they coexisted; sometimes one prevailed. Workers' reactions to both of them, together, or to either one of them, were, however, influenced by absorbing both methods and the ideas behind them whether they were concurrently present or not.

The essence of Taylorism, once favored by V.I. Lenin, is to maximize labor output by a strictly measured quota system. It was believed that because the working class was the master of the state and owner of the enterprises, Taylorism did not conflict ideologically with socialist ideas. Many Chinese managers indeed used so-called “scientific management” to control production speed and quality. One of my interviewees, who worked for a state-owned textile factory during the 1960s in Shanghai, was selected as a municipal model worker because of her productivity and product quality despite the fact that she was never a party member. She remembered how managers and other coworkers surrounded her with a stopwatch and calculated the exact time she spent on an operation. Her working methods were publicized and recommended.Footnote ⁷

Another interviewee, also in her seventies, who used to work for a collectively owned factory assembling electric tools in Shanghai, told me that the daily quota at that time was high, and she was often too busy to go to the bathroom. Her boss actually supervised how much time people spent going to the bathroom.Footnote ⁸ These two examples show that labor discipline was tight from time to time and workers had hardly any power to directly challenge the quota set by management.

However, things do not always remain fixed. Another managerial apparatus implemented along with “scientific management” in about the same time period seems more unique. It was Chairman Mao Zedong who promoted this managerial style. Originally it was called “Two-One-Three” (ercan yigai sanjiehe) and later changed into the “Angang Constitution” (Angang Xianfa) by Mao. The Angang Constitution maintained that workers should participate in management and managers should participate in actual production. While the workers' participation in management was clearly intended to mobilize the workforce, the core of this managerial device gradually eroded the hierarchy between managers and workers. It marked a clear departure from the Soviet “one-man management” as well as from Taylorism. During the Cultural Revolution in the late 1960s through the 1970s, management's participation in production was institutionalized while workers' participation in management became more and more a formality and thus less and less effective. It became a routine practice in which all the administrative, managerial, and technical staff were required to participate in manual labor from cleaning jobs to machine operating at least once or twice a week. The significance of managers' participation in production, in my opinion, was not as a challenge to a conventional division of labor among management, technology, and production so much as an opportunity for workers to directly interact with managers at all levels. The hierarchy remained in place, but the direct interaction enhanced the communication, and the practice did tone down the sharp differences between the two parties. When managers worked on the shop floor, which was really the workers' terrain, they could benefit from workers' skills, experience, and common sense. If the Taylorist scientific management empowered managers in controlling production, workers used this communication to show managers their executive power in the process of production. More importantly, they showed managers what they considered to be a good manager with whom workers would voluntarily cooperate. Workers in their early forties or older have been quite nostalgic about this system until today. Managers with whom I talked, however, generally dismissed it as one of the absurd policies implemented under Mao.

As we discussed before, the Party-State also relied heavily on ideology not only for control, but also for incentives. For example, since employment was guaranteed by the danwei system, workers who did not meet the monthly quota would not face the loss of their jobs but rather criticism from their managers and team members for being “selfish,” “falling behind,” and “wang en” (failing to be reciprocal to the party's kind treatment). In the same vein, if a worker exceeded the quota, he would be praised for having “advanced consciousness.” These workers would be invited to join the party and promoted to leadership positions. Since any monetary incentives were generally considered “bourgeoisie,” honor from promotion or being a model worker was emphasized more strongly. The above-mentioned textile worker was a rare case in that she was never a party member even though she was selected as a model worker several times in the 1960s and 1970s. I asked her, “Why didn't you join the party?” She said, “I am not very good at talking publicly. There was another municipal model worker in our factory. She was very good at talking with all the fancy words. She joined the party and became a manager. But people said the quality of her products was not as good as mine.”Footnote ⁹ People were set up to compete for political honor, for promotion, and even for material rewards, such as thermoses and towels, of which the honorable implication was underscored.

Instead of assuming that Chinese socialism “organized labor to be a managed entity,” it should be mentioned that the ideological control and incentives had different impacts on different generations of workers. It certainly worked more effectively on the old generation than the young one. In the beginning of the new regime, the workforce in Shanghai (and elsewhere in the country) consisted of people who mostly emigrated from nearby rural areas and had a poor family background and low educational attainment. The workers in this generation group, with whom I talked during my fieldwork, confirmed that concepts such as “master,” “equality,” and “liberation” were as appealing to them as significantly improved working conditions.Footnote ¹⁰ A department director (formerly a worker) in his late fifties told me that a lot of elderly workers at the time seriously believed that the factories were their own. He said one old worker in his factory always went to work at 6 a.m. even though the normal starting time was 7:30.Footnote ¹¹

Most workers in this group did have more incentives to do well at work. The paternalistic “bao-en” (reciprocal obligations) discourse also had more moral meaning for this group of laborers. But similar consent among the younger generation of workers was not so prevalent. Despite the indoctrination of “yi ku si tian” (Remembering hardships of the past and reflecting on the happy life of today) in order to influence people's collective memory and inspire their identity with the regime and its ideology, younger workers, who graduated at least from middle school as the beneficiaries of the compulsory education system under socialism, mostly regarded the benefits given to the SOEs' employees as entitlement. Some workers in this younger group became “activists” (jiji fenzi) who followed management closely and whose political and production performance often prevailed in the workplace.Footnote ¹² However, with little sense of “bao-en,” they had very different motives for their good work than older worker activists. Many of them wanted political advancement and opportunities to go to college; especially during the Cultural Revolution this was the only institutional channel through which to get into college.Footnote ¹³

This brings us back to the argument that management lacked control over workers since it could neither hire and fire workers nor decide their wages and bonuses. My research shows that, although management could not use these methods considered canonical under capitalism to control and maximize labor input and output, it had power over job assignment, upgrades in the eight-degree skill ladder (related to wages),Footnote ¹⁴ selection for training programs and colleges, and promotion from “blue collar” to “white collar” positions within a factory. Managers could turn the lifetime employment around to their own advantage by denying the requests of workers to transfer to other work units. Managers also used group meetings, public criticism, and home visits to exert moral pressure on their subordinates. Thus, one former worker (now a chemical technician) told me that, in general, workers feared that managers might “give them tight shoes to wear” (chuan xiaoxie, meaning the possible revenge given by one's superiors). He himself was an activist, first promoted to a position of foreman and finally moved “upstairs” to be an office worker.Footnote ¹⁵

The focus so far, however, is on the exchanges between managers and a few worker activists. After all, how could a manager mobilize a majority of workers with only limited positions for promotion? What methods did he use to draw voluntary cooperation from the majority of workers? Walder argued that managers could distribute apartments and coupons for hot commodities not available in the market, such as sewing machines and bicycles, to attract workers' loyalty.Footnote ¹⁶ My interviewees told me differently. Even if managers made this kind of attempt, in reality, it did not work the way managers would have hoped. Distribution was one area that people watched closely. If the Party-State used ideology for control and aspiration, workers could use it for their own advantage. That is, they used the official discourse about their status as management's political equals as a weapon to fight against any perceived unfairness and despotic ways of ruling.Footnote ¹⁷

A manager could use his distribution power once to take care of those loyal to him, but he might not be able to do it often. If he argued that his favorites were those activists who performed better than others, workers would invoke the political and moral standards of the time by arguing that activists were supposed to work harder for their political ideals, not for materialistic gains. Activists who really had greater political ambitions should by all means avoid creating suspicion among and drawing criticism from fellow workers about their motives. If an activist was rewarded only because he brownnosed his manager in front of, or was perceived so by, other workers, the manager would be accused of favoritism and doomed to fail. In one factory, for example, after the manager decided to give a very quiet worker the monthly bonus of three yuan, which was the lowest in a three-level scale, she disappeared without telling anyone. Her parents reported her missing to the police and asked to have the factory director held responsible. After the worker finally returned, she got the bonus of five yuan, the highest bonus, next month.Footnote ¹⁸

Consequently, to avoid workers' suspicions and criticism, management often left decisions about distributing bonuses or other benefits to the workers. One interviewee told me that the government created a general wage hike for workers after a wage freeze of many years. It should not have resulted in any dispute because the distribution was based on seniority. However, there was a small difference among the people who had the same seniority: forty-three yuan versus forty-one yuan. Knowing that this difference was under scrutiny, the management decided to let the workers determine who should get forty-three yuan and who forty-one. Although the difference was not big, workers seemed to take it as a symbol of fairness. It stirred workers to fight among themselves and between work teams. As mentioned above, there were three levels of monthly bonuses, which resumed in the mid-1970s after the abolition of the bonus system in the mid-1960s: three, four, and five yuan. Again, the differences were only one yuan apart. Most managers let work teams discuss among themselves how to distribute the bonuses fairly. In many factories, I was told, bonus distribution finally came to using a rotation to avoid workers' fighting either with managers or among themselves. Workers successfully pushed for an egalitarian bonus distribution, which defeated its original intention, for now it was no longer a performance-based reward system that management could use for control. For other material goods, managers often used a lottery or chose the neediest according to workers' total family incomes and the number of dependents in their families.

In the late 1970s, before the reforms took place, the shop floor became the arena where workers gained more leverage through bargaining with their work team leaders on production quotas, flexible work hours, loose labor discipline, etc. My interviews with two former team leaders revealed that after political and ideological mobilization lost its momentum in stimulating workers' morale to work hard, and after production performance lost its importance among most workers, the only way to get workers to listen to team leaders was the personal quality of the team leaders. One former team leader said that he had to work extremely hard in order to get others to follow. Skills were also important for a team leader to win respect. I asked almost everyone with whom I spoke whether a team leader had power over the distribution of housing, and they all gave me a negative answer. One former leader said that in some cases, upper management would ask team leaders for appraisals. But team leaders did not know how seriously management would take their recommendations; more importantly, they were afraid that their negative comments might hurt their leadership if they were heard by the workers.

In sum, a socialist factory was as much a contested terrain as its capitalist counterpart, even though the organizational structures and ideology behind management differed in significant ways. It is too simplistic to conclude that managers either had absolute control or did not have any control, given all the structural and ideological conditions within factories. We find that both parties often maneuvered within the system and used the official discourse for their own gains. In the course of interaction between the system and actors and between managers and workers, conflicts arose and became influential forces behind the development of socialist labor relations.

Bargaining in the Reform Era

The economic reforms launched in the late 1970s brought structural transformation to industrial organizations, especially to the well-known danwei system. Researchers in the field have identified three main structural changes as important factors in reconstructing labor relations in today's China: (1) firm ownership, (2) the employment system, and (3) the welfare distribution system.

The faltering SOEs have been rapidly restructuring ownership since the 1990s. In many enterprises, the state is no longer the sole owner or even an owner at all. While some scholars believe that ownership is one of the main causal variables that shape labor relations,Footnote ¹⁹ I argue that the more significant change that actually precedes the ownership shift is the state's withdrawal from decisionmaking related to production and employment. The decentralization of state power actually paved the way for the ownership changes and gave managers more autonomy in controlling production and the labor force within enterprises, including the remaining SOEs.

Regarding the employment system, the most important change is that lifetime employment has been replaced by the labor contract system. Some argue that the labor contract system is a more rational and effective measure for both management and workers because it provides institutionalized certainty and security.Footnote ²⁰ Ideally, this might be true; in reality, however, with an oversupplied labor market, we find that the new labor contract system has actually increased managers' control over the labor force and has enabled them to use it as a potential threat to enhance shop-floor discipline. For most workers, it does not mean security but rather the loss of their so-called “iron rice bowl.”

The wage system has also changed in management's favor. Both piece-rate and fixed-rate wages are used in most factories. If the fixed-rate wages are mainly determined by the labor market, the piece-rate wages are simply up to management's discretion. The bonus is now more closely related to workers' performance as recorded by management. Performance is gauged not only by the quality and quantity of the products, but also by whether a worker is obedient to managerial commands or not.

There is no doubt that the state's withdrawal and the radical changes in both the employment and wage systems have considerably empowered enterprise management. Since the reforms, management has also been able to tighten labor discipline markedly. Many rules and regulations have been established for workers with detailed monetary punishments for violations.Footnote ²¹ As some scholars have argued, enterprise management is by nature despotic.Footnote ²² Nevertheless, it is false to conclude that workers have become completely tamed by the despotic management under crowded labor market conditions. On the contrary, despotic management only results in more resistance from workers. While public protests are often carried out by laid-off workers, the conflicts between management and workers who have jobs usually take place on the shop floor, mostly in the form of bargaining.

One assembly worker, Mr. Xia, at the crushing machine company (I call it Stone), told me a story about his foreman and himself. In order to break a production bottleneck, the foreman, Mr. Wang, asked Mr. Xia to work on Saturday. But Mr. Xia said he could not work on Saturday because he needed to help a friend paint his house but would come to work on Sunday. Mr. Wang then suggested that Mr. Xia come to work on Saturday and help his friend on Sunday. Mr. Xia told me that, considering Mr. Wang had always been nice to him, he agreed to come on Saturday. When I pushed him a little, he admitted he did not have to help his friend on Saturday, but he wanted Mr. Wang to know he had changed his plans for him.Footnote ²³ It was not a hard bargain but was enough to make clear that his compromise was a favor to the foreman.

In one of the two elevator/escalator manufacturers (I call it Triumph), I observed another bargain between a machine operator and his section chief. The section chief assigned the machine operator to test a new machine the company had just bought. The operator only agreed to do the job once the section chief promised to reward him with extra work hours. Although work hours were no longer related to bonus distribution in this particular company, the worker bargained both out of an old habit and for the possibility that he might work less in the future to meet his quota.

Initially, these two incidents appeared so trivial and individual to me as an outsider. I gradually realized, however, that on the shop floor any bargain proposed by workers, won or not, is considered a “loss of face” (a shame or embarrassment) for managers. If a manager's demand was not met by any challenge and was able to elicit workers' voluntary cooperation, he would be considered to “have face” (you mianzi), enjoying his workers' respect. Workers, on the other hand, could decide if they would “give face” (gei mianzi) to their managers or not. “Giving face” was often viewed as a favor to their managers. One day I observed a quarrel between a crane operator, Mr. Zhou, and an assistant workshop director, Mr. Shen, at Stone. Mr. Shen wanted Mr. Zhou to move a few parts from one place to another on the floor. Mr. Zhou suggested he do it later when all parts arrived together. Again, this was not a big deal in my eyes, but Mr. Shen immediately got upset—not because the job was urgent, but because Mr. Zhou had rejected his order in front of many others on the team plus me, to whom he once bragged that his commands never met any objections from his subordinates. Feeling “loss of face,” he blushed and said to Mr. Zhou, “No wonder other people say that we have too many crane operators over here!” Since the possibility of “lay-off” was such a sensitive issue among the workers, who consciously or unconsciously knew it was their vulnerability, words of this nature could be easily taken as an insult and inflame almost every worker at the scene. Mr. Zhou was outraged and demanded to know who the “other people” were to whom Mr. Shen referred.

This kind of iterative practice between shop-floor managers and workers often has, in the long run, a big impact on the values and norms of those involved. Workers' bargaining draws two different kinds of responses from shop-floor managers: to initiate favor exchanges with workers and to eschew picking fights but to record the worker's bargaining as a disobedient behavior on his or her monthly performance evaluation. In both cases, most managers try to avoid constant and overt confrontation with workers. However, the formal rules were altered (as examples will show later) and the expectation of and standard for a “nice boss” were conveyed to the managers in the bargaining process.

Mr. Wang, a workshop director who had worked in the same workshop at Stone for almost thirty years, often took the first approach mentioned above, that is to initiate favor exchanges with workers. To follow the company's rule, he arranged his workers to work overtime on the weekends. Workers, in general, hated overtime, especially on weekends and holidays, because overtime pay was low. Moreover, when overtime was no longer an ad hoc method to meet deadlines and was instead regulated on weekends and holidays by the company, workers viewed overtime only as something that the upper management forced upon them. On the shop floor, however, nobody bargained with Mr. Wang, for he turned overtime arrangements into a give-and-take opportunity. He followed the upper management's requirement to schedule workers on either every Saturday or Sunday, regardless of whether or not there was an urgent task to be done. Overtime remained mandatory as the company management intended, but he gave workers who did overtime regularly more flexibility during work days.

He also offered overtime to those who needed more work hours to be eligible for their monthly bonus. For example, the assembly team in the workshop usually had more than the assigned work hours to collect their monthly bonus, but the machine operators in the same shop often did not have enough work to do and therefore had difficulty collecting even the basic bonus at the end of each month. This was due to the fact that a large number of parts were now sent to outside subcontractors at a lower cost. If calculating on market principles, management should have considered those operators “superfluous” and let them go. Mr. Wang obviously thought differently. He offered help to the operators instead, arranging for them to work overtime for the assembly team on weekends in order to compensate for their possible loss in bonus. He also offered overtime to any of his workers and staff members if the number of days they were absent from work that month had reached a point that they would face a significant cut in their basic salary and bonus according to the company rules.

Mr. Wang's behaviors seem completely “irrational” if judged from a financial point of view, but I believe he followed a different kind of “rationality,” which was very much a reaction to workers' demands. For workers, the gain was not only financial. They preferred to interpret Mr. Wang's gesture as being “nice,” “caring,” and “paying respect” to them. In return, they rarely bargained with him and called it “giving Old Wang face.” In building this kind of cooperative relationship instead of bargaining, workers could choose to do favors for managers.

Here is another example to show how company's rules were possibly altered through shop floor practice. At Stone, the general management once decided that each workshop manager must issue 100 fines per month for product quality and labor discipline problems. If a workshop manager failed to meet the 100-ticket-per-month quota, he would suffer a deduction from his own bonus. Management saw this measure as a means to improve the control over production and labor, but workers regarded it as another absolutely stupid idea that management directed against them. One worker mocked the policy. “How could there be a quota for quality problems? Do you expect people to make exactly 100 mistakes a month?” Mr. Wang, who told me that he disapproved of this provocative rule, had to find a way to satisfy his superiors without offending his workers. In order to gain workers' sympathy, he made it clear to them that the quota was imposed by the upper management and was out of his control. Each time he issued a fine to a worker, he would tell the worker that he also wrote a ticket for himself. The most amazing thing to me was that many workers would voluntarily offer him their help. They told Mr. Wang that if he was unable to fulfill the quota, he could put their names on the tickets, which, of course, meant that in several cases they would be fined for something they did not actually do. Some workers were naturally concerned about the deduction of thirty yuan per ticket from their monthly bonus. An elderly worker comforted them: “Don't worry. Old Wang will definitely give you back this money one way or another.”Footnote ²⁴ (Mr. Wang was in his late forties but had been working at the same factory about thirty years. Workers called him “Old Wang” to show both respect and familiarity.Footnote ²⁵) They seemed to enjoy this “conspiracy” with their Old Wang in defeating the upper management that had tried all sorts of close monitoring.

At Triumph and the third factory I visited (I call it Mars), the shop managers reacted to workers not entirely the same way as Mr. Wang did at Stone. To be sure, there were also some favor exchanges on the shop floor, which would be initiated by the managers at Triumph and Mars. For example, at Triumph, I sat in the section chiefs' office one day, and the workshop manager, Mr. Chen, walked in. Mr. Chen said to one of the section chiefs, “Old Liu, did you notice that Old Cheng (a worker) worked until midnight last night? He lived too far away to go back home and had to sleep here. You'd better remember it!” Mr. Chen apparently wanted to show he was a nice and caring boss with a fair mind. But, because he was still new there, most of the time he used methods such as close monitoring and reducing the work hours assigned to certain jobs to control the work pace. He often provoked confrontation. One day I saw a young worker yelling at him, “You can't watch me all the time! I am not a child!” He then yelled back, “I wouldn't have to if only you would behave yourself!”

Mr. Chen also relied on an evaluation system: each employee received an evaluation every three months and then an annual one, on an A-D scale. The index of the evaluation form suggested that every standard used to judge a worker was measurable, but shop-floor managers included a judgment on whether a worker obeyed managers' orders. This measure, however, was equally provocative, and the section chiefs all told me that confrontations with workers were inevitable around the time an evaluation was given.

At Mars, a Chinese-Japanese joint venture, I observed managers trying to shun confrontation if there were other people around. This does not mean, however, that they ignored any “disobedient behavior;” instead, they recorded workers' “discipline problems,” and these records, along with production records, were used to distribute monthly, seasonal, and annual bonuses. Mars maintained a “fuzzy bonus” system and kept the concrete amount of bonuses confidential. Managers believed that “fuzzy bonuses” were a method of control to thwart bargaining attempts. One manager (also with a last name Chen) said to me:

He admitted that when deciding how much bonus a worker will get, not every standard used for judgment is measurable. These unquantifiable standards are often the gray areas about which workers could argue or bargain with managers. The fuzzy bonus system could not really prevent workers from knowing how much their peers were paid but could block them from talking to managers about any comparison since they were, theoretically, not supposed to know. As one worker put it, “The management probably thought that the fuzzy bonus would reduce conflicts among us. Actually, the conflicts are not reduced but only suppressed.”Footnote ²⁷ That might be true; but I do not think that the management's only goal for a confidential bonus system was to reduce conflicts among workers. What the management intended with differential bonuses was, in my opinion, to convey what management expected from workers besides those written standards, such as the monthly quota. Managers hoped to encourage voluntary cooperation. As in the quotation above, when managers talked about “skills,” they meant an obedient “attitude” toward work and authority. Although the system successfully stopped workers from accusing managers of open favoritism, it is questionable if it could turn the discontent into a complete silence and therefore maximize labor output. As one company director at Triumph admitted to me, “It is about money; but it is not all about money.”Footnote ²⁸

On the surface, bargaining and favor exchanges always seem to exist between shop-floor managers and individual workers, but, when they become daily practice, their impact on the formal rules and on the formation of labor relations is obvious. Many managerial methods are designed to control labor output, and dismissal is indeed a serious potential threat to Chinese workers who face a seemingly endless cheap labor supply. Despite all these macro-level structural and social conditions, workers still have leverage and know how to utilize it in the production process, especially with managers who wish for voluntary cooperation all the time. Shop-floor managers encounter the possible challenges that workers present to the system of rules and regulations as well as to managerial command and production tasks beyond a labor contract.

To the challenges discussed above, there were two main kinds of managerial response. One was to make deals with workers through initiating exchanges of favors and establishing a “benevolent boss” image. Through favor exchanges and other interactions, both parties constructed hidden and tacit agreements. The other response yielded few results. Managers mainly relied on the authority granted by the system and the power of meting out rewards and punishments. Unable to reach a deal in which workers' autonomy and managers' authority were mutually respected, labor relations were rather contentious. In both cases, however, I constantly observed the bargaining that workers took into their own hands, and it often altered the managerial blueprint.

The Significance of Shop-floor Bargaining

Unlike bargaining around the table between management and unions, shop-floor bargaining is hidden in the sense that it is not an organic part of the formal structure and thus not institutionalized. Nevertheless, it is full of dynamics and improvisation. In most cases, shop-floor bargaining aims at material gains and flexible work hours, but it certainly entails more than that. Workers also demand respect and autonomy. Especially under socialism, material incentives were very limited. People closely watched how material goods were handled and eventually turned it into an equalitarian distribution.

To understand the impact of shop-floor bargaining, we need to first locate where the bargaining power comes from. As previously demonstrated, despite the labor contract and many written rules, there is always room for shop-floor managers and workers to maneuver. In the socialist period, things were even more complicated. Due to the conflicting managerial ideologies and methods and the discrepancy between the official discourse and reality, the implementation of rules and regulations often met challenges, and workers could use the same discourse—intended to mobilize them—as the justification for their resistance to managerial demands. Many labor studies have indicated that managers and workers on the shop floor are inevitably interdependent. For example, in the reform era, bottlenecks in production were almost expected in every factory I visited. Overtime thus became unavoidable. Although most of the companies did pay workers for overtime, workers could use it to bargain with managers since overtime pay was usually not written in the contract. Some of the companies, such as Stone, made overtime mandatory to strip workers of leverage in this area. But this method often encountered fierce resistance, which forced the shop manager to alter the rule in one way or another. Testing new machines offered workers another opportunity to bargain with managers for something in return. The problem for managers was not what they compromised but the compromise itself. On the shop floor, workers and managers shared this particular norm: If a manager's request was challenged or became negotiable, the manager would “lose face,” which often pushed managers to initiate an offer and make it a favor bestowed upon workers instead of ending with a forced compromise.

Comparing the two different historical periods, we see that the workplace has experienced profound organizational changes when the state reformed its economic structure and withdrew from the direct involvement in enterprises. The macro-level policy changes did have a great impact on reconstructing authority relations within factories, as managers gained more control over production and employment. Consequently, a different labor regime emerged with tightened labor discipline as well as production quotas. No matter how different the two regimes are, this paper shows that in both time periods, the shop floor was a contested arena of sorts. The constant power struggles between managers and workers marked the important active participation of workers in shaping labor relations. In other words, workers' responses to the management did not always produce the outcomes that policy designers had anticipated. In China, when public protests were not supported by unions and could indeed put workers in political danger and risk their employment, the bargaining forms they utilized on the shop floor often had a more hidden and subtle function in changing many kinds of managerial practices that were invented for control. More importantly, bargaining was imbued with normative influences over management. Using their leverage, workers could choose to voluntarily cooperate with “nice bosses” or not to cooperate with “bad managers.” In doing so, they often successfully sent the message about what they would expect from a nice boss, i.e., that they could not be pushed around and that their autonomy and dignity needed to be respected.

Bargaining is indeed not always about money, even though it may appear that individual material gains are sought most of the time. At Mars, one of the three joint ventures I visited, the annual bonus for workers was six times higher than their counterparts' bonuses at other companies, such as Stone and Triumph. The high income made working at Mars very desirable, and workers wanted to keep their jobs there. However, that does not mean they would subject themselves to managerial demands without complaints, bickering, and all sorts of skirmish. Their high bonus enabled them to reach a middle-class income level. Their rising economic status made them as keen as workers in other companies to defend their individual dignity, if not more so. Ironically, compared with their counterparts in the other two workplaces, it was also harder for the workers at Mars to make managers bend to their demand for respect, since the Mars managers believed that the unusually high pay by the company was enough to buy workers' full surrenders.

Does this kind of individual bargaining on the shop floor in some way stabilize the capitalist system rising in China as the “make-out” game did in Michael Burawoy's renowned study? Empirically, the two cases are not exactly the same. As Burawoy points out, in the mid-1970s in the US, monopoly capitalism itself provided workers a limited but critical freedom from managerial intervention, and workers were able to constitute work as a game.Footnote ²⁹ In the 1990s in postsocialist China, a state that played a noninterventional role in the labor process, the absence of independent unions, overcrowded labor markets, and a dominant ideology that promoted “scientific management” as part of building a modern nation all contributed to the increase and expansion of managerial power within industrial organizations.

A labor regime formed under this kind of circumstance ought to be “despotic” if we use Burawoy's theoretical framework and typology, as some researchers have already done.Footnote ³⁰ In other words, Chinese workers discussed in this paper bargained under despotic management while American workers in Burawoy's study played games in a “hegemonic” labor regime. The significant differences in labor conditions indicate different orientations and outcomes. If the game playing Burawoy discussed generated consent to the rules, the workers' choice here was very much defined by the system. Workers' bargaining on Chinese shop floors was, however, an attempt to redefine the system, alter despotic management, and win respect and dignity. As I have argued above, workers tried to utilize all the possible leverage they might have, and managers found all the possible ways to defeat them (e.g., to make overtime mandatory). The conflict between managers and workers was much tenser and more severe in the Chinese case.

Certainly, compared with organized collective protests on a large scale, the impact of individual shop-floor bargaining on macro-political, legal, and economic systems may not be as big, direct, and visible. But we shall not reduce its function to the influence of only a few individual managers either. Individual bargaining represents a choice workers created for themselves: voluntary cooperation to show respect to managerial authority and “give a manager face” or forced compliance with resentment and discontent. Its effectiveness, however, very much relied on the shop-floor culture workers built together with shop managers. In other words, to make managers care about workers' choice, workers must be able to create collectively a shop-floor ethos in which a manager's despotic attitude not only always provokes confrontation, but is also considered equivalent to incompetence. This normative construction requires time, daily practice, and managers' eventual participation. The age of a manager on the shop floor is thus less important than the years a manager worked with same group of workers in the same workshop. For example, Mr. Wang, the shop manager at Stone, and two managers with the same last name, Chen, at Triumph and Mars, were all in their late forties or early fifties. Wang worked in the same workshop for almost thirty years, and both Chens were new (six months to one year) to their jobs. That mostly explains why Wang was completely sucked into the shop-floor value and normative systems while both Chens were experiencing a more conflicted relationship with their workers.

Studying shop-floor bargaining provides us an opportunity to examine grassroots resistance and its outcomes in two time periods in China. It is not that socialism produced a docile labor force ready for capitalists to use conveniently. Rather, socialism created experienced industrial workers who carried along egalitarian principles and knew how to use the official discourse to their own advantage. They certainly no longer regarded themselves as the masters of the nation, for they had become disillusioned even before the economic reforms; but by actively taking bargaining into their own hands, they asserted that they were not slaves either. The assertion renders an ideological message to those who attempt to rule without consent: the water that carries a ship can also sink it.