THREE: COMPLEXITY OF FEMALE SEX WORKERS' COLLECTIVE ACTIONS IN POSTSOCIALIST CHINA

Abstract

This paper explores labor resistance amongst rural migrant karaoke bar hostesses, many of whom are sex workers, in the northeastern Chinese city of Dalian. I argue that hostesses are subject to exploitation and violence because of hostile public policy and the organization of the sex industry, both of which limit the possibilities for labor organizing based upon localistic networks. While hostesses do form alliances based upon their native place, which supplies them with financial and emotional support, these localistic networks are transient and temporary because hostesses aspire to deflect from their group as a "criminal" group and they also face high internal competitions. [PUBLICATION ABSTRACT]

Headnote

Abstract: This paper explores labor resistance amongst rural migrant karaoke bar hostesses, many of whom are sex workers, in the northeastern Chinese city of Dalian. I argue that hostesses are subject to exploitation and violence because of hostile public policy and the organization of the sex industry, both of which limit the possibilities for labor organizing based upon localistic networks. While hostesses do form alliances based upon their native place, which supplies them with financial and emotional support, these localistic networks are transient and temporary because hostesses aspire to deflect from their group as a "criminal" group and they also face high internal competitions.

Introduction

This paper is based upon my fieldwork on karaoke bar hostesses in the port city of Dalian, in Liaoning Province, China.2 The rapid growth of the city from a fishing village in the 19th century to a metropolis with a population of 6 million3 has made Dalian a magnet for labor migrants. By the year 1998, the most conservative estimate placed the number of the floating population (unregistered migrants from the countryside) in Dalian at around 300,000.4 Institutional (i.e., the household registration policy) and social discrimination force the vast majority of these migrants onto the lowest rung of the labor market. Migrants commonly work as construction workers, garbage collectors, restaurant waitresses, domestic maids, factory workers, and bar hostesses.

A large number of female migrants find employment in Dalian's booming sex industry. These companions, or hostesses, are referred to in Chinese as "sanpei xiaojie," literally young women who accompany men in three ways - generally understood to include varying combinations of alcohol consumption, dancing and singing, and sometimes sexual services. Mainly 17-23 years of age, their services typically include drinking, singing, dancing, playing games, flirting, chatting, and caressing. Beyond the standard service package, some hostesses offer sexual services for an additional fee. Their monthly incomes range from the lowest of 6,000 yuan (U.S. $7505) to tens of thousands of yuan.6 It was roughly estimated that in 1991, more than 800,000 hostesses were involved in sex work (Pan, 1999, pp. 13-14).

The purpose of this paper is to explore and understand the complexity of life and resistance of rural migrant karaoke bar hostesses in the urban Chinese sex industry. I will argue that hostesses in Dalian are subject to exploitation and violence due to the hostile political policy and the organization of the sex industry. My research on hostesses both resonates with and differs from the previous research that argues that migrant women's informal ties based upon localistic networks projects the possibility for labor resistance. On the one hand, hostesses do form alliances based upon their native place, which supplies them with financial and emotional support. On the other hand, these localistic networks are transient and temporary because hostesses aspire to minimize the stigma of membership in a "criminal" group and they also face high internal competitions.

The creation of the informal networks is necessitated by the hostesses' confrontation of such overwhelming exploitation and violence. Hostesses' response to the exploitation and violence is unique in that although some basic factors discourage long-term and formal organization, they do organize informal networks. While these networks are fleeting and unstable in nature, they are critical in allowing the hostesses to survive and even flourish. This paper will illustrate the forms of exploitations and violence hostesses are confronted with and explore their tactics of resistance in their working lives. This paper unfolds in five parts. I will first briefly discuss my fieldwork in Dalian. I will then describe the political policy towards prostitution and the hierarchy of the karaoke bar industry. I will follow these two sections with an account of the kind of violence and exploitation of the hostesses as a result of such political policy and the sex industry. I will then discuss how, within such a hostile environment, hostesses form small-scale and unstable organizations to secure their livelihood and conduct their work.

Living with and Working as a Xiaojie (hostess)

The argument I developed here is based on over 20 months of fieldwork and a research sample of approximately two hundred bar hostesses from three karaoke bars between 1999 and 2002 in Dalian.7 My initial research encountered a series of obstacles. First, my requests to conduct research in karaoke bars were repeatedly turned down by every bar owner I talked to. Second, officials were reluctant to "reveal any secret information" to me. As they contended, their own political positions would be at stake were they to do so. Once, as I was singing songs with a couple of political officials in a karaoke bar, one of the officials suddenly grabbed my purse and started searching the inside. I was too shocked at the moment to ask what he was looking for. In the end, he did not find anything except my keys and a mirror, and then returned the purse to me. He appeared much more relaxed after investigating my purse. Later on, when I recollected the scene, I knew that he was checking my purse to see if I had hidden a tapeor video-recorder there. The nature of my research raised everyone's suspicions of my "real identity" and placed a barrier in my relationship with them.

As time went by, my persistence in conducting the research and my sincerity in interacting with the political officials bore fruit. I could see the suspicions subsiding and the tensions thawing. Some political officials even jokingly told me that they initially believed I was a spy from the United States. Openly acknowledging their suspicion was big progress from their previous silence. At that time, one political official friend asked me to be his daughter's English tutor; I agreed. In exchange, he introduced me to some karaoke bars to conduct my fieldwork. He himself was a regular customer of these establishments and therefore familiar with their proprieties. My identity and my research were fully explained to the bar owners. I assured them that I would not publicize the bar names to the outside world.

Bar owners introduced me to hostesses as a university student from the United States, who had come to do social research. As they explained, I had come to get a "real taste" (tihui) of hostesses' life in order to write a book on hostesses' lives. My presence was greeted lukewarmly. Hostesses found it difficult to understand why anyone would want to study them. One hostess warned me that a female college student such as myself had been raped and murdered by a psychopathic customer while conducting her research. I do not know if this story is true, but it did shake me up a bit.

In the karaoke bars, my initial attempt to interact with hostesses was not very successful. The hostesses ridiculed my student attire, my glasses, my inability to understand or participate in their sex talk and jokes, and disbelieved that I would ever be able to understand their lives. Facing random assaults by the police and hooligans in their dangerous environment, each of them used a fake name, a fake hometown, and a fake personal story. To overcome the barriers, I decided to increase the amount and intimacy of my interactions with them. I handed in the rooming fees to the bar owner and lived and worked with the hostesses as a hostess myself in the karaoke bars. Thereafter, I was intensely involved in every aspect of their lives.

A typical day in the field was like this. We got up around 3 PM and ordered a light meal from a nearby restaurant. The remainder of the afternoon was free for shopping or visiting the beauty parlor. We ordered dinner at around 6 PM. Around that time, the first customers of the evening would begin to trickle into the bar. While waiting to be selected, we sat in the bar lobby watching video compact disks (VCDs) or TV and chatting. Around 12 a.m. we ordered breakfast, and went to bed between 2 to 3 a.m. Visitors were mainly middle-aged businessmen, government officials, policemen, and foreign investors. Hostesses earned obligatory tips in exchange for singing, dancing, drinking, playing games, and conversing with the clients. At times the service could also include kissing and fondling sexual parts, but hostess can legitimately refuse to perform genital or oral sex acts with her customer.

It was not my initial intention to research hostess-client dynamics by directly servicing clients as a hostess. However, objective circumstances mandated that I wait on clients because my personal profile fits within the range of hostesses' typical characteristics. I am Chinese and female. My fieldwork was conducted during my twenty-eighth and twenty-ninth years of life, which put me in the "autumn" years of a hostess's career lifespan. This meant that a customer who saw me sitting in the KTV bar lounge would naturally assume that I was a hostess. I was also obliged to minimize the disruption of my research on the bar's normal business operations. For a hostess to refuse to wait on a customer is simply unheard of. This meant that if a customer chose me to wait on him, it would be very difficult for me to refuse.

To avoid clashes with customers, I asked hostesses for their advice on how to thwart customer's sexual advances. Despite the precautions, I became embroiled in several conflicts with customers. During these conflicts, it took the combined efforts of bar owners, bouncers and hostesses to keep me out of harm's way. I am indebted to them for their advice and, at crucial moments, direct interventions. To extricate me from precarious situations, hostesses dispersed attention and energies that they would have otherwise expended on profit-making matters in order to look after my well-being. Without their sacrifices, my research in the bars would have been too dangerous to continue.

During the fieldwork, I conducted open-ended interviews with urbanites from all walks of life, including political officials, to elicit their perspectives on bar hostesses and information on labor options for rural migrants that are provided by the changing economy. I conducted informal and formal interviews and participated in observation inside and outside of ten different bars and in two hostesses' rural hometowns, embedding myself into the hostesses' working and leisure lives. I documented their consumption habits, the recreational places they frequented, the linguistic forms they emulated or avoided, and their clients' social networks that they attempted to enter. I also recorded the nuanced internal hierarchy among bar hostesses and the informal networks they entered to survive the urban life inside and outside the bar.

I also followed hostess informants during these years to flesh out their longitudinal development throughout their professional and life trajectories. To better situate and comprehend their constructed "urban" identities, I charted in further detail the hierarchical structure of karaoke bars in the city and the hostesses' varying life experiences according to the position of their workplace within this structure. I visited their rural home bases to learn about their rural backgrounds, their social relationships with family members and fellow village residents, perceptions of their transformed identities in the villagers' eyes, and the special role they played in their hometowns as brokers of urban "modernity."

Prostitution and the Post-Mao State

Maoist China declared the eradication of prostitution in 1958, yet the increased leniency of the post-Mao state has opened the way for the reemergence of nightclubs and other leisure sites and erotic services since 1978. To avoid any residual negative connotations left over from the previous era, nightclubs in the current post-Mao period were referred to as karaoke bars, karaoke plazas, or liange ting (literally, "singing practice halls"). These new consumption sites were prominent in the more economically prosperous SEZs (Special Economic Zone) (Jian, 2001, p. 44). Clients could partake of the services offered by hostesses and at the same time engage in "social interactions" (yingchou) that helped cement "relationships" (guanxi) with their business partners or their patrons in the government (Wang, 1999). Hostesses played an indispensable role in the rituals of these male-centered worlds of business and politics.

Like the Maoist state, the post-Mao state enforced an abolitionist policy. The 1991 Two Decisions on Strictly Forbidding the Selling and Buying of Sex and on the Severe Punishment of Criminals Who Abduct and Traffic in or Kidnap Women and Children, the 1992 Law on Protecting the Rights and Interests of Women (Women's Law), the Criminal Law of 1997, and the 1999 Entertainment Regulations link prostitution with trafficking and define both as a violation of women's human rights (Jeffreys, 2004). The "erotic company" of hostesses, pornographic TV shows, erotic performances, and prostitution within karaoke bars are condemned as "cultural trash" that "destabilize state rule and the socialist system."

The state launches periodical nationwide anti-prostitution campaigns - "strike hard" to ensure "cultural purification" and "spiritual civilization."8 The Public Security Bureau employed a complex system of raids to attack karaoke bars. The techniques of raids are self-described as "guerrilla warfare" (da youji) in reference to the heroic efforts of the Communist revolutionaries against the Japanese invaders and nationalists. Raids are divided into several types, including: "regular raids and shock raids, timed raids and random raids, systematic raids and block raids, daytime raids and night raids." Those units and individuals that perform well-measured in the number of arrested hostesses and amount of fines levied-receive high honors and cash bonuses from their municipal government.

The state policy is distorted and derailed by the self-seeking behavior of local officials and police. If hostesses' sexual services are disclosed by the clients to the police, the hostesses would be subject to extreme humiliation, arrest, handsome fines, and incarceration. Indeed, in their everyday lives, the local police constitute their daily fear and terror. Because the police wield arbitrary power, the hostesses find it obligatory to obey their sexual demands without monetary compensation. Local officials' exploitation of hostesses is not just limited to sexual and economic benefits; some of them also maintain a group of 'spy hostesses' that report on bar conditions as well as act as these officials' personal harem. In exchange for these services, hostesses gain immunity from police sanctions.

The anti-prostitution campaign also allows the bar owners to impose more severe regulations and disciplines on the hostesses who otherwise would operate in a more laissez-faire manner. Because the state's anti-prostitution policy is manipulated and usurped by local officials and bar owners for their own ends leading to a violent working environment for the hostesses, hostesses do not disclose their real identity, which makes it more convenient for men to be violent towards them and even to murder them. It was reported that in the city of Shenyang, more than 100 hostesses were murdered in 1999 (Anonymous, 2002). In Dalian, hostesses' bodies were found murdered on the street, but the police could not identify who they were (Sun, 2003).

The Hierarchy of the Bar Industry

Karaoke bars were the most expensive type of bar in Dalian, with prices far exceeding the level in other recreational spots, such as coffee bars, pottery bars,9 soft-drink bars and music bars. Within the karaoke bar industry, bars were ranked vertically according to city locale, exterior and interior furnishings and facilities, organization and management, the number, beauty, educational level and turnover of bar hostesses, its clientele, and its cost level. I conducted field research in ten karaoke bars. Most of my time and effort, however, was concentrated on three bars -Colorful Century (high-tier), Prince (middle-tier) and Romantic Dream (low-tier).

An Upper-tier Karaoke Bar

"Colorful Century" was housed in a four-story building just off Zhongshan square. Zhongshan square was the financial heart of Dalian where the most important banks were located. The city of Dalian was composed of six districts, three county-level cities, a county, and three special administrative zones. In the eyes of the Dalian urbanites, only four centrally located districts were considered the "city" of Dalian, and the rest were adjacent "countryside." These four districts were classified hierarchically according to their locations in the city - from the center to the periphery as Zhongshan, Xigang, Shahekou, and Ganjingzi. Karaoke bars in Zhongshan district had the highest status.

The façade of "Colorful Century," was draped with strings of small and colorful blinking lights that fanned out like the trail on a Western wedding dress. The words, "Colorful Century," in large, electrified Chinese characters hung over the entrance. The doors themselves were fringed with fake plastic ivy and more flashing lights. Two security guards stood at either side of the entrance gate. Over a hundred and fifty seductively dressed women were gathered on the left side of the lobby. They sat on three rows of benches like the audience to an unseen performance. In fact, however, they themselves were the observed. A camera head from the ceiling fed in a live wire to monitors installed inside each private room. Customers could select their escorts from the comfort of couches while sitting in these suites. The stairs leading to the upper floor private rooms were lined with two teams of waitresses, all of the same height and with the same hair-style, dressed in the same dark embroidered miniskirts, their breasts half exposed, and their hips scarcely covered. They simultaneously saluted customers, "Good evening!"

The hostesses in this upscale karaoke bar were taller with greater facial beauty and more shapely bodies than those in lower bars. They were also better educated; four of the hostesses from this group had graduated from high school. They ended up in their current karaoke bars in various ways. Some were introduced to a bar through their friends, previous coworkers or hometown friends. Many brought their sisters to work in the bars. Some simply called a taxi when they arrived in Dalian and had the driver take them to the "red light district." Few were bought by the bar owners from pimps.

The upper stories of this bar were divided into five sections of 10 karaoke rooms each. There was also a fitting room for hostesses that could serve as a hideout during police raids. The karaoke rooms provided privacy. Each room was equipped with foreignmade karaoke equipment, including a 29-inch TV set that continuously played excerpts from erotic western videos. Each karaoke room was furnished with an air-conditioner, rosewood furniture, beautiful window drapes, wallpaper, carpeting, magnificent dim ceiling lights, a big couch, and an end table. There was a space between the TV and the end table where clients could dance with hostesses. They could dance either to each other's singing or to the dance music chosen from a song booklet. The couch could be unfolded into a bed at the request of clients. Many karaoke rooms had adjunct secret bedrooms separated by a curtain camouflaged to look the same texture and color as the wallpaper. This was designed to prevent discovery in case of a police raid.

The karaoke bar was on constant alert for raids. Equipment was provided to ensure the safety of the bar and all its habitants. An alarm button was set up in each karaoke room so that if police made a surprise visit, the alarm could be activated to alert everyone in the karaoke rooms allowing them to hide and destroy any evidence of illegal activity (e.g., refold beds into the couch position). In addition, each staff member carried an interphone with a long antenna that could be used to communicate with others during raids and facilitate the coordination of emergency measures. Each time hostesses "sit on the stage" (zuotai), they received a minimum tip of 200 yuan (U.S. $23) from the clients, twenty yuan (U.S. $2. of which was paid to the bar as the "stage fee" (taifei). If hostesses went offstage, the minimum tip was 400 yuan (U.S. $48) and hostesses had to pay fifty yuan (U.S. $6) for the "off-the-stage fee." In each karaoke room, madams and waitresses each received a tip of 100 yuan (U.S. $12) from the clients.

A Medium-Tier Bar

Similar scale karaoke bars like "Colorful Century" but located in the other more peripheral districts were rated medium level. The medium category karaoke bar was less extravagantly decorated both on its exterior and interior. It was still quite upscale, with three one-story buildings respectively named "Prince Branch 1," "Prince Branch 2," and "Prince Branch 3." The three branches were built next to each other. Each branch had 80-100 hostesses and 23-30 karaoke rooms. Hostesses were housed in a big dormitory building behind the bar. A much smaller percentage of these women, only three, graduated from high school, the highest degree in their group.

The entrance doors were adorned with electric light bulbs, but less extravagantly than in "Colorful Century." As we entered the bar, the first thing we saw was the bar counter. The hallway had a wooden floor instead of a carpet that one saw in the high-tier bar. The hallway led to a long corridor, each side of which was lined with a series of karaoke rooms. Hostesses were sitting, standing, or wandering in the corridor. Usually they gathered in two karaoke rooms watching TV, until newly arrived clients were ushered in and they were evicted. The interior of the karaoke rooms was decorated in a fashion that was similar to those in "Colorful Century," except that there were no foldable couches or adjoining bedrooms. Couches in this bar can sit five to six people and cannot be folded into beds. The karaoke equipment was manufactured in China. There was no minimum charge and the spending level was fifty percent lower than the high-tier bar but fifty percent higher than the low-tier bar. The karaoke room rental fee was also fifty percent lower than in the high-tier bar.

Hostesses working here had to meet four requirements: First, they must have worked as hostesses before - experience was very important. Second, they must be able to consume a great deal of alcohol. Third, they must dress well. Fourth, they had to be sexually open-minded. Hostesses were expected to lead customers in their consumption of beer, hard liquor, and snacks to boost bar revenues. However, there was no fixed minimum charge, which alleviated some of the pressure that forced hostesses in the hightier bar to daily gorge themselves on liquor and food. The abuse of hostesses was relatively less serious than in the high and low tier bars. On average, hostesses' "on-stage tips" fell into the range of 100-400 yuan, (U.S. $12-48) according to the quality of service and their clients' generosity. Their minimum offstage tip was 300 yuan (U.S. $36).

A Low-Tier Bar

The low-tier karaoke bar "Romantic Dream" was located in an enclave of twenty six karaoke bars in a seedy neighborhood huddled under a suspended railroad in Shahekou district. It was known as the "red light district" and was notorious for its polluted conditions, low-quality clients, and aggressive hardcore thugs and criminals. When you entered "Romantic Dream," the open door pulsated with electric light bulbs. One or two hostesses in sexy clothes and heavy make-up sat at the entrance doors, smoking. The open doors gave every passer-by a very clear view of the hostesses who were sitting inside. In "Romantic Dream," prices were negotiable, including hostesses' tips. There was no karaoke room rental fee. There were twenty-seven hostesses in the bar. Only one hostess had graduated from high school, the highest degree in their group.

The "Romantic Dream" had two floors. Upon entering the bar, one immediately encountered the hostesses in the hallway, sitting on the couch in front of the bar counter or on the chairs around the hallway, watching TV or making themselves up. The hallway floor was tiled, and the couches were very dark from grime. There were two corridors of karaoke suites on each side of the bar counter. There were twenty karaoke rooms on the first floor. Three karaoke rooms on the second floor consisted of a waiters' dorm, hostesses' dorm, and a private room for sexual encounters (paofang). The private room was specifically prepared for clients to sleep with hostesses. The lights in these karaoke rooms were as dim as those in the upper- and medium-tier bars. The interior was decorated similarly except that it was filthy. Beer, tea, water, food crumbs, and leftover food were spilled all over the floor. The walls were plastered with discarded dark chewing gum, and the couches were filthy and oozed a strange and unpleasant odor. The bathroom was also very dirty. "Romantic Dream" provided not only a room for private sexual encounters, but also a strip dance show by the hostesses in the karaoke rooms. The bar staff and hostesses tried not to spread the word too widely, lest these two items, which were illegal, attract police raids.

Violence and Exploitation in the Sex Industry

In all these three karaoke bars, clients, madams, bar owners, the entertainment managers, and security guards inflicted violence upon hostesses. In the high-tier bar, security guards' main job was to keep the hostesses from leaving before 12 a.m. and maintain bar security. Many times a team of the security guards rushed upstairs like soldiers to quell fights in the karaoke rooms. The suppression of disturbances itself always involved violence and blood. Unarmed or armed (with beer bottles, knives, and glass) fights between drunken clients and between clients and hostesses were daily routines in the bar.

Hostesses often came downstairs, crying from their injuries: their legs, arms, and breasts black and blue from the hard pinches of some clients. Some hostesses chose to endure whatever abuse they were subjected to, but some opted to quit and, consequently, received no tips for the time they endured. Those who clenched their teeth to see it through with smiles held back their tears and complaints for later when they sent off the clients and returned to the community of hostess. Once a hostess came into the fitting room to change from her dress into slacks because her client had been pinching her legs, and while she was changing her clothes, the madam came in, furious. The madam slapped her face very hard, and hit her head with her big interphone, yelling at her, "Don't you know that your client is looking for you? Why are you hiding here? I have looked for you everywhere! I am now ordering you back to your client right away!" The poor girl tried to dodge the madam's blows, as her face reddened from the madam's slaps. In an anguished voice, she agreed, "Sure, sure, I will go back right away." Fleeing the madam's blows and harsh words, she quickly slipped out of the room.

Hostesses not only obeyed the bar staff but also were responsible for ordering enough beer and snacks to meet the minimal expenditure of 400 yuan (U.S. $48). Some bars required that - in addition to other fees - customers consumed snacks and beverages of a certain monetary level. Known as the minimum charge, these requirements created an onerous burden for hostesses. Hostesses were held responsible for ensuring that customers' expenditures reach the mandated level. To stimulate customers' consumption, hostesses themselves had to continuously consume, especially alcohol.

This problem was especially pronounced in "Colorful Century" where the minimum charge was set very high at 400 yuan (U.S. $48). Here the bathroom was always filled with hostesses vomiting into the toilet, and then returning to their clients to continue drinking. Because of this daily alcohol overdose, most hostesses not only put on weight (which led to other selfdestructive, weight-loss practices) but also developed stomach problems, which, in severe cases, resulted in hospitalization.

Not only the bar owner urged hostesses to attract returning clients, but also the three madams managed and disciplined the hostesses. In the high-tier bar, when the three 35-39-year-old madams - "heads of the group" (lingban) - called on a team of hostesses and led them into clients' karaoke rooms for selection, the hostesses told me that the whole process depended on whether the madam presented you to the clients or not. Clients often looked for a hostess with a beautiful body and face, big breasts, and who was able to drink, sing English songs, sing/dance the best, etc. After inquiring about their preferences, the madams either strongly recommended several hostesses who "fit their tastes the best" or directly pushed some hostesses to those clients' side. So it was very important for a hostess to bribe one of the madams. Once I saw a hostess secretly handing 400 yuan (U.S. $48) to a madam at the entrance gate. The latter accepted it after feigning refusal several times. If the hostess did not bribe the madam, she would not call on her, nor recommend or promote her to the clients. Eventually the hostess would be left with minimal chances to be chosen.

This bribery produced three factions. Each faction was led by one madam, who would call only the hostesses in her faction to the clients' karaoke rooms. New hostesses had to decide which madam seemed nice before bribing her to be accepted into her faction. Similarly, in the medium-tier bar, the male "madam" often showed preference for some hostesses, whom he called more often and recommended more vigorously to customers. However, the favoritism and factionalism found in the high-tier bar was much less severe. While thugs and criminals could be seen in the upper class karaoke bars, their numbers and the level of violence was much greater in lower class bars. Gangsters were involved in attacks upon hostesses, plundering and theft of the bar, the sale of drugs and even murder. This red light district street was first built in 1996. For the first three years of its existence, dead bodies were found lying in the street almost every morning. The police could do nothing about it. During my fieldwork there, I witnessed many bloody fights and saw gangsters roaming the area.

The gangsters of Dalian were tough street fighters who were members of numerous criminal organizations (tuan huo). Government officials told me that there were close connections between gangsters and the local government that fueled crimes and violence in the city. Gangsters were scattered throughout the sex industry in the city working as owners and bouncers of karaoke bars, disco bars, sauna bars, etc.10 They also were involved in the abduction of women from other provinces (Wang, 2003, p. 2), and sold ketamine powder, ecstasy and other kinds of soft and hard drugs in the entertainment places. During the period of the World Soccer Cup, they organized illegal underground gambling activities. These activities were less bothered than protected by the police. However, at times the conflicts between different cliques escalated into violence. A recent violent incident occurred between two cliques in a hotel in Sanba square in Dalian in October of 2004. One was stabbed to death, one was shot to death, and a number of people were seriously injured. When the police arrived, gangsters resisted arrest and attacked the police with knives. In self-defense, the police shot one person to death. After the event, the remaining gangsters escaped. It was these gangsters that hostesses were dealing with on a daily basis.

Each bar on this street had to hire a thug as the barkeeper. This barkeeper had to be good at fighting, otherwise the bar would be forced to close down due to the harassment of roaming gangsters and thugs on the street. During my research in the bar, I witnessed numerous bloody fights between the bar keeper Bing, bar waiters and gangsters, clients, and passers-by. I saw Bing and bar waiters throw heavy stones and chairs at clients and some passer-bys' heads until blood streamed down their heads onto their faces. The bar owner told me that Bing, after having killed and severely injured many men in previous fights, was once sentenced to death. The bar owner spent a great deal of money to finally get Bing out of prison before hiring him as the bar guard. The mere presence of Bing in the bar kept many gangsters and thugs away. According to the owner, if Bing were not in the bar, it would definitely be a mess: all of the hostesses would flee in fear, and everything would be plundered away by gangsters. She entrusted me to Bing and the bar managers to take care of my safety.

Gangsters and other bar owners often came to visit. When they saw pretty hostesses, they dragged them upstairs and raped them. When they saw less pretty hostesses, they slapped them and beat them up. Hostesses were extremely scared of some of the toughest gangsters and thugs. They would run as fast as they could to escape them. Once I fled along with the other hostesses. We climbed up the back wall to the railroad behind the bar, losing our shoes and cutting our feet in the process. It was a very unpleasant experience. Most of the bar hostesses have been raped one or more times by the gangsters. Twice the gangsters came in and started to pull me into a karaoke room. Luckily they were stopped by Bing and the bar managers, who said, "She is not a hostess here. She is my friend." That saved me from imminent danger, but the lingering fear remained.

Bing always beat up the hostesses who did not obey him. At times, when other bar owners came to borrow hostesses, he ordered some hostesses to go. Once when a hostess did not want to go, he shut her in a room and beat her up. Other times, when some hostesses slipped away to other bars to sit onstage, they always received a good beating when they returned. Bing yelled at the hostesses, "If anyone does not listen to me, I will beat you to death!"

Hostesses in "Romantic Dream" were often discontent with the slack business and their low status. They were encouraged to go elsewhere by clients and hostess friends. However, the bar bouncer Bing warned everybody that he knew everyone in the karaoke bar circle throughout the city. He said, if anyone dared to leave for other bars, he would have them beaten to death. Hostesses who did manage to switch bars had ties with one or more gangsters as their protectors. Otherwise, no one dared to leave. Almost all of the hostesses, to protect themselves, were connected with one or two gangsters on the street to gain protection. They frequently joked, "We hostesses are relatives of the underworld."

Hostesses' Resistance: Informal Networks

Facing such violent and exploiting environment, do hostesses rise up and form collective protests? Research on sex workers in areas such as Madras in India and Brazil explains why in these areas, formal and informal organizations are not possible (Roche, Neaigus & Miller, 2005; Asthana & Oostvogels, 1996; De Meis, 2000). In Madras, for example, Asthana et al. attribute the lack of sex workers' formal or informal organization to their highly diverse backgrounds, high dependence upon the brothel establishments, and high mobility. In their research on sex workers in New York City, Roche et al. observe that the presumed temporary nature of sex work, isolation, and denial of their identity as sex workers preclude developing an informal support network. Likewise, in Brazil, De Meis points out that the sex workers' inward disdain and denial of their profession forms the crucial impediment to their potential organization.

Chinese sociologist Pan Suiming, in his work on hostesses in hair salons, argues that in China, informal networks are nonexistent. Pan Suiming relates that hostesses characterize each other as "sly" because they lie to each other and do not help each other (Pan, 2000, p. 186). Pan observes that it is not only because of the fierce competition, but also because hostesses are always mobile - they "meet by chance" (pingshuixiangfeng) and they "constantly come and go" (jusanwuchang). This nomadic lifestyle robs them of any sense of responsibility or duty towards each other. Hence they "do not have the leisure to help others" (Pan, 2000, p. 186).

Although hostesses in my study show similar reactions to their circles as they are harsh critics of themselves, they do form smallscale unions against their oppressors, although this kind of union is often unstable and fleeting. I argue that it is the centrality of native-place, blood relationship, and mutual benefits that differentiate them from the sex workers in the other areas. At the same time, different from the previous research on migrant workers that document the crucial role native-place networks play in organizing protests and strikes (Hershatter, 1986; Honig, 1986; Lee, 1988; Ngai, 2005; Rowe, 1984, Strand, 1989; Zhang, 2001) my research on hostesses showed that although informal alliances were bound by localist ties, these ties were always trumped by the possibility of gaining competitive advantage. Below I will tease out the complexity of hostesses' active formation of small allies that are significant support networks for their working lives.

Mobilizations?

In the karaoke bar "Romantic Dream," waiter Li served as the "madam." His girlfriend was May, one of the bar hostesses. There were times when Li led only May and her close friend into the karaoke room while leaving the other hostesses outside. This was unfair to the other hostesses, especially when many hostesses had not earned any money for several days because of the lack of customers. Business in "Romantic Dream" was the least prosperous compared to the other two bars. Once, waiter Li again led only May and her friend into the clients' karaoke room. Quite upset with it, the other hostesses were sitting together in the hallway, drinking hard liquor and complaining in low voices. Nobody dared to speak up. All of the staff members, as always, were eating together on the other side of the hallway. I wanted to comfort the hostesses, however, since many of them had not had any clients for several days, I realized how useless my comforting words would be.

Hostess Hui had served clients that night. She appeared very indignant about the injustice done to the other hostesses. She downed several glasses of beer. Given courage by the alcohol, she walked towards the staff circle and pleaded with waiter Li on behalf of the other hostesses. "Didn't you know that some hostesses had not had any clients for several days? Why did you just push your wife and her friend into the room without informing the rest of the hostesses, especially when your wife had already had clients tonight! Why didn't you lead the other hostesses into the room to be chosen?" As she spoke, her voice trembled with fear and tears rolled down her face. The bar managers and the barkeeper tried to calm her down, but waiter Li jumped up and began punching her in the face and head, kicking her severely in her chest and belly, yelling at her, "Didn't you already have clients? Why are you minding the other hostesses' business? During this tirade, none of the others complained. So why are you complaining about me having my wife sit in the karaoke room?"

In the face of such violence, Hui collapsed into a chair from the impact of the kicks and blows. She did not dodge at all, but screamed, "Ok, you beat me! You beat me! Beat me hard! Beat me hard! Beat me good! What I said was true!" The other hostesses tearfully threw themselves in front of Hui to protect her from the kicks and blows, crying, huddled around her, deflecting the blows and kicks onto their own bodies, while pleading to Li, "Stop! Don't beat her any more. Don't beat her anymore! Please don't beat her anymore!!" Seeing this scene moved me to tears. I wished I could have done more than just help Hui to her bed and take care of her afterwards.

In this situation, the madam's injustice was challenged by hostess Hui. While Hui was physically beaten, many other hostesses threw themselves to her, forming a shield from the onslaught of kicks and blows. However, in this case, Hui was the only one who stood out to confront the "madam" for his injustice. The others did not join her to protest against the mistreatment by the "madam." Rather, what the others did was more a collective plea than a collective struggle: they were crying and appealing for mercy from the "madam." Such a collective plea failed to change the operation of the karaoke bar industry and only subjected themselves to a more formidable future - as Bing threatened other hostesses after the event, "Had it not been Hui but any other one of you who spoke up [Hui was the "bar mainstay", the tai zhu zi), I would have broken your legs and driven you away."

Although Bing had raped almost every hostess in the bar, no hostess dared to utter a word about it. As hostesses Dee once said to me,

They beat us up any time they want! You think this (profession) is so easy to do? Who are we? Hostesses! The bar boss and managers belong to the upper class, and we belong to the low class. I know that. Bing can rape us any time he wants, and you cannot utter a word! Sometimes I fought with Bing, but the other hostesses just tolerated it. I think if we could unite together, we would be in a much better situation. However, some hostesses like Lynn fell in love with Bing and flattered him all the time! They get pushed to the clients by Bing and earn handsome tips every day.

Dee's criticism of the lack of large-scale collective action among the hostesses resonated with other hostesses in the bar. The failure to mobilize collectively was not because they had succumbed to the persuasive patriarchal powers. Rather, as I came to conclude, it was because large-scale collective mobilization was severely inhibited by these two key factors: one is the possibility of breaking out of the subordinated group (in game-theory terms, there are incentives to defect rather than cooperate); and the other was the potential cost of cooperation.

There was constant tension between the hostesses' loyalty towards their subordinate group and the allure of potential benefit that could be reaped by defecting from the group. For many hostesses, it was their aspiration to break out of the group and move up in the social hierarchy. As long as a portion of the people had the prospect of defecting from the group, they would not be able to mobilize effectively together. They would not jeopardize their chance of defection by completely submitting themselves to the group through mobilization. Thus, the interest-maximizing choice of individual group members ultimately resulted in the loss of the interest of the whole group. For instance, in the above scenario where many hostesses tried to mobilize together against exploitation, a couple of hostesses stood by the owner for fear of losing their status and interest in the bars - and thus their chance of upward mobility.

The possibility of mobilization was not only impeded by hostesses' incessant repositioning between groups (and outright denial of group affiliation), but was also curtailed by internal group competition. As I was told, their main competitors were other hostesses. They struggled to discard their rural roots and competed with other rural migrants to break out of the subordinated group and attain upward mobility. For instance, because it often happened that hostesses stole each other's clients, hostesses learned to guard against each other (e.g., they do not introduce their own clients to others). Meanwhile, they competed with each other over on-stage times, clients' status, and benefits earned from their clients.

Hostess Hong and I were very good friends. She invited me over to her rented apartment to stay many nights when we confided to each other about our personal lives. However, as close as we were, she never introduced her boyfriend to me. When her boyfriend rented an apartment for both of them, she asked me to help her move to the new apartment. She told me to come at 8am Sunday morning, adding casually that her boyfriend would join us as well. I went at the requested time, but never saw her boyfriend. Around 9am when the moving was over, she asked me to leave because her boyfriend was coming soon. Only then did I learn that she had arranged for him to come at 9:30am so that he and I would never meet. Hong's cautionary attitude was well justified in the hostesses' group where snatching each other's clients and boyfriends was common.

In another example, hostess Shi and Ying came from the same place - Anshan in Liaoning province. They came together to the bar as very good friends and they were always together. Like everyone, I thought they were best friends. One night Ying was crying in a karaoke room alone. I asked her why, she said,

Everyone thought Shi and I were best friends because we are locals (laoxiang), but in fact, I am better friends with Dee and the others than I am with Shi. Shi and I are just not compatible because we have different characters. Shi has always cheated and hurt me. Yesterday Shi's client (laogong) asked me to go to a sauna bar with him. I can assure you that I was just sitting there in the hall with him and did not do anything. In fact, I was trying to say good words about Shi because he was criticizing Shi. Later, Shi heard about our excursion and came over to the bar. She sat beside the man and tried to irritate me. She was talking with the man loudly and laughing all the time, totally ignoring me. I did not have any interest in that man, and now she is doing all this to irritate me. In the end I said I am leaving.

Ying was crying the whole time when she was telling the story. In another room everyone was gossiping about Ying and Shi, "If Ying did not do anything with that man, how could Shi be angry with her? That man is Shi's client (laogong) and loved Shi so much, how come she [Ying] interjected and went with him to the sauna bar? She is crying there so much, but nobody has any sympathy for her! Ying has all the schemes (xinyan) in her heart." Others continued,

Yes, you can make good friends with Shi because she is very easy-going, but not Ying at all. Remember when she first came here, she looked like an old grandma (lao da ma), extremely ugly. Now she looks much better than before. When they first came here, she always spent Shi's money. Remember every time when they had meals, it was always Shi who paid the bill. But when Shi borrowed money from Ying, Ying immediately demanded a return.

Apparently the hostesses were used to the scheme of friends stealing each other's men and felt indignant about the fact that Ying went out with Shi's client, regardless of her original intention. In spite of the emphasis placed upon localist networks in the previous research on migrant workers, my research of hostesses showed that although Shi and Ying appeared to be best friends bound by localist ties, these ties were always trumped by the possibility of gaining competitive advantage.

In addition, substantial costs associated with cooperation inhibited collective action. If hostesses mobilized together to defend their rights, their stigma as a hostess would be publicized. It would severely endanger rather than promote their upward mobility. Also, they hope that hostessing would be a springboard for them to move on to more respectable entrepreneurial professions. Therefore, any kind of mobilization or unionization amongst hostesses could serve only as a barrier to their ultimate objective. Finally, because their work sites were subject to a series of police raids regularly each year, mobilization was highly circumscribed by political repression and hostesses' lack of political resources. In view of these significant costs, it was not surprising that hostesses evinced little motivation for mobilization.

Complexity of Informal Alliances

However, hostesses did ally themselves in small groups based on blood relations, rural identity, and exchange of benefits. The fact that their small cliques were based on native-place and blood relations is not a new story. Research in China highlights the centrality of native-place for migrant workers' political alliances and networks. Numerous significant ethnographic and historical studies of migrant workers in Tianjin, Shanghai, Subei, Hankou, Guangzhou, Shenzhen, and Beijing document the crucial role native-place networks play in providing each other financial and emotional support and in organizing protests and strikes. Calling this alliance a 'localist network,' Ching Kwan Lee (1998) illustrates its significance in the lives of the women workers in Shenzhen.

This network assists women in securing jobs, teaching them work skills, and offering them emotional and financial help. Similarly, Pun Ngai's research on women workers in Shenzhen emphasizes kin-ethnic enclaves developed in the workplace in a 'honeycomb' pattern, that is, with the core kin in the middle and other family members and relatives on the periphery (Ngai, 2005, p. 56). She elucidates how kinship and ethnicity intersect in cementing group formation and in providing the best type of assistance in the hostile city. In fact, migrant women's informal ties are so strong that it projects the possibility of labor resistance. As Ngai states, it "silently render[ed] support to the social revolution from below by migrant workers in the city" (Ngai, 2005, p. 60). To control production, managers find it necessary to break up kin-ethnic groups by inserting outsiders into the work group (Ngai, 2005, p. 123). As I illustrated above, in my own research on hostesses, although these localistic networks did provide emotional and financial support for the members, they were subject to dissolution because competition between hostesses for scarce resources provided strong incentives to betray group members.

Many hostesses either arrived in the city accompanied by relatives or were later reunited with relatives when they came to the city. It was not uncommon, then, to find sisters and cousins working as hostesses in the same bar. Hostess Huang and her cousin Ling bumped into each other on the street and Huang learned that Ling was working as a hostess in a nearby sauna bar. Although the two of them had years of animosity and misgivings against each other in the village, in the city of Dalian, they strategically formed a temporary alliance based on exchange of benefits and blood relationships. When Huang came back from Shanghai, she did not have a place to live. She contacted Ling for help. Ling's client lover Liu always brought her to a nearby hotel where Ling made friends with May, the manager of the hotel and the second wife of Liu's colleague. Soon Ling and May became good friends and shared an apartment together. When Huang asked Ling for help, Ling introduced May to Huang and arranged for Huang to stay with them in their apartment. Huang stayed with them for two weeks while searching for another apartment.

In each of these three karaoke bars, there were biological sisters who came to Dalian and worked as hostesses in the same bar. They provided each other with an immense amount of emotional and financial support. For instance, in the low-tier karaoke bar, hostess Dee's 17-year-old little sister Bai followed Dee to the karaoke bar. Since Dee had already worked there for a year, she had a great deal of experience to offer Bai, who was a novice in this field. When Bai first arrived, Dee bought her all the necessities and took very good care of her. Dee always warned Bai not to over-consume alcohol, not to go out with unscrupulous clients, and not to get bullied by others. During the daytime, Dee took Bai out for clothes-shopping and hair-styling and offered advice on the techniques to extract more tips from clients. Later on when Bai fell in love with a client and was distraught and miserable, Dee was always there providing emotional support and teaching Bai to be strong and resilient.

In Dalian, while consanguinity and native-place served as a ready basis for instrumental and temporary coalition, coalitions based on rural identity and exchange of benefits were also significant. Members of small alliances not only offered advice and suggestions in dealing with clients and boyfriends, but also provided emotional, professional, physical and economical support to each other.

Hostesses' group-identification was highly mercurial and contextsensitive, i.e., it changed with changes in circumstances. At times of collision or conflict with urbanites, hostesses with rural backgrounds temporarily united to confront the threat. In one instance, an urban hostess 'stole' a rural hostess' client by displaying her urban ID and denigrating the other hostess' rural background in front of the clients. The rural hostesses responded by collectively criticizing, cursing and threatening violent retaliation against the group of three urban hostesses in the bar. It should be noted, however, that this alliance was transient and did not precipitate cooperation in other areas of endeavor.

At other times, rural hostesses rejected with their rural relatives when they came to visit them. For instance, hostess Lin's cousin sought her help to work as a hostess in the bar. Lin brought her over to the bar around 6pm to chat with everyone in the bar and explore job opportunities. The next day Lin said that her cousin was too tuqi (hick-ish) to work as a hostess there. Everyone else agreed that there was simply no way to cover up her rural aura. Lin said, "Whatever she puts on, she still has the tuqi taste. She is just not suitable for this job." Lin's use of words and tone of voice sounded exactly like an urban woman denigrating a rural woman. In the end, through her network with a client, Lin found her cousin a job at a restaurant. On other occasions, hostesses ruthlessly commented on rural people and their village folks as stupid (ha le) with "peasant consciousness" (nongmin yishi) or "small peasant psyches" (xiao nongmin xinli). Under these circumstances, hostesses emphasized their own cosmopolitan attributes against others' rural backwardness.

Alliances based on exchange of benefits were equally strong. A hostess in one group not only supported the interests of her group members, but also brought them to join her in her new job, perhaps in a more upscale bar. For instance, hostess Ling and Hong were good friends. A client lover of Hong introduced Hong to a madam in an upscale bar in the city and enabled Hong to work there. Hong, in turn, introduced her friend Ling to the madam and managed to have Ling work there as well. Both Hong and Ling benefited from forming such an alliance. Hong was gratified because she could rely on a network for emotional support in a totally strange and daunting environment; and Ling felt fortunate because she could enter an upscale karaoke bar and earn more money.

These examples also show that hostesses are not always completely instrumental with each other in their alliances. For instance, a hostess in an emotional crisis (e.g., abandoned or cheated on by her boyfriend) was often accompanied and comforted by members in her group. Hostesses in one group sometimes mobilized to rebel against wicked clients who had maltreated their members. A nice example of this type of coalition is when hostesses are abused by their patrons. Hua's client lover mistreated her, so Dee and several hostess friends admonished him together and demanded that he apologize to Hua. In the end, pressured by all these women, he did. In another story, hostess Min was broken-hearted to catch her client lover sleeping with another woman in a sauna bar. Min's hostess friends took her out to a restaurant to drink the whole night, and then went to a disco so that she could forget him. Everyone was telling her that he was not worth it, and that she should pay all her attention and energy to earning money through going out with clients. Min followed their advice and went to hotels with clients.

However, the biggest complaint that hostesses had of one another was the lack of loyalty. Even in the absence of loyalty, hostesses opted to perpetuate the network for their mutual benefits. Hostess Di related to me how her local friend of many years-a hostess in another bar from her hometown - had abandoned her when she fell ill in the aftermath of an abortion. Di spoke painfully about how she was forced to ask a neighbor to help change the transfusion bottle that dripped medicine into the vein of her arm. Although Di swore to break out of the relationship, she eventually chose to resume it because she needed her friend's help to gain access to the madam who ran a Japanese bar. In this case, there was an instrumental reason for resuming the network.

The following example epitomizes the fleeting nature of all such coalitions. Hostess Huang and Dong were best friends even though they came from different areas. They called each other sisters. However, the informal ties between Huang and Dong did not exist without rifts. To save rent, Dong moved into Huang's apartment, which was rented by Huang's client lover Yu, a general manager of a shipment company. Dong promised that she would sleep at a sauna bar if Yu came back to stay with Huang. One day Yu called and asked Huang if she was alone at the apartment. Huang responded that three hostesses - Dong and her two friends were there. Yu did not come back.

After several weeks Huang called Yu urging him to come back for a night, and paid about 200 yuan (U.S. $24) to settle Dong and her two friends at a sauna bar for a night. Although Huang treated Dong like her own sister, Dong did not reciprocate. In fact, Dong was quite jealous of Huang for her numerous lovers. Whenever seeing a man with Huang, Dong would ask the man, which one of us looks younger? Because Dong was plump and looked older, men always politely commented that she looked a little older than Huang. Dong would immediately cry out, "Look at her! She is already a mother of an eight-year-old boy! How can I be older than her?!"

Time and time again Dong made sure that every man Huang was with knew the fact that Huang was a mother. Dong always asked Huang, "Home come you always have good times, but I don't? Why don't you introduce me to some managers under Yu's leadership and see if we could become lovers as well?" To help her out, Huang made Yu bring over three managers to meet with Dong and her three girlfriends. One of Dong's girlfriends hooked up with one of the managers, and another girlfriend stole Yu away from Huang. Huang was completely in the cold. Since Yu did not come back any more, Huang decided to move out.

On the day she was moving out, the woman living downstairs said to her, "You are too foolish! Do you know who has stolen your lover?" Huang shook her head with puzzlement, "Who?" The woman said, "The one you have helped the most." Huang was so angry that she confronted Dong immediately, "Do you know who has stolen (qiaozou) my husband (laogong) away?" Dong pretended that she did not know. Huang said, "You don't know? Stop lying to me! My close friend is doing this to me. How can you treat me this way? Have you ever had a sister like me who offered you a place to live, money to stay in the sauna bar, and introduced you to men? I treat you better than your own sister, but you have totally betrayed me!" Dong had nothing to say but murmured, 'I have warned you of losing him before, but you didn't listen." Huang said to her, "From now on, we are no longer friends. You have hurt me too much." Later when Dong called up asking her to go to Shanghai together, Huang replied, "In Shanghai, I can be a good friend of yours, but in Dalian, we are not friends." Because they were going to work in Shanghai where neither had allies, they needed each other for support, causing their "friendship" to be at least temporarily reestablished.

As the example illustrates, regardless of the basis of clique formation, all coalitions were unstable. Competition between hostesses for scarce resources provided strong incentives to betray group members. For instance, hostess Fang found out that hostess May reaped a tip from a client by promising to introduce him to hostess Fang. Or, hostess Cheng sheltered and took care of hostess Hong for a month and a half during Hong's abortion period. Hostess Cheng was extremely unsatisfied with the low compensation fee of 200 yuan (U.S. $24) that hostess Hong offered her. There were other cases of hostesses' snatching away each other's boyfriends and not looking after the sick ones and so on. When such betrayal outweighed cooperative behavior, the small group dissolved. There was a constant tension in their networks between betrayal and cooperation. Whether the group could maintain its cohesiveness depended on the overall proportion of members' cooperation vis-à-vis betrayal.

Conclusion

In this paper, I have explored the violent working environment of the rural migrant karaoke bar hostesses due to the political policy and the sex industry, and the complex nature of hostesses' resistance. Within such a hostile institutional and cultural environment, my ethnographic research revealed the complexity of hostesses' collective actions. On the one hand, large-scale mobilization is inhibited by their unstable coalitions (or complete disaffiliation), the incentive to select non-cooperative strategies, and the costs and risks associated with collective action. These factors make it quite difficult, if not totally impossible, to overcome the barriers of personal interest and danger that impede large-scale collective mobilization. On the other hand, in contrast with researches on sex workers in other areas such as India and Brazil, hostesses in my study do form unstable small-scale group cliques and rely on these networks for help and security. These informal networks are facilitated and enabled by blood relationships, common rural background, native-place, and mutual benefits.

As illustrated in the paper, these informal networks are transient and temporary due to the internal competition and the costs and risks of protests. The lack of large-scale collective action is mainly because hostesses aspire to defect from the hostess group as a criminal group. Research in other countries has revealed to us that mobilization and collective actions against hostile policies and for basic human rights are much easier when the state allows unionization and/or decriminalizes sex work. For instance, in Taiwan, after implementing an abolitionist policy in 1938, the provincial council legalized prostitution in 1949 by distributing license to prostitutes and establishing the legal prostitutes' house. In 1986, when the city government suddenly changed the policy to abolish prostitution, hundreds of prostitutes in Taipei organized a "self-rescue unit" demanding their rights to work and combating social stigma (Xia, 2000). Their protests lasted a year and seven months and finally triumphed. In Bangladesh where prostitution is legal (Qurratul-ain-Tabmina, 2004), prostitute organizations resist social rejection, lobby for legislative changes, and gain social recognition of their profession. They observed that they could not expect the government or the society to resolve the problems unless they unite a strong force and demand a change.

It is not difficult to note the overlapping characteristics shared by these two successful cases of Taiwan and Bangladesh. That is, both prostitution and unionization are legal in both countries. In Taiwan, upon the change of policy, prostitutes realized that the advantages bestowed to them by the favorable policy would no longer exist. As they contended, legal prostitute houses offered them a fair share of payment, legal protection, security, and freedom. Under the auspice of the law, they were no longer harassed by violent clients nor were they subjected to random arrests, fines, or obligatory services to clients. They enjoyed the safe, healthy, and free environment and were committed to winning it back (Xia, 2000).

In both countries, their unionization is supported by women's organizations and NGOs. In Taiwan, local women's organizations and trade unions educated the prostitutes that their job is justified and that they should unionize. In Bangladesh, a central factor in the success of the prostitutes' organization is the support of the human-rights and women's rights organizations, the NGOs of HIV/AIDS prevention and healthcare programs in the brothels. They played a pivotal role in forming the self-help groups and opening a channel through which prostitutes could ventilate their grievances.

Because countries adopting the abolitionist policy tend to lack these two prerequisites of legalization of sex work and organizational support for sex workers, unionization is rather difficult. In Vietnam, for example, although the government, bureaucracies, and the police profit immensely from prostitution, prostitutes suffer from illegal identities and random arrests and harassments (Walters, 2004). Similarly, in Cambodia, the government intermittently closes down brothels and cracks down on prostitution, subjecting prostitutes to a violent environment (Derks, 2004). In China, as I have demonstrated in the paper, although operating in an antagonistic and hostile system, hostesses still manage to mobilize small-scale collective actions. Large-scale collective mobilization and unionization will become much easier when China decriminalizes sex work and/or has a proliferation of strong legal organizations that support prostitutes' unionization.

Footnote

2 I am grateful for the anonymous reviewer's constructive, helpful, and detailed comments that have helped improve the quality of the paper. This paper previously appeared in China: An International Journal (March) 6 (1): 69-95. © [copyright World Scientific Publishing Company]

3 This was the official figure of the city population (in the central four districts) in 2006.

4 Municipal officials interviewed in 2001 estimated Dalian's floating population at 1 million people.

5 One U.S. dollar could be exchanged for 8.4 yuan at the time of my research.

6 A typical wage of female migrants in the city is around 800 yuan ($100) per month.

7 I conducted my fieldwork in karaoke bars in Dalian during four segments of 1999, 2000, 2001 and 2002. I am grateful for the supportive funding provided by the Yale Center for International and Area Studies, and the Council on East Asian Studies of Yale University, without which this research would be impossible.

8 The materials on the anti-prostitution campaign, state policy, and the impact on hostesses are obtained from my ethnographic research at the karaoke bars and my interviews with local officials in Dalian.

9 Pottery bar, a bar where drinks are served together with all kinds of games such as cards, Chinese chess, and military chess. In an adjacent room, a tutor helps you make all kinds of pottery.

10 These people took advantage of the state's liberalization of economy in the 1980s and procured fast money in informal economic sectors, when many people still favored life-long security despite low wages in state employment sectors. Underlying their violence against sex workers is the antagonistic legal and political system against sex workers. As illustrated in the paper, sex workers are not only the target of police raids and police arrests, but also the target of unscrupulous men who believe that they can do anything to these women without any consequences. The fact that sex work is illegal not only invites violence, but also drives sex workers to search for ways to exit the profession instead of committing to collective action within the profession. I have been working on this population of sex workers for over ten years, and all the sex workers I worked with in the low-tier bar at the beginning of my research left the profession after six to eight years. Some of them married their clients, had kids, and stayed in the city, some of them returned to the countryside and married there, some of them remained single and lived as mistresses in their own apartments in the city. Their agency is not manifested by their collective actions against violence inflictors or the hostile system, which could further their social and institutional stigmatization. Rather, as my other works have demonstrated, their agency is exhibited by their manipulation of their relationships with clients in accessing clients' social and economic resources as their bids for social upward mobility.

References

REFERENCES

Anonymous. (2002). Sanpei xiaojie de falu baohu wenti [Legal protection of hostesses]. Shenzhen Fazhi Bao [Shenzhen Law Newspaper], 6/25.

Asthana, S. & R. Oostvogels. (1996). Community participation in HIV prevention: Problems and prospects for community-based strategies among female sex workers in Madras. Social Science & Medicine, 43 (2): 133-148.

De Meis, C. (2002). House and street: Narratives of identity in a liminal space among prostitutes in Brazil. Ethos, 30 (1/2): 3-24.

Derks, A. (2004). The broken women of Cambodia. In E. Micollier (Ed.), .), Sexual cultures in East Asia: The social construction of sexuality and sexual risk in a time of AIDS (pp. 127-155). London: RoutledgeCurzon.

Hershatter, G. (1986). The workers of Tianjin, 1900-1949. Stanford: Stanford University Press.

Honig, E. (1986). Sisters and strangers: Women in the Shanghai cotton mills, 1919-1949. Stanford: Stanford University Press.

Jeffreys, E. (2004). China, sex and prostitution. London: RoutledgeCurzon.

Jian, P. (2001). Caifang shouji: Jingyan Dalian [Interview memoirs in Dalian]. Xinzhoukan [New Weekly], 10: 44.

Lee, C. K. (1998). Gender and the South China miracle. Berkeley: University of California Press.

Ngai, P. (2005). Made in China: Women factory workers in a global workplace. Durham: Duke University Press.

Pan, S. (2000). Sheng cun yu ti yan [Subsistence experience]. Beijing: Zhongguo Shehui Kexue Chubanshe [Chinese Social Science Publishing House].(1999). Cun zai yu hunag niu: Zhong guo di xia xing chan ye [Existence and irony: A scrutiny of the Chinese underground sex industry. Beijing: Qunyan Chubanshe [Qunyan Publishing House].

Qurratul-ain-Tabmina, S. (2004). Sex workers in Bangladesh: Livelihood at what price? Dhaka, Bangladesh: Society for Environment and Human Development.

Roche, B., A. Neaigus & M. Miller. (2005). Street smarts and urban myths: Women, sex work, and the role of storytelling in risk reduction and rationalization. Medical Anthropology Quarterly, 19 (2): 149-170.

Rowe, W. (1984). Hankow: Commerce and society in a Chinese city, 1796- 1889. Stanford: Stanford University Press.

Strand, D. (1989). Rickshaw Beijing: City people and politics in the 1920s. Berkeley: University of California Press.

Sun, S. (2003). Dalian e mo ba xiaoje fenshi shiyi kuai [A man in Dalian divided a hostess' dead body into eleven pieces]. Dongbei xinwen wang [Northeastern News Net], 12/12.

Walters, I. (2004). Dutiful daughters and temporary wives: Economic dependency on commercial sex in Vietnam. In E. Micollier (Ed.), Sexual cultures in East Asia: The social construction of sexuality and sexual risk in a time of AIDS (pp. 76-97). London: RoutledgeCurzon.

Wang, G. (1999). Conspicuous consumption, business networks, and state power in a Chinese city. Dissertation, Department of Anthropology, Yale University.

Wang, J. (2003). Qiujiu duanxin fasong chenggong, jinfang jieujiubei guaimai funu [Successful "SOS" short message, the police rescued abducted women] Liaoshen Wanbao [Liaoshen Evening Newspaper], 5/16: 2.

Xia, L. Q. [2000]. Ririchun [Daily Spring]. Taibei: Taibeishi Ririchun Guaihuai Huzhu xiehui.

Zhang, L. (2001). Strangers in the city: Reconfigurations of space, power, and social networks in China's floating population. Stanford: Stanford University Press.