The rapid growth of foreign direct investment (FDI) in China has raised interest in China’s institutional and cultural characteristics and their impact on organizations (Ahlstrom, Chen, & Yeh, 2010; Tsui & Lau, 2002). In particular, human resource management (HRM) has attracted much attention. Employee turnover, an important topic in HRM, is particularly salient in China (Chen, 2002). Talent management is a vexing problem for firms as it is becoming more difficult to attract and retain talent in China (Farrell & Grant, 2005; Li, Ahlstrom, & Ashkanasy, 2010). For example, it has been reported that the average tenure for Chinese middle to senior level executives in joint ventures—people in significant demand—is only 8–12 months (Goodall & Burgers, 1998). Locating and retaining executive talent in China has only gotten more difficult in recent years; even state-owned enterprises have started to recruit overseas (Yu, 2010).

Turnover is very costly to an organization, particularly in tight labor markets. It has thus received substantial attention in the literature (Chen, 2002; Chen & Francesco, 2000; Chen, Hui, & Sego, 1998; Zhang & Agarwal, 2009). Drawing on March and Simon’s (1958) early work, previous research usually focused on why people leave, using perceived ease and desirability of leaving one’s job to predict turnover (e.g., Griffeth, Hom, & Gaertner, 2000; Mobley, 1977). More recently, Mitchell, Holtom, Lee, Sablynski, and Erez (2001) examined factors that encourage people to remain in their organizations and in doing so introduced a new construct entitled “job embeddedness.”

Regarding research on the newer construct of job embeddedness, most studies were conducted in North America (e.g., Mallol, Holtom, & Lee, 2007; Mitchell et al., 2001). Although there has been some work conducted in the Chinese context, researchers have seldom paid attention to regional differences within China (or other large and diverse countries, for that matter). Similar to the cultural differences between the East and West, regional variation within China can be considerable and may lead to significant variations in attitudes and behavior. As Schlevogt (2001) noted, organizational scientists have rarely considered the impact of different subcultures in China, and may be thus missing a potentially crucial driver of individual behavior and performance.

A few studies have touched on this issue by analyzing enterprises in different Chinese cities (e.g., Huo & Randall, 1991; Peng, 1997; Ralston, Yu, Wang, Terpstra, & Wei, 1996), though regional differences were not the focus. Using Beijing and Shanghai as contrasting cases, Schlevogt (2001) examined regionally and culturally distinctive factors that influence organizational performance in China’s North and South. However, these studies were mainly concerned with performance at the organizational level. There is still a lack of research that focuses on the specific influence of regions at the individual level. Responding to the call for more research on regional differences (Schlevogt, 2001), as well as multiple levels of analysis (Hitt, Beamish, Jackson, & Mathieu, 2007), this paper assesses the possible effects of regional differences and cultural diversity in China by examining dialect similarities and differences in work settings there.

Most of China’s Eastern and Southern provinces are areas where local dialects predominate (Yuan, 1980). Although nearly everyone in China can speak the national language, Mandarin, local dialects such as Cantonese or Shanghainese are still very popular and prevalent in their respective regions. Although dialect has long been recognized as a basis for relationships (and conflict) among Chinese people (Earl, 1837; Erbaugh, 1995; Purcell, 1947, 1951), its impact on other individual outcomes is less well studied (Luo, 2007). Carrying clear regional attributes, dialect plays a role similar to culture in forming individuals’ identification and influencing their perception of organizations. Therefore, it can be asked, does the dialect someone speaks impact his or her willingness to remain in an organization? More specifically, if an employee in China can regularly communicate in a familiar dialect in the work setting, then will that employee feel more affinity with his or her organization and be more likely to stay longer?Footnote 1

This question has considerable relevance at a theory level in that it suggests the importance of understanding regional differences in a large country such as China. It also has practical implications for reducing turnover in an organization as many Chinese organizations encourage employees to speak Mandarin, or they restrict the speaking on the job altogether, as with Foxconn, which is reported to have contributed to serious problems of employee alienation.Footnote 2 This understanding that regional differences such as dialect can impact firm operations in China may also have relevance for other similarly large countries with major internal differences such as India.

In addressing this question, this article assesses whether Chinese employees’ dialect speaking skill can influence their job embeddedness and turnover decision. Different from previous cross-cultural research on job embeddedness, which treated culture as a moderator (Mallol et al., 2007; Ramesh, 2007), this paper explores whether regional differences within China can lead to variations of employees’ perception, and how could dialect affect different dimensions of job embeddedness via social identification process and finally influence their intention to stay. By doing this, this paper not only examines the generalizability of a newly developed construct in China, but introduces an indigenous antecedent of job embeddedness (Ahlstrom et al., 2010). Meanwhile, by recognizing an area largely unstudied in China research, the article raises the issue of controlling regional and (local) cultural differences in conducting managerial studies in China while incorporating indigenous variables into management research (Bhagat, McDevitt, & McDevitt, 2010).

Literature review

Dialect in China

Modernization of a society is often accompanied by population movement from rural to urban areas, and from less-developed to more-developed regions, which leads to contact between people of diverse linguistic or dialect backgrounds. Dialect is something that has a long tradition in China and it has not been significantly weakened by the emergence of Mandarin as the national language (Erbaugh, 1995). The prevalence of dialect can be traced back to Chinese history where several major waves of migration took place due to geographical and political reasons as dialect groups tended to isolate themselves, both socially and geographically (Erbaugh, 1995; Ge, 1991; Purcell, 1947). China has long remained a place where people need journey only 100 km from their homes to encounter a different, and largely unintelligible Chinese dialect (Erbaugh, 1995; Pan, 1990). When migration out of China to Southeast Asia occurred, those studying this migration found that the Chinese people left home not so much as Chinese, but as members of rather different dialect groups that all happened to be from China. These included the Cantonese, Hainanese, Hakka, and Hokkien who grouped together (and did business together) in their new Southeast Asian domiciles (Pan, 1990; Purcell, 1951).

Explorer George Windsor Earl was among the first to document the importance of regional dialect in China. Travelling through Asia the early 1800s, Earl (1837) noticed that there was much friction among the various Chinese dialect groups in Southeast Asia, almost as if they came from different countries and cultures. Sinologist and historian Victor Purcell (1947), writing just after World War II, observed that different Chinese dialect groups in Southeast Asia had been brought into a proximity not experienced in China. This led to a great deal of friction, not cooperation, in their adopted homelands. People groups speaking different dialects regarded one another almost as foreigners and conflict was common (Erbaugh, 1995; Pan, 1990; Purcell, 1947, 1951). Some dialect groups even blamed others (such as the Hakka or Cantonese) for fanning the Taiping Revolution and subsequent upheavals in China (Erbaugh, 1992; Yu, 2002).

The differences among Chinese dialects are indeed considerable, and it is often said that they are closer to being different, albeit related, languages (Norman, 1988). Norman (1988) points out that the Chinese language group is like a set of closely related languages that has evolved over some 3,500 years, rather than a single language displaying simple regional variations. Although some dialects are similar, the speakers of the national language, Mandarin, have a lot of trouble understanding Cantonese, though those two dialects are close. Other dialects, such as the Fujian or Hakka, are considerably different and tough to casually understand for most people in China, much in the way Dutch shares a lot with English, but is difficult for English speakers to comprehend.

To better understand dialect and its effect in China, it is helpful to recognize that the Chinese language can be roughly divided into seven different dialect groups (Yuan, 1980). These groups’ boundaries correspond to the sharpest zones of linguistic variation. The Mandarin dialect, which is also China’s national language, roughly occupies the Northern and Western regions of China. If one were to draw a steep diagonal line from the Northeast around Beijing down toward Southwestern China, the space North and West of that line would represent this primarily Mandarin speaking area. The area to the South and East of that line holds most of the main Chinese dialects. These would include provinces in China’s center, such as Sichuan and Hunan, those to the South such as Guangdong and Hainan, and the Southeast, such as Fujian, Jiangxi, Zhejiang, and Jiangsu (the latter two provinces being just west of Shanghai). From North to South, the main Southeastern dialects are formally known as Wu (spoken in the Shanghai region), Gan (spoken in Jiangxi), Xiang (spoken in Hunan), Min (also spoken in Taiwan and Hainan islands), Hakka (or Kejia—spoken primarily in Southern China and Taiwan), and Yue (Cantonese) (see Figure 1).Footnote 3

Figure 1
figure 1

Map of Chinese dialects in Southern and Eastern China Adopted from: http://en.wikipedia.org/wiki/File:Map_of_sinitic_languages-en.svg (copyright free)

Full size image

Those terms for the dialects usually refer to a single dialect, though sometimes they can refer to a group of related dialects, such as the Wu dialect of the Shanghai-Zhejiang region. Many people still speak dialects in China as their first language. In spite of the national efforts to inculcate Mandarin, dialects are still the language of the street and the factory in many parts of China—particularly the busy Southeast and coastal regions—and some still function as trading languages around East Asia. Yuan (1980) found that over 30% of the Chinese population is innate dialect speakers. A comparison between the ethnolinguistic area and the population distribution in China reveals that many densely populated provinces are dialectical areas, in particular the coastal provinces such as Jiangsu, Zhejiang, Fujian, and Guangdong (China Economic Information Bureau, 2005; Miao & Li, 2006).

These are not trivial local languages spoken by a few villages. Many of these dialects have millions who speak the dialect as a first language that are comparable to European countries. For example, in Guangdong province alone (and neighboring administrative regions such as Hong Kong, Shenzhen, and Macau), almost 90 million people speak Cantonese. Many people in the adjacent province of Guangxi also speak Cantonese. And large numbers of the 50 million overseas Chinese around the Pacific Rim and elsewhere also speak Cantonese. For Fujian province, not only do about 36 million people there speak that dialect, but a close variation of it is spoken in Taiwan as Taiwanese, and large numbers speak Fujian dialect in Southeast Asia, where is it commonly known as Hokkien.

Since the various regional dialects of Chinese are largely mutually unintelligible, urban migrants are faced with great linguistic challenges in adapting to the life in the host community. When living in their hometown, they mostly use the native dialect although Mandarin may also be spoken in formal settings. Once they move to a different region, the local dialect of the host community can be an important factor constructing their sense of integration into the local life. Hence, bilingualism may result from linguistic adaptation on the part of either the local residents or the immigrants, or both (Beer, 1985; Grosjean, 1982).

In China, a major force for encouraging the continued use of local dialects is China’s economic growth. Economic development and the resultant population mobility in China helped to bring about a rising prestige of certain dialects and their diffusion through movies and popular culture (Chen, 1999; Erbaugh, 1995). Southern dialects, such as Cantonese and Hokkien, once spoken by people so poor that they fled overseas to make a living, now have unprecedented prestige because of economic integration and globalization. Most Hong Kong manufacturing has moved North of the border. Many Taiwanese have invested in Mainland China, investing billions in China conducting business, inextricably linking their economy to that of the Mainland. In 2006, total FDI in China from Hong Kong reached US$20 billion while FDI from Taiwan was US$2 billion, listing these economies as first and ninth among all the outside investors respectively. Of the FDI from Hong Kong and Taiwan, US$7 billion was invested to Guangdong province (China Statistics Press, 2007). But the Hong Kong Cantonese often speak only a little Mandarin. Even though the regional Overseas Chinese investors have started to speak more Mandarin recently, the topics are still limited to rather basic communications (Li, 1990).

The Guangdong and Fujian governments have also recognized the importance of dialect ties. In their outreach programs to potential investors, they tout the advantages of using one’s mother tongue to supervise workers (Erbaugh, 1995). Accordingly, while the status of Mandarin across China has been greatly enhanced (Chen, 1999; Chen, Fu, Miao, & Liao, 1999), the government has shown a benign attitude toward local dialects. For example, Guangdong, Shanghai, and Fujian have a long history of having TV programmes broadcast in dialects (in contrast with Singapore, which has long required the use of Mandarin for its Chinese-language broadcasts, even dubbing Hong Kong Cantonese programming into Mandarin). The first language of the majority of the citizens in Guangdong and Shanghai is still local dialect (Miao & Li, 2006). In Wenzhou, a major region within Zhejiang province, dialect is very important such that Mandarin-speaking outsiders can be marginalized. In those regions, informal conversations among employees are usually conducted in dialect if both parties are locals. Salespersons in the shops usually speak dialect first to customers and customers who speak the dialect usually receive better service. So employees who speak the right dialect may be advantageous in this situation and many Chinese take pride in speaking more than one dialect (Erbaugh, 1995).

To date, most research done in China focuses on testing the external validity of the constructs established in the West. In doing so, researchers in China generally use the same set of demographic variables, such as gender, age, marriage status, and education level. Although there are researchers looking into the effects of the demographic variables in China (Chen & Francesco, 2000), little effort has been made in finding indigenous variables such as dialect that may leave a unique impact in the Chinese context. On the relationship between demographic features and language use, studies in the past showed that an individual’s social network is an important factor influencing his/her language behavior. Gal (1978) investigated language shift in the German-Hungarian community of Oberwart, Austria, and observed that there was a close correlation between the peasant-status of one’s social network and one’s language choice. Milroy (1987) added that stylistic variations in the speech community of Belfast were to a large extent regulated by features of a person’s social network (e.g., in terms of multiplexity vs. uniplexity of network ties). Considering that being embedded means getting enmeshed in certain networks, an examination of the relationship between language use and job embeddedness in China’s urban communities will give us insights into the utility of dialect and the key retention construct in China.

Job embeddedness and its relation to turnover

The study of turnover has a rich theoretical and empirical history in which multiple models have been advanced to understand its complexity (Griffeth et al., 2000). Over the years, researchers have determined that, given alternatives, most people stay if they are satisfied with their job and committed to their organization, while they leave if they are not. However, a meta-analysis by Griffeth and colleagues (2000) indicated that organizational and attachment variables have relatively little explanatory power, explaining only four to five percent of the variance in turnover behavior. The consistent but weak links among attitude, perceived alternative, search, and turnover suggest that many other meaningful topics have been neglected (Mitchell et al., 2001). Drawing from the perspectives of embedded figures and field theory (Lewin, 1951), Mitchell and colleagues (2001) developed the concept of job embeddedness, which describes the factors that keep an individual from leaving the organization, in spite of experiencing situations that might lead to thoughts of leaving.

The job embeddedness construct is comprised of three elements. First, links refer to connections an employee and his or her family have in a social, psychological, and financial web that includes work and non-work friends, groups, the community, and the physical environment in which he or she lives. The higher the number of links between the person and the organization, the more an employee is bound to the job and the organization both functionally and emotionally (Mitchell et al., 2001). Second, fit is defined as an employee’s perceived compatibility or comfort with an organization and with his or her environment. The theory of job embeddedness (Mitchell et al., 2001) suggests that an employee’s personal values, career goals, and plans for the future must fit with the larger corporate culture and the demands of his or her immediate job. A person will also consider how well he or she fits the community and surrounding environment: The better the fit, the higher the likelihood that an employee will feel professionally and personally tied to the organization. Studies have shown that person-organization fit (O’Reilly, Chatman, & Caldwell, 1991) and person-job fit (Chan, 1996) predicted actual turnover. Finally, sacrifice captures the perceived cost of material or psychological benefits that may be forfeited by leaving one’s job (Mitchell et al., 2001). The more an employee will have to give up when leaving, the more difficult it will be to terminate employment with the organization. Studies also found that the attractiveness of the pay and benefits packages, as well as job stability and advancement in the organizations, are negatively related to turnover intention (Shaw, Delery, Jenkins, & Gupta, 1998).

These three elements are important in the organizational and community context. Job embeddedness can be work related, (e.g., positive relationships with supervisor and co-workers, good health benefits, etc.) or non-work related (e.g., spouse works in the same area, parents live in the same community). Thus, as shown in Table 1, job embeddedness can be conceived as a 3 × 2 matrix in which links, fit, and sacrifice are associated with the organizations in which employees work and the communities in which they live.

Table 1 Dimensions of job embeddedness.
Full size table

Reflecting the situation of people’s being integrated in a social web, job embeddedness is a retention construct that captures a sizable portion of the decision to leave the organization (Lee, Mitchell, Sablynski, Burton, & Holtom, 2004). Its utility in predicting turnover was supported in various studies (Allen, 2006; Crossley, Bennett, Jex, & Burnfield, 2007). Mitchell and colleagues (2001) provided initial empirical support for the negative relationship between job embeddedness and turnover. They found that aggregated job embeddedness correlated with intention to leave and predicted subsequent voluntary turnover after the effects of gender, satisfaction, commitment, job search, and perceived alternatives had been controlled. Extending Mitchell and colleagues’ (2001) study, Lee and colleagues (2004) looked into the impact of different dimensions of job embeddedness on withdrawal behaviors, and confirmed the effect of job embeddedness on reducing turnover.

In a more recent study, Holtom and Indderieden (2006) integrated the unfolding model of turnover (why people leave) and job embeddedness (why people stay) to obtain a more comprehensive picture of organizational attachment. In a large national study of stayers and leavers across hundreds of employers, Holtom and Indderieden (2006) found that stayers were found to have the highest levels of job embeddedness. Later researchers tested the effects of job embeddedness on turnover using different samples and in different settings (Crossley et al., 2007; Mallol et al., 2007; Ramesh, 2007). Hom and colleagues (2009) verified the utility of job embeddedness as a mediator for the relationship between employment relationships and turnover. Wheeler and Halbesleben (2009) tested the effect of social support on job embeddedness in China. These studies primarily tested the utility of job embeddedness in China, but none of them considered regional differences in China. Different from previous research which assumed China as a unitary culture, this article offers an example of how can regional differences can influence individual outcomes.

Theoretical framework and propositions

Dialect and turnover

Research on relational demography has provided convincing evidence that similarity or difference in demographic profiles among members of a dyad or team can affect turnover (O’Reilly, Caldwell, & Barnett, 1989). Riordan and Shore (1997) found that similarity in race-ethnicity affected individuals’ attitudes toward their work group, and influenced their turnover intention. One important underlying mechanism in this relationship is the process of identification (O’Reilly et al., 1991), in that attachment to an individual, object, group, or organization results from identification with the attitudes, values, or goals of the model. According to social identity theory, people derive their identity largely from the social categories to which they belong (Tajfel, 1982). Demographic attributes are usually the bases for self-identity and an initial impression of an individual. Based on demographic cues, people are categorized into various social groups. Those who are similar are classified into the same social group as self (in-group), and those who are different are classified to a different group (out-group). Once a person’s group identity is determined, attraction between in-group members may lead to increased interaction, communication, and friendship formation (Tsui & Farh, 1997), resulting in a higher level of group social integration, which is negatively associated with turnover (O’Reilly et al., 1989). Research found that that the social identity model also applies to people in China (Li & Hsu, 1995; Tong, Hong, Lee, & Chiu, 1999).

Regionality and local roots are by tradition highly respected by most Chinese because regionality is associated with clan ties (Freeman & Habermann, 1996). As Tan (2006) stressed in connection with minorities in Singapore, regional origin is a salient category affecting many forms of social behavior and thinking. The activities of regional “friendship associations” in many Chinese communities (such as the old and large Zhongshan fraternal organization founded more than 150 years ago among the Hawaiian sugar cane plantations) showed that this clan factor has a greater priority than a similar one in most modern societies of the West.Footnote 4 Even within a Chinese context, people tend to prefer those with similar origins, who share their particular language or dialect, and have concerns about the same affairs in their home provinces. Bearing a salient regional trait, dialect is a common label for categorizing people into different regional groups in China. Thus we propose:

Proposition 1

In China, employees’ focal dialect speaking skill will be negatively related to turnover.

Job embeddedness captures the degree to which an employee feels attached to an organization or community (Mitchell et al., 2001). In the following section, we will examine how dialect influences job embeddedness in the Chinese context.

Dialect and fit

Dialect’s influences on job embeddedness can be further explained by the social categorization and social identification process (Tajfel, 1982; Turner, 1987). An organization, team, or work group can represent a social category with which individuals can identify themselves (Hogg & Terry, 2000). Demographics also are important bases of self-identity for individuals (Tsui & Farh, 1997). If a person uses a dialect at a work place where this dialect is widely spoken, the social identity derived from regionality may enhance the individual’s organizational identity, and lead to a sense of fit to the organization. Following the principles of attitude formation (Eagly & Chaiken, 1993), identification is an employee’s general attitude towards his or her organization. This broad attitude influences how certain specific aspects of the job are perceived and evaluated, resulting in lower (for the weakly identified) or higher (for those identified more strongly) job satisfaction. Individuals who are strongly identified with their organization will perceive their actual work situation more positively. Someone who is strongly identified perceives neutral or even negative conditions as less detrimental because he or she is more likely to see the necessity of these conditions in order to achieve the organization’s overall goals.

Similarly, being highly identified should lead one to think more positively of the organization in general. This, in turn, will lead an individual to actively and/or unconsciously ignore and suppress negative aspects. For example, Van Dick (2004) found that organizational identification is positively related to job satisfaction. Zenger and Lawrence (1989) reported that engineers who were different from others in a project team on age and tenure engaged in less frequent technical communication with other project team members. O’Reilly and colleagues (1989) found that, in addition to a unit-level effect of demography on turnover, individuals most distant from others on demographic attributes were the least socially integrated into the group and the most likely to exit. Therefore, speaking a dialect in an organizational setting and thus giving an indication of one’s natal origin with colleagues, can promote the development of in-group identity, and lead to stronger identification with work unit or organization (Tong et al., 1999), making employees more likely to develop a sense of attachment and commitment to an organization. Thus we propose:

Proposition 2

In China, employees’ focal dialect speaking skill will be positively related to their fit to organization and community.

Dialect and links

When a social category is a meaningful source for social identity, an individual will have a strong social identification with the group, and that group becomes a meaningful psychological group (Cialdini, 2001; Tsui & Farh, 1997; Turner, 1987). The ability to speak the focal language, whether innate or acquired, accelerates the integration process of the employee. It provides a depersonalized group-based attraction derived from demographic similarities. When an individual speaks a dialect not widely used at work place, such a Shanghai person working in an organization in Guangzhou, he or she may feel closer to other people who also speak Shanghainese, and automatically identify other Shanghai natives in the organization as in-group members. Thus, links can be established based on dialect, which signals a common geographical base (Freeman & Habermann, 1996). Xin, Farh, Cheng, and Tsui (1998) further found that speaking the same dialect was significantly related with a high quality supervisor-subordinate relationship. Researchers found employees in China are more frequently engaged in non-work related exchange within the workplace developed through social interaction, such as home visits (Hui & Lin, 1996). Leaders have both in-group and out-group relationships with their employees. In-group, high quality relationships are characterized by high levels of information communication, mutual support, informal influence, and trust whereas out-group, low quality relationships have formal supervision, little support, and mistrust (Chen & Tjosvold, 2007; Fairhurst, 1993).

Other research has found that supervisors who were similar to their subordinates in sex, race, education, and job tenure gave more favorable performance evaluations (O’Reilly et al., 1989). Researchers found that favor exchanges in organizations outside the private lives of family and friends often involve the use of organizational properties and positions (Chen & Chen, 2004). The social gathering or interaction engenders group bonding and promotes social harmony (Tsui & Farh, 1997). Therefore, Chinese employees may have more opportunities to communicate with supervisors in non-work related settings. Speaking the same dialect, in this situation, may grant subordinates greater advantage in building closer relationship with supervisor.

Dialect not only serves as one of the initial cues by which a person considers herself or himself as a member of the group, but also an indicator of whether a person is recognized as in-group or out-group by surrounding people when first joining a company or community. An individual may be favorably inclined toward people who are demographically similar, even though these people are otherwise “strangers” at an interpersonal level. Ample research conducted in the West found that in-group identification leads to interpersonal attraction, communication, and friendship formation (Cialdini, 2001; Tajfel & Turner, 1986; Tsui & O’Reilly, 1989), even without personal knowledge or friendship, as long as others share the same category membership. When an employee speaks the dialect widely used at work place, they may receive more favorable evaluation from others than those who speak Mandarin only, making the employees easier to establish links. It was observed in several studies that speakers of Cantonese and certain other dialects (e.g., Shanghainese) held a less favorable attitude toward Mandarin than speakers of other dialects (Bai, 1994; Kalmar, Zhong, & Xiao, 1987; Tong et al., 1999; Zhou, 2001). Bai (1994) noted that it was because dialects were regarded as an important symbol of local loyalty. In addition, Miao and Li (2006) found that Cantonese and non-Cantonese speakers viewed Cantonese differently in that the Cantonese, but not the non-Cantonese, attached covert prestige to speaking mediocre Mandarin or having a heavier Cantonese accent. Therefore, for employees who work in cities like Guangzhou and Shanghai where the local dialect prevails over Mandarin, speaking the local dialect may make it easier for them to be accepted in the community and build more connections at work place. Therefore, we propose:

Proposition 3

In China, employees’ focal dialect speaking skill will be positively related to their links in organization and community.

Dialect and sacrifice

As defined by Mitchell and colleagues (2001), sacrifice captures the perceived cost of material or psychological benefits. Results indicated that relation demography could influence individuals’ perceptions of advancement opportunities (Riordan & Shore, 1997). Turner (1960) developed a typology that identifies two norms of mobility affecting individual’s career advancement: contest mobility and sponsored mobility. Contest mobility refers to a system of open and fair contest in which each individual can compete for upward mobility. In contrast, under sponsored mobility, career progression is achieved based largely on some criteria of supposed merit as decided by the authority or its agents (Turner, 1960). It is a controlled system and people tend to believe that some individuals have gained sponsorship from the organization or its representatives in the elite selection process. China has long been a country with a rule of man rather than a rule of law. The rule of man has always been treated as beyond laws and institutions, which has remained true even after it began to adopt the Western legal systems. Relationship with authorities or managers will be more relevant to employees’ mobility in an organization in China (Zhang & Ma, 2009). In Mainland China, if a person is to function well in an organization, having personal bonds in the organization will grant that person great advantage because managers have limited time and energy, they can only develop close work relationships with a few subordinates upon whom they use their personal and positional resources to help perform (Farh, Tsui, Xin, & Cheng, 1998; Loi & Ngo, 2010). As discussed in the previous section, speaking dialect to supervisors can be an advantage to establish close relationship, especially when the dialect is not commonly used in the organization. The advantage may be firm specific, so leaving the current company may lead to a sense of direct sacrifice of a salient base for establishing meaningful relationships.

Moreover, Brewer and Miller (1984) captured the importance of demography-based social categories in their discussion of self-identity by emphasizing how people’s identity is both differentiated and based in part on membership in (and emotional ties to) social categories and relational cliques. The demographic factors such as regionality, especially dialect-based, that are central to the individual’s self-identity in China are likely to be important categories that the individual will use to categorize the self and others. As people from all over Greater China show an increasingly public pride in local culture (Miao & Li, 2006), leaving for an environment where local dialect is not applicable may in part lead to a psychic loss of self-identity (Miao & Li, 2006), making withdrawal personally costly to individuals. Admittedly, whether or not being able to use dialect at work place may not be a major consideration when it comes to making decisions on turnover, but it serves as a basis for establishing relationships and defines the nature and content of the relationships. Therefore:

Proposition 4

In China, employees’ focal dialect speaking skill will be positively related to their sacrifice for organization and community.

Although the idea of relational demography may be a universal concept that has relevance for understanding work behavior in difference cultural settings, the concept could be particularly important in the Chinese context where socioeconomic background and family origins of individuals are highly emphasized in interpersonal relationships (Tsui & Farh, 1997). As Fei (1992) noted, Confucian ethics cannot be divorced from the idea of discrete centers fanning out into a web-like network, which closely coincides with the core idea of job embeddedness. We believe that job embeddedness may help to connect demographic influences to culture in explaining additional variance in turnover. As disused above, dialect may influence employees’ job embeddedness through social identification and self-categorization process (Tajfel, 1982; Turner, 1987), which in turn affects turnover. We thus propose job embeddedness as a mediating construct between dialect and turnover:

Proposition 5

In China, job embeddedness mediates the relationship between employees’ focal dialect speaking skill and turnover.

Discussion

To date, very little research exists that examines the impact of a specific cultural heritage on individual-level outcomes of interest to organizations such as turnover. As an initial effort to fill this need, this paper uses dialect as a proxy for regional variation in China to assess the effects of such variation on individual behavior. Dialects, which originated from different parts of China and evolved with only minimal contact with each other (DeFrancis, 1984), bear the distinctive attributes of those regions. Using a dialect at a workplace where this dialect is widely spoken can trigger individuals’ self-categorization process, thereby leading to higher identification with their organizations. Meanwhile, whether innate or acquired, speaking a dialect may be associated with recognition from other colleagues resulting from the relational demography. This two-way identification may lead to higher sense of fit, more informal links among employees, and better relations with supervisors. This all leads us to conclude that employees who are able (and permitted) to speak their dialect at work to colleagues and supervisors will have higher job embeddedness and other positive outcomes.

By explicitly examining the impact of dialect on job embeddedness in China, this article contributes to both the research and practice of management. In addition to examining the generalizability of job embeddedness in China, this paper identifies an indigenous antecedent of job embeddedness, which is an additional important contribution of developing economies research (Ahlstrom et al., 2010). More importantly, the present research draws attention to the effects of regional differences on individual outcomes within Mainland China. This paper reminds future researchers to consider the effects of dialect and other regional and cultural variations when conducting research in China and other large, diverse countries. Accompanying urbanization and modernization, population mobility in China has greatly increased, necessitating more research on regional cultural adaptation, especially for big cities such as Guangzhou and Shanghai where local dialect has a profound influence on people’s daily life (Miao & Li, 2006). Controlling the impact of regional differences on individuals’ behavior and understanding the processes they go through in handling the differences may help explain additional variance in job embeddedness and turnover, and hopefully, also increase the validity of the other indigenous HRM research.

For managers, this article highlights the potential problems of cultural integration within organizations, particularly in cities where local dialects enjoy high prestige. In order to increase employees’ job embeddedness and reduce turnover rate, some organizations engage in established HR practices such as profit sharing and providing on-site child care. We argue that in addition to these practices, organizations should also carry out practices aiming at reducing regional differences within the firm.

This article has certain limitations which suggest avenues for future research. First, as of yet, there is no systematic typology for regional differences. This renders a more comprehensive review difficult. Because the current article only serves as an initial exploration of dialect’s influence on job embeddedness, it does not include the specific nature of the various dialects in China. For example, many cities have their own culture and dialect, but the degree of local loyalty and attitude to immigrants may vary greatly. As such, studying the impact of dialect on individuals’ attitude formation would call for a systematic mapping of the regional and cultural differences within China. Second, dialect is not the only representation for regional differences. There are other psychological clashes resulting from regional differences. Empirical tests of dialect and its effect on key attitudes such as job satisfaction and engagement and intention to leave are needed to clarify the importance of dialect and other regional factors. In addition, with the recent problems in China with employee morale, particularly given the restrictions on speaking during work on the assembly line, research is needed on the factors that can reduce turnover and improve other morale problems (Li et al., 2010). Given the problems in China and other fast growing countries with employee morale recently at firms such as Foxconn and elsewhere, the imperative to better understand the impact of workplace forces impacting morale and attitudes in high production, high stress environments is increasingly clear.

It is hoped that this article will spur other refined research into job embeddedness, as well as other constructs in China. As suggested by Budhwar and Debrah (2009), research needs to be conducted in Asia so as to highlight the context-specific nature of personnel functions and relevant HRM system(s) for each country (c.f. Fang, 2010). Except for systematically categorizing the regional differences in China, further research might empirically test the effects of dialect on job satisfaction, organizational commitment, and organizational citizenship behavior to further establish dialect as a control variable in studies in China. It would also be helpful for researchers to define and examine the many regional cultures within China, if they follow dialect lines, and how they differ from each other and impact behavior in organizations. It would especially be of significant value for researchers to incorporate job embeddedness into HR practices, while controlling for culture. Although job embeddedness offers researchers an alternative perspective to studying turnover, the practical implication of the construct is still not well established. Therefore, how to improve employee loyalty through HR practices is a topic that is of continuing appeal, particularly in the challenging and diverse labor market of China.