INTRODUCTION
Political networking, the activities of developing and exploiting network ties with political actors, is an important strategy adopted by emerging economy firms to navigate weak institutions (Schuler, Shi, Hoskisson, & Chen, Reference Schuler, Shi, Hoskisson and Chen2017; Sun, Mellhai, & Wright, Reference Sun, Mellahi and Wright2012a; Xin & Pearce, Reference Xin and Pearce1996). Government power is widely regarded as one of the most influential, complex, and least predictable environmental factors. It is thus critical for entrepreneurs to maintain a ‘disproportionally greater contact’ with government officials (Child, Reference Child1994: 154), so as to buffer against government extortions and gain access to financial and regulatory resources (Li & Zhang, Reference Li and Zhang2007; Park & Luo, Reference Park and Luo2001; Peng & Luo, Reference Peng and Luo2000).
Prior studies, however, have not been well-developed to address the following research question concerning this strategy. Given the generic nature of political networking in emerging markets, what drives the variation of corporate investment in networking with political actors? Although the difference between private and state-owned enterprises in political networking has attracted plenty of attention from the early literature (Park & Luo, Reference Park and Luo2001; Xin & Pearce, Reference Xin and Pearce1996), there is a dearth of research on the ways in which organizational and environmental factors affect the intensity of political networking undertaken by indigenous private ventures. With the deepening of institutional transitions, this question warrants more investigation because political networking can become more of a rational, strategic choice than a necessity for private ventures (Sun, Zhang, & Mellahi, Reference Sun, Zhang and Mellahi2012b; Zhang, Zhao, & Zhang, Reference Zhang, Zhao and Zhang2016b).
Our article provides an in-depth inquiry into the forces that drive the varying intensities of political networking from the lens of resource dependence theory (RDT henceforth, Pfeffer & Salancik, Reference Pfeffer and Salancik1978/2003). RDT predicts the intensity of a firm's political networking – a typical co-option strategy – to be in close association with the degree of its interdependence with the government (Hillman, Withers, & Collins, Reference Hillman, Withers and Collins2009). This relationship involves not only the dependence of business firms on the government, but the dependence of the latter on the firms, which is in turn influenced by specific interests/objectives of government agencies and politicians (McDonnel & Werner, Reference McDonnell and Werner2016; Sun, Wright, & Mellahi, Reference Sun, Zhang and Mellahi2010b; Wang, Du, & Marquis, Reference Wang, Du and Marquis2019).
Concretely, we identify important determinants of political networking that shape the dyadic business-government interactions. They include (1) size-based dependence indicated by firm size, (2) connection-based dependence shown by firms’ existing political connections, and (3) location-based dependence reflected by whether a firm is located in an environment with better market-supporting institutions – business development zones. First, firm size implies a firm's economic resources/capital and thus affects the focal firm's economic-based power relationship with the government. Prior work has suggested that firm size is ‘the most prevalent’ firm-level antecedent of corporate political activity (Hillman, Keim, & Schuler, Reference Hillman, Keim and Schuler2004: 839). However, the relationship between firm size and political networking remains inconclusive in the extant literature (e.g., Park & Luo, Reference Park and Luo2001; Peng & Luo, Reference Peng and Luo2000).
Second, existing political connections signal a firm's sociopolitical capital and thus affect the focal firm's power relationship with the government. Prior research, however, is largely silent on whether political capital embodied through existing connections will substitute for or further stimulate political networking activities. In this article, we draw upon the recent theoretical advance in political tie heterogeneity (Sun, Reference Sun, Grosse and Meyer2019; Sun, Mellahi, Wright, & Xu, Reference Sun, Mellahi, Wright and Xu2015; Zhang, Marquis, & Qiao, Reference Zhang, Marquis and Qiao2016a) to address this omission. Specifically, we distinguish between embedded political ties and achieved political connections on the basis of our empirical context. The former concern network ties embedded in dense sociopolitical relationships held by entrepreneurs and their employees, who can be former government officials or the relatives of government officials (Haveman, Jian, Shi, & Wang, Reference Haveman, Jia, Shi and Wang2017; Zhang et al., Reference Zhang, Marquis and Qiao2016a). The latter refers to the appointment of entrepreneurs or senior executives to state organs such as legislative bodies (Wang et al., Reference Wang, Du and Marquis2019; Zhang et al., Reference Zhang, Marquis and Qiao2016a). We argue that these two types of political connections have differential impacts on business-government interdependence, thus leading to different political networking intensities.
Finally, firm location represents the most proximate institutional environment in which private ventures interact with government agencies and officials (Armanios, Eesley, Li, & Eisenhardt, Reference Armanios, Eesley, Li and Eisenhardt2017). When local governments make strong efforts to develop a friendly environment to attract and nurture business activities (World Bank, 2011), the power dependence relationship between private businesses and the government can be rather different from that in other locations. As such, the business-friendliness of the local government influences the necessities for corporate political networking. In our article, we examine whether the location in a business development zone makes a difference for a firm's political networking behavior.
We develop and test hypotheses regarding how size-, connection-, and location-based dependencies drive political networking through a survey of Chinese private enterprises. Despite being the most dynamic element of the Chinese economy, the domestic private sector has a disadvantaged socioeconomic status in comparison with its state-owned peers (Tsai, Reference Tsai2007). They are more sensitive to resource dependence relationships with the government (Marquis & Qiao, Reference Marquis and Qiao2020) and tend to rely more on networking to gain legitimacy and resources. Empirically, the survey dataset contains firm-level business entertainment expenses (e.g., eating, drinking, gift-giving, and bribes in disguised form) (Cai, Fang, & Xu, Reference Cai, Fang and Xu2011; Xu, Zhou, & Du, Reference Xu, Zhou and Du2019) and the respective ratios of those spent on government officials and other stakeholders.
Our study constitutes an important contribution to research on corporate political activity (CPA) in emerging economies. While recent years have witnessed burgeoning studies on the contingent/complex impacts of CPA on firm outcomes (Mellahi, Frynas, Sun, & Siegel, Reference Mellahi, Frynas, Sun and Siegel2016; Sun, Reference Sun, Grosse and Meyer2019; Zhang et al., Reference Zhang, Marquis and Qiao2016a), antecedents to CPA have been treated as a well-researched area and not been in the spotlight of the recent nonmarket strategy scholarship. Conventional wisdom contains some ambiguous and inherently inconstant predictions: While resource (in terms of economic or relational capital) -rich firms are predicted to have more capability to engage in CPA (Hillman et al., Reference Hillman, Keim and Schuler2004; Li, Yao, Sue-Chan, & Xi, Reference Li, Yao, Sue-Chan and Xi2011), resource-poor firms are believed to have stronger incentives and urgency to undertake CPA (Xin & Pearce, Reference Xin and Pearce1996). Our article offers a reconciliation of this tension by unraveling deeper and more nuanced power dependence relations between private ventures and the government in China.
Specifically, we identify an inverse U-shaped relationship between firm size and political networking, such that medium-sized firms invest more in political networking than either small or large firms. This non-monotonic relationship reflects firms’ varying power dependence relations with the government. Meanwhile, we find that while embedded political ties substitute for political networking, achieved political connections stimulate more political networking activities. This highlights the roles of different types of political ties in shaping business-government power dependence relations (Sun et al., Reference Sun, Mellahi, Wright and Xu2015; Zhang et al., Reference Zhang, Marquis and Qiao2016a). Finally, firms located in business development zones are found to exhibit a lower political networking intensity than those outside the zones. This demonstrates that the local institutional environment also reflects different patterns of business-government interdependence.
Our study sheds light on how RDT can be utilized to deepen the CPA research. Pfeffer and Salancik (Reference Pfeffer and Salancik1978/2003: xvi) once lamented that RDT had become little more than a ‘metaphorical statement about organizations’, and most studies that use this theory cited it only ceremonially (Wry, Cobb, & Aldrich, Reference Wry, Cobb and Aldrich2013). While the application of this theory has yielded numerous insights into the consequences of CPA (Hillman, Reference Hillman2005; Sun, Hu, & Hillman, Reference Sun, Hu and Hillman2016), our article suggests that RDT can reveal nuanced, underlying factors that shape the business-government interdependence and subsequently the motivation and necessity of CPA. Taking account of the interactive nature of the business-government relation, we unpack the general notion of interdependence into the size-based, connection-based, and location-based dependence relationships that drive the formulation of a firms’ political networking strategy.
THEORETICAL BACKGROUND AND HYPOTHESES DEVELOPMENT
RDT (Pfeffer & Salancik, Reference Pfeffer and Salancik1978/2003) provides critical insights into the nature of the business-government relationship and various ways used by firms to manage it. Business organizations, the theory suggests, are facing a web of interdependencies with external stakeholders. Organizational survival and growth, therefore, are dependent on their ability to obtain resources from and manage uncertainties caused by external constituents. To better manage these interdependencies, organizations can employ a variety of tactics, including co-option (Hillman, Reference Hillman2005) and constraint absorption (e.g., mergers and acquisitions, Casciaro & Piskorski, Reference Casciaro and Piskorski2005). Networking with powerful external stakeholders is a key co-option tactic, through which a focal firm engages in the exchange of valuable resources such as information, friendship, and status with these stakeholders. In so doing, potentially hostile elements of environmental uncertainties can hopefully be averted or absorbed into the focal firm.
Hillman and colleagues (2009: 1412) recognize that the government is ‘one of the most difficult environmental dependencies to control’. This is particularly true in China, where political actors control enormous resources and can create considerable uncertainties for private businesses, such as issuing licenses/permits, granting subsidies and tax breaks, and providing access to government-funded projects and government procurement (Nee & Opper, Reference Nee and Opper2012). Moreover, they can wield power over a focal firm's other crucial stakeholders, such as creditors, suppliers, and buyers, thus indirectly affecting the flow of resources to the focal organization (Sun, Mellahi, & Thun, Reference Sun, Mellahi and Thun2010a).
On the other hand, the government also relies on firms to contribute to fiscal revenues and generate employment, thus helping politicians secure their tenure or obtain promotion (Li & Zhou, Reference Li and Zhou2005). The interdependent relationship between firms and the government can be unequal: some firms may be more reliant on the government for the latter's resources than the government on the firms; alternatively, political actors can be more reliant on some firms for their economic/financial contributions than the other way round. In short, the necessities of political networking in the focal firm hinge critically upon the degree of their interdependence.
Further developments in RDT have unpacked the concept of interdependence by distinguishing two theoretical dimensions, namely power imbalance (or dependence asymmetry) and mutual (or joint) dependence (Casciaro & Piskorski, Reference Casciaro and Piskorski2005; Gulati & Sytch, Reference Gulati and Sytch2007). Specifically, power imbalance results from power differentials between two actors in a dyad, which result from the difference between the dependence of actor A on B and that of actor B on A. Mutual (or joint) dependence refers to the sum of actor A's dependence on B and actor B's dependence on A, irrespective of whether the two parties’ dependencies are balanced or not. As such, by taking into account both power imbalance and mutual dependence, we are able to better understand power dependence relationships in the business-government dyad. While a highly imbalanced power relationship gives the weak party (the private firm) strong motivations to engage in co-optive, networking activities, the feasibility of achieving intensive political networking depends on the degree of mutual dependence between both parties.
In this article, we contend that the power balance and mutual dependence between private businesses and the government have to do with the economic and sociopolitical capital possessed by a focal firm and the institutional environment where the firm is located. In particular, we identify size- (economic capital), connection- (sociopolitical capital), and location-based dependences to understand the drivers of political networking in China. In what follows, we develop specific hypotheses with illustrative interview quotes from our field research provided where necessary.
Size-Based Dependence and Political Networking
Organizational size represents one of the most studied antecedents to CPA in the developed economy context (Hillman et al., Reference Hillman, Keim and Schuler2004; Lux, Crook, & Woehr, Reference Lux, Crook and Woehr2011). Indicating a firm's available economic resources/power, size has long been found to have a positive association with key political tactics such as lobbying and campaign contributions. When it comes to political networking in China, while some prior studies suggest that political networking activities are more essential to small firms than large ones owing to the former's legitimacy/resource disadvantages (Haveman et al., Reference Haveman, Jia, Shi and Wang2017; Li & Zhang, Reference Li and Zhang2007; Peng & Luo, Reference Peng and Luo2000), empirical findings in this regard remain equivocal (e.g., Park & Luo, Reference Park and Luo2001). We believe that this ambiguity entails deeper theorization of the networking patterns between firms with different sizes and the government. Conceptually, the association between firm size and political networking can be better understood by considering both the necessities and the capabilities of managing the business-government interdependence, which is shaped by two facets of the power-dependence relationship – power imbalance and mutual dependence.
First, the necessities of a firm to invest in political networking are an increasing function of its power imbalance with the government. Further, firm size affects the magnitude of such power imbalance. For small private businesses, there is salient power imbalance as their dependence on government agencies is far greater than government agencies on them. This, in turn, suggests a stronger need for them to engage in political networking to reduce this power disparity.
However, mutual dependence, the second dimension of business-government interdependence, is unlikely to be significant in the interactions between small firms and political actors. This is because the magnitude of relational rents to be generated by the two parties is limited. Since the presence of considerable mutual dependence is a necessary condition for focal firms’ active engagement in co-option activities (Casciaro & Piskorski, Reference Casciaro and Piskorski2005), small firms will not devote significant efforts to political networking despite their power disadvantages. For instance, small organizations may not have the necessary resources to invest in intensive political networking (Aldrich & Auster, Reference Aldrich and Auster1986), nor do they have the necessary bargaining power, or the ability, to appropriate sufficient relational rents from networking with government officials (Sun et al., Reference Sun, Wright and Mellahi2010b). Therefore, small firms would not devote significant efforts/resources to political networking if the rewards are too limited to cover the corresponding cost.
Consistent with the foregoing argument, our field research in China reveals a passive attitude of small firms to political networking. Senior executives told us that their intermittent dealings with government officials occurred either when they needed to get through administrative procedures or when they ran into troubles. One business owner recounted the following during our interview:
When I registered the firm, I had to get all necessary permits and licenses from various government branches. Since I had no connections, I had to spend money entertaining those officials in charge. But once I got what I needed, I did not bother to contact them anymore. As you know, nobody (officials) pays attention to you when your business is small. So the best strategy is just to focus on your business.
As firms grow larger, resource constraints become less salient than those of smaller peers. What is more, medium-sized firms will have to interact more than their smaller peers with government agencies for critical resources (such as land, loans, and various permits) as well as other stakeholders, such as creditors, suppliers, distributors, and customers. In emerging markets dominated by nonmarket forces, these constituents can be under heavy influence by political institutions and actors (Okhmatovskiy, Reference Okhmatovskiy2010; Sun et al., 2010a). For example, firms may need to transact with state-owned or politically-connected suppliers or customers, compete with state-owned enterprises or politically-connected firms in certain market segments, and seek loans from state-owned banks to finance their further growth (Nee & Opper, Reference Nee and Opper2012). The increasing level of political embeddedness on the part of medium-sized firms enhances business-government mutual dependence, thus necessitates a much greater level of political networking to sustain their growth. An owner of a garment factory explained this during our interview:
After five years of operation, I came to a threshold to grow my business. For that purpose, I had to get land to build a new factory. As you know, land is controlled by the government. I made lots of efforts and spent a lot in getting land. After getting land, there were dozens of permits and licenses awaiting me. It took me almost a year to get everything ready. Of course, wining and dining or even bribing officials were a must.
In the case of large firms, the necessities of networking with government officials may eventually decline, despite abundant resources and the strong capabilities available for the networking activities. A general prediction of RDT (Pfeffer & Salancik, Reference Pfeffer and Salancik1978/2003) is that largeness reduces power imbalance between focal organization and the external environment. China's institutional context further strengthens the power advantage of large firms: Local governments display a significant developmental orientation, as government officials in different regions compete with each other for their own promotion in the political regime (Li & Zhou, Reference Li and Zhou2005; Xu, Reference Xu2011; Zhang, Reference Zhang2008). And this developmental orientation tends to benefit large firms more significantly.
Large firms, due to their position in the local economy and their contributions to tax, employment, and economic growth, are often more favored and protected by government officials than their smaller peers. While large firms need policy favors to sustain their scale and performance, politicians also rely on large businesses to improve their economic performance and, ultimately, their career prospects. The relatively low power imbalance between large firms and government officials will lead to a lower necessity of political networking, thereby freeing their time and resources to cultivate relationships with their suppliers, customers, and other key business partners. Nee and Opper (Reference Nee and Opper2012: 239) documented this scenario in their interview, ‘it was not us who approached government. We were already the number one company in our sector, so the government was proud of us and wanted to meet us. Then you just manage your relationship with government as you do with customer relations’.
Finally, firms with different sizes are not equally vulnerable to potential political predation. While self-interested officials are generally hesitant to expropriate many resources from large firms under their jurisdiction due to their economic and social clout, small firms do not have adequate resources and visibility to attract intensive rent-seeking activities either. It is medium-sized firms that are most susceptible to officials’ opportunistic activities; they have more resources available and higher visibility (i.e., greater mutual dependence with the government) than small firms, but much lower bargaining power (i.e., greater power imbalance with the government) than large firms. A tax bureau director told us his experiences in selecting candidates for tax investigation:
Large firms are hard to punish since the owners do have bargaining power with my superiors. Thus, it is risky to offend large business owners. Small firms are too poor to exploit – they have fewer resources available, so it is not necessary to squeeze them. Comparatively, middle-sized firms are good targets for our investigation. They have money, while their connections are not powerful enough to prevent our investigation.
Therefore, for medium-sized private businesses, investment in political networking represents an important co-option strategy for sustaining their growth and guarding against predation. Summing up the preceding arguments, we propose:
Hypothesis 1: There will be an inverted U-shaped relationship between organizational size and political networking investment, such that medium-sized firms will invest more in networking with government officials than will either small or large ones.
Connection-Based Dependence and Political Networking
Besides the economic capital/power indicated by size, a firm's sociopolitical capital also influences the power imbalance and mutual dependence in the business-government dyad. In this study, we focus on the sociopolitical capital rooted in a firm's existing political connections and examine how it affects the focal firm's political networking activities. Reflecting on the origin and nature of political connections in China's institutional context, we distinguish two types of corporate political connections: embedded political ties and achieved political connections. We argue that they represent different patterns of business-government interdependence relationships and thus result in different political networking behaviors.
Embedded political ties refer to managerial and employee linkages embedded in longstanding network ties to the political institutions and actors. They range from entrepreneurs’ and senior executives’ prior government working experiences[Footnote 1] (Lester, Hillman, Zardkoohi, & Cannella, Reference Lester, Hillman, Zardkoohi and Cannella2008; Michelson, Reference Michelson2007) to the hire of the relatives of government officials as firm employees (Wank, Reference Wank1999). These political ties embody the cultivation of long-term, reciprocal relationships between business people and the government through common social backgrounds and family/kinship ties. Rather than governed by arm's length transactions in which each party seeks to maximize short-term benefits, these dense interpersonal relationships allow trust and mutual indebtedness to develop over time (Sun et al., Reference Sun, Mellahi, Wright and Xu2015).
According to RDT, embedded political ties highlight the co-option of political forces by private ventures under weak institutions (Sun, Reference Sun, Grosse and Meyer2019; Sun et al., Reference Sun, Mellahi and Wright2012a). As sociopolitical capital in stock, these embedded ties equip connected firms with insider information/knowledge about the policy process, financial resources, and potential business opportunities. They also shield firms from political interference and various forms of government expropriation. For example, Nee and Opper's (Reference Nee and Opper2012) survey in Yangtze Delta region reveals that entrepreneurs with former cadre positions and those who have relatives with cadre positions were more likely to get bank loans, government contracts, and land use rights.
As such, embedded political ties serve to reduce power disadvantages on the part of the connected firms, thereby reducing the necessity for undertaking intensive political networking activities. This is because these firms may already obtain essential sociopolitical capital for their survival and growth. In the case of semi-forced charitable donations in China, Zhang and colleagues (2016a) find that firms with executives having prior government working experiences are more capable of buffering themselves from local governments’ pressures to donate.
In brief, embedded political ties reduce the power imbalance between the focal firm and the government. Therefore, we expect the presence of embedded ties to obviate the need to engage in time-consuming and oftentimes controversial connection-building activities. Thus, we propose:
Hypothesis 2a: There will be a negative association between the intensity of a firm's political networking investment and the presence of embedded political ties, i.e., embedded political ties and political networking are substitutes.
Achieved political connections refer to linkages established by entrepreneurs’ and managers’ prestigious appointments to political institutions, such as legislative bodies, which suggest business people entering politics after they have been successful (Wang et al., Reference Wang, Du and Marquis2019; Zhang et al., Reference Zhang, Marquis and Qiao2016a). This type of political connection involves not only firms’ intention to capture political institutions, but also the government initiatives to co-opt focal firms. Specifically, the government can offer political recognition, social status, and prestige to the business people, and as a response, firms are expected to meet the government's requirements/demands (Marquis & Qian, Reference Marquis and Qian2014). From the government's perspective, the appointments act as a co-option strategy to incorporate certain business elites into the political system. By so doing, the government can leverage firms’ resources to facilitate local economic/social development, and officials can capture their own private benefits (Dickson, Reference Dickson2003; Tsai, Reference Tsai2007).
From the firm's perspective, the appointments are achieved by corporate efforts to engage actively with the government's demands, such as contributing to government-sponsored projects and making charitable donations (Ma & Parish, Reference Ma and Parish2006; Mellahi et al., Reference Mellahi, Frynas, Sun and Siegel2016). Consequently, achieved political connections can serve as a bridge through which government officials can further influence businesses and seek rent, while private businesses aim to capitalize on their political positions through more intensive engagement with the government. That is, the close business-government exchange relation can foster their mutual dependence, which in turn leads to more intensive political networking activities and investment.
In addition, achieved political connections are less capable of reducing the power imbalance between the two parties than embedded political ties. The political appointments are certainly replaceable if the connected firms fail to maintain a stable favor-exchange relationship with the government. The fear of losing these positions also motivates firms to maintain political networking investment because such appointments are subject to competition from many other successful entrepreneurs. Hence, we propose:
Hypothesis 2b: There will be a positive association between the intensity of political networking investment and the presence of achieved political connections, i.e., achieved political connections and political networking are complements.
Location-Based Dependence and Political Networking
Aside from firm-specific characteristics, firm location is an important environmental factor that shapes the power dependence relationship between private businesses and the government. Relative to networking with market actors that may spread out geographically, political networking is more affected by a firm's location since most regulatory/administrative authorities are located in the place of the firm's operation. While prior research has long recognized the role of subnational regions in shaping firm strategies/outcomes, most of these studies in the Chinese context focused on provincial contingencies (e.g., Deng, Yan, & Sun, Reference Deng, Yan and Sun2020; Shi, Sun, & Peng, Reference Shi, Sun and Peng2012; Sun, Qu, & Liao, Reference Sun, Qu and Liao2018; Zhang et al., Reference Zhang, Marquis and Qiao2016a). In this article, we identify a firm's most proximate locational factor – business development zones – to see if and how business-government relations can be restructured in the zones.
Here, the behaviors and efforts of the government make a difference in the dyadic business-government relations. In the Chinese context, local governments play an active role in developing the local economy through building infrastructure, attracting investment, and promoting businesses. The intense interregional competition facilitates a developmental orientation, such that local governments compete with each other in attracting domestic and foreign investments (Li & Zhou, Reference Li and Zhou2005; Marquis & Qiao, Reference Marquis and Qiao2020; Nee & Opper, Reference Nee and Opper2012; Xu, Reference Xu2011).
As a part of the government's developmental efforts, local governments have established various types of development zones, such as high-technology parks, tax-free zones, and industry parks, to attract investment and promote certain strategic, high-tech industries (Armanios et al., Reference Armanios, Eesley, Li and Eisenhardt2017; World Bank, 2011). Although these development zones were originally designed to achieve certain missions (e.g., incubating high tech, fostering export, etc.), most zones evolved into useful vehicles for local governments to boost economic activity in their jurisdictions, ‘taking whichever route provides the fastest growth’ (Heilmann, Shih, & Hofem, Reference Heilmann, Shih and Hofem2013: 903).
Under this intense interregional competition, the last two decades have witnessed rapid growth of the number of development zones. There had been 438 national-level development zones as of the year 2017 (Shen, Reference Shen2018), and more than 10,000 development zones at provincial, city, and county levels by the mid-2010s (Liu & Yang, Reference Liu and Yang2014). Various policy incentives are present to attract investment, including tax breaks, streamlined regulatory procedures, and better financing access (Heilmann et al., Reference Heilmann, Shih and Hofem2013; Zhang & Sonobe, Reference Zhang and Sonobe2011). For instance, the Wuxi municipal government arranged start-up capital and free office space for their target firms (Heilmann et al., Reference Heilmann, Shih and Hofem2013).
Besides the policy favors, business development zones provide focal firms with crucial institutional buffers from red tapes and potential political predatory activities. As Heilmann et al. (Reference Heilmann, Shih and Hofem2013) observed, a key function of development zones is to provide a protective ‘umbrella’ to those firms located inside the zones. In other words, the establishment of business development zones involves creating and nurturing a more market-friendly environment outside the old institution. For instance, the establishment of development zones in the city of Dalian pitted a moribund state sector in the urban core against a vibrant market sector outside of the city (Rithmire, Reference Rithmire2013).
In sum, both the policy favors/inducements and the business-friendly environment created in development zones serve to reduce power imbalance between private businesses and the government. After all, building development zones originated from government efforts, and the success of the zones depends on the investment by and the growth of firms inside the zones. As such, compared to those outside the business development zones, firms located in the zones are subject to much smaller power disadvantages over political agencies and actors. Therefore, the motivations and necessities for political networking investment are expected to be significantly weaker for firms located in development zones than their off-zone peers.
Hypothesis 3: Firms located in business development zones are likely to invest less in political networking than those located outside the zones.
METHODS
Data Collection
In 2008 and 2009, we conducted extensive field research, consisting of more than 40 interviews with business executives from various private ventures. Our interviewees provided detailed information about their own understanding of business-government interactions, and we accumulated a wealth of qualitative information to understand the underlying mechanisms that lead to the variation of political networking behaviors. Based on our literature review and interviews, we developed a questionnaire for a survey that was conducted in 2010.
Considering the large size and high degree of heterogeneity across the country as well as the huge number of private businesses, we adopted a two-step sampling procedure: joint use of purposeful sampling and random sampling. We first chose six provinces: three in coastal regions (Guangdong, Fujian, and Zhejiang), and three in inland regions (Hubei, Sichuan, and Shanxi). Second, one municipality (an administrative level below province) was chosen in each province, from which one district or county was selected as the final research site. We had good access to government officials in these sites, which facilitated the participation of local business executives in the survey. In each research site, we randomly selected a sample of firms from the list of business registration in the local Industry and Commerce Administration Bureau. 400 firms in total were randomly selected for our research.
Since it is challenging to make business executives cooperative through mail surveys, we developed clear interview protocols and trained our research assistants to conduct face-to-face interviews for questionnaire collection. On-site interviews also helped us gain access to the right respondents and ensured the correct understanding of items. All the respondents were business owners or senior executives who had access to the sensitive information involved in their own firms. We promised respondents beforehand to offer them detailed research feedback based on their responses to ensure that they answered the questions more carefully and honestly. In the end, 179 out of the 400 sampled firms cooperated. Our further check of the business registration information did not show systematic non-response bias.
Further, we took various measures to ensure reliability. First, a follow-up check was conducted after we received the completed questionnaires in the survey. Specifically, 20 firms were selected randomly, and telephone interviews were carried out with other top managers who did not participate in the initial survey. All of the correlations of matched variables between the two raters were above 0.92, indicating strong interrater reliability.
Second, we conducted Harman's single-factor test to check the presence of common method variance (Podsakoff, MacKenzie, Lee, & Podsakoff, Reference Podsakoff, MacKenzie, Lee and Podsakoff2003). If common method variance was a serious problem, we would expect a single factor to emerge from a factor analysis or a general factor to account for the majority of the covariance among the independent and dependent variables. We performed a factor analysis of all items, which extracted five factors whose eigenvalues were greater than one. No general factor was apparent in the unrotated factor structure, with the first factor accounting for about 16% of the variance. Thus, potential common method bias can be ruled out.
Variables
The dependent variable measures the relative spending on networking with government officials versus other external stakeholders in the sample firms. In China, there is an accounting item – business entertainment expenses – that encompasses a wide range of corporate networking expenditures on banquets, gifts, club memberships and may include certain bribes in disguise (Cai et al., Reference Cai, Fang and Xu2011; Du, Guariglia, & Newman, Reference Du, Guariglia and Newman2015; Xu et al., Reference Xu, Zhou and Du2019). In the survey, we asked the respondents to report not only their firms’ total business entertainment expenses in 2009, but also the percentages of the expenses spent on government officials, suppliers, distributors, customers, business partners, and bank managers. The percentage spent on government officials is used as the dependent variable. This measure is ideal since the total entertainment expenses as a percentage of sales may decline with firm size, but the percentage of entertainment expenses on officials may not naturally decline.
This measure of political networking complements traditional perceptual ones used in prior studies (Park & Luo, Reference Park and Luo2001; Peng & Luo, Reference Peng and Luo2000) by alleviating potential subjective bias in business executives. While previous studies generally examined the utilization of business ties and political ties in a parallel or composite form (e.g., Sheng, Zhou, & Li, Reference Sheng, Zhou and Li2011; Zhang et al., Reference Zhang, Zhao and Zhang2016b), our study tries to understand how private firms evaluate the relative importance of political networking and make strategic decisions in allocating resources accordingly. To the best of our knowledge, this is among the few studies that collect objective information on firms’ detailed allocation of networking expenses. It captures the weight of political networking in firms’ overall networking activities or the relative importance of political ties to that of general business ties.
Explanatory variables include firm size, a firm's existing political connections, and the location of a firm. Firm size is measured as the natural logarithm of the total number of employees (Li & Zhang, Reference Li and Zhang2007). Embedded political ties are generated by two items: (1) Whether the firm has employees (including but not limited to the entrepreneur) who had prior government working experience; (2) whether the firm has employees who are the relatives of government officials. A binary variable takes a value of 1 if the firm meets either of the above conditions, 0 otherwise. This measure is more comprehensive than most current studies that only capture whether executives had prior government working experience (Zhang et al., Reference Zhang, Marquis and Qiao2016a). Achieved political connections are measured by a binary variable indicating whether a firm has employees (mainly the entrepreneur) who have political appointments to political organs such as the People's Congress or the Chinese People's Political Consultative Conference (Zhang et al., Reference Zhang, Marquis and Qiao2016a). Firm location is measured by a binary variable indicating whether a firm is located in a business development zone (1 = yes, 0 = no). In China, firms located in these zones enjoy more developed market-supporting institutions and various policy favors.
With respect to control variables, we first control for a firm's total business entertaining expenses (in the natural logarithm form), which indicates the overall investment in stakeholder networking. Firm age is calculated by using the year 2010 minus the founding year of the firm (logged). We use a binary variable hybrid to indicate whether a sample firm has state shares. Whether a firm is identified as a high-tech firm (tech) by the local government is also controlled for, since such firms may receive special favorable treatment from the government.
Perception of the importance of political networking is measured by answers to the following question in the questionnaire: ‘It is always said that it is impossible to do business in China without guanxi; to what extent do you agree with this statement when it comes to relations with government officials’? The respondents provided their own ratings with a 7-point Likert scale, ranging from ‘not important at all’ (coded 1) to ‘extremely important’ (coded 7).
Besides firm-level controls, Industry is coded into four categories: Manufacturing, service, real estate, and mining/agriculture, with firms in the manufacturing sector as the reference group. We also control for the quality of provincial-level institutions by using National Economic Research Institute (NERI) Index of Marketization in China's provinces (Fan, Wang, & Zhu, Reference Fan, Wang and Zhu2011). The NERI indices capture the depth of institutional transition in China's provinces and have been widely used in the previous literature (e.g., Sun et al., Reference Sun, Hu and Hillman2016; Zhang et al., Reference Zhang, Marquis and Qiao2016a).
After including all the dependent, explanatory, and control variables in our estimation models, the sample size shrank to 118 because of the missing values in some of the variables mentioned above. We compared the firms in our final sample, with firms dropped for missing values on total assets, sales, and employment. Tests for differences in means did not reveal significant differences between the two groups, indicating that sample attrition is not a problem in our final sample.
RESULTS
Table 1 presents descriptive statistics and correlation matrix. On average, firms spent about 20.62% of their total entertainment expenses on government officials, with a considerably large standard deviation. This demonstrates a considerable variation on the part of the dependent variable that needs to be explained by organizational and environmental factors. The table shows that political networking investment is positively correlated with achieved political connections but negatively correlated with business development zones. Its correlations with firm size and embedded political ties lack statistical significance. Finally, the pair-wise correlations among the explanatory variables are largely negligible, except for a positive one between size and achieved political connections, which implies the aforementioned co-option of large private businesses by the local government.
Table 1. Descriptive statistics and correlations
Notes: N = 118. * and ** denote significance levels of 0.05 and 0.01 respectively.
We use fractional logit regression models to examine the determinants of political networking investment, with results reported in Table 2. Since our dependent variable – the proportion of a firm's entertainment expenditure on government officials – is a fractional response variable, Papke and Wooldridge (Reference Papke and Wooldridge1996) recommend the fractional logit model using a quasi-likelihood estimation method. This model is believed to be preferable to the conventional OLS model in terms of dealing with dependent variables (e.g., proportions) ranging between 0 and 1.
Table 2. Fractional logit regression models examining the drivers of political networking investment
Notes: N = 118. Standard errors are reported in parentheses. +,*, and **denote significant levels of 0.1, 0.05, and 0.01 respectively.
In the baseline model 1, only control variables are included. We find that firms in real estate and mining/agriculture sectors assign a larger proportion of their entertainment expenses on political networking than their manufacturing peers, as these firms may rely more on government officials to get critical resources (particularly land use rights and mineral resources) and various administrative permits. Conversely, high-tech firms have a much lower intensity of political networking than the other firms.
We then introduce linear and quadratic terms of firm size respectively in models 2 and 3, add the two variables of political connections in model 4, and finally add firm location in model 5. Model 5 is the full model. The model fit improves after explanatory variables are added.
H1 predicts an inverted U-shaped relationship between firm size and entrepreneurial political networking. This hypothesis receives strong support. While the results in model 2 suggest a monotonic relationship between firm size and political networking investment, results in model 3 show a strong and significant curvilinear effect with a significantly positive linear term and a significantly negative quadratic term. The explanatory power also increases from 0.082 (model 2) to 0.094 (model 3).
Following recommendations from Haans, Pieters, and He (Reference Haans, Pieters and He2016), we perform a wide range of checks for the inverted U-shaped curve and reported additional results in Table 3. First, we use the U test developed by Lind and Mehlum (Reference Lind and Mehlum2010), and find that the U-shape curve is significant at 5% level. Model 3 in Table 2 suggests that the inflection point occurs at 6.7 (logged firm size, or equivalently 812 employees), which is within one standard deviation (1.617) above the mean (5.413). We also find that the marginal effects of firm size on political networking investment at both the minimum and maximum of the firm size are strong enough to qualify an inverted U-shaped relationship.
Table 3. Tests of the inverted U-shaped relationship between firm size and political networking
Notes: Standard errors are reported in parentheses. +,*, and ** denote significant levels of 0.1, 0.05, and 0.01 respectively.
Meanwhile, our further analysis in Column 1 of Table 3 suggests that the relationship is not consistent with an S-shaped relation. After adding a cubic term in the original model, all the estimated coefficients related to firm size become insignificant. In columns 2 and 3, we split the sample by the value of inflection point and show that while the relationship between firm size and political networking investment is positive in the sub-sample to the left of the inflection point, the relationship becomes negative in the other sub-sample.
Further, we use a semiparametric approach developed by Ai and Chen (Reference Ai and Chen2013) to examine the relationship between firm size and political networking investment. The approach uses a partial linear model where all other variables except for the firm size exhibit a linear relationship with political networking investment, and then estimates the residuals with firm size in a nonparametric fashion without imposing functional form of the relationship. The results are displayed in Figure 1, confirming that the relationship is indeed inverted U-shaped.
Figure 1. The U-shaped curve based on the semiparametric method
H2 posits that embedded political ties substitute for political networking (H2a) and achieved political connections complement political networking (H2b). Results in model 3 of Table 2 lend strong support for this set of hypotheses. The estimated coefficient of embedded political ties is significantly negative (p < 0.05). Others being equal, firms with embedded political ties spend 10.9% less on political networking than their peers. Likewise, the estimated coefficient of achieved political connections is significantly positive (p < 0.01). Also, others being equal, firms with achieved political connections spend 7.0% more on political networking than their peers.
H3 suggests a negative relationship between a firm's location in a business development zone and its political networking investment. This hypothesis receives strong empirical support in model 5 of Table 2. The estimated coefficient of development zones is significantly negative (p < 0.01). Other things being equal, firms located in business development zones spent 10.2% less on political networking relative to those outside the zones.
Further Analyses
To ensure robustness, we also use OLS regression models to test our hypotheses about the determinants of political networking investment, with similar results reported in Table 4. The only notable difference from Table 2 is the result in model 2, where the estimated coefficient of firm size lacks significance.
Table 4. OLS models examining the drivers of political networking investment
Notes: N = 118. Standard errors are reported in parentheses. +, *, and ** denote significant levels of 0.1, 0.05, and 0.01 respectively.
Second, achieved political connections can be endogenous as entrepreneurs might anticipate more political networking and then make a decision to enter politics by taking key political positions. To address this concern, we use the propensity score matching (PSM) approach (Rosenbaum & Rubin, Reference Rosenbaum and Rubin1983) to distinguish between entrepreneurs with political positions and those without. Specifically, we use a Probit model to predict whether an entrepreneur has a political position on the basis of the control variables and then obtain propensity scores and weights. We then use PSM weights for subsequent analysis, with results shown in Table 5.
Table 5. The PSM approach examining the drivers of political networking
Notes: Standard errors are reported in parentheses. +, *, and ** denote significant levels of 0.1, 0.05, and 0.01 respectively. 86 observations remain after propensity score matching.
The first column shows the first-stage Probit model results of the PSM analysis, and the second column shows results regarding the matching quality. After matching, (1) differences of all variables between treatment (firms with political appointment) and control groups (firms without political appointment) are insignificant, (2) pseudo-R square is only 0.003, and (3) median bias is 4.9% -- below the 5% threshold. Taken together, the matching quality is satisfactory, and thus we have alleviated the self-selection problem. After matching, 86 firms remain in our sample. The last five columns show results of our hypotheses and are similar to those reported in Table 2. Therefore, our results remain robust after taking account of the potential endogeneity issue.
Finally, we cluster the standard errors at the provincial level since the location-based power dependence relationship suggests that different firms within the same locality could interact with each other and thus violate the assumption of independence among firms. We reported results in Table 6, which are similar to what is reported in Table 2 and still support our hypotheses.
Table 6. Fractional logit models examining political networking investment with clustered standard errors at the provincial level
Notes: N = 118. Clustered standard errors are reported in parentheses. +, *, and ** denote significant levels of 0.1, 0.05, and 0.01 respectively.
DISCUSSION
This article explores important antecedents of entrepreneurial political networking in China, i.e., the forces driving a private firm's allocation of its limited resources to networking with government officials relative to other external stakeholders. Unpacking the strategic interdependence between firms and the government in the dimensions of power imbalance and mutual dependence, we examine how the economic and sociopolitical capital on the part of private ventures interact with the government to determine the necessities and capabilities of political networking. In concrete terms, we identify size-, connection-, and location-based power dependence relations and reveal their impacts on political networking. The empirical findings contribute to the hitherto underexplored research on the management of political networking investment by private entrepreneurs and shed fresh light on the study of CPA from the RDT lens.
Theoretical Implications
First, our identification of the non-monotonic relationship between firm size and political networking resonates with the call for more in-depth studies of firm behavior from the RDT perspective (Hillman et al., Reference Hillman, Withers and Collins2009; Wry et al., Reference Wry, Cobb and Aldrich2013). Previous studies either indicate that large firms are more politically active because of their resource abundance or suggest that small firms rely more on the government and are thus more motivated to engage in political networking (Hillman et al., Reference Hillman, Keim and Schuler2004; Lux et al., Reference Lux, Crook and Woehr2011; Peng & Luo, Reference Peng and Luo2000; Xin & Pearce, Reference Xin and Pearce1996). Our study contributes to reconciling this tension by revealing a more nuanced relationship between firm size and political networking.
Such an inverted U-shaped relationship originates from and reflects China's institutional environment in which firms with different sizes have different resource dependence relationships with the government. By examining both power imbalance and mutual dependence in the business-government dyad and taking into account both necessities and capabilities of political networking, our research helps to address prior inconsistent findings on the size-networking relationship.
For example, while Peng and Luo (Reference Peng and Luo2000) found a stronger positive performance effect of managerial ties with government officials in smaller firms than in larger ones, Park and Luo (Reference Park and Luo2001) failed to find support for their hypothesized negative association between firm size and the utilization of ties with government authorities. They went on to assert that ‘firms in China develop and maintain good connections with government authorities regardless of their resources’ (471, italics added). Our study provides a deeper rationale for the non-monotonic size-networking relation. Future studies can combine the power imbalance and mutual dependence logics to examine organizational antecedents to other nonmarket strategies in various institutional contexts.
Second, we contribute to the CPA literature by exploring the interrelationship between the sociopolitical capital embodied in firms’ existing political connections and their strategic political networking activities. Given that no CPAs are cost-free, firms may choose from or combine various tactics from their political strategy toolbox. However, we know little about how firms develop and/or manage such a CPA portfolio (Ridge, Ingram, & Hill, Reference Ridge, Ingram and Hill2017), despite a burgeoning stream of research on the complementarity between CPA and strategic corporate social responsibility (Mellahi et al., Reference Mellahi, Frynas, Sun and Siegel2016; Werner, Reference Werner2015). On the basis of the political tie heterogeneity perspective (Sun et al., Reference Sun, Mellahi, Wright and Xu2015; Zhang et al., Reference Zhang, Marquis and Qiao2016a), our differentiation between embedded political ties and achieved political connections reveals subtle differences in the nature and function of political ties and cast fresh light on the complementarity/substitutability between the stock of existing sociopolitical capital and the actions of political networking.
This finding also informs RDT-based strategy research by suggesting potential interactions among various types of dependence-reducing strategies. Pfeffer and Salancik (Reference Pfeffer and Salancik1978/2003) once laid out a variety of dependence-reducing strategies. In reality, firms often engage in multiple strategies to reduce their dependence on the external environment. However, prior research did not adequately consider the interaction of these strategies. Hillman and colleagues (2009) then suggested that research examining resource interdependence investigate the potential substitution effect on the part of the different forms of dependence-reducing strategies. Our study echoes this call and highlights the interactive patterns of different types of co-optive political tactics. Future research can continue this line of inquiry by examining the interactions and compositions of various types of nonmarket strategies that manage the interdependence between firms and the external nonmarket environment.
Finally, our research suggests the importance of examining the incentives/objectives of political institutions and actors when we study CPA through the RDT lens. A majority of prior CPA studies concentrated on the strategic initiatives of the business side while treating the government side as given. However, it can be hard to accurately understand the power dependence relationship by focusing only on the business side (McDonnell & Werner, Reference McDonnell and Werner2016; Wang et al., Reference Wang, Du and Marquis2019). In this article, we take into account the government's incentives and efforts in attracting and co-opting private businesses, which in turn shape the power balance and mutual dependence between the two parties.
Specifically, the government plays an important role in shaping the institutional environment surrounding private ventures. Our finding of the negative association between a firm's business development zone location and political networking indicates that Chinese private firms respond to their changing power relationships with the government by strategically allocating their networking expenditures among government officials and other external stakeholders. The impacts of government behaviors are also reflected by the organizational determinants of political networking. For instance, one important reason why large firms are not pressured to make extra political networking investment is related to the developmental orientation of the Chinese local governments. Similarly, the bridging role of achieved political connections results from government efforts to incorporate business elites into the political system. By taking a holistic view of the business-government interactions, we are better able to understand the nature of their power dependence relationships and, thus, their impacts on firm behaviors.
Practical Implications
Our article also offers crucial managerial implications for doing business in China. Under time and financial constraints, senior executives in all types of business organizations need to carefully decide whom they network with and the associated networking intensity to secure their firms’ survival and growth.[Footnote 2] Our research findings suggest that in emerging economies like China, firms are likely to devise and execute networking strategy based on their organizational attributes (e.g., economic clout and sociopolitical capital) and environmental characteristics in their vicinity.
Particularly, firms should adjust their political networking intensity during different stages of their growth and across different types of locations. Firms can also strategically choose and combine various political activities. While firms with embedded political ties may need to avoid overinvestment in political networking, firms with achieved political connections need to be cognizant of the underlying rationales for continued networking investment in political relationships. In a word, there is no one-size-fits-all political networking strategy for private businesses. From the government's perspective, establishing development zones has a positive effect on the reduction of firm burdens on entertaining government officials, whereas incorporating successful business people into the state organs seems to have an opposite effect.
Limitations and Future Research Implications
This research has several notable limitations, which can motivate future studies. First, as we are interested in exploring the impact of business-government power dependence relations on political networking, we identify size-, connection-, and location-based dependences to implement this research idea. However, we do not claim that these factors exhaust the typology of this power dependence relationship. Future studies can explore and identify other critical antecedents that can affect the power imbalance and mutual dependence between business and government.
Second, in our dataset, we failed to identify any discernable interaction effects between the three types of dependencies on political networking. While we are not certain whether there are theoretically compelling reasons to expect any specific interactive patterns, we encourage future studies to use a larger scale survey and more developed measures to explore the potential interactions between multiple dependence variables.
Third, we recognize that political networking activities can involve both legal and illegal practices. Unfortunately, our measurement of political networking cannot distinguish the two types of activities, as in practice, it can be challenging to draw a clear line between them. Future research can try to untangle normal, legitimate networking activities and illegal/illegitimate, corrupt ones and examine if and how the antecedents/dynamics of the two types of networking activities may differ.
CONCLUSION
In closing, political networking is a widely used firm strategy in emerging economies to navigate their challenging institutional environments. This article applies the RDT insights to the context of political networking in China and reports novel findings on the drivers of this strategy. We hope this study can spark additional thought-provoking pieces to the broader intellectual project of understanding business-government relations in emerging economies.
