Introduction

The topic of corporate responsibility (CR) has become increasingly relevant not only for academics but also for practitioners and policy makers. In this paper, we address two questions. First, what are the different patterns or configurations of CR practices associated with high financial performance? Second, do these CR practices share similarities or differences across societal contexts, and if so, why? We sought to answer these two questions in a comparative study of CR practices in three Chinese societies, namely, Mainland China (China), Hong Kong, and Taiwan.

This study aims to contribute to current knowledge about the evolution and dynamics of CR practices in a number of respects. First, we investigate the extent to which attaining high financial performance is facilitated or hindered by different ‘strategic’ patterns or configurations of stakeholder practices (Porter and Kramer 2006). An instrumental approach to stakeholder management emphasizes the financial benefits of gaining legitimacy and social capital from organizational stakeholders (Aguinis and Glavas 2012; Donaldson and Preston 1995; Taneja et al. 2011). However, there have been mixed research findings regarding the relationship between CR and financial performance (e.g., Falkenberg and Brunsæl 2011; Margolis and Walsh 2003; Orlitzky 2011; Orlitzky et al. 2003; Peloza 2009). One identified reason is that the CR construct has been narrowly and differently operationalized across studies (cf Margolis and Walsh 2003; Montiel 2008; Peloza 2009). For instance, numerous studies on CR and financial performance have focused on practices associated with only one stakeholder group such as communities (Brammer and Millington 2008; Wang and Qian 2012) or the environment (Dixon-Fowler et al. 2013; Russo and Fouts 1997). Other studies have treated various domains of CR practices as conceptually equivalent by using simple summated scores (e.g., Hillman and Keim 2001) or using aggregated scores based on subjectively assigned weights to different CR domains (e.g., Waddock and Graves 1997). As such, there remains considerable uncertainty regarding what happens when firms prioritize the interests of certain stakeholder groups over those of other stakeholder groups (Fransen 2013; Freeman 2010).

To address these issues in the CR literature, we take a more holistic and integrative approach to identify the prevalence of five domains of CR practices (i.e., employee, customer, investor, community, and the environment) and their associations with high financial performance. We follow the instrumental approach to stakeholder management that links firms’ internal and external CR activities to their performance (Jones 1995). We argue that firms do not implement various CR practices separately, but that firms prioritize various CR issues and practices in order to maximize the effectiveness of their stakeholder management programs. The concept of equifinality accepts that there are different ways leading to the same outcome (Gresov and Drazin 1997). We bring the equifinality concept to the CR literature by investigating whether different configurations of CR practices may have similar contributions to high firm performance.

This study also contributes to the debate regarding the convergence–divergence of CR practices across societal contexts. Two institutional theory traditions provide different rationales for understanding such processes (Tempel and Walgenbach 2007). In particular, new institutionalists propose that standardized practices become dominant as a result of isomorphic mechanisms (DiMaggio and Powell 1983; Meyer and Rowan 1977) that can facilitate the transnational diffusion process (Kostova 1999). In contrast, the national business systems institutional framework emphasizes the nature of “national embeddedness” such that national-level variation in institutional environments engenders the ongoing divergence of organizational practices (Matten and Moon 2008; Whitley 1999). Previous cross-national CR research has primarily focused on CR practices in the US and Europe (e.g., Hopkins 2003; Maignan and Ralston 2002; Matten and Moon 2008). However, we propose that the Greater China area is a unique and appropriate setting to test the competing predictions of these two perspectives in respect to the implementation and benefits of CR practices.

In this comparative study, we examine the relationship between configurations of CR practices and financial performance of firms in three societies: China, Hong Kong, and Taiwan. Although these three societies are similar in respect to their Confucian-based cultural heritage, they represent different economic and political institutional environments (Cheung et al. 2010; Child and Tsai 2005; Chow 2004; Moon and Shen 2010) that may influence societal expectations regarding the role and nature of corporate responsibility. What are the predominant patterns of CR practices that enable firms to achieve high financial performance in these three societies? To answer this question, we take an institutional perspective to investigate whether cultural, regulative, and normative institutional differences and similarities explain the extent to which CR programs in these three East Asian societies share standardized forms or diverge into unique patterns. This cross-societal study of CR practices examines the interactions between various institutions within different societies that may result in the presence of similar or dissimilar configurations of CR practices. As such, we respond to Fransen’s (2013) call for research to understand how different dimensions of CR practices are related to various national institutions.

The remainder of the paper is organized as follows. In the next section, we present the theoretical background about CR practices and institutional influences on the cross-societal diffusion of CR practices, and then develop study hypotheses. This is followed by the methods section which describes data collection procedures, measures, as well as the set-theoretic approach used in data analysis (Fiss 2007, 2011; Ragin 2000, 2008a). The paper ends with a discussion of findings, implications for research and practice, and concluding observations.

Patterns of CR Practices Across Institutions

Patterns of CR Practices

We regard CR practices as “the set of corporate actions that positively affects an identifiable social stakeholder’s interests and does not violate the legitimate claims of another identifiable social stakeholder (in the long run)” (Strike et al. 2006, p. 852). In order to succeed in a turbulent and globalized business environment, firms need to actively engage with a variety of social issues in their own nations as well as outside their geographic boundaries (Scherer et al. 2009). These expectations pose challenges to firms in dealing with a variety of stakeholder interests and demands (e.g., Campbell 2007; Freeman 1984, 1994; Graves and Waddock 2000; Hillman and Keim 2001). In this regard, Mitchell et al. (1997) proposed that the degree of stakeholder salience (perception of importance) is a function of the power, legitimacy, and urgency of different stakeholders. Managers’ interpretation and evaluation of stakeholder salience informs whether (or not) social actions are undertaken to manage the firm’s relationships with a particular set of stakeholders. Managers may conduct cost-benefit analyses (McWilliams and Siegel 2001) and selectively devote firm resources to certain stakeholder groups in order to establish legitimacy or balance stakeholder needs. Further, CR programs may enable companies to minimize risks, improve strategic competitiveness, enhance corporate reputation, and achieve higher financial performance (Barnett and Salomon 2006; Hillman and Keim 2001).

One intuitive way to classify stakeholders considers the business relationships of stakeholders with a firm and thus generates the two types of primary and secondary stakeholders (Clarkson 1995; Freeman 1984). Primary stakeholders refer to customers, employees, or investors who are involved in direct business relationships with a focal firm, whereas secondary stakeholders include those such as community, environment, media, special interest groups, or government, who do not necessarily have a formal transaction relationship with the firm. Primary stakeholders are more powerful, and their issues are more salient for the firm to achieve strategic goals (Mitchell et al. 1997).

Managers may draw a fine line between the complementarities and tradeoffs of various CR practices with regard to primary or secondary stakeholders, which leads to the presence of different configurations of CR practices. On the one hand, firms may treat multiple stakeholders holistically since attention toward one type of stakeholders does not necessarily prevent being actively involved in other social domains (Freeman 2010). These activities may be complementary since they are congruent with business goals and can significantly enhance competitive advantage. On the other hand, corporate responses to stakeholder demands may be constrained by the availability of slack resources (e.g., Bowen 2002) which often necessitates being selective in terms of responding to various stakeholders’ demands combined with managers’ perceptions of different degrees of stakeholder salience. As a consequence, firms may focus on certain core (or primary) CR practices while treating others as peripheral (or secondary) factors. This balancing of stakeholder demands and resource availability results in different patterns or configurations of CR practices for firms.

Understanding Patterns of CR Practices Across Societies

Institutional theory has been used to understand similarities and differences in the cross-national adoption of various types of management practices (e.g., Kostova 1999), including the adoption of CR practices (e.g., Aguilera et al. 2007; Husted and Allen 2006). Institutions refer to those “collections of rules and routines that define actions in terms of relations between roles and situations” (March and Olsen 1989, p. 160). These rules can be formal or informal (North 1990). In particular, Campbell (2007) proposed that the occurrence of CR activities is associated with a variety of institutional conditions, such as regulatory forces and social norms that impact the actions of corporations and stakeholder groups.

New institutionalism and national business systems theories have contrasting predictions regarding the influence of national institutional systems on the cross-national diffusion of management practices. While new institutionalists (DiMaggio and Powell 1983; Meyer and Rowan 1977) predict the global diffusion of standardized practices, proponents of the national business systems approach (Whitley 1999) argue that different institutional arrangements at the national level result in cross-national variation in the adoption and form of management practices. In their review of the two perspectives, Tempel and Walgenbach (2007) proposed that these different conclusions may be due to fundamental differences along key dimensions such as mechanisms of adaptation, unit of analysis, and institutions. Whereas new institutionalism proposes that isomorphism mechanisms (i.e., coercive, mimetic, and normative pressures) play the key role in the diffusion of practices, the business systems approach recognizes that organizations have close links with institutional arrangements at the national level. Further, Tempel and Walgenbach (2007) observed that new institutionalism emphasizes the strong influence of normative and cognitive institutions, whereas the business systems approach focuses more on the role of regulatory institutions rather than the other two institutional pillars.

Consistent with Tempel and Walgenbach (2007), Matten and Moon (2008) brought the two theories together and applied the new framework to examine differences in corporate social responsibility (CSR) in the US and Europe. They focused on the rationale for adopting “explicit” and “implicit” CSR practices in these societies by comparing key differences in national institutions (i.e., political system, financial system, education and labor system, and cultural system) and isomorphism mechanisms relating to coercive, mimetic, and normative pressures. Matten and Moon (2008) also recognized that changes in the national institutional framework in Europe occurred due to changes in isomorphic pressures. Thus, different institutions may influence and co-evolve with each other, resulting in the cross-national convergence or divergence of CR practices. They also argue for the application of this framework to CR research in other countries including those in Asia.

Configurations of CR Practices and High Financial Performance in Three Chinese Societies

This study of CR practices configurations associated with different levels of financial performance is grounded in configurational theories of organizations which emphasize the notion of complementarities or wholeness (Fiss 2007, 2011; Ketchen et al. 1997; Milgrom and Roberts 1995). Our approach differs from previous studies of the relationship between CR and financial performance that focused on an individual type of CR practice or aggregated CR practices scores (Hillman and Keim 2001; Wang et al. 2008). We argue that such approaches have masked the nature of relationships across various CR practices and their differential impact on firms’ financial performance. While complementarities exist between some types of CR practices (e.g., generous employee benefits may motivate employees to increase product quality, which is associated with customer practices), tradeoffs may be the true for other CR practices (e.g., more environmental friendly practices may not directly benefit investors). Thus, a comprehensive examination of the relationship between CR practices and firm performance should recognize that multiple CR practices work as a whole in the form of configurations rather than separately.

A holistic and integrative CR strategy recognizes the interconnections among various stakeholder interests. In this study, we examine the adoption of CR practice configurations with regard to five domains (i.e., employee, customer, investor, community, and the environment) and their associations with financial performance. We propose that CR practices concerned with the primary stakeholders (i.e., employees, customers, and investors) that are closely related to the business function should be present in configurations leading to high financial performance. When compared to secondary stakeholders (i.e., community and environment), primary stakeholders are regarded as more salient in terms of making legitimate claims, holding more power, and proposing urgent issues (Agle et al. 1999). One reason is that primary stakeholder groups play significant roles in providing key resources in the value chain system of the firm and enabling it to achieve high financial performance, whereas secondary stakeholder groups are less likely to have direct and immediate impact on business operations. For example, the application of CR practices related to customers could enhance customer satisfaction and loyalty, which then creates financial value for the firm (Daub and Ergenizinger 2005). Similarly, organizations may accord higher priority to employee groups because strategic HR configurations that focus on enhancing employee well-being engender employee commitment to their organizations (Delery and Doty 1996; Yu and Egri 2005). Investors are a salient stakeholder group because they provide financial capital and resources for firms.

In sum, a configurational perspective of the concept of CR practices recognizes that managers prioritize different types of stakeholders and thus may selectively focus on one or more primary CR practice domains to the exclusion of secondary ones. The configurational perspective allows for the possibility that managers may take a holistic approach toward these primary stakeholders because attention toward one type of stakeholder does not necessarily prevent a firm from being actively involved in other social domains. Hence, we expect that high financial value should accrue to firms that have a consistently high level of responsiveness to the full spectrum of primary stakeholders rather than according differential treatment to these groups. Therefore, we propose:

Hypothesis 1

In configurations of CR practices related to high financial performance, customer, employee, and investor CR practices are given high emphasis.

Similarity Versus Dissimilarity: CR Configurations Across Societies

Aguilera and colleagues (2003, 2007) argue for the role of both formal institutions (e.g., laws, government policies, regulations, financial markets, etc.) and informal institutions (e.g., social norms, cultural values, NGOs) and their interactions to explain differences in CR practices across advanced economies. Thus, we argue that societal archetypes of CR practices may converge or diverge across societies depending on the influence of key institutional factors (i.e., regulative, normative, and cognitive–cultural factors). Table 1 identifies various institutional dimensions in respect to the cultural values, economic systems, and political and legal systems for the three societies in this study: China, Hong Kong, and Taiwan.

Table 1 Institutional characteristics of the three Chinese societies
Full size table

Societal Culture and CR Configurations

The institutional dimension related to social and cultural systems predicts converged (or similar) CR patterns across these three Chinese societies. Although Hong Kong and Taiwanese managers have been more influenced by individualistic Western management styles (Chow 2004; Hofstede and Bond 1984), these three societies are culturally similar due to their common traditional cultural roots in Confucianism, the use of common languages, and geographic proximity (Hicks and Redding 1983; Ip 2009). Further, the World Values Survey (Inglehart and Welzel 2005) shows that all three societies have high secular-rational values that place a low emphasis on religiosity, patriotism, traditional familialism, and respect for authority. Therefore, a societal culture perspective leads to the following similarity hypothesis:

Hypothesis 2a

The configurations of CR practices for high financial performance are similar for firms in China, Hong Kong, and Taiwan.

Normative/Regulative Institutions and CR Configurations

Normative and regulative institutional factors suggest divergent (or dissimilar) CR patterns across these three Chinese societies. First, in terms of normative factors based on economic system requirements, Taiwan and Hong Kong have important similarities. Each society started their industrialization processes in the 1950s and 1960s, respectively, and both have high levels of economic and human development, highly developed market economies, advanced financial systems, and a high proportion of services industries (Chow 2004; US Central Intelligence Agency 2009). After periods of rapid economic development, Hong Kong and Taiwan have undergone restructuring of their economies with several labor-intensive industries relocating to other regions (e.g., Mainland China, Vietnam, etc.) where manufacturing costs are much lower. While the demand for professional staff remains high, recent unemployment rates were slightly higher for Taiwan (5.9 %) and Hong Kong (5.3 %) when compared to that of China (4.3 %) which has a relative surplus in labor supply (especially unskilled labor) (IMD 2010).

In respect to political and regulatory systems, Hong Kong and Taiwan also share more similarities with each other than with China. Hong Kong and Taiwan both have highly or moderately democratic political systems in which government has either minimal interference or a low degree of direct involvement (US Central Intelligence Agency 2009). In respect to the quality of societal governance, Hong Kong and Taiwan have similar moderate to high levels of government accountability, regulatory quality, rule of law, and control of corruption that are substantially higher than those of China (Kaufmann et al. 2009).

In contrast, China’s transition to a market economy started only in the late 1970s. As a transitional economy, China has a less advanced financial system and is dominated by labor-intensive industries. In addition, China’s high economic growth rate suggests that the entrepreneurial spirit may co-exist with, or sometimes surpass, the value of achieving harmony. For instance, Chinese organizations have been characterized as being primarily focused on achieving high economic efficiency while issues related to employee benefits, environment protection, community development, or product safety have been viewed as less critical (Swanson et al. 2001; Tang et al. 2003). Combined with a relatively weak governance system, Chinese organizations may be more opportunistic and instrumental in terms of stakeholder management than their counterparts in Hong Kong and Taiwan.

Overall, although with some variation, Hong Kong and Taiwan would generally fit the profile for relatively high normative and regulatory pressures from the full array of stakeholder groups as would be the case in many Western free market economies (Aguilera et al. 2007; Campbell 2007). Whereas a relatively higher emphasis on CR practices relating to primary stakeholders than secondary stakeholders is expected (Mitchell et al. 1997), CR practices configurations for high financial performance would also allow for the inclusion of community and environmental CR practices. In contrast, China’s weak regulatory environment as well as transitional economy would indicate that CR practices configurations for financial performance would be significantly different for firms in China. Specifically, China’s normative and regulative context suggests that CR practices related to customer and investor stakeholders would be more predominant to the relative exclusion of CR practices concerned with employee, community, and environmental stakeholder groups (Kolk et al. 2010; Xun 2012). Thus, we propose the following dissimilarity hypothesis:

Hypothesis 2b

The configurations of CR practices for high financial performance are more similar for Hong Kong and Taiwan firms, compared to those for China.

Methods

Data and Sample

This study was based on company surveys conducted in 2006–2008. Cross-sectional samples of 1,000 firms with 50 or more employees were randomly selected from the China, Hong Kong, and Taiwan listings in the Dun and Bradstreet (D&B) Global Million Dollar database. The surveys were addressed to the most senior executive (e.g., Chairman, CEO or CFO) named in the database. Each survey questionnaire was sent with a cover letter, self-addressed return envelope, and an offer for respondents to receive a summary of study findings (interested respondents were asked to send their business card in a separate envelope). Two to four weeks after the first mailing, a reminder mailing was sent. As a result, 98 China, 193 Hong Kong, and 175 Taiwan companies responded to the survey (after accounting for undeliverable surveys, the response rates were 14, 13, and 12.2 %, respectively). The sample characteristics for the three societies are presented in Table 2. Chi squared tests showed that the organizational characteristics (company size, ownership type, and industry) of our final sample were representative of companies in the D&B database.

Table 2 Characteristics of the society samples
Full size table

Questionnaire Development

The survey material was in Chinese but originally constructed in English. We employed standard translation and back-translation procedures (Brislin 1970) with two bilingual Chinese academics. To insure validity in a cross-cultural setting, we then provided the two versions of the survey questionnaire to a third Chinese academic to verify the content and format of the Chinese version.

Measures

Corporate Responsibility (CR) Practices

We developed customer, employee, investor, and community CR practice items relating to four groups of social stakeholders (e.g., Aupperle et al. 1985; Clarkson 1995; Maignan et al. 1999) and environmental CR practice items relating to proactive corporate environmental management (Branzei and Vertinsky 2002; Egri and Hornal 2002). Each of the 28 items measured the extent to which a CR practice has been adopted in a respondent’s organization using a nine-point Likert scale (1 = strongly disagree to 9 = strongly agree).

Financial Performance

The financial performance measure was adapted from Samiee and Roth (1992) which measured organizations’ return on investment, return on assets, market share, sales growth, and profit growth when compared with their most relevant competitors over the past 3 years. Responses to all the financial performance items used a nine-point Likert scale (1 = strongly disagree to 9 = strongly agree).

Measurement Model

We conducted confirmatory factor analyses (CFA) to assess the convergent and discriminant validity of the five CR practices and financial performance measures. We focused on model fit indices (CFI, NNFI, RMSEA) that are less systematically influenced by sample size and Type II errors for larger sample sizes (Cheung and Rensvold 2002; Marsh et al. 2004). For model fit comparisons, we used Cheung and Rensvold’s (2002) recommendation that a change in CFI < .010 indicates a nonsignificant difference in model fit.

The first set of CFAs was for the total sample with individual country samples counterweighted to be equal (see Table 3). The initial CFA model for the six factors (33 items) had an acceptable level of model fit [χ 2(480)  = 2907.02, CFI = .956, NNFI = .952, RMSEA = .092] but there were eight items that had low factor loadings or significant cross-loadings. The revised CFA model (6 factors, 25 items) had a slightly better model fit [χ 2(260)  = 1147.43, CFI = .962, NNFI = .956, RMSEA = .075, ΔCFI = + .006]. In respect to the convergent validity of the measurement model, the composite reliabilities (Raykov’s rho) ranged from .76 for investor CR to .97 for financial performance. Per Fornell and Larcker (1981), discriminant validity is indicated when the square root of the average variance extracted (AVE) for each factor is greater than the shared variance estimate between any two factors. The square roots of AVE for the six factors (range of .67–.84) were all greater than their respective paired correlations (range of r = .24–.67). In sum, the composite reliabilities and AVE indicate satisfactory construct reliability and validity as they exceed the commonly used thresholds of 0.60 and 0.50, respectively (Bagozzi and Yi 1988).

Table 3 Confirmatory factor analyses of the measurement model
Full size table

To assess the potential of common method variance-affecting results, we conducted a CFA with an additional unmeasured latent method common (ULMC) factor and a one-factor test (per Podsakoff et al. 2003). The ULMC model showed a nonsignificant change in fit (∆CFI = + .010) whereas the one-factor model had a significantly poorer fit (∆CFI = −.107). In sum, these CFA results indicate that common method variance was not a significant issue for these data.

Cross-National Convergent and Discriminant Validity

Multi-group CFAs were conducted to determine the cross-national validity of the measurement model across the three societies in this study (Steenkamp and Baumgartner 1998). As shown in Table 3, the unconstrained configural model (6 factors, 25 items) had an acceptable level of model fit [χ 2(784)  = 1727.22, CFI = .948, NNFI = .940, RMSEA = .087]. Nonsignificant changes in CFI were found for sequential nested models testing for metric invariance (ΔCFI = −.002), scalar invariance (∆CFI = −.008), factor variance and covariance invariance (∆CFI = −.002), and error variance invariance (∆CFI = −.006). In sum, the multi-group CFA results demonstrated the equivalence of the measurement model across the three societies.

Table 4 presents the society means, standard deviations, and scale composite reliabilities (Raykov’s ρ) for the CR practices and financial performance variables.Footnote 1 The convergent validity of the measures is further demonstrated by all scale reliabilities exceeding the 0.70 cutoff value (Nunnally 1978). The range of scale reliabilities were: customer CR (ρ = 0.76–0.88, 4 items), employee CR (ρ = 0.79–0.85, 5 items), investor CR (ρ = 0.71–0.82, 4 items), community CR (ρ = 0.79–0.91, 4 items), environmental CR (ρ = 0.84–0.86, 4 items), and financial performance (ρ = 0.90–0.92, 4 items). The list of measure items is presented in the Appendix.

Table 4 CR practices and financial performance: means, standard deviations, and scale reliabilities
Full size table

Data Analysis

In this study, we used a novel methodology, a set-theoretic approach (Ragin 2000, 2008a), to investigate the configurations of CR practices and their relationships with financial performance. The set-theoretic approach has been widely used in political science and sociology, and has been recently applied in management research (Crilly 2011; Fiss 2007, 2011; Kogut et al. 2004). This analytic method is particularly useful for conducting comparative studies since it represents a synthesis of qualitative and quantitative methods and is designed to study complex strategic combinations and configurations of constructs. The set-theoretic approach identifies nonlinear and asymmetric relationships by making sets and synthesizing cross-case patterns rather than disaggregating cases into variables. These complex relationships are difficult to examine by using traditional empirical techniques such as linear regression.

To investigate the relationship between CR practices and financial performance, we first recoded all the measures into the set membership of the target set represented by continuous values in the interval between 0 and 1 using the indirect method (Ragin 2008b). With 0.5 as the crossover point, this scheme may have an infinite number of set membership scores, where partitioning could be more fine-grained by using continuous sets (Fiss 2007). The indirect method first qualitatively groups cases into categories by the degree of set membership (from 0 to 1). We chose nine anchors to represent the membership of 0.007, 0.047, 0.119, 0.378, 0.500, 0.622, 0.881, 0.953, and 0.993 in the target set for each construct. Establishing the nine anchors for all the constructs requires the application of existing theoretical and substantive knowledge of the companies in the three societies. In this study, because the survey items by themselves reflect executives’ reports of which set or category they belong to with regard to the value of CR practices and financial performance, set membership anchors were primarily based on survey item scores within each society.

For example, we calculated the measures of customer-related practices by Hong Kong companies and grouped firms into different categories by following the rules proposed by Ragin (2008b) as follows: firms with values in the top 0.7 % of sample firms in Hong Kong had a membership of 0.993 in the target set of firms related to high customer practices. In contrast, firms which were in the bottom 0.7 % of sample firms in Hong Kong had a set membership of 0.007 in the target set of firms with high customer practices. Following similar procedures, we created the threshold values for Hong Kong firms with nine anchor memberships in the set of high customer practices at the interval of 0 and 1. We followed similar procedures for calculating set membership of high customer practices for the China and Taiwan samples. The only difference was that each society has its own anchors for grouping cases into categories based on its own data information to control for cross-society differences. Likewise, since all items for the other four CR practices and financial performance were measured on a nine-point scale, we used the same method to establish calibration anchors for all constructs.

After calibrating the data, we conducted analyses using fuzzy-set/qualitative comparative analysis (fs/QCA) software (Ragin 2006). Given the different sample sizes for the three societies, we set different frequency thresholds for cases to be included in a set or configuration of CR practices for each society rather than use a uniform threshold value. Because Hong Kong and Taiwan companies represent a relatively larger sample size (respectively, 193 and 175 companies) than China (98 companies), we chose frequency thresholds of six observations for Hong Kong and Taiwan firms and a threshold of four observations for Chinese firms.

We used consistency scores to evaluate the relations between antecedents and outcome variables. Consistency measures what proportion of a causal configuration is consistent with an outcome (Ragin 2006). For example, a consistency score indicates the proportion of firms that belong in a set of high business outcome and a high CR practice or CR practices. In this study, we used the minimum threshold value of 0.85 in order to restrict our results to cases of higher consistency by following normal practices (e.g., Ragin 2008a). Another QCA statistic is the coverage score which assesses what proportion of outcome is explained by a configuration of causal combinations (Ragin 2006). A coverage score is similar to the R 2 in regression models. In this study, a coverage score measures the proportion of variance in high performance firms that is explained by the implementation of a configuration of CR practices. We used the fuzzy-set-truth-table method with the fs/QCA software to calculate the membership scores of variables as well as the consistency and coverage scores (Ragin 2008a).

Results

The various QCA statistics for the specific and overall configurations of CR practices associated with high financial performance for each society are presented in Table 5. The QCA procedure generated two solutions across the three societies. Each solution covers the high or low value of individual CR practices and the resulting configurations. Following Ragin and Fiss (2009), we present both the parsimonious (indicated by large symbols) and complex (indicated by small symbols) configuration solutions in terms of presence/absence of high levels of a CR practice.Footnote 2 A blank entry indicates that a CR practice is either present or absent for the found configuration so it is not a differentiating factor (i.e., nonsignificant) in a solution.

Table 5 CR configurations for high financial performance: sets of firms and their consistency and coverage scores
Full size table

The two major QCA statistics of consistency and coverage for the overall and specific configurations of CR practices indicate strong support for their significant relationships with high financial performance across the three societies. In respect to the consistency of CR practices configuration and high financial performance, the overall solution consistency scores were all at or above the 0.85 threshold value (ranging from 0.85 to 0.88), and the unique consistency scores for each configuration ranged from 0.86 to 0.91. Coverage scores indicate the proportion of variance explained by the found configurations in terms of the overall solution coverage and the unique coverage for each configuration. Our analyses showed a high proportion of variance explained for overall solution coverage with scores ranging from 0.39 to 0.66 (i.e., 39–66 % of variance explained). The unique coverage scores indicate the individual contributions of each found configuration to explain variance in financial performance, and these scores ranged from 0.03 to 0.31 across the three societies.

Configurations of CR Practices and Financial Performance

In respect to Hypothesis 1 which proposed the emphasis on primary stakeholder CR practices in configurations leading to high financial performance, the results provided strong support across the three societies. Specifically, high levels of customer, employee, and investor CR practices were significantly related to high financial performance in both sets of configurations for firms in China (C1 and C2) and Hong Kong (C3 and C4). For firms in Taiwan, high levels of customer and employee CR practices were significantly related to high financial performance across both configurations (C5 and C6). However, inconsistent with H1, high levels of investor CR practices were significantly related in only one configuration (C5) with low levels of investor CR practices evident in the other configuration (C6) for firms in Taiwan.

As predicted, secondary stakeholder CR practices had inconsistent associations with high financial performance. For each society, one configuration for high financial performance included community CR being an irrelevant factor (i.e., nonsignificant) for Hong Kong (C3) and Taiwan (C5) firms and a negative factor (i.e., low level) for firms in China (C1) and Taiwan (C6). Similarly, the set of configurations for high financial performance also showed low levels of environment CR for both China (C2) and Taiwan (C6) firms and environment CR being an irrelevant factor for Hong Kong firms (C4).

Societal Differences Across CR Practices Configurations

Hypothesis 2 focused on cross-societal similarities (H2a) and differences (H2b) in configurations of CR practices for high financial performance. Focusing first on primary stakeholder CR practices, consistent with H2a, five of the six found configurations were similar in terms of high levels of primary stakeholder CR practices (customers, employees, and investors). Inconsistent with H2a, one configuration (C6) for Taiwan firms had a low level of investor CR practices associated with high financial performance. Also inconsistent with H2a, we found substantial differences across the three Chinese societies in respect to the relationships of community and environment CR practices with high financial performance. For firms in China, high financial performance was also associated with a low level of either community CR (as in C1) or environment CR (as in C2). In contrast, high financial performance for Hong Kong firms does not accord a low emphasis to either community or environment CR practices. Notably, Taiwan firms have two distinct CR practices configurations with one configuration (C6) being decidedly different with low levels of both community and environment CR practices. In sum, minimal support was found for H2a.

Partial support was found for H2b which proposed more configurational similarities between firms in Hong Kong and Taiwan compared to those in China. Consistent with H2b, the CR practices configurations for China firms were dissimilar (especially in respect to community and environment CR practices) from those found for Hong Kong and Taiwan firms. Also consistent with H2b, Hong Kong and Taiwan firms had one CR practices configuration for high financial performance in common (respectively, C3 and C5). Specifically, this configuration featured the presence of customer, employee, investor, and environment CR practices supplemented by the irrelevance of community CR practices. Nevertheless, the other two found configurations for these two societies (C4 and C6) do not lend support to H2b. In particular, high financial performance for Hong Kong firms can include high levels of investor and community CR practices, whereas low levels of these two CR practices (along with environment CR) can be associated with high financial performance for Taiwan firms. Furthermore, the found parsimonious solutions provide empirical evidence that core CR practices differ between these two societies. Specifically, customer CR practices are a core condition associated with high financial performance for Hong Kong firms, whereas employee CR practices are a core condition for Taiwan firms. In sum, only partial support was found for H2b.

Discussion and Conclusion

In this study, we investigated the prevalence of different CR practices and their relation to financial performance in China, Hong Kong, and Taiwan. By using a novel methodology, we found the presence of six configurations of CR practices that provided strong support for the instrumental role of different patterns of CR practices for engendering high financial performance. We also found similarities and differences in patterns of CR practices across these three Chinese societies which may be attributed to their institutional contexts.

Configurations of CR Practices

Our finding of the presence of different configurations of CR practices associated with positive financial performance is consistent with the literature exploring the impact of CR activities on financial performance (Barnett and Salomon 2006; Hillman and Keim 2001). This study lends support to proposals that CR practices may effectively enable firms to minimize risks associated with business operations, obtain legitimacy from different types of stakeholders, and differentiate them from competitors. More interestingly, we found that a balanced profile of CR practices contributes to high financial performance in that none of the six found configurations is comprised of a single stakeholder group. Although different core causal factors were identified across societal configurations (e.g., customer CR for Hong Kong firms, employee CR for Taiwan firms), our findings indicate that various CR practices need to be implemented simultaneously. In other words, it is the configurations of CR practices that count.

Consistent with the literature on stakeholder salience (e.g., Mitchell et al. 1997), we found variation in the contribution of different CR practices to financial performance. Specifically, high levels of stakeholder CR practices related to customers, employees, and investors jointly contribute to high financial performance for firms in all three societies. Further, our findings showed consensus across all configurations on the key role of customer and employee CR practices for high financial performance. Our study thus confirms the primary status of these two stakeholder groups in Chinese societies (Kolk et al. 2010; Xun 2012).

In contrast, we found considerable cross-societal variation in the contribution of community and environment CR practices to high financial performance. One key finding is that the alternative configurations for high financial performance for firms in China had either low levels of community CR or environmental CR. This suggests that in fast growing economies such as China, there is an ongoing trade-off in terms of these two types of corporate responsibility. This evident trade-off in CR program focus between local community development and environmental conservation may be explained by China’s relatively weak environmental regulatory context (Ma and Ortolano 2000; Marquis et al. 2011) coupled with a still emergent civil society sector (e.g., nongovernmental organizations) that has yet to play a strong role in terms of monitoring business’ social and environmental activities (Tang and Zhan 2008; Yang 2005).

We found two distinctly different configurations for business success for Taiwan firms. Whereas one configuration featured high levels of all but community CR practices which was not a differentiating factor, the other configuration featured high levels of customer and employee CR practices but low levels of both community and environment CR practices, in addition to low levels of investor CR practices. As such, there appears to be a clear divide within the Taiwan business sector in respect to one set of financially successful firms exhibiting high levels of discretionary corporate citizenship and another set of successful firms exhibiting low levels of discretionary corporate citizenship. However, we did not find a similar segmentation of CR practices for firms in the Hong Kong business sector. Instead, the two configurations for business success for Hong Kong firms feature high levels of investor CR practices coupled with either high levels of community CR or high levels of environment CR but no low levels of either. In combination with the findings for firms in China, one explanation is that very high levels of societal governance and regulatory environments that are very conducive for business operations (as is the case in Hong Kong) negate the presence of low levels of a CR practice as a pathway to business success.

Similarities and Differences in CR Practices Across Societies

Overall, we found mixed evidence for predictions regarding cross-societal consistency in patterns of CR practices configurations associated with high financial performance. Across the three societies, the closest to a convergent configuration for high financial performance consisted of high customer, employee, investor, and environment CR practices. Interestingly, the alternative configurations showed community CR to be a positive factor for firms in China and Hong Kong but a negative factor for firms in Taiwan. These mixed results regarding the financial benefits of community CR may be attributed to Chinese cultural factors. The cultural roots of community engagement or philanthropy across the three societies are related to the familialism-base within Confucianism (Ip 2009). It is more critical for private owners to protect their family wealth and heritage than to give back to the society. Therefore, the ‘love’ is more toward people with close blood-ties rather than others. Further, the fear of being frequently requested for more philanthropy from various stakeholder groups in local Chinese culture (i.e., , shùdàzhāofēng, a person in a high position is liable to be attacked) may also explain the inconsistent relationships between community CR practices and financial performance for the firms in this study.

Second, the parsimonious solutions suggest that employee CR practices are core causal conditions for high financial performance for Taiwan firms but not for firms in China and Hong Kong. One explanation could be that Taiwan’s government regulation of the labor market (labor legislation, industrial relations, and role of unions) is more restrictive and comprehensive than that in China and Hong Kong (Chow 2004). For instance, Taiwan’s Labor Standard Act specifically requires equal pay for men and women for equivalent work, and its Gender Equality in Employment Law in 2002 prohibits gender-based employment discrimination. Taiwan’s labor legislation provides that a union be organized in firms that have more than 30 workers, and the collective bargaining rights between unions and employers are clearly stipulated. In contrast, labor legislation in Hong Kong is much less formalized and centralized than in other industrial societies at a comparable stage of economic development. One of the key reasons relates to the nature of governmental regulation which promotes a self-regulating, laissez-faire economy with minimal governmental interference. Furthermore, Hong Kong’s lack of collective bargaining legislation has resulted in a fragmented union presence without significant influence on the guarantee of employee rights. While union presence is high in China, most unions are primarily concerned with the political and ideological education of employees. Although China implemented the new Labour Law in 2008, the lack of collective bargaining legislation has limited the power of unions to effectively bargain for improved working conditions for employees.

Limitations and Future Research Directions

While we sought to have matched samples across the three societies, one possible limitation of this study includes the representativeness of the results due to differences in firm characteristics rather than country or societal differences. The majority of our sample firms were private firms and small- and medium-sized enterprises. While such firms comprise the major part of national economies, different results may be obtained for samples primarily composed of large publicly traded firms. In addition, future research using larger samples as well as samples from additional countries is needed to test the generalizability of our findings.

In this study, we used a quantitative analytic approach for identifying CR configurations associated with business success. Further understanding of these findings would be advanced by within-configuration qualitative cases analyses that provide in-depth information regarding the logics and motivations of firms to implement various CR practices.

Implications and Research Contribution

Overall, this study provides support for the instrumental role of CR practices in China, Hong Kong, and Taiwan. One contribution was that examining CR practices from a configurational perspective directly addresses the question of how firms prioritize different stakeholders. Although numerous studies have considered the major differences across multiple stakeholders in terms of their power, urgency, and legitimacy (Mitchell et al. 1997), the set theoretic approach taken in this study provides a more integrative understanding of how companies rank and accord priority to various stakeholder groups.

As also found in previous research on the cross-societal standardization of management practices (e.g., Tempel and Walgenbach 2007), we did not find evidence of a clear convergence-divergence dichotomy. The complicated cross-societal patterns of CR configurations for business success indicate the co-presence of these CR practices in various forms across these three Chinese societies. Neither the new institutionalism nor national business system approaches provides complete explanations for our study findings. Instead, we propose that the interactions of these formal and informal institutions provide alternative rationales for understanding why some archetypes of CR practices may converge while others may diverge across societies (cf Fransen 2013).

Other authors have called for cross-societal CR research beyond the US and Europe (e.g., Matten and Moon 2008; Tempel and Walgenbach 2007). Hence, a contribution of this comparative study of CR practices in three East Asian societies is to examine the generalizability of existing CR theories to new societal contexts.

In sum, this study investigates CR practices and their impact on financial performance in the context of the three Chinese societies of China, Hong Kong, and Taiwan. Our examination of the role of different institutions to explain patterns of CR practices across these societies suggests a complex picture of the institutional rationale for adopting various CR practices.