Abstract
Exporting is a critically important strategy for firms to grow, yet research in this area has tended to ignore how firms can leverage resource-based capabilities to improve export performance. Building on the resource-based view and institutional theory, the authors develop a novel perspective to explain how a firm can improve export performance by aligning its export channel with its level of market orientation capabilities, contingent on the institutional distance between home and export markets. Using a unique database of Chinese exporters, the authors find that exporters with strong market orientation capabilities prefer hierarchical export channels, while those with weak market orientation capabilities prefer hybrid channels. The analysis also indicates that the institutional distance between China and the export market moderates this relation. Moreover, the authors find that aligning export channel choice with firm-level market orientation capabilities and institutional distance yields better export performance.
Keywords
International diversification through exporting provides firms with an opportunity to expand the market for goods and services and to improve performance (Aulakh & Kotabe, 1997; Campa & Guillen, 1999). Exporting research primarily focuses on three key areas: antecedents to exporting (Hitt, Tihanyi, Miller, & Connelly, 2006), export performance (Sousa, Martinez-Lopez, & Coelho, 2008), and export channel selection (Brouthers & Hennart, 2007). Export channel research chiefly relies on transaction cost theory and indicates that exporters normally choose between hierarchal and hybrid channel structures (Anderson & Gatignon, 1986; Campa & Guillen, 1999; Klein, Frazier, & Roth, 1990).
While transaction cost research provides valuable insights about how firms organize export operations to make them most efficient, these studies tend to suffer from a number of shortcomings. First, they ignore the value creation aspect of exporting. Firms export to increase sales and performance, yet transaction cost models of export channel choice do not consider how firms can create value by leveraging existing resources in export markets (Sharma & Erramilli, 2004). Second, the majority of past studies focus on channel choice, ignoring the performance consequences of that choice (Aulakh & Kotabe, 1997). Finally, with one exception (Campa & Guillen, 1999), export channel selection research does not take into account the institutional differences between countries; this research ignores the heterogeneity of export markets and how these differences can impact the effectiveness of different export channels, value creation, and export performance.
In this article, we contribute to the literature by examining all three of these issues. We add a resource-based view (RBV) to transaction cost export channel selection research and examine the value creation potential of leveraging market orientation (MO) capabilities through export channel choice. MO encompasses a set of processes and routines that encourages firms to generate, disseminate, and respond to information about customers, competitors, and the external environment (Kohli & Jaworski, 1990). We argue that learning capabilities like MO may be particularly valuable when firms export because they help these firms gain knowledge about the foreign market, identify required adjustments to products, and create a strategy that positions products in the foreign market to compete effectively against competitors while gaining legitimacy with consumers and government agencies. The RBV suggests that to garner value from capabilities like MO, firms must create the right organizational structure (Barney, Wright, & Ketchen, 2001); they must select an appropriate export channel. Hence, we bring all these strands of previous research together when we theorize and test the notion that firms that align export channel choice not only with transaction cost factors but also with the level of resource-based MO capabilities they possess will achieve higher export performance.
We also extend previous export channel choice research by exploring the impact of institutional distance on this critical decision. Institutional theory suggests that not all countries are alike (North, 1990) and that differences in institutional settings can have an impact on the value a firm can generate from resource-based advantages (Brouthers, Brouthers, & Werner, 2008; Meyer, Estrin, Bhaumik, & Peng, 2009). We develop theory and test the idea that the institutional distance between home and export markets will moderate the interrelationship between MO and export channel choice and that firms using export channels that take into consideration both MO capabilities and the moderating influence of institutional distance, as well as traditional transaction costs, will have higher export performance.
To test our novel hypotheses, we use a unique, hand-collected database of Chinese exporting companies. Our analysis indicates that firms possessing strong MO capabilities tend to use hierarchical export channels in order to garner greater value from this specific learning resource but that firms lacking strong MO capabilities find hybrid cooperative channels to be more effective when exporting. Our results also suggest that the institutional distance between home and export markets moderates this relation, strengthening the value of using hierarchical channels when firms possess strong MO capabilities or the need for partners when firms lack such capabilities. Most important, our findings indicate that firms choosing export channels that align not only with transaction costs but also with the level of MO capabilities they possess, contingent on the institutional distance, will create more successful export operations than firms whose export channels cannot be explained by these factors.
In sum, we make three important contributions to the literature. First, we add a resource-based perspective to a traditional transaction cost model of export channel choice, extending the focus from cost minimization to take into account value creation in export markets. Second, we consider the heterogeneity of export markets and how differences in institutional environments can influence resource-based value creation and export channel choice. Finally, we make an important advance by linking export channel selection to export performance, noting that the alignment of resource-based and institutional factors play an important role in contributing to superior export performance.
Theory and Hypotheses
While prior export channel research relies on transaction cost theory and has done a good job of determining the most efficient export channel for firms to use (Anderson & Gatignon, 1986; Klein et al., 1990; Shervani, Frazier, & Challagalla, 2007), it tends to ignore a number of important issues. First, these studies do not tie channel choice to export performance (Aulakh & Kotabe, 1997); they provide only a description of firm activities instead of a prescriptive solution. Second, transaction cost research focuses on channel efficiency, but some scholars (Autio, Sapienza, & Almeida, 2000) suggest that exporting may be driven at least in part by the firm’s efforts to create value in export markets by leveraging firm capabilities in organizational learning and information utilization.
Recent scholarship (Hult, Ketchen, & Slater, 2005; Morgan, Vorhies, & Mason, 2009) has highlighted the importance of MO capabilities. MO capabilities focus on learning from customers, competitors, and the external environment; processing that information internally; and utilizing it to achieve success (Kohli & Jaworski, 1990). It has been suggested that a market-oriented firm’s concentration on competition and customers enables it to coordinate the production and delivery of tailored offerings based on market intelligence, which improves customer satisfaction, loyalty, and organizational performance (Kirca, Jayachandran, & Bearden, 2005). Thus, MO capabilities may be particularly important within the context of exporting because these capabilities help firms learn about the foreign market and adjust strategies and products to conform to market demand, which should result in superior export performance.
Finally, prior research efforts tend to ignore the heterogeneity of target markets, even though every country has a unique institutional environment that defines the way firms conduct business and attract customers (Scott, 1995). Institutional differences between home and export markets can make it easier or more difficult for firms to harvest value from firm-specific resources like MO (Brouthers et al., 2008; Sirmon, Hitt, & Ireland, 2007). Given the need to account for country differences in the export channel selection process, including institutional aspects helps further inform our knowledge of exporting above simply transaction cost and resource-based perspectives.
Below, we attempt to address these issues. Building on previous export channel (Campa & Guillen, 1999; Klein et al., 1990), resource-based (Barney et al., 2001; Brouthers et al., 2008), and institutional research (Kostova & Zaheer, 1999; North, 1990), we theorize that the export channel a firm uses and the subsequent export performance it achieves not only will vary because of transaction cost factors but will be dependent on the strength of its resource-based MO capabilities and the institutional distance between home and export country.
RBV and Export Channel Choice
When exporting, the firm manufactures its products at home but needs to understand the foreign market to know how to position these products in that market. Knowledge of the market helps the firm identify changes in products that will lead to greater acceptance and sales. Firms with strong MO capabilities have the ability to tap information in the market (customer, competitor, and external environmental information), process that information internally, and use that information to respond effectively (Kohli & Jaworski, 1990). Such firms will want to use hierarchal export channels so that they can get the greatest value from these MO capabilities in the export market. In contrast, firms with weak MO capabilities do not have this ability to get information, process it, and use it. Instead, they may benefit from partnering with a target market firm that can provide information and advice. Below, we explain why.
Firms with strong MO capabilities will prefer to enter export markets using hierarchical export channels instead of teaming up with a local partner, for several reasons. Because strong MO firms are proficient at information generation, partner firms may simply offer information that replicates the information already captured by the focal firm, thus providing little if any additional benefit. In addition, information generated by a partner needs to be transferred to the focal firm, and transfer of knowledge across firm boundaries is not as efficient as transferring information internally (Barkema & Vermeulen, 1998). Thus, strong MO firms will prefer internalized export channels because they provide a more effective structure through which the firm can gain value from its resource-based learning capabilities.
Even if an external partner could help generate additional information, the added benefit of this information may be outweighed by the need to share the resulting rents and by the potential loss of control over MO capabilities. In hybrid export channels, firms share the rents generated by the resources employed (Anderson & Gatignon, 1986), thus reducing the return that a firm with strong MO capabilities could obtain. This means that the added value created by the additional information provided by a potential partner would have to more than double the rents generated to offset the costs of sharing.
In addition, the RBV maintains that control of capabilities, like MO, is important since losing control may lead to poor value creation and/or rapid imitation by competitors. Firms with strong MO want to control these capabilities to be sure they are obtaining reliable information about potential customers, competitors, and other external parties, and that the response to this information is properly implemented, so that they gain the greatest value from these resources. Sharing control of the venture may result in miscommunication or weak strategy implementation by a partner firm (Wu, Sinkovics, Cavusgil, & Roath, 2007). Further, if the firm sets up a joint operation with a partner, it has to share MO capabilities and facilitate training of local staff in how to use them. Ultimately, training partner organizations to take advantage of the firm’s MO may result in the creation of a competitor and reduce returns to the focal firm.
Therefore, firms with strong MO capabilities will use hierarchical channels of exporting. This structure allows such firms to retain all the rents generated from the MO capabilities and maintain control of the MO capabilities, and it provides a more efficient mechanism for transferring and using tacit target market knowledge within the organization.
In contrast, for firms possessing weak MO capabilities, hybrid export channels may provide benefits that outweigh the drawbacks. Partner firms can help generate information, and for weak MO firms this information would, for the most part, be new and additional (Wu et al., 2007). Although transferring the information from partner to focal firms may not be as efficient as obtaining the information directly from the market, any additional information that weak MO firms can obtain adds value to their foreign market entry. Further, while partner firms cannot disseminate and make changes in products based on the information they generate, they can provide suggestions to the focal firm about potential changes that might improve foreign market performance. We do not suggest that weak MO firms using hybrid channels will have the same level of knowledge about the foreign market as strong MO firms going it alone but that the added knowledge a partner provides improves the focal firm’s understanding of the export market (vs. not having this additional information source). Thus, because weak MO firms have few other choices when it comes to export market information, the information generated and transferred from a partner firm helps improve their understanding of that market and create responses that align better with export market needs.
Weak MO firms also are far less concerned with sharing rents and having control over MO capabilities. Without export market information, firms generate low rents. The added information provided by a partner firm will help weak MO firms gain valuable knowledge about the export market, knowledge that they could not generate on their own, which has a significant impact on target market performance far outweighing the costs of shared rents. Further, since weak MO firms do not have an MO resource-based advantage to protect and control, there is little chance of helping to create a potential future competitor by teaming up with a local partner organization.
In sum, a firm with weak MO capabilities may benefit from partnering with another firm that can generate export market knowledge and add value to the exporting operation. Yet for a firm with strong MO capabilities, the added benefit of expanding with a partner may be offset by the difficulties in transferring information across firm boundaries, the redundancy in the information obtained from the partner, the need to share the resulting rents, and the potential loss of control over the MO capabilities. Thus, our first hypothesis states:
Hypothesis 1: Firms with strong MO capabilities will use hierarchical export channels, while those with weak MO capabilities will use hybrid export channels.
Moderating Role of Institutional Differences
Our arguments above suggest that an exporting firm’s MO capabilities will have a significant impact on its export strategy. However, this relationship is far from being universal; export markets are, compared with domestic markets, far more culturally, politically, and geographically distant and diverse and are characterized by additional barriers such as a complex environment and problems with the availability, accessibility, and quality of export intelligence (Cadogan & Diamantopoulos, 1995). These country-specific attributes are often referred to as its institutional environment (Kostova & Zaheer, 1999; Scott, 1995). The institutional environment impacts how a firm does business, manages people, connects with customers, and interacts with the government (Kostova & Zaheer, 1999; Scott, 1995).
Previous research has begun to explore the direct impact of country-level institutional differences on internationalization strategies (Kostova & Zaheer, 1999; Xu & Shenkar, 2002), including equity entry mode strategies (Chan & Makino, 2007) and export channel choice strategies (Campa & Guillen, 1999). We extend this research by theorizing that differences in institutional settings not only have a direct impact on export channel choice but can moderate the relation between MO capabilities and export channel choice. More specifically, we suggest that a firm with strong MO capabilities will increase its preference for hierarchies when institutional distance increases for a number of reasons.
Strong MO firms have the capability to generate information about the new export market, and Sirmon et al. (2007) suggest that “learning is critical” when institutional environments are highly uncertain. As institutional differences increase, the firm will need to generate information about these differences that can be disseminated and used to develop effective responses (Chan & Makino, 2007). Yet differences in institutional norms mean that information that is important to the focal firm may be ignored by a partner simply because it falls outside accepted bounds (Amit & Schoemaker, 1993). Institutional distance can also create significant barriers that make interfirm communication of market information difficult or inaccurate (Kostova & Zaheer, 1999). Hierarchical export channels help firms with strong MO capabilities overcome these institutional differences by controlling the deployment of this resource to better maintain the credibility of the information generated. Hierarchical export channels also allow strong MO firms to disseminate export market knowledge quickly.
In addition, firms possessing strong MO capabilities have the ability to adjust their product offerings as they generate and process export market information. For these firms, reliance on external partners for information responsiveness activities may lead to suboptimal outcomes for several reasons. First, compared with the exporting firm, these partners may have a different understanding of the market information or different goals and objectives (Klein et al., 1990; Wu et al., 2007). Second, partners in institutionally distant locations are “institutionalized in the host country environment” (Xu & Shenkar, 2002: 613). This means that certain potential strategies may be viewed as “unacceptable,” and therefore implementation decisions are bound by the norms and regulations of accepted practice (Amit & Schoemaker, 1993). Yet the best response might fall outside these normal patterns and, in the case of hybrid channels of exporting, may result in delayed or incorrectly implemented responses.
Thus, as institutional distance increases, firms with strong MO capabilities will benefit from using hierarchical export channels. This channel helps exporters overcome differences in goals and objectives, norms, and institutionally bound practices, providing them with an ability to effectively generate, disseminate, and respond to the demands of the export market.
In contrast, if a firm does not compete using MO (i.e., it has weak MO capabilities), in institutionally distant markets it will profit even more from teaming up with a target market-based firm that possesses knowledge about the local institutional environment. This partnership can help the focal firm successfully navigate the foreign market; it can rely on the partner firm’s host country–based competitive advantage as a substitute for the lack of MO capabilities (Xu & Shenkar, 2002). For a firm with weak MO capabilities, partnering with a local organization, while potentially inefficient because of interfirm communication issues associated with high institutional distance, still provides knowledge that cannot be generated internally but might have a significant impact on the success of the export operation. In institutionally distant export markets, weak MO firms need market information even more urgently so that they can align product offerings with customer needs and competitive conditions (Chan & Makino, 2007). But these firms lack internal information dissemination capabilities; they may be successful only if they use hybrid export channels to partner with local firms that can provide such valuable knowledge (Chan & Makino, 2007). Even though this knowledge may be less than complete, it does help these firms overcome some barriers to success that institutional differences create.
Of course, internalizing the export operation through the use of hierarchical export channels may create another problem because firms will need to deal with the costs and complexities of managing issues arising from differences in external environments as well as internal issues that put pressure on firm subunits to conform to firm values and behaviors (Kostova & Roth, 2002; Rosenzweig & Singh, 1991). Past institutional distance studies tend to suggest that country pressures exert a stronger influence than pressures from inside the firm (Kostova & Roth, 2002; Rosenzweig & Singh, 1991). Yet firms must respond to both sets of pressures to be successful. Capabilities like MO help firms deal with these dual institutional pressures because these capabilities provide firms with the ability to learn about the target market’s institutional environment while maintaining internal consistency. As Rosenzweig and Singh (1991) note, changes in peripheral characteristics may be sufficient to adapt to external institutional differences while an organization’s core characteristics remain consistent with the parent firm. This suggests that internal pressures for firmwide conformity may not be a problem because these firms can change product packaging, color, size, and other peripheral characteristics to gain legitimacy in export markets but maintain internal consistency throughout the firm. In contrast, for firms with weak MO capabilities, hybrid export channels offer an opportunity to improve both the generation of export market information and the responsiveness to otherwise uncertain environments and can be used as a method to reduce internal pressures for conformity (Kostova & Roth, 2002; Rosenzweig & Singh, 1991). Hence, our second hypothesis suggests:
Hypothesis 2: The institutional distance between home and target country will positively moderate the relation between MO capabilities and export channel choice, such that the probability of firms with stronger MO capabilities selecting hierarchical channels increases as institutional differences increase and the probability of firms with weaker MO capabilities selecting hybrid channels increases as institutional differences increase.
MO, Institutional Distance, and Export Performance
As part of the RBV, the resource-structure-performance perspective (Barney et al., 2001; Brouthers et al., 2008; Ray, Barney, & Muhanna, 2004) suggests that the way a firm deploys its capabilities and resources has a significant impact on performance; firms that align their organizational structures with the capabilities/resources they hold will achieve superior performance. Alignment is defined as the fit or match between two related variables. The strategic fit literature specifies the operationalization of a contingency perspective, which follows the general axiom that no resource/capability or strategy is universally superior, irrespective of the organizational context (structure). Venkatraman (1989) notes that the “profile deviation” perspective of fit suggests that if an ideal strategy profile is specified for a particular condition, a firm’s degree of adherence to such a multidimensional profile will be positively related to performance if the firm has a high level of profile–condition alignment. Conversely, deviation from this profile implies a weakness in the alignment, resulting in a negative effect on performance.
However, when it comes to export channel choice, firms must make discreet decisions (Klein et al., 1990; Klein & Roth, 1990) such as establishing a wholly owned sales office or working with a partner. There is no continuum of export channels. In these discreet decisions, alignment is defined as the fit or match between two related variables: predicted and actual (Venkatraman, 1989). This concept of fit and performance is central to the RBV because organizational structures facilitate (impede) the exploitation or enhancement of firm-specific capabilities/resources (Barney et al., 2001). The implications for exporters are clear; the way MO capabilities are deployed when exporting may significantly influence export performance (Ray et al., 2004). Utilizing structures that facilitate information generation, dissemination, and responsiveness will enhance a firm’s ability to offer customers solutions that correspond better to their wants and needs and to offer something different from competitors, all of which should result in superior export performance (Morgan et al., 2009).
Further complicating the resource-structure-performance perspective is that for firms expanding aboard, simply aligning structures and capabilities may not be enough to create high-performing export operations because foreign markets may be institutionally distant; differences in regulative, normative, and cognitive environments can impact the value of capabilities (Priem & Butler, 2001; Sirmon et al., 2007). To achieve better performance in these situations, a firm needs to adapt the structure through which its capabilities are deployed to better align with both the capabilities it possesses and institutional demands (Brouthers et al., 2008; Sirmon et al., 2007). Thus, exporting firms might make different structure choices compared to those of domestic firms simply because of institutional differences that impact the value of firm-specific capabilities. Therefore, a firm expanding to export markets based on MO capabilities will need to consider how institutional differences influence the value of these capabilities and adjust its export channel structure choice accordingly. Failure to do so will result in misalignment and, as a consequence, a reduction in export performance.
To conclude, building on previous RBV and institutional theory research, we contend that the fit or alignment between a firm’s MO capabilities, institutional distance, and export channel selection may be a critical determinant of export performance. We theorize that firms with strong MO capabilities will benefit from internalizing the export structure and capturing the learning benefits that this resource can provide, especially as the institutional distance between home and export country increases. In contrast, a firm with weak MO capabilities targeting institutionally distant markets will generate greater export performance by teaming up with a partner organization that can provide some export market information, help it understand institutional differences, and aid the focal firm in responding in an appropriate manner. Thus, our final hypothesis states:
Hypothesis 3: Firms that align their export channels with the level of MO resources they possess, contingent on the institutional distance between home and host country, will have higher export performance.
Method
To test our hypotheses, we collected data from Chinese exporting firms. We selected China as the research laboratory for a number of reasons. China is one of the world’s biggest and fastest growing economies (Brouthers & Xu, 2002), the second largest international trade country, and the most important manufacturing location for firms with both local and foreign investments (Murray, Gao, Kotabe, & Zhou, 2007). Similar to firms from other emerging economies, Chinese firms lack other resource-based advantages, like brand names and superior technology (Hitt, Dacin, Levitas, Arregle, & Borza, 2000), and therefore have adopted a market-oriented approach (Murray et al., 2007). Thus, using Chinese firms provides a large sample of firms and helps us control for the impact of other resource-based factors.
Data were collected in 2008 using a random sample of 600 manufacturing firms (from a total population of about 7,300 firms) involved in exporting; the sample was drawn from the Exporting Firms Directory of Fujian Province, provided by the customs authorities. After multiple telephone calls and e-mails, 501 firms agreed to participate. Of the 99 firms excluded through this process, 21 could not be contacted because of incorrect contact details, 49 were export intermediaries, 22 refused to take part, and 7 had ceased exporting.
A three-page questionnaire with a cover letter and prepaid postage envelope was mailed to the CEOs of these 501 firms. Respondents were asked to provide information for their organizations’ most important export market (defined as the market in which the firm has its greatest sales). The initial mailing and two following waves of requests produced 285 responses. Of these responses, 90 were not usable (53 firms did not use a hierarchical or hybrid channel or reported using multiple channels; 37 others failed to complete the questionnaire). Thus, the usable data set comprised observations from 195 firms (38.9%) of which 116 (59.5 percent) used hierarchical channels and 79 (40.5 percent) employed hybrid channels.
To assess potential response bias, we followed the procedure outlined by Armstrong and Overton (1977). We compared early and late respondents and noted only one marginally significant difference in terms of number of employees (MO, t = −1.703, p = .855; export performance, t = 0.155, p = .463; R&D percentage, t = 1.533, p = 0.366; number of export markets, t = 1.105, p = .103; export experience, t = 0.274, p = .578; number of employees, t = 1.865, p = .067). Further tests comparing the same characteristics of usable and nonusable respondents and of the population and respondents found no significant differences on any of the items tested. Hence, response bias does not appear to be an issue.
Variables
Export channel choice analysis
For our dependent variable, export channel, respondents were asked to indicate which statement best represented the export channel they used in their most important export market. As in Klein and Roth’s (1990) study, hierarchical channels were assigned a value of 1 and included two types: “We have a wholly owned sales subsidiary in the foreign market” and “We serve the market directly from China, using company personnel.” Hybrid channels were assigned a value of 0 and included two types (Klein et al., 1990): “We are involved in a joint venture with another company to handle sales of this product in this market” and “We use commission agents.” Since our export channel selection variable is a binary variable, we used logit regression analysis to test Hypotheses 1 and 2.
We adopted an 11-item (7-point) measure of market orientation, rooted in the mainstream literature (Kohli, Jaworski, & Kumar, 1993) and shown to be reliable and valid in the Chinese context (Murray et al., 2007). This scale included 4 measures of information generation, 4 measures of information dissemination, and 3 measures of information responsiveness: the three component parts of MO (Kohli et al., 1993). The values for these 11 items were summed and averaged to create our MO construct (Cronbach’s α = .87). For all multi-item constructs, we used confirmatory factor analysis (CFA) to explore their reliabilities.
Berry, Guillen, and Zhou (2010) note that there are different definitions of the institutional environment; here, we build on the work of North (1990) and Scott (1995). The regulatory environment deals with the laws and regulations that promote or restrict the behavior of firms and consumers (Scott, 1995). We used the Economic Freedom Index (EFI) to measure the regulative environment because it captures institutional aspects of the environment that support (hinder) an efficient marketplace, thus impacting the ability of foreign firms to export to a particular market (Meyer et al., 2009). Other measures, like the World Competitiveness indices (Yiu & Makino, 2002) or Euromoney country risk (Brouthers et al., 2008), are useful but tend to capture aspects of the regulative environment that impact foreign direct investments, not exporting activities. Our construct regulative distance (Cronbach’s α = .81) is based on the EFI data published in the year that the firm began exporting into the target market. The construct includes 10 items: business freedom, trade freedom, fiscal freedom, government spending, monetary freedom, investment freedom, financial freedom, property rights protection, freedom from corruption, and labor freedom.
We employed the Mahalanobis method of calculating dyadic distances for all three institutional distance measures because this technique facilitates approaching distance as a construct made up of multiple, partially overlapping dimensions (see Berry et al., 2010). The Mahalanobis distance is estimated as the square root of the product of the remainder of the vectors representing the different dimensions of the institutional indicators, its transposition, and the inverse of the covariance matrix.
With regard to the normative environment, Scott (1995) notes that it specifies how things should be done, or the legitimate ways to pursue an end. This dimension encompasses the normative “management” aspect of the institutional environment that impacts exporting operations. Our normative distance variable (Cronbach’s α = .91) captures managerial attitudes and norms taken from the Global Competitiveness Report (Xu, Pan, & Beamish, 2004) in the year export activities began; it includes seven dimensions: efficacy of corporate boards, pay and productivity, capacity and innovation, degree of customer orientation, extent of staff training, reliance on professional management, and willingness to delegate authority. We believe other measures, like the World Competitiveness index (Gaur, Delios, & Singh, 2007) or World Value survey (Brouthers et al., 2008), approximate aspects of the normative environment that are not as informative for exporters.
As in previous studies (Gaur et al., 2007), our third institutional measure, cultural distance, was created using Hofstede’s (1980) four dimensions of national culture. Data for China and each export market were obtained from Hofstede’s website (http://www.geert-hofstede.com/geert_hofstede_resources.shtml). We then used the values for each of these four items (Cronbach’s α = .81) to calculate the Mahalanobis difference (Berry et al., 2010) between China and the export market to create our cultural distance measure.
Three moderating variables were calculated by centering the values of the MO and the regulative distance, normative distance, and cultural distance measures and then multiplying the MO value by the regulative distance measure, the normative distance measure, and the cultural distance measure (Aiken & West, 1991).
We included a number of control variables that have been linked to export channel choice in previous research (Brouthers & Xu, 2002). We began with transaction cost variables. We measured asset specificity in two ways: with a four-item scale (Cronbach’s α = .84) adapted from Shervani et al. (2007), called asset specificity, and by asking respondents to specify the percentage of R&D to total revenue in the past year, called R&D. Behavioral uncertainty was measured as exporting experience—the number of years the firm had been exporting—and with a single item (7-point scale) that we called internal uncertainty, which asked about the ease or difficulty of measuring the performance of individuals with whom the firm may cooperate. External uncertainty was measured with a four-item semantic differential scale (Cronbach’s α = .78) adapted from Shervani et al. (2007). Frequency (channel volume) was measured by the value for the firm’s annual exports to the most important market.
We included several firm and industry control variables. Firm size was operationalized as the number of employees in the firm. International experience was measured by the number of markets to which the firm exports. Market experience was measured as the number of years of experience in the target export market. Market size was captured by examining the gross domestic product (GDP) for the export market (similar results were obtained when using GDP per capita) obtained from the World Bank website. We created three ownership dummy variables for state-owned enterprises, foreign firms, and private firms. Each dummy takes the value of 1 if the firm’s ownership structure matches the variable and the value of 0 otherwise.
We created four dummy variables for firms representing the primary industries in our sample: domestic articles industry, electrical & electronic industry, clothing industry, and food industry, based on the Standard Industrial Classification of Chinese export commodities (MOFCOM, 2008). For each of these industry dummy variables, a value of 1 means the firm is in the industry, while a value of 0 indicates the firm is not in the specific industry.
Export performance analysis
Because there is no agreement on the best way to assess export performance (Sousa et al., 2008) and Chinese managers are unwilling to offer objective data (Brouthers & Xu, 2002), as in previous export studies (Sousa et al., 2008) we used subjective indicators to measure our dependent variable export performance. Respondents were asked to indicate (on a 7-point Likert-type scale) the level of satisfaction over the past three years with the following items: (1) overall export performance, (2) export sales growth, (3) export profitability, and (4) achievement of the firm’s initial strategic objectives. The values of these four items were then summed and averaged to create our export performance construct (Cronbach’s α = .91).
Our performance hypothesis suggests that alignment of export channel with MO resources and institutional distance is important. Following Venkatraman (1989) and others (Brouthers et al., 2008; Shaver, 1998), to test alignment we compared the predicted export channel (from our regression models) to the actual export channel each firm used. Thus, for each firm we calculated a predicted fit variable that took a value of 1 if the export channel used by the firm matched the export channel predicted by our regression model and took a value of 0 if the actual export channel did not match or align with the predicted export channel.
We included variables to control for firm size, R&D, exporting experience, international experience, and external uncertainty. Because export performance may vary simply because of ownership or industry differences (Brouthers & Xu, 2002), we also controlled for these factors. In addition, some research indicates that the export channel a firm uses is related to export performance (Bello & Gilliland, 1997). Therefore, we used a dummy variable, hierarchical mode, which took the value of 1 if the firm used a hierarchical channel but took a value of 0 if it used a hybrid channel.
Finally, Shaver (1998) suggests that strategy choices (like export channel selection) are endogenous and self-selected based on a firm’s own capabilities and industry conditions; ignoring this can lead to incorrect performance regression results. Heckman two-stage regression is a widely accepted method for dealing with this potential problem (Hult et al., 2008). In the Heckman procedure, an unobserved self-selection correction variable, also known as an inverse Mills ratio, was calculated from the estimated parameters of each selection equation on the basis of a set of probit regression analyses and added to our export performance regression models as a control for the possibility that unobserved firm attributes may impact export performance (Brouthers et al., 2008; Shaver, 1998).
Common Methods Variance
Podsakoff and colleagues (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003) provide researchers with two ways to control for common method biases (a) through the design of the study’s procedures and (b) through statistical controls. We utilized both methods in this study. The results of Harmon’s 1-factor test showed a 12-factor solution in which the largest factor explained only about 14.83% of the variance. In addition, the fit indexes for CFA—Tucker-Lewis index (TLI) = 0.10, comparative fit index (CFI) = 0.19, incremental fit index (IFI) = 0.21, and root mean square error of approximation (RMSEA) = 0.15—suggest a poor model fit. Both tests demonstrate that common method bias is not likely to explain any observed relation between model variables in our study.
Results
Before testing our hypotheses, we looked at the correlations between variables. Table 1 shows the means, standard deviations, and correlations for all our constructs. While we noted high variability in our constructs, there is no sign of multicolinearity. Further, we examined the variance inflation factors (VIF) in our regression tests and found that no VIF score was greater than 4.99, indicating a low probability of colinearity. On average, our respondent firms had almost 1,100 employees, had over 9 years export experience, and exported to over 11 different countries.
Means, Standard Deviations, and Correlations
Note: N = 195.
p < .05. **p < .01.
Export Channel Choice Results
Table 2 shows the six logit models we created to explore our first two hypotheses concerning export channel choice. Model 1 in Table 2 is our base model and was significant (p < .01). The transaction cost and other control variables explain about 39% of the variance in our dependent variable, export channel choice. Asset specificity (p < .01), market experience (p < .05), private ownership (p < .10), and foreign ownership (p < .05) are significantly related to the use of hierarchical channels, while domestic articles industry (p < .05), clothing industry (p < .05), and food industry (p < .01) are related to the use of hybrid channels.
Logistic Regression of Export Channel Choice
Note: N = 195. Hybrid channel = 0.
p < .10. **p < .05. ***p < .01 (based on Wald test).
In regression Model 2 (Table 2) we added our primary independent variable, MO capabilities. Model 2 was significant (p < .01), and we noted that the increase in explanatory power over Model 1 was also significant (p < .01). Model 2 explained about 44% of the variance in our dependent variable, export channel choice. In addition, the MO capabilities variable was significantly related to export channel (p < .01); firms with stronger MO capabilities tended to prefer hierarchical channels of exporting, as predicted in Hypothesis 1.
Models 3, 4, and 5 explore the interactions between MO capabilities and our three institutional distance measures separately, while Model 6 includes all interactions between MO and institutional distance proxies. All the interaction models were significant (p < .01); the increase in explanatory power over the MO capabilities model (Model 2) was also significant (p < .01). The interaction variable Regulative Institutional Distance × MO Capabilities was significant (p < .05) as were the interaction variables Normative Institutional Distance × MO Capabilities (p < .05) and Cultural Distance × MO Capabilities (p < .10). These interaction models independently explained about 47% of the variance in our dependent variable, export channel choice, and collectively explained about 49% of the variance (Model 6), providing support for Hypothesis 2.
Export Performance Results
Table 3 shows the results of our performance analysis. The base model (Model 1) contains transaction cost, other control variables, a Heckman self-selection correction term, and the transaction cost economics (TCE) predicted fit variable. TCE predicted fit takes the value of 1 if our export channel choice regression analysis in Table 2 (Model 1) correctly predicts the export channel used by the firm and takes the value of 0 otherwise. Performance Model 1 indicates that the transaction cost and control variables significantly (p < .01) explain export performance and that the TCE predicted fit variable was positive and significantly (p < .01) associated with export performance.
Regression Analysis of Export Performance
Note: N = 195. TCE = transaction cost economics; MO = market orientation.
p < .10. **p < .05. ***p < .01 (t values).
The MO model (Model 2) contains the same transaction cost and control variables, the Heckman self-selection term, and the TCE + MO predicted fit variable. This TCE + MO predicted fit variable takes the value of 1 if our regression analysis in Table 2 (Model 2) correctly predicts the export channel used by the firm and takes the value of 0 otherwise. Our results show that Model 2 was significant (p < .01) and that the TCE + MO predicted fit variable was significant and positively associated with export performance (p < .05); also, there was an increase in the regression R2 (from .21 to .23).
The final model in Table 3 looks at the moderating impact of institutional distance (all three—regulative, normative, and cultural distance) on MO export channel choice and performance. The TCE + MO/Institutional Distance predicted fit variable takes the value of 1 if our regression analysis (Model 6 in Table 2) correctly predicts the export channel used by the firm and takes the value of 0 otherwise. As Table 3 shows, Model 3 was significant (p < .01), and the TCE + MO/Institutional Distance predicted fit variable was significant and positively associated with export performance (p < .01); also, there was an increase in the regression R2 compared with the TCE model (from .21 to .24) and compared to the TCE + MO model (from .23 to .24). These results indicate that, as suggested in Hypothesis 3, firms that use the export channel that best fits with their level of MO advantage and the institutional distance between home and target country, on average, generate greater export performance compared with firms using export channels that do not align with these features.
As a robustness test, we used structural equation modeling in the AMOS statistical package to reestimate our performance model (Model 3 in Table 3) and found similar results—overall model fit: χ2(61) = 105.65, p < .00; IFI = 0.972, TLI = 0.92, CFI = 0.97, RMSEA = 0.06; R2 = .23. Further, we examined the economic significance of our results. Economically, firms that use export channels predicted by the TCE variables had an average performance of 4.68 (SD = 1.44). In contrast, firms that use export channels predicted by the TCE + MO/Institutional Distance variables had average performance of 4.98 (SD = 1.40). A t test indicated that these two groups are significantly different (t = 2.39, p = .02). This indicates that firms that consider MO capabilities and institutional variables as well as transaction cost variables when selecting an export channel have significantly higher export performance compared with firms whose export channels can be explained only by the traditional transaction cost model of export channel selection.
Discussion, Limitations, and Conclusion
In this article, we extend existing export channel research by adding insights from the RBV and institutional theory to explore strategies a firm can use to create value and improve export performance. Previous research suggests that foreign markets provide a business context different from the home context (Scott, 1995) and that firms venturing abroad may create greater value by leveraging learning resources (Autio et al., 2000). Building on this, we theorize and test the notion that resource-based MO capabilities may be particularly useful for exporters because MO capabilities enhance learning (Kohli & Jaworski, 1990). Drawing on a sample of 195 Chinese exporters, we find that, on average, a firm creates better performing export operations when it aligns the export channel with the level of MO capabilities it possesses, contingent on the institutional distance between home and export country.
Our study makes three important contributions to the literature. First, by adding a resource-based perspective to a traditional transaction cost model of export channel choice we extend the focus from cost minimization to take into account value creation in export markets. The transaction cost perspective that underpins the vast majority of international business and exporting studies assumes that firms will make export channel choices based on the efficiency of a particular export channel (e.g., Anderson & Gatignon, 1986; Klein et al., 1990; Shervani et al., 2007). We argue, from the RBV perspective, that this relationship does not always hold true. Firms with strong learning-based capabilities, like MO, may be more driven to use international expansion structures that help it capture value from these resource-based capabilities.
Our article highlights the unique characteristics of MO as a learning capability, in line with international entrepreneurship research that shifts emphasis from traditional firm-specific assets such as brands and technology to the firm’s ability to learn and utilize knowledge (Autio et al., 2000). Indeed, to expand sales abroad, the firm needs to learn not only about foreign markets but also how to use this knowledge and successfully adjust its product offerings. By focusing on MO, a unique learning capability, we theorize and find that firms possessing MO capabilities make very different export structure choices, resulting in superior value creation, compared with the choice that a traditional transaction cost perspective might suggest. More specifically, learning capabilities are more likely to lead to a hierarchical organization of exporting in institutionally distant markets, which changes one of the most commonly accepted arguments of the transaction cost perspective.
Second, this study makes an important contribution to the export strategy literature by addressing a significant criticism of the RBV: that it ignores the context in which firms operate, assuming that the resource-structure-performance interrelationship applies universally and is not influenced by contextual factors (Priem & Butler, 2001; Sirmon et al., 2007). We make a key contribution by developing a framework that takes into account institutional differences between countries and tailors the resource-structure-performance paradigm (Barney et al., 2001; Brouthers et al., 2008) to be applicable to the exporting activities of firms.
Within this framework, we theorize and test the idea that the value of MO learning capabilities, and hence the way a firm structures its export organization to take advantage of these capabilities, is dependent on the similarity or difference in the context where these capabilities are going to be used. The introduction of the institutional distance concept, we believe, provides a unique and interesting new perspective to exporting research. Our theoretical arguments and empirical results highlight the magnitude of this contribution by showing that consideration of institutional context factors can be usefully integrated with the resource-structure-performance perspective to provide a better explanation of firm performance.
Finally, we make an important contribution by linking export channel selection to export performance and noting that the alignment of resource-based and institutional factors play an important role in contributing to superior export performance. We advance the small but growing RBV literature that provides prescriptive models to explain how a firm can use its resources to achieve superior performance (e.g., Morgan et al., 2009). There is little understanding of how the possession of unique and valuable resources leads to better performance (Morgan et al., 2009; Priem & Butler, 2001). Our study provides a theoretical model to explain how firms that possess learning capabilities, like MO, can leverage these capabilities through export structures to achieve superior performance. Our empirical analysis suggests that, after taking into account institutional context differences, aligning the level of MO capabilities with the export channel has important performance implications. Thus, our study helps improve managerial understanding of actions that firms can take to garner value from the resource-based capabilities they possess to improve performance when expanding abroad.
Limitations and Conclusion
This study is subject to several theoretical and methodological limitations, which may offer additional research opportunities. First, we developed theory to explain how an important learning capability, MO, can influence export channel choice and performance. But we did not consider other resource-based advantages a firm might possess. Future research exploring other resources can explore this issue in more detail and add to our understanding of creating value in exporting. Second, we examined only Chinese exporting firms; therefore, our findings may not be generalizable to firms from other countries. Third, we obtained responses from only one person in each firm. Although we took precautions to avoid common methods bias in the data, using multiple informants or collecting data at two different times might improve our ability to detect any biases that do exist. Given the difficulties of collecting data in emerging markets, the single informant method seemed best, but future research can adopt a multi-informant approach to verify the accuracy of our results.
Fourth, we employed cross-sectional data rather than longitudinal data. Cross-sectional data are appropriate for exploring what is happening at a certain point in time. However, we cannot explain the dynamic processes of developing MO capabilities and export channel selection. There could be a learning process taking place where the firm can improve its MO capabilities through the use of specific export channels. Similarly, a firm that consistently used one specific export channel in the past may have developed a preference for this channel regardless of the institutional environment or level of MO capabilities. Future research may wish to use longitudinal data and explore how firms develop MO advantage, how it changes during the exporting process, and how past export channel selection impacts current choices.
Finally, we restricted our sample to firms that used only a single channel of exporting to their most important markets. Yet, as we found in our questionnaires, some firms use multiple exporting channels. It could be that over time additional export channels are added to these successful markets. To improve our understanding of the exporting activities of firms, future research might explore how the use of multiple exporting channels develops in exporting firms and how these multiple channels influence the ability of firms to exploit or supplement their resource-based advantages in foreign markets.
In conclusion, our study provides important extensions to past strategy research that explores the impact of resource-based advantages on firm performance. We develop new theory to explain how the RBV can be applied to exporting activities, an area of strategy that has received little attention from RBV scholars. Building on the resource-structure-performance concept and institutional theory, we develop a unique perspective to explain how a firm can harvest greater value from the learning capabilities it possesses when exporting to foreign markets. Our article provides a model to explain how learning capabilities can be converted into superior performance by considering both the context in which the capabilities are to be used and the structure through which they are deployed. Thus, we move resource-based and export strategy research forward by addressing the impact of context and learning capabilities on the deployment and performance of export operations.
Footnotes
Acknowledgements
We appreciate helpful comments from the editor, Michael Leiblein, and from two anonymous reviewers. We would also like to thank participants at the Harvard University China Goes Global 2009 Conference, the Academy of Management 2010 Annual Meeting, and the research seminar at Cass Business School, City University London, for helpful suggestions. We thank Yinyao Huang, Yangsheng Chen, and Haiying Li for their assistance with the initial data collection.