INTRODUCTION
The past few decades have witnessed rapid growth in outward foreign direct investment (FDI) by emerging economy multinational enterprises (EMNEs). On the one hand, some studies suggest that EMNEs use international expansion as a springboard to overcome their late-comer status (Child & Rodrigues, Reference Child and Rodrigues2005; Luo & Tung, Reference Luo and Tung2007) by seeking strategic assets to augment their existing competencies (Buckley, Munjal, Enderwick, & Forsans, Reference Buckley, Munjal, Enderwick and Forsans2016; Cui, Meyer, & Hu, Reference Cui, Meyer and Hu2014). On the other hand, EMNEs also exploit their homegrown ownership advantages, such as expertise in mass production, low prices, and efficient utilization of resources in the international market (Buckley, Clegg, Cross, Liu, Voss, & Zheng, Reference Buckley, Clegg, Cross, Liu, Voss and Zheng2007; Luo & Tung, Reference Luo and Tung2007). In spite of the diversity of motives that drive international expansion, very little research has been conducted to examine the antecedents of these motives.
Ahsan, Sinha, and Srinivasan (Reference Ahsan, Sinha and Srinivasan2020) represent one of the few exceptions. They classify the international expansion motives of EMNEs along two dimensions – resource-based considerations and relative local advantages of host countries – and identify three types of motives – market-seeking, opportunity-seeking, and strategic asset-seeking. Based on Luo and Tung's (Reference Luo and Tung2018) springboard theory, factors representing the adaptability, amalgamation, and ambidexterity advantages of EMNEs are tested as antecedents of the diverse motives underlying EMNEs’ international expansion based on a sample of knowledge-intensive Indian firms. Their analysis contributes to the stream of literature on EMNEs by providing a distinct and holistic framework for understanding motives that drive international expansion.
We suggest that replicating Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) study can offer insights into the generalizability and heterogeneity of their results and novel insights into EMNEs’ international expansion by contextualizing the motivations they identify. Contextualization is particularly important for international business research (Foroudi, Gupta, Patel, Batsakis, Vaatanen, & Czinkota, Reference Foroudi, Gupta, Patel, Batsakis, Vaatanen and Czinkota2021; Kostova & Hult, Reference Kostova and Hult2016; Teagarden, Von Glinow, & Mellahi, Reference Teagarden, Von Glinow and Mellahi2018), as cross-country differences between economic, institutional, and sociocultural environments affect the motivations, behaviors, and performance of economic actors in significant ways (Jackson & Deeg, Reference Jackson and Deeg2019; Luo & Bu, Reference Luo and Bu2018; Vasudeva, Spencer, & Teegen, Reference Vasudeva, Spencer and Teegen2013). Empirically, we chose to replicate Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020) in the Chinese context because China and India are the two largest emerging economies, where comparable samples of firms with strong international aspirations can be constructed. Our replication study thereby joins a currently underdeveloped body of research that examines heterogeneity across emerging economies (Xu & Meyer, Reference Xu and Meyer2013) and the various strategic behaviors exhibited by EMNEs (Hu, Cui, & Aulakh, Reference Hu, Cui and Aulakh2019). Similarities and differences between Chinese and Indian firms can demonstrate important boundaries of Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) theoretical arguments and further inspire research regarding how specific contextual factors influence international expansion motives.
In the following sections, we explain our reasons for re-examining the motives behind international expansion on the part of Chinese firms, after which we present the methodology and results of the replication study. Next, we compare our results with those of Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020), summarize the implications for contextualizing EMNE research, and discuss limitations and future research suggestions. The final section concludes our study.
RE-EXAMINING THE MOTIVES BEHIND INTERNATIONAL EXPANSION IN CHINA
Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020) provide a new holistic framework that facilitates the classification of various motives that drive internationalization by EMNEs into market-seeking, opportunity-seeking, and strategic asset-seeking international expansion. Specifically, through market-seeking international expansion, firms exploit their specific advantages in other emerging or least-developed economies, as they seek the benefits of economies of scale (Buckley et al., Reference Buckley, Clegg, Cross, Liu, Voss and Zheng2007). Firms that exploit their specific advantages in developed economies are engaging in opportunity-seeking international expansion, as firms can benefit from the institutional environments in developed economies, which are marked by lower operational risk than they face in their home countries (Witt & Lewin, Reference Witt and Lewin2007). Finally, firms seeking to augment their assets through international expansion into developed economies often do so to obtain strategic assets, including technological know-how, distribution channels, managerial skills, and other assets (Luo & Tung, Reference Luo and Tung2007; Mathews, Reference Mathews2006). Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020) studied firm-level antecedents such as R&D investment, financial slack, ownership structure, and family control as determinants of the motives for market-seeking, opportunity-seeking, and strategic asset-seeking international expansion. We summarize Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) hypotheses in Table 1.
Table 1. Summary of Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020) and replication study results
Internationalization motives that are central to internationalization theory are complex and multi-dimensional (Benito, Reference Benito2015; Sethi, Guisinger, Phelan, & Berg, Reference Sethi, Guisinger, Phelan and Berg2003). Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) framework contributes to the literature on motives for internationalization by integrating the unique aspects of EMNEs based on both resource and location considerations. Therefore, it should prove useful to replicate Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) study in another emerging-country context to determine whether the conclusions based on Indian manufacturing firms remain the same. In this study, we employ the quasi-replication method (Bettis, Helfat, & Shaver, Reference Bettis, Helfat and Shaver2016) to assess the robustness and generalizability of Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) results. We chose China as the empirical setting for our replication study, as it is another important emerging economy with large outward FDI and active EMNEs. China's share in developing economies’ outward FDI in 2019 ranked first globally at 31.4%, followed by Hong Kong and Singapore (UNCTAD, 2020). China accounts for 8.9% of global FDI outflow, ranking as the fourth largest investor after Japan, the United States, and the Netherlands (UNCTAD, 2020). Moreover, it has been demonstrated that Chinese outward FDI positively influences host country economies (Donou-Adonsou & Lim, Reference Donou-Adonsou and Lim2018).
India and China share various similarities in terms of macroeconomic and microeconomic activities. At the macroeconomic level, both India and China have undertaken economic reforms in recent decades. In 1991, India initiated an economic liberalization initiative by implementing pro-market reforms and easing restrictions on private-sector participation (Kumar, Singh, Purkayastha, Popli & Gaur, Reference Kumar, Singh, Purkayastha, Popli and Gaur2020). Since the late 1970s, the Chinese government has provided a series of incentives for internationalization by Chinese firms, enabling the country to move from an ‘open up’ policy to a ‘go global’ policy (Buckley et al., Reference Buckley, Clegg, Cross, Liu, Voss and Zheng2007). These institutional transitions increased competitiveness in key industries and drove firms to pursue business activities abroad (Gaur, Kumar, & Singh, Reference Gaur, Kumar and Singh2014). Moreover, both China and India have large domestic markets that have enabled them to develop substantial local business activity without internationalization (Maksimov & Luo, Reference Maksimov and Luo2021). Last but not least, Duran, van Essen, Heugens, Kostova, and Peng (Reference Duran, van Essen, Heugens, Kostova and Peng2019) argued that both China and India are ranked highly among emerging economies regarding the extent to which an emerging economy can provide favorable institutional environments for family-owned firms. At the microeconomic level, Indian and Chinese EMNEs share some similarities with regard to their FDI preferences, as both invest in advanced countries to seek sophisticated technology and know-how as well as world-class brands and international legitimacy, and both prefer acquisitions and setting up wholly owned foreign subsidiaries (Deng, Reference Deng2009; Peng, Wang, & Jiang, Reference Peng, Wang and Jiang2008; Piscitello, Rabellotti, & Giada, Reference Piscitello, Rabellotti, Giada, Risberg, King and Meglio2015).
In addition to the abovementioned similarities, there are also clear differences between India and China. For example, the Chinese government's ‘go global’ policy, which was officially launched in 2000, has aggressively promoted internationalization by Chinese firms (Gaur, Ma, & Ding, Reference Gaur, Ma and Ding2018). Under this policy umbrella, concrete measures such as low-interest financing, favorable exchange rates, reduced taxation, and subsidies have been provided to facilitate expansion to overseas markets (Peng, Reference Peng2012; Ramamurti & Hillemann, Reference Ramamurti and Hillemann2018). In comparison, similar policies enacted by the Indian government have been less consistent and intensive (Nayyar, Reference Nayyar2008; Popli & Sinha, Reference Popli and Sinha2014).
With regard to corporate internationalization activities, De Beule and Duanmu (Reference De Beule and Duanmu2012) showed that, in the case of acquisitions, Chinese firms, especially in high-tech manufacturing industries, target technological assets. Thus, they seem to prefer acquiring firms that are located in developed countries. On the other hand, Indian firms seem to target markets that are less competitive, aiming to acquire firms in developing countries to exploit their competitive advantages. Luo, Sun, and Wang (Reference Luo, Sun and Wang2011) analyzed, among other characteristics, differences between Indian and Chinese firms in corporate governance. Their findings suggest that Chinese firms prefer going public, while Indian firms are family-centered. Family-owned Chinese firms value personal links, such as their hometowns or business ties (Graham & Lam, Reference Graham and Lam2003), while family-owned Indian firms are more keen on joining business groups to gain political bargaining power (Hu et al., Reference Hu, Cui and Aulakh2019).
Acknowledging similarities and differences between India and China are important when investigating whether contextual factors affect the motives that drive international expansion by Indian and Chinese firms. Understanding these similarities and differences can help us interpret the generalizability and heterogeneity of our results in comparison with Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) results. In other words, the contextual alteration our replication study incorporates enables us to recognize inherent limiting conditions and theoretical boundaries (Whetten, Reference Whetten1989). While working towards this objective, we also aim to offer methodological improvements regarding several issues we have identified as problematic in Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020), specifically the appropriateness of the empirical model and the interpretation of the results.
METHODS
Data Sources and Sample
To be consistent with Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) study, we select firms belonging to four high-tech manufacturing industries – chemical materials and chemicals; pharmaceuticals; automotive; and computer, communication, and other electronic equipment manufacturing – as the sample for this study, based on the 2012 industry classification of firms listed by the China Securities Regulatory Commission (China Security Regulatory Commission, 2012). Based on the availability of data, the observation period for our replication study runs from 2007 through 2016, which differs from the 2003–2013 sample period studied in Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020). After deleting affiliates of foreign firms and observations with missing data, we obtain a final dataset consisting of an unbalanced panel containing 759 firms with 3,503 firm-year observations.
Firm-level financial data are collected from the China Stock Market and Accounting Research (CSMAR) database, a widely used database covering listed companies in China. Data indicating international expansion by Chinese firms are collected from three sources – the CSMAR database, firms’ annual reports, and non-periodic announcements.[Footnote 3] Specifically, information regarding greenfield ventures, acquisitions, and joint ventures is collected directly from the CSMAR database, and information regarding alliances and technology acquisition is gathered by searching annual reports and announcements using keywords such as ‘strategic alliances/cooperation’, ‘technology development’, and ‘technology transfer’. A greenfield venture is a completely new organization formed without accessing local resources (Barkema & Vermeulen, Reference Barkema and Vermeulen1998; Meyer, Estrin, Bhaumik, & Peng, Reference Meyer, Estrin, Bhaumik and Peng2009). Therefore, in the overseas affiliated company database in CSMAR, we identify greenfield ventures by counting new overseas affiliates but excluding joint ventures, associated ventures, and subsidiaries acquired in mergers. After excluding expansions located in offshore financial centers, we identify 1,294 international expansions, which comprise 874 greenfield ventures, 25 joint ventures in foreign countries, and 107 joint ventures in China, along with 185 full and partial acquisitions, 76 alliances, and 27 technology acquisitions.[Footnote 4]
Classification of International Expansion Motives
We follow Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) method of classifying international expansion motives, which involves two steps. In the first step, we determine whether the primary objective of international expansion is asset exploitation or asset augmentation. Asset augmentation is a means of achieving or developing a competitive advantage by acquiring strategic assets, such as technology, marketing, and management expertise, that are available in a host country. Asset exploitation involves leveraging specific existing competitive advantages in a host country (Buckley et al., Reference Buckley, Munjal, Enderwick and Forsans2016; Kedia, Gaffney, & Clampit, Reference Kedia, Gaffney and Clampit2012; Makino, Lau, & Yeh, Reference Makino, Lau and Yeh2002).
We classify greenfield ventures under asset exploitation, while coding technology acquisition as asset augmentation. Information was retrieved from annual reports and announcements to classify the motives behind acquisitions and equity alliances. For instance, asset exploitation is classified manually according to keywords such as ‘entry’ and ‘foothold’. Asset augmentation is defined with reference to keywords related to acquiring assets, such as ‘technology’, ‘brand’, and ‘distribution channel’. Among the 132 joint ventures in our sample, however, 121 lack descriptions that identify specific motives. Theoretically, an EMNE uses a joint venture as an efficient organizational form to acquire assets from a local firm (Delios & Beamish, Reference Delios and Beamish1999; Hennart, Reference Hennart1988; Meyer et al., Reference Meyer, Estrin, Bhaumik and Peng2009). Therefore, from a general theoretical perspective, we classify the 121 joint ventures that lack such descriptions as engaging in direct asset augmentation.
In the second step, we follow Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020) to measure the relative differences between host and home countries in terms of strategic-factor-market development, with China representing the home country in our case. Scores for the quality of infrastructure (railroads, ports, air transport, electricity supplies, and telephone lines), quality of local educational systems, quality of scientific research institutions, and availability of scientists and engineers were taken from the 2015–2016 Global Competitiveness Report. Where such scores are not available, we assign scores carefully according to information provided by countries/regions.[Footnote 5]
The results obtained from these two steps enable us to classify international expansion motives into the three types. The form of international expansion that is driven by the asset augmentation motive and involves expansion into countries where strategic-factor-market development is ahead of the development in China is classified as strategic asset-seeking. Asset exploitation in countries where strategic-factor-market development is ahead of the development in China is classified as opportunity-seeking, while asset exploitation in countries where strategic-factor-market development lags the development in China is classified as market-seeking.
Models for Analysis and Dependent Variables
Consistent with Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020), we conduct two sets of analyses to test the hypotheses (see Table 1, where we list all the hypotheses tested by Ahsan et al., Reference Ahsan, Sinha and Srinivasan2020 and the present study). The first set of analyses tests hypotheses regarding direct relationships between antecedents and motives that drive international expansion (i.e., H1a, H1b, H2a, H2b, H3a, and H4a) with a negative binomial model at the firm-year level using the full panel dataset (759 firms over the 2007–2016 period, with 3,503 firm-year observations). The dependent variable is the total number of yearly expansions driven by various motives for international expansion by a given firm in each year in the sample period.
Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020) adopted the Heckman sample-selection model to analyze their panel data and discussed the model choice in their robustness analysis. In our replication study, however, several considerations have led us to choose an alternative model. First, the dependent variable is a count variable. In the second stage, the Heckman model is a linear model that could produce negative predicted values and does not restrict predicted values to integers. Hence, for binary, count, and ordinal responses, the Heckman method is not appropriate and may lead to incorrect conclusions (Miranda & Rabe-Hesketh, Reference Miranda and Rabe-Hesketh2006). Second, a likelihood ratio test rejected the null hypothesis that the data can be pooled (as seen in Table 5), indicating that a panel data model should be adopted. The Heckman model is not appropriate, however, for analyzing panel data, because it ignores within-firm correlations. Finally, Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020) excluded firms that pursue no international expansion in the second stage of the Heckman sample-selection model. We, however, can use the full sample that contains all the firms that have or have not engaged in international expansion in a negative binomial regression analysis. Therefore, we avoid the problem of sample-selection bias with our model. Given these considerations, we instead adopt a random-effects negative binomial panel data model to test the hypotheses.
The second set of analyses tests hypotheses regarding the probability of pursuing international expansion based on being driven by one motive rather than another (i.e., H1c, H2c, H3b, H3c, H4b, and H4c) at the international expansion level using a pooled sample of 1,294 cases of international expansion conducted by sample firms during the 2007–2016 period. Like Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020), we adopt a multinomial logit (M-logit) model to analyze the effects of the predictor variables on the likelihood of expanding internationally based on being driven by one motive rather than another. The dependent variable is a categorical variable with three values that correspond to the three international expansion motives. If the coefficient of an independent variable in the multinomial logit model is positive and statistically significant, the likelihood that a firm is driven by the motive is indicated by the dependent variable rather than the base motive increases. If the coefficient is negative and significant, the probability that a firm is driven by the motive is associated with the dependent variable rather than the base motive decreases. Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020) interpreted the coefficients derived with the multinomial logit model as if they reflect the strength of the relationships between the hypothesized antecedents and international expansion motives, which we think is inaccurate. Therefore, we reformulate and summarize the corresponding hypotheses, so that they indicate how changes in the predictors relate to the probabilities that one motive rather than another drives international expansion (the differences are reflected in the formulations of the hypotheses that can be found in Table 1).
Independent Variables and Control Variables
We follow the variable definitions in Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020) as closely as possible in our replication study. Because of differences in information disclosure requirements between China and India, however, Chinese data are not available for every variable used in Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020). We, therefore, replace these variables with the most similar variables for which the data are available in the Chinese context. In Table 2, we compare our measurements of the independent and control variables in the replication study with those of Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020).
Table 2. Variable measurement in Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020) and the replication study
Among the independent variables, R&D investment is measured by research and development (R&D) expenditures as a percentage of net sales in Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020), but we measure it as a percentage of total sales. Financial slack is represented by cash asset ratios in our replication study instead of by operating cash flows divided by net sales, as in Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) study. This is because data on operating cash flows are not available in the Chinese dataset. Because a firm's cash asset ratio is the current value of its marketable securities and cash divided by its current liabilities, which indicates the firm's ability to pay its short-term obligations, we adopt this variable to represent financial slack. Ownership concentration is measured by the Herfindahl index for the shareholdings of all shareholders in Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) study, while in this replication study the Herfindahl index is calculated based on the shareholdings of the top ten shareholders only. In Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) study, family control equals one if a family is the largest shareholder in a firm and zero otherwise. In this replication study, we adopt the definition of family firms as specified by the CSMAR, which defines family firms more broadly and from a controlling perspective.[Footnote 6]
Among the eight control variables, five are identical to variables included in Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) study: firm size, firm age, FII shareholding, financial leverage, and market structure. Regarding the other control variables, network ties is measured as the number of ties that a focal firm has through its directors with other listed firms included in Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) study, while, in our study, we calculate the number of concurrent positions that a focal firm's directors, supervisors, and executives occupy in other listed firms. In Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) study, marketing intensity is defined as marketing expenses as a percentage of sales. Because marketing expense data are not available for Chinese listed companies, we adopt the ratio of selling expenses to net sales instead. Selling expenses represent a broader category of expenses incurred in product sales, and they include expenses involved in advertising, packaging, insurance, exhibitions, and so on. Last but not least, the variable business group affiliation that is included in Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020) is not included in our replication study because relevant data are unavailable.
RESULTS
Table 3 shows the correlations between the variables and relevant descriptive statistics using the panel dataset (n = 3,503). The mean value of the opportunity-seeking motive is the highest among the three motive variables, followed by strategic asset-seeking and market-seeking. Table 4 shows the correlations between the independent variables and the descriptive statistics for the sample where each international expansion is an observation (n = 1,170).
Table 3. Descriptive statistics and correlation table for the negative binomial regression model
Table 4. Descriptive statistics and correlation table for the multinomial logit regression model
Notes: The dependent variable for the multinomial logit regression model is a categorical variable. Therefore, we do not report the correlation between it and the independent variables.
The regression results derived from the negative binomial model and multinomial logit model are presented in Tables 5 and 6, respectively, along with Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) results. In Table 5, we report Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) results derived from the negative binomial regressions with models 1, 3, and 5, and our replication results obtained with models 2, 4, and 6. Three dependent variables are constructed to capture the total number of new international expansions by firms in each year based on strategic asset-seeking (models 1 and 2), opportunity-seeking (models 3 and 4), and market-seeking motives (models 5 and 6). In Table 6, we report the results derived from Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) multinomial logit regressions with models 7, 9, and 11, and our replication results with models 8, 10, and 12. The dependent variable in models 7–10 uses the strategic asset-seeking motive as the baseline category, while the dependent variable in models 11 and 12 uses the opportunity-seeking motive as the baseline category. In this article, we chose the 10% significance level.
Table 5. Negative binomial regression results
Notes: aWald χ 2 values are significant at p-values = 0.000.
b χ 2 values of the LR test vs. pooled are significant at p-values = 0.000.
Table 6. Multinomial logit model results
Note: aWald χ 2 values are significant at p-values = 0.000.
The reported results support hypotheses 1a and 1b, as the effects of R&D investment on market-seeking international expansion derived from model 6 (β = 4.51, p = 0.014) and on opportunity-seeking derived from model 4 (β = 2.92, p = 0.000) are positive and statistically significant. Hypothesis 1c is also supported, as the effects of R&D investment on strategic asset-seeking international expansion derived from model 2 (β = 2.30, p = 0.084) are positive and marginally significant. Moreover, the coefficients of R&D investment derived from models 8 (β = 10.52, p = 0.000) and 10 (β = 7.24, p = 0.003) are both positive and significant. These results show that an increase in R&D investment increases the likelihood that firms choose opportunity-seeking and market-seeking international expansion over strategic asset-seeking international expansion. The positive and significant coefficient of financial slack derived from model 2 (β = 1.40, p = 0.017) supports hypothesis 2a, which proposes that financial slack is positively associated with strategic asset-seeking international expansion. Hypothesis 2b, which proposes a positive association of financial slack with opportunity-seeking international expansion, is not supported (β = 0.66, p = 0.110). The first statement of hypothesis 2c regarding a positive association of financial slack with market-seeking international expansion is not supported by model 6 (β = 0.23, p = 0.849). The clause in the second statement of hypothesis 2c that proposes that firms with greater financial slack are less likely to pursue market-seeking international expansion over opportunity-seeking international expansion is not supported by model 12 (β = −1.46, p = 0.150), as the coefficient is not significant. The clause in hypothesis 2c that proposes that firms with greater financial slack are less likely to pursue market-seeking international expansion than strategic asset-seeking international expansion is supported, as the coefficient derived from model 8 is negative and significant (β = −1.90, p = 0.083). In all, hypothesis 2c is partially supported.
The negative and significant coefficient of ownership concentration derived from model 2 (β = −2.39, p = 0.011) supports hypothesis 3a, indicating that ownership concentration has a negative impact on strategic asset-seeking international expansion. Hypothesis 3b is not supported, however, as the coefficient of ownership concentration derived from model 10 (β = −2.26, p = 0.007) is negative and significant, which means that firms that operate under more highly concentrated ownership are more likely to choose strategic asset-seeking international expansion than opportunity-seeking international expansion. The first statement of hypothesis 3c, which proposes that there is a negative effect of ownership concentration on market-seeking international expansion, is not supported either, as the coefficient derived from model 6 (β = −1.64, p = 0.253) is not significant. The clause in the second statement of hypothesis 3c that proposes that firms with high ownership concentration are more likely to choose market-seeking international expansion than strategic asset-seeking international expansion is not supported by model 8 (β = −2.98, p = 0.014). Indeed, the negative and significant results indicate that firms with more highly concentrated ownership are less likely to choose market-seeking international expansion than strategic asset-seeking international expansion. In addition, the clause in the second statement of hypothesis 3c that proposes that firms with high ownership concentration are more likely to choose market-seeking than opportunity-seeking international expansion is not supported either, as the coefficient derived from model 12 is not significant (β = −0.72, p = 0.552). Overall, hypothesis 3c is not supported.
Hypothesis 4a, regarding the inclination on the part of family-controlled firms (as opposed to non-family-controlled firms) to pursue market-seeking international expansion, is supported, with a positive and significant coefficient of family control derived from model 6 (β = 1.09, p = 0.002). Hypothesis 4b, which proposes that family-controlled firms are more likely to conduct market-seeking international expansion than opportunity-seeking international expansion, is not supported, as the coefficient derived from model 12 is not significant (β = 0.07, p = 0.802). Finally, the first statement of hypothesis 4c, which proposes that there is a positive effect of family control on strategic asset-seeking international expansion, is supported, as the coefficient derived from model 2 (β = 1.04, p = 0.000) is positive and significant. Moreover, the coefficient derived from model 10 (β = 0.38, p = 0.045) shows that family-controlled firms are less likely to choose strategic asset-seeking international expansion than opportunity-seeking international expansion, which lends support to the corresponding clause in the second statement of hypothesis 4c. However, the other clause in the second statement of hypothesis 4c, which proposes that family-controlled firms are less likely to conduct strategic asset-seeking international expansion than market-seeking international expansion, is not supported, as the coefficient of family control derived from model 8 is not significant (β = 0.45, p = 0.113). Overall, hypothesis 4c is partially supported.
In the discussion below, we compare the findings of our replication study with those reported in Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020) regarding the four main independent variables.
DISCUSSION
Comparing the Current Findings with Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) Findings
R&D investment
Both Chinese and Indian firms that invest more robustly in R&D are more likely to pursue market-seeking and opportunity-seeking than strategic asset-seeking international expansion. These results manifest the exploitative nature of market-seeking and opportunity-seeking international expansion. In line with Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) study, we found support for the argument that EMNEs possess specific technological ownership advantages that can be exploited in the overseas market.
The main difference between Chinese and Indian firms is that R&D investment has a significantly positive impact on strategic asset-seeking international expansion for Chinese firms but not for Indian firms. This result may reflect differences in absorptive capacity that differentiate Chinese from Indian firms. Because many strategic assets are patents and proprietary technologies (Cui et al., Reference Cui, Meyer and Hu2014; Luo & Tung, Reference Luo and Tung2007), EMNEs need sufficient absorptive capacity to make productive use of such assets. R&D investment plays an important role in building this capacity (Cohen & Levinthal, Reference Cohen and Levinthal1990). As seen in Table 3, the R&D intensity of Chinese EMNEs averages 5% of total sales, which is five times than that of an Indian firm's (see Ahsan et al., Reference Ahsan, Sinha and Srinivasan2020, Table 2).[Footnote 7] As Chinese firms are more R&D-intensive than Indian firms, we suspect that Chinese EMNEs, on average, feature stronger absorptive capacity than their Indian counterparts, enabling Chinese EMNEs to recognize, assimilate, and use external knowledge in international strategic factor markets more effectively. This greater efficacy in the pursuit of strategic asset-seeking may explain the positive association between R&D intensity and Chinese EMNEs’ motivation to engage in strategic asset-seeking international expansion.
Financial slack
Consistent with the evidence from India presented by Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020), financial slack is found to have no significant relationship with opportunity-seeking or market-seeking international expansion on the part of Chinese EMNEs. This finding suggests that, insofar as opportunity-seeking and market-seeking activities are associated with relatively low or moderate levels of risk, especially when compared with strategic asset-seeking (Figueira-de-Lemos, Johanson, & Vahlne, Reference Figueira-de-Lemos, Johanson and Vahlne2011; Johanson & Vahlne, Reference Johanson and Vahlne1977, Reference Johanson and Vahlne1990), maintaining a financial buffer is not essential to low-risk international expansion. We find, however, that in China financial slack is positively related to strategic asset-seeking international expansion, and firms with greater financial slack are more likely to pursue that forms international expansion than market-seeking international expansion. Because strategic asset-seeking international expansion involves higher risk and uncertainty than market-seeking international expansion (Figueira-de-Lemos et al., Reference Figueira-de-Lemos, Johanson and Vahlne2011; Johanson & Vahlne, Reference Johanson and Vahlne1977, Reference Johanson and Vahlne1990), financial slack is essential to Chinese firms seeking to alleviate the risk and uncertainty associated with strategic asset-seeking international expansion. This effect is not, however, supported in Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) study. This may reflect differences of scale and in patterns of strategic asset-seeking FDI conducted by Chinese and Indian firms. Chinese firms are reportedly aggressive in acquiring high-value foreign assets to accelerate their pursuit of global market leadership (Deng, Reference Deng2009; Huang and Sharif, Reference Huang and Sharif2016; Park & Roh, Reference Park and Roh2019; Peng, Reference Peng2012). Such an aggressive pattern of international expansion requires financial buffering. On the other hand, Indian firms appear to be more risk-averse than Chinese firms. Elango and Pattnaik (Reference Elango and Pattnaik2011) find, for example, that Indian firms engage in cross-border acquisitions serially, often acquiring targets of increasing value sequentially as part of a learning approach to managing risks ‘before the big leap’.
Ownership concentration
The concentration of ownership in Indian firms is negatively related only to strategic asset-seeking international expansion, and firms that operate under more highly concentrated ownership are more likely to choose market-seeking international expansion than strategic asset-seeking international expansion. Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020) explain this finding based on the premise that strategic asset-seeking international expansion is riskier than opportunity-seeking or market-seeking expansion (Figueira-de-Lemos et al., Reference Figueira-de-Lemos, Johanson and Vahlne2011; Johanson & Vahlne, Reference Johanson and Vahlne1977, Reference Johanson and Vahlne1990). They conclude that firms featuring more highly concentrated ownership are less likely to pursue risky international expansion. This finding is not replicated precisely in our analysis of Chinese firms. Instead, our analysis shows that ownership concentration is negatively associated with both strategic asset-seeking and opportunity-seeking international expansion. Among Chinese firms that engage in international expansion, though, we found that highly concentrated ownership makes pursuing strategic asset-seeking international expansion more likely than pursuing market-seeking or opportunity-seeking international expansion. These mixed findings suggest that ownership concentration is not necessarily associated with risk avoidance, at least not universally.
We suspect that there are at least two factors contributing to the complexity of the ownership concentration effect. It depends first on the type of concentrated ownership involved. State owners may have different risk preferences and time horizons than private owners. Compared with private owners, state owners may have a greater appetite for risk and are more likely to believe they can afford to pursue long-term objectives over short-term profits (Bruton, Peng, Ahlstrom, Stan, & Xu, Reference Bruton, Peng, Ahlstrom, Stan and Xu2015; Zhou, Gao, & Zhao, Reference Zhou, Gao and Zhao2017). The second factor concerns the relative risk profiles of the various motives that drive international expansion. Strategic asset-seeking may not always be riskier than other types of international expansion, depending on the institutional support provided to expanding firms.
Both ownership structure and risk vary significantly with home contexts. For instance, while state ownership is prevalent in China, family owners dominate Indian firms (Hu et al., Reference Hu, Cui and Aulakh2019). Also, governmental support for internationalization reduces the risk associated with strategic asset-seeking international expansion for Chinese firms (Hitt, Ahlstrom, Dacin, Levitas, & Svobodina, Reference Hitt, Ahlstrom, Dacin, Levitas and Svobodina2004; Yan, Zhu, Fan, & Kalfadellis, Reference Yan, Zhu, Fan and Kalfadellis2018). Through a series of policies, such as low-interest financing, favorable exchange rates, and reduced taxation, Chinese governments reduce the costs that Chinese firms incur when acquiring strategic assets in pursuit of international expansion (Buckley et al., Reference Buckley, Clegg, Cross, Liu, Voss and Zheng2007; Gu & Reed, Reference Gu and Reed2013; Peng, Reference Peng2012). The extent of such institutional support of Indian firms’ international expansion has been less substantial (Nayyar, Reference Nayyar2008; Popli & Sinha, Reference Popli and Sinha2014).
Family control
Family-controlled Chinese firms are more likely than non-family Chinese firms to initiate international expansion, no matter which of the three motives we include in our analysis is involved. By comparison, family-controlled Indian firms are more likely than non-family firms to pursue only opportunity-seeking and market-seeking international expansion. In both the Chinese and Indian contexts, family-controlled firms are more likely to choose opportunity-seeking international expansion than strategic asset-seeking expansion.
To summarize, the only difference between Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) and our findings pertaining to family firms is that family-controlled Chinese firms are more likely than non-family firms to pursue strategic asset-seeking international expansion, whereas family-controlled Indian firms are not more likely than non-family firms to do so. It is possible that this difference reflects differences in home institutional contexts. One critical institutional void that affects emerging markets is the lack of intellectual property protection and a fully functioning domestic strategic factor market. Family-controlled Chinese and Indian firms may, however, respond to this institutional void differently. Research has found that Chinese firms often use FDI to avoid home institutional constraints or deficiencies (Cui & Xu, Reference Cui and Xu2019; Xia, Ma, Lu, & Yiu, Reference Xia, Ma, Lu and Yiu2014). Family-controlled firms in China have relied traditionally on low-cost manufacturing and local network resources and therefore are in greater need of extricating themselves from local dependencies as they seek to upgrade their value-adding competencies (Erdener & Shapiro, Reference Erdener and Shapiro2005). In contrast, family-controlled firms in India are more likely to organize themselves into powerful business groups with political bargaining power vis-à-vis the government (Hu et al., Reference Hu, Cui and Aulakh2019). The group structure and political power allow these Indian family firms to exploit home institutional voids to an extent that weakens their motives to seek strategic assets overseas.
Implications for Contextualizing EMNE Research
Table 7 juxtaposes the findings of this replication study with the findings reported by Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020). It shows both the convergence and divergence of empirical evidence gathered from the Chinese and Indian contexts. The significant divergence between the contexts of these two large emerging economies demonstrates the need to contextualize EMNE research. It is clear that there are substantial heterogeneities across emerging economies (Hoskisson, Wright, Filatotchev, & Peng, Reference Hoskisson, Wright, Filatotchev and Peng2013; Ramanurti & Singh, Reference Ramanurti, Singh, Ramamuti and Singh2009; Ramamurti, Reference Ramamurti2012) that influence the drivers of internationalization motives.
Table 7. Comparison of empirical results reported by Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020) and the replication study pertaining to India and China, respectively
Specifically, we find that, in the Chinese and Indian contexts, several home-based heterogeneities are likely to shape EMNEs’ international expansion. The first involves the homegrown absorptive capacity of EMNEs. A threshold level of absorptive capacity may be required for R&D to have a significant impact on EMNEs’ efforts in seeking strategic assets overseas. In other words, EMNEs exhibit varying levels of technological readiness for international springboarding.
The second heterogeneity involves the risk tolerance of EMNEs, especially with regard to overseas acquisition. Chinese firms acquire overseas assets more aggressively, requiring considerable financial slack, whereas Indian firms acquire assets more incrementally and thus do not require much financial slack. These distinctive patterns at the country level are also reported in prior studies (e.g., Buckley et al., Reference Buckley, Clegg, Cross, Liu, Voss and Zheng2007), indicating country-level institutional isomorphism. We suspect that this isomorphism is largely mimetic, as firms follow their domestic peers’ approaches to risk.
The third heterogeneity involves home institutional environments, which manifest in cross-country differences in the dominant ownership models, institutional support for international expansion, and power dynamics that play out between firms and governments. Our findings echo prior studies such as Hu et al. (Reference Hu, Cui and Aulakh2019) in highlighting the differences between the Chinese and Indian institutional systems.
Limitations and Future Research
This study is subject to one methodological limitation. Because the multinomial logit model does not fit panel data, we treat the data as cross-sectional data, neglecting fixed effects arising from firm-level heterogeneity.
Nevertheless, this replication study highlights several important areas that are worth studying in future research. First, future research should further contextualize EMNE internationalization theory. In addition to changing the country context, future studies could also expand to cover other industries or time periods. Such replication studies would help researchers better characterize country, industry, and time effects on factors that determine international expansion motives. Second, we argue that a sufficient level of absorptive capacity is needed for firms to acquire strategic assets, and both dominant ownership type and home institutional support can influence firms’ international expansion choices. Future research designed to ascertain their effects and provide additional relevant empirical evidence is needed. Last but not least, future research should account for other aspects of host country characteristics. In Ahsan et al. (Reference Ahsan, Sinha and Srinivasan2020) and our replication study, the classification of motives for international expansion is based on resource considerations and the relative advantages of host and home countries. Host country advantages are measured by reference to strategic factor markets. Future research should extend this line of work to investigate how other aspects, such as institutional and cultural factors, affect factors that determine international expansion motives.
CONCLUSION
We replicate Ahsan et al.'s (Reference Ahsan, Sinha and Srinivasan2020) study of factors that determine motives for international expansion using a sample of Chinese high-tech manufacturing firms. Several similarities have been found between Chinese firms and Indian firms. For example, firms in both countries that invest in R&D more robustly are likely to exploit their existing ownership advantages in other countries. Financial slack is not associated with opportunity-seeking or market-seeking international expansion. Firms with highly concentrated ownership are less likely to pursue strategic asset-seeking international expansion in both countries. Family-owned firms are more likely than non-family firms to pursue opportunity-seeking and market-seeking international expansion.
In addition to the abovementioned similarities, our findings reveal two main differences between Chinese and Indian firms. First, we find that R&D investment, financial slack, and family control are positively associated with strategic asset-seeking only for Chinese firms, which may result from their greater absorptive capacity, more aggressive pattern of acquisitions when expanding internationally, and unique approach in responding to institutional voids when compared with Indian firms.
Second, the impact of ownership concentration differs greatly for Chinese firms and Indian firms. In India, ownership concentration is negatively associated with strategic asset-seeking international expansion and positively associated with opportunity-seeking international expansion, while in China ownership concentration is negatively associated with both strategic asset-seeking and opportunity-seeking international expansion. Moreover, Indian firms that operate under highly concentrated ownership are more likely to pursue market-seeking international expansion than strategic asset-seeking international expansion. On the other hand, Chinese firms with highly concentrated ownership are more likely to choose strategic asset-seeking international expansion than opportunity-seeking or market-seeking international expansion. We suggest that at least two factors contribute to this heterogeneous result. In China, state ownership makes undertaking risky and long-term objectives more affordable, providing an advantage that private firms in India lack. The other factor may lie in the differential institutional support provided by the Chinese and Indian governments. Governmental support for international expansion may help to reduce the risk associated with seeking strategic assets, making firms with highly concentrated ownership in China more likely to pursue strategic asset-seeking international expansion.
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are openly available in Open Science Framework at https://osf.io/fhq9n/
