China's OFDI policy announcements and cross-border M&A

Abstract

Purpose

This paper aims to enhance the understanding of the role of Chinese outward foreign direct investment (OFDI) policies for cross-border merger and acquisition (M&A) by distinguishing between coercive and noncoercive OFDI policies.

Design/methodology/approach

The dependent variable is the count of completed M&A transactions, measured monthly. Due to the nature of the study’s data, the author performs a zero-inflated negative binomial (ZINB) regression.

Findings

Separating between coercive and noncoercive policies, the author finds that the latter type shows a stronger supportive effect on the count of M&A deals. Considering firm ownership, the study’s results reveal that announcements of coercive policies have a weaker effect on cross-border M&A for state-owned enterprises (SOEs) than that for private-owned enterprises (POEs). For local SOEs (LSOEs) and central SOEs (CSOEs), this difference becomes even larger with noncoercive policy announcements. The influence on M&A of both policy types gets partially replaced with increasing internationalization experience.

Originality/value

Combining institutional theory with policy change theory, the author argues that international business (IB) research on policy change needs to consider the integration of theoretical policy-level approaches to catch the effects of policy change on firm internationalization appropriately. The findings of the study support this argument by highlighting that the policy effect differs by policy type.

Citation

Utesch-Xiong, F. (2022), "China's OFDI policy announcements and cross-border M&A", International Journal of Emerging Markets, Vol. 17 No. 6, pp. 1522-1542. https://doi.org/10.1108/IJOEM-05-2020-0549

1. Introduction

The Chinese government aims to develop successful multinational enterprises (MNEs) and shifts the competitive advantage of the economy toward sectors that it finds to be of strategic importance (Buckley et al., 2018; Gammeltoft et al., 2010; Nem Singh and Ovadia, 2018), whereby the financial crisis helped China to gain a stronger hold in the world economy. Weakened developed countries gave Chinese MNEs the chance to acquire struggling firms abroad and gain missing knowledge. The foreign direct investment (FDI) openness of host countries was also existent in some economies, which later turned toward a more restrictive stance once Chinese outward foreign direct investment (OFDI) increased (c.f. U.S. Department of the Treasury, 2020). Often, as in the European Union (EU) or the USA, rules for screening inward FDI were laid out in times where South–North investment took place to a lesser extent and hence, they were not prepared for state-backed or subsidies benefiting investments (Hanemann and Huotari, 2015; Tingley et al., 2015). As a result of this unpreparedness, as well as the support of home country policies, Chinese firms were able to take advantage of the situation.

With the vast number of changes in China's OFDI policies, and with that the large-scale internationalization of Chinese MNEs (Luo et al., 2010), also international business (IB) research studying home country policies increased (e.g. Angulo-Ruiz et al., 2019; Chen et al., 2016; GaoYan, 2020). However, we find that three areas of OFDI policy research require extensions: first, despite the recent shift in the focus of IB research, the level of analysis is often on institutions than policies (Hong et al., 2015; Li and Ding, 2017). However, the study of policy influence can hardly be explained by only using institutional theory and explaining the role of policy on market actors is also the task of (IB) policy scholars (Clegg, 2019). Second, we find that when IB research distinguishes between OFDI policies, it does so only based on the direction of policy influence, i.e. supporting or restraining (e.g. Lu et al., 2014; Yang and Stoltenberg, 2014), without taking into consideration different types of policies (e.g. coercive and noncoercive; Cushman, 1937, 1941, p. 3, as cited in Lowi, 2016). However, contrasting characteristics of policies might lead to different pressures on cross-border merger and acquisition (M&A) (c.f. Rottig and de Oliveira, 2019), which cannot be controlled for when merely distinguishing between supportive and restraining policies. Third, IB studies on OFDI policies often cover only one specific policy or its change across time (e.g. Lu et al., 2014; Wang et al., 2012b). Current studies do not compare policies or even policies of different types across time. Thus, we are not able to define the role of home country-supportive policies on cross-border M&A on a broader level.

We aim to address these issues in our article by stating the following research question: how do OFDI-supportive policy types differ in their effect on cross-border M&A? Thereby, we follow the calls to shed more light on the differences between the types of OFDI-supportive home government inventions (Lu et al., 2014), calls for research on emerging economy multinational enterprise (EMNE) responses to OFDI-related policies (Luo et al., 2010) and more IB policy research (Buckley et al., 2018). We merge the theoretical approaches from the policy change literature (e.g. Béland, 2007; Lowi, 1985; Streeck and Thelen, 2005) with institutional theory (DiMaggio and Powell, 1983; Scott, 2014) due to the (long-term) interrelationship between changes in policies and institutions. Combining both theory streams, lets us expand the IB policy literature and introduce a theoretical avenue for future policy research, below the institutional level. We distinguish OFDI-supportive policies between the coercive and noncoercive type. To the best of our knowledge, we are the first to do so when studying OFDI policies. Our results reveal that it is important to distinguish between these two types as the response of firms through cross-border M&A differs. Studying the moderating effects of government ties and M&A experience on the relationship between home country policies and OFDI, we contribute to the IB literature (Cui, 2016; Hong et al., 2015; Luo et al., 2010) by not only covering a longer period and focusing on more than one selected policy, but we also distinguish between policy types. We find that for both the government ties and M&A experience moderators, results differ between coercive and noncoercive policies.

We start this article by introducing the theoretical framework merging institutional theory and theoretical literature on policy change. We then present our hypotheses on the influence of OFDI policies on cross-border M&A, as well as the methodology. The results of our zero-inflated negative binomial (ZINB) regression and the post hoc analysis follow hereafter. We conclude with a discussion of our findings and highlighting the contributions of this paper.

2. Theoretical framework

2.1 Institutions and policies

As highlighted by Buckley et al. (2018), IB has been unclear about the usage of the terminology covering institutions and policies. The ongoing discussion in political science (e.g. Béland, 2007; Streeck and Thelen, 2005, p. 12) shows that IB scholars might be able to make use of the developed definitions from that field. Streeck and Thelen (2005, p. 12) developed the following distinction between policies and institutions: “Policies, that is to say, are institutions in our sense to the extent that they constitute rules for actors other than for the policymakers themselves—rules that can and need to be implemented and that are legitimate in that they will if necessary be enforced by agents acting on behalf of the society as a whole.” Thus, they make a distinction in terms of actors being able to claim for their rights. A similar definition is listed in the Oxford English Dictionary (Hornby et al., 2010), which defines a policy as “a principle or course of action” and an institution as “an established law, custom.” Institutions influence when and how policies are implemented (Greif, 2006, p. 380; Lowi, 1985), i.e. the form and characteristics of policies, which shows that institutions hold an essential part in the policy change process. An example from the Chinese context is on the one hand, the development of the publication of the new Small and Medium Enterprises (SME) Promotion law (NPC, 2017) which sets the broad legal framework and on the other hand, the subsequent issuance of the Development Plan for SME Promotion (2016–2020) that defines specifically how SMEs can receive government support (MIIT, 2016). This relationship of the announcement of amendments to existing laws and the chain reaction of changes in policies reveals that we cannot leave the institutional literature out of the discussion completely, when studying the policy context. Hence, we propose a combination of the policy change and institutional literature, to put policy change into a theoretical context.

Where institutional research on legitimacy pressures (DiMaggio and Powell, 1983) distinguishes between three institutional pillars (Scott, 2014) and the accompanying differences in legitimacy pressures for firms (coercive, normative, mimetic) in policy research, differences have been made between coercive and noncoercive (normative) policy types (Cushman, 1937, 1941, p. 3, as cited in Lowi, 2016) and the pressures thereof. As stated by Lawrence and Suddaby (2006), policymakers equip policies with a mechanism of pressure, to be able to maintain institutions. For firms targeted by policies, these pressures are similar to those of institutions as the policy implementations can also lead to legitimacy pressures on firms, especially in an authoritarian home country context (c.f. Rottig and de Oliveira, 2019). We use the institution–policy relationship as our theoretical foundation to study the home country policy pressures on firm internationalization (Table 1).

2.2 OFDI-supportive policy types

The policy typology (i.e. coercive or noncoercive) does not represent an OFDI-restrictive or supportive position, both might restrict or support OFDI. Policies of the coercive type include formalized measures that articulate the benefits of the policy, next to the requirements that firms need to align to, to be able to take advantage of this policy. For example, if a Chinese firm wants to invest abroad, they are required to go through a registration or approval process (e.g. NDRC, 2014). Also, in comparison to noncoercive policies, coercive ones sometimes even highlight punishment schemes for misbehavior (MOFCOM and MOF, 2013). Noncoercive OFDI policies with their less specific articulation, using soft guidance measures, have no such “gatekeeper” power. They often include expressions of opinions (e.g. State Council, 2010, 2015), listings of industries abroad that are found suitable for home country development and thus, firms investing in these areas abroad shall be supported (e.g. NDRC, 2011) or the outline of strategy plans considering the support for OFDI (e.g. NDRC, 2012; NDRC et al., 2015) – all without defining specific approaches on how firms will be supported in their international endeavors. Hence, due to their type, coercive policies carry the pressure for firms to follow the announced policy (change) similar to the coercive institutional pressure, whereas for noncoercive policies, the form of pressure is related to normative institutions.

3. Hypotheses development

3.1 Home country OFDI-supportive policies

IB literature highlights that OFDI-supportive state interventions influence firm internationalization. For example, Zhang and van den Bulcke (1996) argue that Chinese firms using mostly M&A as a form of internationalization are more influenced by home country policies than those with greenfield OFDI. This finding found support by Du and Zhang (2018) on Chinese M&A in developing countries, as well as Buckley et al. (2016) on developed economies. Furthermore, Buckley et al. (2007) studying China's OFDI policy liberalization with a binary dummy for the year of Deng Xiaoping's South China tour find that OFDI to developed countries increased afterward. Also, Lu et al. (2014) in their study on the decision of Chinese firms to internationalize and Lu et al. (2011) analyzing the motives for OFDI, both, found that home country policies lead to more OFDI. Even though these findings are not specific to Chinese M&A but include all forms of internationalization, they show more broadly the change in the attitude of Chinese firms on OFDI through policy liberalization.

When studying policy change empirically, studies analyzed the implementation of one specific policy or the change of one type of policy over time (Du and Zhang, 2018; Lu et al., 2014; Wang et al., 2012b). However, they did not compare different OFDI-supportive policies of the coercive and noncoercive type. To be able to build a broader argument about the role of supportive policies on Chinese OFDI, we analyze the joint effect of the issuance of OFDI-supportive policies.

H1.

The announcement of OFDI-supportive policies leads to more cross-border M&A.

3.2 Differences in policy types

As in our study with nationwide introduced OFDI policies, the counterfactual situation and thus, control groups are missing, it cannot be analyzed how MNEs would have internationalized without the policy announcement or what the difference-in-difference effect is between groups (Blundell and Dias, 2009). However, as the effect of policy implementation on firm behavior differs between policy types (Prakash and Kollman, 2004), a comparison study seems appropriate. Considering coercive and noncoercive OFDI-supportive policies, one might argue that MNEs react stronger to the coercive type as the probability of benefitting from the issued policies is higher (c.f. Torres and Clegg, 2014). Where for coercive policies, the positive effect of the policy change comes naturally by following the newly defined guidelines, with changes in noncoercive-supportive policies, the firms' benefits might not be clearly defined. Where the literature studied the effect of OFDI-supportive policies on OFDI (see the previous section), they did not distinguish – at least not empirically – between the types of OFDI-supportive policies.

H2.

The effect of OFDI policies on cross-border M&A is more positive for coercive policies than for noncoercive policies.

3.3 Policy announcements and ownership

IB scholars have studied the institutional pressures of home country policies and the process of gaining legitimacy in the context of Chinese MNEs and their distance to the government (e.g. Cui and Jiang, 2012; Hong et al., 2015; Peng et al., 2004). As the literature shows (Li et al., 2014, 2018b), state-owned enterprises (SOEs) closer to the central government (e.g. central state-owned enterprises [CSOEs]) are equipped with different tasks than those being further away (e.g. local state-owned enterprises [LSOEs]), and hence, they are under different legitimacy pressure. As argued by these groups of authors, the legitimacy pressures of CSOEs are largely to act as a policy tool of the home government. As a firm with close ties to the home government, it can thus receive special financial support from governmental agencies (Holtbrügge, 2018) to fulfill the defined aims of the government, e.g. from the State-owned Assets Supervision and Administration Commission (SASAC). This support is not accessible to other firms (Cui and Jiang, 2012; Li et al., 2014; Luo et al., 2010; Sutherland, 2009) [1].

With decreasing influence of the government on the firm's management and increased focus on profit maximization, specific firm support for their OFDI endeavors becomes less likely (e.g. Liang et al., 2017; Voss et al., 2010). As a result, the role of each OFDI-supportive policy is perceived differently between the ownership levels (Cui, 2016) and the dependence on OFDI-supportive policies becomes stronger (weaker), the less (more) the firm can rely on the government to receive government support in addition to the OFDI-supportive policies. Firms with stronger home government ties do not rely on the announcements of noncoercive policies, i.e. policies for which the support mechanism is not formally articulated. Taking benefit of coercive policies does not include this uncertainty. The costs of following a defined set of rules, as included in coercive policies, are low for firms with close government ties due to their experience in going through state administrative procedures (e.g. through chief executive officers [CEOs] previously holding positions in the government) (c.f. Li et al., 2018a; Peng et al., 2004). For private-owned enterprises (POEs), which in opposite to SOEs and especially CSOEs rely mainly on the market economy and do not enjoy the same safety net as firms with closer state connections, noncoercive policy announcements might reveal an opportunity to shine in front of its home government and to gain support in its cross-border investment. Benefitting from coercive policies, in comparison to state-owned or state-controlled firms comes at a larger cost as firms with weaker government ties do usually not own the same experience in handling state administrative procedures (c.f. Li et al., 2018a; Peng et al., 2004).

H3a.

The positive effect of coercive policies on cross-border M&A is greater for firms with higher levels of government ties.

H3b.

The positive effect of noncoercive policies on cross-border M&A is greater for firms with weaker levels of government ties.

3.4 Policy announcements and M&A experience

The literature argues that with larger internationalization experience, MNEs and their home country governments get to know the benefits of one another more profoundly (Luo et al., 2010). Furthermore, studies have also revealed that if firms make use of OFDI-supportive policies, their internationalization experience seems to no longer be seen as crucial for the decision of an additional M&A (e.g. Lu et al., 2014). However, IB research did so far not focus on the substitution effect from the other side of this relationship, i.e. if with larger M&A experience, firms rely less on home country OFDI-supportive policies. Studying this relationship might provide insights into how much or if at all the costs related to the exploitation of coercive and noncoercive OFDI policies (freedom to operate – coercive policies; uncertainty to gain benefit – noncoercive policies) might be a burden for Chinese firms to conduct cross-border M&A and if M&A experience might be able to replace it. As the costs related to both policy types differ, we study the M&A experience separated between coercive and noncoercive policies.

H4a.

Internationalization experience weakens the positive relationship between coercive policy announcements and cross-border M&A.

H4b.

Internationalization experience weakens the positive relationship between noncoercive policy announcements and cross-border M&A.

Summarized, we hypothesize that cross-border M&A is influenced by coercive and noncoercive OFDI policies differently, whereby the level of influence depends upon the level of government control and the M&A experience of the firm. The conceptual model builds the basis whereon we will conduct our empirical analysis, as shown in Figure 1.

4. Methodology

4.1 Sample and data

To capture China's OFDI period of gaining a stronger hold in the world economy (2008: 3% of worldwide OFDI; 2015: 9%; UNCTAD, 2020a), we focus on the years 2008–2015. We used two M&A databases (SDC Platinum [SDC], Bureau van Dijk Zephyr [Zephyr) to search for Chinese cross-border transactions with announcement dates between 2008 and 2015 and acquisition shares of at least 10% (UNCTAD, 2007). We defined the acquirers' parents as being one level below the holding company, respectively, one level below any entity with solely managerial tasks. This approach is in line with the approach followed in the statistical department of the United Nations Conference on Trade and Development (UNCTAD) for the World Investment Report (Trentini, Skype-interview, July 05, 2018). Excluding duplicates, transactions with undisclosed acquirers, acquirers being individuals and transactions with target firms located in Hong Kong or Macao, we were left with 986 M&A deals.

The distribution between ownership types is shown in Figure 2. Overall, we include 570 Chinese parent companies, with POEs counting for the majority of M&A transactions (523), followed by CSOEs (170), LSOEs (144) and other SOEs (79). POEs' annual number of M&A deals climbed almost continuously, leading to an increase from 25 deals in 2008 to 155 in 2015. Throughout the same period, the annual number of completed deals stayed relatively stable for CSOEs and approximately doubled for other types of SOEs.

We include six OFDI-supportive policies in our study, of which, half were of the coercive and half of the noncoercive policy type. Other policies were not considered as they either are specific on the industry – or the firm ownership level – and thus not directly comparable or they have a potentially nonsupportive influence on the M&A decision (e.g. OFDI policies of FDI statistics). As most of China's OFDI policies are supportive (using computer-aided content analysis technique [2], we found that 107 out of 117 Chinese OFDI policies issued between 1980 and 2019 showed a supportive tendency), we focus on the supportive policies. We categorized all policies of the 2008–2015 period based on a thorough analysis of their texts. Thereby, we distinguished between coercive and noncoercive policies by studying the approach of each policy in supporting firm internationalization. If a policy does not formally state the circumstances under which support is given to a cross-border M&A project, the policy is categorized as being of the noncoercive type. In this situation, the policy does not guarantee firms to gain support, even if they would act in the way that the policy states (e.g. the publication of government opinions).

We include a one-month lag between our dependent and all of the explanatory variables to avoid possible endogeneity issues (Yuan and Pangarkar, 2015) and counteract the inclusion of internationalization transactions taking place within the month of policy issuance but before the policy announcement.

4.2 Variables

4.2.1 Dependent variable

Following the literature, our dependent variable is measured by the number of completed cross-border M&A deals (e.g. Li et al., 2018b; Xia et al., 2014). To be able to measure the direct effect of the policy implementation on the firms' internationalization, we use the monthly count of M&A deals (see Table 2). Thus, we can be close to the policy implementation date and do not lose too much of an immediate policy effect (c.f. Cui, 2016; Globerman and Shapiro, 1999) as it would be the case with lagging for a longer period.

4.2.2 Explanatory variables

With our zero-inflated model, we estimate the count of Chinese company's (i) investments abroad in country (j) in a given month (t) as

CNi,j,t=α+β1Xi,tHome+β2Zj,tHost+Industry Fixed Effects+Year Fixed Effects+εi,j,t

Where β1Xi,tHome includes our home country explanatory variables, in β2Zj,tHost, we cover the country control variables.

The explanatory variables on coercive OFDI policies and noncoercive OFDI policies are deemed to be of OFDI-supportive nature (Table 3). As the date of policy issuance is the same as its effective date for all included policies, each policy will be proxied by a dichotomous dummy covering the period after its announcement and until policy expiration. Thus, we can study the policies' potential influence on OFDI.

As a moderator variable, we include firm ownership (Hong et al., 2015; Wang et al., 2012b), whereby we distinguish between different levels of closeness to the central government, from CSOEs being under very strong supervision to other SOEs being under central control but not under the special body (SASAC) for China's most important state firms (Chang and Jin, 2016; Li et al., 2014), to LSOEs and POEs. By including firm ownership as a categorical variable, we follow an established approach of the literature (e.g. Cui and Jiang, 2012; Dau, 2012). As is standard practice for categorical variables, the lowest level of the categorical variable (for us POEs) is defined as a reference point to which the results of the other levels are referred to (StataCorp, 2015). Furthermore, as Lu et al. (2014) found that FDI experience is an important moderator for studying home government support, we follow their approach and include a moderator for the M&A experience of Chinese firms.

4.2.3 Control variables

Following the literature on Chinese OFDI, we control on the firm level for firm age (e.g. Wang et al., 2012a), firm size (e.g. Holtbrügge, 2018; Luk et al., 2008) and foreign ownership (Bhaumik et al., 2019; Karhu, 2015). We control for home country effects, such as the differences in access to central institutions, industrial clusters and infrastructure (e.g. Fan et al., 2011). Moreover, we consider the potential mimetic pressures on a firm's decision to internationalize (e.g. Henisz and Delios, 2001; Li et al., 2018b) and the presidency of Xi Jinping bringing a new approach toward Chinese OFDI (Devinney and Hartwell, 2020). For host countries, we control for high-income countries (Buckley et al., 2014), inward FDI restrictiveness (Lu et al., 2014), as well as tax havens (e.g. Buckley et al., 2015; Fuest et al., 2019).

4.3 Modeling approach

As the dependent variable consists of discrete, nonnegative integers, regarding Greene (2011), a standard Poisson model is not suitable and a zero-inflated count model regression should be applied. Performing information criterion tests (Desmarais and Harden, 2013), our results indicate that the ZINB model is the most appropriate. As our zero-inflation parameter, we use the sum of a firm's internationalization activities (Li et al., 2018b) of the previous 60 months, i.e. we estimated the probability of zero OFDI activities based upon the internationalization behavior within the last 60 months. Further, we included firm-level clustered standard errors in all our estimations to control for within-country correlation.

5. Results

In Table 4, we included the descriptive statistics for the variables of our model. We find that our variables for investment treaties and high-income countries, as well as noncoercive policies and the variable for the presidency of Xi Jinping reveal higher levels of correlation. We tested these combinations for being able to be run together and found no changes in significance or the coefficient sign in our models. The variance inflation factors test shows that we do not have any problem with collinearity (Brown et al., 2018).

The results of the ZINB model are shown in Table 5. Where the first model only includes the control variables, in model 2, we add the OFDI-supportive policy dummy. Model 3 splits the dummy covering all OFDI-supportive policies into coercive and noncoercive policies. While model 4 covers the interaction terms on ownership types and coercive policies, model 5 includes noncoercive policies. This is repeated in models 6 and 7; however, we replace ownership with M&A experience.

Hypothesis 1 predicts that the announcement of OFDI-supportive policies has a positive effect and thus leads to more cross-border M&A. The significant positive coefficient of our OFDI-supportive policies variable (model 2) supports our hypothesis (b = 0.152, p = 0.1). Hypothesis 2 (model 3) predicts that the announcement of coercive policies (b = 0.091) leads to more M&A than the announcement of noncoercive policies (b = 0.202, p = 0.1) cannot be supported. Turning toward our hypotheses studying the moderation effect of firm ownership on the relationship between policy announcements and cross-border M&A, we want to highlight that our different ownership types are part of a categorical variable (firm ownership). Hence, the results need to be interpreted in relative terms to the base of the categorical variable (POEs). Against the prediction of Hypothesis 3a (model 4), our results show that for MNEs with the highest home government ties (CSOEs), the policy effect is with the smallest (b = −0.377, p = 0.01), next to other SOEs (b = −0.454, p = 0.01). None of the state ownership categories shows a more positive coercive policy announcement effect than POEs. With a focus on noncoercive policies (Hypothesis 3b: model 5), we predict that with weaker levels of government ties, the policy announcement has a stronger effect on cross-border M&A. This gains support by our results as not only with higher levels of government connection, the coefficients of the noncoercive policy interaction terms turn increasingly negative, from LSOEs (b = −0.337) to other SOEs (b = −0.497) and CSOEs (b = −0.537), but all interaction terms are also significant at the one percent level. Finally, our results reveal that Hypothesis 4a (model 6), where we predict that with more M&A experience, the announcement of coercive policies has less effect on cross-border M&A, can be supported. The interaction variable between coercive policies and M&A experience is negative and significant (b = −0.075, p = 0.01). Focusing on noncoercive policies, we also confirm our Hypothesis 4b (model 7) that increasing internationalization experience replaces some of the effects of noncoercive policy announcements (b = −0.085; p = 0.01).

6. Post hoc analysis

Following Brown et al. (2018), we control the robustness of our ZINB regression with a change in the outcome variable to a binary variable – being 1 for the month that the firm completes an M&A deal and 0 otherwise. With the dichotomous variable leading to M&A being a rare event, we perform a firthlogit regression (Firth, 1993). The results confirm the findings of our ZINB models.

In addition to the logit approach, we analyzed different periods for measuring a firm's internationalization experience (three years, five years, since 1980) and the firm's mimetic pressure to internationalize. For the latter, we also used different combinations between home country regions/provinces, industries and ownerships. For our investment treaty variable, we also considered only bilateral investment treaties, excluding other investment treaties. The results confirm the robustness of our findings in Table 5.

7. Discussion, conclusions and research limitations

7.1 Discussion and conclusions

In this article, we explore how OFDI-supportive policy types are different in their effect on cross-border M&A and elaborate on the moderating role of government ties and M&A experience. By studying the Chinese context, we focus on a market environment with large government policy intervention. Nevertheless, we believe that our findings can be applied to other emerging economies revealing a similar relationship between the state and cross-border M&A. Using policy change-specific literature from the political sciences in our research, we define an avenue for forthcoming IB policy studies and make the following contributions:

First, our article contributes to the literature on OFDI policies with the comparative analyses of policy types. So far, no study empirically tested the difference in the announcement effect of OFDI policies on cross-border M&A, between coercive and noncoercive policies. While some studies focus on a selected policy (e.g. Du and Zhang, 2018; Lu et al., 2011, 2014) and were able to show that OFDI-supportive policies lead to more OFDI, our findings add to this literature by revealing that the announcement effect on cross-border M&A differs between policy types. With noncoercive policies having a larger effect on cross-border M&A than coercive policies, we argue that host country governments might increase their inward FDI restrictiveness in line with coercive policy developments in China (c.f. Cui, 2016; Raess, 2020; Yin, 2020). With a clearly defined support mechanism in coercive policies, host governments find formal ground for developing protective “countermeasures,” whereas with noncoercive policies, this basis becomes more unstable.

Second, we contribute to research on the moderating role of ownership types on the relationship between home country policies and firm internationalization. The literature (e.g. Cui, 2016) argues that across ownership types, the response to home country OFDI policies might differ. Hong et al. (2015) find that the effect of state ownership on the amount of a firm's OFDI is especially strong in industries that receive government support. Our study contributes to this literature by finding that with decreasing government ties, the influence of coercive and noncoercive FDI policy announcements on cross-border M&A becomes greater. Our findings also show that while POEs and SOEs in China from many perspectives appear similar (Milhaupt and Zheng, 2015), the policy pressures to internationalize differ.

Third, our results add to the literature on the role of internationalization experience for cross-border M&A, by revealing that the effect of coercive and noncoercive policy announcements gets partially replaced with increasing experience. This supports the argument of Luo et al. (2010) that with more M&A experience, firms increase their knowledge on the value (and costs) of their home country support when investing abroad. At the same time, our results also confirm the argument of Lu et al. (2014) that with using OFDI-supportive policies, the internationalization experience loses in importance.

Fourth, as our results show, when studying home country government support in the form of policies, the mere focus on institutional theory (e.g. Lu et al., 2011) might not be sufficient to explain differences in results between policy types (c.f. Clegg, 2019). By integrating the policy change literature from the political sciences, we provide another stream theoretical perspective to the IB scholarship. We argue that this is required as with increasing cross-border M&A from countries of strong state support, the politicization of this form of internationalization also gains traction and we need to appropriately frame the picture.

With our findings, we provide valuable insights for policymakers of the home and host country. The former, for example, can take away that coercive and noncoercive policies are not alike in their influence on cross-border M&A and that host countries might more strongly or merely react to the announcement of coercive policies. This could influence how the home country policymakers formulate policies in the future. For host country policymakers, our findings on the relationship between firm ownership types and the influence of coercive policies on cross-border M&A might be of specific interest as this might help in formulating appropriate inward FDI screening mechanisms.

7.2 Limitations and future research

We acknowledge that our study does not come without limitations. For example, the vague language used in policy statements leaves room for interpretation. Furthermore, our policies are included as binary variables which do not provide insights into the policy-embedded tools through which the government supports OFDI. Besides, as we focus on different home country OFDI policy types, factors of the host country are only included to a limited extend. Here, future research might study the effect of OFDI policies from a host country perspective. As more countries implement inward FDI screening mechanisms, this offers great research potential.

Figures

Figure 1

Conceptual framework

Figure 2

Chinese cross-border M&A and OFDI flows, 2008–2015

Table 1

Policies and institutions

Policy type	Coercive/regulatory	Noncoercive/normative
Policy intention	Impose obligations	Impose positions
Policy formalization	Highly formalized	Less formal/less specific
Policy comparison level	Power	Privileges
Institutional pillars	Regulative	Normative
Basis of compliance	Expedience	Social obligation
Basis of order	Regulative rules	Binding expectations
Mechanisms	Coercive	Normative
Logic	Instrumentality	Appropriateness
Basis of legitimacy	Legally sanctioned	Morally governed

Source(s): Adapted from Lowi (1985), Scott (2014, p. 60)

Table 2

Variable description and data sources

Variables	Descriptions and data sources
M&A	Dependent variable. M&A is defined as the monthly count of completed cross-border M&A deals (source: SDC; Zephyr)
M&A experience	Moderator variable. Count of a firm's M&A deals completed within the previous 60 months (source: SDC; Zephyr)
Firm ownership (CSOEs, LSOEs other SOEs, POEs)	Moderator variable. Categorical variable with POEs as the base level. CSOEs are defined regarding the official list of SASAC (SASAC, 2017), other state-owned or controlled enterprises (other SOEs) are either under the control of ministries or the state's sovereign wealth fund (China International Corporation), LSOEs are defined as being under the control of the local government, and POEs are not owned or controlled by any state authority (sources: SDC; Orbis; annual reports; company websites)
Firm age	Control variable. Difference between an individual month and the year of establishment (sources: Orbis; annual reports; company websites)
Firm size (small, medium, large, very large firms)	Control variable. Grouping firms into four different size categories based upon their operating revenue, total assets and the number of employees (source: Orbis)
Foreign shareholders	Control variable. Share of all foreign shareholders on a firm's total shareholders at the end of the year (source: Orbis)
Closeness to industrial clusters (eastern region, central region, western region)	Control variable. Grouping home regions into eastern, intermediate and western region (source: National Bureau of Statistics, 2018)
Closeness to central state institutions (Beijing)	Control variable. Binary variable for being located in the capital city (sources: Orbis, SDC; annual reports; company websites)
Mimetic pressures	Control variable. Count of completed M&A projects of Chinese MNEs with the same ownership and industry, covering the previous 60 months (sources: Orbis, SDC; annual reports; company websites)
Presidency of Xi Jinping (Xi Jinping)	Control variable. Binary variable being 1 for the months following Xi Jinping's presidential inauguration, before 0
Knowledge acquisition (high-income countries)	Control variable. Counting the investments into high-income countries; monthly (source: World Bank, 2020)
Host country inward FDI restrictiveness (investment treaties)	Control variable. Binary variable being 1 if the investment is into a host country China has to the point of the M&A an effective bilateral or other type investment treaty with, otherwise 0 (source: UNCTAD, 2020b)
Tax havens	Control variable. Counting the investments into tax haven countries; monthly (source: Fuest et al., 2019; OECD, 2001)
OFDI-supportive policies (OFDI-supportive policies)	Main independent variable. Count variable covering coercive and noncoercive OFDI policies
Coercive OFDI policies (coercive policies)	Main independent variable. Count variable covering coercive OFDI policies
Noncoercive OFDI policies (noncoercive policies)	Main independent variable. Count variable covering noncoercive OFDI policies

Table 3

OFDI-supportive policies included in our regression

Coercive policies	Issued	Citation
Decentralizing the Power of Examination and Approval in OFDI Projects Summary: The policy aims at meeting the needs of overseas investment development under the new economic situation. It increases the threshold value for which an OFDI project approval is needed from central government institutions and decentralizes decision-making	February 14, 2011	NDRC (2011)
Measures for the Administration of Special Funds for Foreign Investment Cooperation Summary: To continuously improve China's “going global” strategy and regulate the foreign economic cooperation management, an investment fund covering pre-implementation expenses of OFDI projects is implemented	June 14, 2013	MOF and MOFCOM (2013)
Centralized Operation and Management of Foreign Exchange Funds of Multinational Companies (for Trial Implementation) Summary: With this policy, the Chinese government aims to improve its current approaches to cross-border investments and upgrade its industry. By fulfilling specific criteria, Chinese companies can use foreign exchange funds abroad and in their home country	April 18, 2014	SAFE (2014)

Noncoercive policies	Issued	Citation
National Work of Overseas Investment and Cooperation for 2010 Summary: To improve the development level of “going global”, it highlights the OFDI development objectives and guidelines for 2010	February 26, 2010	MOFCOM (2010)
“Twelfth Five-Year Plan” Utilization of Foreign Capital and OFDI Planning Summary: Informing companies on forthcoming OFDI-related changes, the policy describes how OFDI shall be conducted in the forthcoming five-year period and which government procedures shall be improved to support OFDI	July 17, 2012	NDRC (2012b)
Developing a New System of Open Economy Summary: The efforts to attract foreign investment into China shall be organically combined with the efforts of “going global”. To additionally encourage the use of Renminbi in making OFDI this policy introduces a new (OFDI encouraging) era of economic (policy) reform with 2021 (the 100th anniversary of the founding of the Communist Party of China) and 2049 (the 100th anniversary of the founding of the People's Republic of China) as key dates	May 5, 2015	State Council (2015)

Table 4

Correlation matrix and descriptive statistics

Variable	Mean	SD	1	2	3	4	5	6	7	8	9	10	11	12	13
1. M&A	0.018	0.148
2. LSOEs	0.140	0.347	0.007
3. Other SOEs	0.063	0.243	0.009	−0.109
4. CSOEs	0.091	0.288	0.040	−0.135	−0.087
5. Firm age	13.576	9.622	0.025	0.140	0.090	0.235
6. Medium firms	0.030	0.169	−0.009	−0.074	−0.047	−0.059	−0.112
7. Large firms	0.089	0.284	−0.011	0.024	−0.082	−0.078	−0.095	−0.055
8. Very large firms	0.679	0.467	0.019	0.061	0.095	0.163	0.126	−0.284	−0.491
9. M&A experience	0.685	1.349	0.072	0.050	0.073	0.291	0.215	−0.047	−0.089	0.114
10. Eastern region	0.828	0.377	0.010	−0.198	0.016	0.095	−0.005	−0.008	−0.057	0.063	0.033
11. Central region	0.079	0.270	−0.005	0.072	−0.018	−0.045	−0.008	−0.010	0.079	−0.067	−0.010	−0.639
12. Beijing	0.246	0.430	0.028	−0.116	0.159	0.417	0.072	−0.088	−0.085	0.111	0.177	0.247	−0.149
13. Mimetic pressure	3.549	6.112	0.027	−0.025	−0.106	0.021	−0.010	−0.021	0.030	−0.020	0.098	0.033	−0.031	0.054
14. High-income countries	0.015	0.129	0.018	0.006	0.003	0.030	0.018	−0.008	−0.009	0.013	0.154	0.008	−0.002	0.021	0.031
15. Investment treaties	0.012	0.123	0.014	0.010	0.003	0.034	0.025	−0.006	−0.006	0.016	0.150	0.005	−0.003	0.018	0.024
16. Foreign shareholders	0.011	0.073	0.003	−0.032	0.034	−0.026	0.013	−0.029	0.019	0.037	0.013	0.048	−0.024	−0.036	−0.039
17. Tax havens	0.001	0.037	0.003	0.000	0.005	0.001	0.001	−0.002	−0.003	−0.003	0.029	0.004	0.000	0.006	0.002
18. Xi Jinping	0.250	0.433	0.033	−0.004	−0.001	−0.005	0.145	−0.001	0.010	−0.014	0.158	−0.003	0.001	−0.001	0.256
19. OFDI-supportive policies	2.379	1.996	0.037	−0.006	−0.003	−0.006	0.184	0.000	0.008	−0.016	0.212	−0.002	0.002	0.001	0.327
20. Coercive policies	1.137	1.148	0.035	−0.006	−0.002	−0.006	0.176	0.000	0.009	−0.016	0.201	−0.003	0.003	0.000	0.313
21. Noncoercive policies	1.242	0.926	0.035	−0.006	−0.003	−0.006	0.177	0.000	0.007	−0.015	0.208	−0.001	0.002	0.001	0.316

Variable	14	15	16	17	18	19	20
15. Investment treaties	0.754
16. Foreign shareholders	0.003	0.003
17. Tax havens	0.224	0.022	0.001
18. Xi Jinping	0.038	0.031	0.013	0.005
19. OFDI-supportive policies	0.040	0.034	0.012	0.001	0.802
20. Coercive policies	0.039	0.033	0.012	0.003	0.861	0.970
21. Noncoercive policies	0.037	0.032	0.011	−0.002	0.660	0.952	0.849

Table 5

ZINB regression analysis of Chinese cross-border M&A deals

	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6	Model 7
Control variables
M&A experience	0.075*	0.075*	0.075*	0.088***	0.095***	0.234***	0.261***
	(0.041)	(0.041)	(0.041)	(0.034)	(0.033)	(0.070)	(0.037)
LSOEs	0.233**	0.232**	0.232**	0.564***	0.734***	0.220**	0.222**
	(0.096)	(0.096)	(0.096)	(0.159)	(0.201)	(0.099)	(0.098)
Other SOEs	0.280**	0.280**	0.279**	0.850***	0.965***	0.261**	0.268**
	(0.126)	(0.126)	(0.126)	(0.208)	(0.236)	(0.125)	(0.127)
CSOEs	0.461***	0.466***	0.467***	0.960***	1.209***	0.369**	0.381***
	(0.134)	(0.134)	(0.134)	(0.160)	(0.187)	(0.147)	(0.142)
Firm age	0.003	0.003	0.003	0.002	0.002	0.002	0.002
	(0.004)	(0.004)	(0.004)	(0.004)	(0.004)	(0.004)	(0.004)
Medium firms	−0.078	−0.077	−0.077	−0.083	−0.080	−0.097	−0.092
	(0.096)	(0.096)	(0.096)	(0.096)	(0.095)	(0.101)	(0.099)
Large firms	−0.091	−0.090	−0.091	−0.094	−0.095	−0.094	−0.096
	(0.098)	(0.098)	(0.098)	(0.096)	(0.096)	(0.102)	(0.101)
Very large firms	0.124	0.124	0.124	0.125	0.124	0.122	0.119
	(0.078)	(0.078)	(0.078)	(0.078)	(0.078)	(0.083)	(0.082)
Foreign shareholders	0.502	0.503	0.504	0.419	0.431	0.441	0.430
	(0.404)	(0.404)	(0.404)	(0.416)	(0.413)	(0.414)	(0.414)
Eastern region	0.173*	0.173*	0.172*	0.176*	0.175*	0.187**	0.182**
	(0.094)	(0.094)	(0.094)	(0.091)	(0.090)	(0.090)	(0.090)
Central region	0.069	0.069	0.068	0.061	0.059	0.051	0.051
	(0.121)	(0.121)	(0.121)	(0.115)	(0.114)	(0.106)	(0.108)
Beijing	0.134*	0.134*	0.134*	0.134*	0.133*	0.125*	0.126*
	(0.075)	(0.075)	(0.075)	(0.073)	(0.073)	(0.075)	(0.075)
Mimetic pressure	0.013*	0.012*	0.012*	0.007	0.006	0.013*	0.012*
	(0.007)	(0.007)	(0.007)	(0.007)	(0.007)	(0.007)	(0.007)
Xi Jinping	0.844***	−0.005	0.046	0.619	0.433	0.512	0.251
	(0.208)	(0.498)	(0.502)	(0.421)	(0.358)	(0.420)	(0.358)
High-income countries	0.354	0.352	0.351	0.319	0.319	0.344	0.365
	(0.319)	(0.320)	(0.320)	(0.308)	(0.310)	(0.297)	(0.303)
Investment treaties	−0.132	−0.133	−0.133	−0.128	−0.131	−0.124	−0.143
	(0.325)	(0.327)	(0.327)	(0.314)	(0.318)	(0.300)	(0.307)
Tax havens	−0.008	−0.010	−0.005	0.024	0.013	−0.037	−0.039
	(0.567)	(0.564)	(0.566)	(0.562)	(0.568)	(0.553)	(0.566)
Independent variables/Interactions
OFDI-supportive policies		0.152*
		(0.084)
Coercive policies			0.091	0.226*		0.196
			(0.123)	(0.127)		(0.124)
Coercive p.*LSOEs				−0.230**
				(0.093)
Coercive p.*other SOEs				−0.454***
				(0.118)
Coercive p.*CSOEs				−0.377***
				(0.097)
Noncoercive policies			0.202*		0.371***		0.323***
			(0.120)		(0.128)		(0.122)
Noncoercive p.*LSOEs					−0.337***
					(0.128)
Noncoercive p.*other SOEs					−0.497***
					(0.126)
Non-coercive p.*CSOEs					−0.537***
					(0.110)
Coercive p.*M&A experience						−0.075***
						(0.024)
Noncoercive p.*M&A experience							−0.085***
							(0.024)
Constant	−4.723***	−4.731***	−4.733***	−5.073***	−5.276***	−5.316***	−5.485***
	(0.676)	(0.684)	(0.684)	(0.663)	(0.684)	(0.979)	(0.244)
Observations	50,554	50,554	50,554	50,554	50,554	50,554	50,554
Industry	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes	Yes

Note(s): *, **, *** mean significant at, respectively, 10%, 5%, 1% levels; robust standard errors clustered at the firm level in parentheses

Notes

1. For exceptional cases in which POEs also received financial no-policy state support in their OFDI endeavors, see Milhaupt and Zheng (2015).

2. We used KH Coder software (version 3.01) from Higuchi (2020) to perform a content analysis on all 117 Chinese OFDI policies published between 1980 and 2019. Defining keywords for the OFDI-supportive and restrictive tendencies, their number of appearances were counted per policy. While this approach has clearly some shortcomings, it tackles the subjectivity issues of defining policy tendencies through a human-coding (Gaur and Kumar, 2018; Ostergard, 2000).