Alliance Network Position,Embeddedness and Effects on the Carbon Performance of Firms in Emerging Economies

Abstract

This article explores the effect of network structure on the carbon performance of firms in emerging economies at an ego network level. Building on the theoretical framework of social networks, we posit that an ego firm’s network position, structural embeddedness, and structural constraint affect carbon performance. We examine the research hypotheses using a panel data set made up of 44 Indian firms that entered into alliances under the Clean Development Mechanism over the 2005-2009 period. Our main results show that the central position of the focal firm in the network and its degree of network embeddedness exert a positive effect on its carbon performance. This article contributes to the literature on climate strategy by exploring the influence of the structural characteristics of the firm’s ego network in the carbon market on its environmental performance.

Keywords

carbon performance environmental alliance network structure carbon market climate strategy emerging economies

Introduction

In this article, we investigate the effect of alliance network embeddedness on the carbon performance of emerging economies’ firms, which operate in a carbon market. Improving the carbon performance, defined here as a firm’s reduction of greenhouse gas emissions (Busch & Hoffmann, 2011), is a key challenge faced by emerging economies’ firms today as they can no longer ignore stakeholder pressures and the costs associated with buying the rights to pollute in the carbon constrained future (Brohe, Eyre, & Howarth, 2009; Busch & Hoffmann, 2011; Stern, 2007). We argue and demonstrate that alliance network embeddedness, defined as a pattern of ties between firms, affects carbon performance.

Literature on carbon performance has highlighted and developed various institutional and strategic factors leading firms to reduce their greenhouse gas emissions (Pinkse, 2007). In this literature, carbon performance has been linked with regulatory and normative forces (Porter & Linde, 1995), existing capabilities (Christmann, 2000), competitive resources (Hart, 1995), management attitude (Sharma, 2000), stakeholder pressures (Jennings & Zandbergen, 1995), or concern to be efficient, and legitimate (Bansal & Roth, 2000), in response to competitive and institutional pressures (Delmas & Toffel, 2008). We contend, however, that a carbon-constrained future, which may lead to a crisis for firms lacking environmental capabilities (Busch & Hoffmann, 2011), necessitates not only managing institutional pressures, or accessing competitive and scarce resources, but also forming environmental alliances (Koontz & Thomas, 2006; Wassmer, Paquin, & Sharma, 2012) and assembling them into alliance networks.

Literature on carbon performance has clearly acknowledged the key role of environmental alliances, defined as “arrangements between a firm and one or more other organizations with the goal of reducing negative or generating positive environmental impact” (Wassmer et al., 2012, p. 3). Environmental alliances enable firms to jointly tackle the complexity of reducing their individual and collective environmental footprints under uncertainty (Lin, 2012a, 2102b). Studies in this stream of literature have empirically demonstrated the positive alliance effect on environmental performance of partnering firms (Albino, Dangelico, & Pontrandolfo, 2012; Lin, 2012a, 2012b). However, these studies have been conducted using a dyadic or a single-alliance perspective. To our knowledge, this literature has not addressed the firm’s carbon performance from an alliance network perspective. Such a perspective emphasizes the importance of managing the relationships and connections with all the firms embedded in an alliance network so as to access new information, to gain a stronger central position within the network, to negotiate with partners on better terms (Burt, 1992, 1995; Wassmer et al., 2012), or to enhance trust and collaboration (Coleman, 1990; Meschi & Wassmer, 2013) within the network to perform more effectively. To fill this gap, we address two interlinked research questions: Does participation in an alliance network affect the carbon performance of the firm? What, if any, network characteristics affect the carbon performance of the firm?

These two questions are studied in the context of carbon market strategies of firms in emerging economies, that is, countries with “a rapid pace of economic development, and government policies favoring economic liberalization and the adoption of a free-market system” (Hoskisson, Eden, Lau, & Wright, 2000, p. 249). Our focus on this specific context is motivated as follows: First, the majority of the extant literature is concentrated on non–market climate strategies of emerging economies’ firms (He, Tian, & Chen, 2007), despite the important role the carbon market plays (Brohe et al., 2009; Kolk & Mulder, 2011) in providing firms with resources, and a platform to interact, share, and influence each other’s behaviors. Second, we contend that adopting such a network perspective is particularly relevant in the context of emerging economies. This is because their high inherent uncertainty (Hoskisson et al., 2000) renders the understanding of determinants of carbon performance particularly difficult (Jeswani, Wehrmeyer, & Mulugetta, 2008; Lin, 2012a, 2012b), as they do not face quotas to emit greenhouse gases, and are relatively at a disadvantage due to lack of access to information, technologies and resources (Jeswani et al., 2008; Lin, 2012b). By being embedded in a network of environmental alliances and managing this particular pattern of ties, emerging economies’ firms can access key information and technologies, build reputation and attract resourceful partners, and consequently improve their carbon performance.

Responding to the call of certain researchers (Roome, 2001; Wassmer et al., 2012), with this article, then, we aim to contribute to the literature on carbon performance. First, by delineating the effect of network embeddedness on the carbon emission reduction, our article contributes toward the development of the network theoretical framework and demonstrates the dynamics of alliance networks related to carbon performance. Building on the existing literature on carbon market strategies and interorganizational networks, we argue in this article that adopting the multialliance network level of analysis will help in better understanding and explaining the carbon performance outcomes and their determinants. Second, we analyze the characteristics and structure of the ego network, defined here as, the alliance network of focal emerging economies’ firms (or egos), and relationship to its carbon emission reduction.

The article is organized as follows: In the first section, we present a literature review on key concepts related to carbon performance and the carbon market. The second section proposes a network theoretical framework (Borgatti & Halgin, 2011). Building on this framework, we develop research hypotheses predicting the relationship of three key dimensions of a focal firm’s (or an ego’s) network structure—namely, network position (or centrality), structural embeddedness, and structural constraint—to its carbon performance. In this section, our approach to alliance networks differs from existing approaches that focus on government and stakeholder stimuli and information mechanisms as socially constructed antecedents of environmental performance. Here, we view alliance networks as objective information channels and influencing mechanism that shape a firm’s carbon strategy and performance. In the third section, we introduce our panel data set, comprising 44 focal Indian firms engaged in carbon market alliances over the 2005-2009 period, and the selected variables. Subsequently, we test the hypotheses and present the statistical results. The main findings and limitations of the article, as well as directions for future research, are discussed in the final section.

Literature Review

Environmental Alliances, Alliance Network, and Carbon Performance

A rich body of literature has demonstrated that corporate carbon performance affects financial performance (Berchicci & King, 2007; Lee, 2012; Wang, Li, & Gao, 2013). Given this important role of carbon performance, various studies have theorized and empirically highlighted its determinants. Scholars have traced the motives for increasing carbon performance mostly through institutional pressures (Bansal & Roth, 2000; Delmas & Toffel, 2011; Porter & Linde, 1995) and resource-based, endogenous competitive reasons (Bansal & Roth, 2000; Hoffman, 2005). For example, Sariannidis, Zafeiriou, Giannarakis, and Arabatzis (2013) demonstrate that coercive pressures from investors motivate firms to reduce carbon emissions. Delmas and Toffel (2011) explain that market pressures (from investors, competitors, and customers) and nonmarket pressures (from nongovernmental organizations [NGOs], and regulators) affect the firm’s environmental strategy (including carbon performance) and that organizational characteristics moderate this influence. Whereas this stream of literature views corporate carbon strategy as the outcome of institutional and competitive pressures, and despite enlightenment of the factors that affect the firm’s carbon performance, most of these studies are firm specific and do not use ties between firms as unit of analysis. In fact, firms’ ties and their patterns and configurations (otherwise defined as network structure) are objective channels of influence (Borgatti & Foster, 2003; Emirbayer, 1997). Understanding how this objective influence mechanism plays out to enhance the firm’s carbon performance is far from being fully understood and thus merits further attention.

Few other studies demonstrate the effect of interorganizational relationships on the carbon performance of firms. For example, Lin (2012a, 2012b) shows that alliances with diverse set of partners enhance the firm’s carbon performance. Similarly, Albino et al. (2012) demonstrate that alliances between suppliers, customers, government agencies, and NGOs positively affect the firm’s carbon performance. However, these studies take the individual alliance as the unit of analysis and do not take into account the effect of a pattern of multiple ties between firms or alliance network on carbon performance. In the conclusion of an article on existing contributions and future directions for research on environmental alliances, Wassmer et al. (2012) summarize how the firm’s engagement into an alliance affects its carbon performance and point out that exploring a firm’s position in its network of ties in relation to its subsequent environmental performance is an area that needs further scholarly consideration. The few and sparse studies that investigate such a network effect include Pulver (2007) who demonstrate that variation in firms’ responses to address climate change “are best explained by the different scientific networks and regional and national policy fields in which corporate decision makers in each company were embedded” (p. 73). Akiyama (2010) shows that middle managers create a cohesive network, within the Japanese Skisui House Group, which ensures “socialization” and “externalization” (p. 240) of responsibility values and tacit knowledge and consequently helps achieve environmental management targets. Both of these studies (Akiyama, 2010; Pulver, 2007) appreciably demonstrate the effective role of networks at the manager level. However, what remains to be explored is whether networks studied at firm level affect corporate carbon performance. In this article, we will examine this relationship, that is, the influence of the firm’s alliance network on its carbon performance.

Carbon Market and Carbon Performance of Firms From Emerging Economies

Consistent with the above research gap, we contend that another related area requiring scholarly attention is the question of which factors affect environmental behavior and the performance of emerging economies’ firms (Wang et al., 2013). Most of the existing literature has studied carbon strategy and performance for firms from developed countries’ (Kolk & Pinkse, 2005), which in many cases are required by regulations to reduce their carbon footprint. However, conclusions drawn from this stream of literature cannot hold true for firms from emerging economies. Due to different business dynamics in the emerging economies (Hoskisson et al., 2000), firms are not constrained, through local and formal regulations, to behave responsibly. The literature suggests that emerging economies’ firms, from the Asia Pacific region in particular, behave responsibly as such behavior is rewarded by carbon market trading (Bansal & Knox-Hayes, 2013). The carbon market provides emerging economies’ firms with the opportunity to procure innovative environmental technologies from developed countries (Petersen, Escobar, Espinoza, & Vredenburg, 2006; Schneider, Holzer, & Hoffmann, 2008). There are a number of other advantages that are also demonstrated in the literature on the carbon market. Several authors (Schneider et al., 2008; Seres, Haites, & Murphy, 2009) have shown that the carbon market facilitates the transfer of human skills and institutional capacity to manage projects from developed to developing countries (Seres et al., 2009). They lower the barriers of environmental projects’ commercial viability by increasing internal rate of return, spreading information through stakeholders and intermediaries, and accessing capital (Schneider et al., 2008). They also give first-mover advantages to emerging economies’ firms (Petersen et al., 2006), and as a consequence of the emerging value chain associated with the carbon market (Kolk & Mulder, 2011; Schneider, Hendrichs, & Hoffmann, 2010), the latter potentially facilitates the sharing of knowledge and the formation of alliances that can enhance the firm’s innovative capabilities. While literature on the advantages and complexities of the carbon market abounds, many questions remain unanswered. This stream of literature and associated research put into evidence, for example, the fact that the carbon market is beneficial to emerging economies’ firms. It is assumed that by teaming up with carbon market actors, emerging economies’ firms would tap into accrued opportunities and resources and consequently improve their carbon performance. Worthy of note however—and one that this article seeks to address—is the fact that such research efforts demonstrate focus at both single-firm and single-alliance levels, with little or no focus at the network level.

Theoretical Framework and Hypotheses

Network Theoretical Framework

In this article, we build on network theory to explore the relationship of alliance network embeddedness of emerging economies’ firms to their carbon performance. Network theory aims at understanding the dynamics of social actors¹ and the ties between them. It is defined as “the mechanisms and processes that interact with network structures to yield certain outcomes” (Borgatti & Halgin, 2011, p. 1) for social actors.

Research on the outcomes of network, defined as “the pattern of relationships that exist among a set of actors” (Phelps, 2010, p. 890), can be traced back to the works of Moreno and Jennings (1934), who analyzed the link between runaway schoolgirls and their positions in their social network (Borgatti, Mehra, Brass, & Labianca, 2009). Later, the sociologist Durkheim argued that “reasons for social regularities were to be found not in the intentions of individuals but in the structure of the social environments in which they were embedded” (Durkheim & Simpson, 1951, cited in Borgatti et al., 2009, p. 892). Today, network theory encompasses various (sociological, organizational, and economic) approaches to comprehend the relationships between actors’ networks and their behavior.

In the following pages, we will draw arguments from three key concepts of network theory—centrality (Freeman, 1978), embeddedness (Granovetter, 1985), and structural holes (or structural constraint; Burt, 1992)—to explore the effect of networks on the carbon performance of emerging economies’ firms.

Centrality, Embeddedness, and Structural Holes

Centrality

Centrality is defined as the focal firm’s position in a network and its number of direct ties (Freeman, 1978; Wasserman & Faust, 1994). It determines in part which channels of information and what resources the focal firm has access to within its network. Emerging economies’ firms are at a relative disadvantage due to lack of access to technology, information, and know-how (Jeswani et al., 2008; Schneider et al., 2008) as to how to improve their carbon performance. As a consequence, emerging economies’ firms demonstrating a central position in the network are likely to enjoy relative competitive advantage in terms of access to information, knowledge, and technology. Research (Ahuja, 2000) has shown that knowledge and complementary skills are shared with centrally positioned firms in a network, making them more innovative. The more central position a firm occupies, the more new valuable information, reputation, power, and other benefits it enjoys (Ahuja, 2000; Burt, 1995). Central firms, due to their direct ties in their ego network, have better and more timely access to information compared to their counterparts in the industry (Ahuja, 2000; Gulati, 1995). Moreover, connecting with other actors in a network increases visibility and reputation. As a consequence, central firms have more chance to attract more resourceful partners, and form stable alliances (Polidoro, Ahuja, & Mitchell, 2011), which in turn translate into enhanced corporate performance.

Direct ties in the carbon market can provide central firms with access to new methods of reducing carbon emissions, resources, information, and technology by either tapping into the information and knowledge base in the network or attracting and forming alliances with firms with more resources. Consequently, we formulate the following hypothesis:

Hypothesis 1: For emerging economies’ firms operating in the carbon market, a central position in an ego network has a positive relationship to its carbon performance.

Structural Embeddedness

Firms contract long-term partnerships to reduce carbon emissions (Bansal & Knox-Hayes, 2013). Such an arrangement thus necessitates some mechanisms to ensure that alliance members respect laid down rights and maintain smooth functioning of the partnerships. In the carbon market, powerful partners, that is, those that occupy central positions in the alliance network, may constrain the behavior of other peripheral actors, and assert terms favorable to central firms only. However, such a difference in structural status, between central and peripheral positions, of an alliance network’s members can be expected in the carbon market as firms from developed countries are mostly carbon intermediaries. By definition, these carbon intermediaries have ties to many actors and enjoy relative powerful bargaining (central) positions, which may constrain the behavior of emerging economies’ firms.

Alliance network position accrues certain benefits to firms; however, if partners enjoy different status brackets, relative advantageous position in a network may increase friction between partners due to power (Greve, Baum, Mitsuhashi, & Rowley, 2010) and reputation asymmetries (Podolny, 1993). Similarly, if focal firms are peripheral in a given network, costs resulting from alliances with resource-rich central firms may overshadow the benefits of such collaborations. Structural embeddedness, defined as the “process by which social relations shape economic actions” (Uzzi, 1996, p. 674), can mitigate the negative side effects of having asymmetric partnerships between central and peripheral firms. Granovetter (1985) highlights that economic actors’ embeddedness in social networks play a role in generating order, trust, and rich information to transact with known and reputed actors, and that opportunistic behavior of partners to an economic exchange is “policed by quick spread of information about instance of malfeasance” (p. 492). As network members are structurally embedded, norm-breaking behavior gets greater visibility (Polidoro et al., 2011). Structural embeddedness also helps firms overcome friction in their alliances (Polidoro et al., 2011), which may arise due to resource scarcity and incompatibility (Greve et al., 2010). Network members mitigate the friction that may arise by associating with resource-rich nonsubstitutable firms, and by further embedding via common third-party firms (Bae & Gargiulo, 2004; Gulati & Gargiulo, 1999). Furthermore, embeddedness helps firms access reliable information about potential partners’ capabilities and evaluate the possible opportunistic behavior of partners (Gulati, 1995, 1999; Uzzi, 1996). Given the evidence that those emerging economies’ firms need resources and technology (Jeswani et al., 2008) to implement carbon reduction strategy, embeddedness, through creating order and trust (Coleman, 1990), augments the likelihood that proprietary information and technology will be shared in the network, which thus can increase the carbon performance of embedded firms. Based on the above developments, we hypothesize the following:

Hypothesis 2: For emerging economies’ firms operating in the carbon market, structural embeddedness in an ego network has a positive relationship to its carbon performance.

Structural Constraint

In the previous developments, we have argued that if firms occupy different alliance network positions (central vs. peripheral), there is a strong possibility that centrally positioned firms assert and dictate terms creating distrust, and friction between alliance members (Greve et al., 2010; Polidoro et al., 2011). However, structural embeddedness can mitigate the friction between central and peripheral actors, and generates trust between network members, which positively affects their carbon performance.

In this section, we argue that the benefit of being embedded in the alliance network can be overshadowed by the negative consequences on carbon performance of a specific network structure where members are densely connected with one another. This specific network structure, in which partners of the focal firm are strongly interconnected, is defined as structural constraint (Battilana & Casciaro, 2012; Burt, 1992, 2005). In such a network structure, where members are considered as structurally constrained, they cannot take advantage of access to nonredundant information (Burt, 1992). Furthermore, structurally constrained firms cannot control the collusive behavior of other network members by pitting them against one another to obtain benefits (Burt, 2005).

Investments to reduce carbon emissions are commissioned unilaterally (Lutken & Michaelowa, 2008) by emerging economies’ firms; and firms from developed countries join in later. Such a situation means that firms from emerging economies need to have diversified channels of information and knowledge about how to enhance their carbon performance. In addition, they must have a strong bargaining position to obtain favorable terms with firms from developed countries. Furthermore, the uncertainty in the carbon market (Kolk & Mulder, 2011) requires that emerging economies’ firms have nonredundant and disconnected contacts that could give them leverage in negotiating carbon prices and controlling the collusive behavior of their partners. Given this, we posit that in the absence of such a network structure the carbon performance of firms may be negatively affected. We thus hypothesize the following:

Hypothesis 3: For emerging economies’ firms operating in the carbon market, structural constraint has a negative relationship to its carbon performance.

Research Method

CDM Projects

We selected Kyoto Protocol’s Clean Development Mechanism (CDM), as an empirical context, as this provides a rich site to analyze relational influence on the behavior of emerging economies’ firms that participate in it. CDM helps developed countries meet their emission quotas at lower than domestic abatement costs. Trading of carbon credits generated through this flexible mechanism (Lecocq & Ambrosi, 2007; Paulsson & von Malmborg, 2004) is used as a strategy to compensate costs incurred by organizations in developed countries in order to adapt to climate change (Kolk & Pinkse, 2005). The process of commissioning a CDM project starts with approval from a designated national authority of host organizations that approves project design documents, namely, formal documents detailing technical aspects, prospective emission reductions, and sustainable development benefits. Project owners, organizations or NGOs, develop projects themselves, which are termed either as unilateral projects (Lutken & Michaelowa, 2008) or as specialist intermediaries, project developers, or consultants. Similarly, project participants, that is, legal CDM project partners, from developed countries, can also approach the approving authority and find suitable projects to invest in, as well as partners from developing countries (Bansal & Knox-Hayes, 2013). Once a project is approved by the host country, and validated by an authorized intermediary, it is then registered by the CDM Executive Board. Certified Emission Reductions (CERs, i.e., tradable carbon credits generated through CDM projects) are issued to project participants after verification and certification of operational entities.

Research Sample and Data

The CDM is primarily a project-based market. As of August 1, 2011, of a total of 3,337 CDM projects registered worldwide, 3,109 were at validation stage, and 113 were in the process of registration. China was a major producer of CERs, with 54.7% of the CERs’ share, from 2,727 projects, followed by India with a 15.4% share from 1,698 projects.

We concentrated only on projects in which two or more firms entered into partnership to reduce and sell carbon emissions under CDM. Our focus was the long-term (greater than 5 years) partnerships between project participants. These partnerships beside being commercial were strategic in nature (Jeswani et al., 2008; Schneider et al., 2008) as regards exchange of knowledge, management processes, ideas, and so on.

In this article, we focused exclusively on Indian firms involved in CDM projects. India is the second largest CDM project host. We used a three-stage selection procedure to constitute the research sample. First, as our aim was to measure the carbon performance of firms over a long period of time, and as the Kyoto Protocol took full effect from 2005, we selected all Indian firms that entered into one or several alliances aiming to reduce carbon emissions from 2005 to 2009 inclusive. We selected 2005-2009 as the period of observation with a view to capturing performance variance across two different carbon emission trading regimes (Betz & Sato, 2006). On this basis, the initial sample translated as 79 Indian firms. Second, we retained in the sample only Indian firms for which financial and complete project level data were available. Third, we traced all ties of these focal firms and removed all those entries in which parent firms had ties with their subsidiaries. As a result, our final sample comprised 44 Indian focal firms (or egos; see Table 1), their 55 Indian and international partners, and at the second tier level, 1,293 alters of egos’ partners. As a consequence, we analyzed a total of 1,504 CDM projects and 5,819 alliance ties in the CDM market between 2005 and 2009.

Table 1.

Sample Indian Firms.

No.	Firm name	Main Standard Industrial Classification code	No.	Firm name	Main Standard Industrial Classification code
1	Aditya Birla Nuvo	2281	23	Kalpataru Power Transmission	3629
2	Amarjothi Spinning Mills	2281	24	MSPL	3325
3	Andhra Pradesh Paper Mills	2281	25	My Home Power	4911
4	Anik Industries	2023	26	Nahar Spinning Mills	2281
5	Ashok Leyland Project Service	8712	27	Nuziveedu Seeds (NSL Renewable Power)	139
6	BF Utilities	4911	28	Orient Cement/Orient Paper & Industries	2621
7	Bharat Forge	3462	29	Oswal Woolen Mills	2211
8	CESC	4911	30	Philips Carbon Black	2895
9	Clarion Power	4911	31	Rajasthan State Mines and Minerals	1475
10	DCM Shriram	2875	32	Rajshree Sugars & Chemicals	2064
11	Deepak Spinners	2299	33	Rithwik Energy Systems	6512
12	Ganpati Sugar Industries	2061	34	SatyaMaharshi Power	4911
13	Godavari Sugar Mills	2061	35	Shree Bhawani Paper Mills	5111
14	Graphite India	3624	36	ShriRenuga Textiles	2281
15	Gujarat Alkalies & Chemicals	2812	37	SRF	2899
16	Gujarat Fluorochemicals	2813	38	Subhash Kabini Power	4911
17	Gujarat Heavy Chemicals (GHCL)	2899	39	Suzlon Energy	4931
18	Gujarat JHM Hotels (GJHM)	7021	40	Tamilnadu Newsprints and Papers	2621
19	Ind-Bharath Energies	4911	41	Tata Motors	3711
20	Indian Acrylics	2823	42	Triveni Engineering & Industries	2062
21	ITC	2111	43	United Phosphorus	2879
22	JSW Steel	3312	44	Usha Martin	3357

To build the sample firms networks, data on CDM projects were obtained from the Institute for Global Environmental Strategies (http://www.iges.or.jp/en/cdm/index.html). We triangulated all project level information with hundreds of available project documents, relevant financial and sustainability reports, and written exchanges between project participants and the CDM secretariat. For this purpose, data for cross-checking and verification were downloaded from the CDM Pipeline (http://cdmpipeline.org/), and the United Nation Framework Convention on Climate Change (http://cdm.unfccc.int/). Firm-level financial information was obtained from the ORBIS database, which is a global source of economic and financial information on private firms.

The sample ego network grew from a loosely coupled network in 2005 (see A in Figure 1) to one with increased connectedness in 2009 (see B in Figure 1). Connectedness, which signifies the degree to which firms connect disconnected groups, increased from 0.378 in 2005 to 0.977 in 2009, whereas network density increased from 0.017 to 0.051 over the same period.

Figure 1.

2005-2009 Evolution of an ego network structure of CDM projects.

Dependent Variable

Carbon Performance

We chose carbon emissions reduction under CDM to compute our dependent variable. Carbon emissions reductions can simultaneously capture carbon performance and financial performance of firms (Busch & Hoffmann, 2011). We selected total carbon emissions reductions per year (rather than emissions relative to company size) following Earnhart and Lizal (2006). However, we controlled for other firm-level factors to account for differential emissions. To reduce skewness, we logarithmically (natural log) transformed this variable.

Independent Variables

Network Position (Centrality)

To measure network position, we used degree centrality (Freeman, 1978) as it is considered a suitable measure for ego networks (Swaminathan & Moorman, 2009). Degree centrality of an ego measures how many alters are connected with that ego.

It takes on average 670 days to register a CDM project, and then many more months to receive the first transfer of carbon credits. We therefore also took into consideration alliances associated with CDM projects that were under registration, or under validation, as information or knowledge from such alliances would influence carbon emissions reductions from their registered projects.

We computed alliances using binary coding, and following Huisman (2009), we imputed zeros for all firms that did not have alliances or partners in a given year. To account for ego networks of different sizes, we computed normalized degree centrality scores for each firm using UCINET 6 (Borgatti, Everett, & Freeman, 2002).

Structural Embeddedness

Following Gulati and Gargiulo (1999) and Polidoro et al. (2011), we operationalized structural embeddedness by measuring the “extent to which a given pair of organizations shares common partners from past ties” (Gulati & Gargiulo, 1999, p. 1463), by computing the number of common partners of CDM project participants prior to the observation year.

Structural Constraint

This variable is computed following Burt (1995). It measures the degree to which an ego’s alters are connected with each other, and indicates fewer structural holes for egos. Higher values indicate more structural constraint and less structural holes (Burt, 1995). UCINET 6 (Borgatti et al., 2002) was used to compute this variable.

Control Variables

As we were interested in examining the relationship of network structure to the firm’s carbon performance, we needed to control for the carbon market and other organizational factors that could influence performance outcomes of the focal firms.

We controlled for the effect of carbon price in secondary markets as it may affect the focal firm’s environmental investments. CERs price information before 2008 was not available as CERs only started to be traded in the secondary carbon markets in late 2007. We took an average of the spot EUA (European Union Allowance) prices, on which CERs prices are based, from 2005 to 2009, as proxy for CER prices. Data were obtained from Bluenext (European Emissions Exchange).

We also controlled for the effects of the focal firm’s characteristics and financial resources (George, 2005; Moon & deLeon, 2007; Polidoro et al., 2011; Uzzi, 1996) by including in the statistical analysis the following variables: firm age (measured by calculating the difference between the observation year and the firm’s founding year), firm listing (measured as a dummy variable: 1 for publicly listed firms and 0 otherwise), return on assets (net profit-to-assets ratio calculated for the year prior to the observation), firm size (total sales estimated for the year prior to the observation), and organizational slack (debt-to-equity ratio calculated for the year prior to the observation).

Model Specification and Estimation

Our primary concern was to measure carbon performance of firms over two different emission trading periods (Betz & Sato, 2006). As firms have differential strategic choices to join the carbon market network (Lutken & Michaelowa, 2008), this may bias the relationship of our computed network measures to a firm’s carbon performance. To address this possible self-selection bias, we followed Earnhart and Lizal (2006) and implemented the Heckman (1979) two-stage procedure. First we ran a Probit model to compute the firm’s probability of joining the carbon market network, by constructing a dummy variable that takes the value of 0 if in a particular year the given firm has an alliance with others, and 1 otherwise (see Table 2). We used carbon price, firm size, firm age, return on assets, and firm listing, as independent variables. In addition to these variables, we also included two variables, diversification degree (coded 0 if the firm has only one business activity, 1 the firm’s business activities are in the same three-digit Standard Industrial Classification (SIC) code, 2 in the same two-digit SIC code, 3 in the same one-digit SIC code, and 4 if the firm’s SIC codes are totally different) and CDM project size (measured as a dummy variable: 1 for large project size and 0 otherwise), that are likely to influence the firm’s probability of joining the carbon market network. These latter two variables are excluded from the second stage models of Table 4. This procedure allows to differentiate between the first-stage Probit and second-stage models. The first stage Probit model (χ² = 46.41, p < .01) predicts a firm’s probability of joining the market network with a success rate of 75% (see Table 2).

Table 2.

Results of the Probit Model Predicting Firm’s Probability of Joining the Carbon Market Network.

Variables	Model	% Change^a
Carbon price^b	−0.245 (0.046)**	−0.078
Firm age^b	−0.439 (0.425)	−0.139
Firm size^b	0.273 (0.131)*	0.087
Return on assets	−1.097 (0.490)*	−0.348
Organizational slack	−0.012 (0.060)	−0.004
Diversification degree	0.073 (0.074)	0.023
Project size	−0.264 (0.234)	−0.086
Firm listing	−0.138 (0.239)	−0.043
Constant	−0.454 (0.944)
Percentage to network	70
Percentage correctly predicted	75
Pseudo R²	.17
χ²	46.406
Log likelihood	−111.187

Note. We dummy coded the dependent variable as follows: It takes the value of 0 if in a particular year, the focal firm has an alliance with others, and 1 otherwise. Number of firms = 44; number of firm-year observations = 220; observation period: 2005-2009. Standard errors are in parentheses.

Change in probability (network) per unit change in predictors.

Logarithmic transformation.

†

p < .1. *p < .05. **p < .01.

Second, based on resulting coefficient estimates, we computed inverse Mills ratio for each firm in each year, and included this as regressor in all estimated models to control for self-selection bias (see Table 4). In the second-stage models, the sample of 44 Indian focal firms/egos included all annual observations within the 2005-2009 period. This expansion resulted in a panel data set of 220 firm-year observations from 2005 to 2009. The correlation matrix suggested relatively high correlations (at p < .05) between some variables, in particular amongst network position, structural embeddedness, firm size and firm age, and at the same time between carbon performance and structural embeddedness (see Table 3). However, variance inflation factors for all variables, and in all second-stage models were below the rule-of-thumb value of 10 (Phelps, 2010), that is, there was no issue of multicollinearity.

Table 3.

Descriptive Statistics and Correlations.

No.	Variable	Mean	SD	Minimum	Maximum	1	2	3	4	5	6	7	8
1	Carbon performance^a	10.18	1.65	5.44	15.99	—
2	Network position	0.36	1.04	0.00	10.15	.26**	—
3	Structural embeddedness	0.50	1.73	0.00	15.00	.34**	.23**	—
4	Structural constraint	0.55	0.45	0.00	1.13	−.07**	−.03**	−.15**	—
5	Carbon price^a	0.84	2.70	−3.90	3.11	−.11	.12**	−.07	−.30	—
6	Firm age^a	1.42	0.32	0.70	2.08	−.16**	.01	.10^†	−.09	−.04	—
7	Firm size^a	7.97	0.97	5.11	10.14	.17^†	.07	.11	.07	−.08	.52**	—
8	Return on assets	0.05	0.10	−0.22	0.95	.13	−.01	.02	.05^†	−.05	.15	.24**	—
9	Organizational slack	1.75	1.99	0.00	14.74	−.10	−.06	−.05	.02	.00	.01	.01	−.34**

Note. Number of firms = 44; number of firm-year observations = 220; observation period: 2005-2009.

Logarithmic transformation.

†

p < .1. *p < .05. **p < .01.

We ran fixed-effects panel regression, as suggested by the Hausman specification test (χ² = 265.56, p < .01). The Wald, Wooldridge, and Pesaran tests suggested issues of heteroskedasticity (χ² = 3.6, p < .01), autocorrelation (F = 184.68, p < .01), and contemporenous correlation (p < .01), respectively, in our panel data. Such issues, though common features in panel data but rarely addressed (De Hoyos & Sarafidis, 2007; Hoechle, 2007), were expected in our case due to sluggishness in network data (Gujarati & Porter, 2008), common shocks (De Hoyos & Sarafidis, 2007) due to inclusion of two different emission-trading phases (Betz & Sato, 2006), and in particular due to network effect (Steglich, Snijders, & Pearson, 2010). To address these issues, we estimated models using both fixed-effect panel regression and method proposed by Hoechle (2007). In fixed-effects panel regression, we used a cluster robust option controlling for the firm industry (measured by the main four-digit SIC code). Clustered standard errors are robust to heteroskedasticity and autocorrelation only, whereas Driscoll–Kraay standard errors are robust to cross-section dependence as well (see method proposed by Hoechle, 2007). As a consequence, both types of robust standard errors in the model estimation were applied.

Following Earnhart and Lizal (2006), we included respectively year dummies in all models to control for the unobserved exogenous effects and inverse Mills ratio to control for self-selection bias. We firstly estimated a model with control variables, inverse Mills ratio, and year dummies using fixed-effects panel regression (see Model 1). In the next stage (see Model 2), the full model was estimated with main effects, that is, network position, structural embeddedness, and structural constraint. Finally, we estimated Model 3 using the Hoechle (2007) method. In this last model, we computed Driscoll–Kraay robust standard errors. We ran a postestimation Ramsey test (F = 2.40, p < .1), which suggested that there was no omitted variable bias in either of our full models (2 and 3).

Results

In the following pages, we will explain the results of Model 3, (R² = 0.533; F = 171.25, p < .01), which is an optimal model relative to the baseline Model 1 (R² = .423; F = 13.81, p <.01), or Model 2 (R² = .533; F = 15.59, p < .01), which does not account for cross-section dependence (see Table 4).

Table 4.

Results of Fixed-Effects Panel Regressions Predicting Firm’s Carbon Performance.

Variables	Model 1^a	Model 2^a	Model 3^b
Carbon price^c	−0.276 (0.118)*	−0.485 (0.147)**	−0.485 (0.085)**
Firm age^c	−4.669 (3.871)	−5.836 (3.891)	−5.836 (3.964)
Firm size^c	0.834 (0.572)	0.468 (0.448)	0.468 (0.156)**
Return on assets	−0.261 (0.289)	−0.320 (0.381)	−0.320 (0.434)
Organizational slack	0.065 (0.029)*	0.040 (0.024)	0.040 (0.011)**
Network position		0.545 (0.221)*	0.545 (0.095)**
Structural embeddedness		0.205 (0.102)^†	0.205 (0.044)**
Structural constraint		−0.328 (0.207)	−0.328 (0.228)
Inverse Mills ratio	0.455 (0.401)	0.792 (0.363)*	0.792 (0.382)*
Year dummy included	YES	YES	YES
Constant	9.184 (6.379)	12.950 (6.122)*	12.950 (6.651)^†
R ²	.423	.533	.533
Adjusted R²	.30	.49	.49

Note. Number of firm-year observations = 154; number of firms = 44; observation period: 2005-2009.

Clustered standard errors are in parentheses.

Driscoll-Kraay standard errors are in parentheses.

Logarithmic transformation.

†

p < .1. *p < .05. **p < .01.

Hypothesis 1 predicted a positive relationship between a firm’s network central position and carbon performance. We found full support for this relationship as network position affects carbon performance significantly and positively (β = 0.545, p < .01). Results also supported Hypothesis 2 as structural embeddedness was found to have a significant and positive relationship with carbon performance (β = 0.205, p < .01). Hypothesis 3 predicted a negative relationship between structural constraint and carbon performance. Results did not support this hypothesis.

As regards control variables, carbon price (β = −0.485, p < .01) was found to negatively influence carbon performance, whereas firm size (β = 0.468, p < .01) and organizational slack (β = 0.039, p < .01) affect carbon performance positively (see Table 4 for more details).

We estimated the model with alternative specifications to check the robustness of the results. We ran models with density, while controlling for network size following Phelps (2010), to estimate the effect of structural constraint on carbon performance; we also checked nonlinearity of main effects using both fixed-effects and the Hoechle (2007) method. The results were found to be consistent as reported in Models 2 and 3, without significant improvement in the models.

Following Earnhart and Lizal (2006), we ran Granger causality tests to establish the direction of causality. We firstly implemented the Fisher unit root test to confirm that all variables followed one trend (Baltagi, 2009). We then ran the Granger causality tests by including three lagged periods of all variables. Based on this combination, we found no evidence that carbon performance Granger-cause network structure and other predictors. However, we found that some predictors Granger-cause the dependent variable. In particular, we found that structural embeddedness and organizational slack Granger-cause (at p < .05) carbon performance, while some other predictors and control variables provide weak evidence of causal links with the dependent variable.

Discussion and Conclusion

In this article, we have introduced ego-network structure—and specifically the concepts of centrality, structural embeddedness, and structural constraint—into the literature on carbon performance of the firm, and chosen the CDM market to test our hypotheses. Our main results show that a central position of the focal firm in the carbon market and its degree of network embeddedness affect its carbon performance positively. It was not possible, however, to confirm whether structural constraint negatively affects carbon performance.

Contrary to our expectations, we found a negative relationship between carbon price in the secondary market and the carbon performance of firms. In other words, if the carbon price drops, carbon supply will increase. Such a finding may corroborate the arguments of Kolk and Pinkse (2005) that most of the carbon buyers in the market use emissions trading as a compensation strategy as they gather carbon credits to offset the cost of reducing emissions themselves.

From a theoretical perspective, the research presented here contributes to the understanding of the influence of network characteristics on carbon performance of the firm. This has been done via investigation into ego network ties and resulting responsible behavior that were hitherto underexplored areas in the literature on corporate environmental performance (Wassmer et al., 2012). Contrary to some research that found no evidence of correlation between network embeddedness of firms and their responsible behavior (O’Shaughnessy, Gedajlovic, & Reinmoeller, 2007), we empirically demonstrate, while controlling for contextual contingencies, the relationship between network embeddedness of firms and their resulting carbon performance, thereby affirming the effectiveness of the network lens in devising or evaluating strategies for climate change.

Moreover, the burgeoning interest of academics to acknowledge the role of the carbon market in providing a diverse set of advantages to actors has so far focused on the pragmatic implications of that market. Previous research on the antecedents and consequences of carbon strategies has largely ignored the need to explore the relational and network theoretical underpinnings and consequences for the carbon market–based strategy that organizations pursue. In this article, we set out to expand the theoretical base for such attempts by introducing the relational and network paradigm, and to put into light emerging carbon market actors’ behaviors and their relational determinants.

Relating to the CDM literature specifically, using a relational and network approach, our findings contribute by providing improved structural understanding of a firm’s carbon performance. Schneider, Hoffmann, and Gurjar (2009), for example, limited their analysis to the effects of the firm’s characteristics alone on their emission reduction under CDM. In this article, we extend this approach by including certain network determinants of emission reduction in the specific setting of climate strategies of emerging economies’ firms. We demonstrate that social influence does not necessarily operate through and depend on actors alone, neither do actors act on islands pursuing self-centered objectives to improve carbon performance. On the contrary, our findings clearly show that social relations or pattern of ties between actors can and do positively or negatively affect carbon performance. We therefore contend that antecedents of carbon market actors’ behavior can be traced through an actor’s objective relational network, rather than more subjective social influence or actor characteristics. Such a view is supported by the relational paradigm (Emirbayer, 1997), which gives primacy to social relations between actors to comprehend social processes and actors’ behavior.

The above theoretical developments and associated findings are of importance also to practitioners, from a managerial perspective, in that they provide previously unidentified guidelines on one hand for the improved design of climate strategies per se―by government, for example―and on the other for firms involved in such schemes, and associated performance issues and implications. We demonstrate that a central network position in the carbon market brings with it advantages to influence and to access information. Furthermore, the negative effects due to asymmetric partnerships could be avoided by being more structurally embedded or building more common third-party ties. This suggests that beside forging formal alliances through joint ventures to access technology and other competitive resources, organizations may achieve their objectives to improve their environmental performance via building and managing the alliance network, and the targeting of specific network characteristics. Such a view implies expanding managerial focus from that of developing environmental resources to that of proactively managing the relationships with partners rather than simply responding to stakeholders.

Our research has some limitations. We chose to focus on alliances between project participants in the CDM market. In an attempt to be clear as to chosen boundaries, we concentrated on the focal firms’ carbon performance and selected long-term interorganizational partnerships of our sample firms in the CDM primary market. The possibility that focal firms’ alliances beyond the CDM market may influence their carbon performance cannot be ruled out. This is a network boundary issue, a constant challenge in network research. Furthermore, organizational learning is a continuous process, and carbon performance can be taken as representing a process innovation or an “ecopreneurial venture” (Schaltegger & Wagner, 2011). It may consequently be necessary to control for the effect of a firm’s absorptive capacity (Ahuja, 2000; Cohen & Levinthal, 1990) as this may cause differential innovation outcomes. This possibility cannot be ruled out. We nonetheless attempted to address this issue by including ego firms’ financial and organizational characteristics. For future research, additional access to information on the sample firms’ R&D expenditures and patents would help here.

Recent interest by researchers (Kolk & Mulder, 2011) highlights the opportunities the CDM market offers to all stakeholders. The network perspective we have developed in this article may prove helpful for professionals and managers to look beyond the firm-specific and dyadic levels (Rowley, 1997) in devising strategies for climate change. The carbon market offers funding opportunities to entrepreneurs to realize their ventures in addressing social issues in emerging economies while generating profits. The network theoretical framework offers an additional lens to better understand the structural dynamics of emerging economies’ entrepreneurial motivations and their rewards, which may later be applied in devising “bottom of the pyramid” strategies (Rivera-Santos & Rufín, 2010) for multinational firms.

Our study, for the purposes of future research directions, can be placed in the “network flow” theoretical framework (Borgatti & Halgin, 2011). It contributes from a structural perspective and focuses on exploring the consequences of network structure on firm performance. This is usually considered as a deterministic approach (Kilduff & Brass, 2010) as it ignores the impact of actors’ chosen strategy on economic outcomes. Further research is needed to explore the role of agency in shaping the networks and network influence (Dacin, Dacin, & Tracey, 2011; Emirbayer & Goodwin, 1994), rather than just adopting a structural perspective. Our proposed network theoretical framework suggests one of the possible carbon strategies that can be implemented, without recourse to financial resources only, that is, actively manipulating and structuring the carbon market network to one’s own advantage. We implicitly assume actors’ strategic choices created the very network that enabled or constrained them. However, as we focused only on the structural influence of the firm’s behavior, we did not theorize and model for chosen corporate strategies and network formation processes. Some scholars (Archer, 1995) demonstrate analytical dualism of structure and agency—that is, both are causal mechanisms. This, together with our proposed network theoretical framework, suggests one more possible direction for future research: to explore the interplay, and the dynamics of the firm’s strategies for climate change and the resulting network. Network researchers argue, for instance, that network dynamics could be driven by transitivity, homophily, and assortative matching (Rivera, Soderstrom, & Uzzi, 2010). New stochastic actor-based models (Snijders, Van de Bunt, & Steglich, 2010; Steglich et al., 2010) for network dynamics also offer opportunities to distinguish which mechanisms play a major role in the coevolution of actors’ network characteristics and their attributes or behaviors.

Moreover, our result shows that structural constraint positively affects carbon performance. The literature puts forth convincing, although competing, arguments in favor of structural constraint (or network closure) in generating trust and tacit knowledge (Coleman, 1990) and about the merits of structural holes (less structural constraint) for new information and control (Burt, 1992). However, some authors put into evidence that both constraint (and network closure) and structural holes (absence of structural constraint) affect performance, with certain contingencies (Battilana & Casciaro, 2012; Soda, Usai, & Zaheer, 2004). Future research should look into whether past or current structural constraint affects carbon performance, and whether this is contingent on the purpose, and degree of innovation to reduce carbon emissions.

Footnotes

Acknowledgements

We thank Professor MarkStarik, Editor in Chief of Organization & Environment, and the two anonymous reviewers for their constructive remarks and helpful feedback. We are also grateful to Emmanuelle Reynaud, Anne Norheim-Hansen, and Ulrich Wassmer for their insightful comments on an earlier draft of this article. This article draws on a portion of a paper selected in the Best Paper Proceedings of the 71st Academy of Management Conference, Boston, August 2012.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research project is partly funded by the Research Chair “Responsible Purchasing in the Network Environment” (KEDGE Business School, France).

Notes

Author Biographies

Naeem Ashraf is an Assisant Professor of Strategy at Lahore University of Management Sciences (Pakistan). His research interests include inter-organizational networks, corporate social responsibility, and markets and ethics.

Pierre-Xavier Meschi is Professor of Strategy at Aix Marseille Université and Skema Business School (France). He is currently the head of the CERGAM research center (Aix Marseille Université). His research interests include international expansion strategy, entry modes (international joint ventures and M&As), foreign direct investment in emerging economies, and country risk.

Robert Spencer is Professor of Marketing and Purchasing, and Chair in Responsible Purchasing at Kedge Business School (France). His main research interests include BtoB networks, market-shaping, and responsible purchasing.

References

Ahuja

(2000). Collaboration networks, structural holes, and innovation: A longitudinal study. Administrative Science Quarterly, 45, 425-455.

Akiyama

(2010). CSR and inter-organisational network management of corporate groups: Case study on environmental management of Sekisui House Corporation Group. Asian Business & Management, 9, 223-243.

Albino

Dangelico

R. M.

Pontrandolfo

(2012). Do inter-organizational collaborations enhance a firm’s environmental performance? A study of the largest U.S. companies. Journal of Cleaner Production, 37, 304-315.

Archer

M. S.

(1995). Realist social theory: The morphogenetic approach. Cambridge, England: Cambridge University Press.

Bae

Gargiulo

(2004). Partner substitutability, alliance network structure, and firm profitability in the telecommunications industry. Academy of Management Journal, 47, 843-859.

Baltagi

B. H.

(2009). A companion to econometric analysis of panel data (1st ed.). New York, NY: Wiley.

Bansal

Knox-Hayes

(2013). The time and space of materiality in organizations and the natural environment. Organization & Environment, 26, 61-82.

Bansal

Roth

(2000). Why companies go green: A model of ecological responsiveness. Academy of Management Journal, 43, 717-736.

Battilana

Casciaro

(2012). Change agents, networks, and institutions: A contingency theory of organizational change. Academy of Management Journal, 55, 381-398.

10.

Berchicci

King

(2007). Postcards from the edge: A review of the business and environment literature. Academy of Management Annals, 1, 513-547.

11.

Betz

Sato

(2006). Emissions trading: Lessons learnt from the 1st phase of the EU ETS and prospects for the 2nd phase. Climate Policy, 6, 351-359.

12.

Borgatti

S. P.

Everett

M. G.

Freeman

L. C.

(2002). UCINET for Windows [Software for social network analysis]. Harvard, MA: Analytic Technologies. Retrieved from https://sites.google.com/site/ucinetsoftware/home

13.

Borgatti

S. P.

Foster

P. C.

(2003). The network paradigm in organizational research: A review and typology. Journal of Management, 29, 991-1013.

14.

Borgatti

S. P.

Halgin

D. S.

(2011). On network theory. Organization Science, 22, 1168-1181.

15.

Borgatti

S. P.

Mehra

Brass

D. J.

Labianca

(2009). Network analysis in the social sciences. Science, 323, 892-895.

16.

Brohe

Eyre

Howarth

(2009). Carbon markets: An international business guide. New York, NY: Routledge.

17.

Burt

R. S.

(1992). The social structure of competition. In Nohria

Eccles

R. G.

(Eds.), Networks and organizations: Structures, form and action (pp. 397-429). Cambridge, MA: Harvard Business School Press.

18.

Burt

R. S.

(1995). Structural holes: The social structure of competition. Cambridge, MA: Harvard University Press.

19.

Burt

R. S.

(2005). Brokerage and closure: An introduction to social capital. Oxford, England: Oxford University Press.

20.

Busch

Hoffmann

V. H.

(2011). How hot is your bottom line? Linking carbon and financial performance. Business & Society, 50, 233-265.

21.

Christmann

(2000). Effects of “best practices” of environmental management on cost advantage: The role of complementary assets. Academy of Management Journal, 43, 663-680.

22.

Cohen

W. M.

Levinthal

D. A.

(1990). Absorptive capacity: A new perspective on learning and innovation. Administrative Science Quarterly, 35, 128-152.

23.

Coleman

(1990). Foundations of social theory. Cambridge, MA: Belknap.

24.

Dacin

M. T.

Dacin

P. A.

Tracey

(2011). Social entrepreneurship: A critique and future directions. Organization Science, 22, 1203-1213.

25.

De Hoyos

R. E.

Sarafidis

(2007). Testing for cross-sectional dependence in panel-data models. Stata Journal, 6, 482-496.

26.

Delmas

M. A.

Toffel

M. W.

(2008). Organizational responses to environmental demands: Opening the black box. Strategic Management Journal, 29, 1027-1055.

27.

Delmas

M. A.

Toffel

M. W.

(2011). Institutional pressures and organizational characteristics: Implications for environmental strategy. In Bansal

Hoffman

A. J.

(Eds.), The Oxford handbook of business and the natural environment (pp. 229-247). Oxford, England: Oxford University Press.

28.

DiMaggio

P. J.

Powell

W. Z.

(1983). The iron cage revisited: Institutional isomorpohism and collective rationality in oragnizational fields. American Sociological Review, 48, 147-160.

29.

Earnhart

Lizal

(2006). Effects of ownership and financial performance on corporate environmental performance. Journal of Comparative Economics, 34, 111-129.

30.

Emirbayer

(1997). Manifesto for a relational sociology. American Journal of Sociology, 103, 281-317.

31.

Emirbayer

Goodwin

(1994). Network analysis, culture, and the problem of agency. American Journal of Sociology, 99, 1411-1454.

32.

Freeman

L. C.

(1978). Centrality in social networks conceptual clarification. Social Networks, 1, 215-239.

33.

George

(2005). Slack resources and the performance of privately held firms. Academy of Management Journal, 48, 661-676.

34.

Granovetter

M. S.

(1985). Economic action and social structure: The problem of embeddedness. American Journal of Sociology, 91, 481-510.

35.

Greve

H. R.

Baum

J. A. C.

Mitsuhashi

Rowley

T. J.

(2010). Built to last but falling apart: Cohesion, friction and withdrawal from interfirm alliances. Academy of Management Journal, 53, 302-322.

36.

Gujarati

Porter

(2008). Basic econometrics (5th ed.). New York, NY: McGraw-Hill/Irwin.

37.

Gulati

(1995). Social structure and alliance formation patterns: A longitudinal analysis. Administrative Science Quarterly, 4), 619-652.

38.

Gulati

(1999). Network location and learning: The influence of network resources and firm capabilities on alliance formation. Strategic Management Journal, 20, 397-420.

39.

Gulati

Gargiulo

(1999). Where do interorganizational networks come from? American Journal of Sociology, 104, 1439-1438.

40.

Hart

S. L.

(1995). A natural-resource-based view of the firm. Academy of Management Review, 20, 986-1014.

41.

Tian

Chen

(2007). Performance implications of nonmarket strategy in China. Asia Pacific Journal of Management, 24, 151-169.

42.

Heckman

J. J.

(1979). Sample selection bias as a specification error. Econometrica: Journal of the Econometric Society, 47, 153-161.

43.

Hoechle

(2007). Robust standard errors for panel regressions with cross-sectional dependence. Stata Journal, 7, 281-312.

44.

Hoffman

A. J.

(2005). Climate change strategy: The business logic behind voluntary greenhouse gas reductions. California Management Review, 47(3), 21-46.

45.

Hoskisson

R. E.

Eden

Lau

C. M.

Wright

(2000). Strategy in emerging economies. Academy of Management Journal, 43, 249-267.

46.

Huisman

(2009). Imputation of missing network data: Some simple procedures. Journal of Social Structure, 10, 1-29.

47.

Jennings

P. D.

Zandbergen

P. A.

(1995). Ecologically sustainable organizations: An institutional approach. Academy of Management Review, 20, 1015-1052.

48.

Jeswani

H. K.

Wehrmeyer

Mulugetta

(2008). How warm is the corporate response to climate change? Evidence from Pakistan and the UK. Business Strategy and the Environment, 17, 46-60.

49.

Kilduff

Brass

D. J.

(2010). Organizational social network research: Core Ideas and key debates. Academy of Management Annals, 4, 317-357.

50.

Kolk

Mulder

(2011). Regulatory uncertainty and opportunity seeking: The case of clean development. California Management Review, 54, 88-106.

51.

Kolk

Pinkse

(2005). Business responses to climate change. California Management Review, 47(3), 6-20.

52.

Koontz

T. M.

Thomas

C. W.

(2006). What do we know and need to know about the environmental outcomes of collaborative management? Public Administration Review, 66, 111-121.

53.

Lecocq

Ambrosi

(2007). The Clean Development Mechanism: History, status, and prospects. Review of Environmental Economics and Policy, 1, 134-151.

54.

Lee

S. Y.

(2012). Corporate carbon strategies in responding to climate change. Business Strategy and the Environment, 21, 33-48.

55.

Lin

(2012a). Cross-sector alliances for corporate social responsibility partner heterogeneity moderates environmental strategy outcomes. Journal of Business Ethics, 110, 219-229.

56.

Lin

(2012b). Strategic alliances for environmental improvements. Business & Society, 51, 335-348.

57.

Lutken

S. E.

Michaelowa

(2008). Corporate strategies and the clean development mechanism: Developing country financing for developed country commitments? Cheltenham, England: Edward Elgar.

58.

Meschi

P.-X.

Wassmer

(2013). The effect of foreign partner network embeddedness on international joint venture failure: Evidence from European firms’ investments in emerging economies. International Business Review, 22, 713-724.

59.

Moon

S. G.

deLeon

(2007). Contexts and corporate voluntary environmental behaviors examining the EPA’s green lights voluntary program. Organization & Environment, 20, 480-496.

60.

Moreno

J. L.

Jennings

H. H.

(1934). Who shall survive? A new approach to the problem of human interrelations. Washington, DC: Nervous and Mental Disease Publishing Co.

61.

O’Shaughnessy

Gedajlovic

Reinmoeller

(2007). The influence of firm, industry and network on the corporate social performance of Japanese firms. Asia Pacific Journal of Management, 24, 283-303.

62.

Paulsson

von Malmborg

(2004). Carbon dioxide emission trading, or not? An institutional analysis of company behaviour in Sweden. Corporate Social Responsibility and Environmental Management, 11, 211-221.

63.

Petersen

H. L.

Escobar

L. F.

Espinoza

J. L.

Vredenburg

(2006). Financing technological improvements and firm competitive advantage through the Kyoto protocol’s clean development mechanism (CDM): A Latin American example. Latin American Business Review, 6(4), 23-43.

64.

Phelps

C. C.

(2010). A longitudinal study of the influence of alliance network structure and composition on firm exploratory innovation. Academy of Management Journal, 53, 890-913.

65.

Pinkse

(2007). Corporate intentions to participate in emission trading. Business Strategy and the Environment, 16, 12-25.

66.

Podolny

J. M.

(1993). A status-based model of market competition. American Journal of Sociology, 98, 829-872.

67.

Polidoro

Ahuja

Mitchell

(2011). When the social structure overshadows competitive incentives: The effects of network embeddedness on joint venture dissolution. Academy of Management Journal, 54, 203-223.

68.

Porter

M. E.

Linde

C. V. D.

(1995). Toward a new conception of the environment-competitiveness relationship. Journal of Economic Perspectives, 9, 97-118.

69.

Pulver

(2007). Making sense of corporate environmentalism: An environmental contestation approach to analyzing the causes and consequences of the climate change policy split in the oil industry. Organization & Environment, 20, 44-83.

70.

Rivera

M. T.

Soderstrom

S. B.

Uzzi

(2010). Dynamics of dyads in social networks: Assortative, relational, and proximity mechanisms. Annual Review of Sociology, 36, 91-115.

71.

Rivera-Santos

Rufín

(2010). Global village vs. small town: Understanding networks at the Base of the Pyramid. International Business Review, 19, 126-139.

72.

Roome

(2001). Conceptualizing and studying the contribution of networks in environmental management and sustainable development. Business Strategy and the Environment, 10(2), 69-76.

73.

Rowley

T. J.

(1997). Moving beyond dyadic ties: A network theory of stakeholder influences. Academy of Management Review, 22, 887-910.

74.

Sariannidis

Zafeiriou

Giannarakis

Arabatzis

(2013). CO2 emissions and financial performance of socially responsible firms: An empirical survey. Business Strategy and the Environment, 22, 109-120.

75.

Schaltegger

Wagner

(2011). Sustainable entrepreneurship and sustainability innovation: Categories and interactions. Business Strategy and the Environment, 20, 222-237.

76.

Schneider

Hendrichs

Hoffmann

V. H.

(2010). Navigating the global carbon market: An analysis of the CDM’s value chain and prevalent business models. Energy Policy, 38, 277-287.

77.

Schneider

Hoffmann

V. H.

Gurjar

B. R.

(2009). Corporate responses to the CDM: The Indian pulp and paper industry. Climate Policy, 9, 255-272.

78.

Schneider

Holzer

Hoffmann

V. H.

(2008). Understanding the CDM’s contribution to technology transfer. Energy Policy, 36, 2930-2938.

79.

Seres

Haites

Murphy

(2009). Analysis of technology transfer in CDM projects: An update. Energy Policy, 37, 4919-4926.

80.

Sharma

(2000). Managerial interpretations and organizational context as predictors of corporate choice of environmental strategy. Academy of Management Journal, 43, 681-697.

81.

Snijders

T. A. B.

Van de Bunt

G. G.

Steglich

C. E. G.

(2010). Introduction to stochastic actor-based models for network dynamics. Social Networks, 32, 44-60.

82.

Soda

Usai

Zaheer

(2004). Network memory: The influence of past and current networks on performance. Academy of Management Journal, 47, 893-906.

83.

Steglich

Snijders

T. A. B.

Pearson

(2010). Dynamic networks and behavior: Separating selection from influence. Sociological Methodology, 40, 329-393.

84.

Stern

(2007). The economics of climate change: The Stern review. Cambridge, England: Cambridge University Press.

85.

Swaminathan

Moorman

(2009). Marketing alliances, firm networks, and firm value creation. Journal of Marketing, 73(5), 52-69.

86.

Uzzi

(1996). The sources and consequences of embeddedness for the economic performance of organizations: The network effect. American Sociological Review, 61, 674-698.

87.

Wang

Gao

(2013). Do greenhouse gas emissions affect financial performance? An empirical examination of Australian public firms. Business Strategy and the Environment. Advance online publication. doi:10.1002/bse.1790

88.

Wasserman

Faust

(1994). Social network analysis: Methods and applications. Cambridge, England: Cambridge University Press.

89.

Wassmer

Paquin

Sharma

(2012). The engagement of firms in environmental collaborations: Existing contributions and future directions. Business & Society. Advance online publication. doi:10.1177/0007650312439865