1In a context of increasing constraints in terms of environmental performance, firms have developed a range of innovative practices not only to meet regulatory requirements, but also to signal their commitment to the environment. In this matter, the adoption of environmental management systems (EMS) has the advantages of constituting a device that rationalizes and improves management control (Frondel et al., 2007; Wagner, 2008; Wagner, Llerena, 2011) while providing a tool for signaling the firm’s environmental commitment to its external stakeholders (Johnstone, Labonne, 2009; Laperche, Picard, 2013). In this context, we can assume that there is a need to develop appropriate organizational structures to support EMS (Lopez-Fernandez, Serrano-Bedia, 2007) and that EMS adoption is likely to generate organizational changes (e.g., learning by standardizing, new forms of authority by using the codified practices) that can affect managerial practices. In this sense, EMS may help to solve these two issues. Consequently, a better understanding of organizational changes associated with EMS adoption is crucial for managers and EMS promoters.

2The analysis of EMS adoption is a way to understands organizational processes — that could be deep — associated with the growing development of new environmental practices within the industrial firms. This study of environmental management systems (EMS) and organizational changes is also relevant because it has little been studied by works that focused on the technical dimension of environmental innovations.

3By definition, an EMS provides the requirements for an organization’s structure, responsibilities, practices, procedures, processes and resources for environmental management that institutionalizes, in the organization, the responsible corporate environmental management (Bansal, Bogner, 2002), such that a firm can reduce its negative environmental impact while improving management control (Nishitani, 2009). The two dominant standards of certifiable EMS, are a worldwide standard (ISO 14001) and an European standard (EMAS, Eco Management and Audit Scheme). Usually, the wider the standard is, the less stringent the requirements are. EMAS requires a full disclosure of firm’s environmental objectives and effective results, whereas ISO 14001 only demands disclosure of the firm’s environmental policy (Bansal, Bogner, 2002).

4The research on EMS has mainly focused on the analysis of the determinants of its adoption (Darnall et al., 2008), as well as the relationship between its implementation and firms’ environmental and economic performances (Bracke et al., 2008; Nishitani, 2011). However, the literature on the interaction between EMS and the organization of firms remains relatively scarce. Yet, this kind of innovation is important because it could imply many organizational changes. As various scholars, particularly in management studies, have noted, EMS and specifically certified standards, such as ISO 14001 or EMAS, are characterized by strong managerial and organizational requirements. These standards involve a commitment on the part of management, planning of the improvements, implementation of corrective and preventive actions, staff training and other activities (Rondinelli, Vastag, 2000; Reverdy, 2006). More generally, an EMS concerns various aspects of governance and the coordination of the different actors (Bloom et al., 2010; Martin et al., 2012), which are far from insignificant in terms of internal organization and external coordination arrangements.

5The objective of this study is to analyze EMS adoption and the associated organizational changes on an original and nationwide sample of French industrial firms. Our main hypothesis is that the organizational changes associated with EMS adoption include a process of internal change and adaptation, as well as interactions between firms and their external environment (Delmas, Toffel, 2008). With regards to the dynamics of internal change, the literature on environmental innovation provides an analytical framework for understanding the forms of governance adopted by organizations, as well as their informational and decision-making structure. These dynamics also involve external coordination mechanisms and more broadly, raise the question of the influence of a firm’s external environments in the process of adoption, particularly its industrial and regulatory environments (Depret, Hamdouch, 2009).

6From an empirical point of view, the influence of the adoption of an EMS on the organization has often been examined in management studies but seldom tested in economic models due to, among other factors, the lack of country-wide data on the characteristics of organizational changes. The originality of this paper is related to the uniqueness of the database used, which is the “Organizational Change and Computerization” survey (COI-TIC 2006) conducted in France by INSEE (National Institute for Statistics and Economic Studies) and the CEE (Centre for Employment Studies). Merged with the EAE 2005 survey (Annual Business Survey, INSEE), this original database provides information on the structures and strategies of French industrial firms, as well as on the processes of organizational change they have experienced between 2003 and 2006. The COI-TIC survey examines a sample of 3647 firms that is representative of all French industrial firms. The objective is to test, using a probit model, the relation between the adoption of an EMS and the different types of organizational changes undertaken by industrial firms. From the dataset, a distinction is made between the changes that occurred in firms that adopted an EMS before 2003 and firms that adopted an EMS between 2003 and 2006.

7The following section presents the theoretical framework and the hypotheses. Section 2 describes the data and presents the variables and econometric model. Section 3 is dedicated to the presentation of the results. Finally, the last section concludes the paper and discusses its contributions and limitations.

Theoretical framework and hypotheses: organizational changes and EMS adoption

8In line with the Porter Hypothesis (Porter, van der Linde, 1995), most studies on the impact of EMS have intended to analyze the impact of adoption on environmental performance and its often ambivalent effect on economic performance (Rehfeld et al., 2007; Rennings, Rammer, 2011; Wagner, 2008). However, the literature on the interaction between EMS and the organization of firms remains relatively scarce [1]. Yet, the adoption of an EMS constitutes an innovation that brings changes, depending on the intensity of the firm’s environmental goals (Bloom et al. 2010), the nature of the EMS (whether it involves a certification or not; Johnstone, Labonne, 2009) or the degree of complementarity between the environmental innovations and other types of innovations undertaken by the firm (Antonioli et al., 2013; Milgrom, Roberts, 1995). Laperche and Picard (2013) show notably the importance of the development of product-services associated to the management of the firms’ environmental strategies. Indeed, as pointed out by Avadikyan et al. (2001) and Antonioli et al. (2013), changes may be a necessary condition for the implementation of an EMS and to generate/ensure the environmental performance. Thus, EMS and organizational changes could have a bidirectional causality, in the sense that not only EMS adoption may produce organizational changes, but can be resulted from such changes in firm’s organization.

9An organizational change is defined as any change in the distribution of power, skills, information management and lines of communications (Greenan, 2003; Murphy, 2002), as well as in arrangements with external partners (Benghozi, 2001). This implies different changes in the hierarchical and communication structure, work and production organization, a reallocation of responsibilities and different changes in the firm’s external relationships (Antonioli, Mazzanti, 2009; Greenan, 2003). These external relations are particularly important given the fact that stakeholders’ requirements play a key role in the environmental practices of firms (Bracke et al., 2008; Nakamura et al., 2001; Nishitani, 2009; Gilli et al., 2014).

10In what follows, we highlight two sets of hypotheses based on the relationship between organizational changes and EMS adoption according to the firm’s internal organization and the coordination modes with external partners. A third set of factors is added, as control variables, in order to take into account the role of regulation and the external environment.

Data and methods

Empirical model

32We use a binary probit model to analyze organizational changes associated to EMS adoption. First, we estimate the factors characterizing the adoption of EMS (a general model), then the adoption of EMS before 2003, finally the adoption after 2003, i.e. during the 2003-2006 period. The model consists of a EMS adoption equation (1) which relates to the probability of adopting an EMS, by a firm i, that can be described as follows:

33

34The binary decision EMS_i to adopt an EMS, by a firm i, is modeled as the outcome of an unobserved latent variable EMS^*_i, and we observe that a firm adopts an EMS (EMS_i = 1 when EMS^*_i > 0. It is assumed that EMS^*_i is a linear function of a vector of internal organizational changes variables, IOC_i; the inter-firms relationships, IF_i; control variables related to the firm’s external environment, EE_i; other firm-specific characteristics, x_i; and finally, unobserved factors u_i, with u_i → N(0,1) (Greene, 2003).

35Probit models assume to follow a standard normal distribution of errors. This is the main difference with for instance, logit models, which assumed a logistic distribution of errors. In practical terms, the results of both these models for binary outcomes are very similar. We can obtain the coefficients of a logit by multiplying those of a probit by π/√3. However, in order to calculate the marginal effects, probit models are more appropriate. We have verified this by testing both models and the results of the logit model are quite similar.

Variables

36Dependent variable. From the COI-TIC database, we are able to test the dependent variable “EMS adoption” which indicates whether a firm possesses an environmental management system, as the ISO 14001 norm. As shown in table 1, 3221 firms have adopted an EMS by 2006 (15.23% of French industrial firms). A large majority of the adoption process took place before 2003 (11.43%) compared to those after 2006 (4.38%).

Table 1

Organizational characteristics of French industrial firms (in %)

Total industrial firms Total EMS adopters EMS adopters before 2003 EMS adopters during 2003-2006 Number of firms 21 151 3221 2417 926 % 100 15.23 11.43 4.38 Hierarchical levels: Stable 87.50 83.06 84.15 80.38 Increase 7.06 8.58 6.35 14.54 Decrease 5.44 8.35 9.51 5.09 Hierarchy implication in tasks: Stable 93.15 92.54 94.57 87.58 Increase 3.55 2.73 2.03 4.84 Decrease 3.30 4.73 3.40 7.58 Cross-sectional work teams to improve: new product conception 34.68 50.58 53.55 45.74 quality and production 48.65 64.40 64.31 64.23 customer relationships 29.80 42.40 42.60 44.02 Quality certification (during 2003-2006) 5.74 8.55 2.17 25.43 Offshoring 7.98 13.02 14.46 12.41 Increase in logistical arrangements 13.99 18.57 12.06 36.12 Increase of computer-based functions 8.80 18.57 6.08 11.10 Increase in contracts with customers 4,08 2.71 1.77 4.81 Increase in contracts with suppliers 8.11 8.43 6.70 12.45 Downstream dependency 38.16 47.82 48.00 48.68 Upstream dependency 31.00 30.09 31.27 26.90 Largest customer demands a quality norm 63.87 87.28 89.55 79.39 Largest supplier complies with quality norm 65.32 83.55 84.54 79.42 Size: 20 – 49 employees 54.12 31.79 32.27 33.56 50 – 249 35.63 38.82 37.80 41.01 250 – 499 5.82 13.03 12.63 13.30 500 or more 4.42 16.37 17.30 12.14 Group 50.74 68.79 68.78 67.14 Sector : Agro-food 11.48 15.09 14.29 15.30 Consumer goods 19.74 12.71 11.28 18.38 Automobile 2.59 5.30 5.70 3.57 Equipment goods 22.66 16.25 17.57 15.59 Intermediate goods 43.54 50.65 51.16 49.16

Organizational characteristics of French industrial firms (in %)

37Independent variables. The complete description of variables is shown in table 2. The variables in this table correspond to internal organizational changes, the coordination modes with external partners and others control variables. The questions asked in the survey best capture the point of view of company’s representatives and give a clear overview of the organization’s structure, management tools and organizational changes.

Table 2

Description of variables^*

Variables Definition Dependent variables EMS = 1 if the firm has an Environmental Management System (EMS), 0 otherwise EMS03 = 1 if the firm has adopted an EMS before 2003, 0 otherwise EMS06 = 1 if the firm has an EMS during 2003-2006 period, 0 otherwise Independent variables Internal organizational changes Hierarchical levels Qualitative variable of 3 modalities: Stable (reference), increase or decrease of the number of hierarchical levels in the firm during the 2003-2006 period Increase hierarchy’s implication Qualitative variable of 3 modalities: Stable (reference), increase or decrease of the implication of the hierarchy on firm’s tasks and monitoring of results during the 2003-2006 period Cross-sectional work teams to improve: -new product conception = 1 if the firm has implemented a cross-sectional working team to improve the conception of new products and R&D, 0 otherwise -quality and production = 1 if the firm has implemented a cross-sectional working team to improve quality and production, 0 otherwise -customer relationships = 1 if the firm has implemented a cross-sectional working team to improve its customer relationships, 0 otherwise Quality certification = Qualitative variable of 3 modalities: none (in reference), 1 if the firm declared to have carried a quality certification (ex. ISO 9001) before 2003, or 2 if the adoption took place during the 2003-2006 period Increase in logistical arrangements = 1 if the firm’s number of managerial practices (Traceability, Just in time, Value engineering, Supply Chain management) has increased during the 2003-06 period, 0 if stable or decrease Increase num. of computer-based functions = 1 if the number of functions managed by ICT systems (conception, purchases, sales, production human resources, finance, etc.), 0 if stable or decrease Offshoring = 1 if the firm has, during the 2003-2006 period, relocate part of its activity abroad, 0 otherwise Changes in inter-firms relations Increase in contracts with customers = 1 if the firm has increased its contractual practices with customers (certification, delivery delays or a customer service contract), 0 if stable or decrease Increase in contracts with suppliers = 1 if the firms has increased its contractual practices with suppliers (long term contracts, delivery delays or specifications), 0 if stable or decrease

Downstream dependency = 1 if the top three customers represent more than 50% of the firm’s total revenue, 0 otherwise Upstream dependency = 1 if the top three suppliers represent more than 50% of the firm’s total purchases, 0 otherwise Largest customer quality norm = 1 if the firm’s largest customers imposes its quality norm to the firm, 0 otherwise Largest supplier quality norm = 1 if the firm’s largest supplier complies with a quality norm demanded by the firm, 0 otherwise External environment and control variables Market concentration Logarithm of the C4 concentration ratio: cumulated market shares of the first four firms in the sector (at the NAF 220 level*) Exports Logarithm of the firm’s exports rate: exports/total revenue Industrial agglomeration effects Logarithm of fraction of workers in the “département” that an industry represent over total workers in the same geographical area, relative to the share of the whole industry in national employment Head office location Qualitative variable with 3 modalities for the location of the firm’s head office: Urban area (reference), Peri-urban area, Rural area Changes in regulation and norms = 1 if the firm’s activity has been strongly or very strongly affected by a change in regulations and norms, 0 if little or no effect Size Logarithm of the number of employees in the firm. Group = 1 if the firm is a subsidiary of a group, 0 if independent Sector Qualitative variable with 5 modalities: Agro-food (reference), Consumer goods Automobile, Equipment goods, and Intermediate goods

Description of variables^*

* NAF is the French nomenclature of activities, version 2003, first revision.

38More precisely, first we consider internal organizational changes related to the governance structure changes in Hierarchical levels and the Hierarchy’s implication, labor structure (the creation of Cross-sectional groups for new product conception, to improve quality and production and to improve customer relationships) and managerial, logistical and informational structures: Quality certification (before and after 2003), the increase of Logistical arrangements (traceability, just-in-time, value engineering, supply chain management) and the increase of Computer-based functions (conception, purchases, sales, production human resources, finance, etc.). We also consider whether a firm has relocated a part of its activities Offshore.

39Second, we capture the pressure and the changes in the coordination modes with external partners by the increase in Contracts with customers; the Downstream and Upstream dependencies (whether the firm’s three largest customers / suppliers represent more than 50% of the firm’s total revenue/ purchase), and the requirement of a Quality norm by customers and to Suppliers.

40Moreover, a third group of variables characterizes the roles of the external environment: the C4 Market concentration ratio; the Exports ratio and the Industrial agglomeration effects and the Head office location in order to control for the spatial factors. The effect of regulatory environment is captured by the Changes in regulations and norms. Finally, we introduce the size of the firm, whether they belong to a group and the sector as controls in the regressions.

Results

41The Organizational Change and Computerization survey (COI-TIC 2006) analyzes the changes that took place over the period 2003-2006 in all French industrial firms. These organizational changes pertain to firms’ internal organizational structures (hierarchical and decision-making structures and organizational practices) and external structures (inter-firms relationships). The three models test the dependent variables EMS = 1 (adoption behavior of all adopters, model 1), EMS03 = 1 (adoption process of adopters before 2003, model 2) and EMS06 = 1 (adopters after 2003, model 3).

42The adoption of an EMS is associated with different forms of governance dynamics. While there is a dominant tendency towards an increase in the number of hierarchical levels, once we take into account the period during which the adoption took place, we can bring to light a twofold change: a growing number of hierarchical layers in firms that adopted an EMS during the 2003-2006 period (model 3) and a reduction of hierarchical levels for those that adopted prior to that period (model 2).

43Thus, for the firms that adopted an EMS during the period 2003-2006, the adoption was associated with an increase in hierarchical levels and in the implication of the hierarchy in tasks. Some studies hypothesize that this increase is linked to temporary organizational changes such as the introduction of an environmental manager (Reverdy, 2005). As for the firms that adopted EMS before 2003, we observe that the changes in governance arrangements that are significantly associated with the adoption of an EMS are a decrease of hierarchical levels, but without a significant change in the hierarchy’s implications in the firm’s tasks. These different results highlight the role of EMS as a mechanism for reducing information costs, motivated by the need to reduce “problem solving” costs in the sense meant by Garicano (2000), and reveals a process of “flattering” hierarchical structures associated with an organizational learning process.

Table 3

Environmental Management System and organizational changes

Model 1 Model 2 Model 3 Dependent variables EMS EMS before 2003 EMS 2003-2006 M.E. S.E. M.E. S.E. M.E. S.E. Internal organizational changes Hierarchical levels: Stable Ref. Ref. Ref.  Increase 0.069*** (0.012) 0.023* (0.012) 0.046*** (0.009) Decrease 0.012 (0.012) 0.035** (0.012) -0.026*** (0.006) Hierarchy implication in tasks: Stable Ref. Ref. Ref.  Increase 0.008 (0.017) -0.024 (0.016) 0.033** (0.013) Decrease 0.043** (0.016) 0.004 (0.016) 0.014 (0.010) Cross-sectional work teams:  for new product conception 0.019* (0.008) 0.047*** (0.007) -0.011* (0.005) to improve quality and production -0.013 (0.008) -0.024*** (0.007) 0.004 (0.005) to improve customer relationships 0.034*** (0.008) 0.017* (0.007) 0.018*** (0.005) Quality certification: 0 (in reference)  Before 2003 0.228*** (0.007) 0.175*** (0.006) 0.067*** (0.004) 2003-2006 0.293*** (0.014) 0.040* (0.018) 0.312*** (0.019) Increase in logistical arrangements 0.010 (0.008) -0.044*** (0.008) 0.055*** (0.006) Increase of computer-based functions -0.036*** (0.011) -0.014 (0.011) -0.008 (0.006) Offshoring 0.035** (0.011) 0.052*** (0.010) 0.015* (0.007) Changes in inter-firms relations Increase in contracts with customers -0.070*** (0.015) -0.064*** (0.015) -0.027*** (0.007) Increase in contracts with suppliers -0.011 (0.010) -0.021* (0.010) 0.002 (0.006) Downstream dependency 0.032*** (0.006) 0.023*** (0.006) 0.015*** (0.004) Upstream dependency 0.004 (0.007) 0.012 (0.006) -0.010* (0.004)

Model 1 Model 2 Model 3 Dependent variables EMS EMS before 2003 EMS 2003-2006 M.E. S.E. M.E. S.E. M.E. S.E. Largest customer quality norm 0.097*** (0.007) 0.095*** (0.007) 0.009 (0.005) Largest supplier quality norm 0.018* (0.008) 0.017* (0.007) 0.004 (0.005) External environment and control variables Market concentration 0.033*** (0.005) 0.033*** (0.005) -0.001 (0.003) Exports 0.084*** (0.016) 0.053*** (0.015) 0.057*** (0.010) Industrial agglomeration effects 0.002*** (0.000) 0.000 (0.000) 0.001*** (0.000) Head office location: Urban area Ref. Ref. Ref.  Peri-urban -0.019* (0.008) -0.019* (0.007) 0.009 (0.006) Rural area -0.010 (0.008) -0.013 (0.007) 0.013* (0.005) Changes in regulation and norms 0.063*** (0.006) 0.047*** (0.006) 0.019*** (0.004) Size 0.053*** (0.003) 0.040*** (0.003) 0.011*** (0.002) Group 0.006 (0.007) -0.011 (0.007) 0.018*** (0.004) Sector: Agro-food Ref. Ref. Ref.  Consumer goods -0.001 (0.012) -0.030** (0.011) 0.040*** (0.010) Automobile 0.046* (0.020) 0.035 (0.019) -0.004 (0.012) Equipment goods -0.109*** (0.009) -0.083*** (0.009) -0.028*** (0.006) Intermediate goods -0.024* (0.009) -0.029*** (0.009) 0.005 (0.006) Observations 3647 3647 3647 Weighted data 21 151 21 151 21 151 % of correct predictions 79.16 82.48 93.36 Log likelihood -6972.01 -5892.39 -3088.58 Pseudo R2 (McFadden’s) 0.227 0.216 0.188 BIC 14190.06 12030.82 6423.216

Environmental Management System and organizational changes

M.E.: Marginal effects. S.E.: Standard errors. Significance levels. ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001

44Finally, the results show that the evolution of the number of hierarchical levels is associated with the implementation of formalized collective arrangements (“team-based”), whose objective varies according to the production-related issues to be solved (Garicano, Wu, 2012). Thus, the adoption of an EMS is, for late adopters, positively related to the creation of cross-sectional working teams to improve external relationships with customers. For firms that adopted an EMS prior to 2003, the establishment of cross-sectional working teams is broader and intended to both improve the design and R&D of new products and customer relationships. Thus, the changes in hierarchical levels (hypothesis 1) resulting from the introduction of an EMS is also complementary with the development of collective labor practices within the firm (hypothesis 2). These results are controlled by the size of the firm, which shows a positive and linear relationship to EMS adoption, regardless of the period. This positive effect of size with adoption process has been shown by the literature (Christmann, Taylor, 2001). Moreover, one could notice that belonging to a group has only a positive effect for firms adopting EMS after 2003.

45The other results are related to the formalization of organizational arrangements, such as quality certification, which are associated with EMS adoption for all firms and, more particularly, for the late adopters. The results confirm the joint effect of adopting a quality certification and EMS (hypothesis 3). For those firms that have adopted an EMS before 2003, the adoption of quality certification has a very positive and significant correlation when this adoption of quality certification takes place before 2003. We can conclude that the quality certification was achieved at the same time or probably before the adoption of their EMS.

46Regarding other informational mechanisms, the results tend to show a correlation between the implementation of an EMS and the formalization of practices (i.e. significant effect of increased logistical arrangements), which is coherent with the rationale of complementarity motivating the adoption of various organizational mechanisms within the firm during the observation period.

47Concerning the firm’s modes of coordination with external partners, which appear to have an increasingly structuring effect on the management of environmental issues (hypothesis 5), they seem strongly conditioned by the relationship with customers. The process of adopting an EMS by new adopters (Model 3) is particularly influenced by the fact that the firm is very dependent on its customers (i.e., the three main customers represent more than 50% of their turnover). The results also reveal the strong influence of the main client’s demand for quality standards. This effect is present for the entire population of firms adopting EMS. More generally, it is interesting to note that this dependence does not involve an increase in contractual relations and that an increase in contracts with customers (as in Rosen et al. 2000) and, to a lesser extent with suppliers, has a negative effect. These results are controlled by the sector in which firms operate. Firms operating in the consumer goods sector show a positive effect of EMS adoption after 2003, but this effect is negative (and significant) if the adoption took place before 2003. Firms in the equipment goods and intermediate goods sectors have similar results when the adoption takes place before 2003.

48With regards to the external environment as a factor of adoption, the survey allows us to determine whether the firm’s activity was affected by changes in its regulatory, market or spatial environment. Thus, whether the adoption took place before or during the observation period, we observe that changes in regulations and norms are the main factor for adopting an EMS. The firm’s location has also a significant influence on EMS. For new adopters, industrial agglomeration effects are found to have a positive effect on EMS. We also note that for new adopters, being located in rural areas has a positive effect, which is not the case for firms that adopted prior to 2003. Moreover, EMS adoption is strongly correlated with the firm’s openness to international markets. This openness plays in two ways. On one hand, exporting has a positive effect (which corresponds to the signal theory), in that an EMS is considered an environmental organizational innovation serving to signal the firm’s environmental performance. On the other hand, a relocation of some of the firm’s activities offshore is positively associated with the adoption of EMS. Taking into consideration the fact that organizational changes are influenced by the regulatory environment, stricter in terms of environment in developed countries such as France, we can assume that the implementation of an EMS may lead the firm to relocate its most polluting activities offshore to respond to institutional regulations in terms of environmental protection (hypothesis 4).

Conclusions

49The objective of this article was to study organizational changes associated to EMS adoption in firms. Our hypothesis is that this question requires the analysis of both the process of change and adaptation within firms but also the interactions between the organization of the firm and its environment. Empirically, the analysis is based on data from the “Organizational Change and Computerization” survey (COI-TIC 2006), conducted by the French Statistical Institutes on a representative sample of all French industrial firms.

50The results highlight various forms of organizational change associated with the adoption of an EMS, which concern both the internal organization of firms and their forms of coordination with external partners. In terms of governance, the adoption processes are correlated with, among other things, changes in the hierarchical levels associated with a development of collective labor practices. The formalization of practices also has an influence in terms of logistical arrangements and adoption of certification systems, which appears to be complementary with the adoption of EMS. Results from model 3 — through which we can simultaneously observe the adoption and organizational changes in the firm — highlight the correlation between firms’ interactions with their upstream and downstream partners and their involvement in a process of formalization of practices such as the adoption of quality certifications or other logistical arrangements.

51The adoption of an EMS appears to be closely related to organizational changes intended to respond to external demands from clients and partners in the same industry. The strong influence of customers and firms’ integration in international markets, which is correlated with a high sensitivity to changes in standards and regulations, relates to firms’ desire to signal their environmental quality in international markets and to various stakeholders (e.g., public authorities, customers). Organizational dynamics associated with the process of EMS adoption clearly seem to be related to coercive and normative mechanisms, which reveal the weight of institutional isomorphism processes in the adoption of EMS (Delmas, Toffel, 2008; DiMaggio, Powell, 1983).

52This study, conducted on a sample of all French industrial firms, highlights that EMS adoption is associated with organizational changes. It sheds light the importance of organizational processes and their coevolution with environmental innovative practices. This result is particularly relevant in policy making towards eco-innovation, because these policies are mostly oriented towards technical innovation in detriment of organizational innovations. Further research should focus on whether the intensity of these changes varies depending on the weak or strong perspective of the firm’s environmental performance (Bloom et al., 2010). This question of intensity and radicality of environmental innovation and its impact on organizational design — seldom analyzed in economic literature — constitutes a subject of future research and represents a major challenge in terms of industrial policy in the current economic and ecological context.