The nexus of environmental innovation and circularity: Evidence from European economies

ABSTRACT

This paper shows the importance of environmental innovation on the circularity performance of each country that creates favorable conditions to help countries move towards the circular economy. Six measures are used to reflect environmental innovation, including the percentage of enterprises that invest in environmental innovation, the percentage of enterprises implementing environmental innovation activities, the number of ISO 14001 certificates, patents related to environmental innovation, the total R&D personnel and researchers, and the amount of green early-stage investments. We provide empirical evidence that innovation in environmental activities is crucial for an economy’s circularity performance, especially in the long term. Environmental innovation is also an essential enable of circularity.

KEYWORDS:

• environmental innovation
• circularity performance
• economic circularity
• European countries

SUBJECTS:

• G3
• F2
• O16
• C33

1. Introduction

As humans continue to abuse and misuse natural resources, ecosystems worldwide are on the verge of collapse. The environment has already reached several tipping points, increasing the possibility that irreversible changes could occur in the future (Rockström et al., 2009). There has been a growing recognition of the necessity to decouple social and economic development from the extraction of resources and waste disposal in recent decades. In the 21st century, one of the most challenging issues appears to be balancing economic growth among competing countries while raising the standard of living for a population projected to exceed 10 billion in the mid-century. The OECD (2012) reports that the sustainable development of global biodiversity cannot be compromised by a limited number of natural resources.

A circular economy is gaining traction because of the limitations of conventional economies. Business organizations, industrial organizations, and national agendas have been urged to place closed-loop thinking at the core of the ‘circular economy’ (CE) (Preston, 2012). An economic theory derived from natural ecosystems, CE proposes the shift towards a permanently regenerative economy as a result of moving from a linear economy (which consists primarily of extractive, distributing, consuming, and disposing of goods that are unidirectional). CE is concerned with designing processes and products that minimize negative impacts on the environment and society, reduce non-renewable resources, eliminate dangerously hazardous substances, extend product life, and maximize the likelihood of recycling and reclaiming materials. As a result of the loop restoration and renewal, a sustainable economic development model is proposed for creating value. The goal is to shift the focus from ownership and material production to providing ‘services’ instead of residual waste or input into other processes. (EMF, 2012).

Tackling new problems may not always produce better results from a welfare or sustainability standpoint (Soete, 2013). Despite technological feasibility, ethically desirable behavior and environmental sustainability are not always synonymous. During the 20th century, carbon-intensive and extraction-based mass-production technologies raised fundamental questions regarding the concept of ‘progress’ in retrospect. As a result, some innovation concepts may be revised or even destroyed, as well as some intellectual property. A critical component of transition is changing ‘how we innovate’ (Schot & Steinmueller, 2018). An effective transition involves multiple actors, discrete actions, and extended activities over an extended period. The period during which new business models, products, and organizational structures emerge is characterized by an interrelated sequence of technological and non-technological innovations. Innovation concepts emerged around issues relating to transitions and societal challenges in general as the environment became a significant policy concern (Boons et al., 2013; Rennings, 2000). During the industrial era, there was no scope for this growing ‘pro-environment’ innovation agenda (Freeman & Soete, 1997).

With the help of organizations like the European Union (EU) (UNEP, 2014), CE has gained more significance as a research topic in recent years (European report, 2021). However, there remains confusion over what changes need to be made and how they can be achieved (Schulte, 2013), even though change is necessary for businesses and governments. While the sustainability transition is defined as a core driver of change, ‘eco-innovation’ (EI) plays a critical role in fostering change. According to Carrillo et al. (2010), innovation can take a variety of forms, including products, processes, marketing, and organizational benefits. Essentially, sustainability has been acknowledged by the European Commission as an integral component of achieving ‘environmental benefits’, which include more efficient use and consumption of resources (EC, 2012) and more competitive business models and technologies (Al-Ajlani et al., 2021). In the policy realm, transitioning to a circular economy has been referred to as a ‘catalyst’ (Russo, 2018).

Despite this, some studies have explored the importance of EI to a CE, as opposed to the CE and EI intersection approaches. To achieve a tech-economic paradigm shift, innovation activities should be aligned with sustainable paths (Mirata & Emtairah, 2005). EI can, in its current form, be used to facilitate the changes that are needed to deploy and reinforce the CE framework. However, there is still an insufficient discussion about how EI will facilitate a pro-CE transition. This paper contributes to the debate regarding sustainable transition by exploring a relationship between environmental innovation and circularity. The study is the first to provide a theoretical foundation for exploring the effect of countries’ EI and the performance of circularity. Environmental innovation is measured by six different measures, including the percentage of investments in environmental innovations, the percentage of the implementation of environmental innovations by enterprises, the number of ISO 14001 certifications, the number of green innovation-related patents, the number of researchers and R&D personnel, and the total amount invested in companies at the beginning of their development. Following Hong Nham and Ha (2022), we use a variety of measures in this article to evaluate the performance of circularity. There are a number of measures that are used in this article to assess the efficiency of the circularity process. Waste amount, patents relating to circularity, circular materials used, recycling rate, biowaste recycling, and e-waste recycling are measures of circularity. The sample of European countries between 2012 and 2019 is analyzed by combining various techniques and practical strategies. This database is chosen because this is the only region that provides comprehensive measures of circularity. Circularity performance is not represented by the other measures since they are quite out-of-date and can not reflect the diversity of circularity.

Research of a broader scope seems instrumental to understanding the dynamics of such a transformation, which will be supported by an integrated, dynamic business model and public policy welcoming to CE. As actors and institutions become more familiar with EI, their CE efforts can be better adjusted and calibrated. As a result of this analysis, a new generation of business actors will be capable of redesigning their business models in a sustainable manner from day one. Taking uncertainty and feedback loops into account is necessary for policymakers to implement initiatives that take into account EI, its relationship with CE, and its effect on education. To link the CE and EI, this paper proposes the notion of ‘clean congruence’ based on a conceptually-driven literature review.

A brief outline of the remaining sections is organized as outlined below. A literature review is conducted in Section 2, followed by hypotheses formulation. In Section 3, a model, data, and estimation strategy analysis is carried out. Besides the empirical results, a discussion is given in the following section. Research findings will be presented, and their policy implications will be analyzed in the final phase.

2. Literature review: circular economy and eco-innovation

2.1. A brief review of the link

The relationship between CE and EI is intuitively apparent, and it seems unlikely to achieve CE by itself. The specific details of this, however, remain to be established. In fact, not every aspect of EI is connected with a CE, as well as not every aspect of CE requires innovation. There is, however, bound to be some overlap between the two. A deeper understanding of these two concepts is essential in order to determine which innovations will be most appropriate for CE models and what technological changes are necessary to achieve a CE.

Since Joseph Schumpeter’s seminal writings in (1934), innovation has been recognized as more than newness (Schumpeter, 1934). This is, instead, the result of combining new ideas with production factors. Innovating goes beyond technical sophistication and involves adapting to a specific context, i.e. introducing an ingenious idea into an economic or institutional setting that is unique and sometimes quirky (Fagerberg et al., 2010). To put it another way, innovation does not only refer to scientific and technological advancements. Innovation has been viewed from this perspective as neither an automatic result of increased research and development nor passively influenced by market forces (Caraça et al., 2009).

Innovation does not always produce better results from a welfare or sustainability standpoint: Tackling new problems may not always produce better results (Soete, 2013). Technological feasibility does not always equate with ethically desirable behavior or environmental sustainability (UNEP, 2011). Carbon-intensive and extraction-based mass-production technologies of the 20th century raise fundamental questions regarding the concept of ‘progress’ in retrospect. This can lead to some revision of innovation concepts or even some destruction of intellectual property. A critical component of transition is changing ‘how we innovate’ (Schot & Steinmueller, 2018). A transition is a complex dynamic process in which several actors participate, discrete actions are taken, and activities continue for an extended period. Throughout this period, new business models, products, and organizational structures emerge, replacing or complementing incumbent ones characterized by an interrelated sequence of technological and non-technological innovations. Innovation concepts emerged around Issues relating to transitions and societal challenges in general as the environment became a significant policy concern (Boons et al., 2013; Rennings, 2000). During the industrial era, there was no scope for this growing ‘pro-environment’ innovation agenda (Freeman & Soete, 1997).

Innovation studies have seen some lexical variation as the environmental perspective has entered and diffused. The concept of innovation increasingly emerged as a dynamic process involving many different activities and evolving in real history (Castellacci et al., 2005). Besides research and development from suppliers in the ‘high-tech’ sector, social and cultural factors contributed to its development. Several studies have been conducted on sustainable development and transition that have been useful to innovation studies (Smith et al., 2010). As a socially embedded process, science and technology are emphasized in sustainability and transition studies; knowledge and mental maps are interconnected based on user skills, institutions/regulations, and infrastructure, and consumers’ expectations are shaped (Markard et al., 2012).

Sustainability transition approaches lack consensus regarding how they should be operationalized. Many viewpoints coexist together, and many theoretical approaches are relevant, such as evolutionary economics, management of niches (Kemp et al., 1998), and technology innovation systems or approaches that utilize multiple levels of analysis to understand sociotechnical changes (Geels, 2011) or innovative ecosystems are some examples. Even though creativity in terminology creativity can be viewed as an indication of the concept of restlessness, the ‘Proliferation of labels’ may hinder development in a field (Barney, 2017). “Environmental innovation refers to innovation that benefits the environment (van den Bergh et al., 2011). The notion of ‘sustainable innovation’ indicates an innovation that takes into account ecological, economic, and social factors, thus taking into account spatial, temporal, and cultural contexts, in addition to product and process innovations (Clark & Charter, 2007), as well as organizational models. According to Cuerva et al. (2014), ‘green innovation’ involves developing, improving, or creating new products or processes that promote environmental sustainability. In recent years, ‘Business model innovation’ has also gained traction within the semantic field. In this case, business model innovation enhances the organization’s ability to produce, deliver, and capture value to maximize its contribution to society (Bocken et al., 2014).

As the scope of (EI) has recently been expanded beyond its initial focus on ‘end of pipe’ technologies. As defined today, economic performance is defined as enabling an economic environment in which sustainable development is not hindered (i.e. economically, ecologically, and socially sustainable). The EC (European Commission) defines it more positively as a means of achieving or aiming for tangible progress in the pursuit of sustainable growth, which includes reducing the environmental impact, strengthening the resilience of the environment, or using natural resources more efficiently and responsibly (EC, 2011a). Furthermore, EI can facilitate increased competitiveness without adverse impacts on society and the environment (OECD, 2009) and is essential to sustainability (EC, 2011b). Few attempts to simplify and consolidate might be useful in this situation, despite some irreducible variability. For the purposes of this paper, environmental innovation will be used to refer to a standardized and comprehensive term for all environmentally sensitive innovations following the preference in the paper. Environmentally-conscious innovations have positive ecological effects and/or address environmental concerns. In light of the redesigning of studies on ‘transformative innovation’, we take EI as well as CE together as a fulcrum of realizing a new clean, coherent paradigm for tech-economic development (Schot & Steinmueller, 2018).

2.2. Refocusing and untangling the concept of eco-innovation

Understanding EI and its dimensions are necessary for policy-making, entrepreneurship, and academic research. An innovation that prevents, mitigates or recovers environmental damage can be defined as a broad, applicable, operational definition. According to Barbieri et al. (2022) and Veefkind et al. (2012), a definition of this type must include: (i) Environmentally friendly innovation (e.g. green innovation); (ii) Creating clean and efficient results for the market (e.g. environmental innovation); (iii) Socially responsible and lasting benefits (e.g. sustainable innovation); (iv) Change of the whole business model (i.e. innovation of a sustainable business model).

Besides analyzing EI from multiple perspectives, this definition provides a robust platform for identifying other sides of the issue. According to existing EI typologies and the innovation guidelines in the Oslo Manual (OECD, 2009), EI is examined from three different perspectives: innovations (targets), mechanisms of change (mechanisms), and the overall effects (impacts). Thus, incorporating this discussion into operational terms, EI can be defined as an innovation that: a) has a positive impact on the environment, b) provides cost-efficiencies, market enhancements, or regulatory considerations while preventing natural capital damage; c) creates new or enhanced products, services, technologies, organizational or marketing strategies; d) reflects radical change or incremental growth, and; e) involves a single or multiple actors.

2.3. Changing to a sustainable approach through CE-friendly approaches: A family of CE-friendly concepts

Global economies are largely modeled as exploitation of resources on an open-ended systems (take-make-dispose) in linear flow of contracts and regulations from tangible production to intangible contracts. Despite the fact that future-oriented debates have existed for quite some time (see Mendonça (2017), there has been no significant challenge to this linear model so far, in spite of growing awareness that resources on the earth cannot be infinitely used and a growing awareness of corporate social responsibility, among other concepts. As a result, moving away from existing models will be challenging because long-held technical systems, such as special interests and risk avoidance, will become even stronger (Markard et al., 2012; Schulte, 2013).

A high level of expectation has been established following the Paris Climate Conference 21 with the signing of the initial agreement by 175 governments (including the US and China as the original signatories), the European Union, and 174 countries (UNFCCC, 2016). In a fiercely dynamic global market, various actors have not aligned their interests as they strive to maintain competitiveness while dealing with the consequences of continuing environmental degradation. The new systems would focus on extending the lifespan of materials, reusing, remanufacturing, and recycling while decoupling development from resource consumption (UNEP, 2011). While it is increasingly recognized that change is necessary, the specific pathway of transition is much less clearly defined. Literature has proposed several approaches to frame the CE discussion, which has shaped our understanding of it today.

CE is based on several concepts that are not new. Utilizing waste materials from animals (e.g. pelts, blood, and bones) for clothing, shelter, weapons and jewelry dates back to the Neolithic period (Desrochers, 2000). Similar to today, cooperative arrangements between manufacturers and consumers were already in use during the 19th century through the exchange of byproducts in exchange for services (Desrochers, 2000, 2002). It emerged during the latter half of the twentieth century, along with concerns about resource exhaustion on a planetary scale that emphasized the necessity of finding a new equilibrium within the ‘System of ecological cyclicity’; Georgescu-Roegen’s (1986) view of economic complexity as a function of entropy. A full chapter by Pearce and Turner (1989) discusses CE as a label for the first time, arguing that environmental values are economically practical in light of Boulding and Georgescu-Roegen’s arguments that natural systems are capable of producing waste, but they absorb and recycle it as opposed to traditional open-ended economies. In the man-made economy, the authors argue for material flow circularly. With a system of loops that mimics nature, new inputs will be reduced, and the materials and waste of the environment will be depleted slower (by sourcing materials and eliminating waste). Resource consumption should not result in simply littering the environment or creating products that will be obsolete over time; instead, new resources should be created so that old one can be recycled.

Frosch and Gallopoulos (1989) popularized the concept of CE in industrial ecology, particularly in the US (1998). According to the literature on industrial ecology, there is an explicit recommendation that industrial organizations should mimic the strategies of natural systems. A key component of the theory is the concept of ‘material symbiosis’, which uses waste byproducts to produce materials for other processes and businesses (Kiefer et al., 2019). In Europe, many institutions have taken up the industrial symbiosis concept, gaining popularity. In this paper, the main idea is on identifying a ‘systems integration’ approach whereby firms share byproducts and close their material cycles, and this approach has been viewed as a direct component of the implementation of CE.

In response to the critique of ‘industrial capitalism’, opponents provided the notion of ‘natural capitalism’ in the late 1990s and early 2000s, which paradoxically both threatens and uses the environment for its resources (Hawken et al., 2010). By using more efficient manufacturing processes, better valuing materials, and reusing and recycling them, we can achieve environmental and economic benefits.

Among the key elements of CE is the ‘Cradle to Cradle’ approach, which emphasizes strict manufacturing needs to be transformed into self-feeding services nexus that work together to create jobs, save resources and reduce waste (W. R. Stahel & Reday-Mulvey, 1981). Eventually, it was realized that services have the potential to revitalize the economy. In 2010 and 2013, W. Stahel (2010, 2013) argued that by providing ‘services’, manufacturer can maximize the life of their products by minimizing the use of new inputs. As a result, both producers (who can maintain control their assets) and consumers (who only pay for the benefits) benefit from this.

In addition, several concepts that follow the 3Rs of ‘reduce, reuse, and recycle’ are discussed, as well as the conception of a closed loop, such as zero emissions or a system where everything is used, and cycles of nature are performed (Pauli, 2010). McDonough and Braungart (2010) have furthered the ‘cradle to cradle’ concept; and the concept of ‘zero waste’. A concept that emerged from multiple ideas and schools of thought, it has a broad range of applications today. As the CEs have different meanings and different responsibilities for different stakeholder groups, all of these contributions should be considered in their respective contexts (EIO, 2016).).

2.4. The formal concept of circular economy

As CE gained a new boost when it entered the policy arena. As an early adopter of CE initiatives, Germany used the ‘Closed Substance Cycle and Waste Management Act’ to eliminate waste disposal practices that harm the environment. Japan has established a legal framework that encourages a society based on recycling through the Basic Law for Establishment of the Recycling-Based Society of 2000. Moreover, it became more practical when the discussion began in China in 1998 and afterward when the central government formally adopted it in 2002. In the country, explicit policies were enacted regarding the CE for the first time.

During 2005 and 2007, circular economy pilots in two batches were conducted with the aim of ‘promote circular economy theory into action through industry sectors, key economic areas, and urban demonstration projects’. The Circular Economy Promotion Law was passed in 2008 and was in force in 2009 to achieve sustainable development, protect the environment, and promote energy efficiency (Geng et al., 2012). A key message of the 12th Five-Year Plan (2011E15) was the importance of eco-industrial.

In 2015, the EU launched a circular economy action plan (EC, 2015) to enter the western policy arena. In the last few years, think tanks and private institutions have been heavily involved in the CE study, including the Ellen MacArthur Foundation (EMF). There are several factors that appear to be contributing to this momentum. It has been noted that the CE concept provides both economic benefits and business solutions at the same time, making it more useful and tangible than other environmental conservation approaches (Sauvé et al., 2016). Several countries, including Finland, France, the Netherlands, Spain, and Sweden, have highlighted its potential to create jobs, improve resource productivity, improve the trade balance, and reduce CO2 production. The Groupo Interministerial Economia Circular, 2017 indicates that Portugal has also begun consultations concerning its CE Action Plan for 2017 through 2020. In addition to the concept being taken as a policy enactment device, available opportunities of funding under the EU Circular Economy Action Plan (EC, 2017) have been taken into account.

As a result, the CE cannot be considered a consensus concept, nor can its definition be agreed upon (Kirchherr et al., 2017). Organizations such as the United Nations, non-governmental organizations, and academic institutions have provided many different definitions. While the definitions highlight certain elements that define what the CE entails, they also highlight a set of core components: i) reducing inputs and utilizing renewable resources efficiently (minimizing waste, utilizing renewable materials); ii) extending the life cycle and reconceptualizing systems (repairing, reconditioning, and remanufacturing; procuring and outsourcing; innovating on business and procurement models); iii) Valorization of output reductions and waste minimization (recycling, recovering, and valuing byproducts and waste).

The CE process encompasses all aspects of the value chain that replaces the linear economy’s end-of-life concepts of reuse, restoration, and renewal with circular flows that are intentionally designed to be restorative. The CE improves efficiency, productivity, and resource allocation, thus enabling competitiveness. By decreasing negative externalities, CE reduces negative externalities, and by creating employment opportunities, it also creates new concepts of consumption (EMF, 2012, 2013).

While CE has a broad scope, it remains difficult to define it concisely. It might be possible to define CE in terms of its approach toward sustainable development in accordance with the reviewed strands of analysis. This approach is accomplished through the implementation of several strategies that reorganize social and production systems to generate closed, regenerative circuits that are environmentally friendly. The main characteristics of this kind of product are the minimization of resources and wastes and the design of manufacturing and consumption processes that are efficient, reusable, repairable, and recyclable from the start.

In an attempt to extend Freeman and Soete (1997)‘s work, de Jesus et al. (2018) propose the notion of clean convergence, i.e. convergence between technological and social subsystems capable of overcoming the problems associated with Fordist, depletion-prone, carbon-intensive times. CE is the most durable, self-reinforcing, and enduring among all positive congruences.

In order to understand the level of granularity of CE implementation, three levels of analysis are required (Ghisellini et al., 2016). An individual actor may be targeted by CE at a micro level, especially a company (Zhu et al., 2010). Environmentally friendly production strategies are also examples, as are efforts to reduce resource consumption, the use of labels, and eco-friendly consumption and production methods. On a meso level, it is about actor interactions, especially between firms: management of green supply chains and industrial symbiosis (Zhu et al., 2010). On a national or global level, in general CE has been viewed as laws, regulatory impact analyses, zero waste regimes, and recycling-oriented societies.

In this context, CE is regarded as: a) Towards ‘clean congruence’ through the guidance of new institutional arrangements balancing environmental concerns with socioeconomic performance while simultaneously promoting technological development that is unreliant on finite resources; b) the development of multi-level frameworks (microscopic, mesoscale, and macroscale) for reshaping and reorienting resilience and sustainability in production and business models; c) a concept that encompasses the growing of new business models, reducing the extraction of resources, maximizing the reuse of resources, increasing efficiency, and enhancing waste recycling.

2.5. The circular economy and eco-innovation: Connecting the two

Cheap resources drove the 20th century for widening markets, the early 21st century has brought heightened volatility in prices and uncertainty in the geoeconomic arena (Barney, 2000). Despite recycling being considered as essential, production of waste is still relatively unchecked (WWF, 2014). By 2050, global consumption is expected to triple due to a dramatic rise in consumption over the last two centuries (Kevin van Langen et al., 2021). Environmental standards are becoming tighter, and consumers are becoming more aware of climate change. To reduce waste of materials and energy in a competitive and dynamic but finite world, a closed-loop economic model that promotes innovation along the entire value chain can be an alternative solution, supported by UNEP (2006, 2011, 2012).

The EU has assumed a global leadership role in developing and implementing sustainable economies and societies since the Lisbon Strategy was adopted in 2000. Many flagship projects and action plans have been recently developed by the EC to promote a transition to sustainability, including EI (EC, 2011a); and efficiency in resource using (EC, 2011c). It is now firmly established that the EU is committed to implementing a circular economy, as the Circular Economy Action Plan confirms, highlighting the close relationship between CE and innovation, but also EI (EC, 2017). According to this argument, the CE depends on implementing eco-innovation systematically, which includes all actors involved in value and supply chains (EC, 2016).

It is a dynamic enterprise that continuously rewires overlaps between various activities in order to transform production routines and consumption habits (EMF, 2013, 2012). Matschoss et al. (2014) identify EI as a key means to accomplishing this, owing to the introduction of new technologies as well as the introduction of new business models and structures. Hence, it is possible to explore the EI-CE links, the primary aim of the study.

While both concepts have complex connections, there are nevertheless many similarities between them. The boundaries of both are somewhat vague and encompass several related terms. It remains to be seen in what ways CE and EI are connected, even if it seems intuitively obvious that they are and that CE cannot be achieved without EI. The CE is not the only area in which EI has effects, and EI can also affect the CE in other areas. In spite of the fact that there is an undeniable relationship between the two, a deeper analysis may utilize all the aspects that have already been discussed (horizontal axis) as well as the micro and macro levels of. In addition to providing supporting evidence, the following literature review distills practical insights. When a closed-loop, production-utilization congruence is achieved, specific types of change must be implemented through a understanding the overlap between CE and the EI on a deeper level. In the quest for transition strategies to CE, policy and decision-makers can monitor and map these self-reinforcing patterns. In the term of a new (green, innovative) techno-economic paradigm, the idea of ‘clean congruence’ addresses mismatches between ecological and economic sustainability from multiple perspectives.

3. Empirical methodology

A model explaining the nexus between environmental innovation performance (EPI) and circularity performance (CIR) is as follows:

(1) $CI{R}_{it}={\beta }_0+{\beta }_{1}EP{I}_{i,t}+{\beta }_{2}E{G}_{i,t}+{\beta }_{3}T{S}_{i,t}+{\beta }_{4}FD{I}_{i,t}+{\beta }_{5}IN{D}_{i,t}+{\beta }_{6}EP{I}_{i,t}+{\beta }_{7}N{R}_{i,t}+{\beta }_{8}D{M}_{i,t}+{\phi }_t+{\omega }_i+{\epsilon }_{ijt,}$(1)

where i and t respectively represent country i and year t. ${\phi }_t$ and ${\omega }_i$ are added into the model to capture the country and year-fixed effects, and ${\epsilon }_{ijt,}$ is the error term.

3.1. Circularity performance (CIR)

Based on Kristensen and Mosgaard (2020), there is no single, widely accepted method for measuring the circular economy. Several attempts have been made over the years. A literature analysis indicates, however, that most of the considered indicators focus on a single aspect, usually inputs and outputs, or only consider a few aspects of circularity (Moraga et al., 2019). According to Kristensen and Mosgaard (2020), recycling, end-of-life management, and regeneration are trend topics for circularity indicators, while few studies have addressed dismantling, extending the useful life, maximizing resources, or recycling for reuse. Scheepens et al. (2016) developed an LCA-based metric to measure circularity in products, but the methodology focuses on reducing externalities rather than measuring circularity. In Genovese et al. (2017), supply chain performance was compared using a lifecycle perspective to integrate circular economies within sustainable principles. Maio and Rem (2015) developed the Circular Economy Index (CEI), which takes into account environmental and economic factors. The CEI focuses on the recycling process, excluding the recovery of materials. A CE performance indicator for industrial products was developed by Cayzer et al. (2017). A performance indicator based on longevity has been developed by Franklin-Johnson et al. (2016), that is, the length of time that a resource is kept in use. DiMaio et al. (2017) demonstrated the difference between the resource efficiency of a process and of a product within supply chains from a lifecycle perspective. In this paper, we follow Nham & Ha (Hong Nham & Ha, 2022) to use a variety of measures in this article to evaluate the performance of circularity. These measures include the amount of municipal waste, the number of circularity patents, the amount of circular material used, the rate of recycling waste, the rate of recycling biowaste, and the rate of recycling e-waste.

To measure the performance of circularity in European countries, we use six distinct measures, including the proportion of the municipal waste generated per capita (kilograms) is measured by the CIR_MW; the number of patents related to recycling and secondary raw materials (CIR_PA); circular material usage (CIR_MA) calculated as the circular material use rate (%); CIR_RW performance based on all waste excluding major mineral waste (%); the recycling rate of biowaste (CIR_RB); CIR_RE measures the recycling rate of e-waste (%). Statistics are taken from Eurostat from 2012 to 2019.

3.2. Key explanatory variable

Following Al-Ajlani et al. (2021), six measures are used to capture the performance of EI in European countries, namely: the percentage of enterprises implementing EI investments (% of surveyed firms); EI activities are represented by the percentage of enterprises that are participating in EI activities (e.g. implementing resource efficiency actions, sustainable products, or ISO 14,001 certifications) based on the share of certified firms among surveyed firms; a number of enterprises with 14,001 registration (EI_ISO) measured as the share of surveyed firms (per min population); and a number of EI related patents (EI_PATENT) (per min population). To shed light on this link, we further indicate the mechanism by studying the impacts of total investments in R&D personnel and researchers (EI_RD) measured as a share of total employment; and the total value of green early-stage investments per capita (EI_GREEN). These variables are sources from the OECD Statistics from 2011 to 2019.

3.3. Control variables

For the determination of the control variables, we used literature-based empirical studies, especially Bu et al. (2019) and Nham & Ha (Hong Nham & Ha, 2022). We have taken economic growth (EG), trade share (TS), industrialization level (IND), and democratization level (DM) as explanatory variables. We also use the percentage of net FDI inflows (FDI) in our theoretical model. Following Le and Nguyen (2019), we consider the effects of natural rents (NR), while a level of democratization (DM) is also added. These variables are taken from World Development Indicators (WDI). Information and statistics about all variables are provided in Table 1. The final sample contains cleaning data from 18 countries during the 2012–2020 period. As shown in Table 2, all variables that suggest a positive correlation between different measures of CE and EI are included in the correlation matrix. A cross-sectional dependence test can be performed on the data. A stationarity test based on cross-sectional dependence (CD) tests proposed by Pesaran (2021) is used to determine whether the data show stationarity. Im-Pesaran-Shin unit root test was developed by Im et al. (2003). The findings are presented in Table 3.

Table 1. Description of variables

Variable	Definition	Measure	Source	Obs	Mean	SD	Min	Max
CE_MW	Municipal waste per capita	Generation of municipal waste per capita (Kilograms per capita)	Eurostat	171	478.49	129.45	239.00	830.00
EI_ENTER	Enterprises with Environmental innovation	The percentage of enterprises implementing environmental innovation investment (% of surveyed firms)	OECD.Stat	162	77.60	36.21	0.00	155.00
EI_ACT	Environmental innovation activities	The percentage of enterprises implementing environmental innovation activities (e.g. implementation of resource efficiency actions, sustainable products, or ISO 14,001 certificates) (% of surveyed firms)	OECD.Stat	162	94.09	32.91	38.00	171.00
EI_ISO	Enterprises with new 14,001 registration	Number of ISO 14,001 certificates (per min population)	OECD.Stat	162	93.31	46.77	0.00	168.00
EI_PATENT	Environmental innovation-related patents	Environmental innovation-related patents (per min population)	OECD.Stat	162	127.40	72.89	0.00	322.00
EI_RD	Investments in R&D personnel and researchers	Total R&D personnel and researchers (% of total employment)	OECD.Stat	162	113.98	60.47	25.00	226.00
EI_GREEN	Environmental early-stage investments	Total value of green early-stage investments (USD/capita)	OECD.Stat	162	105.91	114.01	0.00	422.00
EG	Economic growth	The real GDP per capita (constant 2010 US dollars).	WDI	171	36.93	27.75	1.02	111.15
TS	Trade share	The proportion of GDP.	WDI	171	1.29	0.70	0.55	4.08
FDI	Net inflow of foreign direct investment	The proportion of GDP.	WDI	171	0.04	0.27	−1.54	1.63
IND	Industrialization level	The value added to GDP.	WDI	171	0.23	0.06	0.11	0.37
EPI	Environmental performance index	The score is scaled between 0 and 100, where 0 and 100 mean worst and best performance, respectively.	YCELP	171	71.81	7.41	60.18	82.86
NR	Natural rents	The share of the sum of coal rents, mineral rents, natural gas rents, and forest rents to GDP (%).	WDI	171	0.88	1.73	0.01	10.97
DM	Level of democratization	The index of democratization	FSSDA	171	1.60	0.49	1.00	2.00

Table 2. Correlation coefficients

	CE_MW	EI_ENTER	EI_ACT	EI_ISO	EI_PATENT	EI_RD	EI_GREEN	EG	TS	FDI	IND	EPI	NR	DM
CE_MW	1
EI_ENTER	0.0957	1
EI_ACT	0.0190	0.744***	1
EI_ISO	0.125	0.615***	0.754***	1
EI_PATENT	0.332***	0.456***	0.610***	0.733***	1
EI_RD	0.612***	0.371***	0.457***	0.624***	0.817***	1
EI_GREEN	0.611***	0.232**	0.226**	0.314***	0.634***	0.757***	1
EG	0.628***	0.463***	0.334***	0.450***	0.695***	0.830***	0.706***	1
TS	0.124	0.0145	−0.165*	−0.183*	0.0298	0.161*	0.212**	0.461***	1
FDI	0.183*	0.0843	0.0887	0.107	0.0771	0.136	−0.0605	0.163*	0.0787	1
IND	−0.594***	−0.0247	0.210**	0.0273	−0.146	−0.406***	−0.480***	−0.598***	−0.260***	−0.131	1
EPI	0.582***	0.376***	0.496***	0.301***	0.332***	0.404***	0.378***	0.626***	−0.217**	0.245**	−0.407***	1
NR	−0.194*	−0.431***	−0.395***	−0.262***	−0.255**	−0.317***	−0.0970	−0.384***	−0.0893	−0.0546	0.239**	−0.305***	1
DM	−0.512***	−0.395***	−0.393***	−0.381***	−0.374***	−0.434***	−0.452***	−0.433***	−0.123	−0.195*	0.335***	−0.459***	0.175*	1

Table 3. Cross sectional dependence tests and stationary tests

Variable (in level)	CD-test, Pesaran (2004)	Im-Pesaran-Shin test (Z-bar)	Variable (in difference)	Im-Pesaran-Shin test (Z-bar)
CE_MW	7.712***	4.135	DCE_MW	−2.524***
EI_ENTER	4.561***	−0.453	DEI_ENTER	−4.251***
EI_ACT	0.56	5.055	DEI_ACT	−4.224***
EI_ISO	19.481***	−0.376	DEI_ISO	−5.212***
EI_PATENT	0.184	−0.068	DEI_PATENT	−5.050***
EI_RD	8.126***	0.871	DEI_RD	−3.483***
EI_GREEN	5.672***	−0.237	DEI_GREEN	−4.722***
EG	20.739***	−1.559**	DEG	−5.225***
TS	42.070***	3.007	DTS	−3.698***
FDI	14.973***	0.463	DFDI	−3.241***
IND	0.103	−4.056***	DIND	−4.653***
EPI	7.381***	0.247	DEPI	−3.663***
NR	12.463***	1.136	DNR	−3.219***
DM	0.034	9.771	DDM	−3.370***

Notably, both short-term and long-term effects of the project are taken into consideration. Accordingly, Pesaran and Smith (1995) developed the ARDL method. Considering the potential presence of endogeneity in this model, a pool mean group (PMG) is used as a means of accounting for causal relationships between variables and heteroscedasticity across countries. To begin estimating the relationship between the two variables, we use the Kao (1999), Pedroni (2004), and Westerlund (2005)’s tests. In Table 4, we find that economic complexity, product proximity, and circularity are all co-integrated over the long run.

Table 4. Cointegration test

Model: f(Environmental innovation and circularity)	Kao test	Pedroni test
	Dickey-Fuller test	Phillips-Perron t
CE_MW
EI_ENTER	−2.33***	−23.52***
EI_ACT	−3.28***	−18.88***
EI_ISO	−4.52***	−11.86***
EI_PATENT	−4.38***	−22.95***
EI_RD	−4.23***	−17.95***
EI_GREEN	−3.47***	−18.93***

To shed the light on the distinct effects of environmental innovation and, we use alternative econometric techniques, namely a two-step system General Method of Moment, known as Two-step GMM to deal with a potential issue of endogeneity. The use of lags for persistent dynamic panels may result in weak instruments for the first differenced variables, thereby producing biased results, as argued by Bond et al. (2001). As a solution to this problem, we follow Arellano & Bover (1995) and Blundell & Bond (1998) in developing the two-step GMM. As we have seen in this article, this method is suitable for short dynamic panels (Roodman, 2009). The two-step GMM is also essential to resolve the unobserved heterogeneity and endogeneity that arise in our model (Blundell & Bond, 1998; Roodman, 2009). As part of this study, the two-step system GMM approach is used, along with the Windmeijer (2005) correction and a collapsed instrument set as in Roodman (2009). As a result of the extra instrument sets used in the two-step system GMM, it is possible to correct the bias estimation. In order to determine the validity of the instruments, several tests are conducted, including the Hansen test of over-identifying restrictions (Parente & Silva, 2012), and the Difference-in-Hansen test of the extra instruments (Roodman, 2009).

In brief, since our sample is characterized by a small number of countries (N) and a small-time interval (T), a check of the existence of CD and stationarity of included variables are also proven. Based on Beck and Katz (1995), we use the PCSE model. The FGLS is also employed as a robustness check when this method can deal with the issue of heterogeneity. In addition, to deal with endogeneity resulting from the simultaneity between EI and CE, a one-year lag of all independent variables is utilized. For a robustness check, the two-step GMM is employed.

4. Empirical results

4.1. Environmental innovation and circularity

Table 5 demonstrates the impact of environmental innovation-related activities on circularity security performance (CIR). By using the PCSE estimate and FGLS estimate, we mainly focus on investigating the relationship between variables. When evaluating the impact of EI-related activities (EI_ENTER, EI_ACT, EI_ISO, EI_PATENT, EI_RD, EI_GREEN) on municipal waste per capita (CE_MW). In Panel A and Panel B, FGLS and PCSE estimates reach quite similar conclusions. In particular, the results indicate that all the impacts of EI_ENTER, EI_ACT, EI_ISO, and EI_PATENT on municipal waste per capita (CE_MW) are at a 1% significance level and negative. To be specific, the coefficients of EI_ENTER, EI_ACT, EI_ISO, and EI_PATENT are −0.95, −1.34, −1.77, and −0.53, respectively. The result means that an increase the environmental innovation-related activities, such as the percent of investment in enterprises implementing EI, the percentage of enterprises implementing EI activities, the share of certified firms among surveyed firms, and the number of EI-related patents increase, municipal waste per capita will also experience a decline about environmental innovation will improve the environment quality. Our findings are consistent with. Similarly, Rodríguez-Espíndola et al. (2022) also provide empirical evidence on the link between EI and circularity performance. A systematic literature on this association is also provided by Suchek et al. (2021).

Table 5. Linear impacts of environmental innovation on the circulation

	PCSE estimates
	(1)	(2)	(3)	(4)	(5)	(6)
VARIABLES	Enterprises with EI	EI activities	Enterprises with new ISO 14,001 registration	EI-related patents	Green R&D	Green Investment
L.EI	−0.95***	−1.34***	−1.77***	−0.53***	0.55***	0.48***
	(0.275)	(0.271)	(0.099)	(0.087)	(0.148)	(0.067)
L.EG	2.50***	2.50***	2.05***	3.35***	1.04***	0.17
	(0.434)	(0.353)	(0.374)	(0.209)	(0.292)	(0.418)
L.TS	−20.91***	−18.08***	−54.03***	−30.44***	−12.40**	−13.01**
	(7.792)	(5.792)	(5.269)	(4.117)	(5.603)	(6.010)
L.FDI	15.17	10.44	21.75	1.10	17.08	61.32***
	(15.325)	(12.219)	(16.285)	(11.313)	(13.361)	(23.764)
L.IND	−612.34***	−280.03*	−367.45***	−546.77***	−932.56***	−776.22***
	(109.844)	(147.497)	(49.057)	(48.323)	(49.484)	(53.899)
L.EPI	3.93***	7.12***	2.27***	4.40***	2.45***	1.85*
	(0.744)	(0.866)	(0.702)	(0.672)	(0.910)	(1.067)
L.NR	−11.48	−2.26	−6.63	30.76**	22.13	−1.18
	(17.707)	(14.354)	(11.701)	(14.512)	(14.498)	(14.117)
L.DM	−6.11	5.98	−156.25***	−9.62	12.28	−8.07
	(9.549)	(8.904)	(17.277)	(8.998)	(9.191)	(13.013)
Observations	144	144	144	144	144	144
Number of economies	18	18	18	18	18	18
	FGLS estimates
	(7)	(8)	(9)	(10)	(11)	(12)
L.EI	−0.95***	−1.34***	−1.77***	−0.53***	0.55*	0.48***
	(0.318)	(0.356)	(0.278)	(0.195)	(0.329)	(0.110)
L.EG	2.50***	2.50***	2.05***	3.35***	1.04	0.17
	(0.790)	(0.759)	(0.675)	(0.966)	(0.839)	(0.790)
L.TS	−20.91	−18.08	−54.03***	−30.44	−12.40	−13.01
	(18.251)	(17.727)	(17.689)	(19.440)	(18.317)	(17.336)
L.FDI	15.17	10.44	21.75	1.10	17.08	61.32**
	(27.691)	(27.121)	(25.201)	(27.796)	(28.730)	(29.360)
L.IND	−612.34***	−280.03	−367.45**	−546.77**	−932.56***	−776.22***
	(199.265)	(235.492)	(182.359)	(215.111)	(190.397)	(176.601)
L.EPI	3.93*	7.12***	2.27	4.40*	2.45	1.85
	(2.372)	(2.446)	(2.180)	(2.380)	(2.658)	(2.358)
L.NR	−11.48	−2.26	−6.63	30.76*	22.13	−1.18
	(20.893)	(18.291)	(16.375)	(17.442)	(17.581)	(17.593)
L.DM	−6.11	5.98	−156.25***	−9.62	12.28	−8.07
	(31.635)	(31.237)	(37.352)	(31.835)	(33.856)	(30.616)
Observations	144	144	144	144	144	144
Number of economies	18	18	18	18	18	18

4.2. Robustness checks

4.2.1. Short-run and long-run effect: dynamic fixed effect ARDL model

In a subsequent stage, we conduct robustness checks to validate our previous findings. Firstly, we distinguish the influence of environmental innovation on circularity security performance in the long term and short term. The results are summarized in Table 6. In the short term, we find that only EI_ISO has a statistically significant impact on EC_MW, which implies that increasing the number of enterprises registered 14001 (EI_ISO) will lead to a decrease on municipal waste per capita. All the coefficients of other EI variables are not statistically meaningful, suggesting that the EI_ENTER, EI_ACT, EI_PATENT, EI_RD, and EI_GREEN and municipal waste per capita are not empirically apparent in the short term. In the long term, the impact of enterprises with environmental innovation, environmental innovation activities, enterprises with new 14001 registration, environmental innovation-related patents, investments in R&D personnel and researchers, and environmental early-stage investments patents on municipal waste per capita is statistically significant at 5% and 1% and positive with the coefficients being 31.69, 0.85, 4.52, 0.52, 1.45, 0.12, respectively. The results suggest a favorable effect of environmental innovation, but it is more likely to exist in the long term.

Table 6. The influence of economic complexity on circularity: Short-run and long-run effects

	(1)	(2)	(3)	(4)	(5)	(6)
VARIABLES	Enterprises with EI	EI activities	Enterprises with new ISO 14,001 registration	EI-related patents	Green R&D	Green Investment
The short-term effect
EC terrm	−0.14*	−0.55***	−0.50***	−0.52***	−0.22**	−0.64***
	(0.083)	(0.107)	(0.118)	(0.119)	(0.103)	(0.111)
D.EI	1.82	0.40	−2.16***	−0.26	−0.44	0.10
	(1.330)	(0.280)	(0.555)	(0.265)	(0.439)	(0.160)
The long-term effect
EI	31.69**	0.85***	4.52***	0.52***	1.45***	0.12**
	(12.394)	(0.230)	(0.242)	(0.068)	(0.254)	(0.049)
Observations	144	144	144	144	144	144

Standard errors in parentheses.

*** p<0.01, ** p<0.05, * p<0.1.

PMG-ARDL is employed.

4.2.2. Endogeneity control: two-step GMM

Finally, endogeneity may arise from a simultaneous innovation in the environment and circularity. To address this issue, we follow Hong Nham and Ha (2022) to use a two-step GMM estimate and represent in Table 7. EI_ENTER and EI_ACT have a statistically significant impact on municipal waste per capita in Panel A. In Panel B, we use the squared term of EI, the result indicate that the impact of all measures for EI on circularity has statistical and negative, except for the impact in term of all EI on circulation patents which is significant at 10% and positive. In addition, the adverse impacts of environmental innovation are reported as the coefficient of the squared term of EI on circulation patents is statistically negative. On the other hand, we also find the evidence on the favorable effects of product proximity.

Table 7. Two-step GMM estimates

	(1)	(2)	(3)	(4)	(5)	(6)
VARIABLES	Enterprises with EI	EI activities	Enterprises with new ISO 14,001registration	EI-related patents	Green R&D	Green Investment
EI	6.1208*	1.5678*	−0.0876	0.3121	−1.1656	0.2207
	(3.306)	(0.946)	(0.121)	(0.668)	(0.721)	(0.295)
EG	6.3886	−0.9706	7.9766*	1.7347	3.0484	5.7487
	(8.505)	(0.666)	(4.523)	(9.207)	(2.317)	(7.192)
TS	1.519*	−13.596	−2.459	−1.195	37.1681	−13.478
	(0.827)	(23.018)	(158.179)	(2.186)	(40.344)	(16.903)
FDI	−4.4666	345.5854**	52.6791	53.3517	14.0081	78.2295
	(291.994)	(145.034)	(33.056)	(66.930)	(14.834)	(77.018)
IND	−8.445*	−3.017**	3.852*	8.071	1.549	1.625
	(4.635)	(2.363)	(2.910)	(1.880)	(0.122)	(1.236)
EPI	−17.3024	−15.9063***	−3.2292	−4.2989	−0.6645	12.6224
	(26.060)	(4.057)	(3.266)	(8.384)	(3.829)	(19.992)
NR	−8.5707	−2.9853	−7.3860*	31.9744	2.6319	62.9445
	(7.7281)	(2.273)	(4.1802)	(48.146)	(48.223)	(104.564)
DM	2.1036	−4.4842*	−1.9892*	−5.5932	−1.1561**	1.3243
	(2.968)	(2.3611)	(1.0964)	(5.8813)	(0.7025)	(5.1929)
Observations	144	144	144	144	144	144
Number of economies	18	18	18	18	18	18

5. Conclusions

We are the first to clearly distinguish the importance of environmental innovation on the circularity performance of each country. By using various variables to capture different issues of circularity, our study is expected to analyze the enablers of circularity in depth. This study shows the heterogeneous nonlinear effects of environmental innovation on circularity. However, we provide empirical evidence that innovation activities involving environment play an important role in raising a country’s circularity performance, especially in the long term. Environmental innovation is also an essential enable of circularity.

On the policy front, the findings presented above strongly imply that environmental innovation helps limit emissions to an economic environment and is more convenient for stabilizing the circulating economy. Hence, one of the top priorities for any economy is to improve the quality as well as the number of enterprises investing in environmental innovation activities. In addition, it is advisable for the government to build sophisticated abilities to support economic upgrades. Facilitating businesses to invest and develop a green environment easily is a more effective development strategy.

Compliance with Ethical Standards

• Disclosure of potential conflicts of interest

• Research involving Human Participants and/or Animals

• Informed consent

Disclosure statement

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability statement

Data available on request due to privacy/ethical restrictions. https://ec.europa.eu/newsroom/rtd/items/725730

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This paper was supported by National Economics University.

Notes on contributors

Nguyen Khac Quoc Bao

Doctor Nguyen Khac Thai Bao received his Ph.D from University of Economics Ho Chi Minh City (UEH). He is currently a lecturer at University of Economics Ho Chi Minh City (UEH). His main research areas are financial economics, macroeconomic analysis, international economics, financial stability, and corporate performance analysis.

Le Thanh Ha

Doctor Le Thanh Ha received his Ph.D in Policy Analysis from National Graduate Institute for Policy Studies. He is currently a lecturer at the Faculty of Economics, National Economics University. His main research areas are digitalization, government issues, international economics, macroeconomic analysis, international economics, financial stability, and corporate performance analysis.

Appendix

Table A1: Countries in the sample

EU countries
Austria	Hungary	Portugal
Belgium	Iceland	Slovak Republic
Bulgaria	Ireland	Slovenia
Czech Republic	Italy	Sweden
Denmark	Lithuania
Spain	Luxembourg
Estonia	Latvia
United Kingdom	Malta
Greece	Netherlands
Croatia	Poland