See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/332558353

A Study on the Determinants of Eco-Innovation of Korean Manufacturing

Firms

Preprint · April 2019

DOI: 10.20944/preprints201904.0218.v1

CITATIONS READS

4 57

2 authors, including:

Cheon Yu
Mokpo National University
37 PUBLICATIONS 227 CITATIONS

SEE PROFILE

All content following this page was uploaded by Cheon Yu on 20 April 2020.

The user has requested enhancement of the downloaded file.

Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 19 April 2019 doi:10.20944/preprints201904.0218.v1

1 Article
2 A Study on the Determinants of Eco-Innovation of
3 Korean Manufacturing Firms
4 Cheon Yu and Yun Seop Hwang *
5 1 Trade Research Institute, Kyung Hee University, Seoul 02447, Korea; [email protected]
6 2 Department of International Business and Trade, Kyung Hee University, Seoul 02447, Korea;
7 [email protected]
8 * Correspondence: [email protected];
9
10
11 Abstract: The move to a low carbon economy is very important for enhancing international
12 competitiveness. The eco-innovation is the critical factor of the green paradigm. This study is
13 designed to investigate deeply the determinants of eco-innovation of manufacturing firms in Korea
14 by suggesting anticipated regulations, self-regulations, and industry-specific characteristics as
15 external factors and open information sources as internal factors. The data used in the analysis is
16 1,946 sample firms from Korean Innovation Survey 2010 based on the Oslo Manual. Using the multi-
17 variate probit analysis and the zero-inflated negative binomial (ZINB) regression analysis, we have
18 found out that the anticipated regulations and self-regulations have significant influences both on
19 eco-process innovation and eco-product innovation, while industrial characteristics have no effects.
20 The empirical results also show that the breadth of information sources has a positive effect on
21 businesses in implementing eco-innovations. Our findings show that the Korean government should
22 provide a good platform where firms can better understand the future trends of environmental
23 policies, particularly policies on anticipated and self-regulations. At the same time, Korean firms
24 should establish a voluntary system to control environmental activities so that they can improve eco-
25 innovations through integrating external information.

26 Keywords: eco-innovation; anticipated regulation; self-regulation; industry-specific characteristics;

27 information sourcing openness; multivariate probit model; zero inflated negative binomial model

28 1. Introduction
29 As the Paris Convention concludes, greenhouse gas reductions are becoming a global obligation,
30 and eco-innovation pressures are increasing for firms. Since securing low-carbon capabilities of firms
31 is a very important factor in strengthening international competitiveness, governments in each
32 country are encouraging firms’ eco-innovation through not only regulation but also support [1]. The
33 transition to a low-carbon economy is a big challenge for highly carbon-dependent manufacturing
34 companies. Large-scale investment is needed in new facilities and technology development, and it is
35 difficult to create competitive advantage with existing technologies, production methods and
36 products due to changes in consumer awareness and preference [2]. Many studies have been
37 conducted on the factors that determine the eco-innovation of firms.
38 Previous studies on the determinants of eco-innovation have been limited in geographically and
39 scholarly concerns. From the geographical point of view, much of the research is focused largely on
40 the developed countries such as Germany [3], Britain [4], Spain [5], and overall European countries
41 [6]. Asian countries, such as China [7] and Taiwan [8], have recently been studied, but there is little
42 research conducted on Korea.
43 From the scholarly concerns, firm size, industry characteristics, and market pressure are typical
44 and most discussed determinants of eco-innovation. However, in order to more efficiently estimate
45 the factors that determine the eco-innovation of firms, it is necessary to consider self-regulation
46 activities, industry-specific factors, anticipated-regulations, and information sourcing openness.
1

© 2019 by the author(s). Distributed under a Creative Commons CC BY license.

Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 19 April 2019 doi:10.20944/preprints201904.0218.v1

47 Information sourcing openness is newly suggested for eco-innovation. This is related to open
48 innovation framework [9] and sheds new light on that firm’s effort for open information encourages
49 eco-innovation and consequently it endows the competitiveness of a firm. Although some of the
50 mentioned determinants have been suggested by a few studies [3], they lack an explanation of logic
51 and causation for eco-innovation activities of firms.
52 The purpose of this study is to identify the determinants of eco-innovations of Korean
53 manufacturing firms. Compared to previous research, we concentrate more on the factors that have
54 not been scrupulously treated before. Anticipated-regulations, self-regulations, and industry-specific
55 characteristics are introduced as external determinants, and information sourcing openness is
56 introduced as an internal determinant with logic and causation derived from various theories. These
57 variables, together with regulatory pull/push and market pull, also play an important role in
58 determining eco-innovation activities of a firm. More importantly, if the determinants proved to be
59 powerful driving forces of eco-innovation, then it contributes to fill the research gap between general
60 innovation and eco-innovation. Besides, Korean manufacturing firms’ characteristics can be shown
61 and compare our results with those of previous research done in other countries.
62 The remainder of the paper is organized as follows: Section 2 reviews the relevant literature and
63 the hypotheses are proposed. Section 3 the research model and estimation method is presented.
64 Section 4 illustrates the empirical and measurement results. Section 5 provides the conclusions and
65 implications and the future research direction.

66 2. Literature review and Hypotheses

67 2.1 Concepts of Eco-innovation

68 Eco-innovation is not limited to innovations in products, processes, marketing methods, and
69 organizational methods, but also includes innovations in social and institutional structures [10]. Most
70 of researches, however, focus on products and processes related innovation activities of firms and
71 are so in eco-innovation researches. According to Negny et al. [11], eco-process innovation is defined
72 as a newly-introduced elements on the production process of eco-friendly products. And eco-product
73 innovation is defined as an introduction of new or ground-breaking eco-friendly products. These two
74 types of eco-innovations will be treated as dependent variables in representing eco-innovation
75 outcomes. Eco-process innovation is measured in terms of consumed material per output unit in the
76 process, emitted carbon-dioxide, contaminated material, reduction of harmful and wasted material
77 and water, and recycle of material. Eco-product innovation is measured in terms of energy
78 consumption reduction from using concerned products, reduction of water and air pollutants, and
79 improvement of recyclability after use. In the research eco-process and product innovations are
80 separately discussed and measured on Korean manufacturing firms.

81 2.2 Determinants of eco-innovation and its effect on performance

82 Scholars discussed about the determinants of eco-innovation in various ways. Horbach [12]
83 grouped them into supply, demand, and environmental policy, and Horbach et al. [3] grouped into
84 regulation (Pull/Push), market factors (Pull), technological factors (Push), and firm-specific factors.
85 Cai and Zhou [7] divided the determinants into internal drivers such as technological capability,
86 organizational capability, and CSR, and external drivers such as regulation, customers, and
87 competition, and took integrating capability as an mediating variable and then analyzed the
88 relationship with eco-innovation. Following Cai and Zhou [7], in our study, external and internal
89 factors are analyzed. The external factors include anticipated regulations, self-regulations, industry-
90 specific factors and other control variables. In dealing with internal factors, the focus is placed on
91 external information openness, and other factors (technology, organizational capability, etc.) are used
92 as control variables.

2
Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 19 April 2019 doi:10.20944/preprints201904.0218.v1

93 Contingency theory explains the mechanism of the influence of outer circumstances on firm’s
94 internal organization activities and decision making. It seems adequate for explaining the influence
95 of anticipated regulation, self-regulation, and industry-specific factors on eco-innovation
96 performance. One of the issues we are interested in is whether the regulation, including anticipated
97 and self, has some influence on corporate eco-innovation activity. From the view point of contingency
98 theory, anticipated- and self-regulation is outside circumstances which have direct effect on firm
99 activity. Traditionally scholars insisted that regulations have negative influence on corporate
100 innovation activities and its performance [13]. Contrary to this view, Porter and van der Linde [14]
101 argued that companies would carry out more innovations to prevent environmental pollution under
102 stricter future regulations. Industry specific factors also are external environment that has direct
103 influence on firm activities. In this paper, we focus on energy consumption of an industry which
104 consequently incurs more regulation of the industry.

105 2.2.1 External determinants of eco-innovation

106 Anticipated regulation
107 Preemptive responses for anticipated regulations can save costs that are possibly incurred in the
108 future, by applying the technologies related to the anticipated future regulations in advance to their
109 production processes and products [15]. The firms that succeeded in innovation beforehand can
110 prevent their current or potential competitors from easily entering the market by utilizing their own
111 competitive superiority, for example, by setting the environmental regulation standards in the
112 industry [16]. Although there still exist organizational coordination problems [17], a company will
113 collect information about the anticipated regulations and take strategic actions if it is aware of these
114 advantages.
115 Recent empirical results have proved that anticipated regulations spur eco-innovation [3,18].
116 The path of its impact can be explained by two ways. One is firm’s recognition path. If a firm
117 anticipates that there will be new regulations in the future and recognizes them as threat, it will take
118 various actions in response to the regulations. They can reset its strategic direction and make new
119 frames according to new conditions [19]. The second path is from learning effect of late-movers. The
120 first-movers’ response to the regulation can encourage late-movers’ participation in eco-innovation
121 activities. Under some circumstances late-movers can respond more actively than first-movers [20].
122 Through these two paths the anticipated regulation can accelerate the eco-innovation activities.
123 Furthermore, first-movers’ response to the regulation will encourage late-movers’ participation, so
124 late-movers can respond more actively than first-movers [20].
125 The Korean government has introduced various ways such as public hearing or prior notice
126 system to help companies adapt to new regulations. As a member of the Paris Convention in 2016,
127 Korea has an obligation to largely reduce its CO2 emissions. Under these circumstances, firms make
128 endeavors to come up with sustainable and active measures to respond to anticipated regulations.
129 Based on the discussion above, we propose the following hypothesis.
130
131 H1. The anticipated regulations will positively affect Korean firms’ eco-innovation performance.
132
133 Self-regulations
134 Another external factor that influences eco-innovation is self-regulations within industry.
135 Regulations are generally imposed by the government. On the contrary, the stakeholders, as
136 themselves are the object of regulation, regulate their own activities and advancing information
137 technologies also encourage self-regulation. Advanced information technologies, however, also
138 encourage self-regulation, leading stakeholders to autonomously control their practices. Self-
139 regulation includes establishing financial exchanges, licensing professionals, setting safety standards,
140 controlling entertainment content, advertising restrictions, and voluntarily reducing pollution, etc.
141 [21], which will continuously change along with the development of technology and society.
3
Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 19 April 2019 doi:10.20944/preprints201904.0218.v1

142 Self-regulations are formed under the following conditions — 1) when there is market failure, 2)
143 when it is hard to adjust the market failure or the failure accompanies huge costs, and 3) when self-
144 regulations are more efficient than public (governmental) regulations [22,23]. Among these
145 conditions, the third one is relevant to benefit-costs and differences between public regulations and
146 self-regulations. Ogus [23] explains that self-regulation is advantageous from the following
147 perspectives: First, self-regulation agencies have many experts and experiences of technological
148 innovation trials in certain areas. They can also establish standards with less information costs, so
149 firms under self-regulation have greater innovation potential; Secondly, companies can reduce costs
150 in monitoring and enforcement through creating reliability among stakeholders. Lastly, relatively
151 less formal nature of self-regulation enables lower costs for enforcement and standard revision.
152 Self-regulation has both positive and negative effects. From the legislative view, self-regulation
153 is a type of contemporary ‘corporatism’ which is empowered not ty the formal institutional process,
154 but by the informal interest groups. Such groups have the potential to abuse the power or tendency
155 to administer not in right way. Potential abuse/misuse of the power, in virtual, is activated when the
156 third parties experience with negative effect [22]. Furthermore, self-regulation related groups are
157 tend to be very generous to illegal actions of the group members [24]. Rent-seeking hypothesis
158 explains that the establishment of self-regulation agency, by itself, hinders or making barriers in
159 voluntary establishment of self-regulate organizations. Since established self-regulation agency feels
160 that it owns vested rights in this field and does not want to share the rights with other interest groups.
161 This situation itself creates the inefficiency.
162 This gives important implications to both the government and firms. If it is more efficient for
163 firms to do voluntary environmental regulations, there is no reason for the government to set rules
164 on the industry while incurring social costs. Self-regulation is also beneficial because unintended
165 negative effects caused by government regulations will not be a problem anymore. Voluntary
166 regulation also enables firms to reduce burden from forceful government regulations, boost
167 environmental competitiveness, and appeal to the market more effectively with their
168 environmentally responsible behaviors [25].
169 It can be expected that firms can benefit from self-regulation thanks to regulatory flexibility,
170 preemption of existing regulations, and improved anticipation of future regulations. Firms are also
171 expected to actively engage in eco-innovation and achieve better performance. In Korea, corporate
172 social responsibility (CSR) is regarded very important for firms, and eco-friendly corporate image is
173 a valuable asset for corporate performance. Based on the discussion so far, the following hypothesis
174 can be established:
175
176 H2. Self-regulations will positively affect Korean firms’ eco-innovation performance.
177
178 Industry-specific factors
179 Most eco-innovation related research empirically analyzed the determinants of eco-innovation
180 activities of firms [6]. However, few studies considered industry-specific characteristics, which can
181 also influence eco-innovation activities. According to Horbach [12], high energy-consuming
182 industries, such as automobile and machinery, show a positive relationship with eco-innovation, but
183 industries which consume less amounts of energy do not show any significant relationship. Machiba
184 [26] chooses three sectors such as the automotive and transport industry, the iron and steel industry,
185 and the electronics industry as examples and analyze the application of eco-innovation. And the
186 result indicates that eco-innovations in these industries are more active because energy-intensive
187 industries are more likely to pursue eco-innovation to increase energy efficiency.
188 Regarding industry-specific factors, few studies regarded these factors as the determinants of
189 eco-innovation. Most Korean studies about the relationship between industry-specific factors and
190 eco-innovation assert that there is a positive relationship between them. When treating climate
191 change issues, energy consumption of an industry is most widely used as a proxy for industry-

4
Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 19 April 2019 doi:10.20944/preprints201904.0218.v1

192 specific factors. In this paper, we expect that energy consumption of an industry has a positive
193 relationship with eco-innovation activities of Korean manufacturing firms.
194
195 H3. Energy consumption as an industry-specific factor will positively affect Korean firms’ eco-
196 innovation performance.

197 2.2.2 Internal determinants of eco-innovation

198 Information sourcing openness for eco-innovation
199 Chesbrough [9] argued that innovation outcomes can improve through greater openness to
200 obtaining and using external information. When a firm becomes more open to the use of external
201 information, it can achieve eco-innovation and higher performance. Firms should not only become
202 more open to external information, but also consider the importance of such information. Companies
203 with more sources of gathering helpful and important information can have more opportunities to
204 achieve higher performance. De Marchi and Grandineti [27] argued that the information obtained
205 from external partners such as research institutions, colleges, or competitor firms is more important
206 to eco-innovation than to other types of innovations.
207 Laursen and Salter [28] categorized external information collected for innovation activities in
208 terms of breadth and depth. For the breadth of sourcing, Ghisetti et al. [29] suggested two reasons
209 why firms need a broader source of external information. Firstly, it is difficult to respond to various
210 changes only with their own internal resources and capabilities, especially under the global climate
211 change situation. Secondly, eco-innovation requires a huge volume of information in achieving
212 multiple objects, not only improving the productivity and quality of common innovation but also
213 reaching environmental targets. Rennings and Rammer [30] proved that the German firms use
214 relatively more diverse sources of information to obtain innovative outcomes, and Horbach et al. [31]
215 argued that eco-innovation needs more external information sources.
216 Meanwhile, firms need more professional and in-depth information guides in order to carry out
217 innovative activities. In this study, we use the importance of information, instead of using the concept
218 of depth. It seems that importance is similar to depth in that it is understood as usefulness and
219 sincerity of outsourced information. However, the importance evaluates the importance level of
220 external information to the firm, while the depth of information evaluates the degree of specialty and
221 cooperation of partner companies. Resource-based view (hereinafter referred to as RBV) insists that
222 the way a firm keeps its competitive advantage depends on the uniqueness, rarity, and worthiness of
223 its resources, and on whether it is difficult to imitate its resources or not [32]. Capabilities based on
224 accumulated information can be one of the key resources that cannot be easily imitated by
225 competitors. The complexity and rapid changes of the recent external circumstances urge firms to
226 gather important information from outside as well as inside the organization [9].
227 It is expected that a company will be more likely to achieve innovation, including eco-
228 innovation, when it acquires greater amount of important information from various external sources.
229 Korean firms are making efforts to strengthen their competitiveness by introducing information from
230 various external sources. In a study on the effects of external information search on the Korean ICT
231 sector, Hwang and Lee [33] found that external information search is relevant to incremental
232 innovation and productivity. From the point of supply chain management, Woo et al. [34]
233 investigated communication capability and external green integration for the financial and green
234 performance of construction providers in Korea. Their results show that the greater capability a firm
235 shows in sharing information with other organizations, the more likely it is for the firm to acquire
236 superior positions in taking environmentally cooperative actions, in making green cost reduction,
237 and in securing higher competitiveness. It means that the efforts to get valuable information from
238 external organizations can lead to better performance of Korean manufacturing firms. According to
239 these discussions, the following hypotheses are proposed.
240

5
Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 19 April 2019 doi:10.20944/preprints201904.0218.v1

241 H4. Information sourcing openness have a positive influence on eco-innovation performance of
242 Korean manufacturing firms.
243 H4a. When information sourcing becomes broader, the performances of eco-innovation by
244 Korean manufacturing firms increase.
245 H4b. When information sourcing is more important to receiving firms, the performances of eco-
246 innovation by Korean firms increase.

247 2.2.3. Control variables

248 This study sets the control variables as follows. The control variables as external factors are
249 market pull, regulatory pull and push. Following the demand pull hypothesis, the market demands
250 or needs promote technological innovations [35,36]. Firms prefer to take competitive advantages
251 through eco-innovation as they are aware of the market demands for eco-friendly goods triggered by
252 the need for energy saving and environmental preservation [18,37-39]. Regulatory push and pull are
253 also important influential factors for eco-innovation. There is a conventional view that governmental
254 mandatory regulations (regulatory push) hamper economic growth by aggravating cost loads [13,40-
255 42]. On this account, firms try to avoid cost increase and reduce innovative actions which require
256 capital expenditure. Conversely, there are also cases where the government’s environmental
257 regulation is an important factor in driving firms to pursue eco-innovation [3,43]. Stricter regulations
258 drive firms to lead innovations to reduce environmental pollutants, which enable profits to exceed
259 costs [14]. On the other hand, Governmental supporting policies (regulation pull factor) have a
260 positive influence on firms’ eco-innovation [12,44,45].
261 The control variables as internal factors are innovative capability, technology push, firm size and
262 age. Eco-innovation requires more professional knowledge compared to general innovation [46,47].
263 It means that general innovation capability is necessary in successfully accomplishing eco-
264 innovation. General innovation capability includes technological innovation capability for products
265 and processes and non-technological innovative capability for markets and organizations. Generally,
266 higher general innovation capability leads to higher outcomes of eco-innovation. Technology push
267 like R&D investment is also important determinants for eco-innovation. Firms’ investments into R&D
268 produce new technological information and promote technological innovation by strengthening
269 internal capability of assimilation and exploitation [48]. R&D investment refers to technological and
270 absorbing capabilities, a determinant of internal technological level of a firm [49]. It is also true for
271 eco-innovation [12].

272 3. Research Model and Data

273 3.1 Research Model

274 The research model of this study is presented in <Figure 1>. The dependent variables are divided
275 into eco-process innovation and eco-product innovation. Eco-process innovation consists of six items
276 and eco-product innovation consists of three items. Whether eco-innovation is executed or not is
277 analyzed by using the binary dependent variables, and the sum of all areas, from 0 to 9, is used as the
278 aggregated variable for eco-innovation performance.
279

6
Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 19 April 2019 doi:10.20944/preprints201904.0218.v1

280
281 Figure 1. Research Model

282
283
284

285 3.2 Definition of the data and the variables

286 The data used in this study is “Korean Innovation Survey 2010: Manufacturing sector.” The
287 statistics come from Science and Technology Policy Institute (STEPI)’s investigation into firms’
288 innovation activities from 2007-2009 based on the Oslo Manual developed by OECD and Eurostat
289 Community Innovation Survey (CIS). The survey is approved by Statistics Korea (KOSTAT) and
290 recognized to have high reliability, validity, and international comparability. The size of the
291 population is 41,485, and the samples of 3,925 firms were selected by using the stratified sampling
292 method, and 1,964 firms remained after the deletion of missing data. The definitions of the variables
293 are summarized in <Table 1>.
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308

7
Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 19 April 2019 doi:10.20944/preprints201904.0218.v1

309 Table 1. Definition of Variables

Variables Label Definition Type
Anti-
Anti-Regu The future predicted environmental regulations or taxes Binary
Regulation
Indep Self-regulation Self-Regu Voluntary conventions or agreements within the industry Binary
enden
Energy Within an industry, the averaged energy consumption of
t Eng-Cons Log
Consumption the belonged firms
varia
bles Information Breadth The number of used information sources 0~12
Sourcing Averaged degree of information use(=sum of importance of
Openness Importance 0~5
used information/the number of used sources)
Pre-Regu Existing environmental regulations or taxes Binary
Regulatory
Using subsidiaries from the government or financial
Pull/Push Subsidy Binary
benefits related to eco-innovation
Market Market demands for eco-innovation from the current or
Contr Mkt-Pull Binary
Pull future consumers
ol
Innovation The number of innovations on products, processes,
varia Inno-Capa 0~4
Capability organizations, and marketing
bles
Technology In-R&D % of internal R&D expense to sales volumes 2007~2009 Log
Push Ex-R&D % of external R&D expense to sales volumes 2007~2009 Log
Size Log (The number of full-time employees in 2007) Log
Age The age of the firm (=2014-the founded year) count
D1 Reduction of material consumption per output unit Binary
D2 Reduction of energy consumption per output unit Binary
Eco-Process D3 Reduction of CO2 emissions Binary
Innovation D4 Replacing polluting or harm matters Binary
Depe
D5 Reduction of soil, water, noise, and air pollutants Binary
ndent
D6 Recycling wastes, water, and materials Binary
D7 Reduction of energy consumption Binary
Eco-Product
D8 Reduction of soil, water, noise, and air pollutants Binary
Innovation
D9 Improving recyclability after product use Binary
Total Number D10 (1)+(2)+(3)+(4)+(5)+(6)+(7)+(8)+(9) 0~9

310

311 3.3 Analysis Model

312 As mentioned before, there are two special kinds of dependent variables in our research model.
313 One is binary (0, 1) variable, and another is count variable. The former is on whether a firm engages
314 in eco-innovative actions, including six eco-process innovative actions and three eco-product
315 innovative actions. The latter is the overall number (count number) of the innovative actions.
316 Considering the special data types, we estimate multivariate probit models addressing the nine
317 innovative actions at the same time. In our research, the dependent variables will be 0 if the firm does
318 not engage in eco-innovative action, and will be 1 if the firm engages in eco-innovative action. And
319 the multivariate probit model is a generalization of the probit model used to estimate several
320 correlated binary outcomes jointly. In this research, the multivariate probit model allows a
321 simultaneous estimation of the nine types of eco-innovation. The existing evidence about the
322 likelihood of complementarities between different types of innovative activity makes us consider that
323 unobserved firms’ characteristics may jointly influence the nine types of eco-innovation [50]. If these
324 correlations were neglected, parameter estimates would be biased and inconsistent. The multivariate
325 probit models can be written as:
326
8
Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 19 April 2019 doi:10.20944/preprints201904.0218.v1

327 𝑦∗ = 𝛽 + 𝛽 𝑥 +𝛽 𝑥 + ⋯+𝛽 𝑥 + 𝜀
328 𝑦 = 1 𝑖𝑓 𝑦 ∗ > 0 𝑎𝑛𝑑 0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒.
329
330 Where i represents individual firms; 𝑦 are nine binary eco-innovation dependent variables
331 (m= 1, 2, …, 9); 𝑥 are indepent variables and control variables (p= 1, 2, …, 12); and 𝜀 are error
332 terms distributed as multivariate normal, each with a mean of zero, and variance-covariance matrix
333 V, where V has values of 1 on the leading diagonal and correlations 𝜌 = 𝜌 as off-diagonal
334 elements. Also, we estimate the zero-inflated negative binomial regression model for the overall
335 number of eco-innovation activities. There are two basic methods for modeling count variables with
336 excessive zeros, zero-inflated poison (ZIP) regression and ZINB regression. There is an issue that
337 should be considered when using Poisson models which are a Poisson distribution and constrain the
338 variance to be equal to the sample mean. This is a problem in our case given that the sample is much
339 skewed. In contrast, ZIP does not have this constraint. By testing, we ultimately estimate zero-inflated
340 negative binomial models in recognition of the high number of “zero” responses in the dependent
341 variables. According to Table 2, the ZINB is preferred over ZIP. Thus, ZINB models are used to
342 estimate the determinants of the total number of eco-innovations.

343 Table 2. Tests and Fit Statistics

ZIP BIC= -7962.343 AIC= 3.425 Prefer Over Evidence

vs. ZINB BIC= -8053.391 dif= 91.048 ZINB ZIP Very strong
AIC= 3.375 dif= 0.050 ZINB ZIP
LRX2= 98.622 prob= 0.000 ZINB ZIP p=0.000
344
345 The ZINB regression model assumes that there are two distinct data generation processes. The
346 results of a Bernoulli trial are used to determine which of the two processes. For observation 𝑖, with
347 probability 𝜋 , the only possible response of the first process is zero counts, and with the probability
348 of (1 - 𝜋 ), the response of the second process is governed by a negative binomial with mean 𝜇 . The
349 zero counts are generated from both the first and second processes, where a probability is estimated
350 for whether zero counts are from the first or the second process. The overall probability of zero counts
351 is the combined probability of zeros from the two processes. A ZINB model for response 𝑌 can be
352 written as:
⎧ 𝑘
⎪ 𝑃(𝑌 = 0) = 𝜋 + (1 − 𝜋 ) ·
𝜇 +𝑘
353
⎨ Γ(𝑌 + 𝑘) 𝑘 𝑘
⎪𝑃(𝑌 = 𝑛) = (1 − 𝜋 ) · · · 1−
⎩ Γ(𝑘)Γ(𝑌 + 1) 𝜇 + 𝑘 𝜇 +𝑘
354
355 Where k is the over dispersion parameter; Γ is the gamma distribution, and n is a natural
356 number larger than 0. We can model 𝜋 and 𝜇 as a function of a set of explanatory variables. For
357 𝜋 , it is common to use a logistic regression with a logit link function, as it describes a binomial
358 process:
359
360 𝑙𝑜𝑔𝑖𝑡(𝜋 ) = 𝑒 ⋯

361
362 Where α is the intercept, 𝛽 … 𝛽 are the model parameters to estimate, and 𝑋 … 𝑋 are a
363 set of independent variables. We can also model the dependence of 𝜇 on a different (or same) set of
364 explanatory variables with the aid of a log link function:
365
366 𝑙𝑜𝑔(𝜇 ) = 𝜆 + 𝛿 𝑍 + 𝛿 𝑍 + ⋯ + 𝛿 𝑍
367

9
Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 19 April 2019 doi:10.20944/preprints201904.0218.v1

368 Where 𝜆 is the intercept, 𝛿 … 𝜹𝒏 are the model parameters to estimate, and 𝑧 … 𝑧 are a
369 set of independent variables.

370 4. Results of the Empirical Analyses

371 <Table 3> shows the results of multivariate probit models, which analyzed the effects of
372 determinate factors on the nine types of eco-innovation of the 1,964 manufacturing firms in Korea.
373 The corresponding Wald tests analyzing the explanatory power of the entire model indicate that the
374 null hypothesis that all parameters of the explanatory variables are zero can be clearly rejected at all
375 common levels of significance for all probit models. The last column in the <table 3> reports the
376 determinants on the extent to which firms engage in eco-innovation which can be seen as outcome of
377 eco-innovation.

378 4.1. The Results of Empirical Analyses of external factors

379 Among the external factors, the future anticipated regulations have positive influences on eco-
380 process and eco-product innovation at the 1% significant level. Also for the eco-innovation outcome
381 performance by how many kinds of innovation activities are enacted, there is a significant positive
382 influence. Therefore, the hypothesis 1 is supported. This result shows that firms make efforts to win
383 competitive superiority by differentiating their products and cutting down expenses by
384 implementing environment-related innovation activities preemptively in response to the future
385 predicted regulations [15, 16]. It means that the argument by Porter and Van der Linder [14] can also
386 apply to Korea. Also in Khanna et al. [46], current and anticipated regulatory pressures, as proxied
387 by penalties, inspections, hazardous air pollutants and non-attainment have a statistically positive
388 impact on an adoption of a new pollution prevention measure. But, due to data limitation, it is
389 difficult to figure out whether companies regard future predicted regulations as an opportunity or a
390 threat.
391 Not only the governmental regulations, but also firms’ voluntarily engaged conventions (self-
392 regulation) have influence on their eco-innovation activities. According to the results of analyses,
393 firms which are voluntarily engaged in conventions show higher possibilities of innovation in all
394 activities which belong to eco-process and eco-product innovations. Such firms also have higher
395 numbers of eco-innovation activities than others. From this result, hypothesis 2 is adopted. It means
396 that, as in the argument of Nash and Ehrenfeld [51], eco-innovation activities are also realized under
397 non-public control in Korea, and that voluntarily engaging firms are more likely to use eco-
398 innovation than non-voluntary firms. Despite the vulnerability of voluntary conventions being
399 loosely operated compared to the government regulations without explicit punishment on free-ride
400 or opportunistic behaviors, voluntary conventions promote eco-innovations of Korean firms.
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
10
Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 19 April 2019 doi:10.20944/preprints201904.0218.v1

416 Table 3. Determinants of Eco-innovation by different environmental types & Total Number of Eco-innovation

Number of obs = 1946 Eco-Process Innovation

Wald chi2 (117) =
2125.87 D5
D1 D2 D3 D4 D6
Prob > chi2 = 0.000 Soil/Water/
Material Energy CO2 Danger Recycling
Log likelihood = - Noise
6014.5327
Anti-Regu 0.7583*** 0.9308*** 0.8141*** 0.9138*** 0.7982*** 0.7374***
(9.44) (11.73) (10.17) (11.43) (10.20) (9.22)
Independent variables

External

0.7230*** 0.8251*** 0.8399*** 0.6958*** 0.7872*** 0.9280***

Self-Regu
(8.39) (9.83) (9.81) (8.02) (9.19) (10.85)
0.0188 0.0363* 0.0549** 0.0029 0.0543*** 0.0750***
Eng-Cons
(0.88) (1.73) (2.57) (0.14) (2.63) (3.66)
0.0287** 0.0232* 0.0181 0.0129 0.0198 -0.0068
Breadth
Internal

(2.30) (1.92) (1.45) (1.06) (1.64) (-0.58)

0.0880* 0.0653 0.0272 0.0380 -0.0312 -0.0020
Importance
(1.68) (1.28) (0.52) (0.75) (-0.62) (-0.04)
0.6855*** 0.6268*** 0.6615*** 1.0226*** 1.0763*** 1.0579***
Pre-Regu
(7.69) (7.18) (7.48) (11.51) (12.51) (12.04)
External

0.7103*** 0.6787*** 0.6907*** 0.7343*** 0.7590*** 0.8178***

Subsidy
(5.30) (5.20) (5.08) (5.18) (5.59) (5.78)
1.1900*** 1.1461*** 1.0380*** 1.0793*** 1.0554*** 1.1152***
Mkt-Pull
(15.82) (15.63) (13.72) (14.57) (14.23) (15.44)
Control variables

0.1523*** 0.0908** 0.0973** 0.2085*** 0.0968** 0.1287***

Inno-Capa
(3.91) (2.43) (2.51) (5.48) (2.58) (3.50)
-0.1199 -0.0023 -0.0554 -0.1452 -0.1497 -0.3538***
In-R&D
(-1.04) (-0.02) (-0.49) (-1.23) (-1.32) (-2.95)
Internal

0.1294 0.0309 -0.0637 0.3998 0.0138 0.1461

Ex-R&D
(0.43) (0.11) (-0.20) (1.28) (0.05) (0.50)
0.0405 0.0198 0.1088*** 0.0225 -0.0016 0.0087
Size
(1.21) (0.61) (3.30) (0.68) (-0.05) (0.27)
0.0005 0.0068** -0.0019 0.0027 0.0085*** 0.0021
Age
(0.16) (2.37) (-0.63) (0.91) (2.97) (0.75)
-2.7555*** -2.4701*** -2.9788*** -1.9467*** -2.5577*** -2.2967***
_cons
(-3.94) (-3.66) (-4.04) (-2.73) (-3.72) (0.6741)
417 Note. Standard deviation in parenthesis. * p < 0.10; **p < 0.05; ***p < 0.01.

11
Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 19 April 2019 doi:10.20944/preprints201904.0218.v1

418 Table 3. Determinants of Eco-innovation by different environmental types & Total Number of Eco-innovation
419 (continued)

Total Number
Negative Binomial
Eco-Product Innovation
Regression Part
Number of obs = 1946 ( >0 eco-innovation)
Wald chi2 (117) = 2125.87 Number of obs = 1946
Prob > chi2 = 0.000 Nonzero obs = 1081
Log likelihood = D8 Zero obs = 865
D7 D9
-6014.5327 Soil/Water LR chi2 (13) = 946.36
Energy Recycling
/Noise Prob > chi2 = 0.000
Inflation model = logit
Log likelihood = -3361.498
Anti-Regu 0.5214*** 0.645*** 0.5274*** 0.8968***
(6.52) (8.40) (6.75) (14.14)
Independent variables

External

0.8705*** 0.7181*** 0.7931*** 0.9503***

Self-Regu
(10.34) (8.53) (9.70) (13.60)
-0.0029 0.0270 -0.0119 0.0320**
Eng-Cons
(-0.14) (1.30) (-0.58) (2.04)
0.0213* 0.0245** 0.0313*** 0.0212**
Breadth
Internal

(1.80) (1.97) (2.62) (2.35)

-0.0522 0.0543 0.0168 0.0291
Importance
(-1.06) (1.06) (0.34) (0.78)
0.6239*** 0.7478*** 0.7016*** 1.0457***
Pre-Regu
(7.22) (8.54) (8.17) (14.61)
External

0.7125*** 0.6853*** 0.4359*** 0.8006***

Subsidy
(5.13) (5.06) (3.23) (7.47)
1.2471*** 1.1636*** 1.0428*** 1.3937***
Mkt-Pull
(17.39) (15.43) (14.63) (23.52)
Control variables

0.1468*** 0.1404*** 0.1164*** 0.1611***

Inno-Capa
(4.00) (3.63) (3.18) (5.67)
-0.1359 -0.1329 -0.3950*** -0.2145**
In-R&D
(-1.21) (-1.10) (-3.14) (-2.43)
Internal

0.1907 0.4419 -0.3264 0.1140

Ex-R&D
(0.65) (1.51) (-1.00) (0.49)
0.0629* 0.0450 -0.0777** -0.0014
Size
(1.94) (0.46) (-2.45) (-0.06)
-0.0033 0.0004 0.0016 0.0019
Age
(-1.14) (0.15) (0.54) (0.85)
-1.955*** -1.9133*** -3.4034*** -1.4686
_cons
(-2.91) (-2.84) (-4.58) (-2.77)
Inflated (Probit Selection
-18.8153
Part)
(-0.04)
_cons
-0.4044***
/lnalpha
(-6.10)
alpha 0.6674
420 Note. Standard deviation in parenthesis. * p < 0.10; **p < 0.05; ***p < 0.01.

421
422
Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 19 April 2019 doi:10.20944/preprints201904.0218.v1

423 Regarding industry-specific factors, there are interesting results. Industry-specific factors
424 positively influence eco-process innovations, but influence eco-product innovations insignificantly.
425 Among eco-process innovations, when average energy consumption in an industry is higher,
426 possibility of innovation is higher in areas including (2) reducing energy consumption per output
427 unit, (3) reducing carbon dioxide emission, (5) reducing soil, water, noise, and air pollution, and (6)
428 recycling of wastes, water, and materials. This means that firms in a high energy consumption
429 industry are focusing more on the benefit from cutting down their energy consumption rather than
430 consumers’ energy consumption in product use. Also in manufacturing processes, they concentrate
431 more on the innovative actions which bring reduction outcomes in a relatively short period of time.
432 But in the processes which need investment and research for a relatively longer period of time, such
433 as replacement of pollutants or harm matters, innovative actions are less likely to happen at least in
434 short term. Industry-specific factors have influences on eco-innovation possibilities of companies, but
435 not on eco-innovation outcomes.

436 4.2. The Results of Empirical Analyses of Internal Factors

437 As one of our focusing variables, let us see how openness to information affects eco-innovation.
438 First, when the scope of obtaining information is broader and more diverse, the diversity of eco-
439 innovation improves. This result is consistent with Laursen and Salter [28] which studied general
440 technology innovation. Hypothesis 4-1, therefore, is supported. In Korea, as many researchers
441 suggest, eco-innovation requires capabilities to deal with different dimensions, which can be
442 acquired from diverse external information sources [27,29,31]. So, the firms with broader information
443 sources can be said to get higher outcomes of eco-innovation than those with narrower sources.
444 Also in eco-process and eco-product innovation activities, it has a significantly positive influence
445 on the activities of (1) material, (2) energy, and all activities in eco-product. From this result, firms
446 obtaining information from broader and more diverse sources are more likely to take actions on eco-
447 innovation. This result has great implications on both policy and academic study. In terms of
448 academic study, it suggests the necessity of expense-related studies, as it is possible that more diverse
449 sources of external information accompany higher costs in their handling and management. In terms
450 of policy, the method to raise diversity of external information should be proposed.
451 The importance of openness to information has not showed any significant results in the
452 outcomes and individual items of innovation activity, which is far from the expectations of this study.
453 Therefore, hypothesis 4-2 is not supported. It is probably because of firms’ lack of absorbing
454 capability, which refers to firms’ ability to effectively use outside information [52,53]. The value of
455 external information depends on firms’ potential absorptive capabilities [53-55]. But this result should
456 be strengthened by empirical analyses.

457 4.3. The Results of Empirical Analyses of Control Factors

458 Regarding the results about regulatory pull/push, Korean firms show the same results as the
459 previous studies [3,43]. The present regulations positively affect eco-innovation at 1% significance
460 level. Also, governmental subsidiaries show the same results as the present regulations. Both
461 regulatory push and pull have a significantly positive influence on each dimension of eco-process
462 and eco-product innovation. The existences of the present regulations and governmental support
463 have significantly positive relationships with all activities of eco-process and eco-product innovation.
464 Also in the market-pull factors, higher demands on eco-friendly products have greater positive
465 influences not only on eco-process and eco-process innovation, but also on eco-innovation outcome
466 performance. This is consistent with the empirical research results of Kammerer [18]. Next, let us see
467 how influential internal factors are on eco-innovation of Korean firms. First, innovation capability
468 has a significant effect at 1% level. It means that for Korean firms, higher innovation capability
469 promotes eco-innovation. When seeing technology push factors, internal R&D investments have
470 significantly negative relationships with eco-innovation. The reason is that business R&D investment

13
Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 19 April 2019 doi:10.20944/preprints201904.0218.v1

471 has a negative effect on productivity growth in short run, as Guellec and Potterie [56] argued. They
472 presented that it takes some time for those outcomes to have relationships each other, and before they
473 do, they can show negative relationships. In Seop et al. [57], R&D intensity has a negative relationship
474 with firms’ revenues in short run. Meanwhile, external investment is not in significant relationships.
475 The size of firms, which is represented by the number of employees, has a significant effect only
476 partly on eco-innovation. Only eco-innovation activities for carbon dioxide reduction have been
477 proportionally influenced by firm size. In eco-product innovation, the activities regarding energy use
478 reduction have been positively influenced by the size of firms at 10% significance level, though firms’
479 size negatively affects recycling of products. Firm's age, which represents firm's experience, has a
480 positive relationship with the overall outcomes of eco-innovation.

481 5. Conclusions
482 This study has empirically proved the determinant factors of the eco-innovation of Korean
483 manufacturing firms, divided into external and internal factors, mainly by using the variables that
484 have not been much proved. Both anticipated regulations and self-regulations, which are external
485 factors, promote eco-innovation. On the dimension of industry-specific characteristics, firms which
486 belong to a more energy-consuming industry are more concentrated on eco-process innovation. In
487 information sourcing openness, which is an internal factor, breadth is a factor with a more significant
488 influence on eco-innovation than importance is. It means that the capability of a firm to make
489 connections with diverse bodies (businesses, colleges, research institutions, etc.) and apply their
490 information itself has an influence on its choice of eco-innovation. The three variables, anticipated
491 regulations, self-regulations, and the breadth of information sourcing, also have significantly strong
492 influences on firms’ eco-innovation outcomes, which are represented by how frequently they engage
493 in eco-innovation activities.
494 According to the empirical results, the following academic and policy implications can be
495 provided. First, it has been already confirmed by a lot of researchers that the government’s
496 environmental regulations promote eco-innovations [14,58,59]. This study analyzes how firms react
497 to anticipated environmental regulations imposed by the government, which has not been much
498 addressed in previous studies. As shown in the results, firms take positive response strategies rather
499 than passive defending actions, as anticipated regulations imply that their future competition
500 circumstances will be different. Particularly, the argument that late comers take anticipated
501 regulations as an opportunity to secure competitive advantages and become leading firms, and take
502 more vigorous actions for technological innovations strengthens the validity of the results of this
503 study [60].
504 When firms clearly see the direction of environmental policies, they decide on their strategies
505 and respond to anticipated circumstantial changes. The consistency and continuation of the
506 government’s environment policies, therefore, are important. But even when firms are aware of
507 external circumstantial changes, they cannot be immune from the problem of uncertainty. Firms are
508 more likely to take passive actions when uncertainty is high, while they are more likely to take active
509 actions when uncertainty is low and there is improved predictability. When it comes to activities
510 related to eco-innovation, the predictability of future regulations can be explained by the same theory.
511 Due to the limitations of our data, however, future studies should be conducted based on related
512 studies.
513 Second, apart from governmental regulations, self-regulations have a positive influence on
514 firms’ engagement in eco-innovation activities and their performance. There is a lack of concrete
515 discussion about the mechanism through which self-regulations have such influences. In this study,
516 negative and positive influences of self-regulations are presented, and the empirical analyses are
517 carried out to prove them. As shown from the results, the positive relationship of self-regulation with
518 eco-innovation suggests that self-regulations have positive effects in Korea, and firms should be
519 induced to take voluntary actions on environment-related innovation activities by themselves
520 through related policies. Especially, the realization of self-regulation leads to the preparation of
14
Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 19 April 2019 doi:10.20944/preprints201904.0218.v1

521 effective methods to replace non-elastic governmental regulations [23]. From the government’s
522 perspective, self-regulations can reduce external effects from the establishment of rules and cut down
523 the costs on persuading various interest groups. For these purposes, it is important to prepare a
524 foundation on which sound self-regulations can be established, and the government and firms should
525 make joint efforts to do so. Third, concrete identify of the relationship between industry-specific
526 factors and eco-innovation. Although the results of some studies have presented that firms which
527 belong to a high energy-consuming industry are active for eco-innovation, they might be missing a
528 concrete influential relationship. As the result of this study shows, industry-specific factors do not
529 have a significant influence on eco-product innovation, but have a significant influence on eco-
530 process innovation. Particularly, it is significant in some kinds of eco-innovation, such as energy
531 consumption reduction, carbon dioxide reduction, soil/water/noise/air pollutants reduction, and
532 wastes/water/material recycling. It suggests that firms that consume a lot of energy practice more
533 process-centered eco-innovation policies. Fourth, the major contribution of this study is that it has
534 figured out that various information sources could help firms win a competitive advantage through
535 eco-innovation. Since eco-innovation intrinsically has a complex development process and multiple
536 purposes, using diverse external information is important [27,29]. Also, multiple dimensions such as
537 design, user involvement, product and service, or governance should be taken into consideration in
538 the process of eco-innovation [27]. It is also well explained in this study as it presents the positive
539 relationship between the breadth of information sources and eco-innovation. In particular, the
540 positive relationship with overall eco-innovation suggests that firms which want to win competitive
541 superiority through product differentiation need to actively use external resources. Meanwhile, from
542 firms’ point of view, the importance of external information does not influence eco-innovation
543 engagement and outcomes. It seems to have something to do with absorbing capability, though there
544 should be related further studies in the future.
545 Despite the implications presented above, this study has several limitations. First, it could not
546 fully take into consideration the temporal effect of each factor considered in the analyses on eco-
547 innovation. In future studies, temporal research should be done on the same firms analyzed in this
548 study. Second, since it could not figure out whether firms take predicted regulations as opportunities
549 or threats due to the limit of data, the changes of firms’ behaviors were not explained either. If more
550 research is done on this issue, it could provide important implications on how the strategy
551 formations, actions, and outcomes of companies will change depending on their awareness of
552 anticipated regulations.
553
554
555 Author Contributions: Conceptualization: C.Y. and Y.S.H.; Formal analysis: C.Y.; Project administration: Y.S.H.;
556 Supervision: Y.S.H.; Writing original draft: C.Y.; Writing—review and editing: Y.S.H.

557 Funding: This study received no external funding.

558 Acknowledgments: This work was supported by the Ministry of Education of Korea and the National Research
559 Foundation of Korea (NRF-2016S1A5B6925462)

560 Conflicts of Interest: The authors declare no conflicts of interest.

561
562 Reference
563
564 1. Hart, S. L.; Dowell, G. Invited editorial: a natural-resource-based view of the firm: fifteen years after. Journal
565 of management 2011, 37(5), 1464-1479.
566 2. Porter, M. E.; Reinhardt, F.L. A strategic approach to climate. Harvard Business Review 2007, 85(10), 22-26.
567 3. Horbach, J.; Rammer, C.; Rennings, K. Determinants of eco-innovations by type of environmental impact-
568 The role of regulatory push/pull, technology push and market pull. Ecological Economics 2012, 78, 112-122.

15
Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 19 April 2019 doi:10.20944/preprints201904.0218.v1

569 4. Kesidou, E.; Demirel, P. On the drivers of eco-innovations: Empirical evidence from the UK. Research Policy
570 2012, 41, 862-870.
571 5. Cuerva, M.C.; Triguero-Cano, A.; Corcoles, D. Drivers of green and non-green innovation: empirical
572 evidence in Low-Tech SMEs. Journal of Cleaner Production 2014, 68, 104-113.
573 6. Triguero, A.; Moreno-Mondejar, L.; Davia, M.A. Drivers of different types of eco-innovation in European
574 SMEs. Ecological Economics 2013, 92, 25-33.
575 7. Cai, W.; Zhou, X. On the drivers of eco-innovation: empirical evidence from China. Journal of Cleaner
576 Production 2014, 79, 239-248.
577 8. Cheng, C.C.; Yang, C.; Sheu, C. The link between eco-innovation and business performance: a Taiwanses
578 industry context. Journal of Cleaner Production 2014, 64, 81-90.
579 9. Chesbrough H. The Era of Open Innovation. MIT Sloan Management Review 2003, 44(3), pp. 35-41
580 10. Rennings, K. Redefining innovation—eco-innovation research and the contribution from ecological
581 economics. Ecological economics 2000, 32(2), 319-332.
582 11. Negny, S.; Belaud, J.; Robles, C.; Reyes, R.; Ferrer, B. Toward an eco-innovative method based on a better
583 use of resources: application to chemical process preliminary design. Journal of Cleaner Production 2012, 32,
584 101-113.
585 12. Horbach, J. Determinants of environmental innovation – new evidence from German panel data sources.
586 Research Policy 2008, 37, 163-173.
587 13. Iraldo, F.; Testa, F.; Melis, M.; Frey, M. A literature review on the links between environmental regulation
588 and competitiveness. Environmental Policy and Governance 2011, 21(3), 210-222.
589 14. Porter, M.E.; Van der Linde, C. Green and competitive: ending the stalemate. Harvard Business Review 1995,
590 73 (5), 120-134.
591 15. Christmann, P. Effects of “best practices” of environmental management on cost advantage: The role of
592 complementary assets. Academy of Management Journal 2000, 43(4), 663-680.
593 16. Dean, T.J.; Brown, R.L. Pollution regulation as a barrier to new firm entry: Initial evidence and implications
594 for future research. Academy of Management Journal 1995, 38(1), 288-303.
595 17. Russo, M.V.; Harrison, N.S. Organizational design and environmental performance: Clues from the
596 electronics industry. Academy of Management Journal 2005, 48(4), 582-593.
597 18. Kammerer, D. The effects of customer benefit and regulation on environmental product innovation.
598 Empirical evidence from appliance manufacturers in Germany. Ecological Economics 2009, 68, 2285-2295.
599 19. Reid, E.M.; Toffel, M.W. Responding to public and private politics: Corporate disclosure of climate change
600 strategies. Strategic Management Journal 2009, 30(11), 1157-1178.
601 20. Cohen, P. The regulation of protein function by multisite phosphorylation–a 25 year update. Trends in
602 biochemical sciences 2000, 25(12), 596-601.
603 21. Maxwell, J.W.; Lyon, T.P.; Hackett, S.C. Self-Regulation and Social Welfare: The Political Economy of
604 Corporate Environmentalism. The Journal of Law and Economics 2000, 43(2), 583-618.
605 22. Cane, P. Self regulation and judicial review, Civil Justice Quarterly 1987, 6, 324-347.
606 23. Ogus, A. Rethinking self-regulation. Oxford Journal of Legal Studies 1995, 15(1), 97-108.
607 24. Scott, C.; Black, J.; Cranston, R. Cranston's Consumers and the Law, 3rd ed.; Cambridge University Press:
608 Cambridge, U.K., 2000; ISBN 10 0521605970.
609 25. Delmas, M.A.; Terlaak, A.K. A framework for analyzing environmental voluntary agreements. California
610 Management Review 2001, 43(3), 44-63.
611 26. Machiba, T. Eco-innovation for enabling resource efficiency and green growth: development of an
612 analytical framework and preliminary analysis of industry and policy practices. International Economics and
613 Economic Policy 2010, 7(2-3), 357-370.
614 27. De Marchi, V.; Grandinetti, R. Knowledge strategies for environmental innovations: the case of Italian
615 manufacturing firms. Journal of Knowledge Management 2013, 17(4), 569-582.
616 28. Laursen, K.; Salter, A. Open for innovation: the role of openness in explaining innovation performance
617 among U.K. manufacturing firms. Strategic Management Journal 2006, 27, 131-150.
618 29. Ghisetti, C.; Marzucchi, A.; Montresor, S. The open eco-innovation mode, An empirical investigation of
619 eleven European countries. Research Policy 2015, 44, 1080-1093.
620 30. Rennings, K.; Rammer, C. Increasing energy and resource efficiency through innovation – an explorative
621 analysis using innovation survey data. Czech Journal of Economics and Finance 2009, 59 (1), 442-459.

16
Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 19 April 2019 doi:10.20944/preprints201904.0218.v1

622 31. Horbach, J.; Oltra, V.; Belin, J. Determinants and specificities of eco-innovations. An econometric analysis
623 for the French and German industry based on the Community Innovation Survey. Industry and Innovation
624 2013, 20 (6), 523-543.
625 32. Penrose, E. The theory of the Growth of the firm, 4th ed.; Oxford University Press: NY, USA, 2009; ISBN
626 9780199573844.
627 33. Hwang, J.; Lee, Y.; External knowledge search, innovative performance and productivity in the Korean ICT
628 sector. Telecommunications Policy 2010, 34(10), 562-571.
629 34. Woo, Y.B.; Cho, S.; Kim, J.; Kim, B.S. Optimization-based approach for strategic design and operation of a
630 biomass-to-hydrogen supply chain. International Journal of Hydrogen Energy 2016, 41(12), 5405-5418.
631 35. Schmookler, J. Invention and economic growth. Harvard University Press. Cambridge, Mass., USA, 1996; ISBN
632 9780674432833.
633 36. Rosenberg, N. Science, invention and economic growth. The Economic Journal 1974, 84, 90-108.
634 37. Cleff, T.; Rennings, K. Determinants of environmental product and process innovation. European
635 environment 1999, 9(5), 191-201.
636 38. Hicks, C.; Dietmar, R. Improving cleaner production through the application of environmental
637 management tools in China. J. Clean. Prod. 2007, 15, 395-408.
638 39. Pujari, D. Eco-innovation and new product development: understanding the influences on market
639 performance. Technovation 2006, 26, 76-85.
640 40. Gray, W.B. The cost of regulation: OSHA, EPA and the productivity slowdown, The American Economic
641 Review 1987, 77(5), 998-1006.
642 41. Jenkins, R., 1998. Environmental regulation and international competitiveness: a review of literature and
643 some European evidence. Discussion Paper Series #9801. United Nations University. Institute for New
644 Technologies: Maastricht, Netherlands, 1998.
645 42. Smith, J.B.; Sims, W.A. The impact of pollution charges on productivity growth in Canadian brewing. The
646 RAND Journal of Economics 1985, 16(3), 410-423.
647 43. Del Rio Gonzalez, P. The empirical analysis of the determinants for environmental technological change: a
648 research agenda. Ecological Economics 2009, 68, 861-878.
649 44. Arranz, N.; Fdez de Arroyabe, J.C. The choice of partners in R&D cooperation: An empirical analysis of
650 Spanish firms. Technovation 2008, 28 (1), 88-100.
651 45. Lerner, J.S.; Tetlock, P.E. Accounting for the effects of accountability. Psychological bulletin 1999, 125(2), 255.
652 46. Khanna, M.; Deltas, G.; Harrington, D.R. Adoption of pollution prevention techniques: the role of
653 management systems and regulatory pressures. Environmental and Resource Economics 2009, 44(1), 85-106.
654 47. De Marchi, V. Environmental innovation and R&D cooperation: Empirical evidence from Spanish
655 manufacturing firms. Research Policy 2012, 41, 614-623.
656 48. Cohen, W.M.; Levinthal, D.A. Innovation and learning: the two faces of R & D. The economic journal 1989,
657 99(397), 569-596.
658 49. Cohen, W.M.; Levinthal, D.A. Fortune favors the prepared firm. Management science 1994, 40(2), 227-251.
659 50. Reichstein, T.; Salter, A. Investigating the sources of process innovation among UK manufacturing
660 firms. Industrial and Corporate Change 2006, 15(4), 653-682.
661 51. Nash, J.; Ehrenfeld, J. Codes of environmental management practice: Assessing their potential as a tool for
662 change. Annual Review of Energy and the Environment 1997, 22(1), 487-535.
663 52. Fabrizio, K.R. Absorptive capacity and the search for innovation. Research Policy 2009, 38(2), 255-267.
664 53. Koza, M.P.; Lewin, A.Y. The co-evolution of strategic alliances. Organization science 1998, 9(3), 255-264.
665 54. Xia, T.; Roper, S. From capability to connectivity—Absorptive capacity and exploratory alliances in
666 biopharmaceutical firms: A US–Europe comparison. Technovation 2008, 28(11), 776-785.
667 55. Xia, T.; Roper, S. Unpacking Open Innovation: Absorptive Capacity, Exploratory and Exploitative
668 Openness, and the Growth of Entrepreneurial Biopharmaceutical Firms. Journal of Small Business
669 Management 2016, 54(3), 931-952.
670 56. Guellec, D.; De la Potterie, B.V.P. The internationalisation of technology analysed with patent data. Research
671 Policy 2001, 30(8), 1253-1266.
672 57. Seop, H. Y.; Min, K. H.; Sik, K. H. Industry openness, firm capacity, and performance of Korean
673 SMEs. Journal of Korea Trade 2013, 17(3), 77-105.

17
Preprints (www.preprints.org) | NOT PEER-REVIEWED | Posted: 19 April 2019 doi:10.20944/preprints201904.0218.v1

674 58. Palmer, K.; Oates, W.E.; Portney, P.R. Tightening environmental standards: The benefit-cost or the no-cost
675 paradigm? The Journal of Economic Perspectives 1995, 9(4), 119-132.
676 59. Wagner, B.A. Learning and knowledge transfer in partnering: an empirical case study. Journal of Knowledge
677 Management 2003, 7(2), 97-113.
678 60. Sharma, S.; Pablo, A.L.; Vredenburg, H. Corporate environmental responsiveness strategies the importance
679 of issue interpretation and organizational context. The Journal of Applied Behavioral Science 1999, 35(1), 87-
680 108.

18

View publication stats