SMEs’ External Technology R&D Cooperation Network Diversity and Their Greenhouse Gas Emission Reduction and Energy Saving: A Moderated Mediation Analysis

Abstract

The purpose of this research is to empirically reveal the effect of external technology R&D cooperation network diversity (ETRDCND) on the greenhouse gas (GHG) emission reduction and energy saving of small and medium-sized enterprises (SMEs). Besides this, this study aims at analyzing the roles of production time reduction and absorptive capacity in the relationship between SMEs’ ETRDCND and their GHG emission reduction and energy saving. GHG emission and energy usage have been playing a crucial role in aggravating global warming. Global warming results in big problems such as worldwide unusual weather and health disorders. SMEs play a substantial role in the industrial growth of the global economy, which increases GHG emission and energy consumption. By performing the ordinary least squares regression with the data of 3300 South Korean SMEs, this research reveals four points. First, ETRDCND positively influences SMEs’ GHG emission reduction and energy saving. Second, production time reduction perfectly mediates the relationship between SMEs’ ETRDCND and their GHG emission reduction and energy saving. Third, the mediating role of production time reduction in this relationship is moderated by SMEs’ absorptive capacity. Fourth, ETRDCND significantly influences SMEs’ GHG emission reduction and their energy saving only if SMEs possess their own absorptive capacity.

1. Introduction

This research aims at empirically revealing the impact of external technology R&D cooperation network diversity (ETRDCND) on the greenhouse gas (GHG) emission reduction and energy saving of small and medium-sized enterprises (SMEs). In addition to this, this study aims at looking into the roles of production time reduction and absorptive capacity in the relationship between SMEs’ ETRDCND and their GHG emission reduction and energy saving. GHG emission and energy usage have been playing a critical role in worsening global warming [1]. Global warming is a serious obstacle to sustainable human development [1], changing the climate [2] and threatening human health across the world [3]. SMEs make a substantial contribution to the industrial growth of the global economy [4], which facilitates GHG emission and energy consumption [5]. Therefore, it is necessary to reveal factors in positively influencing SMEs’ GHG emission reduction and energy saving, which has created a high demand for studies to empirically investigate them. To make a good contribution to satisfying this demand, this study attempts to empirically look into the influence of ETRDCND on SMEs’ GHG emission reduction and energy saving.

Research into open innovation has identified SMEs’ external technology research and development (R&D) cooperation as a determinant factor of their innovation performance [6]. In line with this, recent studies on SMEs’ open innovation have confirmed ETRDCND as a critical factor in creating SMEs’ various innovation performances. For example, Hau [7] empirically showed that ETRDCND positively influenced SMEs’ productivity improvement and cost reduction. Hau [8] pointed out that ETRDCND was a significant factor in positively impacting SMEs’ import substitution. Gu, Jiang, and Wang [9] revealed that cooperation networks had a positive effect on SMEs’ revenue from new products and services.

Although recent studies on SMEs’ open innovation have revealed the impact of ETRDCND on SMEs’ various innovation performances, they seem to have several limitations. Recent studies on SMEs’ open innovation tend to aim mainly at the influence of ETRDCND on SMEs’ innovation performance creating economic values. So, they provide little knowledge about the impact of ETRDCND on such innovation performance as SMEs’ GHG emission reduction and energy saving creating environmental values. Furthermore, they shed little light on the roles of production time reduction and absorptive capacity in the relationship between SMEs’ ETRDCND and their GHG emission reduction and energy saving. Production time reduction is important for GHG emission reduction and energy saving in the perspective of energy efficiency [10]. Absorptive capacity is crucial for SMEs’ searching for and digesting valuable external knowledge from their external technology R&D cooperation network [11]. However, little is known about the roles of production time reduction and absorptive capacity in the relationship between SMEs’ ETRDCND and their GHG emission reduction and energy saving. Therefore, in order to overcome the limitations of recent studies on SMEs’ open innovation, this study addresses three research questions as follows:

(1)

What is the effect of ETRDCND on SMEs’ GHG emission reduction and energy saving?

(2)

What is the role of production time reduction in the relationship between SMEs’ ETRDCND and their GHG emission reduction and energy saving?

(3)

What role does absorptive capacity play in the relationship between SMEs’ ETRDCND and their GHG emission reduction and energy saving?

This article is organized into seven sections including this introduction: Section 2 provides a literature review on open innovation and SMEs as well as recent studies on SMEs’ open innovation. Section 3 constructs three hypotheses on the theoretical basis of open innovation. Section 4 describes the research methodology used to test the three hypotheses. Section 5 reports the hypothesis testing results. Section 6 presents theoretical and practical implications from this study. Section 7 provides suggestions for further studies.

2. Literature Review

2.1. Open Innovation and SMEs

Open innovation has drawn a lot of attention since Chesbrough coined it in 2003 [12]. Open innovation is a technology R&D paradigm emphasizing that enterprises can effectively create innovations by using their external knowledge network as well as internal knowledge network [13]. Knowledge network indicates a set of entities—individuals, groups, or organizations—in social interrelationships in which knowledge is created, shared, and advanced [14]. Diverse knowledge from the external knowledge network is one of the effective soils in creating and growing innovations for enterprises [15]. Enterprises pursuing closed innovation use knowledge only from their internal knowledge network, and the soils for innovations are confined to their internal knowledge network [16]. However, enterprises implementing open innovation can use more various knowledge from not only their internal knowledge network but also their external knowledge network to increase innovations [13].

Compared to large enterprises, SMEs are confronted by practical problems in technology R&D [6]. SMEs are limited in financial investment to support technology R&D [6]. Moreover, they have slim chances of recruiting experts to deepen or widen their internal knowledge network [17]. For these problems, open innovation can make a significant breakthrough [6]. It is substantially effective in surmounting the problems to receive heterogeneous knowledge transfusion from their external knowledge network [6]. Accordingly, SMEs have increasingly adopted open innovation as their technology R&D paradigm [17].

2.2. Recent Studies on SMEs’ Open Innvovation

External technology R&D cooperation is one of the effective open innovation strategies [18]. It enables enterprises to effectively learn valuable external knowledge from cooperation partners [19]. Moreover, it is useful for significantly increasing SMEs’ innovation performance [6]. Therefore, recent studies on SMEs’ open innovation explore the impact of external technology R&D cooperation on SMEs’ various innovation performances. However, in spite of their major findings summarized in

Table 1. Summary of recent studies on SMEs’ open innovation.

References	Major Finding Related to This Research	Targeted Innovation Performance
[11]	Absorptive capacity significantly mediates the relationship between knowledge inflows from various external knowledge sources including alliance partners and SMEs’ innovation performance.	Product innovation, service innovation
[24]	R&D cooperation plays a significant role in creating SMEs’ responsible innovation.	Product, service, or business model innovation mitigating or solving social or environmental issues
[25]	It is effective in making above average innovation performance for SMEs to implement R&D cooperation with research institutes, universities, and private businesses at the same time.	Patents, process innovation, product innovation
[18]	The broad and intensive cooperation with partners positively impacts SMEs’ performance.	Relative sales, new product development, market share
[26]	External technology collaboration network diversity positively influences SMEs’ technology management capabilities.	New technology development capability, technology commercialization capability
[9]	Cooperative networks positively influence SMEs’ innovation performance.	Revenue from new products and services
[8]	Exterior R&D network has a positive effect on SMEs’ performance, moderated by their absorptive capacity.	Import substitution
[22]	External technology cooperation network diversity has a positive influence on SMEs’ diversification.	Diversification into new business fields
[20]	Diverse external technology R&D collaboration network has a positive impact on SMES’ technology R&D performance.	Production process improvement, cost reduction
[27]	External technology collaboration network diversity positively influences SMEs’ technology-related capacity.	Technology commercialization capability
[7]	Diversity of technology R&D collaboration network is positively related to SMEs’ R&D performance.	Productivity improvement, cost reduction
[21]	The more diverse SMEs’ external technology R&D cooperation network is, the better their performance is.	Export growth, employment increase

, they seem to have three limitations as follows:

First, the recent studies aim mainly at innovation performance under the economic perspective, paying little attention to innovation performance under the climate protection perspective. Most of them concentrate on product, process, or service innovation resulting in economic values such as revenue increase [9], cost reduction [20], export growth [21], import substitution [8], or diversification into new business fields [22]. So, they can hardly shed a light on the influence of ETRDCND on SMEs’ GHG emission reduction and energy saving.

Second, the recent studies present little empirical evidence about the role of production time reduction in the relationship between SMEs’ ETRDCND and their GHG emission reduction and energy saving. Enterprises’ production time is important because their long production time can lead them to consume more energy and emit more GHG in accordance with the time spent in producing their goods in terms of energy efficiency [10]. However, the recent studies seem not to empirically examine the role of production time reduction in the relationship between SMEs’ ETRDCND and their GHG emission reduction and energy saving.

Third, the recent studies provide little knowledge about the role which absorptive capacity plays in the relationship between SMEs’ ETRDCND and their GHG emission reduction and energy saving. Absorptive capacity is referred to as an enterprises’ capability of searching for useful knowledge and effectively imbibing it [23]. The recent studies point out the significant role of absorptive capacity in not only mediating the impact of the external knowledge network on SMEs’ innovation performance [11] but also moderating the impact of it [8]. However, the recent studies seem not to empirically illuminate the role which absorptive capacity plays in the relationship between SMEs’ ETRDCND and their GHG emission reduction and energy saving. Therefore, this study makes three hypotheses to overcome the three limitations of the recent studies and empirically test them. Table 1 reports the summary of the recent studies on SMEs’ open innovation in terms of their major findings related to this study and targeted innovation performance.

3. Theoretical Basis and Hypothesis Construction

This study constructs three hypotheses on the theoretical basis of open innovation. Hypothesis 1 deals with the positive and direct impact of ETRDCND on SMEs’ GHG emission reduction and energy saving. Firms’ technology R&D results in various innovation performances for climate protection [28] as well as industrial growth [29]. However, SMEs are limited in their internal technology R&D resources and capabilities to support technology R&D [6]. According to Pfeffer and Salancik [30], enterprises are limited to their environmental constraints but they can change the constraints through cooperation in their interfirm network. In line with this, open innovation points out that external technology R&D cooperation with various partners is an effective way of overcoming SMEs’ limitations in their internal technology R&D resources and capabilities [6]. It is useful for increasing SMEs’ innovation performance to make external technology R&D cooperation with various external knowledge sources [25]. This leads to the following Hypothesis 1:

Hypothesis 1:

SMEs’ ETRDCND positively influences their GHG emission reduction and energy saving.

Hypothesis 2 treats the mediating role of production time reduction in the relationship between SMEs’ ETRDCND and their GHG emission reduction and energy saving. Innovating production processes is one of the effective ways of enabling enterprises to reduce their production time [29]. External technology R&D cooperation enables firms to extend the scope of their internal knowledge base for process innovation [13]. ETRDCND has a positive impact on SMEs’ production process improvement [20]. Accordingly, ETRDCND can exert a positive influence on SMEs’ production time reduction.

Enterprises’ long production time can lead them to consume more energy and emit more GHG in accordance with the time spent in their production of goods [10]. Greater production time reduction results in greater GHG emission reduction and energy saving [31]. Therefore, production time reduction can positively influence SMEs’ GHG emission reduction and energy saving. Considering the impact of ETRDCND on SMEs’ production time reduction and the effect of production time reduction on SMEs’ GHG emission reduction and energy saving, this study constructs Hypothesis 2 as follows:

Hypothesis 2:

Production time reduction mediates the relationship between SMEs’ ETRDCND and their GHG emission reduction and energy saving.

Hypothesis 3 deals with the moderating effect of absorptive capacity on the mediating role of production time reduction in Hypothesis 2. Absorptive capacity is necessary for enterprises to successfully sense and digest valuable external knowledge for technological innovation [29]. Absorptive capacity enables enterprises to successfully apply useful knowledge from external technology R&D cooperation to their technological innovation [23]. ETRDCND does not significantly influence SMEs’ innovation performance without absorptive capacity [8]. Accordingly, it can be difficult for ETRDCND to significantly impact SMEs’ production time reduction without absorptive capacity. According to the Baron and Kenny test [32] to examine the mediation effect, the significant influence of ETRDCND on production time reduction is one of the essential conditions for production time reduction to become a significant mediator. Accordingly, the mediating role of production time reduction in Hypothesis 2 can depend on SMEs’ absorptive capacity. This generates the following Hypothesis 3:

Hypothesis 3:

Absorptive capacity moderates the mediating role of their production time reduction in the relationship between SMEs’ ETRDCND and their GHG emission reduction and energy saving.

In addition to considering the impact of ETRDCND and the roles of production time reduction and absorptive capacity in Hypotheses 1 through 3, this study takes the effect of SMEs’ size into consideration by using it as a control variable for testing the hypotheses due to the positive influence of enterprises’ size on their innovation performance resulting from technology R&D [6].

4. Research Methodology

4.1. Data

This study used the 3300 data in the 2017 SMEs’ Technology Statistics (2017 SMETS) to empirically test the three hypotheses. The 2017 SMETS is a data set resulting from a survey concerning South Korean SMEs’ technology R&D management and performance in 2016. The survey was carried out by the Ministry of SMEs and Startups and the Korea Federation of SMEs (KBIZ) in the Republic of Korea.

Table 2. The profile of data (n = 3300).

Item	Option	Frequency (%)
Enterprise Type	General SMEs	1691 (51.2%)
	Venture SMEs	936 (28.4%)
	Other	673 (20.4%)
Technology Sector	Chemical	632 (19.2%)
	Machinery•Material	941 (28.5%)
	Bio•Medical	268 (8.1%)
	Energy•Resource	128 (3.9%)
	Information & communication	305 (9.2%)
	Other	1026 (31.1%)
Location	Seoul	484 (14.7%)
	Other	2816 (85.3%)

reports the profile of the 3300 data used in this study in terms of enterprise type, technology sector, and location.

4.2. Measurement

This study used five variables, such as ETRDCN, production time reduction, absorptive capacity, GHG emission reduction, and energy saving, and SMEs’ size to test the hypotheses.

Table 3. Operational definition and measurement for variables.

Variable	Type	Operational Definition	Measurement
ETRDCND	Independent variable	The number of different kinds of partners with which SMEs made external technology R&D cooperation in 2016	Eight-point scale
Production time reduction	Mediating variable	The degree of SMEs’ production time reduction resulting from technology R&D in 2016	Five-point scale
Absorptive capacity	Moderating variable	The existence of SMEs’ technology R&D organization	Binary scale
GHG emission reduction and energy saving	Dependent variable	The degree of SMEs’ GHG emission reduction and energy saving resulting from technology R&D in 2016	Five-point scale
SMEs’ size	Control variable	The size of SMEs in terms of their total sales in 2016	Total sales in 2016

summarizes them in terms of their type, operational definition, and measurement.

This study gauged SMEs’ ETRDCND by adapting Tsai’s [33] measurement of collaborative network diversity for the context of South Korean SMEs’ technology R&D. SMEs can make external technology R&D cooperation with various partners [22]. These partners can be classified into seven types as follows: (1) university, (2) public or national research institute, (3) private research institute, (4) large-sized enterprise, (5) medium-sized enterprise, (6) small-sized enterprise, and (7) foreign enterprise or organization. This study measured the number of different kinds of partners with which SMEs made external technology R&D cooperation in 2016. For example, if an SME did not make any external technology R&D cooperation with the partners, its ETRDCND was evaluated to be zero. If an SME made external technology R&D cooperation with all of types of the partners, its ETRDCND was evaluated to be seven.

This study measured production time reduction based on the five-point scale from Hau [34]. The value of one in it indicated “very little degree” of SMEs’ production time reduction resulting from technology R&D in 2016 but the value of five stood for “very much degree” of it.

The existence of technology R&D organization for SMEs is a proxy for their absorptive capacity [8,35]. Therefore, this study checked it by using a binary scale with either the value of one meaning the existence of technology R&D organization or the value of zero indicating the nonexistence of it.

In a very similar way to the measurement of production time reduction, GHG emission reduction and energy saving were measured based on the five-point scale from Hau [36]. In this scale, the value of one meant “very little degree” of SMEs’ GHG emission reduction and energy saving resulting from technology R&D in 2016. The value of five in the scale indicated “very much degree” of it.

Enterprises’ size can be measured with their sales [37]. Therefore, the total sales of SMEs in 2016 were gauged for the control variable for this study.

4.3. Analysis Method

This research performed ordinary least squares (OLS) regression [38] with the IBM SPSS version 23 to test the three hypotheses. Regression analysis is one of the most preferred statistical methods in various fields [39]. It is effective in analyzing the direct, mediating, or moderating relationships between variables being studied [40]. This study needed to test the direct relationship in Hypothesis 1, the mediating relationship in Hypothesis 2, and the moderating relationship in Hypothesis 3. Therefore, this research statistically tested the significances of the Hypotheses 1 through 3 based on the OLS regression equations as follows:

$$\mathrm{OLS}\mathrm{regression}:\mathrm{Y}={\alpha }_0+{\alpha }_{1}X+{\alpha }_{2}C+{\epsilon }_1$$

(1)

$$\mathrm{OLS}\mathrm{regression}:\mathrm{Y}={\beta }_0+{\beta }_{1}X+{\beta }_{2}M+{\beta }_{3}C+{\epsilon }_2$$

(2)

$$\mathrm{OLS}\mathrm{regression}:\mathrm{M}={\gamma }_0+{\gamma }_{1}X+{\epsilon }_3$$

(3)

In the regression equations, Y stands for SMEs’ GHG emission reduction and energy saving, X for SMEs’ ETRDCND, $M$ for SMEs’ production time reduction, and C for SMEs’ size in terms of sales. The ${\alpha }_0$, ${\beta }_0$, and ${\gamma }_0$ denote the constant term in each regression equation. The ${\epsilon }_1$, ${\epsilon }_2$, and ${\epsilon }_3$ are the error terms in the regression equations. The ${\alpha }_1$, ${\beta }_1$, and ${\gamma }_1$ are the regression coefficients for X, ${\beta }_2$ for $M$, and ${\alpha }_2$ and ${\beta }_3$ for C in the regression Equations (1) through (3).

This study took three steps to test the Hypotheses 1 through 3 by using the OLS regression equations as follows:

In the first step, the direct effect in Hypothesis 1 was tested with the OLS regression Equation (1) based on the total data (n = 3300).

In the second step, this research carried out the Sobel test [41] and the Baron and Kenny test [32] to examine the mediation impact in Hypothesis 2 by using the OLS regression Equations (1) through (3) with the total data. This study calculated the z-value from the Sobel test [41] by using the following equation. In this equation, the $S_{{\beta }_2}$ and $S_{{\gamma }_1}$ denote the standard error of ${\beta }_2$ and ${\gamma }_1$, respectively.

$$\mathrm{Z}=\frac{{\gamma }_1\times {\beta }_2}{\sqrt{{\gamma }_1^2{S}_{{\beta }_2}^2+{\beta }_2^2{S}_{{\gamma }_1}^2}}$$

(4)

The Baron and Kenny test [32] requires four conditions necessary for empirically confirming the mediation effect in Hypothesis 2: The first condition is that the ${\alpha }_1$ must be significant in OLS regression Equation (1). The second condition is that ${\beta }_2$ must be significant in OLS regression Equation (2). The third condition is that ${\gamma }_1$ must be significant in OLS regression Equation (3). The fourth condition is that ${\beta }_1$ must be insignificant or smaller than ${\alpha }_1$. The $M$ in OLS regression Equation (2) proves to be a partial mediator if ${\beta }_1$ is smaller than ${\alpha }_1$ [32]. The $M$ turns out to be a perfect mediator if ${\beta }_1$ is insignificant [32].

In the third step, this research tested the moderated mediation effect in Hypothesis 3. The moderated mediation effect turns out to be significant if the mediation effect relies on the moderator [42]. Therefore, in the third step, this study performed the Sobel test [41] and the Baron and Kenny test [32] to check the significance of the mediation effect in Hypothesis 2 according to two different groups: one with technology R&D organization (n = 2020) and the other without it (n = 1280).

5. Hypothesis Testing Result

The OLS regression results supported Hypothesis 1 at the level of significance at 0.1. The empirical analysis results based on OLS regression Equation (1) with the total group (n = 3300) indicated that ETRDCND had a positive impact on SMEs’ GHG emission reduction and energy saving (${\alpha }_1$ = 0.028, t-value = 1.730).

Hypothesis 2 was supported. As illustrated in Figure 1, the analysis results showed that production time reduction significantly mediated the relationship between SMEs’ ETRDCND and their GHG emission reduction and energy saving (the z-value from the Sobel test [41] = 4.675). ETRDCND positively influenced SMEs’ production time reduction (${\gamma }_1$ = 0.152, t-value = 5.395). SMEs’ GHG emission reduction and energy saving was positively impacted by their production time reduction (${\beta }_2$ = 0.092, t-value = 9.460). The analysis results met the four necessary conditions required by the Baron and Kenny test [32]. Moreover, ${\beta }_1$ was insignificant in the OLS regression Equation (2) (${\beta }_1$ = 0.014, t-value = 0.867). This confirms that production time reduction is a perfect mediator according to the Baron and Kenny test [32].

Hypothesis 3 was supported. In the group with technology R&D organization, production time reduction was found out to be a significant mediator as illustrated in Figure 2. However, it turned out to be an insignificant mediator in the group without technology R&D organization. This reveals that the mediating effect of production time is significant only in the group with technology R&D organization. This proves that absorptive capacity represented by technology R&D organization significantly moderates the mediating role of production time reduction, confirming the significance of the moderated mediation effect in Hypothesis 3 [42].

More specifically, in the group with technology R&D organization, the analysis results indicated that the z-value from the Sobel test [41] was 4.159, confirming the significant mediating role of production time reduction. In line with this, they satisfied the four conditions for the Baron and Kenny test [32]. According to the results from this test, production time reduction proved to be a perfect mediator as seen in Figure 2. However, in the group without technology R&D organization, ETRDCND was found out not to have a significant impact on SMEs’ GHG emission reduction and energy saving (${\alpha }_1$ = 0.027, t-value = 0.743). This did not satisfy the first condition for the Baron and Kenny test [32], proving that production time reduction was not a significant mediator in the group without technology R&D organization. This reveals that the mediating impact of production time reduction is moderated by SMEs’ technology R&D organization representing their absorptive capacity, confirming the significant moderated mediation effect in Hypothesis 3 [42].

In addition, ETRDCND significantly and positively influenced SMEs’ GHG emission reduction and energy saving in the group with technology R&D organization (${\alpha }_1$ = 0.033, t-value = 1.910). However, the positive impact of ETRDCND was insignificant in the group without technology R&D organization (${\alpha }_1$ = 0.027, t-value = 0.743).

Variance inflation factors (VIFs) were examined to check the potential for multicollinearity in the OLS regressions for this study. The range of the VIFs in this research was from 1.001 to 1.011, showing few signs of multicollinearity in this study.

Table 4. Summary of hypothesis testing results.

Hypothesis: Type of Effect	Result	Empirical Analysis Result
Hypothesis 1: direct effect	Supported	${\alpha }_1$ = 0.028 * in TG 1
Hypothesis 2: mediation effect	Supported	ZVFST 2 = 4.675 *** in TG 1
		${\alpha }_1$ = 0.028 * in TG 1
		${\beta }_2$ = 0.092 *** in TG 1
		${\gamma }_1$ = 0.152 *** in TG 1
		${\beta }_1$ = 0.014 ns in TG 1
Hypothesis 3: moderated mediation effect	Supported	GWTRDO 3	GWOTRDO 4
		ZVFST 2 = 4.159 ***
		${\alpha }_1$ = 0.033 *	${\alpha }_1$ = 0.027 ns
		${\beta }_2$ = 0.111 ***
		${\gamma }_1$ = 0.149 ***
		${\beta }_1$ = 0.017 ns

¹ Total Group (n = 3300); ² Z-value from the Sobel Test [41]; ³ Group with Technology R&D Organization (n = 2020); ⁴ Group without Technology R&D Organization (n = 1280); * P < 0.1; ** P < 0.05; *** P < 0.01; ns = not significant.

provides the summary of hypothesis testing results.

6. Discussion

6.1. Summary of Findings

Based on the OLS regression analysis results by using the 3300 data of South Korean SMEs, this research provides four findings as follows: First, ETRDCND positively and significantly influences SMEs’ GHG emission reduction and energy saving. Second, production time reduction plays a significant role in perfectly mediating the relationship between SMEs’ ETRDCND and their GHG emission reduction and energy saving. Third, the mediating role of production time reduction in this relationship is significantly moderated by SMEs’ absorptive capacity. Fourth, ETRDCND significantly influences SMEs’ GHG emission reduction and their energy saving only if SMEs possess their own absorptive capacity.

6.2. Theoretical Implication

The findings from this study are expected to present meaningful theoretical implications which recent studies on SMEs’ open innovation have seldom provided up to now as follows: First, this study shed a new light on the significant and positive impact of ETRDCND on SMEs’ GHG emission reduction and energy saving. The recent studies have aimed mainly at the relationship between SMEs’ ETRDCND and innovation performance creating economic values. For example, they have concentrated mainly on innovation performance such as cost reduction [20], relative sales [18], import substitution [8], export growth [21], and diversification into new business fields [22]. However, this study empirically reveals an important implication that ETRDCND positively influences SMEs’ innovation performance creating environmental values such as GHG emission reduction and energy saving.

Second, this study illuminates the role of production time reduction in the relationship between SMEs’ ETRDCND and their GHG emission reduction and energy saving. Production time reduction is crucial for GHG emission reduction and energy saving in terms of energy efficiency [31]. However, little is known about the role which production time reduction plays in the relationship between SMEs’ ETRDCND and their GHG emission reduction and energy saving. Concerning the role of production time reduction in this relationship, this research provides a fresh implication. This study empirically shows that production time reduction is a perfect mediator in the relationship between SMEs’ ETRDCND and their GHG emission reduction and energy saving.

Third, this study reveals that absorptive capacity significantly moderates the mediating role of production time reduction in the relationship between SMEs’ ETRDCND and their GHG emission reduction and energy saving. The importance of absorptive capacity has been emphasized by recent studies on SMEs’ open innovation [8]. However, they have rarely explored the role which absorptive capacity plays in the relationship between SMEs’ ETRDCND and their GHG emission reduction and energy saving. This study illuminates the critical role of SMEs’ absorptive capacity in the relationship between them by empirically revealing that the mediating role of production time reduction is significant only if SMEs have their own absorptive capacity. Furthermore, this study highlights the importance of SMEs’ absorptive capacity by empirically showing that SMEs’ ETRDCND does not significantly influence their GHG emission reduction and energy saving without absorptive capacity.

6.3. Practical Implication

The findings from this research are expected to provide useful practical implications as follows: First, this study highlights the importance of increasing SMEs’ ETRDCND for reducing their GHG emission and energy usage. GHG emission and energy usage have been playing a critical role in aggravating global warming [1]. Global warming results in big problems such as worldwide unusual weather [1] and health disorders [3]. SMEs substantially contribute to the industrial growth of the global economy [4], which promotes GHG emission and energy consumption [5]. Therefore, SMEs’ strategic efforts to reduce them is important for mitigating global warming [2]. This study reveals that increasing ETRDCND can be an effective strategic effort to reduce SMEs’ GHG emission and energy consumption.

Second, this research points out that external technology R&D cooperation can be a useful breakthrough for SMEs’ limitation in internal resources and capacities by empirically proving the positive and significant impact of ETRDCND on SMEs’ production time reduction. Production process improvement can require technology R&D resources and capacities beyond SMEs’ internal knowledge network [20]. SMEs do not have enough internal technology R&D resources and capabilities [6]. However, Pfeffer and Salancik [30] and Chesbrough [13] point out that SMEs can surmount their limitation in them by using technology R&D cooperation with various partners in their external knowledge network. In line with this, this study provides empirical evidence that SMEs’ production time reduction is positively influenced by their ETRDCND.

Third, this study reveals the important role which absorptive capacity plays in the relationship between SMEs’ ETRDCND and their GHG emission reduction and energy saving. This study empirically shows that ETRDCND positively influences SMEs’ GHG emission reduction and energy saving only in the group with absorptive capacity. Therefore, it is necessary for SMEs to possess their own absorptive capacity in order to make their GHG emission reduction and energy saving successful through external technology R&D cooperation.

7. Suggestion for Further Studies

Suggestions for better studies are provided on the basis of the limitations of this research as follows: First, this study used only a network measure such as ETRDCND. It will be better for future studies to use other network measures such as the centrality or density of external technology R&D cooperation network to widen their analysis results.

Second, this research focused only on GHG emission reduction and energy saving as innovation performance creating environmental values. It will be effective in producing more implications about the relationship between SMEs’ ETRDCND and their environment-friendly innovation performance to consider water protection or soil conservation in future studies.

Third, this study analyzed the impact of ETRDCND on enterprises’ GHG emission reduction and energy saving only in the context of SMEs. However, it will be very useful for making good insights to empirically compare the difference in the impact of ETRDCND on them according to enterprises’ size.

Fourth, this study performed the analyses by using the cross-sectional data, which makes it impossible to consider the change of impacts of ETRDCND on SMEs’ greenhouse gas emission reduction and energy saving as time progresses. So, analyses with simulation or times series data will be more useful for further studies.

Fifth, the data analyzed in this study is based on South Korean SMEs. If further studies use the data from SMEs in various countries, it will be useful for increasing the generalizability of findings from their analysis results.