Impact of Energy Enterprise Exports on Technological Innovation: PSM Analysis Based on Chinese Industrial Enterprise Database

Abstract

Export is an important way to promote the innovation and growth of energy enterprises by learning new knowledge and technology. To explore the impact mechanism of energy enterprises’ export behaviour on technological innovation, this article uses the micro data of 8,548 energy enterprises in the Chinese Industrial Enterprise Database from 2004 to 2007 for propensity score matching analysis, from the two dimensions of technological innovation willingness and technological innovation capability. The empirical results show that (a) the export behaviour of energy enterprises has a significant promotion effect on the technological innovation willingness and capability of energy enterprises; export behaviour has a continuous promotion effect on the technological innovation willingness of energy enterprises, and this promotion effect has gradually increased. There is a time lag in the promotion effect of export behaviour on the technological innovation capability of energy enterprises, and this promotion effect is gradually increasing. (b) Compared with non-state-owned energy enterprises, export behaviours have a more immediate effect on the technological innovation willingness of state-owned energy enterprises; compared with small and medium-sized energy enterprises, export behaviours have a more positive impact on the technological innovation willingness and capabilities of large energy enterprises.

Keywords

Energy enterprise export behaviour technological innovation willingness technological innovation capability PSM

Introduction

The energy industry is the foundation of industrial development. The rapid global industrial development is accompanied by serious problems, such as high energy consumption, environmental pollution and resource shortage. The energy sector, including the activities undertaken by its constituents, is considered to be particularly environmentally sensitive, mainly attributable to the use of fossil fuels. To give an example, it is estimated that 35% of the global GHG emissions stem from the operations of the energy sector (Efren et al., 2022). The climate challenge and energy environment crisis caused by it have intensified, which increases the urgency of innovation in energy enterprises (Zhong et al., 2022). The technological innovation of energy enterprises not only helps promote economic benefits but also provides guarantees for global energy security, thereby enhancing environmental benefits (Brutschin & Fleig, 2016; Danish & Recep, 2022). In the process of enterprise technological innovation, technological innovation willingness represents the importance and desire of enterprises for technological innovation, and directly affects the level of enterprise innovation investment (Jiang & Liu, 2021); technological innovation capability can promote the research and development of new products into the market, promote the improvement of production efficiency and reduce production costs (Wang & Luo, 2020). With the rapid change in the industry, energy enterprises are also facing the challenge of the global market, which requires enterprises to continuously improve the level of technological innovation to adapt to the development of international market and make faster responses (Muoz et al., 2022). A large number of studies show that there is a strong correlation between export behaviour and technological innovation (Love & Roper, 2016). Export is not only a strong driving force for business growth, but also provides opportunities for opening up new markets and accepting new competition. Enterprises that actively enter the international market will participate in multinational competition, which is conducive to encouraging local enterprises to carry out R&D and innovation. At the same time, enterprises entering the international market can access the knowledge, technology and management experience of overseas markets, and improve their innovation awareness and absorption capacity in the face of international market pressure to improve the level of technological innovation (Grossman & Helpman, 1990; Grossman & Helpman, 1991; Salomon & Shaver, 2005). The continuous advancement of the process of energy reform requires the technological renewal of energy enterprises to have a global vision. Therefore, the research on the export behaviour and technological innovation of energy enterprises highlights its significance in the times.

This article uses the micro-data of 8,548 energy enterprises in the Chinese Industrial Enterprise Database from 2004 to 2007 to study the impact of energy enterprises’ export behaviour on technological innovation. The contributions of this article are as follows: first, it complements the confirmation of the role of export behaviour in promoting technological innovation in the energy industry and getting industry-specific results. Second, from the perspective of technological innovation willingness and technological innovation ability, this article analyses the impact of energy enterprises’ export behaviour on technological innovation, establishes a model according to different enterprise ownership and enterprise scale and expands the relevant research. Thirdly, this article uses the propensity score matching (PSM) method to evaluate the impact of export behaviour on technological innovation and separates the relevant factors affecting technological innovation due to sample selection in the model, to confirm that export behaviour is a causal effect of technological innovation of energy enterprises rather than a simple correlation effect.

The structure of this article is as follows: the second part is the literature review, which expounds on the theoretical basis of the impact of export behaviour on technological innovation; the third part is the research design, which introduces the PSM model used in this article; the fourth part is the empirical test; the fifth part is the conclusions and policy implications and the sixth part is the limitations and future research directions.

Literature Review

Technological innovation plays an important role in the development of the energy industry. However, technological innovation is affected by many factors, including the enterprise’s own features, such as investment in R&D (Shefer & Frenkel, 2005; Xu et al., 2021), human capital (Lenihan et al., 2019), absorptive capacity (Benhayoun et al., 2020; Harris et al., 2021) and business innovation modes (Parrilli et al., 2020). At the same time, technological innovation is also affected to a certain extent by the process of interaction between the enterprise and the external environment, such as government subsidies (Su & Li, 2021; Wu et al., 2020), supply chain collaboration (Hofman et al., 2020), policy support (Yang et al., 2020), FDI (Ahmad et al., 2020; Yang et al., 2021) and export behaviour (Zhou et al., 2020. Among them, export behaviour is considered a key factor affecting enterprise technological innovation. Previous studies generally believe that entering the export market has a better performance in technological innovation (Huang et al., 2018); from the trade theory, the possibility of the existence of this relationship can be explained, one is the self-selection effect, the other is export learning effect.

The concept of self-selection effects is that innovation will increase the productivity of enterprises and affect their export decisions, that is, enterprises with a higher level of technological innovation will choose to export (Clerides et al., 1996). Enterprises with stronger technological innovation can undertake the fixed cost of overseas marketing to enter the export market. Enterprises entering the overseas market can observe the gap between their own innovation level and the international market, thereby eliminating the relatively poor innovative enterprises (Melitz, 2003), which means technological innovation is a necessary condition for export behaviour (Dai et al., 2020; Muoz et al., 2022). In the actual production process, technological innovation maybe not only an important driving factor for export enterprises, but also an important driving factor for non-export enterprises. Non-export enterprises need to compete with export enterprises at home and are also preparing to compete with foreign competitors (Gerschewski et al., 2019); the self-selection effect proves that export behaviour drives the improvement of the technological innovation level of the whole industry (Melitz, 2003).

Another theory is the export learning effect, which means that enterprises gain experience in the fierce competition in the international market through export to promote enterprise innovation (Cintio et al., 2020; Grossman & Helpman, 1990; Grossman & Helpman, 1991; Muoz et al., 2022). From the perspective of technological upgrading, in order to meet the delivery requirements of foreign consumers (orderers), export-oriented absorption capacity enhancement or advanced equipment transfer will be adopted by export enterprises, and the increased R&D expenditure will form ‘induced technology upgrading’, so as to improve the potential of technological innovation (Dai & Yu, 2012; Gkypali et al., 2021, Peluffo, 2016). From the perspective of knowledge spill over, export enterprises can get access to knowledge and technology in overseas markets (Salomon & Shaver, 2005). The higher the knowledge stock is, the more convenient the knowledge exchange between professions will lead to the faster emergence of new knowledge and technology (He & Huang, 2021), thus forming a virtuous circle. Backward individuals can also obtain huge benefits from the external diffusion of technology in the advanced international trade market. The spill over effect of technology promotes export enterprises to improve the level of technological innovation (He & Huang, 2020; Shapiro & Walker, 2018). From the perspective of enterprise decision-making, the process of export learning makes them better understand the needs and technical obstacles of the global market, face richer information and complex market pressure (Grossman & Helpman, 1990; Grossman & Helpman, 1991) and enterprises experience ‘search and learning’ to absorb external knowledge that better matches their own strategy (Salomon & Jin, 2010), adjust the technical state and make innovative decisions in order to improve their competitive advantage and market power (Wu et al., 2020).

Furthermore, scholars have mainly used the following methods to study the relationship between export behaviour and technological innovation: structural equation modelling (Gkypali et al., 2018), multiple regression analysis (Falk & Figueira, 2019; Montalbano & Nenci, 2018), nonlinear regression analysis (Rodil et al., 2016), factor analysis (Boso et al., 2019), etc.

In conclusion, the literature has already analysed the theoretical basis of the relationship between exports and technological innovation. However, these studies mainly analyse how the export behaviour affects technological innovation through the self-selection effect and export learning effect, but less confirm that export behaviour is the inherent causal effect of technological innovation. At the same time, few scholars have carried out such research on the energy industry.

Research Design

Method

Considering that export enterprises have a stronger level of technological innovation, enterprises with a higher level of technological innovation are more inclined to export trade. This may lead to selectivity bias in the sample and bias in the result of parameter estimation. Because the principle of PSM method is to construct propensity scores according to multidimensional characteristic variables and find the closest sample from the control group to match with the treatment group, it is used to deal with the samples in this study to analyse the causal relationship. The basic idea of PSM is as follows: based on the existing control group (energy enterprises that do not conduct export activities), the individual characteristics of the treatment group (energy enterprises that conduct export activities) are matched as closely as possible with those of the new control group. The principle of such matching is that the control group and treatment group are different in whether they perform export behaviour, and other covariates that determine individual characteristics are as similar as possible. The purpose of this research is to examine the net effects of export behaviour on technological innovation after eliminating other related factors that may affect technological innovation. The control group and treatment group are two samples with very similar individual characteristics as follows:

E (Y_{i}) = E (Y_{}^{1}_{i}^{} | S = 1) - E (Y_{}^{0}_{i}^{} | S = 1)

(1)

Among them, Y_i is the outcome variables, regardless of the export behaviour of energy enterprises, and are the embodiment of their technological innovation. S is a processing variable, which indicates whether the energy enterprise conducts export behaviour; 1 means that there is export behaviour, and 0 means that there is no export behaviour. Y⁰_i means that the outcome variable of the enterprise that conducts export behaviour under the assumption of no export behaviour cannot be observed, but variable set X can be constructed and matched with Y⁰_i individuals to form a new control group as follows:

E (Y_{}^{1}_{i}^{} | S = 1, X = {x_{i}}) = E (Y_{}^{0}_{i}^{} | S = 0, X = {x_{i}})

(2)

Propensity score matching is used to transform a multidimensional variable set into a one-dimensional variable:

E (Y_{i}) = E (Y_{}^{1}_{i}^{} | S = 1, P_{}^{1}_{i}^{} (X = {x_{i}})) - E (Y_{}^{0}_{i}^{} | S = 0, {\tilde{P}}_{}^{1}_{i}^{} (X = {x_{i}}))

(3)

P¹_i represents the probability of an individual in the treatment group performing an export behaviour in the variable set X, and P¹_i represents the probability of an individual in the control group being closest to the individual in the treatment group, so that the two groups of samples can avoid sample selection bias as much as possible. N₁ denotes the number of samples in the treatment group, and the control group matched with the treatment group at S = 0 will finally form the average treatment effect of participants, which is used to evaluate the differences in technological innovation between energy enterprises with export behaviour and those without export behaviour:

A T T = \frac{1}{N_{1}} \sum_{i; S_{i} = 1} [(Y_{i}^{}_{}^{S = 1} - Y_{0 i}^{}_{}^{S = 1})]

Data Sources and Variable Selection

Since 2008, the ‘Chinese Industrial Enterprise Database’ has no longer stored data on technological innovation, and thus, this article selects all energy enterprises in the database from 2004 to 2007. The OECD and IRA believe that fossil fuels, nuclear energy and hydroelectric power generation should be divided into either economic or social energy systems (Bointner, 2014). Based on this principle, this study selected all energy enterprises under industry codes 06, 07, 25, 44 and 45 in the Chinese Industrial Enterprise Database as the sample selection objects because the export delivery value and corporate characteristics disclosed in 2004 were the most complete and the indicators related to technological innovation disclosed in 2005 to 2007 were relatively complete and consistent in statistical calibre. Therefore, this article uses the export delivery value of energy enterprises in 2004 as a proxy variable for export behaviour (Zhu & Zhao, 2017; Yu et al., 2011) to examine the impact of export practices on innovation in energy technologies over the next three years. Due to the lack of indicators in the database, abnormal data and differences in the enterprises covered in different years, this article matches, eliminates and filters the original data and finally obtains 8,548 energy enterprises for empirical research.

According to the research background and existing research experience, the relevant variables for text selection mainly include the following: (a) The outcome variable is technological innovation, the measure of which includes two dimensions, namely, technological innovation willingness (Willingness) and technological innovation capability (Capability). Technological innovation willingness is measured by the proportion of R&D expenditure in the total industrial sales output value of an enterprise. This variable represents the degree of importance that an enterprise attaches to technological research and development and is used to characterise the intensity of the technological innovation willingness of energy enterprises. Technological innovation capability is measured by the output value of new products and industry measured by the proportion of sales output value. This variable can reflect the level of absorption and transformation of the new knowledge of an enterprise and the ability of energy enterprises to innovate in commercial value transformation (Kang, 2018). (b) The processing variable is export behaviour (Exp), which is represented by a dummy variable. If the energy enterprise conducts an export behaviour, that is, there is an export delivery value, the value is set to 1; otherwise, it is set to 0. (c) Covariates: Labour productivity (L) is measured by the ratio of the total industrial output value of energy enterprises to the total number of employees. Financing constraint (Fin) is measured by the ratio of energy enterprises’ interest expenditures to fixed assets. Ownership structure (Str) is measured by the ratio of state-owned paid-in capital to total paid-in capital. Operating time (age) is measured by the difference between the current year and the year when the enterprise was founded. Government subsidy (S) is set as a dummy variable; if the energy enterprise receives government subsidies, then the value is set to 1; otherwise, it is set to 0. Labour union construction (Lab) is set as a dummy variable; if the energy enterprise has labour union expenditures, then the value is set to 1; otherwise, it is set to 0. Employee training (Edu) is set as a dummy variable; if the energy enterprise has employee education expenditures, then the value is set to 1; otherwise, it is set to 0. This article controls for the industry effect in the empirical process. The variable measure and data source are shown in Table 1.

Table 1.

Variable Measure and Data Source

	Variable	Measure	Data Source
Outcome variable	Technological innovation willingness (Willingness)	The proportion of R&D expenditure in the total industrial sales output value of an enterprise.	Chinese Industrial Enterprise Database
	Technological innovation capability (Capability)	The output value of new products and industry measured by the proportion of sales output value.	Chinese Industrial Enterprise Database
Processing variable	Export behaviour (Exp)	A dummy variable. If the energy enterprise conducts an export behaviour, the value is set to 1; otherwise, it is set to 0.	Chinese Industrial Enterprise Database
Covariate	Labour productivity (L)	The ratio of the total industrial output value of energy enterprises to the total number of employees.	Chinese Industrial Enterprise Database
	Financing constraint (Fin)	The ratio of energy enterprises’ interest expenditures to fixed assets.	Chinese Industrial Enterprise Database
	Ownership structure (Str)	The ratio of state-owned paid-in capital to total paid-in capital.	Chinese Industrial Enterprise Database
	Operating time (age)	The difference between the current year and the year when the enterprise was founded.	Chinese Industrial Enterprise Database
	Government subsidy (S)	Dummy variable; if the energy enterprise receives government subsidies, the value is set to 1; otherwise, it is set to 0.	Chinese Industrial Enterprise Database
	Labour union construction (Lab)	Dummy variable; if the energy enterprise has labour union expenditures, the value is set to 1; otherwise, it is set to 0.	Chinese Industrial Enterprise Database
	Employee training (Edu)	Dummy variable; if the energy enterprise has employee education expenditures, the value is set to 1; otherwise, it is set to 0.	Chinese Industrial Enterprise Database

Source: Chinese Industrial Enterprise Database, http://microdata.sozdata.com

Empirical Analysis

PSM Balance Test

In this article, PSM is adopted for analysis. Each PSM result should contain four parts: propensity score estimation, PSM balance result, p-score fitting degree and ATT value.

Table 2 reports the regression results for the covariates. To verify the rationality of the variables set by the model, this article adopts the probit and logit methods to estimate the propensity score. It can be seen from the regression results that the results obtained by the two regression methods are approximately the same, and most of the covariates involved are significant, indicating that the variable settings in this article are reasonable. At present, both probit and logit methods have been widely used in the research using PSM, but in this model, the logit method has a better regression result than the probit method. Thus, this article uses logit regression for PSM processing.

Table 2.

Regression Results of the Probit and Logit Methods

Method	Probit		Logit
Method	Coefficient	Z	Coefficient	Z
L	5.76e^-6***	3.30	1.03e^-5***	3.53
Fin	0.0042**	1.80	0.0095**	2.30
Str	–0.2302***	–2.58	–0.5207**	-2.39
Age	0.0055**	2.54	0.0128***	2.58
S	0.5734***	7.27	1.371***	7.61
Lab	0.1615	1.60	0.4336*	1.73
Edu	0.0999	1.01	0.2759	1.12

Source: The authors (calculation by Stata).

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

Whether the PSM test conclusion is reliable is closely related to whether the ‘independence hypothesis’ can be satisfied; that is, there is no significant difference in covariates between the sample of energy enterprises with export behaviour and that of the control group without export behaviour after matching. If this hypothesis is not satisfied, then it means that the individual characteristics of the two are quite different and that the sample selection is improper. Therefore, before PSM, a matching balance test is needed. Based on the above propensity score estimation, PSM was conducted. Table 3 reports the situation of samples supported by PSM. Because the control group had a sufficient number of samples, the treatment group was matched to a complete degree. Table 4 reports the balance test results of the covariates under PSM. The results show that except for the Edu variable in the technological innovation willingness model and the L variable in the technological innovation capability model, the deviations in other variables after matching are all less than 10%, and for the matched sample under the t-test, there was no significant difference between the two groups. The p-score fitting before and after sample matching is shown in Figure 1, and the fitting degree is ideal. Therefore, the selection of matching variables in this article is appropriate, and PSM is suitable for this research.

Table 3.

Common Support for PSM Processing Samples

Match	Common Support	Total
Control group	8,394	8,394
Treatment group	154	154
Total	8,548	8,548

Source: The authors (calculation by Stata).

Table 4.

Results of the Balance Test

–		Technological Innovation Willingness					Technological Innovation Capability
–	Match	Mean Value		Bias/%	Reduction Bias/%	t-Value	Mean Value		Bias/%	Reduction Bias/%	t-Value
–	Match	Control	Treatment	Bias/%	Reduction Bias/%	t-Value	Control	Treatment	Bias/%	Reduction Bias/%	t-Value
L	Before	7177.2	579.6	16.0	66.6	8.01***	7177.2	579.6	16.0	33.3	8.011***
L	After	7177.2	4974.6	5.3	66.6	0.38	7177.2	2775	10.7	33.3	0.83
Fin	Before	3.986	2.0879	11.5	22.7	2.67***	3.986	2.0879	11.5	80.1	2.67***
Fin	After	3.986	2.5179	8.9	22.7	0.79	3.986	3.609	2.3	80.1	0.16
Structure	Before	0.2958	0.3467	–11.5	78.4	–1.36	0.2958	0.3467	22.5	50.3	–1.36
Structure	After	0.2958	0.3068	–2.5	78.4	–0.22	0.2958	0.3211	–5.7	50.3	–0.51
Age	Before	33.312	29.414	22.5	93.5	3.16***	33.312	29.4140	54.6	83.3	3.16***
Age	After	33.312	33.058	1.5	93.5	0.12	33.312	33.963	–3.8	83.3	–0.29
S	Before	0.3571	0.1306	54.6	88.5	8.19***	0.3571	0.1306	27.6	85.7	8.19***
S	After	0.3571	0.3831	–6.3	88.5	–0.47	0.3571	0.3896	–7.8	85.7	–0.59
Lab	Before	0.7922	0.6711	27.6	65.2	3.18***	0.7922	0.6711	26.5	82.6	3.18***
Lab	After	0.7922	0.7500	9.6	65.2	0.88	0.7922	0.7711	4.8	82.6	0.45
Edu	Before	0.7857	0.6683	26.5	55.7	3.07***	0.7857	0.6683	–34.4	66.8	3.07***
Edu	After	0.7857	0.7338	11.7	55.7	1.07	0.7857	0.7467	8.8	66.8	0.81
Industry	Before	Control					Control
Industry	After	Control					Control

Source: The authors (calculation by Stata).

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

Figure 1.

Source: The authors (calculation by Stata).

PSM Test Results

From the empirical results in Table 5, we can see that the ATT effects of export behaviour on technological innovation willingness in the next three years are 0.0015, 0.0026 and 0.0029, which are all positive numbers, indicating that export behaviour has a positive effect on the increase in technological innovation willingness. From the perspective of significance level, the influence of export behaviour on technological innovation willingness passed the 1% significance test. From the above analysis, it can be seen that export behaviour will significantly promote technological innovation willingness in the next three phases, and with the increase in time, the promotion effect of export behaviour on technological innovation willingness will also increase. This shows that when energy enterprises enter the international market, exports enable them to access knowledge and technologies in overseas markets and face more complex market situations and more advanced technology competition, which will stimulate their awareness of technological innovation; facing international market pressures, energy enterprises will raise their awareness of technological innovation, increase capital investment in technology research and development and focus on improving their overall innovation strength (Salomon & Shaver, 2005). To increase profit margins, energy enterprises will adopt more advanced technologies, produce ‘leading technology upgrades’ and conduct advanced research and development to better learn from exports. The higher the degree of foreign trade is, the greater the willingness of energy enterprises to update their knowledge and technologies (Brandt et al., 2012; Dai & Yu, 2013).

Table 5.

The ATT Effect of Export Behaviour on Technological Innovation

Phase	Matching	Technological Innovation Willingness		Technological Innovation Capability
Phase	Matching	ATT	t-Value	ATT	t-Value
1	Before	0.0015	2.84***	0.0140	2.93***
1	After	0.0014	2.59***	0.0134	1.53
2	Before	0.0025	9.10***	0.0236	5.62***
2	After	0.0026	2.62***	0.0241	2.39***
3	Before	0.0025	2.25***	0.0425	11.56***
3	After	0.0029	2.77***	0.0429	3.32***

Source: The authors (calculation by Stata).

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

The ATT effect of export behaviour on technological innovation capability in the next three years is 0.0134, 0.0241 and 0.0429, which are all positive numbers, indicating that export behaviour has a positive role in promoting technological innovation capability. From the perspective of significance level, the effect of export behaviour on technological innovation capabilities passes the 1% significance test from the next phase. From the above analysis, it can be seen that with the increase in time, the role of export behaviour in promoting technological innovation capacity also will have a tendency to increase, and the role of export behaviour in promoting technological innovation capacity will be significantly improved from the next phase. Export behaviour will enable energy enterprises to access innovative product markets, imitate and recreate advanced knowledge and technologies, obtain more foreign advanced technical information, including information on competitive products, consumer preferences and related production technologies and encourage them to strengthen cooperation with trading partners in China’s R&D cooperation framework to increase productivity (Huang et al., 2018; Xie & Ding, 2018). This will play a certain role in promoting the output value of new products in the next year. However, a certain production cycle is required for new products from research and development to the formation of output value, and due to the limitations of factors such as digestion and absorption capacity and production cycle, it needs to produce the promotion effect of the output nature of technological innovation capability in the next two years.

In summary, export behaviour has an immediate effect on the promotion of the technological innovation willingness of energy enterprises, and at the same time, there is a certain degree of continuity. The promotion effect is gradually strengthened, export behaviour has a time lag in the promotion of energy enterprises’ technological innovation capability, and there is also a certain continuity. The promotion effect gradually increases.

Sample Testing and Expansion Analysis

Considering the transformation background of China’s industrial economy at its present stage, there will be differences in the development path and policy support degree of energy enterprises with different ownership structures. Therefore, it is necessary to test the impact of export behaviour on the technological innovation willingness and capability of energy enterprises with different ownership structures. The total sample used in this article is divided into state-owned and non-state-owned energy enterprises. The common support of the sample distribution after matching is shown in Table 6.

Table 6.

Common Support of the PSM Processing Samples of the Different Ownership Structures of Energy Enterprises

–	Matching	Common Support	Total
State-owned energy enterprises	Before	3,928	3,928
	After	45	45
	Total	3,973	3,973
Non-state-owned energy enterprises	Before	4,466	4,466
	After	109	109
	Total	4,575	4,575

Source: The authors (calculation by Stata).

The results of the ATT effect of the PSM treatment are shown in Table 7. From the perspective of the PSM results for technology innovation willingness, the ATT effect of state-owned energy enterprises increased year by year and was significantly greater than that of non-state-owned energy enterprises; from the perspective of the level of significance, the impact of the export behaviour of state-owned energy enterprises on technological innovation willingness is significant in the next three phases, while that of non-state-owned energy enterprises is significant after the next two phases.

Table 7.

ATT Effect of Energy Enterprises with Different Ownership Structures on Technological Innovation

–	Phases	Matching	Technological Innovation Willingness		Technological Innovation Capability
–	Phases	Matching	ATT	t-Value	ATT	t-Value
State-owned energy enterprises	1	Before	0.0033	2.42**	0.0220	2.88***
	1	After	0.0037	2.87***	0.0235	1.43
	2	Before	0.0070	14.13***	0.0327	4.90
	2	After	0.0071	2.20**	0.0360	1.82**
	3	Before	0.0072	5.70***	0.0358	6.84
	3	After	0.0071	2.17**	0.0370	2.00**
Non-state-owned energy enterprises	1	Before	0.0008	2.66***	0.0103	1.65*
	1	After	0.0008	1.48	0.0110	1.03
	2	Before	0.0006	1.84**	0.0195	1.84*
	2	After	0.0006	1.99**	0.0236	2.33**
	3	Before	0.0004	0.23	0.0447	8.79***
	3	After	0.0009	1.64*	0.0474	2.86***

Source: The authors (calculation by Stata).

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

It can be seen that export behaviour has a timely and obvious effect on the technological innovation willingness of state-owned energy enterprises, which have a natural connection with the government. Under the guidance of national innovation-driven policy, when energy enterprises accept the incentives of dealing with international market pressure, they have a stronger awareness of technological innovation than non-state-owned energy enterprises, so they can improve their R&D investment and actively apply for innovation fund projects or subsidies for scientific and technological research and development, thus increasing the total amount of innovation investment of energy enterprises in time.

From the PSM results for technological innovation capability, the ATT effect of state-owned energy enterprises is similar to that of non-state-owned energy enterprises. Export behaviour will significantly promote technological innovation capability after the next two phases. It can be seen that the different ownership structures of enterprises will not play a significant role in learning, absorbing and transforming new knowledge.

Considering the different scales of energy enterprises, the capital operation ability, export strategy formulation and knowledge digestion and absorption capacity of such enterprises are different; these factors will have a different impact on export behaviour in terms of the technology innovation willingness and technological innovation capability of energy enterprises. Therefore, it is necessary to examine the influence of export behaviour on the technological innovation willingness and capability of differently sized energy enterprises. The total sample of energy enterprises is divided into large, medium-sized and small enterprises. The common support of the sample distribution after matching is shown in Table 8.

Table 8.

Common Support of the PSM Processing Samples of Differently Sized Energy Enterprises

–	Matching	Common Support	Total
Large energy enterprises –	Before	296	296
	After	62	62
	Total	358	358
Medium-sized energy enterprises –	Before	2,240	2,240
	After	44	44
	Total	2,284	2,284
Small energy enterprises –	Before	5,858	5,858
	After	48	48
	Total	5,906	5,906

Source: The authors (calculation by Stata).

The results of the ATT effect of the PSM treatment are shown in Table 9. From the perspective of the PSM results for technology innovation willingness, the ATT effect value of large energy enterprises is positive, which is far greater than that of small and medium-sized energy enterprises, and it will be significant at the level of 5% after the next two phases. Large energy enterprises have a high market share and a strong capital base. When such enterprises enter the international market, they are more likely to resist R&D risks and are more willing to explore emerging markets. Therefore, the willingness to research and develop innovative products is more active. Therefore, export behaviour has a more significant impact on the technology innovation willingness of large energy enterprises. However, the export behaviour of small and medium-sized energy enterprises has no significant effect on the willingness for technological innovation. Small and medium-sized energy enterprises have weak innovation consciousness, and their business objectives are more concentrated on operation, production and traditional market share. Therefore, the total amount of capital flow to innovation and development is less in these enterprises, and their technological innovation willingness will not be increased due to export behaviour, especially for medium-sized energy enterprises in the stage of pursuing the expansion of large-scale production, and even have an insignificant negative impact trend.

Table 9.

ATT Effect of Export Behaviour of Differently Sized Energy Enterprises on Technological Innovation

	Phase	Matching	Technological Innovation Willingness		Technological Innovation Capability
	Phase	Matching	ATT	t-Value	ATT	t-Value
Large energy enterprises	1	Before	0.0023	1.70*	0.0129	2.54**
	1	After	0.0012	0.53	0.0127	1.50
	2	Before	0.0048	3.57***	0.0176	2.34**
	2	After	0.0049	2.01**	0.0166	1.39
	3	Before	0.0051	2.87***	0.0319	5.16***
	3	After	0.0060	2.31**	0.0319	2.64***
Medium-sized energy enterprises	1	Before	-0.0003	-0.20	0.0096	1.29
	1	After	-0.0011	-1.29	0.0102	0.64
	2	Before	-0.0001	-0.25	0.0227	2.88
	2	After	-0.0011	-1.15	0.0256	1.25
	3	Before	-0.0011	-0.33	0.0296	4.86
	3	After	-0.0001	-0.34	0.0298	1.39
Small energy enterprises	1	Before	-0.0001	-0.18	0.0191	2.08
	1	After	0.0001	0.67	0.0180	0.82
	2	Before	0.0000	-0.06	0.0303	4.07***
	2	After	0.0001	0.43	0.0317	1.36
	3	Before	-0.0002	-0.21	0.0666	9.78***
	3	After	0.0001	0.71	0.0691	2.06**

Source: The authors (calculation by Stata).

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

From the perspective of the PSM results for technology innovation capability, the export behaviour of large and small energy enterprises will start to show a significant promotion role in the next three phases. This is because large energy enterprises have a strong sense of technological innovation, it is more likely to enter advanced fields, participate in international competition, contact more new knowledge and technology and have stronger talent reserve and absorption capacity. The new knowledge introduced into the international market is easier to be applied and transformed. After a certain period of production transformation, with the improvement of their own infrastructure, it is easier for the large enterprises to promote the formation of new products and improve their technological innovation capability. However, the export behaviour of small and medium-sized enterprises has no significant effect on technological innovation ability. This is because the market competitiveness of small and medium-sized enterprises is weaker, it is difficult for them to participate in the international market with higher technical barriers, and the degree of knowledge spill over is lower, which is not conducive to the acceptance of new knowledge and technology. In addition, enterprises have weak absorptive capacity due to old equipment and lack of R&D personnel, so it is difficult to directly improve their technological innovation capability through export behaviour.

Conclusions and Policy Implications

This article uses the Chinese Industrial Enterprises Database and PSM to deal with potential self-selection and missing variable biases to improve the current lack of research on ‘export learning effects’ in academia, and to reveal the causal effect of export behaviour on the technological innovation willingness and capability of energy enterprises. The results are as follows: (a) The export behaviour of energy enterprises plays a significant role in promoting technological innovation willingness and capability. The promotion effect of export behaviour on the technological innovation willingness of energy enterprises is immediate, there is a certain sustainability and the promotion effect is gradually enhanced. The promotion effect of export behaviour on the technological innovation capability of energy enterprises has a time lag but also has a certain sustainability, and the promotion effect is gradually enhanced. (b) There is no significant difference between export behaviour and the technological innovation capability of energy enterprises with different ownership structures; export behaviour has a more positive impact on large energy enterprises and has no significant influence on small and medium-sized energy enterprises to promote the technological innovation capability of large and small energy enterprises in the next three phases.

Based on the above analysis, this article puts forward the following management implications:

The government should establish and improve the export incentive policy for energy enterprises. Energy enterprises will increase their technological innovation willingness in the process of exporting, but the proportion of energy enterprises that actually export is relatively small, and thus, the government should encourage more energy enterprises to ‘go global’ to accept international competition. The government should adjust its policy incentives and guidance on innovation direction according to the type of enterprise; through policy guidance or advocacy, the transformation of the commercial value and added value of energy technological innovation will be promoted, and the exporting of energy products with high energy efficiency and low energy consumption will be increased. We should eliminate the concerns that technological innovation is imitated in the process of export, improve the protection mechanism of energy intellectual property rights, actively participate in the formulation of international trade rules for energy and create a good trade environment for local energy enterprises.

The export service management mechanism of energy enterprises is actively implemented in the industry. Relevant departments should actively carry out international energy trade and cooperation exchange and negotiation meetings, make full use of AI technology to build a cutting-edge knowledge exchange platform between domestic energy enterprises and international advanced enterprises and establish a big data knowledge base in the innovation field of international energy enterprises to facilitate China’s energy enterprises to learn advanced technologies and new knowledge as well as to narrow the technological gap. The relevant functional departments should vigorously promote the process of the trade facilitation of energy enterprises and establish special departments or associations for the export supervision of energy enterprises to solve the demand of large, medium-sized and small energy enterprises for talent, technology, capital and other innovative resources to different degrees and resolve the mismatch and distortion of resources caused by institutional factors. Doing so would allow export behaviour to become the core power for enhancing the technological innovation willingness and capability of energy enterprises.

Energy enterprises should formulate effective export strategies. Different types of energy enterprises should formulate export trade strategies that are suitable for their own development at different stages, especially large and state-owned energy enterprises. These enterprises should play a pioneering role in the industry and promote the technological progress of the whole energy industry through the diffusion of knowledge within the industry; establish and improve the risk resistance mechanism of the international market, introduce advanced enterprise management modes, actively accept the challenge of the international market and effectively form the transformation of new knowledge and new products; and improve service systems, including R&D design, information systems, strategic consulting, maintenance services, etc., promote the development of new products and services with innovative ideas, improve the international competitiveness of China’s energy products, enhance the added value of new export products, and realize the transformation and upgrading of China’s energy products in foreign trade.

Limitations and Future Research Directions

Our research expands the existing literature and enhances our understanding of the causal relationship between export behaviour and technological innovation in energy enterprises. However, we would like to highlight some limitations. First, the export behaviour in this article refers to the general export volume. However, on the premise of available data, future research can divide the export behaviour into multiple dimensions for further discussions, for example, the export behaviour can be classified into some dimensions such as OEM, ODM, OBM, etc. Second, research shows that export behaviour plays an important role in promoting the willingness and ability of energy technology innovation. Whether it will restrict the promotion and cultivation of the independent innovation ability of Chinese local enterprises, in terms of enterprise strategy, it will often be suppressed by the externality of the global market, leading to technology monopoly. At this time, how Chinese energy enterprises overcome external pressure and improve their independent innovation ability will be a possibility for future research.

Footnotes

Acknowledgements

This research has been partially supported by the National Natural Science Foundation of China (72074059) and the Social Science Foundation of Heilongjiang (20GLB120).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Ahmad

, Khattak

S. I.

, Khan

, & Anwar

(2020). Innovation, foreign direct investment (FDI), and the energy–pollution–growth nexus in OECD region: A simultaneous equation modeling approach. Environmental and Ecological Statistics, 27(1), 203–232.

Benhayoun

, Dain

M. A. L.

Dominguez-Pery

, & Andrew

(2020). SMEs embedded in collaborative innovation networks: How to measure their absorptive capacity? Technological Forecasting & Social Change, 15, 1–25.

Bointner

(2014). Innovation in the energy sector: Lessons learnt from R&D expenditures and patents in selected IEA countries. Energy Policy, 73, 733–747.

Boso

, Adeola

, Danso

, & Assadinia

(2019). The effect of export marketing capabilities on export performance: Moderating role of dysfunctional competition. Industrial Marketing Management, 78, 137–145.

Brandt

, Biesebroeck

J. V.

, & Zhang

(2012). Creative accounting or creative destruction? Firm-level productivity growth in Chinese manufacturing. Journal of Development Economics, 97(2), 339–351.

Brutschin

, & Fleig

(2016). Innovation in the energy sector: The role of fossil fuels and developing economies. Energy Policy, 97, 27–38.

Cintio

M.D.

, Ghosh

, & Grassi

(2020). Direct or indirect exports: What matters for firms’ innovation activities? Applied Economics Letters, 27(2), 92–103.

Clerides

, Lach

, & Tybout

(1996). Is learning-by-exporting important? Micro-dynamic evidence from Colombia, Mexico and Morocco. Finance and Economics Discussion Series, 113(3), 903–948.

Dai

, Liu

, Lin

, & Greenaway

(2020). How innovation impacts firms’ export survival: Does export mode matter? The World Economy, 43(1), 81–113.

10.

Dai

, & Yu

(2012). R&D input, export and productivity improvement before export: Evidence from Chinese manufacturing enterprises. China Economic Quarterly, 11(1), 211–230.

11.

Dai

, & Yu

(2013). Firm R&D, absorptive capacity and learning by exporting: Firm-level evidence from China. The World Economy, 36(9), 1131–1145.

12.

Danish, & Recep, E. (2022). Analyzing energy innovation-emissions nexus in China: A novel dynamic simulation method. Energy. 10.1016/j.energy.2021.123010

13.

Efren

, Pia

, Nuria

E. H.

, & Blanca

L. D.

(2022). Internationalization and environmental innovation in the energy sector: Exploring the differences between multinational enterprises from emerging and developed countries. Energy Policy. 10.1016/j.enpol.2022.112867

14.

Falk

, & Figueira

(2019). Complementarity of R&D and productivity in SME export behavior. Journal of Business Research, 96, 157–168.

15.

Gerschewski

, Scott Ennel

, & Rose

E. L.

(2019). Ready to export? The role of export readiness for superior export performance of small and medium-sized enterprises. World Economy, 43(2), 1253–1276.

16.

Gkypali

, Arvanitis

, & Tsekouras

(2018). Absorptive capacity, exporting activities, innovation openness and innovation performance: A SEM approach towards a unifying framework. Technological Forecasting & Social Change, 132, 143–155.

17.

Gkypali

, Love

J. H.

, & Roper

(2021). Export status and SME productivity: Learning- to-export versus learning-by-exporting. Journal of Business Research, 128, 486–498.

18.

Grossman

G. M.

, & Helpman

(1990). Trade, knowledge spillovers, and growth. Social Science Electronic Publishing, 35(2–3), 517–526.

19.

Grossman

G. M.

, & Helpman

(1990). Innovation and growth in the global economy. Mit Press Books, 1(2), 323–324.

20.

Harris

, Krenz

, & Moffat

(2021). The effects of absorptive capacity on innovation performance: A cross-country perspective. Journal of Common Market Studies, 59(3), 589–607.

21.

L.Y.

, & Huang

(2020). Processing trade and energy efficiency: Evidence from Chinese manufacturing firms. Journal of Cleaner Production, 276, 1–11.

22.

L.Y.

, & Huang

(2021). How can export improve firms’ energy efficiency? The role of innovation investment. Structural Change and Economic Dynamics, 59, 90–97.

23.

Hofman

P. S.

, Blome

, Schleper

M. C.

, & Subramanian

(2020). Supply chain collaboration and eco-innovations: An institutional perspective from China. Business Strategy and the Environment, 29, 2734–2754.

24.

Huang

X. H.

, Jin

Z. C.

, & Yu

L. H.

(2018). Exports, innovation and markup ratio: An analysis from the perspective of factor intensity. World Economy, 41(5), 127–148.

25.

Jiang

, & Liu

(2021). Research on the impact of venture capital and information diaphaneity on enterprise innovation willingness. Chinese Journal of Management, 18(8), 1187–1194.

26.

Kang

Z. Y.

(2018). Influence of government science and technology innovation Fund on enterprises’ product innovation. R&D Management, 30(2), 103–113.

27.

Lenihan

, Helen

, & Kenvin

(2019). Driving innovation: Public policy and human capital. Research Policy, 48(9), 1–19.

28.

Love

J. H.

, Roper

(2016). SME innovation, exporting and growth: A review of existing evidence. International Small Business Journal, 33(1), 28–48.

29.

Melitz

M. J.

(2003). The impact of trade on intra-industry reallocations and aggregate industry productivity. Econometrica, 71(6), 1695–1725.

30.

Montalbano

, & Nenci

(2018). Energy efficiency, productivity and exporting: Firm-level evidence in Latin America. Energy Economics, SI, 97–110.

31.

Muoz

, Galvez

, Enjolras

, Camargo

, & Alfaro

(2022). Relationship between innovation and exports in enterprises: A support tool for synergistic improvement plans. Technological Forecasting and Social Change, 117, 1–12.

32.

Parrilli

M. D.

, Balavac

, & Radicic

(2020). Business innovation modes and their impact on innovation outputs: Regional variations and the nature of innovation across EU regions. Research Policy, 49, 1–25.

33.

Peluffo

(2016). The role of investments in export growth. Small Business Economics, 47, 115–137.

34.

Rodil

, Vence

, & Snchez

M. D. C.

(2016). The relationship between innovation and export behaviour: The case of Galician firms. Technological Forecasting & Social Change, 113(96), 248–265.

35.

Salomon

, & Shaver

J. M.

(2005). Learning by exporting: New insights from examining firm innovation. Journal of Economics & Management Strategy, 14(2), 431–460.

36.

Salomon

, & Jin

(2010). Do leading or lagging firms learn more from exporting? Strategic Management Journal, 31(10), 1088–1113.

37.

Shapiro

J. S.

, & Walker

(2018). Why is pollution from us manufacturing declining? The roles of environmental regulation, productivity, and trade. American Economic Review, 108(12), 3814–3854.

38.

Shefer

, & Frenkel

(2005). R&D, Firm size and innovation: An empirical analysis. Technovation, 25(1), 25–32.

39.

, & Li

(2021). Interaction effects of government subsidies, R&D input and innovation performance of Chinese energy industry: A panel vector autoregressive (PVAR) analysis. Technology and Analysis & Strategic Management. 10.1080/09537325.2021.1979205.

40.

Wang

, & Luo

(2020). Has technological innovation capability addressed environmental pollution from the dual perspective of FDI quantity and quality? Evidence from China-science direct. Journal of Cleaner Production, 258, 1–13.

41.

, Wei

, & Wang

(2020). Disentangling the effects of business groups in the innovation-export relationship. Research Policy. 10.1016/j.respol.2020.104093.

42.

Xie

J. G.

, & Ding

(2018). Export and industrial R&D an empirical study based on micro data of China. Review of Industrial Economics, 26(3), 112–126.

43.

, Wang

X. H.

, & Liu

(2021). Government subsidies, R&D investment and innovation performance: Analysis from pharmaceutical sector in China. Technology and Analysis & Strategic Management, 33, 535–553.

44.

Yang

, Zeng

, & Zhou

Z. D.

(2021). Growth or innovation? The impacts analysis of FDI on performance of China’s high-tech zones. Journal of the Asia Pacific Economy. 10.1080/13547860.2021.1961988.

45.

Yang

, Elsner

, Zhang

, & Berger

(2020). Collaborative innovation and policy support: The emergence of trilateral networks. Applied Economics, 52, 3651–3668.

46.

H. X.

, Gong

L. T.

, & Chen

Y. Y.

(2011). Financial constraints and export behaviour. Economic Research, 4, 55–67.

47.

Zhong

, Huang

, & He

(2022). The technological innovation efficiency of China’s lithium-ion battery listed enterprises: Evidence from a three-stage DEA model and micro-data. Energy. 10.1016/j.energy.2022.123331

48.

Zhou

, Liu

, Li

, & Zhang

(2020). Foreign equity, exporting and firm innovation: An emerging market perspective. The Journal of Technology Transfer, 7. 10.1007/s10961-020-09813-6.

49.

Zhu

, & Zhao

Y. L.

(2017). Resource misallocation and firms’ exporting behaviour: An empirical study based on the Chinese industrial firms data. Financial Research, 11, 49–64.