Drivers of the development of global climate-change- mitigation technology: a patent-based decomposition analysis

Abstract

The development of the climate-change- mitigation technology has received widespread attention from both academic and policy studies. Nevertheless, very few studies have explained how and why economies contribute differently to global development. This paper decomposed the development of the global climate-change-mitigation technology, proxied by patent-based indicators, from 1996 to 2015 into several predefined factors. The results show that the worldwide surge of climate-change-mitigation-technology patents from 1996 to 2011 is driven by increased concentration on green invention, improved research intensity, and enlarged economic scale, while the falling of patent counts from 2011 to 2015 is predominantly due to less concentration on green invention. Among different climate-change-mitigation technologies, the type-specific development is attributed to different dominant factors, and the resulting priority change can reflect the shift of both global research and development (R&D) resource and market demand. Regarding regional contributions, the resulting economy-specific contributions to each driving factor can be used to design the policies to promote the development of the global climate-change-mitigation technology.

Graphical abstract

Keywords

climate change mitigation / technology development / logarithmic mean Divisia index / green patents

1 Introduction

After repetitive arguments and discussions about the potential threat of global climate change, at the 2015 Paris Climate Conference, the human community finally reached a consensus to act in concert over the matter of climate change mitigation [1–4]. Among all the potential solutions such as stalling growth and shutting down pollution firms, promoting the innovation and diffusion of the climate change mitigation technology (henceforth, CCMT) is the most desirable one concerning the balance of economic and climate goals [5–7]. According to the International Energy Agency calculation, the efficiency technology and the renewable energy technology contribute 40% and 35% respectively to the reduction of energy-related CO₂ emissions from 2014 to 2060, given the 2°C scenario [8]. Moreover, the stringent climate goals do not only entail accelerating growth of CCMT in developed economies but also require substantial innovative activities in underdeveloped economies mimicking the advanced technology, as well as absorbing and localizing the foreign technology [9,10]. Therefore, to inquire whether economies across the globe could combat climate change, it is necessary to observe the development of CCMT from a global perspective.

With the ongoing climate change issues, there are emerging studies focusing on the development of CCMT. For instance, Fujii and Managi [11] collected the patent information of China’s sustainable green technology and investigated the different development trends over three different periods. Dechezleprêtre and Martin selected 19 types of low-carbon technologies and counted the patents from UK inventors [12]. Considering that climate change is a global issue, a large body of this literature focused on the global development pattern of CCMT [13–16]. For instance, Albino et al. [13] selected the low-carbon energy technology patents granted at U.S.PTO. from 1971 to 2010, and subdivided the patents into subgroups of technology. Their research adds to the literature by providing detailed country-and-year-specific contributions to each type of low-carbon energy technology. Rather than patent data, Gallagher et al. [15] used the data of governmental expenditure on energy research, development, and demonstration. The results show that the nuclear energy technology is the largest portion of the overall expenditures of IEA member countries, while the fossil fuel technology still accounts for an important part. Dechezleprêtre et al. [14] were among the first to focus on climate change mitigation. They identified 13 types of CCMT patent and found a noticeable fluctuation throughout 1978–2005. Given the previous studies showing different technology development trends among economies, Yan et al. [16] focused on the convergent trend of global technology. Using a panel of data, including 72 economies over the period of 1990–2012, they found no convergent trends of low-carbon technology for all economies. However, there were several convergent clubs where economies within each club converged to a common trend. In addition to the development trend of CCMT, many existing studies explored the impact factors of specific types of CCMT, e.g., clean technology in automobiles, and these studies are typically at the firm level [17] or sector level [18]. In contrast, the studies from the global perspective are relatively few [19]. For instance, Verdolini and Galeotti [20] found empirical evidence that changes in green technologies were driven by energy price, technological opportunity, and technology spillover between countries. Dechezleprêtre et al. [14] found the changes in the global CCMT from 1978 to 2005 were related to worldwide policy, such as the Tokyo Protocol and country-specific policies.

Existing studies have paved the way for the research conducted in the present paper. Concerning the urgency of collective action for climate change mitigation, the cross-regional studies typically found different technology development trends among economies [13,14,16,21]. Moreover, the underlying reasons for such heterogeneity may be the different dominant factors of CCMT across economies. Thus, to understand the global development of CCMT, it is necessary to calculate the contributions of each economy to the overall development and understand the roles of specific regions or economies in each type of driving factors. Thus, this paper aims to investigate the driving factors of CCMT, taking technology difference and economy heterogeneity into consideration.

To the best of the authors’ knowledge, this paper is among the first to investigate the driving factors of fully defined CCMT from a global perspective. In contrast, previous studies typically view the development of CCMT from specific types of technologies or given regions [11,13]. Besides, this paper focuses on technology differences, uncovering different driving factors of specific technologies, and calculating the priority change of research activities on CCMT. Given the existing studies discovering technological development trends [13], this paper provides more information on the underlying driving forces. Furthermore, this paper considers heterogeneity among regions, calculating different contributions of specific economies to each driving factor of CCMT.

2 Methods and data

The methodology adopted in this paper consists of patent statistics which indicates the global development of CCMT and the LMDI method which decomposes CCMT development change into different driving factors.

2.1 Patent statistics as indicators of technology development

This paper uses the patent counts to proxy for technology development. Since the broad concept of technological development involves various processes, such as R&D input and output, diffusion, adoption, and application [22], this paper assumes that patent-based indicators can reflect the development trend in a technological field⁾. The main reason for this is that, as the output of knowledge production, the number of patent filings in a given year can reflect the active degree of the research in a given field. The granted patents can represent intellectual property rights. Thus, the patented invention typically has a technological extension or advantage over the past technology, and it may potentially create commercial value in addition to the existing technology. For an economy, if more patents are filed in a technology field, the more massive technology progress is making by inventors in this economy [23,24]. Therefore, the patent count is a widely accepted indicator for technology development [25–29].

Compared with other frequently used indicators, the patent-based indicator has the following advantages⁾. First, patent data are available worldwide. The applicants file the patent in different patent offices, of which the bibliographic data used to illustrate the intellectual rights are typically available to interested users. In contrast, the R&D expenditures are only available for limited numbers of economies, e.g., the Energy Technology R&D Budget Database of IEA. Second, the patent counts are comparable across economies. The statistic scope of the patent-based indicator is generally unique, given the unified worldwide system of patent classification. In contrast, the R&D input is typically difficult to compare among economies since the concept of R&D input varies dramatically with different accounting principles worldwide. Third, the patent-based indicator has a significant advantage in the subdivision. According to the patent classification system, it is methodologically easy to count the patents of a technology class, its subclass, and its further detailed technology groups.

The statistical strategy of patent counts consists of four steps in sequence. First, this paper selects the patents of invention belonging to the classification of CCMT from the world patent database. The utility model and design patent are excluded since this paper focuses on technological progress. The CCMT patent is defined according to the Cooperative Patent Classification (CPC) system, as shown in Table 1.

Second, this paper defines the origin of a patent by referring to the country or region codes of its inventor. For instance, if a patent is filed in the USPTO but residential country (or region) of its inventor is the UK, it is defined as a patent of the UK concerning that it is the people from the UK who has created the patent-related knowledge. It should be noted that, in the worldwide patent database, there is a large scale of missing information and inaccurate information on the residential country (or region) of the inventor. To improve the accuracy and completeness, the allocation procedure proposed by Pasimeni is employed in this paper [30], which uses the standardization information on patent ID and person name (i.e., DOCDB database), provided by the European Patent Office (EPO), as the correction of missing or inaccurate data on the residential country (or region) of the inventor. After the above correction, the still missing data on the country/region code of the inventor are replaced by the country (or region) code of the applicant. The reason for this is that when the inventors were applying for the patents in their own countries, they probably would think it is unnecessary to fill in the information on the country (or region) of residence [14].

Third, this paper calculates the fractional count of CCMT patent as the CCMT development indicator in an economy. That is, to avoid double-counting, one patent was only counted once. For a patent with inventors from more than one economy, this paper calculates the regional ratio as the count of each economy, i.e., the numbers of inventors from an economy to the overall inventor numbers of the patent.

2.2 LMDI analysis

2.2.1 Decomposition of overall CCMT patent counts

Based on the patent-based indicators, this paper uses the widely used LMDI method to analyze the change of CCMT [31–34]. As noted by a seminal work of Ang, the LMDI method has several merits, i.e., factor reversal, time reversal, and zero-value robust, making it a practical tool for users from academia, public administration, and other fields. Nevertheless, this method has its limitations [35]. By assumption, changes in the variable of interest are attributed to several pre-defined driving factors, potentially omitting some influential factors or including some unnecessary factors. Thus, the basic settings of the LMDI model should be well supported by the literature.

Referring to Fujii and Managi [11], the global CCMT patent counts (contributed by N economies) at time t (P_G,t) can be expressed as

(1)\begin{array}{l}{P}_{\mathrm{G},t}={\displaystyle \sum _i^N{\displaystyle \frac{P_{\mathrm{G},it}}{T_{it}}}{\displaystyle \frac{T_{it}}{R_{it}}}{\displaystyle \frac{R_{it}}{Y_{it}}}{\displaystyle \frac{Y_{it}}{Y_t}}{Y}_t}\\ ={\displaystyle \sum _i^t{\mathrm{GREEN}}_{it}\times {\mathrm{EFFI}}_{it}\times {\mathrm{RND}}_{it}\times {\mathrm{REGI}}_{it}\times {\mathrm{SCAL}}_t},\end{array}

where GREEN_it is the share of CCMT patent (P_G,it) in total patent counts (T_it) in economy i of year t, indicating the focus of the research activity of an economy on CCMT; EFFI_it is the ratio of the R&D output (T_it, i.e., total patent counts) to the R&D input (T_it, i.e., R&D expenditures), indicating the efficiency of R&D activity; RND_it is the share of R&D expenditure (T_it) in the GDP (Y_it) of an economy, indicating the R&D intensity of an economy; REGI_it is the share of the total GDP of economy i, indicating the regional economic development structure [36]; and SCAL_it is the total GDP of sample economies (Y_t), indicating the overall economic scale.

According to Ang [35], the change of global CCMT patent between time t and τ can be decomposed into several factors as

(2)$$\begin{gathered} D_{\mathrm{t}\mathrm{o}\mathrm{t}}^n=P_{\mathrm{G},t}^n-P_{\mathrm{G},r}^n={\displaystyle \sum _i(P_{\mathrm{G},t}^n-P_{\mathrm{G},\tau }^n)}\ln ({\mathrm{G}\mathrm{R}\mathrm{E}\mathrm{E}\mathrm{N}}_{i,t}^n/{\mathrm{G}\mathrm{R}\mathrm{E}\mathrm{E}\mathrm{N}}_{i,\tau }^n)+{\displaystyle \sum _i(P_{\mathrm{G},t}^n-P_{\mathrm{G},\tau }^n)}\ln ({\mathrm{E}\mathrm{E}\mathrm{F}\mathrm{I}}_{i,t}^n/{\mathrm{E}\mathrm{E}\mathrm{F}\mathrm{I}}_{i,\tau }^n) \\ +{\displaystyle \sum _i(P_{\mathrm{G},t}^n-P_{\mathrm{G},\tau }^n)}\ln ({\mathrm{R}\mathrm{N}\mathrm{D}}_{i,t}^n/{\mathrm{R}\mathrm{N}\mathrm{D}}_{i,\tau }^n)+{\displaystyle \sum _i(P_{\mathrm{G},t}^n-P_{\mathrm{G},\tau }^n)}\ln ({\mathrm{R}\mathrm{E}\mathrm{G}\mathrm{I}}_{i,t}^n/{\mathrm{R}\mathrm{E}\mathrm{G}\mathrm{I}}_{i,\tau }^n) \\ +{\displaystyle \sum _i(P_{\mathrm{G},t}^n-P_{\mathrm{G},\tau }^n)}\ln ({\mathrm{S}\mathrm{C}\mathrm{A}\mathrm{L}}_{i,t}^n/{\mathrm{S}\mathrm{C}\mathrm{A}\mathrm{L}}_{i,\tau }^n) \\ =D_{\mathrm{G}\mathrm{R}\mathrm{E}\mathrm{E}\mathrm{N}}^n+D_{\mathrm{E}\mathrm{E}\mathrm{F}\mathrm{I}}^n+D_{\mathrm{R}\mathrm{N}\mathrm{D}}^n+D_{\mathrm{R}\mathrm{E}\mathrm{G}\mathrm{I}}^n+D_{\mathrm{S}\mathrm{C}\mathrm{A}\mathrm{L}}^n, \end{gathered}$$

where L(P_{\mathrm{G},t}^n-P_{\mathrm{G},r}^n) is the logarithmic mean weight that is calculated as

(3)L(P_{\mathrm{G},t}^n-P_{\mathrm{G},r}^n)=(P_{\mathrm{G},t}^n-P_{\mathrm{G},r}^n)/(\ln P_{\mathrm{G},t}^n-\ln P_{\mathrm{G},r}^n)\mathrm{.}

According to Eq. (2), the change of overall CCMT patent counts is caused by the changes in green focusing (GREEN), research efficiency (EFFI), R&D intensity (RND), regional GDP structure (REGI), and overall economic scale (SCAL).

2.2.2 Decomposition of patent counts of specific CCMT

For a specific type of CCMT, referring to Fujii and Managi [11], this paper assumes the change of technology as

(4)\begin{array}{l}{P}_{\mathrm{E},t}={\displaystyle \sum _i^N{\displaystyle \frac{P_{\mathrm{E},it}}{P_{\mathrm{G},it}}}{\displaystyle \frac{P_{\mathrm{G},it}}{T_{it}}}{\displaystyle \frac{T_{it}}{R_{it}}}{\displaystyle \frac{R_{it}}{Y_{it}}}{\displaystyle \frac{Y_{it}}{Y_t}}{Y}_t}\\ ={\displaystyle \sum _i^N{\mathrm{PRIO}}_{it}\times {\mathrm{GREEN}}_{it}\times {\mathrm{EFFI}}_{it}\times {\mathrm{RND}}_{it}\times {\mathrm{REGI}}_{it}\times {\mathrm{SCAL}}_t},\end{array}

where PRIO is the patent counts of a specific type of CCMT (\left(P_{\mathrm{E},it}\right)) divided by the overall patent counts of CCMT (\left(P_{\mathrm{G},it}\right)), indicating the priority of specific technology. The increase of the PRIO factor means that the specific type of technology grows faster than the overall CCMT. The other variables are the same as those in Eq. (1). With the same procedure as that in Eqs. (2) and (3), this paper decomposes the change of patent counts of a specific type of CCMT into factors like changes in priority (PRIO), green focusing (GREEN), research efficiency (EFFI), R&D intensity (RND), regional GDP structure (REGI), and overall economic scale (SCAL).

2.3 Data

The patent bibliographic data used to calculate technology development indicators are obtained from the EPO worldwide patent database (PATSTAT 2018Autumn Edition). PATSTAT contains the patent information from more than 80 patent offices worldwide, amounting to over 80 million patent applications since the 19th century. With the PATSTAT database, this paper employs the patent statistics of CCMT and obtains an unbalanced panel of data covering more than 100 economies. However, the LMDI framework requires balanced data. This paper obtains the R&D expenditure data from the World Bank database and the real GDP data from the Penn World Table (Version 9.1). Regarding the data availability, there are frequently missing data of R&D expenditure before 1996. Therefore, this paper chooses the top 20 economies according to the ranking of accumulated CCMT patent counts from 1996 to 2015. These 20 economies as a group account for about 92% of the global CCMT patents over the period. Thus, it is reasonable to conclude that the samples can reflect the global development trend of CCMT. As a result, the sample for LMDI analysis is a balanced panel of data consisting of the top 20 inventor economies from 1996 to 2015.

3 LMDI results of global CCMT development

3.1 Global development trends of CCMT

Following the statistical strategy in Section 2, this paper depicts the global development of CCMT in Fig. 1. To be specific, Fig. 1(a) and Fig. 1(b) shows the year-by-year change of patent counts, decomposed into technological fields and regions, respectively.

3.1.1 Overall development trends

In Fig. 1(a), the number of CCMT patents continuously increased from 1996, peaked in 2012, and gradually decreased afterward. Compared with the figure of 1995, the total number of CCMT patents in 2015 increased by 277%, reflecting a high growth rate on average.

3.1.2 Technology-specific development trends

According to Fig. 1(a), the energy-related technology, the production-related technology, and the transportation-related technology contribute 31.5%, 25.3%, and 24.5% of the overall CCMT patents in 2015. In contrast, the buildings-related technology, the waste-related technology, and the GHG-related technology account for 10%, 7.4%, and 1.3%, respectively. During the research period, the ratio of the GHG-related technology (1.4% on average), the buildings-related technology (9.8% on average), and the transportation-related technology (23.2% on average) hold steady at given levels. In contrast, the ratio of the energy-related technology increased from 22.5% to 31.5%. In comparison, the ratio of the production-related technology decreased from 31.2% to 25.3%, and the ratio of the waste-related technology decreased from 15% to 7.4%. In general, such dynamics reflect the shift of the global R&D activities in CCMT from the production-related technology and the waste-related technology to energy-related technology.

3.1.3 Regional development trends

Regarding the regional contribution of CCMT patent counts shown in Fig. 1(b), the summed ratio of the sample economies remains almost unchanged at 92% from 1996 to 2015. This ratio indicates that innovative activities are quite concentrated in these selected economies. After classifying the sample economies geographically, it is shown that East Asia is the largest origin of CCMT patents over the research period, accounting for 54% of the total in 2015. In sequence, the contributions of North America, Europe, and Oceania in 2015 are 19.0%, 19.0%, and 0.5%, respectively. During the research period, the contribution of East Asia increased by near 10%, that of North America remains stable, but that of Europe and Oceania contributions decreased by 8.3% and 0.8%, respectively.

3.2 Drivers of global CCMT development

Using the LMDI method, this paper analyses the driving factors of the change in global CCMT. In Fig. 2, the global CCMT patent changes are decomposed into five factors, i.e., green focusing, research efficiency, R&D intensity, regional structure, and economic scale. To observe the dynamics of the driving factors in more detail, the research period is divided into four stages. From 1996 to 2001, the patent counts of CCMT increased by 64.6% in the first stage. However, the growth rate fell to 37.7% in the second stage. After 2006, the global patent counts dramatically increased by 90.9% in the third stage, but it decreased by 11.9% in the last stage.

Among all drivers, the green focusing is the dominant factor for the change in the global patent counts. The contribution of green focusing, inclining from 20.4% every five years in the first stage to 73.3% in the third stage, resulted in the booming of CCMT patent during 2006–2011. However, this driver declined dramatically to –32% in the fourth stage, leading to a decrease in overall CCMT patent counts. From 1996 to 2011, the global patent activity is increasingly focusing on the climate-friendly technology, but such a tendency ceased and reversed after the global financial crisis.

The R&D intensity keeps contributing to the growth of CCMT patent counts for all four stages, although its contribution fluctuates. Its maximum driving contribution, i.e., 14.4%, emerged in the third stage, while the minimum contribution was in the last stage, with a value of 5%. Thus, intensive R&D activities can bring about innovative outputs, which would, with a certain probability, induce CCMT patent growth.

The scale factor also continuously drives the growth of CCMT patent for all stages. In the first and fourth stages, its contributions are 20.6% and 8.8%, respectively, ranking first among all factors. In the second stage and the third stage, it is the second-largest driving factor. Thus, the growing income worldwide is an important driver that stimulates economies to innovate in CCMT.

The contributions of the efficiency factor fluctuate. It is positively related to the patent growth of CCMT in the first stage and the last stage, while it is a negative driving factor in the second stage and the third stage. Referring to the statistical data, the average growth rate of R&D efficiency is –2% by year in the second stage, compared with a growth rate of 4% in the first stage. Thus, the falling efficiency of research activity can potentially impede the growth of CCMT patents. Besides, the regional output structure is a limited influencing factor, given that the absolute value of its contribution ranks at the bottom of each stage. Moreover, its negative contributions in most stages indicate that the uneven economic development across economies might hinder the development of CCMT.

3.3 Drivers of specific types of CCMT development

Considering the existence of different development trends among various technologies, we investigate drivers of a specific technology. Figure 3 depicts the LMDI results employing the strategy of Eq. (2). The bar charts with different colors show the accumulated changes, from 1996 to the corresponding year, of CCMT patent due to the change of different factors. The line chart shows the overall change of CCMT patent considering all influential factors. In addition to the analysis of overall CCMT, the decomposition analysis shown in Fig. 3 contains the priority indicator. It can reflect the relative concentration of global research activity among various technologies.

Referring to the six panels, the changes in patent counts resulted in the priority change of CCMT research (factor priority) are different by year or by type of technology. For instance, during 2000–2010, the change in the priority indicator led to the sustained growth of CCMT patent related to energy. However, it is the largest negative driving factor for the change of CCMT patent related to waste. It reflects the research activity shifted from waste to energy and other CCMT during that period. For all the six types of technologies, the expansion of economic scale (factor scale), the improvement in R&D intensity (factor R&D rate), and the focus of patent activity on climate change mitigation (factor Green) strongly contribute to the growth of CCMT patent during the research period. For instance, these three factors contribute to over 95% of the cumulative growth of CCMT patent related to buildings, GHG, and energy. In contrast, the change of regional economic structure (factor GDP share) negatively relates to the growth of CCMT patent for almost every type of technology. Regarding the efficiency indicator, it contributed to the growth of the CCMT patent in each type before 2005. Nevertheless, after 2005, the improved research efficiency led to an increase of CCMT patent in industrial production and waste. Regarding the different development patterns among technologies from 2011 to 2015, the decreases in priority and green indicators led to the falling of the patent counts of building-related and energy-related technologies. The decrease of green indicator led to the falling of GHG-related patent counts. In contrast, the increased priority indicator helped maintain the patent counts of transportation-related and industrial-process-related technologies. Moreover, the sustainable growth of waste-related patent counts is due to the improved efficiency, more intensive R&D activity, and improved priority.

Since the year-by-year cumulative indicators are not convenient to observe the effects of specific factors in each period, especially for priority indicator, Fig. 4 depicts the change of patent counts due to priority change in each period. To get the percentage figures for each type of CCMT, we divide the accumulated change of patent counts due to priority change in a period by the patent counts in the first year of the period.

In the first period, the change of the priority indicator contributed to the growth of CCMT patents related to GHG, energy, transportation, and production. It is worth mentioning that 56.2% of the changes in patent counts in CCMT related to GHG is driven by priority change over this period. In the second period, the priority indicator contributed to the increases of CCMT patents related to buildings, energy, and transportation. The ratio of contribution was larger for both energy and production compared with that of the first period. However, it did not drive the growth of CCMT patents related to GHG and production as it did in the first period. Thus, the research priority shifted from GHG and production to buildings, energy, and transportation in the second period. In the third period, the priority indicator only contributed to the growth of CCMT patents related to GHG and energy. In the fourth period, the roles of priority indicator for most types of technology reversed, except for CCMT related to buildings. The priority of research shifted from GHG-related and energy-related technologies to transportation-related, production-related, and waste-related technologies.

4 Regional contributions to the rise and fall of global CCMT development

4.1 Booming CCMT patents by drivers and regions

In Fig. 2, the CCMT patent grew steadily from 1996 to 2011, and then the patent counts decreased from 2011 to 2015. The analysis in Section 3.2 decomposed such changes into different factors at the global level. However, there may be heterogeneous development of the technology among economies [13], which entails an investigation into regional contributions to the rise and fall of the global CCMT development. To this end, this paper uses the LMDI framework and observe the extent to which an economy contributes to the specific drivers of technology development. Figure 5 depicts the top 10 contributors of the CCMT patent growth from 1996 to 2011, and their specific contributions to each of the three patent-booming drivers during the period.

Figure 5(a) depicts regional contributions to the overall change of CCMT patent from 1996 to 2011. During this period, Japan is the largest CCMT inventor economy, contributing about 21% to the overall count of the sample economies. Figure 5(b) depicts the contributions of specific economies to the green indicator, which is the most prominent factor during that period. The green-focusing-induced increase of CCMT patent mainly came from Japan, the United States, Republic of Korea, and Germany. Among the sample economies, Japan contributed the most to the green indicator, indicating that it is the leading economy of the world with a focus on green inventions. Figure 5(c) focuses on the R&D rate indicator. The East Asian economies, i.e., China, Republic of Korea, and Japan, ranked at the top of the list regarding the growth of CCMT patent driven by increased R&D intensity. Figure 5(d) focuses on the GDP indicator, representing the increase of CCMT patent due to the economy-specific GDP growth. While most economies contributed moderately to the GDP-growth-induced patent increase (i.e., below 5%), China’s contribution amounted to roughly 34%. From 1996 to 2011, China’s average GDP growth rate is about 9.9%, referring to China Statistical Yearbook, such a scale effect is prominent among all economies worldwide.

From the above analysis, it is evident that several factors drove the booming CCMT patent from 1996 to 2011, and that different economies have heterogeneous contributions to each factor. For instance, China contributed by substantial growth in GDP and by increasingly intensive R&D activity. Thus, it shows to be scale driven. In contrast, Japan contributed by sustained R&D inputs and (more importantly) by increasing the focus of patent activity on green technology, although it did not experience a continued economic expansion over that period.

4.2 Falling of CCMT by drivers and regions

For the global CCMT patent from 2011 to 2015, the only negative driving factor, i.e., green, outweighed the sum of positive driving factors, making the global CCMT count decrease, as shown in Fig. 2. During this period, economic growths are not satisfying for major economies (especially in European countries who were suffering the debt crisis). The factor of economic growth did not appear to be a strong driving force as it was in the previous periods. To further explore the reason for the falling of CCMT patent counts, this paper decomposes the contribution of each economy to the overall change into different factors. Figure 6 demonstrates the LMDI results of the top 10 contributors of the CCMT patent decrease in sequence. The stacked bar charts represent the driver-specific change of CCMT in each economy. The tagged line charts represent the overall change of CCMT in each economy from 2011 to 2015. The GDP indicator is the sum of the factors GDP share and Scale, representing the change of patent counts due to the GDP change of the specific economy.

In Fig. 6, the decrease in the concentration on CCMT happens in all economies. Especially in Panel A, Japan’s decreased green focusing contributed to a 6.41% decrease of global CCMT patent from 2011 to 2015, followed by the US, Republic of Korea, and Germany. Besides, the decreased green focusing of the US and Republic of Korea has led to a 5.66% and 4.53% decrease of global CCMT patent, respectively, which surpassed the corresponding 3.36% and 1.94% growths due to the increasing factors of the two countries. Thus, the leading economies have been less focused on the green technology, which is the main reason for decreasing global CCMT patents over this period. Additionally, the well-performance factors and poor-performance factors are different among economies. There are considerable CCMT patent growths due to regional GDP growths (GDP) in most economies except Denmark and Italy. The improved research efficiency (efficiency) contributed to CCMT patent growth in the US, Canada, and the Netherlands. The decreased R&D intensity (R&D rate) led to considerable reductions in CCMT patent counts of Spain. In contrast, the increased R&D intensity potentially caused a substantial increase in CCMT patent counts in Korea, France, Italy, and the Netherlands. However, its positive effect was offset by corresponding negative effects in each economy. Taking into account of all influencing factors, only China exhibited CCMT patent growth among the sample economies. It contributed about a 10.7% increase to the global CCMT patent, and the negative contributions of other economies offset its positive contribution. Thus, the overall CCMT patent counts of the sample economies fell between 2011 and 2015.

5 Discussion on CCMT-related policy

Given the ‘double externality problem’ [37], the expected development of CCMT typically entails support from policy tools, especially when the CCMT at the early stage of technological development has a significant cost disadvantage with its rivals. With the LMDI results from Section 4, the CCMT-related policy can be discussed in detail

This paper finds different patterns in the development of CCMT among various technological fields. For instance, the decreased priority and the less focusing on the green technology dominated the falling of CCMT patent counts related to energy from 2011 to 2015. However, the increased priority and improved research efficiency dominated the increase of CCMT patent counts related to waste during the same period. Thus, to cease the currently decreasing CCMT patent trend, the policy-making procedure should take into account specific drivers of different technologies. The database and methodological framework adopted in this paper make it possible to further obtain policy implications regarding CCMT development within a country (or region) and subdivide them into more specific technology groups. Additionally, all subclasses of CCMT have in common three positively influencing factors. To stimulate the patent growth of CCMT, it is necessary to stimulate income growth, encourage more intensive R&D activities, and direct the R&D activities to the fields that contribute to climate change mitigation.

Regarding the relative change of specific types of CCMT, the results show that the priority of CCMT-related research varies in different periods. The changes in research priority can reflect the relative changes in the international demand for different technologies [11]. For policymakers intending to promote CCMT development, knowing the changing demand of technology can help them shape the policies. Additionally, profit-driven investors are unwilling to invest in highly risky technologies, making CCMT the field with suboptimal investment [37]. Thus, tracking the priority change of CCMT development will help the government direct technological progress to balance the economy and the environment. On the one hand, policymakers could encourage the development of technologies that are currently the priorities of technology investment, meeting the global technology demand. On the other hand, policymakers should also invest in the technologies that are currently not priorities but can significantly combat climate change in the future, e.g., CCS technology as a particularly important measure of climate change mitigation [8]. In general, methodological framework in this paper combined with the economic or technological analysis of a specific technology field, e.g., oil price distortion in the transport sector [38], will make it possible to propose practical policy suggestions for each economy.

Additionally, the results also highlight the necessity of multilateral policies of green growth. For either of the four research periods, the factor of regional income structure has contributed to the increase of global CCMT patents, which indicates the importance of regional structure optimization. If the share of economies with a high concentration on the green technology, high research efficiency, and high R&D intensity increase, there will be an increase of global CCMT patents as the outputs. Thus, to strengthen the worldwide technological capability of combating climate change, the global economic policy should be multilateral, balancing the income growth across the globe and enabling all economies to contribute their solutions. Moreover, economies should design policies to encourage technological diffusion, spillover, and cooperation. The LMDI results show that the different economies have contributed to the CCMT patent surge in different ways, indicating the economy-specific comparative advantages in green development. Thus, there might be no universal standard for promoting CCMT; economies should design policies based on their circumstances. First, economies could construct their policies based on the economy-specific performance of driving factors. It is the first priority for an economy to improve its poor-performance factors that contribute to a considerable increase of CCMT patents in other economies with similar economic and technological situations. For instance, China’s booming CCMT during the research period is mostly due to its expanding economic volume and sustained R&D input. However, the performances of its research efficiency and green focusing are not as good as those of other leaders in this technology field [39]. It is also necessary for an economy to maintain its well-performance factors. For instance, Japan was one of the leaders in the worldwide CCMT development before 2011 because of its continuing R&D input and focusing on green innovations. Second, multilateral policies should be designed to direct the knowledge and talents flow across the globe efficiently. For instance, through joint research programs, different economies can learn from each other, and the R&D resources can be distributed among economies utilizing their comparative advantages. In general, cooperative and diversified CCMT policies will help the world construct the technological capability of climate change mitigation.

6 Conclusions

With the growing literature focusing on the development of the climate change mitigation technology, this paper focuses on the reasons for CCMT development, taking into account the technology difference and economy heterogeneity. With a patent-based indicator of CCMT development in the top 20 inventing economies from 1996 to 2015, the results show the driving factors of the development of overall CCMT and its specific types. Moreover, this paper exhibits the contributions of specific economies to the rise and falling of global CCMT development.

The global patent counts of CCMT surged from 1996 to 2012 and then gradually decreased. The LMDI results indicate that the surge of CCMT patent is mainly resulted from the concentration of research activity on the green technology, the increased R&D intensity, and the global income growth. However, the factors driving the decrease of CCMT patent after 2011 were predominantly worldwide less focusing on the green technology. Shocked by the global financial crisis started around 2009, both governments and researchers made the R&D investment less risky, shifting from CCMTs which exhibited ‘double externality problem’ to other technologies (Ref. [37]). Given the enormous demands of CCMT innovation with global effort to meet the 1.5°C and 2°C targets, it is necessary to stop the currently ongoing decrease of CCMT patents. That is, governments should direct the research activities to CCMT through policies such as the governmental CCMT R&D expenditures and the supply of new credit to borrowers undertaking CCMT research activities. The LMDI framework of this paper, with a focus on technology difference and economy heterogeneity, makes it possible to formulate an economy-and-technology specific policy on CCMT.

References

[1] Batten J A, Kinateder H, Szilagyi P G, Addressing COP21 using a stock and oil market integration index. Energy Policy, 2018, 116: 127–136

[2] Costa-Campi M T, García-Quevedo J, Martínez-Ros E. What are the determinants of investment in environmental R&D? Energy Policy, 2017b, 104: 455–465

[3] Duan H, Zhang G, Wang S, Robust climate change research: a review on multi-model analysis. Environmental Research Letters, 2019, 14(3): 033001

[4] Geng Y, Fujita T, Chiu A, Responding to the Paris Climate Agreement: global climate change mitigation efforts. Frontiers in Energy, 2018, 12(3): 333–337

[5] Buchholz W, Dippl L, Eichenseer M. Subsidizing renewables as part of taking leadership in international climate policy: the German case. Energy Policy, 2019, 129: 765–773

[6] Costa-Campi M T, del Rio P, Trujillo-Baute E. Trade-offs in energy and environmental policy. Energy Policy, 2017, 104: 415–418

[7] Shao S, Yang L, Gan C, Using an extended LMDI model to explore techno-economic drivers of energy-related industrial CO₂ emission changes: a case study for Shanghai (China). Renewable & Sustainable Energy Reviews, 2016b, 55: 516–536

[8] International Energy Agency (IEA). Energy Technology Perspectives 2017. Paris, France, 2017

[9] Dussaux D, Dechezleprêtre A, Glachant M. Intellectual property rights protection and the international transfer of low-carbon technologies. London: Grantham Research Institute on Climate Change and the Environment Working Paper No. 288, 2018

[10] He J. Situation and measures of China’s CO₂ emission mitigation after the Paris Agreement. Frontiers in Energy, 2018, 12(3): 353–361

[11] Fujii H, Managi S. Decomposition analysis of sustainable green technology inventions in China. Technological Forecasting and Social Change, 2019, 139: 10–16

[12] Dechezleprêtre A, Martin R. Low-carbon innovation in the UK: evidence from patent data. Policy Paper of the Grantham Research Institute on Climate Change and the Environment and the Centre for Climate Change Economics and Policy. London School of Economics and Political Science, 2010

[13] Albino V, Ardito L, Dangelico R M, Understanding the development trends of low-carbon energy technologies: a patent analysis. Applied Energy, 2014, 135: 836–854

[14] Dechezleprêtre A, Glachant M, Haščič I, Invention and transfer of climate change–mitigation technologies: a global analysis. Review of Environmental Economics and Policy, 2011, 5(1): 109–130

[15] Gallagher K S, Holdren J P, Sagar A D. Energy-technology innovation. Annual Review of Environment and Resources, 2006, 31(1): 193–237

[16] Yan Z, Du K, Yang Z, Convergence or divergence? Understanding the global development trend of low-carbon technologies. Energy Policy, 2017, 109: 499–509

[17] Aghion P, Dechezleprêtre A, Hémous D, Carbon taxes, path dependency, and directed technical change: evidence from the auto industry. Journal of Political Economy, 2016, 124(1): 1–51

[18] Ghisetti C, Quatraro F. Beyond inducement in climate change: does environmental performance spur environmental technologies? A regional analysis of cross-sectoral differences. Ecological Economics, 2013, 96: 99–113

[19] Johnstone N, Haščič I, Popp D. Renewable energy policies and technological innovation: evidence based on patent counts. Environmental and Resource Economics, 2010, 45(1): 133–155

[20] Verdolini E, Galeotti M. At home and abroad: an empirical analysis of innovation and diffusion in energy technologies. Journal of Environmental Economics and Management, 2011, 61(2): 119–134

[21] Gallagher K S, Anadon L D, Kempener R, Trends in investments in global energy research, development, and demonstration. Wiley Interdisciplinary Reviews: Climate Change, 2011, 2(3): 373–396

[22] Shao S, Hu Z, Cao J, Environmental regulation and enterprise innovation: a review. Business Strategy and the Environment, 2020, 29(3): 1465–1478

[23] Li K, Lin B. Impact of energy technology patents in China: evidence from a panel cointegration and error correction model. Energy Policy, 2016, 89: 214–223

[24] Sun H, Edziah B K, Sun C, Institutional quality, green innovation and energy efficiency. Energy Policy, 2019, 135: 111002

[25] Griliches Z. Patent statistics as economic indicators: a survey. Journal of Economic Literature, 1990, 28: 1661–1707

[26] Haščič I, Migotto M. Measuring environmental innovation using patent data. Paris: OECD Environment Working Papers, No. 89. OECD Publishing, 2015

[27] Jaffe A B, Palmer K. Environmental regulation and innovation: a panel data study. Review of Economics and Statistics, 1997, 79(4): 610–619

[28] Lanjouw J O, Pakes A, Putnam J. How to count patents and value intellectual property: the uses of patent renewal and application data. Journal of Industrial Economics, 1998, 46(4): 405–432

[29] Popp D. The effect of new technology on energy consumption. Resource and Energy Economics, 2001, 23(3): 215–239

[30] Pasimeni F. SQL query to increase data accuracy and completeness in PATSTAT. World Patent Information, 2019, 57: 1–7

[31] Shao S, Liu J, Geng Y, Uncovering driving factors of carbon emissions from China’s mining sector. Applied Energy, 2016, 166: 220–238

[32] Wang Q, Hang Y, Su B, Contributions to sector-level carbon intensity change: an integrated decomposition analysis. Energy Economics, 2018, 70: 12–25

[33] Zha D, Zhou D, Zhou P. Driving forces of residential CO₂ emissions in urban and rural China: an index decomposition analysis. Energy Policy, 2010, 38(7): 3377–3383

[34] Zhang X, Zhao X, Jiang Z, How to achieve the 2030 CO₂ emission-reduction targets for China’s industrial sector: retrospective decomposition and prospective trajectories? Global Environmental Change, 2017, 44: 83–97

[35] Ang B W. The LMDI approach to decomposition analysis: a practical guide. Energy Policy, 2005, 33(7): 867–871

[36] Zheng J, Mi Z, Coffman D M, Regional development and carbon emissions in China. Energy Economics, 2019, 81: 25–36

[37] Ley M, Stucki T, Woerter M. The impact of energy prices on green innovation. Energy Journal, 2016, 37(1): 41–75

[38] Wang X, Bai M, Xie C. Investigating CO₂ mitigation potentials and the impact of oil price distortion in China’s transport sector. Energy Policy, 2019, 130: 320–327

[39] Yang Z, Shao S, Li C, Alleviating the misallocation of R&D inputs in China’s manufacturing sector: from the perspectives of factor-biased technological innovation and substitution elasticity. Technological Forecasting and Social Change, 2020, 151: 119878

RIGHTS & PERMISSIONS

Higher Education Press