The dynamics of the studies of China’s science, technology and innovation (STI): a bibliometric analysis of an emerging field

Since 1978, alongside China’s rise as a leading country in science, technology and innovation (STI), the studies of the country’s STI have been emerging as a field attracting increasing scholarly attention. Using the bibliometric method and the data from the Web of Science (WoS), this paper seeks to provide a comprehensive picture of the studies of China’s STI. The findings show that scholarly interests in China’s STI started in 1995 and have since developed rapidly; institutions in China, the U.S. and the U.K. are main contributors to the field, contributing 50%, 27.2% and 12% of the scholarship respectively, with Tsinghua University, Zhejiang University and the Chinese Academy of Sciences being three major institutional contributors. Seminal works have been focused on STI issues at the macro or national, meso or industrial and regional, and micro or organizational and firm levels. A possible agenda for further research is to develop new theories based on China’s practice paying specific attention to issues including R&D expenditure, S&T performance evaluation, regional innovation ecosystem, SOEs in innovation and the role of the Chinese Communist Party in innovation.


The Dynamics of the Studies of China's Science, Technology and Innovation (STI): A Bibliometric Analysis of an Emerging Field 1 Introduction
Globalization, together with localization, blurs national boundaries but does not take critical roles away from the nation-state (Dicken 2007). The national interest still is of prime importance within global governance (Sun and Grimes 2016), which is best exemplified by the recent frictions between China and the United States in trade and technology. Over the past decade, China has prioritized transforming from the world's factory to a leading technological and innovation power. Consequently, studying science & technology and innovation (STI) activities in China is central to understanding its international competitiveness in the knowledge-based economy.
In 1978, China embraced a "spring of science" and launched the reform and open door policy. Some forty years later, in addition to displaying impressive achievements such as the increasing investment in research and development (R&D), the emergence of a very large talent pool whose quality also has been improving, and a steady rise in the contributions of its scientists to international publications, China's STI system has produced some major accomplishments in national defense, as well as in certain fields of basic research and technologies.
All these demonstrate that China has the institutional capacity to mobilize talent and the financial and material resources required to achieve high-priority, national-security objectives (Xue 1997;Suttmeier 1981).
However, the general inefficiency of transferring R&D achievements to production, even amid the reform, also makes it clear that structural reform of China's STI system is imperative if the system were to meet the demand for successful innovation in an increasingly market-oriented and knowledge-based economy. Then, the question becomes why China cannot innovate or China has diligently pursued R&D activities but breakthroughs are still rare (Abrami et al. 2014). To answer the question, scholars have investigated China's STI development (e.g. Zhou and Leydesdorff 2006;Mu and Qu 2008;Hu and Mathews 2008;Fan 2014); international organizations such as the OECD, World Bank, and UNESCO also have produced comprehensive reviews (OECD 2008;Zhang et al. 2009;Cao 2015).
Moreover, scholars have reviewed the literature on national innovation and on China as a country. On the former, Fagerberg and Sapprasert (2011) examined the important role of the "national innovation system (NIS)" as a new approach within the innovation studies. Teixeira (2014) explicitly addressed the roots, evolution and influence of the NIS literature using bibliometrics. Expanding the study of the NIS to that of national innovation, and also using bibliometrics, Sun and Grimes (2016) identified the most significant countries and institutions, major journals, seminal contributions and contributors, and clusters in national innovation studies.
On the latter, Liu et al. (2015a) indicated that China's publications in the Social Science Citation Index (SSCI) database have been rising in terms of volume, world share, and global ranking, but China has yet to make contributions to the social science's literature. Liu et al. (2015b) also examined China-related SSCI publications bibliometrically, concluding that the contributors of these publications are largely limited to China's affluent regions and some of China's geographically proximate neighbors. The research interests of such studies have gradually shifted to the country's environmental issues, public health and the economy.
Unfortunately, there appears to be a lack of a systematic review of the studies of China's STI, a field at the intersection of the national innovation studies and China studies and one of rising interests with tremendous policy implications. Indeed, answering such basic questions as what research has been done, who the major contributing countries and institutions and the leading contributors are, in which journals the research has been published, and especially what the research focuses are is useful to understanding this emerging field as well as its direction for future research.
In conducting such a review, we aim to achieve a twofold goal. Theoretically, we seek to fill the gap in the review of the literature on China's STI by integrating the national innovation studies and China studies.
Particularly, sketching the evolving studies of China's STI could help draw a more accurate and comprehensive picture of the literature of the field, and generalize the dynamic structure and pattern of the STI studies in an emerging economy. We also try to deepen our understanding of China's STI development reflected in the STI studies.
The rest of this paper is structured as follows: Section 2 presents methodology and data. Section 3 intends to draw a systematic and dynamic map through laying out major contributing countries/economies and institutions; seminal contributions and their contributors, and major journals; and research agendas, respectively. Section 4 discusses our findings and concludes the paper.

Methodology and data
The paper takes a quantitative approach to analyze the literature of China's STI. In particular, we use the bibliometric method to analyze citations to the publications on the subject under study. Doing so allows us to process a large amount of bibliometric data and to describe the dynamic or evolving structure of a research field.
The limitations of the approach are also obvious as the data retrieved, based on keyword search, may contain a certain amount of noise by either including some unrelated literature or excluding some important publications.
But we are able to compensate for the problem by our more than 20 years of experience as researchers in and contributors to the studies of China's STI. Our interaction with scholars in the field also helps us to identify seminal contributions and leading contributors that the bibliometric analysis might have failed to do.

Data
Clarivate Analytics's Web of Science (WoS) is selected as the data source of this study. We consider the WoS to be a more appropriate database for our study for several reasons. First, as one of the global citation databases and comprehensive platforms, the WoS can track ideas across disciplines and time from its over 159 million records and over 1.7 billion cited references. Second, comparative and longitudinal studies have shown a consistent and reasonably stable quarterly growth for both publications and citations in the WoS. By comparison, Scopus is a new database including citation information of articles published since 1996, and Google Scholar is a free database but the level of accuracy of its citation counts has been seriously doubted (Levine-Clark and Gil, 2008;Jacsó 2010;Harzing and Alakangas, 2016). Third, previous studies have used the WoS to track progress in specific research fields (Zhou and Leydesdorff 2006;Liu et al. 2015aLiu et al. , 2015bSun and Grimes 2016). This paper focuses on publications related to the studies of China's STI. Science Citation Index Expanded (SCIE) and Social Science Citation Index (SSCI) datasets in the WoS were used as our data source. SCIE covers over 9,200 journals across 178 scientific disciplines, having more than 53 million records and 1.18 billion cited references. SSCI covers over 3,400 journals across 58 social sciences disciplines, having more than 9 million records and 122 million cited references.
We extracted data from the WoS through several steps. First, we searched keywords -"China + innovation" or "China + science and technology" or "China + S&T" -as "topics" in the web page of "basic search" in the database of the "Web of Science TM Core Collection." These keywords likely cover most of the literature in this field but would possibly miss some papers related to China's STI that are not captured by these keywords. Second, we selected the "timespan" between 1978 and 2015 and the database settings were "Science Citation Index Expanded (SCI-EXPANDED) -1978-present" and "Social Sciences Citation Index (SSCI) -1978-present" in the "Web of Science Core Collection: Citation Indexes." Doing so excludes records in the Conference Proceedings Citation Index -Science (CPCI-S) and Conference Proceedings Citation Index -Social Science & Humanities (CPCI-SSH), though they partially overlap with SCIE and SSCI (Bar-Ilan 2009). In fact, as a given work included in conference proceedings may later appear as a journal paper, our approach prevents the double-counting problem. Meanwhile, the studies of China's STI related to social science is not a field where conference proceedings are major venues like in computer science (Bar-Ilan, 2010).
Third, we refined search results in "Research Domains" and selected following research domains: "Management/Business/Economics/Planning Development/Information Science Library Science/Multidisciplinary Sciences/Geography/Area studies/Political Science/Education educational research/Public Administration/Urban Studies/Social Sciences Interdisciplinary/Law/Business finance/Agricultural economics policy/Social Issues." The studies of China's STI as an interdisciplinary research domain may include publications in the natural sciences and social sciences. In this paper, we only pay close attention to papers in the research domains of the social sciences such as management, business and economics, and exclude publications in specific scientific and technical fields such as chemistry and information technology.
We cleaned the raw data carefully to ensure their accuracy. For example, institutions may be under different names. We performed a name disambiguation exercise and designated each institute by a unique name. The publication records are defined as a knowledge domain (collection) in HistCite. Our resulting aggregated database is defined as the research collection of China's STI (RCSTI) including publications' references and citations inside and outside of the domain. The citations of a cited publication are calculated by the frequency in citing publications' references. The collection includes 2,041 published records in 506 journals and 71,255 cited references between 1978 and 2015. The paper will investigate the number of publications annually in analyzing the trend of the studies.

Research method
We used HistCite to perform the bibliometric analysis on the WoS data. As a good tool for a historical analysis of literature, HistCite is commonly used to analyze and visualize the development of a research field and to explore its evolutionary characteristics (Garfield et al. 2002;Lucio-Arias and Leydesdorff 2008;Garfield 2009).
HistCite uses the databases of publications with cited references from the WoS or other similar databases to produce various tables and graphs showing informetric indicators of a research field (Garfield et al. 2006).
The number of received citations is a basic scientometric indicator reflecting the impact of a publication record, and citations can show the evolution of a theme within a domain by describing the relationship between publication records. Particularly, local citations are calculated based on the citation frequency within the basic collection RCSTI, and global citation measures how often each paper was cited in the entire WoS realm (Garfield et al. 2002). This paper considers only local citations but not global citations. Therefore, local citation scores include only the citations within the basic collection RCSTI, and global citations scores include those within the SCIE and SSCI so well as those within the WoS realm. Consequently, global citation scores are higher than local ones. Global citations export domain-specific insider-knowledge into other knowledge domains and the scores can show multidisciplinary impacts. But there may be little relatedness of specialized knowledge between these cross-referenced articles (Garfield et al. 2006). Our main concerns were these publications' impacts within the basic collection RCSTI, which reflect the close communication between academic peers and relatedness of specialized knowledge in the same field. Obviously, one limitation of the approach is that we cannot explore the actual outreach of the knowledge domain of China's STI studies in other domains within the SCIE and SSCI realm, let alone that within the WoS realm. LCS/t means the local citations scores per year or the average citation score every year. LCSe denotes local citation scores in the last three years of the collection timespan, and LCSb local citation scores in the first three years of the paper published. LCS (e/b) equals LCSe divided by LCSb. A greater than 1 LCS (e/b) means that citations tend to increase in recent years; when LCS (e/b) is less than 1, citations tend to decrease.

Results
Applying the bibliometric method to the data from the WoS and combining our literature research and research experience, we have reached some results (see Table 1-6).

Descriptive statistics
We are interested in the distributions of publications by economies and institutions, seminal articles, leading contributing journals and main contributors. Table 1 shows descriptive statistics of the main indicators.
[ Table 1 about here] On a full-accounting basis, China-based scholars contributed half of the publications in the studies of China's STI and scholars from seventeen economies contributed at least one percent of the publications.
Scholars from Tsinghua University and Zhejiang University contributed 90 or more publications and received 232 local citations while there are institutions that contributed a publication but did not receive any local citations.
Only about a third of the publications received local citations by the means of LCS; all publications being 1.52. Thus, the local citations in the studies of China's STI is small. The greatest number of papers published in a journal is 101 articles and the greatest number of papers an author published is 27 articles.

The growth of China's STI studies
The first paper that studies China's STI is "Education, Science, and Technology in China which included EST as one of the papers. Although it is a personal reflection of an editor of Science, an outcome of science tourism as it was known, rather than an academic paper, it describes China's scene of S&T in 1979.
At that time, most research was carried out at the institutes affiliated with the Chinese Academy of Sciences and a lesser amount at universities. China's leadership had little experience in the effective integration of advanced research and development into major industrial complexes (Abelson 1979). EST also for the first time raised a very critical question: "Why hasn't China developed faster and more extensively?" In fact, the question has hovered over the studies of China's STI for the decades to come. China's R&D intensity, or its R&D expenditure as a percentage of GDP, reflecting to some extent their relatedness. It can be argued that China's R&D investment has promoted its S&T development and innovation, which in turn has attracted domestic and international academic community's attention to the studies of China's S&T and innovation.

Contributions to the studies of China's STI
This section traces the growth of the studies of China's STI and analyzes contributing economies and institutions to the field. It consists of three parts. The first reviews the seminal contributions of China's STI studies. The next two parts analyze main journals publishing China's STI studies and leading contributors. Table 2 shows the records of full-counting-based papers published by scholars from various economies between 1978 and 2015. China is the undeniable leader, accounting for 50% of the 2,041 records that explicitly analyzed STI in the country, followed by the U.S. and the U.K., which together contributed 89.2% of the total. 1 It stands to reason that China-based scholars have paid most attention to the STI issues in their own country. Some of them have international visions and academic competence in the social sciences, and their research on China's STI has been accepted by the international academic community and published in international journals.

The contributing economies
[ Table 2  Geographically, South Korea and Japan have been and will continue to be influenced directly by China's development in STI. However, scholars from the two countries have paid more attention to their own nation's STI issues than those of China (Sun and Grimes 2016). Despite their much smaller economies, both Singapore and Malaysia have seen their scholars paying very close attention to STI in China. India, another emerging economy, did not show much interest in China, with contribution of its scholars to the studies of China's STI only being 1.7% of the total. The remaining nine economies accounting for 14.7% of the contributions are located at continental Europe. This confirms that outside of Greater China, North America and Europe are key centers for the studies of China's STI, which is consistent with the findings from the national innovation studies (Sun and Grimes 2016). Despite an increasingly diverse profile of contributing economies, the substantial rise of China's STI research has been largely limited to the key centers, with some of China's geographically proximate neighbors contributing to the field, differing from the research focusing on China as a country (Liu et al. 2015a(Liu et al. , 2015b. Although China as a rising STI power has attracted North America and Europe's attention, the studies of China's STI have yet to become a mainstream field on a global scale. For example, most of the work by scholars in East Asia, including China, has appeared in local journals with local languages (Sun and Grimes 2016). But language is not the primary reason deterring scholars of these economies from publishing their work internationally. Rather, the publication phenomenon may suggest that China has not generated enough influence and has not attracted sufficient global attention. As China's rise in the technological term could gradually change the geographical configuration of global R&D and innovation, North America and Europe are the first to feel China being their competitor and collaborator. Japan and South Korea, as two major R&D and innovation centers in Asia as well as in the world, still are closely watching the U.S. and Europe rather than neighboring China.

Leading contributing institutions
We identified top 20 of the 1,289 institutions that contributed to the studies of China's STI in terms of records and TLCS (Table 3). These are the institutions with which the primary authors -first and corresponding authors -are associated.

Seminal contributions
The number of citations that a publication received is related to factors including the type of documents (review articles attracting more citations), the number of authors, and others, while high citations are correlated with peer judgments about scientific excellence and the importance of the contributions (Garfield 1979).
Generally, the earlier a paper was published, the more citations it would receive, controlling for the quality of the paper and other factors. Considering that the timespan between the publication date and 2015 would influence the total number of citations, LCS/t -the average local citations scores per year since the publication date -is a more effective indicator than either TLCS or TLCS/x -total citation scores, excluding self-citations, in identifying seminal contributions. Only a small number of the 2,041 articles can be regarded as "seminal," which is measured by LCS/t. Table 4  [ Table 4 about here] Among the topics of the seminal contributions were national innovation systems, innovation capacity and policies, regional innovation strategy and capability, and firm's product innovation, capability and performance.
Measured by LCS/t, TLCS and TLCS/x, the paper on top of the seminal paper list is "Comparing innovation systems: A framework and application to China's transitional context," mentioned above, published in RP by Liu and White. In fact, this paper not only examines China's innovation system in transition but also extrapolates directly from system modules to functions to address a fundamental weakness of the national innovation system research -"the lack of system-level explanatory factors." It focuses therefore on the five

Leading contributing journals
The 2,041 papers were published in 506 journals, with ten journals publishing 25.6% of them, showing a success-breeding-success phenomenon (Price, 1976). Measured by the number of records, the most important journals are Scientometrics (SCIM, 4.9% of the publications), the International Journal of Technology Management (IJTM, 3.4%), and Technological Forecasting and Social Change (3.1%), indicating a wide and diversified outlet (Table 5). SCIM is concerned with the quantitative features and characteristics of S&T activities, mainly analyzing publication and patent statistics. Emphasis is placed on statistical and mathematical methods used to analyze the development and mechanism of S&T; or plenty of China's STI studies are data driven.
[ Table 5 about here] In addition to the number of records, the average annual LCS since the publication date of papers (LCS/t) provides an indication of the impact of journals in the field of China's STI studies (Sun and Grimes 2016). has a longer history than CMS. Obviously, the international academic community studied the Chinese economy earlier than its management.
The impact factor of SMJ is higher than those of TIM journals, although it is difficult to compare the impact factors of journals in different fields because of their varying missions and scopes (Dorta-González and Dorta-González 2013). TIM is a subfield of strategic management, and SMJ, founded in 1980, is the world's leading journal for research in strategic management.

The leading contributors
We used number of records, TLCS and LCS/t to identify the seminal contributors, which are important for understating a field. A total of 3,473 authors had contributed to China's STI studies, but name ambiguation implies the number of authors may be more. Bibliometrics has not been able to offer a valid overall alternative because of the almost overwhelming difficulty in identifying the true author(s) of each publication (D'Angelo et al. 2011). This is especially in our case because of the ambiguity of Chinese names (Cornell 1982).
Let's use a single example to illustrate the severity of name ambiguity. A contributor "Liu Y" ranks the third place contributing 23 records in our database. A manual reading turned out that the single entry "Liu Y" refers to nine different individuals. Liu Yi from Xi'an Jiaotong University contributed 10 records, Liu Yun from the Beijing Institute of Technology contributed 2 records, two Liu Yangs -one from the South China University of Technology contributing 4 records and the other from the University of Science and Technology of China with 2 records, and Liu Ying from Tsinghua University, three Liu Yus -one each from Peking University, the University of Texas, El Paso and Capital Medical University contributing 1 record, and finally, Liu Ye from the China University of Geoscience also had 1 record. By comparison, the most common surname/initials combination is only shared by 1.2% of the authors of that surname in the Western case (Cornell 1982).
[ Table 6 about here] In order to identity prolific authors, we manually disambiguated leading authors by taking their affiliations, co-authors, cited authors into consideration (see Table 6

Main topics in the studies of China's STI
Identifying main topics is useful for deepening the understanding of any academic field. Topics identified based on citation network analysis reflect knowledge connection between publications and clear relatedness in the timespan of the publications, although co-citation network, bibliographical coupling, and co-word analysis can also identify topics of the publications. Here, we analyze the main topics in the studies of China's STI based on citation network, seminal contributions and our research experience.
First, main topics are essentially the ones that a large number of scholars pay attention to and publish extensively. Seminal contributions can come from main topics. Citation network is helpful in generating most highly cited publications and their citation linkages, thus shedding some light on the evolution of a field and how linkages have identified over time. Therefore, citation network is a useful but insufficient tool to identify the main topics. Some papers that can be regarded in the same topic had no citation linkages, which might be partly attributed to negligence or bias on the part of authors. For example, in the citation network, both node 57 and node 148 are related to the reform of the innovation system, but the node 148 did not cite the node 57; nodes 258, 293 and 336 all are about Zhongguancun Science Park, but there were no citation linkages between them.
In addition, cross-referencing effects exist between different topics (Sun and Seamus 2016). Certainly, we can also achieve clustering by using other tools such as co-keywords and co-citations. However, similar to citations, the existence of co-keyword and co-citation relations does not necessarily suggest relations between research contents; what is intrinsically related may not necessarily have the above quantitative relations. Therefore, we also incorporated judgments based on our experience as participants in this field and especially our familiarity with the literature and leading contributors.
Second, as mentioned, the studies of China's STI is an epitome of studying China's STI development coevolving with such development. There was only one node in 1994 and 1997 respectively. The main contributions in the studies of China's STI appeared after 2000. Thus, seminal contributions could reflect the main topics of the field (see Table 4). Third, some papers excluded in the network for few citations have certain significance and influence on China's STI enterprise so as to be included in our discussion.
[ Figure 2 about here] CiteNetExplorer generates 50 most cited papers based on LCS in the citation network (see Figure 2 and Table 7). In Figure 2

Macro
At the early stage of reform and opening-up, China focused on S&T activities. The concept of "innovation" was first introduced in 1987 when Ye (1987) argued that China should seize the opportunity of the new technological revolution, accelerate industrial technological development, and transfer the world's new technology to China.
Obviously, Ye's discourse extended the innovation from technological to economic development.

S&T reform and innovation system development
The reform of China's S&T system began in 1985 following the economic reform in 1978. Scholars started reviewing and assessing the reform some ten years later, and this line of work has continued to be active, generating a number of publications. Of them, in a pioneering work, Xue (1997) (record 57) pointed out that despite significant progress in the reform of its S&T system, China still faced challenges such as weak industrial innovation before achieving a true transition to an enterprise-centered innovation system. A seminal work by Liu and White (2001) (record 148) indicates that with the reform, China's innovation system had been more effective than that under central planning in introducing, diffusing and exploiting innovations.
In the following year, Sun (2002) (record 177) indicated that the state plays a crucial role in the development, restructuring, and performance of China's innovation system during the transition period.
Updating 's work, Sun and Liu (2010) found that China's R&D funding has shifted from a government-led to an enterprise-centered model, while the central government leads in reforming its innovation system with "Chinese characteristics." In addition, scholars have compared innovation systems between Mainland China and Taiwan (Chang and Shih 2004)  ineffectiveness in macro-level coordination that influences distribution of resources at the meso level; and a weak culture of performance evaluation at the micro level with an overwhelming "publish-or-perish" orientation. Further, in their review of China's post-2012 S&T and innovation system reform, Cao and Suttmeier (2017) pointed out that new reform policies promise new problems while ignoring a deeper underlying obstacle, the state's role in the pursuit of innovation.

Innovation policy and indigenous innovation strategy
As a developmental state, China has formulated a slew of innovation policies, which in turn have played a central role in the development of S&T and innovation. Huang et al. (2004)

Meso
The amount of papers on the geography of STI in China reflects the importance of knowledge stickiness in space and industry as well as the industrial context of innovation.

The territorial distribution of innovation
Region is important for STI development in a vast territory like China. Research on territorial distribution of innovation has considerable potential (Sun 2016). Sun (2000) (record 118) found that patents are highly clustered in coastal provinces with rapid economic growth and inland provinces with a large population base, but the degree of regional concentration of overall patents is declining over time.  (record 558) indicated that regional disparity in innovation comes from not only regions' different level of R&D investment but also their inefficient innovation process.
Territorial distribution of the ICT industry is another interesting topic. Wang and Lin (2008) (record 458) revealed that ICT manufacturing has a strong tendency to conglomerate in the eastern coast, but these is no significant relationship between spatial agglomeration and innovation and economic performance. Their subsequent work confirmed that frequent and intensive production linkages among firms in Shenzhen's ICT cluster were unable to guarantee outstanding innovative performance of these firms ) (record 746).

Zhongguancun Science Park
Moving from territorial distribution to a certain region, Zhongguancun Science Park in Beijing, and indeed, science parks in China as a whole, has become the focus of attention. In 2004, Cao (2004) asked whether Zhongguancun's transition represents a process of "growing pains" or "premature senility." He concluded that the overwhelming role of government had impeded the efforts of Zhongguancun from duplicating the Silicon Valley model.  (record 258) examined the changing patterns of behavior and interaction among the state, domestic and multinational corporations during three different stages of Zhongguancun Science Park's development. Liefner et al. (2006) (record 293) found that companies in Zhongguancun are linked to both sources of knowledgeforeign companies and domestic universities and public research organizations. Tan (2006) (record 336) found that Zhongguancun has played a vital role in facilitating technology transfer and innovation since its inception. However, within a relatively short period, Zhongguancun started to show signs of premature aging and decline, especially when compared with its role model, Silicon Valley, a conclusion that is similar to Cao (2004).

Technological learning and catching-up in industry
An emerging country like South Korea has gone through a technological learning and catching-up process (Lee and Lim 2001). China is not different. Lu and Lazonick (2001) (record 139) demonstrated state's strategic role in the increasing importance for economic success of the integration of investment and organizational learning in Chinese electronic publishing industry. Modifying and applying Lee and Lim (2001) to the study of China's telecommunication industry, Mu and Lee (2005)

(record 265) found that the important factors in
China's catching-up process are the "market for technology" strategy, knowledge diffusion from MNCs to domestic firms, and government-driven industrial upgrading.

Micro
Innovation has been extended from enterprises to other organizations of the innovation system. So, it's clear, the STI research at the micro level supplements and complements that at the macro and meso levels.

The determinants of enterprises' innovation and performance
Published work tries to determine the factors impacting an enterprise's innovation and performance. Pappy and Song (1994) (record 32) indicated that relative product advantage and the acquisition of marketing information were highly correlated with new product success based on a survey of 129 state-owned enterprises (SOEs) in China. As China has witnessed its enterprises' innovation performance rising continually, studies have shown the impacts of several crucial factors.
First, personal relationship, or guanxi, has become a very important construct in explaining a firm's performance in the Chinese context. Park and Luo (2001) (record 142) provided strong evidence that Chinese firms develop guanxi as a strategic mechanism to overcome competitive and resource disadvantages. Gao et al. (2008) (record 472) found that absorptive capacity moderates the effect of managerial ties on a corporation's innovativeness. Sheng et al. (2011) (record 857) indicated that business ties have a stronger positive effect on firm performance than political ties, and both effects depend on institutional and market environments.
Second, organizational capability is central to innovation. Yam et al. (2004) (record 232) suggested that Chinese firms consider a more balanced focus on their harmonizing capabilities of technological innovation between R&D, resources allocation, learning, and organizing. Chen et al. (2005) (record 254) concluded that conflict management based on cooperation promotes productive and top management team's effectiveness, which in turn leads to organizational innovation. Zhou and Wu (2010) (record 723) proposed that technological capability has differential effects on exploitative and explorative innovation.
Third, ownership matters in innovation.  (record 146) indicated that managers from large SOEs are not as innovative and are less willing to take risks than entrepreneurs from small privately-owned enterprises (POEs). Zhou et al. (2017) showed that the state ownership enables a Chinese firm to obtain crucial R&D resources but makes the firm less efficient in using those resources in innovation, and that a minority state ownership is an optimal structure for innovation development in this context.
Finally, public policy can help create an environment conducive to innovation. Zhao (2006) (record 327) argued that weak intellectual property right protection leads to low returns to innovation and underutilization of innovative talent, and MNCs that possess alternative mechanisms for protecting their intellectual properties therefore find it attractive to conduct R&D in China. Hu and Jefferson (2009) (record 609) found that amendments to China's patent law that favor patent holders and ownership reform that has clarified the assignment of property rights have led to China's patent boom. Guan and Yam (2015) indicated that government financial incentives such as special loans and tax credits positively influenced firm's innovative economic performance; however, direct earmarks not only failed to enhance and sometimes negatively affected such performance.

Enterprises' indigenous innovation capability
Research on Chinese firm's indigenous innovation capability had already begun before China launched the indigenous innovation strategy in 2006.  (record 171) revealed that in-house R&D efforts, rather than imported technologies, are the primary sources of industrial innovation in China. Zhou (2006) (record 312) found that an innovation strategy performed better over an imitation strategy in new product development.
Clearly, both in-house R&D and indigenous innovation capability are more important than imported technologies and imitation.
After 2006, the studies of China's STI have paid more attention to the issue of indigenous innovation. Guan et al. (2009) (record 586) found that manufacturing firms had already started moving away from a reliance on imported technology and equipment and using indigenous R&D efforts to innovate in the market economy. Wang and Kafouros (2009) found that international trade, FDI and R&D do not always lead to a firm's positive innovation performance, and their effects are moderated by technological opportunities and foreign presence. Li et al. (2010) (record 739) suggested that firms that invest in R&D and marketing activities benefit further from access to foreign knowledge due to increased absorptive capacity.

University spin-offs and Ventures
University's central role in an STI system is also reflected on spinning off firms, collaborating with enterprises and so on. Just as Chen and Kenney (2007)  continued to find that successful agency business activity is positively related to new venture performance but negatively related to its product innovation efforts in Chinese high-tech new ventures. Yiu et al. (2007) (record 373) examined empirically that relationship between firm-specific ownership advantages and international venturing is moderated by the degree of home industry competition and export intensity. Li and Zhang (2007) (record 375) demonstrated that managers' political networking and functional experience are positively related to new venture's performance.
Finally, Zhang and Li (2010) (record 700) proposed that new ventures' ties with service intermediaries enable the ventures to plug into these networks and contribute to the ventures' product innovation by broadening the scope of their external innovation search and reducing their search cost.

Summary
The studies of China's STI have been rapidly emerging since 1995, during which some hot topics have emerged.
At the macro level, scholars have paid more attention to China's S&T reform and innovation system, innovation policy and indigenous innovation strategy, and national innovation capability. As the government's "unlimited power" is the nature of innovation system with "Chinese characteristics" (Fang 2010), the comprehensive reform of the S&T system will depend on the further reform of China's political system. China's innovation policy was shifting to a more market-based system through critical financial, tax, and fiscal measures; China's innovation capacity has improved measured by the indicators of publications and patents.
At the meso level, territorial distribution of innovation, Zhongguancun Science Park in particular, and technological learning and catching-up in industry have attracted significant scholarly attention. China's innovation tends to concentrate on coastal provinces, which has been fundamentally driven by R&D investment, industrial specialization and innovation efficiency. There is no significant relationship between spatial agglomeration and innovation in China's ICT industry. Zhongguancun Science Park has started to show signs of premature aging and relative decline in China's innovation with the rise other high-tech zones such as Shenzhen.
And China's technological catching-up depends on in-house R&D development at enterprises to build innovation capability and governments' promotion.
At the micro level, scholars are interested in learning the determinants of Chinese enterprises' innovation and performance, enterprises' indigenous innovation and university's spin-offs and venturing. Guanxi and managerial ties lead to higher firm innovation performance, innovation at Chinese firms depends on learning, technological capability, and ownership and public policy as important institutional factors influencing innovation. Indigenous and foreign innovation efforts are complementary, and the assimilation of foreign technology depends on firms' in-house R&D and absorbing ability. Universities have played an extremely important role in innovation. However, theoretical contributions coming out of China's experience and context are few and far between, except guanxi as a concept of management studies bringing such experience to the international scholarly community.

Conclusion and Discussions
Since 1978, the studies of China's STI have been emerging as a new field attracting extensive scholarly attention alongside the country's rise as a leading country in STI. This paper attempts to provide a comprehensive and synthetic picture of China's STI study literature both qualitatively and quantitatively. The study includes 2,041 papers published by scholars from 1,289 institutions in 506 journals with 71,255 references and citations, from which we have found the following key findings. Third, we identified several key research areas at the macro, meso and micro levels of the studies of China's STI, which also prompt our consideration of agendas for future research. At the macro level, potential topics of research include examining the differences between China and developed economies in terms of their institutional environment of STI development, practically the role of state in the STI system. China's model deviates from the developmental state or the state-led innovation system. As a transitional economy, China has been shifting from a centrally planned economy to socialist market-oriented economy, and from a closed to open economy. In this context, we want to know more about China's system of R&D expenditure, such as structure of government funding and corporate investment on R&D, allocation mechanism of R&D funding, performance and efficiency of R&D expenditure, which is at the core of its innovation system determining China's innovation capacity; China's mechanism of S&T evaluation, the foundation of S&T governance, whose reform is central to improving China's efficiency of S&T activities.
At the meso level, there is a lack of theorization of the Chinese innovation practice, as scholars still prefer to use existing theories like Lee and Lim (2001) Finally, our bibliometric analysis may suffer from two limitations. First, we limited our data to the publications indexed by SCI-E and SSCI included in the Web of Science, which favors English-language journals and excludes books. While probably incomprehensive, our analysis might not have missed that much as many scholars also had published their journal articles before tuning out book-length, more systematic analysis.
Our study also excluded the database of conference proceedings -CPCI-S and CPCI-SSH, among others -in the WoS. For example, Jin Chen and his group have published at least 14 papers at conference proceedings included in CPCI of the WoS since 2006. The second limit has to do with the drawback of the bibliometric method. For example, name ambiguity has failed us to identify all authors, although we managed to manually identify some leading contributors; citation analysis may not capture the organic linkage and evolution of the literature. Nevertheless, we tried to compensate such a problem with a careful reading of the entire body of the literature so as to figure out its essence at the macro, meso, and micro levels. Notes: These descriptive statistics were analyzed on the basis of software-HistCite statistics, and the problem of name ambiguation was not taken into account in the statistics, although we manually identified the major contributors. Name ambiguation will influence the results of LCSx and the records of contributors a little. Notes: First, the main contributors ranked by their contributions in studies of China's S&T and innovation, and by the number of papers they published, without distinguishing whether they were first author, correspondence authors or other authors. Second, the author's institution is the latest institution signed in their latest publication during the period of this study, and these do not represent their current and previous institutions. For example, now Yuan Li works in Tongji University. Third, indeed, there are only 19 authors whose number of publications equal to or more than nine.