Introduction

Innovation is the reorganization of existing ideas or the application of new ideas to new processes and products, which has received widespread attention from researchers (Matos et al., 2022). Successful economic development is closely related to a country’s ability to acquire, absorb, disseminate, and apply modern technology, as reflected in its national innovation system (NIS) (Metcalfe and Ramlogan, 2008; Kim and Lee, 2022). In the NIS conceptual framework (Freeman, 1987), government and institutional structures promote the generation and diffusion of innovation in coordinating the national economy (Watkins et al., 2015; Ulmanen and Bergek, 2021). The NIS encompasses all economic, political, and social factors affecting national innovation. The NIS can enhance environmental sustainability and play a decisive role in globally coordinated efforts to create a sustainable future (Fernandes et al., 2022). Countries that have successfully integrated these innovative factors perform well and achieve greater economic prosperity (Kwon and Motohashi, 2017; Khan, 2022). Within the NIS is a group of institutions jointly and individually dedicated to developing and disseminating new technologies (Metcalfe, 1995). In other words, the NIS depends on not only the roles of individual institutions but also how internal system factors interact (Calia et al., 2007). Therefore, in the innovation process, the connection between internal elements of the NIS is key to improving a country’s innovation capability.

Scholars have extensively researched national innovation capability (NIC) from the perspective of the NIS and analyzed the conditional configurations that lead to high NIC (Khedhaouria and Thurik, 2017). However, there has been no in-depth discussion of which specific configurations can stably generate high NIC. First, NIC research has not been integrated (Fagerberg and Srholec, 2008). Some studies cover only the conditions considered to affect innovation (Khayyat and Lee, 2015), and may thus overlook other potential influencing factors and the interactions between them. However, the NIS is systematic and evolutionary (Lundvall et al., 2002), and scattered literature is not conducive to researchers and managers fully understanding national-level innovation activities. Second, research on the configuration of conditions for NIC has used cross-sectional data, analyzing the innovation paths of different countries in a given year (Crespo and Crespo, 2016; Huarng and Yu, 2022). Although these studies adopted a systematic perspective, cross-sectional data cannot fully explain the temporal differences in NIC and the continuity of configurations. In other words, whether the innovation paths of different economies have evolved over time remains to be determined. Accordingly, this study tackles the following research questions:

RQ1: Do the five elements of the NIS individually affect a country’s NIC? How does the combination of factors affect the NIC of high- and upper middle-income economies? Specifically, which paths generate high NIC?

RQ2: Do these paths have time- or cross-sectional effects? Are they stable?

RQ3: Are there configurations that can simultaneously drive both economies to generate high NIC?

Cross-border comparisons based on the NIS, especially analyzing non-highly industrialized countries, are a trend in NIS research (Balzat and Hanusch, 2004). This study analyzes the individual and combined influence of NIS elements on NIC in high- and upper-middle-income economies and explores whether the configuration of high NIC has temporal and spatial effects. High-income economies are the focus of most theoretical and empirical analyses of the NIS (Edquist, 2001; Khan, 2022), and analyzing the innovative development path of high-income economies can help other economies learn from their successful practices. Upper middle-income economies include many emerging economies that have shown great potential and growth momentum in innovation activities (WIPO, 2022). In addition, NIS theoretical research has gradually expanded its focus to include “imitation” and “learning” activities in middle-income economies (Casadella and Uzunidis, 2017; Hu et al., 2017). Studying the current status of NIC in upper middle-income economies is conducive to promoting global innovation development. High-income and upper middle-income economies play important roles in global innovation. Therefore, analyzing and comparing their innovation paths can support international cooperation and knowledge exchange, promote innovation path sharing, and advance global innovation progress.

This study explores how five elements of the NIS (institutions, human capital and research, infrastructure, market sophistication, and business sophistication) affect the NIC of high- and upper middle-income economies and the stability of their influence over time. This study has significant theoretical contributions. Firstly, this study improves the transparency and replicability of innovation paths in different economies by analyzing the time effects of NIS factors on the necessity and combination effects of NIC and broadens the research perspective of NIS-related theories. Furthermore, this study incorporates temporality into the fsQCA method, revealing the stable allocation of NIS elements, characterizing the evolution trend of the global innovation landscape, and providing robust data support for various economies to adjust innovation strategies. The findings suggest that high- and upper middle-income economies should prioritize the role of infrastructure in NIC. Additionally, a configuration perspective should be applied to integrate NIS elements and thereby enhance NIC, focusing especially on the interaction between human capital and research (HCR), market sophistication (MS), and business sophistication (BS). Overall, the research findings provide key insights for decision-makers in countries with different economic levels to adjust innovation policies and integrate innovation resources, especially those committed to improving NIC and global competitiveness.

The remainder of this paper is organized as follows. The next section reviews literature on the NIS, its constituent elements, and NIC. Next, we detail the research methodology, then report the results of using fsQCA with panel data from the Global Innovation Index (GII) Report, 2011–2022, to explore what combinations of NIS elements produce high NIC. Finally, we explore the study’s theoretical and practical significance, analyze its limitations, and propose future research directions.

Literature review

Innovation can drive economic growth, technological progress, and global competitiveness. We take the NIS as our theoretical framework and believe that its internal institutions influence the growth of NIC. First, we review research on the NIS and NIC. Second, we consider the five innovative elements outlined by the GII Report as the framework for evaluating the institutional structure of the NIS.

National Innovation System (NIS)

The NIS was developed by Freeman as a conceptual framework positing that innovation activities are not just the work of enterprises or individuals (Freeman, 1987). Governments and institutional structures play crucial roles in coordinating innovation activities within a national economy (Watkins et al., 2015; Crespo and Crespo, 2016). Important research findings on the NIS are as follows. First, within the NIS, a country’s key organizational structures and systems interact, with the core activities being the co-creation, storage, and transmission of new knowledge and technology (Alcorta and Peres, 1998; Erzurumlu et al., 2022), thereby upgrading NIC (Numminen, 1996; Jankowska et al., 2017). Second, the growth of knowledge and technology in the NIS comes mainly from intangible national investments in education, infrastructure, technology research and development, business connections, and other domains (Crespo and Crespo, 2016). Third, government policies are responsible for coordinating the cooperation of various institutional structures in the NIS (Prokop et al., 2021). Fourth, economic globalization has gradually blurred the boundaries of innovation activities, increasing the mobility of knowledge in various fields and regions, and thus providing opportunities for developing countries to catch up. Therefore, developing countries should observe and learn from the NIS institutional structure of high NIC countries (Rakas and Hain, 2019; Lee et al., 2021; Khan, 2022). The NIS has become a popular theoretical framework for describing how interactions between economic participants generate technological innovation and economic growth (Fernandes et al., 2022). This study investigates how the interaction between NIS internal factors impacts NIC in high-income and upper middle-income economies.

National Innovation Capability (NIC)

Based on the above conceptual framework, this study regards NIC as the ability of various actors and institutions within a country to interact, engage in innovative activities, produce innovative technologies, and promote national economic development. Existing research uses the NIS as a theoretical framework for discussing the NIC of different economies (Acs et al., 2017; Kashani and Roshani, 2019). Scholars have not yet reached a consensus on the framework of NIS elements, as different economies have specific economic foundations and innovative institutional structures. The biggest challenge in enhancing NIC is determining how the government can reasonably allocate the internal institutional structure of the NIS (Freeman, 1987; Crespo and Crespo, 2016). Meanwhile, since innovation is a dynamic evolutionary process, analysis of a country’s allocation methods of NIS elements needs to consider time effects (Erzurumlu et al., 2022). Therefore, this study incorporates time effects into the analysis and explores the configurations of NIS factors in economies of different income levels from 2011 to 2022, as well as the ways in which these factors affect NIC.

NIS elements

The GII Report provides an important reference for evaluating the NIS institutional framework and measuring NIC. The index was jointly developed in 2007 by INSEAD Business School, Cornell University, and the World Intellectual Property Organization (WIPO). It comprises five innovation input indicators—institutions, human capital and research, infrastructure, market sophistication, and business sophistication—and ranks countries based on the average of each indicator of innovation input and output (WIPO, 2022). The GII has been used to study the NIS and NIC and is believed to provide an effective NIS framework that can measure a country’s NIC (Crespo and Crespo, 2016; Maruccia et al., 2020). Therefore, this study suggests that an NIS comprises the same five elements as the GII and that their combined effects impact NIC.

Institutions (INS) have always been considered a determining factor for a country’s NIC (Nelson, 1988). Within an economy, the responsibilities and relationships of actors are governed and regulated by laws, policies, and other institutional factors. The political, regulatory, and business environment set the conditions for innovation in different economies (Watkins et al., 2015), which vary in their ability to accommodate the high degree of uncertainty associated with innovation activities (Nelson, 2008). Countries with strong institutional support for innovation provide comprehensive policy services and a stable political environment for enterprise operations, featuring intellectual property protection, tax exemptions, and other policies to reduce operational risks and encourage innovative activities by enterprises and research institutions (Erzurumlu et al., 2022). The regulatory environment reflects government efforts to reduce conflicts between private sector development and economic actors and to cultivate more stable cooperative relationships (Furman et al., 2002; Yu et al., 2020). Meanwhile, the business environment reflects the government’s efforts to reduce uncertainty for corporations, for instance, by helping new entrepreneurs to start enterprises, resolving bankruptcy and tax issues, and encouraging the competitiveness necessary for innovation (Schot and Steinmueller, 2018).

Human capital and research (HCR) are valued as key elements of NIC (Castellacci and Natera, 2013). Having a well-established education system can create the human capital required to build innovation capabilities (Khedhaouria and Thurik, 2017). Higher education promotes innovation by directly supplying human capital and generating useful knowledge to support innovation (Saad et al., 2015). In addition, the research and development (R&D) activities of human capital are key to promoting innovation. Supporting internal R&D activities and improving the quality of scientific and research institutions are conducive to the assimilation, absorption, and creation of knowledge products (Chen et al., 2020).

Infrastructure (INF) refers to the goods and services provided by a country for the transfer and dissemination of technological knowledge that can enhance the country’s ability to absorb, adopt, and implement advanced foreign technologies (Castellacci and Natera, 2013). Infrastructure improvements can enhance innovation abilities and speed. The most common infrastructure in the NIS is information and communication technology (ICT), which enhances NIC by widely disseminating new ideas and investing in new applications (Lee et al., 2016). In addition, other infrastructures such as public utility networks (Bronzini and Piselli, 2009), power infrastructure and transportation systems (Lopez-Claros and Mata, 2010), and ecological sustainability infrastructure (Rakas and Hain, 2019) can also reduce economic costs and strengthen NIC. Communication and transportation logistics infrastructures can also promote innovation through promoting and disseminating innovative achievements (Fakhimi and Miremadi, 2022).

Market sophistication (MS) reflects the quality of the market, including finance, investment, trade, market size, and competition, and is an essential element of innovation (Filippetti and Archibugi, 2011). The financial system provides trade, competition, and sufficient domestic market size for enterprise prosperity and innovation (Helveston et al., 2019). Fierce competition encourages enterprises to continuously improve the quality of products and services to achieve continuous innovation and increase their market share and competitiveness (Wen et al., 2022).

Finally, business sophistication (BS) reflects the level of business conditions related to knowledge workers, the quality of business clusters and networks, and the absorptive capability of enterprises (Furman et al., 2002). Innovative talents with high-level skills are essential for enterprises to achieve innovation and improve their competitive advantage (Marvel et al., 2020). Employees with higher levels of education, experience, and skills are more eager to create new knowledge through communication and cooperation (Al-Omoush et al., 2022). Establishing innovative connections with joint ventures and strategic alliances can promote knowledge dissemination and accelerate innovation (Erzurumlu et al., 2013). In addition, the liberalization of international markets is conducive to the absorption and flow of knowledge (Khedhaouria and Thurik, 2017), and domestic innovation can be achieved by opening the NIS to international markets to attract foreign investment or by importing high-tech to absorb knowledge (Castellacci and Natera, 2013). Based on the above discussion, this study proposes the following:

Proposition 1: The existence of institutions (INS) leads to high NIC.

Proposition 2: The existence of human capital and research (HCR) leads to high NIC.

Proposition 3: The existence of infrastructure (INF) leads to high NIC.

Proposition 4: The presence of market sophistication (MS) leads to high NIC.

Proposition 5: The presence of business sophistication (BS) leads to high NIC.

The theoretical framework of this study is shown in Fig. 1.

Fig. 1
figure 1

Theoretical framework.

Full size image

Research methodology

Method selection

This study used fsQCA to analyze the configurations of NIS elements generating high NIC, according to GII Reports. QCA utilizes Boolean logic and algebra to analyze cases by exploring the common influence of multiple antecedent interactions on specific phenomena (Ragin, 2008). Unknown facts can be understood using information that is known to people (Thomann and Maggetti, 2020). FsQCA is suitable for this study for three reasons. First, fsQCA aims to explain the multiple pathways that lead to specific outcomes (Beynon et al., 2019) and analyze the impact of different combinations of antecedent conditions on those outcomes (Ferrer et al., 2023). It can thus delve deeply into the mechanisms of action between NIS elements and NIC, providing a clearer explanation for high NIC. Second, fsQCA is gradually being applied to research in the field of innovation (Dabić et al., 2021), particularly with GII Reports as the data source (Crespo and Crespo, 2016; Huarng and Yu, 2022). Third, existing studies have shown that fsQCA is suitable for small to medium-sized samples (López-Cabarcos et al., 2022), such as the one used in this study.

Prior studies have mostly used fsQCA to analyze only cross-sectional data, thus neglecting temporal dynamic effects (Huang et al., 2022). This prevents researchers from determining whether the configuration results are stable. With advances in research methods, fsQCA is gradually being used to process panel data (Guedes et al., 2016; Beynon et al., 2020; Piñeiro-Chousa et al., 2023). By incorporating time effects into fsQCA, this method can effectively evaluate the set theory relationship of panel data by comparing consistency and coverage across time and cases (Garcia-Castro and Ariño, 2016). This can compensate for the limitations of analysis based solely on cross-sectional data and thus provide new insights. This study follows the steps of Beynon et al. (2020): calibration, fsQCA analysis, and panel data analysis.

Data collection

This study uses GII to measure the NIC of various economies worldwide. Data were sourced from the GII Reports from 2011 to 2022. The sample comprises 44 high-income and 31 upper middle-income economies. The World Bank categorizes economies into different groups based on their per capita national income, including low-income, middle-income, and high-income economies. Due to the wide range of middle-income, it is further divided into two categories: lower middle and upper middle. For example, according to the classification criteria of the World Bank for 2020, a per capita national income of $4096–12,695 is classified as upper middle-income countries, and a per capita national income of US$12,695 or more is classified as high-income economies (the World Bank, 2024). However, this standard is not fixed and will be adjusted according to the global economic situation. This study groups economies based on the development trends of their per capita national income. Therefore, each economy always belongs to the high-income or upper middle-income economic group. The economies covered by high-income and upper middle-income economic groups are listed in the appendix. Table 1 shows the detailed sources of the data.

Table 1 Data sources.
Full size table

Calibration and measurement

In fsQCA, calibration is the process of assigning collective members to cases in which the calibrated collective members are between 0 and 1 (Ragin, 2008). As suggested by Ragin (2008), this study used the three thresholds of 0.95, 0.5, and 0.05 for direct calibration to determine the membership level of each case in the fuzzy set. Table 2 reports the anchor point calibration.

Table 2 Calibration of variables.
Full size table

Results

Necessity analysis

This study used fsQCA software to detect the necessary conditions, which are conducive to, but do not guarantee, the occurrence of results. Following existing literature (Amara et al., 2020; Beynon et al., 2020), the consistency threshold for necessity analysis was set to 0.9. The necessity analysis results are shown in Figs. 2 and 3. None of the antecedent conditions constituted a necessary condition for generating high NIC. Therefore, it is necessary to further analyze the combinations of the five conditional variables to identify paths producing high NIC.

Fig. 2
figure 2

Necessity test results for high-income economies.

Full size image
Fig. 3
figure 3

Necessity test results for upper middle-income economies.

Full size image

Sufficiency analysis

Research often uses consistency thresholds within the range of 0.75–0.85 established by Ragin (2006), or natural discontinuity values that truncate consistency scores (Crilly et al., 2012). In addition, to ensure the minimum number of cases used to evaluate relationships, a frequency threshold must be set. Referring to the recommendations of Fiss (2011), this study set the consistency threshold to 0.8, the proportional reduction in inconsistency (PRI) to 0.75, and the case frequency threshold to 2 to obtain a truth table. The core solution is generally used to determine the number of configurations and variables, whereas the reduced solution is used to determine core conditions. According to Fiss (2011), a condition that appears in the reduced solution is a core condition, indicating a strong causal relationship with the outcome variable, while a condition that appears in the core solution but not in the reduced solution is a peripheral condition, only weakly related to the outcome variable. As reported in Table 3, the configuration results indicate that from 2011 to 2022, four configurations can be considered sufficient conditions for generating high NIC in high-income economies, while there were two such configurations for upper middle-income economies.

Table 3 Configuration of high NIC.
Full size table

Considering first the four configurations that can generate high NIC for high-income economies, configuration H1 included INS, HCR, and MS as the core conditions, and the presence or absence of INF and BS did not affect the results. The consistency score was 96.4%. In configuration H2, the core conditions were INS, MS, and BS, and the presence or absence of HCR and INF did not affect the results. The consistency score was 97.3%. The core conditions in configuration H3 were HCR, MS, and BS, and the presence or absence of INS and INF did not affect the results. The consistency score was 97.9%. Finally, configuration H4 included INS, HCR, INF, and BS as the core conditions, and the presence or absence of MS did not affect the results. The consistency score was 97.7%.

Turning next to the two configurations that can generate high NIC for upper middle-income economies, configuration H5 included INS, INF, and BS as the core conditions, and the presence or absence of HCR and MS did not affect the results. The consistency score was 95.3%. In configuration H6, the core conditions were HCR, MS, and BS, and the presence or absence of INS and INF did not affect the results. The consistency score was 96.1%.

In addition, the sufficiency analysis results indicated that configurations of H3 and H6 are the same, meaning that the combination of HCR, MS, and BS is beneficial for generating high NIC in both high- and upper middle-income economies. Additionally, H3 and H6 had the highest coverage in high-income group and upper middle-income group, indicating that the explanatory power and representativeness of the configuration are relatively good.

Panel data breakdown of FsQCA results

Since the development of NIC is dynamic and continuous, cross-sectional data are not sufficient to explain how the NIS elements affect NIC over time. Therefore, this study adopts the methods and techniques of Beynon et al. (2020) to explore the development of elements and configurations over time, aiming to improve understanding of the interaction between NIS elements and NIC. Results are reported for the time effect of necessary conditions and the time and cross-sectional effects of configurations for high NIC.

Time effect for necessary conditions: INF shows an upward trend

As shown in Figs. 2 and 3, no NIS element constitutes a necessary condition for high NIC. Notably, however, the between consistency (BECONS) adjusted distance of INF for high NIC was greater than the threshold of 0.2. This indicates a significant time effect of INF on high NIC in both high- and upper middle-income economies. The necessity consistency of INF for high NIC increased annually (see Figs. 4 and 5).

Fig. 4
figure 4

Necessity consistency of NIS elements for high-income economies.

Full size image
Fig. 5
figure 5

Necessity consistency of NIS elements for upper middle-income economies.

Full size image

Time and cross-sectional effects for high NIC configurations

Configuration analysis is the focus of fsQCA. Table 4 shows the pooled consistency (POCONS), between consistency distance (BECONS distance), within consistency distance (WICONS distance), and adjusted distance for all the panel data. The pooled consistency (POCONS) of all configurations that generate a high NIC is >0.8, indicating that POCONS has good explanatory power (Guedes et al., 2016). These six configurations can be considered sufficient conditions for high NIC generation.

Table 4 Results of panel data.
Full size table

BECONS measures the consistency of cross-section t in panel data every year (Beynon et al., 2020). The BECONS values for each configuration from 2011 to 2022 are shown in Figs. 6 and 7, demonstrating their evolution trends over time. The consistency of H1 declined slightly in 2018, then showed an upward trend in 2020. H2 and H3 exhibited significantly higher consistency (from 2011 to 2022). Similar to H1, H4 showed a slight downward trend in 2018. H5 showed significantly high consistency from 2011 to 2018, followed by a downward trend after 2019. In contrast, the consistency of H6 was relatively stable. The BECONS adjusted distance tests the time effect in each case. The results show that the BECONS adjusted distance for all six configurations was <0.2, indicating that there was no significant time effect (Garcia-Castro and Ariño, 2016).

Fig. 6
figure 6

BECONS values for configurations in high-income economies.

Full size image
Fig. 7
figure 7

BECONS values for configurations in upper middle-income economies.

Full size image

WICONS measures the vertical consistency of set–subset connections of cases in a panel (Garcia-Castro and Ariño, 2016) and was measured at the national level. Figures 8 and 9 show the evolution trends in WICONS values from 2011 to 2022. The results reveal strong consistency in respective configurations H1–H4 in ~30 countries, together with strong consistency in configurations H5 and H6 for ~20 countries. The figures also present the WICONS values for each configuration in other countries. In addition, the WICONS adjusted distance tests the heterogeneity of cases. The results show that the WICONS adjusted distances of H1–H6 were <0.2. This indicates that there was no significant cross-sectional effect in H1–H6 (Garcia-Castro and Ariño, 2016).

Fig. 8
figure 8

WICONS values for configurations in high-income economies.

Full size image
Fig. 9
figure 9

WICONS values for configurations in upper middle-income economies.

Full size image

As previously mentioned, H3 and H6 have the same configuration. This configuration did not exhibit significant time effects (BECONS adjusted distance <0.2) or cross-sectional effects (WICONS adjusted distance <0.2) in the high- and upper middle-income economies. There is no significant difference in the explanatory power of this configuration between the two economy types.

Discussion

The fsQCA results reveal four paths to high NIC in high-income economies (H1–H4) and two paths to high NIC in upper middle-income economies (H5 and H6). These six paths indicate how the five elements of NIS interact to produce high NIC, and reveal that diverse combinations of antecedent conditions can produce the same result (Fagerberg and Srholec, 2008). This study thus supports the view that innovation is complex (Crespo and Crespo, 2016). To achieve high NIC, different income economies should strive to develop multiple elements of the NIS (Khedhaouria and Thurik, 2017; Huarng and Yu, 2022). This study also supports the view that no single element of the NIS can be a necessary condition for high NIC.

The panel data results show a time effect in the necessity of INF for NIC, both for high- and upper middle-income economies. Previous studies have shown the importance of INF for NIC (Lee et al., 2016) but have not pointed out its time effect. This study thus enriches understanding by suggesting that the impact of INF on NIC gradually increases over time. In other words, INF provides the necessary technological structure for knowledge transfer and diffusion, which is increasingly important for improving the NIC. INF includes Information and communication technologies (ICTs), General infrastructure and Ecological sustainability (WIPO, 2022). Among them, analyzing the performance of ICT can help us understand the importance of INF for NIC. Especially with the spread of the digital revolution into multiple fields, the impact on NIC of ICT—an important INF component—will increase further (Lee et al., 2022). China could be a representative example of the INF role within the NIC framework. The development of ICT has changed the composite pattern of innovation elements and strengthened NIC (Liu and Lee, 2021). A panel data analysis showed that ICT penetration affects economic growth by promoting knowledge dissemination and innovation (Vu, 2011). On the one hand, the extensive network structure of INF generates an economic foundation for innovation by promoting transactions, reducing costs, and increasing market access (Dutta et al. 2010). As enterprises transform and upgrade, ICT helps them collaborate with other participants and promote open innovation practices (Zhou et al., 2019). The role of ICT in upper middle-income economies includes not only supporting the innovation activities of enterprises but also stimulating entrepreneurial activities. With the rise of the internet in China, an increasing number of people are doing business online, and the number of small and medium-sized enterprises relying on online trading platforms is gradually rising. The trend of entrepreneurs adopting ICT used by their peers will expand due to “social influence” (Afawubo and Noglo, 2022). ICT provides hardware support for green technology innovation, and the improvement of green innovation in enterprises and cities has driven the upgrading of existing network infrastructure (Tang et al., 2021). In summary, INF promotes innovation, and improved innovation drives infrastructure upgrading. The role of INF in innovation gradually increases over time.

Combining the results of the fsQCA and panel data analysis reveals one path to generating high NIC in both high- and upper middle-income economies (see Fig. 10). Configurations H3 and H6 are the same, and there were no significant time or cross-sectional effects between 2011 and 2022. This means that the combination of HCR, MS, and BS has a long-lasting and stable association with high NIC in both types of economies. HCR includes performance in the aspects of Education, Tertiary education, Research and development (R&D). MS includes performance in the aspects of Credit, Investment, Trade, diversification, and market scale. BS includes performance in the aspects of Knowledge workers/Innovation linkages/Knowledge absorption. Due to limitations in research methods, we are unable to present the sequential effects of HS, MS, and HCR on high NIC. Notably, the panel data results show that H3 covers the Switzerland case from 2011 to 2022, while H6 covers the Chinese case over the same period, indicating that HCR*MS*BS has strong explanatory power for Switzerland and China. Analyzing the situation in Switzerland and China helps to understand how HCR*MS*BS drives NIC. According to the 2022 GII Report (WIPO, 2022), Switzerland and China rank first among high- and upper middle-income economies, respectively. Switzerland has ranked first for 12 consecutive years, while China is the only developing country with a top 30 GII ranking. As a representative case of HCR*MS*BS in a high-income economy, Switzerland has shown excellent and stable NIC (WIPO, 2022). In the 2022 GII report, Switzerland ranked 4th globally in HCR, 8th in MS, and 7th in BS. In terms of HCR, Research and Development (R&D) has performed exceptionally well in Switzerland. It has a strong education system and is continent-leading in the workforce size and R&D expenditures of knowledge-intensive enterprises—a persistent advantage of the Swiss NIS (Marxt and Brunner, 2013). In terms of MS, credit and investment performed excellently. Switzerland is also closely connected to global markets, and its comprehensive tax system encourages regions to actively promote innovation (Erzurumlu et al., 2022). In terms of BS, the performance of Innovation linkages is very high, mainly manifested in University–industry R&D collaboration, Patent families, and State of cluster development and depth (WIPO, 2022). Similar to the situation in Switzerland, China’s Research and Development (R&D) has performed outstandingly in HCR. The National Bureau of Statistics of the People’s Republic of China pointed out that in 2022, the Chinese government’s R&D investment exceeded 3 trillion yuan (The State Council of the People's Republic of China, 2023). China’s emphasis on basic research and major technological infrastructure is increasing. In addition, thanks to the large-scale market scale, China’s MS has performed excellently. MS drives innovation in emerging enterprises in various ways. For example, green investments in China drive enterprises to engage in green innovation activities by alleviating financing constraints (Zhang et al., 2023). This is conducive to stimulating the innovation vitality of enterprises and improving NIC. On BS, multiple government departments in China have introduced various measures to reform intellectual property rights, demonstrating commitment to further enhancing the country’s innovation capability (The State Council Information Office of the People’s Republic of China, 2022). In addition, both countries have performed well in INF. As the common combination of high and upper-middle-income countries, the HCR*MS*BS combination needs relatively more emphasis, as well as INF.

Fig. 10
figure 10

Common path to high NIC.

Full size image

Theoretical implications

First, this study improves the transparency and replicability of innovation paths in different economies by analyzing the time effects of NIS elements on the necessity and combinatorial effects of NIC and broadens the perspective of NIS-related theoretical research. This study finds that no single element is a necessary condition for high NIC and that the necessity consistency of INF for NIC increases over time. In other words, the importance of INF for NIC in high- and upper middle-income economies is gradually increasing. In addition, four configurations can generate high NIC for high-income economies, while two configurations can do the same for upper middle-income economies. The combination of HCR, MS, and BS was found to generate high NIC in both types of economies. Previous studies have shown that the combination of INS, HCR, INF, and BS can generate high NIC (Khedhaouria and Thurik, 2017). This study further expands the timeliness of the impact of this configuration: from 2011 to 2022, it can steadily drive some high-income economies to achieve high NIC. Overall, this study characterizes the interdependence and evolutionary path of NIS elements, providing reference for different economies to optimize resource allocation worldwide.

Second, this study makes another important contribution regarding research methods. Previous studies have only analyzed configurations that generate high GII within a given year (Crespo and Crespo, 2016; Khedhaouria and Thurik, 2017; Huarng and Yu, 2022). Limited by Cross-sectional data, the evolution trend of innovation paths in different economies has not been further revealed. Since the global innovation landscape is constantly evolving, researchers need to investigate the impact of time on fsQCA results. This study builds on innovative attempts to incorporate time effects into the fsQCA method (Garcia-Castro and Ariño, 2016; Guedes et al., 2016; Beynon et al., 2020), thereby revealing a stable configuration of NIS elements that generate consistently high NIC. As an extension of previous studies on GII and NIC, this study emphasizes the time effect of INF’s necessity consistency and that there is a path for generating high NIC in both high- and upper middle-income economies. In summary, this study takes an important step in advancing research methods, which provides a robust and universal interpretation of global innovation trends.

Management implications

This study has practical implications for different economies. First, the increase in INF’s necessity consistency for high NIC indicates the rising importance of this element. High- and upper middle-income economies should further strengthen infrastructure construction; optimize infrastructure layout, structure, functions, and development models, and promote further improvements in innovation capabilities.

Second, the results show that the HCR*MS*BS configuration can promote the generation of high NIC in both types of economies. Therefore, managers should pay more attention to the interaction mechanisms among these three factors. In particular, it is important to improve the education system and provide talent resources to activate innovation. In addition, diversification of the trade market should be promoted by optimizing the credit system and venture capital projects. Similarly, managers should prioritize the role of knowledge-based BS in improving NIC, and promote knowledge absorption and creation by increasing connections between industry, academia, and research while also optimizing research and development efficiency.

Finally, the results indicate that no single NIS element can generate high NIC. Four configurations can drive high-income economies to generate high NIC, while two configurations can do the same in upper middle-income economies. Therefore, different economies should focus on the combined effect of NIS elements and the interaction mechanisms between them. To efficiently improve NIC, government decision-makers should choose the paths best suited for their current development level.

Limitations and future study directions

This study has some limitations. First, we only analyzed the configurations that generated high NICs from 2011 to 2022 and did not consider the configurations that generated low NIC. The latter is also worth studying to alert countries with lagging innovation. Accordingly, future research could further analyze the configurations that cause low NIC, and provide a more comprehensive interpretation of GII Reports and the global innovation environment.

Second, although this study used fsQCA with panel data to analyze NIC, and addressed the lack of time effects in previous studies employing fsQCA, we did not analyze the order of the role of the elements. Future research could combine other methods and tools to conduct a more detailed analysis of this complex problem.

Third, this study uses innovation inputs from the GII Report as NIS elements. Future research could further expand the analytical perspective by incorporating other possible key elements, thereby increasing understanding of how the NIS influences NIC.

Conclusion

This study explores the dynamic impact on NIC of the five NIS elements: institutions, human capital and research, infrastructure, market sophistication, and business sophistication. The fsQCA results reveal four stable configurations that can generate high NIC in high-income economies, and two stable configurations producing the same outcome in upper middle-income economies. One path (HCR*MS*BS) can continuously drive both types of economy to achieve high NIC. The panel data results show a time effect on the necessity consistency of INF for high NIC, indicating that it plays an increasingly important role. This study depicts the evolution trend of the global innovation landscape, improves the transparency and replicability of innovation paths for different economies, and provides robust data support for adjusting innovation strategies for each economy. Therefore, further developing NIS to enhance NIC is a key measure for achieving sustainable development in different income economies. The research results will help different economies adjust innovation policies, optimize the allocation of innovation factors, and improve innovation capability. Overall, the research findings greatly contribute to improving understanding of this exploratory topic.