Analyzing the configuration of the National Innovation System for Innovation Capability: evidence from Global Innovation Index reports

The study examines how elements of the National Innovation System (NIS) shape national innovation capability (NIC) over time and across economies. While prior research often assessed static, cross-sectional effects of individual NIS elements, it remains unclear whether these factors exert stable, enduring impacts across countries and years. Grounded in the NIS framework, the paper argues that interactions among institutions, human capital and research (HCR), infrastructure (INF), market sophistication (MS), and business sophistication (BS) are central to innovation outcomes. The authors articulate three research questions: (RQ1) Do the five NIS elements individually affect NIC, and how do combinations of factors affect NIC in 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 economy groups to high NIC? The study focuses on high- and upper middle-income economies to compare innovation paths, inform policy learning, and assess temporal stability of NIS configurations from 2011 to 2022.

Using the NIS as the theoretical lens, the paper reviews how interactions among national institutions, organizations, and policies co-create, store, and diffuse knowledge and technology, shaping NIC. Key insights include the roles of intangible investments (education, R&D, infrastructure, business connections) and government policy coordination; and how globalization enables knowledge mobility and catch-up opportunities. NIC is defined as a country’s ability to mobilize actors and institutions to engage in innovation and drive economic development. Given heterogeneity in national contexts, there is no consensus on the exact set of NIS components, and time dynamics matter for allocation of innovation resources. Adopting the Global Innovation Index (GII) as an operational framework, the study uses five innovation inputs as NIS elements: Institutions (INS), Human capital and research (HCR), Infrastructure (INF), Market sophistication (MS), and Business sophistication (BS). The paper details each element’s theorized contribution to NIC and proposes five propositions that the presence of each element (INS, HCR, INF, MS, BS) leads to high NIC.

Design: The study applies fuzzy set qualitative comparative analysis (fsQCA) to identify configurational paths of NIS elements leading to high NIC and to assess their temporal stability using panel data techniques. Data: GII Reports 2011–2022. Sample comprises 44 high-income and 31 upper middle-income economies (classified following World Bank income group trajectories). NIC is the outcome; INS, HCR, INF, MS, and BS are antecedent conditions (all from GII innovation input pillars). Calibration: Direct calibration with three anchors at 0.95, 0.50, and 0.05 membership thresholds (Ragin, 2008). Separate anchor values for high-income and upper middle-income groups are reported (e.g., for high-income: INS 93.465/80.300/60.470; HCR 64.800/49.800/31.070; INF 65.230/54.550/37.870; MS 75.900/53.700/37.470; BS 63.630/46.100/24.700; NIC 62.043/49.400/33.470; analogous anchors for upper middle-income are provided). Analysis steps: (1) Necessity analysis using 0.90 consistency threshold; (2) Sufficiency analysis via truth tables with thresholds: consistency 0.80, PRI 0.75, case frequency 2; identification of core vs peripheral conditions per reduced/core solutions (Fiss, 2011). (3) Panel fsQCA following Beynon et al. (2020): pooled consistency (POCONS) and coverage, between consistency distance (BECONS) and adjusted distance for time effects, within consistency distance (WICONS) and adjusted distance for cross-sectional heterogeneity. Software: fsQCA software (noted; version not specified).

- Necessity analysis: No single NIS element is a necessary condition for high NIC in either group. However, infrastructure (INF) exhibits a temporal pattern in necessity consistency, increasing annually across 2011–2022. - Sufficiency/configurations: 2011–2022 results identify six sufficient configurations for high NIC: High-income economies (four paths): H1: INS • HCR • MS (core) [INF, BS irrelevant]; consistency 0.964; raw coverage 0.705. H2: INS • MS • BS (core) [HCR, INF irrelevant]; consistency 0.973; raw coverage 0.707. H3: HCR • MS • BS (core) [INS, INF irrelevant]; consistency 0.979; raw coverage 0.708 (highest coverage in this group). H4: INS • HCR • INF • BS (core) [MS irrelevant]; consistency 0.977; raw coverage 0.625. Upper middle-income economies (two paths): H5: INS • INF • BS (core) [HCR, MS irrelevant]; consistency 0.953; raw coverage 0.483. H6: HCR • MS • BS (core) [INS, INF irrelevant]; consistency 0.961; raw coverage 0.545 (highest coverage in this group). - Common configuration: H3 and H6 are identical (HCR • MS • BS) and robustly associated with high NIC in both economy groups. - Panel stability: POCONS > 0.80 for all six configurations; BECONS adjusted distances < 0.20 (no significant time effect on configurations’ consistency); WICONS adjusted distances < 0.20 (no significant cross-sectional heterogeneity). By contrast, INF shows a notable upward trend in necessity consistency (BECONS adjusted distance > 0.20 for necessity test), indicating increasing importance over time. - Illustrative cases: Switzerland (high-income) and China (upper middle-income) exemplify the HCR•MS•BS configuration, aligning with their strong GII performances (e.g., Switzerland’s high ranks in HCR, MS, BS; China’s sustained R&D investment, large market scale, and strengthening of IP regimes).

The findings address RQ1 by demonstrating that while no single NIS element is necessary for high NIC, multiple alternative configurations of elements are sufficient, and these differ by income group. High-income economies exhibit four viable paths, reflecting diversified routes to high NIC; upper middle-income economies exhibit two paths, reflecting more constrained but effective routes. For RQ2, panel fsQCA shows that these paths are temporally stable and do not exhibit significant cross-sectional heterogeneity; however, the necessity consistency of INF increases over time, implying that infrastructure’s role in enabling knowledge transfer and innovation diffusion is becoming progressively more critical. For RQ3, the study identifies a shared, stable configuration—HCR•MS•BS—effective across both economy groups. This cross-group path underscores the centrality of investing in human capital and R&D, deepening markets (credit, investment, trade, market scale), and strengthening business sophistication (knowledge workers, linkages, absorption) to sustain high NIC. The results reinforce the systemic nature of innovation and the importance of complementarities among NIS elements, offering transparent, replicable innovation paths that policymakers can tailor to developmental stages.

The study maps dynamic, configurational pathways through which NIS elements drive NIC across 2011–2022. Four stable configurations lead to high NIC in high-income economies and two in upper middle-income economies, with HCR•MS•BS forming a common, enduring path for both groups. Panel evidence indicates rising necessity consistency of INF over time, highlighting its growing strategic importance in innovation ecosystems. These insights improve transparency and replicability of innovation paths and support evidence-based policy to optimize NIS resource allocation for sustainable development and competitiveness. Future work should analyze low-NIC configurations, sequence effects among elements, and broaden the NIS construct beyond GII inputs.

- Scope of outcomes: The analysis focuses on configurations leading to high NIC (2011–2022) and does not examine configurations associated with low NIC; studying the latter could inform lagging countries and provide a fuller picture. - Temporal ordering: While incorporating time effects into fsQCA, the study does not establish causal sequencing among NIS elements; complementary methods are needed to unpack temporal order and causal mechanisms. - Construct coverage: NIS is operationalized via GII input pillars; future research could extend the framework by incorporating additional elements and contextual factors to enrich understanding of how NIS influences NIC.