Regional Integration, University Resources, and Firm Performance: Evidence from the Yangtze River Delta in China

Abstract

Universities play a critical role in knowledge creation and technological innovation, serving as key drivers of regional development. However, existing research has paid limited attention to the mechanisms through which university innovation inputs translate into firm-level performance, particularly in the context of science and technology corridors in emerging economies. This study investigates how university innovation resources affect enterprise performance in the G60 Science and Technology Corridor within China’s Yangtze River Delta, one of the country’s most dynamic innovation regions. Using a panel dataset of 55 universities across nine cities from 2008 to 2017, we employ spatial analysis and fixed-effects panel regression models to examine the relationship between university innovation inputs and firm performance and further explore the mediating roles of local human capital and firm R&D investment. The results show that university innovation inputs significantly enhance enterprise performance, although excessive human resource inputs exhibit a negative effect on both short-term and long-term outcomes. Local human capital and firm R&D investment serve as key mediating mechanisms, with input and output resources influencing enterprise performance through distinct pathways. Heterogeneity analysis reveals that non-state-owned enterprises and small- and medium-sized enterprises derive greater long-term benefits from university resources. These findings contribute to the literature by clarifying the conceptual distinction between university innovation inputs and outputs, and by demonstrating the micro-level mechanisms—R&D investment and human capital—through which university-generated knowledge affects firm performance. The results also provide empirical evidence from an emerging economic context, extending the applicability of knowledge spillover and absorptive capacity theories. Policy implications include optimizing university human resource allocation, strengthening university–enterprise collaboration, and providing targeted support for non-state-owned enterprises and SMEs. Future research may extend the analysis to include institutional factors and university heterogeneity.

1. Introduction

The intersection of globalization and the knowledge economy has spurred considerable academic attention on the role of universities and enterprises in regional innovation systems (Couchman et al., 2008). Central to this discourse is the “triple helix” model, which emphasizes the crucial role of universities in fostering collaborative innovation alongside industry and government (Etzkowitz & Leydesdorff, 1997). This model portrays universities as pivotal drivers of economic innovation, contributing not only to knowledge generation but also to the formation of strategic partnerships that facilitate technological advancement (Henderson, 2023). In parallel, regional innovation systems (RIS) theory underscores the importance of innovation as a driver of regional development, competitiveness, and economic growth (Leydesdorff & Strand, 2013; Cantner et al., 2019). This dual framework—university-driven innovation and regional economic growth—has become a focal point in understanding how these two spheres interact. Early contributions to innovation system thinking, such as Lundvall (1985), emphasized user-producer interactions as central to innovation processes, laying groundwork for later regional extensions (Lawson, 1999).

While the growing importance of knowledge accumulation in regional development is widely acknowledged (X. Zheng et al., 2025), existing studies tend to focus on the broader economic outcomes of university–industry collaborations, often overlooking the specific mechanisms that drive firm-level performance and the impact of university innovation resources on enterprises. Exemplary science and technology corridors such as Silicon Valley, Boston’s Route 128, and the UK’s M4 Motorway provide critical insights into how such collaborations foster technological and economic progress (Porter, 1998; Cooke, 2002; Mothe & Mallory, 2006). However, most of the literature on such corridors is focused on developed economies, leaving a gap in understanding the role of universities and innovation systems in developing countries. Recent scholarship has begun to address this gap. Rudskaya et al. (2022) map regional efficiency in transition economies, while Li et al. (2024) utilize Natural Language Processing (NLP) to demonstrate how university–industry collaboration specifically mitigates private R&D shortfalls in developing sectors. More broadly, Batterbury and Hill (2005) examined how higher education institutions contribute to regional development, highlighting the importance of understanding these dynamics across different institutional contexts.

This study seeks to fill these gaps by focusing on China’s Yangtze River Delta (YRD) region, an area marked by rapid urbanization, significant innovation capacity, and government-driven regional integration strategies. The YRD region is home to 413 universities and 61,000 high-tech enterprises (Zhang et al., 2021), providing a robust context for exploring the synergies between higher education institutions and enterprises. The development of the G60 Science and Technology Corridor, which spans nine cities along the G60 expressway, presents a unique opportunity to investigate the interactions between universities, research institutions, and enterprises in a dynamic RIS. The G60 Corridor exemplifies the potential for university-driven innovation to foster local economic growth and technological development in emerging regions, particularly as dynamic efficiency in such ecosystems often follows complex “M-shaped” growth trajectories (Z. Zheng et al., 2026).

The existing literature has largely overlooked the nuanced relationship between university resources and enterprise performance, particularly in terms of long-term and short-term dynamics. While universities are commonly regarded as key contributors to regional development, the mechanisms by which their resources—both input and output—translate into measurable firm outcomes remain underexplored. Studies by Liu et al. (2021) and Hou et al. (2021) point to the substantial contributions that universities make to fostering innovation within enterprises. However, these studies have often focused on macroeconomic impacts, leaving the firm-level dynamics largely unexamined. Additionally, the impact of enterprise heterogeneity—such as ownership structures and firm sizes—on the relationship between university resources and enterprise performance has yet to be thoroughly explored.

Borah et al. (2023) argue that universities bear broader innovation responsibilities compared to government or industry, as they not only foster the creation of new knowledge but also facilitate the commercialization and application of that knowledge through multi-helix integration. This positions universities as crucial actors in regional innovation systems, whose contributions go beyond merely conducting research, but also extend to nurturing the innovation ecosystem through collaboration with industry stakeholders. Using Chinese data, Ma (2024) highlights that higher education systems drive innovation and growth, with implications for other countries where the effectiveness of such systems can also be significantly enhanced by high levels of trade openness. In a related study, P. Yang and Liu (2024) examine the expansion of tertiary education in China, highlighting how socioeconomic and politico-institutional factors drive the adoption of vocational colleges, with significant regional and temporal variations shaping the structural diversity of the sector.

This paper seeks to address these gaps by examining the effects of university innovation resources on enterprise performance in the G60 Corridor, with a focus on both short-term and long-term outcomes. By investigating the heterogeneous effects of university–industry collaboration across different types of enterprises, this research contributes to a deeper understanding of how universities’ roles in innovation systems influence local economic development and firm performance (Shang et al., 2021).

The primary objective of this study is to investigate how university innovation inputs affect enterprise performance within the G60 Science and Technology Corridor, with a focus on the mediating mechanisms of local human capital and firm R&D investment, as well as the heterogeneous effects across different enterprise types. The contribution of this study lies in its novel application of spatial data analysis using ArcGIS 8.2.2 and quantitative methods with Stata to explore how university resources affect enterprise performance within a regional innovation framework. This approach provides empirical evidence on the mechanisms of university–industry collaboration, especially within the context of developing countries. It also highlights the impact of different enterprise characteristics, such as ownership and size, on the effectiveness of university-driven innovation initiatives. Through this analysis, the study deepens our understanding of the intricate relationship between universities and enterprises in regional development, thus enhancing the theoretical and empirical foundations of innovation systems research.

2. Literature Review and Hypotheses

2.1. Theoretical Foundations

This study draws on the Triple Helix framework, regional innovation systems (RIS), and innovation production theory to conceptualize how university innovation inputs drive firm performance in the G60 Science and Technology Corridor.

The Triple Helix model emphasizes dynamic interactions among universities, industry, and government in fostering innovation (Etzkowitz, 2003). Universities serve not only as knowledge producers but also as active participants in knowledge commercialization and diffusion. Complementing this perspective, regional innovation systems theory highlights that innovation emerges from region-specific institutional arrangements through which universities, firms, and research institutions engage in knowledge exchange and collective learning processes (Lundvall, 1985; Cooke, 1992; Lawson, 1999).

Innovation production theory frames innovation as a process whereby inputs—such as research funding and personnel—are transformed into outputs, such as patents and technology transfers (Griliches, 1979). This study strictly distinguishes university innovation resources as inputs, while treating outputs as outcomes rather than resources, thereby maintaining conceptual clarity. Recent research further reinforces the foundational role of human capital and R&D intensity in knowledge creation and diffusion (Coutinho & Au-Yong-Oliveira, 2024; Mohajan & Zhao, 2026).

Together, these theoretical perspectives provide a coherent foundation for understanding how university innovation inputs influence firm performance within regional innovation systems.

2.2. University Innovation Inputs and Firm Performance

Knowledge spillover theory provides the primary explanation for how university innovation inputs enhance firm performance (Audretsch et al., 2012). University–industry collaboration serves as the key channel through which such spillovers occur, with firms engaging universities through joint research, technology transfer, and open innovation practices (Gustina et al., 2024; Cohen et al., 2025).

In this study, firm performance constitutes a multidimensional construct capturing both short-term and long-term outcomes. Short-term performance is measured by accounting-based indicators such as profitability, while long-term performance is captured by market-based measures such as Tobin’s Q. This dual approach reflects both current operational efficiency and future growth potential (Saharti, 2025; López-Paredes et al., 2025).

University innovation inputs, particularly research funding and personnel, strengthen the regional knowledge base and enhance firms’ access to external knowledge. Empirical evidence confirms that greater R&D intensity and knowledge accessibility significantly improve firm performance across diverse institutional contexts (Coutinho & Au-Yong-Oliveira, 2024; Mohajan & Zhao, 2026). Although this study focuses on the Chinese context, the underlying mechanisms of knowledge spillovers and university–industry interactions remain broadly applicable.

H1. 

University innovation inputs (measured by research personnel and research funding) have a positive effect on firm performance (measured by short-term accounting performance and long-term market performance) of regional enterprises.

The positive relationship between university knowledge spillovers and firm outcomes is well-documented across different contexts. Feldman (1999) provided early evidence that knowledge spillovers are geographically localized and disproportionately benefit proximate firms. More recently, He and Luo (2025) demonstrated significant regional disparities in the economic impact of industry-university-research collaboration in China, highlighting the importance of local absorptive conditions (Hong & Su, 2013).

2.3. Transmission Mechanisms

The effect of university innovation inputs on firm performance operates indirectly through firms’ internal mechanisms rather than through direct channels. Absorptive capacity theory posits that firms must recognize, assimilate, and apply external knowledge to derive benefits from it.

First, R&D investment represents a primary transmission channel. Exposure to university-generated knowledge reduces technological uncertainty and creates new innovation opportunities, thereby encouraging firms to increase their R&D expenditures (Laursen & Salter, 2006). Empirical studies confirm that R&D intensity plays a crucial role in enhancing firm performance (Saharti, 2025; Bekata & Kero, 2025).

H2a. 

University innovation inputs enhance firm performance by promoting firms’ R&D investment.

The effectiveness of R&D cooperation, however, may vary depending on firms’ collaboration experience and innovation strategies (Lhuillery & Pfister, 2009).

Second, human capital serves as a critical mechanism for knowledge absorption and innovation. Universities contribute to skilled labor development and facilitate knowledge diffusion, which strengthens firms’ innovation capabilities (Santoro & Chakrabarti, 2002). Recent evidence demonstrates that human capital efficiency significantly improves firm performance and sustainability outcomes (Al Frijat & Elamer, 2025), while knowledge management and open innovation practices further reinforce the relationship between knowledge and performance (Martínez-Falcó et al., 2026; Leitner, 2011; Schmiedeberg, 2008).

H2b. 

University innovation inputs improve firm performance by facilitating human capital accumulation within firms.

2.4. Heterogeneous Effects

2.4.1. Ownership Structure

The impact of university innovation inputs on firm performance varies across firms with different ownership characteristics. In the Chinese context, state-owned enterprises (SOEs)—defined as firms ultimately controlled by the state through ownership or control rights—typically maintain closer institutional connections with public universities, which are predominantly state-funded. This institutional proximity facilitates knowledge exchange and collaboration. Moreover, SOEs generally possess stronger financial resources and greater tolerance for long-term innovation risks (Ortiz & Gargallo-Castel, 2025; Lei et al., 2012; Lin & Milhaupt, 2013).

H3a. 

Ownership structure (state-owned vs. non-state-owned enterprises) moderates the relationship between university innovation inputs and firm performance.

2.4.2. Firm Size

Firm size constitutes another important source of heterogeneity. Large firms, defined as those with total assets above the sample median, typically possess superior financial resources, established R&D infrastructure, and stronger absorptive capacity. These advantages enable large firms to more effectively transform university-generated knowledge into economic performance. In contrast, small- and medium-sized enterprises (SMEs), despite resource constraints, may benefit from greater organizational flexibility and responsiveness to technological opportunities (López-Paredes et al., 2025; Maolani et al., 2026; Nieto & Santamaría, 2010; Kalsaas, 2013).

H3b. 

Firm size moderates the relationship between university innovation inputs and firm performance.

2.5. Research Gaps and Contributions

Despite extensive research on university–industry collaboration and knowledge spillovers, several critical gaps persist in the literature.

First, existing studies frequently fail to distinguish clearly between university innovation inputs and outputs when conceptualizing university resources. This conflation creates conceptual ambiguity and obscures universities’ specific roles in the innovation process. Second, prior research has primarily examined direct relationships between universities and firm performance, devoting limited attention to the mediating mechanisms through which university innovation inputs influence firms. The roles of R&D investment and human capital in this transmission process remain underexplored. Third, much of the existing literature derives from developed economy contexts, where institutional arrangements and university–industry collaboration patterns differ substantially from those in emerging economies such as China (Lee, 2011; Xu et al., 2021; Agasisti & Belfield, 2017; Guccio et al., 2016).

This paper addresses these gaps through three primary contributions. First, it establishes a clear conceptual distinction by defining university innovation resources strictly as inputs. Second, it examines the mechanisms—specifically R&D investment and human capital—through which university inputs affect firm performance. Third, by providing empirical evidence from China’s G60 Science and Technology Corridor within the Yangtze River Delta, this study tests the applicability of established theories in an emerging economy context.

3. Data Description

3.1. Sampling Regions

The Yangtze River Delta (YRD) is recognized as one of China’s most economically dynamic regions, distinguished by its robust development, openness, and innovation capacity. Spanning Shanghai, Jiangsu, Zhejiang, and Anhui provinces, the YRD covers a vast area of 358,000 square kilometers, incorporating 27 cities across 225,000 square kilometers. The integration of this region has become a critical component of China’s broader economic strategy.

Since 2007, spatial planning in the YRD has evolved from a “point-axis” approach to a model focused on coordinated development, with major cities positioned as central hubs. As part of the G60 Expressway Development Plan, nine cities—including Shanghai Songjiang, Jiaxing, Hangzhou, Jinhua, Huzhou, Suzhou, Xuancheng, Wuhu, and Hefei—are collaborating to establish the G60 Corridor. This science and technology corridor spans 76,200 square kilometers and serves a permanent population of 49 million within the YRD. As a key driver of regional integration, the G60 Corridor plays an essential role in China’s national integration strategy. The directive to “Accelerate the establishment of the G60 Corridor in the YRD” is explicitly incorporated into the Outline of the 14th Five-Year Plan and 2035 Vision of China (see Figure 1a,b).

Several innovation-related elements have been concentrated in the YRD’s G60 Corridor. According to the 2021 China Statistical Yearbook, the GDP of the nine cities within the G60 Corridor reached 7.55 trillion RMB, accounting for one-fifteenth of China’s total GDP. These cities contribute one-twelfth of national public budget revenues and one-tenth of the country’s high-tech company output. Data from the Ministry of Education of China indicates that, as of 2021, the YRD is home to 495 universities, representing 17% of China’s higher education institutions, with 55 high-quality universities located within the G60 Corridor.

The G60 Corridor, a product of over 40 years of China’s reform and opening-up policies, has become a pioneering example of science and technology corridor planning in the country. The development of this corridor provides significant insights not only for China but also for other developing nations, particularly with regard to spatial planning.

3.2. Control Variables

This study utilizes two performance indicators to assess the regional enterprises: (i) Earnings per Share (EPS), which is calculated as a company’s profit divided by its outstanding common shares, serving as a direct indicator of short-term profitability; and (ii) Tobin’s Q (Tobin), the ratio of a company’s market value to the replacement cost of its assets, which incorporates both future growth prospects and risk adjustment, providing a measure of a company’s potential for future growth and investment (see

Table 1. Descriptive statistics of the key variables for the sample enterprises.

Variable Code	Variable Meaning	Obs	Mean	SD	Min	Median	Max
eps	Earnings per share (CNY/share)	3195	0.415	0.468	−0.545	0.318	2.539
Tobin	Tobin Q	3195	2.177	1.392	0.922	1.728	8.469
Rp	Research project funding (1000 CNY)	3195	14.749	1.774	8.660	15.454	16.176
Sf	senior faculties (person)	3195	8.799	1.308	4.796	9.538	9.879
Sp	Scientific publications (pieces)	3195	9.875	1.510	5.375	10.793	11.094
Tt	Technology transfer contracts (items)	3195	5.438	1.728	0.000	6.082	7.101
lev	Debt-to-assets	3195	22.113	1.440	19.634	21.900	27.121
cashflow	Net cash flows	3195	0.427	0.205	0.053	0.417	0.887
fixed	Fixed assets ratio	3195	0.189	0.352	−0.567	0.118	1.832
growth	Growth rate of operating income	3195	0.192	0.160	0.002	0.150	0.677
size	Enterprise size	3195	0.429	1.258	−0.587	0.109	9.189
age	Enterprise age	3195	2.141	0.827	0.000	2.303	3.219
soe	Enterprise ownership	3195	0.435	0.496	0.000	0.000	1.000
edu	Human capital	3195	10.051	0.866	7.043	10.515	10.680
RD	R&D investment	3195	2.694	4.495	0.000	0.900	48.430

). Tobin’s Q is widely used in corporate finance and innovation studies as a forward-looking performance indicator (B. Yang & Gan, 2021).

Explanatory variables include the innovation resources available at regional universities, categorized into input and output resources (see Table 1). Innovation resource input focuses on financial and human capital, with indicators such as research project funding (Rp) and senior faculty members (Sf). Innovation resource output emphasizes research outcomes and technology transfer, incorporating measures such as scientific publications (Sp) and technology transfer contracts (Tt).

To control for estimation bias due to omitted variables, we include two categories of control variables, as outlined in Table 1. The first category comprises financial indicators directly related to enterprise performance, value, and growth, such as the debt-to-assets ratio (lev), net cash flow (cashflow), fixed assets ratio (fixed), and the growth rate of operating income (growth). The second category consists of enterprise profile variables, including enterprise size (size), age (age), and ownership structure (soe).

Enterprise profiles have frequently been incorporated as control variables in empirical studies, as their relationship with enterprise performance is less direct than that of operating conditions and performance. Enterprise size, for instance, is a contested variable. While some studies suggest that larger enterprises exhibit superior performance (Blundell et al., 1999), others argue for a negative relationship (Hoppe & Lee, 2003). Aghion et al. (2005) propose an “inverted U” relationship, where innovation efficiency first increases and then decreases with enterprise size. Given its critical role in firm performance, size is included as a control variable in this study.

Similarly, enterprise age is considered a control variable, acknowledging that younger firms often face resource constraints and skepticism from suppliers and customers. In contrast, more mature firms, with accumulated experience and knowledge, tend to outperform younger counterparts (García-Quevedo et al., 2014). Furthermore, considering the increasing prominence of state-owned enterprises (SOEs) in China over the past decade (Tan et al., 2015), ownership type is also included as a control variable, reflecting the significant role that SOEs play in China’s economy (Ervits, 2023).

3.3. Data Sources

The data for explanatory variables is primarily derived from the Compilation of Scientific and Technological Statistics of Universities, published by the Ministry of Education of China and accessed through the ESP data platform. This dataset pertains exclusively to public universities, excluding vocational and private institutions. However, it includes “985” and “211” project universities, which are central to China’s research and innovation activities. Therefore, the data on innovation resources in universities provides a representative snapshot of the region. The sample includes 55 universities located across nine cities in the G60 Corridor between 2008 and 2017, namely Shanghai (17), Jiaxing (1), Hangzhou (11), Jinhua (1), Suzhou (4), Huzhou (1), Xuancheng (0), Wuhu (5), and Hefei (15). Data on the explained and control variables is sourced from listed companies in the G60 Corridor during this period, obtained from the Guotaian database. Following the exclusion of incomplete data, descriptive statistics for the dataset are presented in Table 1.

3.4. Methodology

Building upon existing literature, we hypothesize that enterprise performance is intricately linked not only to internal resource inputs but also to external innovation activities, particularly the spillover effects generated by university innovation resources. To examine this relationship, we apply the Griliches–Jaffe knowledge production function within the classical knowledge spillover model, which is formalized in the conceptual framework:

$$P=f\left(KF,Z\right)$$

(1)

Equation (1) reflects the conceptual relationship between enterprise performance and its influencing factors, where $P$ stands for enterprise performance. We use earning per share and Tobin’s Q to measure $P$. $KF$ (key factor) is the primary variable, which includes university innovation resource input $\left(UI\right)$ and university innovation resource output $\left(UP\right)$. Research project funding and senior faculties are used to measure $UI$, and scientific publications and technology transfer contracts are considered to evaluate $UP$. $Z$ represents other factors that affect enterprise performance, including debt-to-assets (lev), net cash flows (cashflow), fixed assets ratio (fixed), growth rate of operating income (growth), company size (size), company age (age) and company ownership (soe).

Based on this conceptual framework, we specify the following econometric models. To account for potential time lags between explanatory and dependent variables, the baseline empirical specification takes the form:

$${EPS}_{st}=C+\alpha {UI}_{st}+\beta {UP}_{st}+\gamma Z_{st}+{\mu }_{st}$$

(2)

$${TobinQ}_{st}=C+\alpha {UI}_{st}+\beta {UP}_{st}+\gamma Z_{st}+{\mu }_{st}$$

(3)

${\mathrm{E}\mathrm{P}\mathrm{S}}_{\mathrm{s}\mathrm{t}}$ represents the earnings per share in year t of listed enterprises in the G60 Corridor. ${\mathrm{T}\mathrm{o}\mathrm{b}\mathrm{i}\mathrm{n}\mathrm{Q}}_{\mathrm{s}\mathrm{t}}$ is the Tobin’s Q in year t of listed enterprises in the G60 Corridor. ${\mathrm{U}\mathrm{I}}_{\mathrm{s}\mathrm{t}}$ is the research project fundings and senior faculties. ${\mathrm{U}\mathrm{P}}_{\mathrm{s}\mathrm{t}}$ represents the scientific publications and technology transfer contracts. ${\mathrm{Z}}_{\mathrm{s}\mathrm{t}}$ is a series of control variables that affect enterprise performance, and ${\mathsf{\mu }}_{\mathrm{s}\mathrm{t}}$ is the usual error term.

To address scale differences across variables, we apply natural logarithmic transformations to strictly positive, continuous variables—including research project funding ($\mathit{Rp}$), senior faculties ($\mathit{Sf}$), scientific publications ($\mathit{Sp}$), technology transfer contracts ($\mathit{Tt}$), firm size ($\mathit{size}$), firm age ($\mathit{age}$), and R&D investment ($\mathit{RD}$). Variables expressed as ratios or containing non-positive values—namely earnings per share ($\mathit{eps}$), Tobin’s Q ($\mathit{Tobin}$), debt-to-assets ratio ($\mathit{lev}$), net cash flow ratio ($\mathit{cashflow}$), fixed assets ratio ($\mathit{fixed}$), and operating income growth rate ($\mathit{growth}$)—are retained in their original form, as logarithmic transformation is not applicable. The final specification thus incorporates these transformations appropriately, with L denoting the natural logarithm for transformed variables only.

$${EPS}_{st}=C+\alpha {LUI}_{st}+\beta {LUP}_{st}+\gamma Z_{st}+{\mu }_{st}$$

(4)

$${TobinQ}_{st}=C+\alpha L{UI}_{st}+\beta {LUP}_{st}+\gamma Z_{st}+{\mu }_{st}$$

(5)

where $Z_{st}$ includes the control variables (lev, cashflow, fixed, growth, size, age, soe), with size and age also log-transformed, and the remaining control variables entered in their original form.

To comprehensively address our research questions, we employ a multi-method approach that combines spatial analysis, panel data econometrics, mediation analysis, and heterogeneity analysis. Spatial mapping using ArcGIS is first conducted to visually capture the distribution patterns of university innovation resources and enterprise performance across the G60 Corridor, providing an intuitive foundation for subsequent quantitative analysis. The fixed-effects panel regression is then applied to estimate the impact of university innovation resources on enterprise performance, as it effectively controls for time-invariant unobserved heterogeneity at the firm level and accounts for common temporal shocks through year fixed effects, thereby strengthening causal inference. To uncover the underlying mechanisms, we adopt a two-stage mediation analysis that traces the pathways through local human capital and enterprise R&D investment, following established practices in the literature. Finally, heterogeneity analysis by ownership and firm size allows us to examine whether the effects vary across different enterprise types, yielding policy-relevant insights.

Nevertheless, each method has inherent limitations. The fixed-effects model mitigates but does not eliminate concerns about reverse causality and time-varying omitted variables; thus, our findings should be interpreted as robust associations rather than definitive causal claims. Florio et al. (2016) demonstrated the value of comprehensive impact assessment frameworks for large-scale research infrastructures, an approach that could inform future evaluations of science and technology corridors such as the G60. The mediation analysis relies on the assumption that mediators are exogenous conditional on controls, and it does not fully capture dynamic feedback loops. The spatial analysis is descriptive and lacks formal statistical tests for spatial dependence. Heterogeneity analysis based on sample splitting may suffer from reduced statistical power in smaller subsamples, and we do not formally test coefficient differences across groups. Additionally, our sample is restricted to listed enterprises, which may limit generalizability to unlisted SMEs. These limitations point to avenues for future research, such as employing instrumental variables, spatial econometric models, and broader enterprise samples.

4. Estimation Results

4.1. Spatial Distribution of University Innovation Resources and Enterprise Performance in the G60 Corridor

To investigate the relationship between university innovation resources and regional enterprise performance in the G60 Corridor, spatial distribution maps were generated using ArcGIS software (version 8.2.2.). These maps illustrate the geographical distribution of both university innovation resources and enterprise performance. Earnings per share (EPS) and Tobin’s Q were divided into seven categories based on the natural breakpoint method. This facilitated the creation of spatial distribution maps reflecting the number of listed companies, EPS, and Tobin’s Q across cities in the G60 Corridor.

As depicted in Figure 2, provincial capitals and municipalities host the highest concentrations of listed companies. Additionally, EPS and Tobin’s Q exhibit a geographically uneven distribution among the nine cities, with stronger performance trends in the eastern region compared to the central region.

Similarly, spatial distribution maps of university innovation resources in the G60 Corridor (Figure 3) reveal a concentration in municipalities and provincial capitals. Xuancheng in Anhui Province, lacking a single public university, forms a “central depression belt.” Among the prefecture-level cities, Suzhou in Jiangsu Province emerges as the leader in terms of innovation resources.

4.2. Baseline Regression

The Hausman and F tests were conducted to determine the appropriate econometric model for data analysis. As reported in

Table 2. Hausman test and F-test results.

	Hausman Test			F Test
	Chi2 Statistics	p-Value	Conclusion	Chi2 Statistics	p-Value	Conclusion
Equation (1)	58.141	0.000	Reject	6.37	0.000	Reject
Equation (2)	190.481	0.000	Reject	4.91	0.000	Reject

, the results of both tests reject the null hypothesis, indicating that a fixed-effects model is most suitable for the regression analysis.

Benchmark regression results based on Equations (4) and (5) are presented in

Table 3. Benchmark Regression Results.

Variables	EPS (Equations (4))	Tobin (Equations (5))
Rp	0.1202 ***	0.1983 **
	(2.78)	(2.09)
Sf	−0.3369 ***	−0.5003 **
	(−5.43)	(−2.35)
Sp	0.1593 ***	−0.0600
	(2.94)	(−0.98)
Tt	0.0407 ***	0.0592 ***
	(4.35)	(4.79)
size	0.1934 ***	−0.8072 ***
	(16.22)	(−13.28)
lev	−0.1467 **	−0.0080
	(−2.02)	(−0.04)
cashflow	0.2376 ***	0.1370 ***
	(14.10)	(2.82)
fixed	−0.5363 ***	−0.2696
	(−13.98)	(−1.53)
growth	0.0145 ***	−0.0110 *
	(6.02)	(−1.96)
age	−0.2082 ***	0.9464 ***
	(−10.69)	(19.48)
soe	−0.0450 **	−0.7958 ***
	(−2.55)	(−18.01)
Constant	−3.8858 ***	20.0648 ***
	(−13.14)	(8.78)
Observations	3195	3195
R-squared	0.145	0.374
Number of groups	485	485
company	YES	YES
year	YES	YES

Note: t-statistics in parentheses *** p < 0.01, ** p < 0.05, * p < 0.1.

. Controlling for time-fixed effects, individual-fixed effects, and seven control variables, the findings demonstrate that all four indicators of higher education innovation resources significantly influence the short-term performance of regional enterprises in the G60 Corridor. Three indicators—research project funding, senior faculty, and technology transfer contracts—also affect long-term enterprise performance. Notably, senior faculty, representing human resource inputs from universities, exert a negative effect on both short-term and long-term enterprise performance.

4.3. Mechanism Analysis

The analysis above suggests that university innovation resources affect both the short- and long-term performance of regional enterprises. To further explore the mechanisms of influence, this study examines two potential pathways: the external environment and enterprise resources. Hypothesis 2 posits that university innovation resources enhance local human capital and increase R&D investment, thereby influencing enterprise performance. This hypothesis was tested using a two-stage regression approach.

Table 4. Mechanism Analysis Involving the Impact of Innovation Resources on Enterprise Performance.

	(1)	(2)
Variables	Edu	RD
Rp	0.2327 ***	1.0566 ***
	(3.39)	(4.46)
Sf	0.2695 **	0.8859 ***
	(2.21)	(2.82)
Sp	0.0446 ***	−2.0088 ***
	(2.79)	(−4.27)
Tt	0.0022	0.2012 **
	(0.25)	(2.07)
size	0.0010	−0.2745 ***
	(0.95)	(−10.05)
Lev	−0.0169 ***	−3.2426 ***
	(−6.34)	(−5.97)
cashflow	−0.0102 ***	0.8057 **
	(−2.64)	(2.19)
fixed	−0.0094	−2.4522 ***
	(−0.41)	(−6.11)
growth	−0.0006	0.0139
	(−0.91)	(0.53)
age	0.0275 ***	−0.9867 ***
	(6.14)	(−12.30)
soe	0.0087 ***	−0.7625 ***
	(5.39)	(−6.98)
Constant	3.6890 ***	8.9554 ***
	(12.23)	(12.15)
Observations	3195	3195
R-squared	0.586	0.253
Number of groups	485	485
company	YES	YES
year	YES	YES

Note: t-statistics in parentheses *** p < 0.01, ** p < 0.05.

shows that increased input of university innovation resources positively correlates with higher levels of human capital and R&D investment among listed enterprises in the region. The findings highlight that innovation resource output differs from input in terms of influence mechanisms. For instance, university scientific publications, a measure of innovation output, significantly enhance regional human capital, but exhibit a significant negative association with R&D investment, suggesting that increased scientific output may, counterintuitively, be associated with reduced firm-level R&D spending in the short term. Conversely, technology transfer contracts significantly boost R&D investment but have no measurable effect on human capital. These findings suggest nuanced differences in how university innovation resource outputs influence enterprise performance through R&D investment and human capital.

4.4. Heterogeneity Analysis

Enterprise ownership and size are two dimensions along which the impact of university innovation resources varies. Ownership was classified as state-owned (dummy variable = 1) or non-state-owned (dummy variable = 0).

Table 5. Ownership of enterprise performance.

	State-Owned	Non-State Owned	State-Owned	Non-State Owned
Variables	Eps	Eps	Tobin	Tobin
Rp	−0.0410	0.1293 **	0.5331 *	0.4137 ***
	(−0.91)	(2.03)	(1.89)	(4.35)
Sf	−0.3195 **	−0.3024 ***	−0.5403	−0.9964 ***
	(−2.27)	(−10.15)	(−0.98)	(−3.76)
Sp	0.2805 ***	0.0845	0.0700	0.2123 ***
	(4.83)	(1.37)	(0.20)	(4.70)
Tt	0.0539 ***	0.0367 ***	−0.0398	0.0582 **
	(7.63)	(7.24)	(−0.58)	(2.33)
size	0.2605 ***	0.1516 ***	−0.5619 ***	−1.0234 ***
	(10.57)	(20.02)	(−9.64)	(−12.83)
lev	−0.1462	−0.1181 ***	−0.8486 ***	0.5751 ***
	(−0.95)	(−4.33)	(−2.94)	(3.07)
cashflow	0.4394 ***	0.1822 ***	0.2836 ***	0.1343 ***
	(12.29)	(18.98)	(5.00)	(3.25)
fixed	−0.5605 ***	−0.5239 ***	−0.0843	−0.7219 ***
	(−4.26)	(−6.64)	(−1.08)	(−2.87)
growth	0.0160 ***	0.0106 ***	0.0037	−0.0217 *
	(3.66)	(3.43)	(0.80)	(−1.85)
age	−0.0611 ***	−0.2190 ***	0.7584 ***	0.9496 ***
	(−5.45)	(−15.20)	(7.85)	(13.75)
Constant	−4.8588 ***	−2.6888 ***	9.3522 ***	23.0556 ***
	(−5.71)	(−10.13)	(3.35)	(9.20)
Observations	1389	1806	1389	1806
R-squared	0.160	0.153	0.145	0.406
Number of groups	167	336	167	336
company	YES	YES	YES	YES
year	YES	YES	YES	YES

Note: t-statistics in parentheses *** p < 0.01, ** p < 0.05, * p < 0.1.

demonstrates that senior faculty and technology transfer contracts significantly influence the short-term performance of both categories of enterprises. However, non-state-owned enterprises exhibit greater long-term benefits from university innovation resources, with all four indicators significantly contributing to their growth.

Enterprise size was analyzed using median-based classification into large enterprises and small- and medium-sized enterprises (SMEs).

Table 6. The effect of enterprise size on enterprise performance.

	Large-Size	Small and Medium Size	Large-Size	Small and Medium Size
Variables	Eps	Eps	Tobin	Tobin
Rp	0.0850	0.1636 ***	0.0520	0.3176 **
	(1.31)	(2.72)	(0.17)	(2.37)
Sf	−0.4440 ***	−0.2660 *	0.1464	−1.8052 ***
	(−7.64)	(−1.70)	(0.30)	(−9.43)
Sp	0.2879 ***	0.0553	−0.2888	0.2553 **
	(4.37)	(1.17)	(−1.20)	(2.53)
Tt	0.0776 ***	0.0013	−0.0057	0.0763 **
	(5.40)	(0.27)	(−0.09)	(2.51)
size	0.2178 ***	0.1851 ***	−0.3985 ***	−1.1934 ***
	(12.59)	(14.32)	(−4.69)	(−18.04)
lev	−0.3890 ***	0.0164	−0.3849	0.1108
	(−3.80)	(0.22)	(−1.16)	(0.60)
cashflow	0.4246 ***	0.1780 ***	0.3901 ***	0.0804 *
	(13.29)	(6.61)	(6.29)	(1.74)
fixed	−0.7003 ***	−0.4473 ***	−0.0510	−0.3726
	(−8.12)	(−7.54)	(−0.17)	(−1.65)
growth	0.0218 ***	0.0080 **	0.0050	−0.0441 ***
	(7.39)	(2.17)	(0.87)	(−3.95)
age	−0.1866 ***	−0.2367 ***	0.6300 ***	1.0665 ***
	(−9.85)	(−5.70)	(5.92)	(13.26)
soe	−0.0330	−0.0092	−0.6384 ***	−0.7717 ***
	(−1.40)	(−0.60)	(−4.40)	(−6.27)
Constant	−4.4114 ***	−3.7344 ***	10.7562 ***	34.3885 ***
	(−19.58)	(−3.57)	(4.11)	(18.45)
Observations	1597	1598	1597	1598
R-squared	0.145	0.198	0.142	0.500
Number of groups	205	280	205	280
company	YES	YES	YES	YES
year	YES	YES	YES	YES

Note: t-statistics in parentheses *** p < 0.01, ** p < 0.05, * p < 0.1.

shows that three indicators of university innovation resources—senior faculty, scientific publications, and technology transfer contracts—significantly affect the short-term performance of large enterprises. For SMEs, research project funding and senior faculty are the key determinants of short-term performance. In the long term, all four indicators significantly influence SMEs’ Tobin’s Q, whereas no such effect is observed for large enterprises. These results indicate that SMEs are more sensitive to university innovation resources in terms of long-term performance.

In summary, the findings suggest that university innovation resources exert a stronger influence on the long-term performance of non-state-owned enterprises and SMEs. Furthermore, the heterogeneity analysis corroborates the benchmark regression results, highlighting the negative impact of university human resource inputs on enterprise performance.

5. Discussion

This study quantitatively examines the relationship between university innovation resources and the performance of regional listed enterprises. The findings reveal that most university innovation resources significantly enhance regional enterprise performance, aligning with prior studies (Hou et al., 2019; Wang et al., 2022). Notably, an important and unexpected result is that human resource inputs in universities negatively impact both short-term and long-term enterprise performance. This contradicts Østergaard’s (2009) finding that university linkages facilitate knowledge flows and contribute to industrial development. A plausible explanation is the limited mobility of academics and scientists within the region, which hinders interregional knowledge flow and the creation of intangible assets for local enterprises (Scellato et al., 2015). In contrast, in Western contexts, tenured professors often hold dual roles in academia and industry, enhancing cross-sector knowledge exchange (Bauder, 2015). However, the faculty profession in Chinese universities remains relatively stable, with low mobility (Yan et al., 2015). This highlights the need to foster greater interaction and collaboration between universities and industry, as well as to ensure a rational and fluid allocation of university human resources. Encouraging such mobility within large cities can enhance regional human capital distribution and mitigate disparities in university resources.

These findings contribute to the literature on university–industry linkages by suggesting that the effectiveness of university innovation resources does not solely depend on the scale of resource inputs but also on the institutional and organizational mechanisms through which these resources are mobilized and transferred to firms. This highlights the importance of considering the structural conditions that shape knowledge diffusion within regional innovation systems.

The analysis also identifies local human capital and enterprise R&D investment as mediating factors between university innovation inputs and enterprise performance, supporting hypotheses H2a and H2b. This is because resource inputs can foster regional agglomeration effects, stimulate economic vitality, and encourage businesses to enhance R&D investments (Romer, 1990; Benhabib & Spiegel, 1994). The finding that scientific publications are negatively associated with R&D investment warrants further consideration. A plausible explanation is that university scientific output may substitute for firm-level R&D in the short term, particularly when firms lack sufficient absorptive capacity to translate academic knowledge into applied innovation. Alternatively, the time lag between knowledge creation and its commercial application may obscure positive effects in the contemporaneous analysis. High-quality human capital with advanced learning and knowledge transformation abilities further drives enterprise innovation and growth (Aghion & Howitt, 1992). Consequently, increasing financial and human resource investments in universities and developing resource evaluation and allocation mechanisms that address regional disparities are imperative.

From a theoretical perspective, this study extends the innovation production framework (Griliches, 1979). It incorporates firm-level absorptive capacity (R&D investment and human capital) as mediating channels, thereby shifting the analytical focus from aggregate knowledge spillovers to the micro-foundations of knowledge transfer. This addresses a gap in the literature, which has predominantly focused on direct effects without fully examining the internal mechanisms through which external knowledge is transformed into firm-level outcomes.

Heterogeneity analysis shows that non-state-owned enterprises (NSOEs) and small- and medium-sized enterprises (SMEs) benefit significantly from university innovation resources in the long term. While prior literature has seldom distinguished between short-term and long-term enterprise performance, some studies have reached similar conclusions regarding heterogeneous firm responses to innovation resources.

This result suggests that firm characteristics play an important role in shaping the effectiveness of university innovation resources. The finding that NSOEs and SMEs benefit more from university innovation resources aligns with prior studies on firm heterogeneity (Laursen & Salter, 2006), which suggest that firms with limited internal R&D capacity rely more heavily on external knowledge sources. Our results extend this literature by distinguishing between short-term and long-term performance effects, showing that the reliance on university resources is particularly pronounced in the long run for these firm types. This distinction is important because it suggests that the benefits of university–industry collaboration may take time to materialize, especially for firms with fewer internal resources.

Overall, the findings of this study enrich the existing literature by providing empirical evidence on how university innovation inputs influence firm performance through internal firm mechanisms and heterogeneous firm characteristics. By integrating insights from knowledge spillover theory and innovation production theory, the study offers a more comprehensive explanation of the pathways linking university innovation resources and enterprise development.

6. Conclusions and Implications

This study investigates the relationship between university innovation resources and the performance of regional listed enterprises in the Yangtze River Delta. The empirical results indicate that most university innovation resources positively influence enterprise performance. In addition, the analysis reveals that local human capital and enterprise R&D investment serve as important mediating mechanisms through which university innovation resources affect firm performance. The heterogeneity analysis further shows that the effects are more pronounced for non-state-owned enterprises and small- and medium-sized enterprises in the long term.

From a theoretical perspective, this study contributes to the literature by clarifying the conceptual distinction between university innovation inputs and outputs by emphasizing the role of innovation inputs as foundational drivers of knowledge creation and diffusion. By focusing on research funding and research personnel as key innovation inputs, this study provides a more precise analytical framework for examining the role of universities in regional innovation systems.

Furthermore, the findings highlight the importance of internal firm mechanisms in transforming external knowledge resources into economic performance. Specifically, R&D investment and human capital act as critical channels through which university-generated knowledge influences enterprise development. This contributes to the broader literature on knowledge spillovers and absorptive capacity by demonstrating the micro-level mechanisms through which university innovation resources are linked to firm performance.

From a practical perspective, the results suggest that policymakers should strengthen support for university research funding and facilitate closer collaboration between universities and enterprises. Encouraging knowledge exchange, promoting academic mobility, and enhancing regional human capital development may further improve the effectiveness of university-driven innovation.

For enterprises, the findings suggest that strategically increasing internal R&D investment can enhance absorptive capacity, enabling firms to better utilize university-generated knowledge. Non-state-owned enterprises and SMEs, which derive greater long-term benefits from university innovation resources, should proactively establish sustained collaborations with universities. Additionally, the negative effect of senior faculty inputs on firm performance highlights the value of fostering more flexible and direct interactions with university researchers through joint projects and talent training programs.

For universities, the findings indicate that simply increasing human resource inputs may not automatically benefit regional enterprises, partly due to limited academic mobility. Policies encouraging faculty engagement with industry—such as incorporating collaboration into promotion criteria and supporting sabbaticals in enterprises—can help address this. Strengthening technology transfer infrastructure and maintaining commitment to basic research are also essential, as these activities support regional human capital development and facilitate the translation of research into industrial applications.

Overall, by providing empirical evidence from one of China’s most dynamic innovation regions, this study enriches the understanding of how university innovation inputs contribute to enterprise development and offers new insights into the role of universities in regional economic growth.

7. Limitations and Future Research

This study has several limitations that warrant further investigation. First, while our fixed-effects panel regression controls for time-invariant unobserved heterogeneity, it does not fully eliminate concerns about reverse causality or time-varying omitted variables. Future research could employ instrumental variable approaches or quasi-experimental designs to strengthen causal claims. Second, our two-stage mediation analysis relies on the assumption that mediators are exogenous conditional on controls. Longitudinal mediation models with lagged structures could better capture dynamic feedback loops and temporal sequencing. Third, the spatial analysis using ArcGIS is primarily descriptive. Spatial econometric techniques such as Moran’s I or spatial autoregressive models would enable formal testing of spatial spillovers and dependencies. Fourth, our sample is restricted to listed enterprises, which may limit generalizability to unlisted small and micro firms. Future studies could incorporate survey data or administrative records covering a broader spectrum of enterprises. Fifth, the temporal scope of our data (2008–2017) may not capture recent developments following the formal incorporation of the G60 Corridor into China’s 14th Five-Year Plan. Extending the analysis to more recent years would reveal whether the identified patterns have persisted or evolved. Finally, enterprise heterogeneity was analyzed primarily through ownership and size. Future research could expand this classification to include technological orientation (e.g., high-tech vs. non-high-tech) or industry-specific characteristics (Hong & Su, 2013; Gao et al., 2024; Du & Seo, 2022). Further exploration of university heterogeneity—such as categorizing universities by quality (e.g., 985 vs. non-985) or ownership (e.g., public vs. private)—would also provide deeper insights into their differential impacts on regional innovation. Additionally, emerging technologies such as artificial intelligence are increasingly shaping university-industry collaboration and green innovation outcomes, as highlighted by Xia et al. (2025), suggesting a promising avenue for future research at the intersection of AI, sustainability, and regional innovation systems.