Sage Journals: Discover world-class research

Abstract

Over the past few decades, “governing by numbers” in higher education has made significant strides. In this context, Chinese universities have actively responded to national policy demands, represented by the “New Quality Productive Force” (NQPF) strategies. Drawing on panel data from 30 Chinese provinces, municipalities, and autonomous regions from 2012 to 2022, this study employs two-way fixed-effects and mediation-effects models to empirically investigate the impact of university science and technology investment (USTI) on regional NQPF (RNQPF). The results demonstrate that USTI significantly promotes RNQPF, with basic research investment having the most pronounced effect. Mechanism analysis reveals that USTI facilitates RNQPF by promoting scientific research output, fostering high-tech industry supply and demand, and driving regional development. However, the mediating effects of talent cultivation, attracting high-level talent, and the transformation of scientific and technological achievements are not significant. This study posits that university functions are characterized by loosely coupled systems in the field of higher education; thus, quantifiable indicators only measure limited objectives of higher education development while overlooking more complex and latent functions. The findings suggest that increasing investment in basic research and optimizing technology resource allocation, focusing on scientific output and the cultivation of high-tech industries, can enhance RNQPF.

Keywords

new quality productive force science and technology investment regional economy higher education policy evaluation

Introduction

In a centralized higher education system, universities are profoundly influenced by national agendas that shape their goals, strategies, and governance structures. As open systems, universities continuously adapt to external pressures, such as political and economic forces, by realigning their internal structures and missions. This adaptability becomes particularly crucial when national agendas are clearly articulated, and resource allocation is tied to compliance with policy objectives (Papadimitriou & Boboc, 2021). In such systems, universities become more strategically oriented, refining their strategies as they respond to policy demands and strengthen their legitimacy (Krücken et al., 2025; Ramirez-Cardona & Calderón-Hernández, 2024). Chinese universities, in particular, are characterized by a highly centralized governance model, with substantial influence from the state and national agendas (Y. Wang et al., 2022).

The New Quality Productive Forces (NQPF), introduced by President Xi Jinping in 2023 and now a central element of China’s national agenda, has had a significant impact on university governance. Similar to policies like “Bidenomics” or the “New Washington Consensus” in the United States (Wei, 2023) and the European Union’s “Economic Security Strategy” (European Commission, 2023), the NQPF seeks to address the uncertainties surrounding future technological innovation needs. It emphasizes innovation-driven, advanced technological, high-efficiency, and high-quality production processes. By promoting industrial upgrades and optimizing resource allocation, it aims to transform discourse and create policy synergy, injecting new momentum into economic growth while ensuring national economic security (Z. Liu, 2024). A key pathway in this process is the Chinese government’s increased investment in university science and technology, leveraging the research strengths and talent cultivation capabilities of higher education institutions to rapidly advance the NQPF and drive sustainable economic development. Under the discourse of the NQPF, the fate of universities is closely linked to their ability to convert government investment into tangible advancements in NQPF.

While the NQPF has become increasingly formalized into a calculable policy framework, positioning it as an authoritative guide in higher education planning, its transformation warrants critical scrutiny. In the context of higher education’s integration into “governing by numbers,” such metrics are not merely neutral tools but become part of a broader power-knowledge complex (Madsen, 2022). The process of converting strategic concepts like NQPF into quantifiable targets risks reifying specific developmental logics while sidelining alternative visions of institutional purpose. As Piattoeva and Boden (2020) argue through the concept of “escaping numbers,” metrics initially designed as tools of control can detach from their original purposes, becoming self-referential and performative across different governance settings. When reduced to a system of indicators, NQPF may inadvertently direct universities down predetermined paths, reinforcing selective priorities, legitimizing narrow investment rationales, and generating unintended allocative consequences. Understanding the relationship between quantified national agendas and university science and technology investment (USTI) requires careful attention to the potential consequences of these measurement regimes.

Bring into consideration of the two highly cited and influential NQPF indexes (Lu et al., 2024; J. Wang & Rongji, 2024) by scholar communities as the dependent variable, and per capita USTI (the total USTI in a province divided by the full-time equivalent (FTE) number of research and development personnel in the province) as the core independent variable, along with other control variables, this study aims to explore the relationship between USTI and the regional NQPF(RNQPF), and thus reveal the above mentioned potential consequences of those quantified measurement regimes.

Literature Review and Hypothesis Proposal

Literature Review on the Relationship Between Higher Education and NQPF

NQPF signifies a qualitative leap from traditional productivity. The theoretical perspective of Marxism and the vision of China’s traditional culture form the theoretical logic for proposing NQPF. China’s economic development achievements and the technological revolution wave constitute the historical logic behind NQPF. NQPF, as a composite concept of “political concept, academic concept, and industry concept” (Zeng & Zeng, 2024), emerges from the interplay of theoretical, historical, and practical logic, distinguishing it from traditional productivity in three aspects. These aspects include high-quality workers characterized by knowledge-based, innovative, and autonomous attributes (Pu & Huang, 2023), natural objects and raw materials with increased technological elements accompanying new scientific discoveries, as well as non-material forms of “new materials” objects of labor (Z. Xie et al., 2024), and “new media” means of labor exemplified by high-precision equipment and digital tools (Shi & Ling, 2024).

Adapting and supporting high-quality development is a fundamental requirement for higher education (S. Wang & Yifei, 2023). The functions of talent cultivation, scientific research, and social services within modern higher education systems are closely linked to NQPF development. By establishing a talent cultivation system aligned with NQPF, strengthening basic research and original innovation, and promoting education evaluation system reform and organized scientific research, higher education can offer robust human resources and technological support for regional economic development (F. Xie, 2023). Huai (2024) emphasize the pivotal role of technological innovation and educational development in fostering NQPF, asserting that promoting NQPF development is a critical mission for higher education.

Research on the Impact of University Scientific and Technological Investment on Regional Economic Development

Scholars in economics and higher education have closely examined the impact of USTI on regional economic development. Feldman (1999) highlights the significance of geographical location on innovative activities, including industry clusters, urban scale economies, and diversified economic activities’ roles in promoting innovation. Feldman also emphasizes the geographical constraints on knowledge spillover, suggesting that knowledge and technology are more likely to be disseminated and applied within their region of origin or adjacent areas. Furthermore, Feldman underscores the role of individuals, particularly “star scientists,” in knowledge and skill transfer. Henrys (2008)“university-government-industry triple helix” model considers universities as economic growth entities parallel to government and enterprises, positing that these three entities together form the basic structure of innovation-driven economic growth. Guo et al. (2023) argue that higher education’s promotion of regional economic development is primarily reflected in human resources, scientific research achievements, infrastructure, and labor consumption. They identify factors hindering higher education’s promotion of high-quality regional economic development, including mismatches between professional settings and regional economic development, discrepancies between teaching staff and social service demands, and the need to optimize mechanisms for university-industry-research collaboration.

Existing empirical research predominantly examines the impact of university science and technology innovation on regional or urban economic development. Luo and Zhiqin (2022) discovered that most provincial universities did not achieve DEA efficiency in technology research and development, and technology achievement transformation stages. While university science and technology research and development have a positive yet insignificant impact on high-quality economic development, the transformation rate of university science and technology achievements significantly impacts economic development. S. Wang and Yifei, 2023, applying ecological thinking and the fsQCA method, identified two driving mechanisms for higher education supporting regional economic development: “talent - achievement transformation - opportunity support type” and “achievement transformation - opportunity - cooperation support type.” Educational achievement transformation and investment deficit were found to be the core conditions hindering higher education’s support for regional economic development. Y. Liu and Jiahui (2023), using natural fracture and spatial measurement methods, uncovered a significant spatial correlation between university science and technology innovation capabilities and high-quality economic development levels in various provinces. Differences were also observed in the basic conditions and innovation speeds of university science and technology innovation among cities in eastern and western regions.

The promotion of technological innovation in universities and colleges has varying effects on economic development depending on factors such as school type differences (Tan & Yibei, 2015), temporal differences (Leng, 2009), and time lags (Tan & Yibei, 2015), and is influenced by intermediary factors such as industrial structure. Nie (2024), using a DID model, found that the “Double First Class” initiative can promote urban economic development through knowledge spillover effects, and that the initiative has a greater impact on the economic development of cities in eastern regions, southern regions, and cities with superior industrial structures.

In conclusion, the emergence of NQPF is grounded in theoretical, historical, and practical logic, imposing new demands on labor, objects of labor, and means of labor. Higher education in this new era should prioritize NQPF development as a key mission, strengthening the nexus between scientific research, teaching, industry integration, and talent cultivation while leveraging universities’ role as primary drivers of basic research. Concurrently, government investment in science and technology, along with university innovation output, is subject to both a general “high input, high output” pattern and factors such as regional economic levels, industrial structures, and universities’ technological transformation capabilities. To optimize the allocation of higher education resources and effectively promote NQPF development, it is crucial to accurately assess the actual effects of regional USTI on science, technology innovation, and regional economic development levels. Moreover, clarifying the relationships between regional science and technology investment, university science and technology innovation, and regional economic development is essential in exploring potential mechanisms for USTI’s impact on RNQPF.

Research Hypotheses

University-based research fosters applied research, addressing economic and industrial challenges. Increased USTI funding allocates resources for basic and applied research, promoting technological advancements and innovation. This facilitates local access to cutting-edge scientific solutions, driving improvements in product innovation, process efficiency, and cost reduction, ultimately enhancing RNQPF. Given knowledge diffusion within and around the region of origin (Lin et al., 2025), USTI directly influence RNQPF. Considering this analysis, the following hypotheses are presented:

H1: USTI can propel the enhancement of RNQPF.

H2: USTI can stimulate scientific research achievement output, ultimately improving RNQPF.

The human capital externality theory highlights significant positive externalities from labor factor agglomeration, particularly advanced labor factors (Glaeser & Lu, 2018). Empirical evidence from various countries shows a strong positive correlation between higher education expansion, notably graduate education, and the proportion of highly-educated individuals in a region. Increased USTI drives higher education expansion, as resources are allocated to improve education infrastructure, resources, and quality. By attracting students to engage in research and innovation through scholarships and expanded enrollment, universities provide vital human resource support for regional economic development. This results in a larger pool of advanced labor forces, thereby promoting RNQPF. Considering this analysis, the following hypothesis is proposed:

H3: USTI can expand the scale of talent cultivation, thereby elevating RNQPF.

The triple helix theory asserts that local enterprises’ assimilation and application of universities’ research and technological achievements can enhance production efficiency and product quality, thereby elevating RNQPF levels. Augmenting USTI facilitate the creation of transformation platforms and service systems, enabling effective collaboration between scientific accomplishments and market demands, just like the application-oriented undergraduate institutions Shenzhen Polytechnic University and Nanjing Institute of Industry Technology are well intertwined with their collaborated enterprises (Zhang, 2025). This helps enterprises address technical challenges and fosters industrial innovation. Considering this analysis, the following hypothesis is proposed:

H4: USTI facilitates the transfer of scientific and technological achievements, enabling universities to fulfill their social service function through technology transfer, patent sales, and other avenues, thereby raising RNQPF levels.

The new era demands drive the transformation toward high-tech industries, characterized by intellect, innovation, strategic significance, and low resource consumption. Unlike the “small science” era, the “big science” era requires substantial investment in research and development for high-tech industries. Adequate funding in science and technology is essential for their emergence and evolution. University investments in science and technology foster breakthroughs in fundamental innovation, creating solutions for significant practical problems, thereby fostering and strengthening high-tech industries (Zhao & Lilan, 2009). Considering this analysis, the following hypothesis is presented:

H5: USTI can foster high-tech industry supply and demand, increase regional high-tech industry revenue, and thereby elevate RNQPF levels.

High-level talent is pivotal in promoting scientific innovation and industry advancement. Possessing extensive research experience, innovation capabilities, and leadership potential, these individuals achieve remarkable research accomplishments while guiding disciplinary development and industrial innovation. Attracting and nurturing high-level talent is vital for high-tech industry growth, requiring not only competitive salaries and research funding but also high-quality research equipment, platforms, and teams. These necessitate significant USTI. In view of this analysis, the following hypothesis is proposed:

H6: USTI can attract high-level talent, stimulate innovative vitality, and thus enhance RNQPF levels.

Research Design

Variables and Data Sources

The variables used in this study primarily originate from the “Compilation of Science and Technology Statistics of Higher Education Institutions” by the Science and Technology Department of the Ministry of Education, the Social Science, Technology, and Cultural Industry Statistics Division of the National Bureau of Statistics, the “China High-tech Industry Statistical Yearbook” by the High-tech Industry Division of the National Development and Reform Commission, the “China Education Statistical Yearbook” by the Development Planning Division of the Ministry of Education, and the official websites of the National Natural Science Foundation of China, the Chinese Academy of Sciences, and the Chinese Academy of Engineering. The data used spans the years 2012 to 2022.

The dependent variable of this study is derived from the comprehensive evaluation index system of RNQPF, as calculated by J. Wang and Rongji (2024). Additionally, this study will use the NQPF index calculated by Lu et al. (2024) as an alternative dependent variable.

The core independent variable of this study is the per capita USTI. Due to differences in population, economy, and the scale of higher education across provinces, using a per capita indicator can control for differences in expenditure inputs caused by scale disparities, allowing for a fairer comparison of the efficiency of USTI between different regions. To further delve into the analysis, this study will also use per capita USTI on basic research, per capita USTI on applied research and experimental development, per capita government USTI, and per capita expenditures USTI by enterprises and institutions as alternative independent variables.

The influence of USTI on RNQPF may be affected by various external factors. Drawing on relevant literature, this study selects the following control variables: (1) the number of colleges and universities in the region; (2) the proportion of colleges and universities at the undergraduate level in the region; (3) the proportion of faculty members with senior professional titles in the region; (4) the total value of fixed assets in colleges and universities in the region. The operational definition of variables and data sources are shown in Table 1.

Table 1.

Operational Definition of Variables and Data Sources.

Types of Variable	Variable	Variable abbreviations	Definition	Data sources
Dependent variable	Comprehensive index for new quality productivity	Npro1	Evaluation index system for NQPF using laborers, labor objects, and labor resources	J. Wang and Rongji (2024)
Dependent variable	Levels of development of new quality productivity	Npro2	Evaluation index system measurement results encompassing three dimensions: scientific and technological productivity, green productivity, and digital productivity.	Lu et al. (2024)
Independent variables	Per capita USTI	IP	University/colleges science and technology expenditures/full-time equivalent (FTE) staff for research and development	Compilation of Science and Technology Statistics of Higher Education Institutions
	Per capita USTI on basic research	FIP	Basic research USTI/number of personnel invested in basic research in the same year
	Per capita USTI on applied research and experimental development	ATIP	Applied and experimental research USTI/number of personnel invested in applied and experimental research in the same year.
	Per capita USTI by government	GIP	Government-funded USTI/Full-time equivalent (FTE) staff for research and development.
	Per capita on science and technology by enterprises and institutions	EIP	USTI commissioned by business entities/full-time equivalent (FTE) staff for research and development.
Mediating variable	Number of graduate students	ST	Number of graduate students engaged in USTI	Compilation of Science and Technology Statistics of Higher Education Institutions
	Number of science and technology articles	TP	Number of articles related to scientific and technological research published by university faculty or students in academic journals
	Number of all published articles	WP	Number of scientific and technological papers published by university faculty or students in foreign or national academic journals
	Number of Technology transfer contracts (items)	TR	Total number of technology transfer contracts signed between universities or research institutions and other entities (such as enterprises and other research institutions)
	Technology transfer income	TRI	Total monetary income obtained by universities or research institutions from other entities through technology transfer contracts
	Number of patent sales contracts	SP	Number of contracts for the sale of patents owned by universities to enterprises or other institutions
	Patent sales income	SPRI	Total monetary income obtained by universities through the sale of patents
	Operating revenue of high-tech industries	HTI	Sum of operating income from medical instrument and pharmaceutical manufacturing, electronic and communication equipment manufacturing, and computer and office equipment manufacturing	China High-tech Industry Statistical Yearbook
	Revenue from new products in high-tech industries	HTP	Sum of new product income from medical instrument and pharmaceutical manufacturing, electronic and communication equipment manufacturing, and computer and office equipment manufacturing	China High-tech Industry Statistical Yearbook
	Number of high-level talents	HT	Total number of recipients of the National Outstanding Youth Science Fund, the Excellent Youth Science Fund, and the number of newly elected academicians of the two academies	Official Websites of the National Natural Science Foundation of China, the Chinese Academy of Sciences, and the Chinese Academy of Engineering.
Control variable	Number of colleges and universities	SC	Number of higher education institutions (including regular universities, vocational universities at the undergraduate and junior college level)	China Education Statistical Yearbook
	Proportion of colleges and universities at the undergraduate level	USP	Number of undergraduate level universities/Total number of universities
	Total value of fixed assets	CAP	Total value of university fixed assets, including teaching and scientific research equipment and facilities
	Proportion of faculty members with senior professional titles	PROP	(Number of professors + Number of other senior technical personnel) /Total number of faculty and staff	Compilation of Science and Technology Statistics of Higher Education Institutions

According to the hypotheses of this study, the following variables were selected as mediating variables: the number of postgraduate students, the quantity of scientific and technological academic papers, the quantity of scientific and technological academic papers in foreign and national publications, the number of technology transfer contracts, technology transfer income, the number of patent sales contracts, patent sales income, the main business income of high-tech industries, and the number of high-level talents. The names, definitions, and data sources of the variables in this study are shown in Table 1.

Model Design

To test Hypothesis 1 of this study, the following regression model (1) is constructed:

\begin{matrix} NPr o_{it} = α_{0} + α_{1} I P_{it} + \sum Contro l_{it} + \sum Y ear \\ + \sum Area + ε_{it} \end{matrix}

(1)

NPro is the dependent variable representing the RNQPF; IP is the independent variable denoting the per capita USTI; Control represents a series of control variables; Year and Area are fixed effect represent the fixed effects for the year and individual, respectively; ε is the random disturbance term; the subscripts i and t represent the province and year, respectively. In this study the individual fixed effects (area fixed effects) pertain to the fixed effects of the 31 provinces in China, which account for the unique characteristics and differences among these regions.

To test Hypotheses 2, 3, 4, 5, and 6, this study refers to Jiang’s (2022) discussion on mediating effect analysis and its testing, and constructs a mediating effect model (2):

\begin{matrix} M V_{it} = β_{0} + β_{1} IP + \sum Contro l_{it} + \sum Y ear \\ + \sum Area + ε_{it} \end{matrix}

(2)

MV represents the mediating variable. During the mediation effect regression analysis, the following variables are used as independent variables: postgraduate training scale (ST), scientific research achievements (TP, WP), scientific and technological achievements transformation (TR, TRI, SP, SPRI), regional high-tech industry income (HTI, HTP), and the number of high-level talents (HT). The dependent variable, fixed effect, random disturbance term, and subscripts in model (2) are consistent with model (1).

Empirical Analysis Results

Descriptive Statistics

Table 2 displays the descriptive statistics of the dependent variable. Between 2012 and 2022, the RNQPF remains on an upward trend, the average RNQPF across the country increased from 0.101 to 0.205, and the median increased from 0.091 to 0.165. In 2012, the maximum value of RNQPF was approximately 4.32 times the minimum value, 1.78 times the average value, and 1.98 times the median. In 2022, these gaps widened to 4.78, 2.48, and 3.10 times, respectively, indicating disparities in the RNQPF among different regions and an expansion of these disparities over the decade. Provinces with relatively high RNQPF in 2022 included Guangdong, Jiangsu, Zhejiang, Shanghai, Beijing, and Shandong, all of which are eastern provinces in China.

Table 2.

Descriptive Statistics of Dependent Variable.

Province	2012 Level	2022 Level	2012 Ranking	2022 Ranking	Ranking Change
Guangdong	0.179	0.509	2	1	1
Beijing	0.163	0.417	4	2	2
Jiangsu	0.179	0.409	1	3	−2
Zhejiang	0.173	0.378	3	4	−1
Shanghai	0.130	0.348	6	5	1
Shandong	0.127	0.292	7	6	1
Mean	0.101	0.205
Median	0.091	0.165

Note. Due to space limitations, only the calculation results for provinces with higher levels of RQNP are displayed.

From 2012 to 2022, the total USTI across all provinces in China increased from 103 billion yuan to 282.9 billion yuan, representing an average annual increase of 17.9 billion yuan, with an average growth rate of approximately 17%. During this period, the per capita USTI grew from 3.19 million yuan to 5.46 million yuan, showing an average annual increase of 2.27 million yuan and an average growth rate of 7.12%. The median value of per capita USTI increased from 2.98 million yuan to 5.25 million yuan across all provinces. In 2012, the maximum value of per capita USTI was 7.63 times the minimum value, but this gap narrowed to 5.38 times in 2022. The average value of per capita USTI was 1.07 times the median value in 2012, and this difference reduced to 1.04 times in 2022. Based on the descriptive analysis of the independent and dependent variables, it is evident that over the past decades, there has been a widening gap in NQPF levels among various provinces, whereas the disparity in USTIs across colleges and universities has contracted.

Over all, the disparity in USTI among provinces in China has narrowed in recent years, with rapid increases in various types of USTI. Notably, the USTI for applied and experimental research, the government USTI have experienced relatively higher growth rates. The collinearity test results of the main variables show that the Variance Inflation Factor (VIF) values range from 1.27 to 4.44, indicating that there is no collinearity problem among the selected independent variables. The specific intermediary variables and their descriptive statistical results can be found in Table 3.

Table 3.

Descriptive Statistical Results for the Intermediary Variables Involved in This Analysis.

Variable	Number	Mean	Median	Std. dev.	Min	Max
ST	330	13,331.12	9,812.5	13,639.09	112	78,007
TP	330	31,499.04	25,366.5	24,888.24	8.504	122,873
WP	330	12,580.08	8,678	13,509.25	65	74,398
TR	330	427,036.4	260	609,133.2	0	5,056
TRI	330	97,237.67	45,696	142,208.2	0	739,252
SP	330	209,063.6	86	419,902.2	0	4,175
SPI	330	111,139.5	19,582.5	305,498.5	0	3,574,191
SPRI	330	44,969.95	10,940	88,771.15	0	577,123
HTI	330	5,047.565	2,119.4	8,667.678	16.9	56,092.89
HTP	330	17,214,430	4,181,489	35,969,999	2,253	245,664,700
IT	330	2,352.455	1,590.5	2,409.783	6	11,407
HT	330	167,363.6	70.5	285,908.5	0	2,042

Benchmark Regression

As shown in Table 4, column (1) presents the direct regression results between per capita USTI (independent variable) and RNQPF (dependent variable); columns (2) and (3) represent the regression results after controlling fixed effects and adding control variables, respectively; and column (4) shows the regression results after simultaneously adding control variables and fixed effects. It can be observed that the coefficient of per capita USTI is significantly positive at the 1% level. Thus, H1 of this study is validated, indicating that USTI can propel the enhancement of RNQPF.

Table 4.

Benchmark Regression Analysis Results.

Variables and statistical coefficients	Npro1
Variables and statistical coefficients	(1)	(2)	(3)	(4)
IP	2.433 604*** (0.157 652 1)	0.805 357 2*** (0.202 688)	0.514 520 4*** (0.112 304 9)	0.682 683 5*** (0.013 391 18)
SC			−7.782 638*** (1.411 901)	−13.226 41*** (3.626 247)
USP			−625.323 (380.382 1)	−625.303 4 (605.436 2)
CAP			0.000 163 4*** (0.000 006 52)	0.000 181 1*** (0.000 009 9)
PROP			863.669 5 (526.070 5)	199 9.227^** (746.879 8)
Constant	397.962 4 *** (106.477 8 )	751.787 4*** (80.875 16)	812.795 2** (241.261)	918.230 7⁺ (542.212 3)
Fixed effects	No	Yes	No	Yes
N	330	330	330	330
R² adjusted	0.415 6	0.416 1	0.781 9	0.674

Note. The symbols ⁺, *, **, and *** represent the 10%, 5%, 1%, and 0.1% significance levels, respectively. The numbers in parentheses are standard deviations.

Robustness Test

Replacing Dependent Variable

To validate the robustness of the findings, this study employs the NQPF index developed by Lu et al. (2024) as an alternative dependent variable. By employing Lu et al.’s NQPF index as the dependent variable in Model (1), the regression results, as shown in Table 5 column (1), demonstrate that the coefficient of per capita USTI is significantly positive at the 1% level. This indicates that per capita USTI still has a positive impact on RNQPF, and thus H1 of this study still validated.

Table 5.

Robustness Test Analysis Results (I).

	Alternative dependent variable	Exclusion of outlier years	Exclusion of outlier provinces
Way of robustness test	Npro2	Npro1	Npro1
Variables and statistical coefficients	(1)	(2)	(3)
IP	0.792 554 6*** (0.241 937 7)	0.339 758 9* (0.140 348 9)	0.768 579 7*** (0.214 012 8)
Constant	−474.105 3 (979.612 1)	711.256 5 (546.866 8)	660.793 1 (787.586 6)
Control variable	Yes	Yes	Yes
Fixed effect	Yes	Yes	Yes
N	330	240	286
R² adjusted	0.645 4	0.757 8	0.719 2

Note. The symbols * and *** represent the 5% and 0.1% significance levels, respectively. The numbers in parentheses are standard deviations.

Replacing Independent Variables

To further investigate the impact of USTI from various sources and for different purposes on RNQPF, this study employs alternative independent variables. Per capita government USTI, per capita enterprise USTI, per capita basic research USTI, and per capita applied and experimental research USTI are utilized in Model (1) for regression analysis. The results in Table 6, column (1) indicates that both government and enterprise USTI contribute to enhancing RNQPF, with enterprise USTI having a larger promotional effect. The results in Table 6, column (2) suggest that per capita USTI, particularly basic research USTI, can effectively promote RNQPF, and all these imply that H1 of this study is validated.

Table 6.

Robustness Test Analysis Results (II).

Variables and statistical coefficients	Npro1
Variables and statistical coefficients	(1)	(2)
Independent variables
GIP	0.566 870 7*** (0.172 622 8)
EIP	1.101 107* (0.497 647 6)
FIP		0.968 679 3*** (0.194 589 2)
ATIP		0.130 077 6 (0.151 453 5)
Constant	770.288 8 (554.856 8)	839.302 4 (550.844)
Control variable	Yes	Yes
Fixed effects	Yes	Yes
N	330	330
R² adjusted	0.697	0.701

Note. The symbols * and *** represent the 5% and 0.1% significance levels, respectively. The numbers in parentheses are standard deviations.

Exclusion of Outlier Years and Provinces Samples

Descriptive statistics indicate that the COVID-19 pandemic had a negative impact on USTI in 2020. To mitigate the potential influence of this outlier, data from the years 2020 to 2022 (3 years in total) were excluded, and Model (1) was re-estimated. As shown in Table 5, column (2), the coefficient of per capita USTI remained significantly positive at the 5% level. Due to the unique nature of USTI in certain provinces, four samples from Ningxia, Qinghai, Hainan, and Inner Mongolia were removed to minimize the influence of extreme outliers. Model (1) was re-estimated, and the results are presented in Table 5, column (3). The adjusted R-squared (R²) value improved from 0.674 in the original model (Table 5, column (4)) to 0.719, and the coefficient of per capita USTI remained significantly positive at the 0.1% level. Which means that H1 of this study still validated.

Endogeneity Test

This study employs an instrumental variable method to address endogeneity concerns. One-period and two-period lagged values of the independent variable are selected as instrumental variables. The rationale for using lagged values as instruments lies in their ability to resolve endogeneity issues by introducing a variable correlated with the independent variable but not with the error term. Lagged values of the independent variable are considered exogenous to the error term for the current period, as they are not influenced by current RNQPF. By employing lagged values as instrumental variables, the correlation between the independent variable and the error term is mitigated, yielding more accurate and robust estimation results.

Columns (1) and (2) of Table 7 present regression results using the one-period lagged independent variable as an instrumental variable, while columns (3) and (4) employ the two-period lagged independent variable as an instrumental variable. The first-stage regression results in columns (1) and (3) of Table 7 indicate that the coefficients of the chosen instrumental variables are positive and significant at the 0.1% level. The first-stage F-test values are greater than the critical value of 10, suggesting that the instrumental variables pass the exogeneity test. Furthermore, the instrumental variables also pass weak identification, under-identification, and overidentification tests, supporting the validity of the one-period and two-period lagged independent variables as effective instrumental variables. The second-stage regression results in columns (2) and (4) of Table 7 demonstrate that the coefficients of per capita USTI are significantly positive at the 0.1% level. This finding suggests that, even after considering the issue of endogeneity, USTI significantly promotes RNQPF.

Table 7.

Robustness Test Analysis Results (III).

Instrumental variable	One-period lagged model		Two-period lagged model
Dependent variable	IP	Npro1	IP	Npro1
Variables and statistical coefficients	(1)	(2)	(3)	(4)
IP		0.971 267 9*** (0.263 970 2)		1.041 545*** (0.473 457)
Instrumental variables	0.535 401 4*** (0.050 588 2)		0.335 431 1*** (0.062 872 1)
Constant	485.045 2* (239.163 2)	752.445 2 (698.473 2)	207.899 (371.868 4)	907.981 9 (949.662 8)
Control variable	Yes	Yes	Yes	Yes
Two-way fixed effects	Yes	Yes	Yes	Yes
N	300	300	270	270
R² adjusted	0.580 7	0.674 9	0.701 7	0.643 4
First-stage F-test	39.46		24.47
Weak identification test	82.179***		26.741***
Weak instrumental variable test	112.011 (Stock-Yogo 10% critical value )		28.464 (Stock-Yogo 10% critical value)
Overidentification test	Yes ***		Yes ***

Note. 1. As indicated by the benchmark regression and robustness test, USTIs have a significant impact on improving RNQPF. Furthermore, compared to investments in applied research and experimental development, basic research investments demonstrate a more pronounced effect in promoting RNQPF. 2. The symbols * and *** represent the 5% and 0.1% significance levels, respectively. The numbers in parentheses are standard deviations.

Mechanism Study

USTIs Can Stimulate Output of Scientific Research, Thereby Improving RNQPF

Table 8, column (1) displays the results employing scientific and technological academic papers as a mediating variable. The findings reveal a significantly positive coefficient for per capita University Science and Technology Investment (USTI) at the 5% level, suggesting a potential mediating effect of scientific research publication in enhancing RNQPF. The Sobel test result (0.029) lends support to Hypothesis 2 of this study.

Table 8.

Mediation Effect Analysis Results (I).

Dependent variable	Npro1
Mediating variable	TP	WP	ST	TR	TRI
Variable and statistical coefficient	(1)	(2)	(3)	(4)	(5)
IP	6.186 061* (2.697 183)	4.256 461*** (2.078 541)	0.3.437 324⁺ (2.065 566)	0.225 642 7 (0.215 756 6)	29.649 09*** (97.038 32)
Constant	294 58.6⁺ (157 05.32)	205 97.3⁺ (121 03.05)	493 32.01*** (120 27.5)	158 6.205 (125 6.321)	−629 240.9 ( 565 040.5)
Control variable	Yes	Yes	Yes	Yes	Yes
Two-way fixed effect	Yes	Yes	Yes	Yes	Yes
Sobel test	0.028 418 87	0.046 271 38	0.230 911 3	0.326 819 01	0.879 094 76
N	330	330	330	330	331
R² adjusted	0.980 4	0.960 5	0.961 7	0.790 4	0.222 2

Note. The symbols +, *, and *** represent the 10%, 5%, and 0.1% significance levels, respectively. The numbers in parentheses are standard deviations.

Table 8, column (2) showcases the results utilizing high-level scientific research output (international and national scientific and technological academic papers) as a mediating variable. The results exhibit a significantly positive coefficient for per capita USTI at the 0.1% level, which passes the Sobel test, further substantiating H2 of this study.

USTIs Can Expand the Scale of Talent Cultivation, But Not Necessarily Improving RNQPF

Table 8, column (3) shows the results using the number of postgraduates as a mediating variable. The results indicate that the coefficient of per capita USTI is significantly positive at the 10% level. However, the Sobel test result of 0.23 does not pass the test, rejecting H3 of this study, which means USTI cannot elevating RNQPF by expand the scale of talent cultivation. The lack of support for H3 can be attributed to the fact that doctoral student cultivation is predominantly concentrated in prestigious universities located in large and medium-sized cities (Luo, et al., 2022). Consequently, doctoral graduates are heavily concentrated in major cities such as Beijing, Shanghai, Guangzhou, and Shenzhen, limiting their distribution to other regions. Additionally, mismatched postgraduate program offerings, regional industrial demands, and homogenization in postgraduate training may hinder students from finding suitable positions in the job market that fully utilize their skills and knowledge (Z. Wang & Yini, 2024).

USTIs Can Promote Technology Transfer, But Not Necessarily Improving RNQPF

Table 8, column (4) presents the results using technology transfer contracts as a mediating variable. The findings show a positive, yet statistically insignificant, coefficient for per capita technological investment by universities. Furthermore, the Sobel test was not passed. Table 8, Column (5) displays the results with technology transfer income as a mediating variable. In this case, the coefficient for per capita technological investment by universities was positive and statistically significant at the 0.1% level but did not pass the Sobel test.

Table 9, column (1), (2) and (3) displays the regression model test results using the sale of patents by universities as mediating variables. The results show that when the number of patents sold, the total amount of patent contracts, and the actual income from patent sales are used as mediating variables, the coefficients for per capita investment in science and technology at universities are positive but fail to pass significance tests. Additionally, none of the results pass the Sobel test. Consequently, it can be concluded that H4 is not supported by the findings of this study.

Table 9.

Mediation Effect Analysis Results (II).

Independent variable	Npro1
Mediating variable	SP	SPI	SPRI	HTI	HTP	HT
Variable and statistical coefficient	(1)	(2)	(3)	(4)	(5)	(6)
IP	0.120 141 5 (0.288 139 3)	138.013 1 (225.434 7)	27.673 98 (62.551 45)	7.331 778*** (1.282 839)	432 38.75*** (695 6.301)	0.680 442 1 (0.134 513 8)
Constant	−432.854 3 (167 7.794)	319 606 (131 267 5)	800 02.54 (364 228.3)	312 42.01*** (746 9.792)	202 000 000*** (405 000 00)	−454 8.173 (459 5.624)
Control variable	Yes	Yes	Yes	Yes	Yes	Yes
Two-way fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
Sobel test	0.736 9	0.805 9	0.834 8	0	0	0.827
N	332	333	334	330	330	330
R² adjusted	0.213 4	0.090 4	0.170 6	0.963 4	0.937 5	0.940 5

Note. The symbol *** represent the 0.1% significance level. The numbers in parentheses are standard deviations.

Several factors may explain the lack of support for H4. Firstly, imperfect patent transfer laws and regulations contribute to inadequate invention disclosure systems in universities and research institutions (Jia et al., 2022). Secondly, universities prioritize preventing the loss of state-owned assets during technology transfer due to the dependence on national projects and the classification of state-funded research results as state-owned assets (He & Jichan, 2013). Thirdly, universities focus on patent applications as an evaluation criterion, leading to an overemphasis on publications and project initiation rather than tangible results and practical applications. Lastly, financial constraints faced by local private enterprises and ineffective collaboration between enterprises and universities hinder the transfer of scientific and technological achievements (X. Wang & Ya, 2023).

Per Capita USTIs Can Foster the Creation of High-Tech Industries, Thereby Improving RNQPF

Table 9, columns (4) and (5) show the results using regional high-tech industry income as a mediating variable. The results indicate that when high-tech industry’s main business income and high-tech industry’s new product income are used as mediating variables, the coefficient of per capita USTIs is significantly positive at the 0.1% level and passes the Sobel test. This suggests that Hypothesis H5 of this study is valid.

Per Capita USTIs Can Attract High-Level Talents, But Not Necessarily Improving RNQPF

Table 9, column (6) shows the results using the number of high-level talents as a mediating variable. The results show that the coefficient of per capita USTI is positive but not significant at the 0.1%, 1%, 5%, and 10% level, and does not pass the Sobel test (0.82), Consequently, it can be concluded that H6 is not supported by the findings of this study. H6 was not supported due to several reasons. Firstly, universities and research institutes are primarily responsible for attracting and cultivating high-level talents in China (Gu et al., 2019), which might limit the scope of talent recruitment. Secondly, using the number of recruited talents and published papers as evaluation indicators may overlook the importance of matching talent introduction with regional economic development needs (Liang & Zhengfeng, 2020). Lastly, regional university science and technology investments play a “ hygiene factor” role by maintaining the living standards and facilitating routine work of high-level talents instead of serving as an “incentive factor” for promoting research output.

Discussions and Research Conclusions

Discussions

The introduction of the NQPF concept serves as China’s solution to address the slowing global economic growth. It establishes a discourse framework driven by technological innovation in China, guiding social entities such as universities and enterprises to adjust their behaviors and promote economic development and scientific and technological innovation. Despite its conversion into a calculable policy framework by scholars like Wang et al. (2024) and Lu et al. (2024), the NQPF indicators does not fully reflect the comprehensive connotations of NQPF and is not a neutral policy tool; instead, it forms part of a broader power-knowledge complex. This calculable policy framework, developed from an economic perspective, does not consider the functional complexity of higher education. The various measurable indicators selected in the field of higher education can measure only limited objectives of higher education development. Some quantifiable indicators, such as talent cultivation, the number of high-level talents, and the transformation of scientific and technological achievements, mentioned in the article, cannot fully support the role of USTI in promoting RNQPF.

In addition to the inherent conceptual limitations of the NQPF framework, internal complexities impacting the relationship between university innovation and regional development must be acknowledged. Universities are diverse, with their role as mediators being highly heterogeneous; thus, a leading research university in a tech hub contributes to NQPF differently than a regional university in a less-developed province due to variations in resources, missions, and local industrial needs. Quantifying long-term impacts of talent cultivation and commercialization on innovation is challenging, and common metrics such as patents often fail to capture the full extent of societal benefits. While rooted in China’s context, the core challenges NQPF presents are globally relevant as governments worldwide steer higher education toward national innovation and economic competitiveness. This tension between top-down, metric-driven policy frameworks and universities’ multifaceted roles is central to higher education policy debates in Europe and North America. Thus, China’s NQPF case study provides valuable insights for international scholars and policymakers examining the instrumentalization of universities for economic goals while emphasizing the importance of these aspects in higher education development.

Over the past few decades, quantification management in higher education has made substantial advancements. The prevalent use of calculable tools for NQPF serves as a compelling testament to this progress. Key research topics in quantification management within higher education focus on the life cycle of quantification and the regulatory nexus in the sector (Hillebrandt & Huber, 2020). Such management approaches often prioritize efficiency, emphasizing the transformation of resources into discernible outcomes by universities. During this transformation process, previously non-quantified indicators are quantified, and existing ones are further optimized to ensure comparability. However, decision-making in higher education, akin to other organizations, experiences an influx of numerical data (Miller, 2001), making it challenging to establish linear relationships between inputs and outputs, similar to those observed in corporate production settings. The diverse functions of higher education are characterized by loosely-coupled (Weick, 1976), interconnected within the broader context of the sector. Consequently, caution should be exercised when associating higher education functions with measurable indicators linked to NQPF assessment tools. In the current landscape, more sophisticated data collection methods encompass macro, meso, and micro levels data, and advanced generative artificial intelligence models such as Chat-GPT and Deepseek facilitate more rational and efficient data analysis. The prospect of suitable analytical tools in the future remains promising, with the expectation that they will offer robust decision-making data support for higher education management and policy decisions.

Research Conclusions

The first notable finding is that USTIs effectively influence RNQPF but with gaps in alignment with policy discourse expectations. Policymakers anticipate rapid, high-quality economic development through comprehensive, large-scale investments in science and technology to achieve an overall enhancement of NQPF across various dimensions. The study reveals that university investments in science and technology positively affect RRNQP by promoting scientific research output and boosting regional high-tech industry income. However, the influence of USTIs on RNQPF is primarily through these two aspects, indicating that the comprehensive benefits of these investments have not been fully leveraged during policy implementation. Consequently, the policy implementation process might require further optimization and refinement to better align with the intended goals and expectations.

The second significant finding is that the mediating effects of talent cultivation, the number of high-level talents, and technology transfer are not significant, deviating considerably from policy expectations. The unsupported findings serve as a strong critique of a “Governing by Numbers” approach and hint that policymakers should recognize USTI is just a necessary but insufficient condition for elevating RNQPF, as the intended mediating channels are structurally blocked. Specifically, the coercive isomorphism arising from the pursuit of symbolic capital, such as the establishment of “double first-class universities” and doctoral degree-granting institutions spoiled the allocation of resources (Zhou, 2023). Technology transfer (H4) fails to cross the “Valley of Death” (Markham et al., 2010) for USTI successfully funds the creation of early-stage IP, but its commercialization is systemically thwarted by institutional mismatches (prioritizing academic metrics over market-ready products) and the low absorptive capacity of regional firms (Cohen & Levinthal, 1990). Concurrently, talent cultivation (H3) suffers from a severe skills mismatch, where investment increases graduate quantity but not the specific, high-end innovative quality demanded by RNQPF industries (Luo & Zhiqin, 2022; Z. Wang & Yini, 2024), thus decreases students’ employability (Brown & Souto-Otero, 2020). Finally, talent attraction (H6) is ineffective because USTI acts merely as a baseline “hygiene factor,” which is powerless against the powerful gravitational pull of established metropolitan ecosystems that offer superior career networks and non-monetary amenities (Hu et al., 2020). Therefore, the investment fails to translate into impact because the critical pathways of commercialization and talent alignment are structurally constricted.

Third, the government-led model for science and technology investments in universities may hinder the normal functioning of the university system. Chinese research universities heavily rely on government funding, leading to more influence from governmental policies in determining research directions and management structures, rather than following the natural logic of the academic community. The excessive focus on the quantity of research achievements overlooks the quality of talent cultivation, affecting teaching quality within universities. This misallocation of resources could disrupt the normal order of teaching and research in universities, hampering their long-term development. These findings indicate the need to reassess the current definition of NQPF and USTI policies. By exploring more reasonable development paths and policy guidance, the positive role of universities in the development of NQPF can be fully realized.

Last, our study is subject to certain limitations in its model design and variable selection. A primary constraint is the lack of stratified analysis across different institutional and disciplinary categories. Firstly, the model does not differentiate by academic discipline. The efficiency and nature of outputs from technology investments are known to vary significantly across fields such as science, engineering, medicine, and agriculture. Our aggregated analysis, necessitated by data availability, was unable to capture these critical distinctions. Secondly, our study does not account for institutional heterogeneity. Significant disparities exist in the productivity of technology investment between different tiers of institutions, such as research-intensive versus application-oriented universities. Similarly, the administrative affiliation of institutions—for example, those managed by central versus local governments—can substantially influence the outcomes of their technological inputs, a factor not considered in our model. Thirdly, the variables selected may not fully encapsulate the complexity of technology investment and output in higher education. Future research should therefore aim to incorporate more precise data to identify which disciplines, university tiers, and administrative affiliations yield a more pronounced impact on new quality productive forces. A deeper investigation into the specific pathways through which these investments foster such productivity is also warranted. Furthermore, subsequent studies should continue to explore the influence of other key factors, such as the scale of talent cultivation, the efficiency of technology transfer, and the concentration of high-level academic talent.

Footnotes

Acknowledgements

The authors would like to thank 2024 National Education Science Planning Project (CN) for supporting this project.

ORCID iD

Jian Chen

Ethical Considerations

These considerations were not relevant for this study type.

Informed Consent

The authors were all consented with the submission and pbulicaiton of this paper.

Author Contributions

Jingxuan Shen conceived the study, contributed to its design, supervised data collection, interpreted findings, drafted the manuscript, and edited it for final submission; Jian Chen contributed to the study design, collected and analyzed data, wrote the initial draft of the manuscript, and edited the final submission; Xiufang Duan advised on the study design, assisted with data analysis.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by 2024 National Education Science Planning Project (CN; Grant Number: WGB240478). The Project Name is “Research on the Basic Public Education Service Supply Mechanism in line with the Coordinated Development of Population.”

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

Data will be provided under reasonable requirement by the corresponding author.

References

Brown

Souto-Otero

(2020). The end of the credential society? An analysis of the relationship between education and the labour market using big data. Journal of Education Policy, 35(1), 95–118.

Cohen

W. M.

Levinthal

D. A.

(1990). Absorptive capacity: A new perspective on learning and innovation. Administrative Science Quarterly, 35(1), 128–152.

European Commission. (2023). Joint Communication to The European Parliament, The Council, The European Economic and Social Committee and the Committee of the Regions. https://eur-lex.europa.eu/legal-content/%20EN/TXT/?uri=CELEX%3A52023JC0020&qid=1687525961309

Feldman

M. P.

(1999). The new economics of innovation, spillovers and agglomeration: Areview of Empirical Studies. Economics of Innovation and New Technology, 8(1-2), 5–25.

Glaeser

E. L.

(2018). Human-capital externalities in China (No. w24925). National Bureau of Economic Research.

Jianping

Xiaoling

(2019). Construction and application of evaluation index system for technology talent policy implementation in China. Human Resources Development of China, 36(4), 100–108. https://doi.org/10.16471/j.cnki.11-2822/c.2019.04.008

Guo

Zhang

(2023). Higher education expansion and graduate employment in China: A supply-demand mismatch perspective. Journal of Education and Work, 36(1), 1–20.

Jichan

(2013). Addressing the inefficacy of China's Bayh-Dole act equivalent: a path forward for university knowledge and intellectual property transfer. Intellectual Property, 23(3), 84–88.

Henry

(2008). The triple helix: University-industry-government innovation in action. Routledge.

10.

Hillebrandt

Huber

(2020). Editorial: Quantifying higher education: Governing universities and academics by numbers. Politics and Governance, 8(2), 1–5.

11.

Huai

(2024). Develop new quality productivity and enhance the cultivation of top innovative talents. https://www.news.cn/politics/20240309/7f5623edd2ed4d36abc9a5e23c5fc2d7/c.html

12.

Liu

Zhang

Dong

(2020). Understanding regional talent attraction and its influencing factors in China: From the perspective of spatiotemporal pattern evolution. PLoS One, 15(6), e0234856.

13.

Jia

Zhimin

Longhua

(2022). Problems and countermeasures of improving technology transfer of universities and research institutions in China. Bulletin of National Natural Science Foundation of China, 36(02), 309–315. https://doi.org/10.16262/j.cnki.1000-8217.2022.02.043

14.

Jiang

(2022). Mediating effects and moderating effects in causal inference. China Industrial Economics, 410(5), 100–120. https://doi.org/10.19581/j.cnki.ciejournal.2022.05.005

15.

Krücken

Götze

Seidenschnur

(2025). Lions, game players and niche providers: Universities' positioning in funding competitions in Germany. Higher Education Quarterly, 79(1), e70005.

16.

Leng

(2009). Scientific and technological investment regional difference and economic growth. Journal of Chongqing University of Posts and Telecommunications (Social Science), 21(3), 112–118.

17.

Liang

Zhengfeng

(2020). Research on the mechanism of “Talent Title” alienation. Forum on Science and Technology in China, 405(09), 125–132. https://doi.org/10.13580/j.cnki.fstc.2020.09.019

18.

Lin

Qiong

Wei

(2025). “Vacating the cage to change birds”: geographical distribution of higher education and its impact on local industrial structure upgrading. Peking University Education Review, 22(01), 41–59. https://doi.org/10.12088/pku1671-9468.202401003

19.

Liu

Jiahui

(2023). Scientific and technological innovation in universities and high-quality economic development - An empirical study based on the spatial Dubin model. Review of Evolutionary Economics and Economics of Innovation, 2, 74–91.

20.

Liu

(2024). New Production Relations driven by New Quality Productive Forces: Trends, challenges and countermeasures. China Finance and Economic Review, 13(4), 45–58.

21.

Ziang

Yuping

(2024). Levels of development of new quality productivity, regional differences and paths to enhancement. Journal of Chongqing University (Social Science Edition), 30(3), 1–17. https://doi.org/10.11835/j.issn.1008-5831.jg.2024.03.002

22.

Luo

Tiyan

Yujian

Yonghong

(2022). Analysis of employment destinations and characteristics of doctoral graduates in China: based on data from 2015 to 2020. Academic Degrees & Graduate Education, 409(01), 53–62. https://doi.org/10.16750/j.adge.2022.01.009

23.

Luo

Zhiqin

(2022). Research on the impact of scientific and technological innovation in universities on high-quality economic development - A sample data analysis of 27 provinces from 2009 to 2018. China University Science & Technology, 36(Z1), 29–34. https://doi.org/10.16209/j.cnki.cust.2022.z1.003

24.

Madsen

(2022). Governing by numbers and human capital in education policy beyond neoliberalism. Springer.

25.

Markham

S. K.

Ward

S. J.

Aiman-Smith

Kingon

A. I.

(2010). The valley of death as context for role theory in product innovation. Journal of Product Innovation Management, 27(3), 402–417.

26.

Miller

(2001). Governing by numbers: Why calculative practices matter. Social Research, 68(2), 379–396.

27.

Nie

(2024). Research on measuring the effect of high-quality economic development in the first round of “double first-class” construction - based on the perspective of knowledge spillover. Jiaoyu Guancha, 13(1), 1–6.

28.

Papadimitriou

Boboc

(2021). Conclusion: Re-imagining higher education institutions public mission initiatives through the lens of organizational perspective. In Re-envisioning Higher Education’s public mission: Global perspectives (pp. 275–290). Springer International Publishing.

29.

Piattoeva

Boden

(2020). Escaping numbers? The ambiguities of the governance of education through data. International Studies in Sociology of Education, 29(1–2), 1–18.

30.

Huang

(2023). Generation logic, theoretical innovation and time value of general Xi Jinping's important expostion on new quality productivity. Journal of Southwest University (Social Sciences Edition), 49(6), 1–11. https://doi.org/10.13718/j.cnki.xdsk.2023.06.001

31.

Ramirez-Cardona

Calderón-Hernández

(2024). Organizational coupling in higher education institutions: Challenges for university governance. International Studies of Management and Organization, 54(2), 105–119.

32.

Shi

Ling

(2024). Major strategic significance and implementation path of accelerating the formation of new quality productivity. Research on Financial and Economic Issues, 46(1), 3–12. https://doi.org/10.19654/j.cnki.cjwtyj.2024.01.001

33.

Tan

Yibei

(2015). Empirical analysis of the impact of government investment on innovation output of universities. Science and Technology Management Research, 35(20), 7–11. https://doi.org/10.3969/j.issn.1000-7695.2015.20.002

34.

Wang

Rongji

(2024). New quality productivity: Index construction and spatiotemporal evolution. Journal of Xi’an University of Finance and Economics, 37(1), 31–47. https://doi.org/10.19331/j.cnki.jxufe.20231124.001

35.

Wang

Yifei

(2023). The multiple paths of high-quality development of regional economy supported by higher education. Chongqing Higher Education Research, 11(2), 99–112. https://doi.org/10.15998/j.cnki.issn1673-8012.2023.02.010

36.

Wang

(2023). Accelerating solutions for bottleneck issues in university science and technology achievement transformation. Scitech China, 09, 41–44. 2023.

37.

Wang

Liu

Chen

(2022). Comparative study on the internal governance models of Chinese and European universities. Asia Europe Journal, 20(2), 115–135.

38.

Wang

Yini

(2024). Key issues and promotion strategies for the classified development of postgraduate education. China Higher Education, 60(05), 30–34.

39.

Weick

K. E.

(1976). Educational organizations as loosely coupled systems. Administrative Science Quarterly, 21(1), 1–19.

40.

Wei

(2023). The “New Washington Consensus” and the transition of US economic strategy. China Int'l Stud, 103, 92.

41.

Xie

(2023). A Re-recognition of the relationship between local higher education and socio-economic development. Jiangsu Higher Education, 39(3), 69–73. https://doi.org/10.13236/j.cnkij.she.2023.03.009

42.

Xie

Jianhua

Ying

(2024). The connotation, value and challenge of digital productivity. Studies in Dialectics of Nature, 39(6), 93–99. 1 0.19484/j.cnki.1000-8934.2023.06.012

43.

Zeng

(2024). Composite concept, development foundation and systematic innovation path of new productivity boosters. Journal of Shenzhen University (Humanities & Social Sciences), 41(2), 5–15.

44.

Zhang

(2025). Research on innovation of industry-education integration model in Chinese applied vocational universities. Pacific International Journal, 8(3), 110–116.

45.

Zhao

Lilan

(2009). The influence of university R&D expenditures on hi-tech firms formations and employment growth. Studies in Science of Science, 27(3), 374–380.

46.

Zhou

(2023). Classified management of colleges and universities in China: Source, practice, and reflection. Peking University Education Review, 21(2), 30–37. https://doi.org/10.12088/pku1671-9468.202302002

Does China’s University Science and Technology Investment Facilitates Its New Quality Productive Force Development Objectives?

Abstract

Keywords

Introduction

Literature Review and Hypothesis Proposal

Literature Review on the Relationship Between Higher Education and NQPF

Research on the Impact of University Scientific and Technological Investment on Regional Economic Development

Research Hypotheses

Research Design

Variables and Data Sources

Model Design

Empirical Analysis Results

Descriptive Statistics

Benchmark Regression

Robustness Test

Replacing Dependent Variable

Replacing Independent Variables

Exclusion of Outlier Years and Provinces Samples

Endogeneity Test

Mechanism Study

USTIs Can Stimulate Output of Scientific Research, Thereby Improving RNQPF

USTIs Can Expand the Scale of Talent Cultivation, But Not Necessarily Improving RNQPF

USTIs Can Promote Technology Transfer, But Not Necessarily Improving RNQPF

Per Capita USTIs Can Foster the Creation of High-Tech Industries, Thereby Improving RNQPF

Per Capita USTIs Can Attract High-Level Talents, But Not Necessarily Improving RNQPF

Discussions and Research Conclusions

Discussions

Research Conclusions

Footnotes

Acknowledgements

ORCID iD

Ethical Considerations

Informed Consent

Author Contributions

Funding

Declaration of Conflicting Interests

Data Availability Statement

References