The role of digital transformation of higher education institutions in promoting regional economic competitiveness

With its cross-border nature and resulting strong contextual attributes, the focus of research in higher education is constantly shifting with the economic development and industrial situation. Especially under the impetus of digital transformation, the construction of higher education cannot be limited to the self-organization and development of the institution as the core, but must take into account the other stakeholders it relies on, and integrate the synergistic and symbiotic relationship between multiple subjects. In this sense, university education is embedded in external social relationship networks and internal member relationship networks, and the impact of digital transformation of universities on the regional economy may be transmitted through various types of relationship embedding, which can be characterized in terms of multi-relationship embedding and synergistic symbiosis of subjects:

On the one hand, colleges and universities are created based on the logic of regional industries, and the digital economy industry is becoming a leading industry in economic development, which pushes the industrial colleges to closely follow the needs of the regional digital economy development, and to innovate the multidisciplinary cross-fertilization and cross-border integration of talent cultivation mode. Therefore, it is necessary to break the traditional college training model and organizational barriers, and to be more open, transparent and flexible in the organizational structure, which is also an inevitable requirement for the digital transformation of colleges and universities. Existing research suggests that the digital transformation of colleges and universities is a gradual process, which requires continuous adjustment of the existing model according to the environmental needs and the trend of technological paradigm leap, and completes the process of open adaptation and evolution from other organization to self-organization.

On the other hand, digital technology has become a visible factor in the development of society, promoting the overall transformation of university education in the direction of digitization and intelligence, and using digital technology to strengthen the strength of the connection between the government, enterprises, universities and research institutions and other subjects. Universities are further molded into a multi-symbiotic body, establishing joint cultivation of multiple resources, multi-channel collaboration, and co-management of innovative talent cultivation platforms. At the same time, multi-party quality resource sharing can also promote technology, talent and other resources from the “one-way flow” to the “two-way interaction” change to the specific industry chain, innovation chain as the cornerstone of the service, to create innovative technology and skill talent chain to support the development of the industry chain, the formation of multi-party participation, cooperation and win-win cooperation between universities and research institutions. Forming multi-party participation, win-win cooperation, and sustainable integration power. In addition, in the practical exploration of colleges and universities, it is also found that industrial colleges can accelerate the accumulation of data resources and improve the efficiency of interaction and use of data resources by building a link network of “industry-position-skill-research-teaching”. Enhance the interaction and efficiency of data resources, fully utilize the agglomeration and scaling effects of data capital, and realize the feeding effect of universities on the digital transformation of education.

In regional economics, regions are generally categorized into homogeneous, nodal, and planning regions according to the correlation between the components of the region in terms of their characteristics. Depending on the cohesion, regions can generally be classified into types such as district natural regions, economic regions, and administrative regions.

The selection of provincial administrative regions as the scale region of the study in this paper not only facilitates the operation in practice and meets the requirements of regional characteristics of intra-regional homogeneity and inter-regional heterogeneity, but also has important practical significance. Provincial regions’ vast market space, unique resource endowments, incipient industrial systems and the strong financial and administrative power of local governments in provincial regions, as well as their increasingly mature and effective regional economic regulation and control capabilities, make China’s provincial economic regions play an indispensable and crucial role in the development of China’s transition economy and emerging market economy. The development of provincial economic zones and inter-provincial economic zones plays an important role in promoting regional economic development, breaking down local protectionism, and optimizing social resources. The cultivation and enhancement of economic competitiveness at the provincial level is of great strategic significance in promoting China’s economic and social development and enhancing national competitiveness.

In view of the specific characteristics of provincial regions, this paper defines provincial regional economic competitiveness as the relative productivity level of enterprises and industries (clusters), and their regional dynamic capacity for continuous improvement, as demonstrated by a provincial region in order to realize sustainable, healthy and rapid regional economic development, and to compete with other provincial regions within the scope of China or in transnational economic regions.

The economic competitiveness of provincial regions is specifically expressed in the ability to attract resources from regions outside the province, the ability to realize the efficient transformation of resources into value, and the ability to optimize the allocation of internal and external resources. The foundation of attractiveness lies in the ability to efficiently transform resources into value. And the ability to convert resources efficiently relies on the ability to optimize the allocation of internal and external resources, which is the source of efficiency. The reconfiguration and sustained enhancement of this resource allocation and conversion capacity, in turn, relies on the rapid integration of internal and external resources and capacities within and outside the province by the dynamic capacity of the region.

The capability system of provincial regional economic competitiveness is shown in Figure 1, where the ultimate goal of wealth creation in the provincial region and improvement of the living standards of the population requires regional competitiveness, especially the attractiveness and market competitiveness of provincial enterprises and industries (clusters) to scarce resources to be realized. And in the competitive environment of the market economy, the attractiveness and market competitiveness of provincial regions to scarce resources will inevitably rely on regional operational capabilities to improve the efficiency of resource allocation and conversion, and furthermore, to obtain an increase in relative productivity to realize this.

Figure 1.

Capacity system of provincial regional economic competitiveness

The advantage in efficiency in a dynamic environment inevitably relies on the continuous improvement and innovation of operational capacity, which requires regional dynamic capacity to rapidly integrate internal and external resources and basic capabilities to reconfigure operational capacity according to the dynamic changes in the environment, and to continuously realize the innovation and enhancement of operational capacity. Regional dynamic capability is the ability of regional subject (mainly refers to the government) to reconstruct and innovate regional operation capability by integrating and configuring internal and external resources and capabilities according to the needs of environmental changes. Its formation depends on government capacity and the basic competitiveness of the region, which is determined by the components of regional competitiveness.

Measuring the digital transformation of colleges and universities is an extremely important part of the study, and accurately, comprehensively and objectively measuring the digital transformation index (UDT) of colleges and universities is the basis for exploring the impact of the digital transformation of colleges and universities on the competitiveness of the regional economy.

Currently, there are three main categories of methods used for quantitative measurement of digital transformation in universities: 1)

Comprehensive Indicator Evaluation Method

This method uses a number of digitization-related indicators to construct a multi-dimensional comprehensive evaluation index system, such as ICT investment, Internet penetration, the degree of informatized teaching, and informatized teaching infrastructure.

The measurement and evaluation of the Regional Economic Competitiveness Index (REC) can be done using factor analysis, which can be applied in two main ways: 1)

To seek the basic structure, simplify the observation system and reduce the number of variable dimensions.

Steps in the application of factor analysis:

There are n samples with p indicators each, so the original data matrix is obtained: 1X=[ X11X12⋯X1pX21X22⋯X2p⋯⋯⋯⋯Xn1Xn2⋯Xnp ]

The first step is to standardize the raw data to avoid the influence of indicator units and magnitudes.

The standardized conversion formula is: 2X′ij=Xij−Xj¯/Sj2

In the formula: 3Xj¯=∑i=1nXij/n ,(i=1,2,3,⋯,n;j=1,2,⋯,p) $$\overline {{X_j}} = \mathop \sum \limits_{i = 1}^n {X_{ij}}/n,\left( {i = 1,2,3, \cdots ,n;j = 1,2, \cdots ,p} \right)$$

Standardized X′ij Satisfaction: 4X¯′ij=0,Var(X′ij)=1

In the second step, the sample correlation matrix R is computed : 5R=[ r11r12⋯r1pr21r22⋯r2p……⋯⋯rn1rn2⋯rnp ]

The correlation coefficient r_ij in the matrix is calculated as: 6rij=∑k=1nX′kiX′kj/(n−1),(i=1,2,3,⋯,n;j=1,2,⋯,p)

In the third step, find the eigenvalues, eigenvectors and contributions of the correlation matrix R.

According to the characteristic equation |R – λI| = 0, find the p characteristic root λ_m(m = 1, 2, ⋯, p); according to the system of equations (R – λ_mI)U_m = 0, find the corresponding eigenvector U_m(m = 1, 2, ⋯, p) of λ_m.

In the fourth step, the cumulative variance contribution of the m factors was found: 7θ(k)=(∑m=1kλm)/∑m=1pλm

From θ(k) ≥ 85%, determine the k value.

In the fifth step, the k factor composite scores were calculated.

The factor variables were expressed as linear combinations of the original variables, i.e: 8Fni=∑j=1nUm⋅X′ij,(i=1,2,3,⋯,n;j=1,2,⋯,p)

The contribution of each factor was used as weights and a weighted sum was performed to obtain the composite value: 9Fi=∑m=1mβmFmi ,(βm=λm/∑i=1nλm) $${F_i} = \mathop \sum \limits_{m = 1}^m {\beta _m}{F_{mi}},\left( {{\beta _m} = {\lambda _m}/\mathop \sum \limits_{i = 1}^n {\lambda _m}} \right)$$

In recent years, in order to promote the integration of the digital economy and the real economy, promote the deep combination of digital technology and various industries, and realize the important strategic goal of China’s digital power, the national government has successively formulated relevant policies and measures, and made efforts to accelerate the digitalization of enterprises and improve the efficiency of transformation. The construction of digital China is a crucial factor in promoting Chinese-style modernization in the digital era, and promoting the digital transformation of education is a widespread trend. In the face of the new environment, digitalization provides an important boost for university education, providing momentum and maintaining the foundation.

Currently, research on the digital transformation of university education and the relationship between it and regional economic development focuses on the following three aspects: 1)

Development trend of educational activities in the context of digital economy

With the continuous maturation of digital technology, the continuous development of the digital economy and its deep integration with educational elements, the networked, intelligent, virtualized, cloud-based and decentralized features of educational and teaching activities have become increasingly obvious.

For colleges and universities, digital transformation is characterized by a long cycle, slow results and high risk, and in the initial stage, it is necessary to use a large amount of funds to purchase the necessary financial and human resources and other innovative elements. If we only rely on the market for resource allocation, it will be difficult to achieve breakthrough progress in digital transformation. The government, as the largest external stakeholder of universities, regulates the allocation of resources by giving government subsidies or services, which can alleviate the pressure on the funding of digital technology research and development, and enable universities to promote the process of digital transformation while guaranteeing their own profitability.

In summary, government grants can facilitate the process of digital transformation of universities and fully stimulate the competitiveness of the regional economy. Therefore, the following hypothesis is proposed:

H2, government grants play a positive moderating role in the process of universities’ digital transformation on regional economic competitiveness.

According to the first law of geography, the closer the distance, the stronger the interconnectedness between things. Regional economic competitiveness has different degrees of interdependence between regions, i.e. spatial autocorrelation [23]. The degree of global autocorrelation is measured by Moran, I with the following formula: 12I=m∑i=im∑j=1mwij(qi−q¯)(qj−q¯)∑i=1m∑j=1mwij∑i=1m(qi−q¯)2 where m is the number of spatial units in the study area; w_ij is the spatial weight matrix; q_i and q_j are the observed values for regions i and j, q¯=1m∑i=1mqi , respectively. The larger the absolute value of I, the stronger the dependence and spatial correlation of the economic attribute with the neighboring regions.

Spatial weights are used to define spatial distances between regions, which are also the basis for spatial measurement. The spatial distance can be set in various ways, such as neighboring distance matrix, geographic distance matrix and so on. In this paper, it is set as neighboring space weight w_ij, using “car” neighbor, i.e., if there is a common boundary between the regions, it will be recorded as w_ij = 1, otherwise, it will be recorded as w_ij = 0.

The results of the spatial Durbin model are computed using partial differentiation for the purpose of decomposing the spatial effects of the independent variables [24]. The basic Durbin model equation is now shifted: 13(Im−λW)y=αlm+Xγ+WXδ+ε 14y=∑σ=1q(Im−λW)−1(Imγσ+Wδσ)Xσ+(Im−λW)−1(αlm+ε) where I_m is the unit matrix, σ is the number of independent variables, and X_σ is each independent variable, where σ = 1, 2, ⋯, q. Now let: 15Rσ(W)=(Im−λW)−1×(Imγσ+Wδσ)

So get: 16(y1y2⋮ym)=∑σ=1q(Rσ(W)11Rσ(W)12⋯Rσ(W)1mRσ(W)21Rσ(W)22Rσ(W)2m⋮⋮⋮Rσ(W)m1Rσ(W)m1⋯Rσ(W)mm)×(x1σx2σ⋮xmσ)+(Im−λW)−1(αlm+ε) where R_σ(W)_ij = ∂y_i/∂x_jσ represents the indirect effect, i.e., the effect of the independent variable σ in city j on the dependent variable in city i. R_σ(W)_ij = ∂y_i/∂x_iσ represents the direct effect, i.e., the effect of the σ th independent variable for i cities on the dependent variable for i cities. The total effect for each independent variable is the sum of the two effects.

Due to the serious lack of relevant data in some prefecture-level cities, in order to ensure the availability and completeness of the data, 291 prefecture-level cities in China are finally taken as the research object, and the regional economic competitiveness promotion effect of digital transformation of universities is analyzed by combining relevant statistical data. Among them, see subsection 2.2 for the measurement methods of regional economic competitiveness and digital transformation of universities. Among the control variables, the number of financial institutions comes from the License Information Inquiry System of China Banking and Insurance Regulatory Commission (CBIRC), the newly registered industrial and commercial enterprises in prefecture-level cities come from Enterprise Search, and the high-speed railroad operation data come from the railroad passenger train timetable of past years. The rest of the data are from the National Bureau of Statistics, China Urban Statistical Yearbook, and relevant statistics of provinces, and the moving average method is used to fill in individual missing data.

The descriptive statistics of the main variables are shown in Table 1, including the explanatory variable of regional economic competitiveness REC_it, the average REC_it value of 291 prefecturelevel cities in China is 0.0091, and the maximum value is 2.3431, which is a significant gap. The core explanatory variables are, the mean value of UDT_it of digital transformation index of universities is 0.0581, the mediator variable is 0.0122 of government subsidy AD_it, and nine control variables.

The Hausman test of the regression model before conducting the baseline regression showed that a two-way fixed effects model was more appropriate. Also, standard errors for clustering at the city level are added to take into account the similarity of city groups. Table 2 reports the regression results of the benchmark model. Columns (1), (3), and (5) control only for the core explanatory variables and city and time fixed effects, while the rest report the results with the inclusion of all control variables.

The results from columns (1) and (2) show that the regression coefficients of the index of digital transformation of colleges and universities (UDT_it) are greater than 0 and pass the 1% significance level test, regardless of whether control variables are added or not. It indicates that the digital transformation of universities can significantly enhance the competitiveness of regional economy (REC_it), which verifies the hypothesis 1 proposed in the previous section.

In terms of sub-samples, models (3) and (4) belong to the resistance period, and the coefficient of digital transformation of colleges and universities is negative at the 5% significance level, which may be that the digital transformation of colleges and universities has not yet released the economic dividend during the resistance period (2008-2016), which has a negative effect on the resistance, which is related to the low level of digitalization of China’s colleges and universities at this time of development. While in the recovery period (2017-2023), the digital transformation of universities has a positive effect on regional economic competitiveness and passes the 1% significance level test, which indicates that the driving effect of the digital transformation of universities is gradually emerging, which relies on the rapid development of digital technology in this period. In conclusion, the benchmark results show that the digital transformation of universities and colleges can significantly improve the competitiveness of the regional economy, which again confirms Hypothesis 1 proposed in the previous section.

Regarding the mediating variables, in model (2), the regression coefficient between government subsidies (AD_it) and regional economic competitiveness (REC_it) is 0.6037 and positive at 1% significance level. This result indicates that the government accelerates the process of accelerating the digital transformation of universities in the form of subsidies, which indirectly promotes the competitiveness of the regional economy and verifies the hypothesis 2 proposed in the previous section.

In terms of control variables, both the level of economic development and the industrial structure are significant factors in enhancing regional economic competitiveness. And according to the proxy indicator of industrial structure, this essentially indicates that the optimization of non-agricultural industries further affects the quality of economic development, which coincides with the transformation of manufacturing and service industries empowered by digital talents. Higher wage levels also significantly contribute to regional economic competitiveness, possibly because higher wage levels contribute to a favorable employment climate, bringing about an increase in effective labor supply. Entrepreneurship, as a source of dynamism for industrial development and technological innovation, significantly enhances regional economic competitiveness, which may depend on the development of entrepreneurial innovation and entrepreneurship to provide new development potential for the region in the face of shocks. And the regression coefficient of the level of financial development is positive but not significant, partly because the current financial system of prefecturelevel cities is mostly insufficient, resulting in the role of finance has not yet been emphasized. It is worth noting that the regression coefficient of the government support variable shows negative, probably due to the fact that the local financial expenditure structure is not reasonable enough and the degree of intervention is high, which needs to change the function of government services. The resilience-enhancing effect of transportation infrastructure upgrading is not significant, probably because the economic effect brought by the opening of high-speed rail is absorbed by industrial structure and entrepreneurship. While the scope of regional administrative division can indeed significantly affect the enhancement of regional economic competitiveness, it may stem from the fact that the wider the economic hinterland is, the more it can buffer the adverse impacts of shocks. In conclusion, the digital transformation of universities has a significant impact on regional economic competitiveness, and the results are in line with expectations.

In this paper, the spatial correlation of the economic competitiveness of 291 prefecture-level cities in China is examined using Moran’s I index, as shown in Table 3. The value of Moran’s I index is in the range of [-1,1], and the value of Moran’s I index > 0 indicates the existence of a positive spatial correlation, while the value of Moran’s I index < 0 indicates negative spatial correlation, and the value of Moran’s I index = 0 indicates random distribution, and no correlation. Overall, from 2008 to 2023, the overall Moran’s I index of China’s 291 prefecture-level cities are all greater than 0, and all pass the 1% significance level test, indicating that the economic competitiveness of China’s 291 prefecture-level cities is not random. Rather, they have significant spatial correlation, and there is an obvious positive spatial correlation, but the value of Moran’s I index is not high, which indicates that the degree of spatial correlation still needs to be improved.

The global autocorrelation Moran’s I index can only reflect whether there is agglomeration or dispersion of economic competitiveness among China’s 291 prefecture-level cities as a whole, but it cannot show the specific location of the distribution of agglomeration. In order to visualize the spatial agglomeration status at a higher granularity, this paper further plots Moran’s scatterplot of the economic competitiveness of China’s 291 prefecture-level cities during the resistance period (2008-2016) and the recovery period (2017-2023) of the digital transformation of colleges and universities. The Moran scatterplot measures the degree of similarity and difference between different units and their neighbors in the global picture, and measures the internal structural correlation within the global picture.

The horizontal coordinates in the scatterplot represent the green economy competitive index of each city, and the vertical coordinates represent the spatial lagged values of green economy competitiveness. According to the high and low values of the observations, the spatial area is divided into four quadrants, which correspond to four different spatial agglomeration types:

The first quadrant is the “high-high” HH agglomeration type, i.e., the city and its surroundings are areas with high levels of economic competitiveness, with a positive relationship.

The second quadrant is the “low-high” LH agglomeration type, where cities with low levels of economic competitiveness are surrounded by other cities with high levels of economic competitiveness, with a negative relationship.

The third quadrant is the “low-low” LL agglomeration type, in which cities and their surroundings are areas with low levels of economic competitiveness, with a positive relationship.

The fourth quadrant is the “High-Low” HL agglomeration type, where cities with high levels of economic competitiveness are surrounded by other cities with low levels of economic competitiveness in a negative relationship.

The number of points distributed in the quadrants characterizes the direction and degree of spatial dependence, with more points in quadrants one and three representing a positive direction of spatial dependence, a positive correlation between units, and spatial similarity. The more points in quadrants 2 and 4 where low observations coexist with units in high observation areas, the more points in quadrants 2 and 4 characterize the direction and degree of spatial dependence, representing a negative direction of spatial dependence and a negative correlation between units.

The Moran’s I scatter of economic competitiveness of 291 prefecture-level cities in China during the resistance period (2008-2016) and the recovery period (2017-2023) of the digital transformation of higher education institutions is shown in Figure 2. It is found that although there is spatial agglomeration in the economic competitiveness of 291 prefecture-level cities in China, there are differences in the characteristic attributes, showing spatial heterogeneity. The distribution of the scatter plot gradually tends to be concentrated from dispersion, and from uneven scattering in the four quadrants to concentration in the “high - high” agglomeration type in the first quadrant and the “low - low” agglomeration type in the third quadrant. This indicates that the economic competitiveness of Chinese prefecture-level cities is highly dependent on their location.

Figure 2.

Moran’s I scatter plot of regional economic competitiveness

Spatially, there is more of a positive correlation, which means that there is a spatial mutual transmission mechanism between the levels of economic competitiveness of cities, and cities with abundant human resources, strong technological innovation capacity, and a high degree of digital transformation of colleges and universities may have a spillover effect on their neighboring cities. Cities have different geographic locations and degrees of development, and their resource endowments are characterized by regional characteristics, but the similarity of regional systems, cultures, and development concepts has been formed in China’s overall game of chess and long-term cooperation and development. Good location conditions have formed the commonality in resource endowment and natural and humanistic conditions among neighboring city clusters, which provides a convenient and free flow of information, capital and resources among cities, and promotes the diffusion of resource effects and spatial overflow within the city clusters.

The above regression results are the results of the spatial econometric model based on the spatial geographic distance weight matrix, and do not take into account that the spatial spillover effect of economic competitiveness will also be affected by other factors, such as economic activities, due to the fact that the smaller the economic disparity between regions is, the greater its spatial correlation. Therefore, based on the per capita GDP of 291 prefecture-level cities in China from 2008 to 2023, this paper constructs a spatial economic weight matrix to do a robustness test, and the results are shown in Table 4. Where models (1) and (4) are dynamic Durbin model estimation results based on spatial economic weight matrix, and models (2), (3) and (5) are spatial estimation results of static Durbin model based on spatial economic weight matrix, respectively. As can be seen from the table, after adopting the spatial economic weight matrix, the spatial term coefficients ρ or λ of both static and dynamic spatial econometric models are greater than 0, which are still significantly positive, and the estimated coefficients of the impact of the digital transformation UDT_it of universities on the regional economic competitiveness REC_it are significantly positive, which pass the 10% and 1% significance levels, respectively. And the dynamic SDM fit is optimal, although the size of the estimated coefficients varies. However, the direction and significance level did not change fundamentally, so the results of the study are robust.