Characteristics of student source and employment networks
Basic statistical characteristics
The number and percentage of students of “double first-class” universities in central China and the employment situation are analyzed in this statistical analysis, and the statistical results are shown in Table 2.
Table 2 highlights variations in enrollment, employment, and location preferences among graduates from central China’s ‘Double First-Class’ universities. Henan, Jiangxi, and Shanxi provinces prioritize local recruitment, with over 50% of students originating locally. Locally employed graduates exceed 30%, led by Henan at 66.10%, reflecting its strong local ties. Within provinces, graduates cluster in provincial capitals and metropolitan areas, such as Wuhan and Xiangyang in Hubei. Conversely, those seeking external employment migrate to economic hubs like Beijing, Shanghai, and Guangzhou, or coastal provinces such as Zhejiang and Jiangsu.
China’s university admission system allocates enrollment quotas provincially, compelling students to prioritize institutional prestige over geographic preferences. For example, Henan Province recruits 74.8% of students locally (Table 2), partly due to its large population and limited quota allocations for out-of-province universities. This quota system creates a path-dependent mobility pattern, where students from provinces with fewer top-tier universities are more likely to attend local institutions. Thus, while path dependence and meritocratic connections influence mobility, institutional constraints like admission policies play a pivotal role in shaping student source networks. Furthermore, the Gaokao score, which determines university admission, often limits students’ ability to choose their preferred study location, as they must select institutions where their scores are competitive. This systemic constraint reinforces the concentration of local students in central China’s ‘Double First-Class’ universities.
Inter-provincial mobility is overwhelmingly driven by disparities in economic opportunities. Guangdong, Shanghai, and Beijing attract 24.09% of graduates (Table 2) due to their robust job markets and higher wages, overshadowing environmental or infrastructural factors. The eastern coastal regions, with their advanced industrial bases and higher income levels, serve as magnets for graduates seeking better career prospects. For instance, Guangdong Province alone absorbed 15.86% of graduates in 2018, despite having fewer local student sources. This trend highlights the significant role of economic pull factors in shaping graduate mobility patterns. Moreover, the concentration of multinational corporations, tech hubs, and financial centers in these regions further enhances their attractiveness to highly educated talent.
Spatial patterns of mobility across provinces
Path dependence and meritocratic connection are crucial perspectives for understanding the spatial distribution of economic activities and talent mobility patterns. Path dependence underscores the influence of historical events, emphasizing how past decisions shape current outcomes. Meritocratic connectivity is an important mechanism for network evolution, emphasizing the Matthew effect, where new nodes will tend to establish connections with dominant nodes. From the perspective of meritocratic connection, talents are more inclined to choose to flow to cities with better economic development, public services, wage level and innovation environment. From the perspective of path dependence, the experience, knowledge and cognition that talents have accumulated about cities help them to build their own preferences for city choices. Both mechanisms emphasize cumulative advantage and self-reinforcement, with meritocratic connections focusing on the attributes of the target city and path dependence emphasizing the preferences of the talent itself, and the combination of the two mechanisms is more useful for understanding the network pattern of talent mobility.
According to the OD directed network matrix depicting student mobility from their hometowns to their study locations and subsequently to their employment destinations, the spatial patterns of employment and student origin from “double first-class” universities in central China are visualized in Figs. 2 and 3. Figures 2a and 3a specifically illustrate the inflow patterns of graduates employed from these universities to 31 provinces across China. Figure 2a reveals that graduates from the central China predominantly migrate to the eastern provinces, including Guangdong, Shanghai, Beijing, Zhejiang, Jiangsu, and Shandong, which collectively absorb 36.05% of these graduates. Notably, Guangdong attracts the largest proportion of graduates for employment, although its share has declined slightly over the past three years: 15.86% in 2018, 14.68% in 2019, and 13.66% in 2020. Shanghai, Beijing, and Zhejiang also emerge as highly attractive destinations, accounting for 4.68%, 4.69%, and 5.18% of total inflows over the three years, respectively, while fewer graduates opt for employment in other provinces. The career choices of college graduates are influenced by their place of origin, as well as characterized by path dependence and meritocratic principles. Path dependence suggests that graduates tend to seek employment locally or return to their hometowns after studying elsewhere for an extended period. Meritocratic connections, on the other hand, drive graduates to cities offering favorable urban environments and higher incomes. In terms of inter-provincial mobility, meritocratic connections exert significant influence, particularly evident in cities like Guangdong, Beijing, and Shanghai, which attract large numbers of graduates despite having fewer student sources locally. Figure 3a depicts the employment flow network of graduates, highlighting how the capital city of the central China serves as a hub, absorbing local graduates while also sending them to economically developed eastern coastal regions such as Beijing, Tianjin, Hebei, the Yangtze River Delta, and the Pearl River Delta.
The proportion of graduates in employment and students.
Cross-provincial flow pattern of “double first-class” universities graduates in central China.
Figures 2b and 3b show the pattern of student mobility of “double first-class” universities in the central China. Predominantly, students originate from within the region itself, with the six central provinces contributing 64.09% of these students. This is primarily due to these universities’ significant enrollments within their respective provinces, coupled with targeted recruitment efforts in neighboring provinces. For instance, Wuhan University enrolls approximately 4200 students annually from Hubei, accounting for 30% of its total intake that year, and admits over 1200 students from Henan annually. Supported by the Central and Western Region Enrollment Collaboration Program, provinces within the Central China maintain robust inter-provincial student exchanges, ensuring a substantial and stable flow of enrollments within the region. Beyond the central China, provinces such as Hebei, Guangdong, Guangxi, Zhejiang, Shandong, and Jiangsu constitute the primary sources of students for “double first-class” universities in the central China. Over the past 3 years, these external contributions have remained relatively consistent. This regional enrollment strategy results in a predominant reliance on local student sources within the central China, leading to a decreased intensity in student flows from the central provinces to surrounding areas. The intensity of student flow shows a decreasing state from the central China to the surrounding provinces and cities, generally showing a swirling network layout with a dense center and sparse surrounding.
Spatial pattern of intra-provincial mobility
While eastern provinces like Jiangsu and Zhejiang have denser economic networks, central China hosts significant higher education resources. For instance, Hubei Province, home to 7 ‘Double First-Class’ universities, surpasses Guangdong and Fujian in institutional count. However, graduates still gravitate eastward due to better career prospects. This paradox highlights the disconnect between educational resources and economic opportunities in central China. Despite having a robust higher education system, the region struggles to retain talent due to the limited availability of high-paying jobs and advanced industries. This trend underscores the need for central China to align its educational strengths with economic development strategies to enhance its attractiveness to graduates. For instance, the proportions of college graduates opting to stay in their respective provinces and cities for employment in Hubei, Hunan, Henan, Anhui, Shanxi, and Jiangxi were 33.37%, 36.55%, 66.10%, 44.05%, 37.59%, and 41.01% respectively. There is a big gap with the average proportion of college graduates in East China who can be employed in the provincial area of more than 60%.
To illustrate the hierarchical flow of graduates within the Central China’s municipalities, the distribution of graduates from colleges and universities across various municipalities within the provinces was analyzed.
As depicted in Fig. 4a, Wuhan, Changsha, Hefei, Zhengzhou, Kaifeng, and Taiyuan emerge as focal points for intra-provincial mobility of graduates within the central China. Notably, Zhengzhou in Henan stands out with over 60% of graduates opting to stay within the province, particularly concentrated in the cities of Zhengzhou and Kaifeng. This reflects a distinct hierarchical distribution of employment opportunities within the province. Beyond the provincial capitals mentioned, their surrounding cities also present significant employment prospects. In Anhui, for example, with Hefei as the center, the ratio of the number of graduates flowing in from surrounding cities such as Wuhu, Lu’an and Fuyang all exceeded 2.5%. Similarly, in Hunan, Changsha acts as a major hub, with cities such as Zhuzhou, Yueyang, Changde, and Huaihua also drawing considerable numbers of graduates. Henan, however, exhibits a unique pattern: Zhengzhou, as the provincial capital, attracts a substantial 38.91% of graduates, while Kaifeng, its secondary center, also garners significant inflows at 19.79%. Together, these cities create a multi-core spatial pattern within Henan province.
Flow pattern of graduates in the province.
Figure 4b illustrates the network pattern of graduate flows from universities in the central China to cities within their respective provinces. Notably, all universities, except in Hunan where Xiangtan and Changsha host four “double first-class” institutions, are situated in provincial capital cities. For example, Xiangtan University has a large flow to Changsha City, while combined with Fig. 4a, staying in Xiangtan for employment accounts for a smaller percentage of the total, which shows that it exports less graduate flow to the surrounding cities, and that the capital city of the province exhibits a stronger polarization effect. Overall, the graduate flow within the Central China forms a spatial pattern radiating outward from provincial capitals. Intra-provincial mobility itself is based on the premise of path dependence compared to inter-provincial mobility, but there is also a meritocratic connection to cities as graduates choose to stay in the Central China for employment.
Path dependence is evident in the concentration of universities in provincial capitals, which historically serve as economic hubs with accumulated development over centuries. These cities offer well-established knowledge, perception, and experiences, making them the preferred employment destinations for local residents. Consequently, provincial capitals generally boast favorable geographical locations and wage levels, making them ideal choices for both local and non-local graduates seeking optimal career opportunities.
Analysis of graduates stickiness rate
Sticky rate analysis
Using the neighboring space weight matrix based on the Rook contiguity principle for the common boundaries of the 31 provinces, and conducting a spatial autocorrelation analysis of the viscosity rate of graduate inflow from the Central China to these areas, we obtained a global Moran’s I index of 0.27 and a Z score of 2.73, with a P value of < 0.01. This indicates a spatial correlation in the employment location stickiness rate of graduates, as confirmed by a one-tailed normal distribution hypothesis test. Table 3 presents the LISA cluster types for the stickiness rate in each province. Anhui, Jiangxi, and Hunan in the central China fall into the “high-high” cluster, with stickiness rates above 3.5%. Conversely, the three northeastern provinces and the northwestern region are in the “low-low” cluster, with stickiness rates below 1.7%. Fujian, on the right side of Central China, belongs to the “low-high” zone and has a lower inflow of graduates compared to other provinces in the region. Table 3 does not display any “high-low” clusters, suggesting a more concentrated and uniform graduate flow in the central China. Although Guangdong, Shanghai, and Beijing have high stickiness rates, their spatial aggregation effects are not significant, and their stickiness rates are spatially isolated.
Analysis of net sticking rate
This paper calculates the net sticking rate of employment by province, defined as the difference between the percentage of graduates entering a province for employment and the percentage originating from that province. This rate reflects the attractiveness of a province for university graduates from the central China. A positive net sticking rate indicates that a province not only retains local graduates but also attracts additional talent, while a negative net sticking rate indicates a loss of graduates to other regions. Overall, Guangdong has the highest net sticking rate at 12.78%, followed by Shanghai and Beijing, with net sticking rates of 4.47% and 4.25%, respectively. Other provinces with relatively high net sticking rates include Zhejiang (2.86%), Jiangsu (1.91%), and Hubei (1.40%). Hubei is the only province in the central China with a high stickiness rate, while Hunan has a positive but low net sticking rate (0.26%). The other provinces in the central China are experiencing a brain drain, with Henan (− 9.08%), Shanxi (− 4.14%), Jiangxi (− 2.43%), and Anhui (− 1.50%) all having negative net sticking rates.
Based on the changes in the net sticking rate over the years 2018, 2019, and 2020, provinces within the central China exhibit varying trends. Henan, Hunan, and Anhui show a positive upward trend in their net sticking rates. Conversely, Hubei, Jiangxi, and Shanxi display a downward trend, indicating a weakening in their ability to attract and retain talent. For other provinces, first of all, East China has a high level of attractiveness to graduates from Central China while its net sticking rate also varies, such as Guangdong, Beijing, Shanghai, Tianjin, etc. The net sticking rate has slightly decreased, while the net sticking rate in Zhejiang, Jiangsu, Shandong, etc. has an upward trend. For the western provinces, the implementation of the Western Development Strategy and initiatives encouraging graduates to work in the west have provided some support to their net sticking rates. Notably, the Tibet Autonomous Region has a positive net sticking rate, and other western provinces have also attracted some graduates from the central China, though their net sticking rates remain relatively low. Despite policies aimed at revitalizing the old industrial zones, the three northeastern provinces continue to struggle with low net sticking rates, indicating a significant insufficiency in their attractiveness to graduates from the central China.
Comparing Table 3 with the LISA aggregation type table of stickiness rate reveals that the central China shows “high-high” aggregation in terms of stickiness rate. This is because the central China retains more graduates due to the large number of local students. However, when focusing on the net stickiness rate, the central China does not exhibit the same level of competitiveness despite its quantitative advantage (Fig. 5). The central China’s lower net stickiness rate is attributed to the economic development, extensive job markets, and higher salary levels in Guangdong, Shanghai, and Beijing. These regions attract graduates from across the country and have fewer local students, resulting in higher net stickiness rates. Consequently, according to formula (2), the Central China’s net stickiness rate is low and often negative, indicating a loss of graduates, while the eastern coastal region has positive and high net stickiness rates.
Annual change in net viscosity rate.
