The Impact of Higher Education and Human Capital Quality on “Local- Neighborhood” Economic Growth Ran Zhao, Yuhong Du Beijing Normal University, Beijing 100875, China Abstract. Based on China’s provincial panel data from 1990 to 2017 and the improved Lucas, Nelson & Phelps model, the Spatial Dubin Model is used to test the spatial effects of higher education and human capital quality. The results showed that high-level human capital, characterized by higher education and urban labor income index, indirectly promoted local economic growth through technological innovation. There was also a “local-neighborhood” synergy effect. The neighborhood effect was manifested in that it affected the economic development of neighbors by promoting technological catch-up. After considering the quality fac- tor, both the local and neighborhood effects were enhanced. From a re- gional perspective, higher education in the Yangtze River Delta, where the level of economic development is relatively high, was manifested as a spatial spillover effect of technological innovation and the neighborhood effect in the northeastern Bohai Rim and the Pearl River Delta was man- ifested as a technological catch-up. Best Evidence in Chinese Education 2021; 8(1):1041-1057. Doi: 10.15354/bece.21.ar22. How to Cite: Zhao, R., & Du, Y. (2021). The Impact of higher education and human capital quality on “local-neighborhood” economic growth. Best Evidence in Chinese Education, 8(1):1041-1057. Keywords: Higher Education, Human Capital Quality, Economic Growth, Spatial Dubin Model Zhao & Du. Higher Education, Human Capital Quality and “Local-Neighborhood” Economy. Vol.8, No.1, 2021 1042 About the Authors: Ran Zhao, Lecturer, Faculty of Education/Institute of Educational Economics, Beijing Normal University, Beijing 100875, China. Email: zhaoran@bnu.edu.cn Correspondence to: Yuhong Du, Professor, Faculty of Education/Institute of Educational Economics, Beijing Normal University, Beijing 100875, China. Email: dyh@bnu.edu.cn Funding: This study was supported by the Ministry of Education’s Philosophy and Social Science Major Research Project (15JZD040) and the Central Government’s Special Fund for Basic Scientific Research in Universities (2019NTSS06). Conflict of Interests: None. © 2021 Insights Publisher. All rights reserved. Creative Commons Non Commercial CC BY-NC: This article is distributed under the terms of the Crea- tive Commons Attribution-NonCommercial 4.0 License (http://www.creativecommons.org/licenses/by- nc/4.0/) which permits non-commercial use, reproduction and distribution of the work without further permission provided the original work is attributed by the Insights Publisher. Zhao & Du. Higher Education, Human Capital Quality and “Local-Neighborhood” Economy. Vol.8, No. 1, 2021 1043 Introduction ROM 1978 to 2010, the Chinese economy maintained an average annual growth rate of more than 10%, creating a growth miracle that attracted worldwide atten- tion. After 2010, the growth rate has gradually slowed down, and the GDP growth rate from 2017 to 2019 was below 7%. The economy has transformed from high-speed growth in the past to medium- and high-speed growth. China’s “demograph- ic dividend” advantage is gradually diminishing. Labor shortage and population aging have gradually become key factors restricting economic growth. As a comprehensive indicator to measure population size, structure, and quality, human capital determines national and regional economic development speed and quality. From 1982 to 2017, the proportion of the population with an associate degree education and above in the Chi- nese labor force showed a significant upward trend increased from 0.9% in 1982 to 18% in 2017. The decline in the number of labor forces will not be conducive to economic growth, but can the improvement in labor quality make up for the growing gap caused by the decline in the number of labor forces? What is the mechanism and realization path of higher education’s influence on regional economic growth? Does this impact show spatial dependence characteristics among regions with significant differences in population size, education quality, and economic development? Previous studies adopted the average years of education or the total years of education, which could not reflect the change in the quality of the labor force. Our study used the revised Lucas and Nelson & Phelps models for spatial econ- ometric analysis by constructing two data sets of labor income index human capital and education human capital (Lucas, 1988); Nelson & Phelps, 1965). This helps to under- stand the spatial mechanism of higher education and human capital quality from a deep- er level and provides a decision-making reference for implementing the coordinated development of China’s higher education and regional economy to a certain extent. Literature Review The Mechanism and Direction of Human Capital in Economic Growth Research under the macro-framework mainly focuses on the mechanism of education and human capital in economic growth and the direction and significance of human cap- ital in economic growth. There are primarily the following views on the mechanism: Firstly, human capital directly participates in the production process as a production factor that improves the quality of labor, and the accumulation of specialized human capital is the decisive factor in promoting sustained economic growth (Lucas, 1988). Secondly, human capital promotes the indirect effects of economic growth by fostering technological innovation and technological catch-up, emphasizing that the stock of hu- man capital is a critical factor in economic growth (Romer, 1990; Ding & Knight, 2011; F Zhao & Du. Higher Education, Human Capital Quality and “Local-Neighborhood” Economy. Vol.8, No. 1, 2021 1044 Li et al., 2016). Thirdly, human capital directly affects output as a factor and indirectly promotes a joint mechanism of economic growth by promoting technological innova- tion and imitating absorption (Engelbrecht, 2003; Du et al., 2014). Fourthly, from a re- gional perspective, some studies have confirmed that education and human capital have positive spatial externalities (Chang & Zhao, 2017; Deng & Ke, 2020), but some studies have reached the opposite conclusion (Fischer et al., 2009). It can be seen that the mechanism of different levels of education and human capital in space still needs to be further tested. “Stock” and “Quality” in Human Capital Measurement In the research on the relationship between education, human capital, and economic growth, a vital issue is distinguishing between “stock” and “quality” in human capital. The more mature human capital measurement methods mainly include cost method, income method, and education characteristic method. The main idea of the cost method is to compare the method of measuring physical capital. Most scholars estimate the stock of human capital in various countries or different periods based on perpetual in- ventory technology (Schultz, 1961; Kendrick, 1976; Qian, 2012; Meng & Wang, 2014). Previous studies on the impact of education on economic growth have focused more on quantitative educational indicators. The commonly used proxy variables include the average years of workers’ education (Psacharopoulos & Patrinos, 2004; Barro & Lee, 1993) and the adult literacy rate (Cai & Du, 2000). Studies have also shown that once the factors of education quality (international test scores, etc.) are considered, the influ- ence of the quantity of education becomes insignificant. In contrast, the quality of edu- cation has a strong positive effect on economic growth (Jamison et al., 2007; Hanushek & Woessmann, 2011). Still, the use of international test scores cannot directly measure the human capital level of the working-age population (Graham & Webb, 1979). There are two ways to use the income method to calculate human capital: one is to use the sum of the present value of future income to measure human capital (Jorgenson & Fraumeni, 1989; Li & Tang, 2015; Dong, 2017). However, it is impossible to eliminate the influence of the annual increase in physical capital on the stock of human capital. The other is to estimate human capital by using the present value income of workers. The most representative is LIHK (Labor-Income-Based Human Capital) labor income method (Mulligan & Sala-i-Martin, 1997). It reflects the difference of different educa- tion levels, the change of education quality over time, and factors such as work experi- ence on human capital accumulation. Simultaneously, it effectively eliminates the influ- ence of physical capital on the measurement, which is more reasonable for analyzing economic growth (Zhu & Xu, 2007; Liang et al., 2015). Based on the concept of unit human capital in the income method, this research constructs two types of panel data: human capital index and educational human capital, including scale, structure, and quality. By establishing a spatial measurement model, the spatial effect mechanism of higher education and human capital quality in economic growth is tested under the premise of fully considering spatial and geographical factors. Zhao & Du. Higher Education, Human Capital Quality and “Local-Neighborhood” Economy. Vol.8, No. 1, 2021 1045 Human Capital Measurement Based on LIHK Income Method According to China’s primary data, we first need to estimate the annual population data according to gender, age, urban and rural areas, and education level. Secondly, we need to evaluate the income of each part of the group. Using the method of the China Human Capital and Labor Economics Research Center (CHLR), we estimated the expanded Mincer income equation (1): (1) Where ln(wage) is the natural logarithm of income. Sch represents the number of years of education. Exp is an individual’s work experience. ave_gdp represents the per capita GDP of the province. indus_gdp indicates the proportion of tertiary industry output value in GDP in the province where it is located. ave_wage represents the aver- age wage level in the province. According to the two databases of CHNS (China Health and Nutrition Survey, 1989-2015) and CFPS (China Family Panel Studies, 2010-2016), the weighted results based on the sample size yield the intercept term, Sch, Exp, and the coefficient of Exp 2 . Then, the linear fitting of the variable to the time trend is made, and the fitting value of the missing year is obtained. Then the human capital index of the individual by sex, age, urban and rural, and education level was calculated. The total human capital stock is obtained by adding up the number of labor forces in each group. Construction of Spatial Panel Data Measurement Model Research Method The two basic setting methods for spatial interaction are: one is Spatial Lag Model (SLM) (Lagged Dependent Variable), and the other is Spatial Error Model (SEM) (The error term contains a Spatial Autoregression process). Spatial Dubin Model (SDM) is a general form of the spatial measurement model and a broader SLM and SEM measure- ment model. Among them, yit is the explained variable, xit is the explanatory variable, and the primary expression of the model is: SLM: (2) Zhao & Du. Higher Education, Human Capital Quality and “Local-Neighborhood” Economy. Vol.8, No. 1, 2021 1046 SEM: (3) SDM: (4) This study draws on the model of Lucas and Nelson & Phelps (Lucas, 1988; Nelson & Phelps, 1965), combined with the improved research framework of Yuhong Du et al. (Du & Zhao, 2018), to test the spatial effect mechanism of higher education and human capital quality on economic growth. We hypothesized that different levels of human capital have different growth paths: basic human capital influences growth by increasing factor accumulation. In contrast, advanced human capital promotes growth by promoting technological innovation and imitating catch-up. We set the production function to the following form: (5) Y represents output, K represents capital, Ha represents high-level human capi- tal, Hb represents basic human capital, L represents labor force. A means technological progress rate or total factor productivity. Drawing on the study of Nelson & Phelps (Nelson & Phelps, 1965), Ha repre- sents technological innovation and means technological imitation and catch-up, as shown in equation (6): (6) Take the per capita form and digitize both sides of equation (5) at the same time, and add the spatial effect after substituting equation (6) to obtain the spatial panel model (7): Zhao & Du. Higher Education, Human Capital Quality and “Local-Neighborhood” Economy. Vol.8, No. 1, 2021 1047 (7) Among them, the explained variable is represented by GDP per capita, and the explanatory variables include physical capital (k), number of laborers (L), basic human capital (hb), advanced human capital (ha), technological imitation, and catch-up (Catch), representing a series of Control variables include foreign trade dependence (open), in- dustrial structure (indus), and government support (gov). c, αi, and εit represent a con- stant term, a regional fixed effect, and a random error term, respectively. wij represents the elements of the spatial weight matrix. This study uses three types of matrices for analysis. The adjacency weight matrix W1 is constructed by studying whether the varia- bles are adjacent to each other by assigning values of 0 or 1. If i = j, it means that prov- ince i and j are adjacent; if i ≠ j, it means that province i and j are not adjacent. (8) Besides, in order to test the robustness of the estimation results, we also con- sider the attenuation of the spatial effect as the distance increases and construct the geo- graphical distance weight matrix W2: (9) At the same time, to objectively express the spatial correlation of the economic development level of the spatial unit, we also construct the economic distance weight matrix W3, where i is the average GDP of the region i over the years, and is the av- erage GDP of the full sample. Zhao & Du. Higher Education, Human Capital Quality and “Local-Neighborhood” Economy. Vol.8, No. 1, 2021 1048 (10) Data Sources and Variables According to the availability of samples, this paper selects the panel data of 30 provinc- es from 1990 to 2017 as the research sample (the Tibet Autonomous Region has been eliminated due to excessive data missing). The data we used to measure LIHK’s human capital came from “China’s 1982 Census”, “China’s 1990 Census”, “China’s 2000 Cen- sus”, “China’s 2010 Census”, “China Demographic Yearbook (1991-2018)”, the pro- vincial census data and statistical yearbooks, the China Nutrition and Health Survey (CHNS, 1989-2015) and the Chinese Family Panel Studies (CFPS, 2010-2016) micro- database; the original data of other variables were from the “China Statistical Yearbook” (1991-2018)”, “China Labor Statistics Yearbook (1991-2018)” and statistical yearbook of each province.  Output Level (Y/y) The output level reflects the economic development status of a province and is ex- pressed by the province’s gross regional product (Y) and per capita gross regional prod- uct (y).  Capital Stock (K) / Capital Stock per Capita (k) As there are currently no official statistics on annual physical capital stock data, the most commonly used calculation method is the Perpetual Inventory Method. The inter- provincial material capital stock data used in this article came from Holz and Dr. Sun Yue (Holz & Sun, 2018).  Human Capital We measured human capital in two forms: years of education and LIHK Human Capital Index. Advanced human capital (ha) is equal to the sum of years of education of all workers with higher education, and basic human capital (hb) is equal to the sum of years of education of all workers with elementary, middle, and high school education. This study also measured the human capital index based on the LIHK labor income method, reflecting the comprehensive effect of labor force size and quality. Considering the conditions in rural areas in China, they usually only have limited opportunities for higher education, and it is difficult for them to have the opportunity to obtain advanced skills from their jobs. Therefore, the main reason for the urban-rural income gap can be attributed to the difference in advanced human capital levels. The measured rural labor income index and urban labor income index were used as substitute variables for basic (hb_LIHK) and advanced human capital (ha_LIHK). Zhao & Du. Higher Education, Human Capital Quality and “Local-Neighborhood” Economy. Vol.8, No. 1, 2021 1049  Other Control Variables Foreign trade dependence (open), industrial structure (indus), and government support (gov) respectively used the proportion of total regional imports and exports to GDP, the ratio of the added value of the tertiary industry to GDP, and the balance of government public budget fiscal expenditures in GDP. Empirical Result Analysis Preliminary Test of Spatial Effect Model Before estimating the model, it was necessary to test whether there was a spatial effect between the quality of higher education and human capital and economic growth (shown in Table 1). Both the Lagrange multiplier test and robust test results rejected the null hypothesis that there was no spatial lag effect or spatial error effect. It showed that there was a significant spatial dependence characteristic between higher education, hu- man capital quality, and economic growth in various regions. The Wald statistics of spatial lag and spatial error were both significant at the 1% statistical level, indicating that the Spatial Dubin Model’s null hypothesis could be transformed into a spatial lag or error model should be rejected, so it was more appropriate to establish the Spatial Dubin Model. Simultaneously, the Hausman test result passes the 1% significance test, and a fixed-effect Spatial Dubin Model (SDM-FE) should be established. Considering that China’s education, human capital level, and economic growth all present significant imbalanced characteristics, the mixed effects and time fixed effects ignore regional structural economic differences. In contrast, the space-time double fixed-effect model can simultaneously consider regional spatial differences and period effects, further ef- fectively distinguish the role of spatial dependence from the impact of spatial heteroge- neity and omitted variables. Studies have pointed out that the parameters obtained using Maximum Likelihood Estimation (MLE) were asymptotically effective (Chen, 2014). Therefore, we used the Spatial Dubin Model with dual fixed MLE time and space. Model Estimation Results The estimation results obtained under the three spatial weight matrices of the national adjacency matrix, inverse distance matrix, and economic distance matrix were relatively stable (shown in Table 2). Most of the coefficient estimates had not changed signifi- cantly, indicating the robustness of the estimated results. Simultaneously, the best fit of the model was assessed under the economic distance matrix, indicating that the spatial effect of education and human capital quality on economic growth was more reflected in economic spatial connections. Therefore, in the following, we mainly analyzed the economic distance matrix W3. The estimated values of advanced human capital (varia- bles ha and ha_LIHK) were both positive at the 1% significance level, indicating that advanced human capital represented by higher education (ha) could significantly pro mote regional economic growth through technological innovation effects and urban Zhao & Du. Higher Education, Human Capital Quality and “Local-Neighborhood” Economy. Vol.8, No. 1, 2021 1050 Table 1. The Spatial Effect Test of Higher Education and Human Capital Quali- ty in Economic Growth. Test W1 W2 W3 Lagrange Multiplier Test (Spatial Lag) 7.857*** (0.005) 9.529*** (0.002) 4.279** (0.039) Robust Lagrange multiplier test (Spatial Lag) 7.047*** (0.008) 8.746*** (0.003) 4.052** (0.044) Lagrange Multiplier Test (Spatial Error) 16.072*** (0.005) 8.031*** (0.005) 3.231* (0.072) Robust Lagrange multiplier test (Spatial Error) 15.262*** (0.005) 7.248*** (0.007) 3.004* (0.082) Wald Test (Spatial Lag) 83.06*** (0.000) 240.39*** (0.000) 100.24*** (0.000) Wald Test (Spatial Error) 82.52*** (0.000) 154.71*** (0.000) 53.03*** (0.000) Hausman Test 346.48*** (0.000) 34.81*** (0.000) 95.33*** (0.000) ***, **, * Represent significant at the statistical level of 1%, 5%, and 10%, respectively. The P value of the coef- ficient is in parentheses. Table 2. Robust Estimation Results of Spatial Dubin Model (SDM). Variable Educational Human Capital LIHK Human Capital SDM-FE SDM-FE SDM-FE SDM-FE SDM-FE SDM-FE Adjacency Weight Matrix Geographic Distance Matrix Economic Distance Matrix Adjacency Weight Matrix Geographic Distance Matrix Economic Distance Matrix Capital Stock 0.3169*** (0.0107) 0.3090*** (0.0112) 0.3364*** (0.0113) 0.2983*** (0.0113) 0.2926*** (0.0122) 0.3059*** (0.0119) Number of Labor -0.7215*** (0.0314) -0.7326*** (0.0291) -0.6796*** (0.0297) -1.0331*** (0.0461) -0.9616*** (0.0437) -0.8552*** (0.0393) Basic Human Capital 0.5145*** (0.0929) 0.3112*** (0.0856) 0.5216*** (0.0855) 0.1160*** (0.0195) 0.0444** (0.0187) 0.0407** (0.0187) Advanced Human Capital 0.1196*** (0.0156) 0.0832*** (0.0143) 0.1461*** (0.0140) 0.4696*** (0.0580) 0.4945*** (0.0462) 0.5389*** (0.0537) Technology Catch-up -0.0383*** (0.0042) -0.0335*** (0.0037) -0.0394*** (0.0044) -0.1411*** (0.0306) -0.1667*** (0.0219) -0.2120*** (0.0261) Spatial effect of Capital Stock 0.0036 (0.0239) 0.0613** (0.0310) 0.1187** (0.0466) -0.0328 (0.0247) 0.0036 (0.0333) 0.2024*** (0.0505) Spatial effect of Labor Number 0.4491*** (0.0566) 0.6311*** (0.0922) 0.1447 (0.1304) 0.8156*** (0.1055) 0.9434*** (0.0994) 0.4077** (0.1333) Spatial Effect of Basic Human Capital -0.3800** (0.1619) 0.6952*** (0.1904) -0.7506** (0.2837) -0.2086*** (0.0382) -0.1922*** (0.0487) -0.1422** (0.0659) Spatial Effect of Advanced Human Capital -0.0658** (0.0297) 0.0785** (0.0322) -0.1696*** (0.0509) -0.4810*** (0.1240) -0.1513 (0.1273) -0.2338 (0.1957) Spatial Effect of Technology Catch-up 0.0483*** (0.0080) 0.0492*** (0.0088) 0.1105*** (0.0121) 0.1479** (0.0564) 0.2158*** (0.0528) 0.2922*** (0.0744) Sample Size 840 840 840 840 840 840 R2 0.311 0.366 0.470 0.164 0.294 0.367 ***, **, * Represent significant at the statistical level of 1%, 5%, and 10%, respectively. This table only reports the estimated results of the core explanatory variables, and the complete results can be obtained from the au- thor. Zhao & Du. Higher Education, Human Capital Quality and “Local-Neighborhood” Economy. Vol.8, No. 1, 2021 1051 labor quality factor (ha_LIHK) included in the income index had a more substantial impact on local economic growth. The estimated value of the spatial effect of advanced human capital is negative, and it is only significant under the human capital index measured by education level. The estimated value of the spatial impact of technological catch-up is positive at the 1% significance level, indicating that although the flow of technology and personnel and other factor resources have produced spatial spillover effects, the spillover effects of high-level human capital are mainly realized as techno- logical imitation and catch-up. Since the estimation results of the SDM model cannot fully reflect the relationship between the explained variables and the explanatory varia- bles, according to the research of Lesage and Pace (LeSage & Pace, 2009) , the use of partial differential methods could better deal with the error of spatial spillover effect estimation. The partial differentiation method decomposed the Spatial Dubin Model (SDM) total spatial spillover effects into direct and indirect effects. The total effect was the sum of the immediate impact and the indirect effect. Table 3 reports the estimated results of the local direct effects, the indirect im- pact of neighbors, and the total effects of higher education and human capital quality under the economic distance matrix. Different levels of human capital have other spatial mechanisms for economic growth. Among them, essential human capital directly af- fects local economic growth through the production of final products, but it will inhibit the economic growth of neighboring provinces. The local effect coefficient of high- level human capital, characterized by higher education and urban labor income index, which promotes economic growth through technological innovation, is significantly positive, confirming the mechanism of high-level human capital indirectly promoting regional economic growth through technological innovation. And for every year of higher education years, the growth effect brought by technological innovation in the region will increase by 0.15%. The level of high-level human capital, including quality factors, will increase by one unit. The economic growth brought by technological inno- vation in the region will increase by 0.55%. It shows that compared with the quantity of education, quality factors can promote economic growth in the region. The spillover effect of advanced human capital through technological innovation in promoting the growth of neighboring areas was not significant or significantly negative, which may be due to the “Matthew effect” existing in the advanced human capital level of different provinces, which makes it impossible to form the spillover effect of technological inno- vation. The spatial spillover effect of advanced human capital was mainly manifested in improving the economic growth level of neighboring regions through the promotion of technological imitation and catch-up. The impact of technological catch-up to promote the growth of neighboring areas was 0.11%-0.29%. Therefore, to realize the leapfrog transformation of the spatial spillover effect of higher education from technological imitation and absorption to high-end technological innovation, it is also necessary to rely on higher levels of education and human capital. Improving the quality of educa- tion is the key to promoting the coordinated development of higher education and the regional economy. Zhao & Du. Higher Education, Human Capital Quality and “Local-Neighborhood” Economy. Vol.8, No. 1, 2021 1052 Table 3. Decomposition of the Spatial Effect of Higher Education and Human Capital Quality in Promoting Regional Economic Growth. Variable Local Effect Neighborhood Effect Overall Effect Education Level LIHK Education Level LIHK Education Level LIHK Capital Stock 0.3355*** (0.0115) 0.3033*** (0.0123) 0.0716** (0.0358) 0.1299*** (0.0388) 0.4070*** (0.0365) 0.4332*** (0.0403) Number of Labor -0.6850*** (0.0269) -0.8687*** (0.0358) 0.2061* (0.1179) 0.4841*** (0.1162) -0.4790*** (0.1133) -0.3846*** (0.1153) Basic Human Capital 0.5351*** (0.0925) 0.0456** (0.0186) -0.7291** (0.3054) -0.1265** (0.0611) -0.1940 (0.2830) -0.0809 (0.0630) Advanced Human Capital 0.1499*** (0.0152) 0.5493*** (0.0578) -0.1715** (0.0528) -0.2953 (0.1848) -0.0216 (0.0487) 0.2540 (0.1878) Technology Catch-up -0.0416*** (0.0050) -0.2212*** (0.0295) 0.1058*** (0.0129) 0.2934*** (0.0752) 0.0642*** (0.0115) 0.0722 (0.0674) Foreign Trade Degree of Dependence 0.0436** (0.0144) -0.0028 (0.0159) 0.0455 (0.0386) -0.0781* (0.0449) 0.0891** (0.0382) -0.0808* (0.0477) Industrial Structure -0.3621*** (0.0639) -0.3022*** (0.0623) -0.0902 (0.2330) 0.0419 (0.2219) -0.4523* (0.2402) -0.2604 (0.2352) Government Support -0.4233*** (0.0623) -0.3146*** (0.0692) -0.3324* (0.1833) 0.3758** (0.1828) -0.7558*** (0.1774) 0.0612 (0.1710) ***, **, * Represent significant at the statistical level of 1%, 5%, and 10%, respectively. Table 4. Estimation of the Spatial Effect of Higher Education by Dividing Eco- nomic Zones. Bohai Sea Economic Belt and Northeast China Yangtze River Triangle Economic Belt Pearl River Delta Economic Belt Local Neighborhood Local Neighborhood Local Neighborhood Capital Stock 0.4932*** (0.0212) -0.1160* (0.0687) 0.1341*** (0.0190) -0.0904** (0.0416) 0.2900*** (0.0261) -0.0411 (0.0845) Number of Labor 0.0311 (0.0800) -0.2317 (0.2230) -0.1039*** (0.0063) 0.0439** (0.0181) -0.0426*** (0.0050) -0.0841*** (0.0240) Basic Human Capital 0.1777*** (0.0211) 0.1016* (0.0537) -0.0480* (0.0248) 0.2370*** (0.0574) 0.0499** (0.0242) -0.1249*** (0.0369) Advanced Human Capital 0.1624*** (0.0290) -0.1099 (0.0753) -0.0239 (0.0358) 0.1607** (0.0591) 0.1853*** (0.0553) -0.1323 (0.0850) Technology Catch-up -0.0674*** (0.0075) 0.1659*** (0.0292) 0.0471** (0.0201) 0.0575 (0.0369) 0.0161 (0.0255) 0.1584** (0.0575) Sample Size 252 112 168 R2 0.883 0.362 0.658 *This table only reports the spatial direct and indirect effects of using W3 matrix and education human capital to estimate core explanatory variables. The complete results can be obtained from the author. In terms of regional division, this article draws on the division method of Tongbin Zhang (2016). Among them, the Bohai Sea Eco- nomic Belt and Northeast China include 9 provinces (cities, autonomous regions), including Beijing, Tianjin, Hebei, Shanxi, Inner Mongolia, Liaoning, Jilin, Heilongjiang, and Shandong; the Yangtze River triangle econom- ic belt includes 4 provinces (cities) of Shanghai, Jiangsu, Zhejiang, and Anhui; the Pearl River Delta economic belt includes 6 provinces of Fujian, Jiangxi, Hunan, Guangdong, Guangxi, and Hainan. Zhao & Du. Higher Education, Human Capital Quality and “Local-Neighborhood” Economy. Vol.8, No. 1, 2021 1053 Further Discussion on the Spatial Spillover Effect of Higher Education About 70% of the world’s entire economy is from the economic output of more than 40 major urban agglomerations globally. In recent years, China’s three major urban ag- glomerations, the Yangtze River Delta, the Pearl River Delta, and Beijing-Tianjin- Hebei have become the three growth poles of China’s economic development and have produced more than 40% of GDP. We refer to the classification standard of Zhang et al. (Zhang, 2016) to further estimate the sub-samples of the Northeast, Yangtze River Del- ta, and Pearl River Delta Economic Belts in the Bohai Rim Economic Zone, observe the spatial effects of higher education within the region and make regional comparisons (Table 4). In terms of regional differences, the Yangtze River Delta has the highest per capita GDP during the sample period. The Bohai Rim Economic Belt has a higher per capita GDP than the Pearl River Delta. The technological innovation spillover effect of advanced human capital represented by higher education is significantly positive in the Yangtze River Delta region, indicating that the “agglomeration effect” of the economic belt has brought new momentum to regional economic growth. The neighborhood effect is not significant in the northeastern Bohai Rim region and the Pearl River Delta region. The neighborhood effect of higher education in the northeastern Bohai Rim and the Pearl River Delta is mainly manifested in technology catch-up and imitation. Comply- ing with the “Guangdong-Hong Kong-Macao Greater Bay Area” policy, along with the flow of talents and technologies, higher education has enhanced the ability to absorb and imitate technology in the Pearl River Delta region, promoted the economic growth of neighboring areas, and strengthened the economic growth of higher education throughout the region. Conclusions and Suggestions Using panel data from China’s provinces and municipalities from 1990 to 2017, com- bined with microdata, two human capital measurement data were constructed using the education stock and labor income index method. And learn from the improved Lucas and Nelson & Phelps models to clarify the mechanism of higher education and human capital quality from the spatial dimension. The results found: (i) Higher education indi- rectly promotes local economic growth through technological innovation, but the spatial spillover effect is mainly manifested in promoting economic growth in neighboring are- as through technological catch-up and imitation. After considering quality factors, the local and neighboring growth effects of advanced human capital are both enhanced. (ii) Basic human capital directly acts on output to promote regional economic growth, but it will inhibit the economic growth of other neighboring provinces and cities. (iii) Higher education exhibits different effects in regions with different economic development levels, and it can, even more, exert a technological innovation spillover effect on eco- nomic growth in areas with high economic development levels. Based on the above research conclusions, the policy implications are as follows: Zhao & Du. Higher Education, Human Capital Quality and “Local-Neighborhood” Economy. Vol.8, No. 1, 2021 1054 Firstly, give full play to the spatial coordination mechanism of higher education that promotes local-neighborhood growth through technological innovation and techno- logical imitation and enhances the contribution of higher education to regional econom- ic growth. When formulating regional development policies, spatial factors should be included. Continuously improve the level and quality of human capital, amplify the ag- glomeration effect and scale effect, ensure the full use of human capital externalities, and constantly increase the contribution of higher education and human capital to re- gional economic growth. Secondly, realize the complementary advantages of provincial education and human capital, and form a regional economic layout with high-quality development. It is necessary to strengthen further the positive spillover effect of human capital in pro- moting technology imitation and absorption in the region and at the same time enhance the technology assimilation and absorption capacity of neighboring areas. Creating an environment conducive to sharing knowledge within the region and stimulating imita- tion, absorption, and independent innovation in neighboring areas and the quality of economic development in the entire area are ultimately improved. Thirdly, implement differentiated policies for regions with different develop- ment levels, and build an innovative system for the coordinated development of region- al higher education and economy. The external environment for technological innova- tion and economic growth should be continuously optimized to enable the spillover of the talent innovation effect to break through the limitations of geographical proximity. Open up the path of high-level talent knowledge spillover, improve regional connectivi- ty, realize the matching interaction between high-level innovative talents and the up- grading of regional industrial structure, and form a talent layout conducive to regional coordinated development. References Barro, R.J., & Lee, J.W. (1993). Interna- tional comparisons of educational at- tainment. Journal of Monetary Econom- ics, 32(3):363-394. DOI: https://doi.org/10.1016/0304- 3932(93)90023-9 Cai, F., & Du, Y. (2000). Convergence and differences in China’s regional econom- ic growth: Implications for the western development strategy. Economic Re- search Journal, 46(10):30-37+80. Chang, X., & Zhao, Y. (2017). An econo- metric research on the economic growth effect of China’s human capital: An empirical analysis based on inter- provincial panel data. Journal of Statis- tics and Information, 32(11):10-20. [Chinese] https://doi.org/10.1016/0304-3932(93)90023-9 https://doi.org/10.1016/0304-3932(93)90023-9 Zhao & Du. Higher Education, Human Capital Quality and “Local-Neighborhood” Economy. Vol.8, No. 1, 2021 1055 https://www.cnki.com.cn/Article/CJFD Total-TJLT201711002.htm Chen, Q. (2014). Advanced econometrics and Stata application (Second Edition). Beijing: Higher Education Press. ISBN: 978-7-04-032983-4. Deng, F., & Ke, W. (2020). Heterogeneous human capital and economic develop- ment: an empirical study based on spa- tial heterogeneity. Statistical Research, 37(2):93-104. [Chinese] https://www.cnki.com.cn/Article/CJFD Total-TJYJ202002009.htm Ding, S., & Knight, J. (2011). Why has China grown so fast? The role of physi- cal and human capital formation. Oxford Bulletin of Economics & Statistics, 73(2):141–174. DOI: https://doi.org/10.1111/j.1468- 0084.2010.00625.x Dong, Z. (2017). Research on the interac- tive relationship between human capital and economic growth: An empirical analysis based on China’s human capital index. Macroeconomics, 37(4):88-98. [Chinese] DOI: https://doi.org/10.16304/j.cnki.11- 3952/f.2017.04.010 Du, W., Yang, Z., & Xia, G. (2014). Re- search on the mechanism of human cap- ital promoting economic growth. China Soft Science, 29(8):173-183. [Chinese] DOI: https://doi.org/10.3969/j.issn.1002- 9753.2014.08.018 Du, Y., & Zhao, R. (2018). The role of edu- cation in economic growth: factor ac- cumulation, efficiency improvement, or capital complementarity? Educational Research, 39(5):27-35. [Chinese] https://www.cnki.com.cn/Article/CJFD Total-JYYJ201805005.htm Engelbrecht, H.J. (2003). Human Capital and Economic Growth: Cross-Section Evidence for OECD Countries. Eco- nomic Record, 79:40-51. DOI: https://doi.org/10.1111/1475- 4932.00090 Fischer, M.M., Bartkowska, M., Riedl, A., Sardadvar, S., & Kunnert, A. (2009). The impact of human capital on regional labour productivity in Europe. Letters in Spatial and Resource Sciences, 2(2):97- 108. DOI: https://doi.org/10.1007/s12076-009- 0027-7 Graham, J.W., & Webb, R.H. (1979). Stocks and depreciation of human capi- tal: New evidence from a present-value perspective. Review of Income and Wealth, 25(2):209-224. Hanushek, E.A., & Woessmann, L. (2011). How much do educational outcomes matter in OECD countries? Economic Policy, 26(67): 427-491. https://www.nber.org/system/files/worki ng_papers/w16515/w16515.pdf Holz, C.A., & Sun, Y. (2018). Physical capital estimates for China’s provinces, 1952-2015 and beyond. China Econom- ic Review, 51:342-357. DOI: https://doi.org/10.1016/j.chieco.2017.06 .007 Jamison, E.A., Jamison, D.T., & Hanushek, E. A. (2007). The effects of education quality on income growth and mortality decline. Economics of Education Re- view, 26(6):771-788. DOI: https://doi.org/10.1016/j.econedurev.20 07.07.001 Jorgenson, D., & Fraumeni, B.M. (1989). The Accumulation of Human and Non- human capital, 1948–1984. Chicago: University of Chicago Press. ISBN: 0- 226-48468-8. Kendrick, J.W. (1976). The formation and stocks of total capital. New York: Co- lumbia University Press for NBER. ISBN: 0-87014-271-2. LeSage, J.P., & Pace, R.K. (2009). Intro- duction to spatial econometrics. New https://www.cnki.com.cn/Article/CJFDTotal-TJLT201711002.htm https://www.cnki.com.cn/Article/CJFDTotal-TJLT201711002.htm https://www.cnki.com.cn/Article/CJFDTotal-TJYJ202002009.htm https://www.cnki.com.cn/Article/CJFDTotal-TJYJ202002009.htm https://doi.org/10.1111/j.1468-0084.2010.00625.x https://doi.org/10.1111/j.1468-0084.2010.00625.x https://doi.org/10.16304/j.cnki.11-3952/f.2017.04.010 https://doi.org/10.16304/j.cnki.11-3952/f.2017.04.010 https://doi.org/10.3969/j.issn.1002-9753.2014.08.018 https://doi.org/10.3969/j.issn.1002-9753.2014.08.018 https://www.cnki.com.cn/Article/CJFDTotal-JYYJ201805005.htm https://www.cnki.com.cn/Article/CJFDTotal-JYYJ201805005.htm https://doi.org/10.1111/1475-4932.00090 https://doi.org/10.1111/1475-4932.00090 https://doi.org/10.1007/s12076-009-0027-7 https://doi.org/10.1007/s12076-009-0027-7 https://www.nber.org/system/files/working_papers/w16515/w16515.pdf https://www.nber.org/system/files/working_papers/w16515/w16515.pdf https://doi.org/10.1016/j.chieco.2017.06.007 https://doi.org/10.1016/j.chieco.2017.06.007 https://doi.org/10.1016/j.econedurev.2007.07.001 https://doi.org/10.1016/j.econedurev.2007.07.001 Zhao & Du. Higher Education, Human Capital Quality and “Local-Neighborhood” Economy. Vol.8, No. 1, 2021 1056 York: CRC Press. ISBN: 978-142-006- 424-7. Li, H., & Tang, T. (2015). Regional differ- ences in labor quality based on human capital. Journal of Central University of Finance & Economics, 35(8):72-80+86. [Chinese] http://www.iyonly.com/xbbjb.iyonly.co m/CN/article/downloadArticleFile.do?at tachType=PDF&id=5802 Li, T., Lai, J.T., Wang, Y., & Zhao, D. (2016). Long-run relationship between inequality and growth in post-reform China: New evidence from dynamic panel model. International Review of Economics & Finance, 41:238 -252. DOI: https://doi.org/10.1016/j.iref.2015.08.00 9 Liang, R., Yu, J., & Feng, S. (2015). The calculation of the contribution of human capital to China’s economic growth. South China Journal of Economics, 33(7):1-14. [Chinese] DOI: https://doi.org/10.19592/j.cnki.scje.2015 .07.001 Lucas Jr, R.E. (1988). On the mechanics of economic development. Journal of Monetary Economics, 22(1):3-42. DOI: https://doi.org/10.1016/0304- 3932(88)90168-7 Meng, W., & Wang, X. (2014). China’s provincial-level of human capital meas- urement-based on the cost method of perpetual inventory technology. Studies in Labor Economics, 2(4):141-160. [Chinese] https://www.cnki.com.cn/Article/CJFD Total-LDJJ201404009.htm Mulligan, C.B., & Sala-i-Martin, X. (1997). A labor-income-based measure of the value of human capital: An application to the United States. Japan and the World Economy, 9(2):159-191. DOI: https://doi.org/10.1016/S0922- 1425(96)00236-8 Nelson, R.R., & Phelps, E.S. (1965). In- vestment in humans, technological dif- fusion, and economic growth. Cowles Foundation Discussion Papers 189, Cowles Foundation for Research in Economics, Yale University. https://ideas.repec.org/p/cwl/cwldpp/18 9.html Psacharopoulos, G., & Patrinos, H.A. (2004). Returns to investment in educa- tion: A further update. Education Eco- nomics, 12(2):111-134. DOI: https://doi.org/10.1080/0964529042000 239140 Qian, X. (2012). A statistical estimation of human capital level. Statistical Re- search, 29(8):74-82. [Chinese] DOI: https://doi.org/10.19343/j.cnki.11- 1302/c.2012.08.013 Romer, P. M. (1990). Endogenous techno- logical change. Journal of Political Economy, 98(5):71-102. https://www.journals.uchicago.edu/doi/a bs/10.1086/261725 Schultz, T.W. (1961). Investment in Human Capital. The American Economic Re- view, 51(1):1-17. https://www.jstor.org/stable/1818907?se q=1 Zhang, T., Li, J., & Zhou, H. (2016). Re- gional knowledge spillovers, collabora- tive innovation, and total factor produc- tivity growth in high-tech industries. Fi- nance and Trade Research, 27(1):9-18. [Chinese] DOI: https://doi.org/10.19337/j.cnki.34- 1093/f.2016.01.002 Zhu, P., & Xu, D. (2007). Estimation of human capital in Chinese cities. Eco- nomic Research Journal, 53(9):84-95. [Chinese] https://www.2002n.com/d/img/upfiles_6 /200849153137197.pdf http://www.iyonly.com/xbbjb.iyonly.com/CN/article/downloadArticleFile.do?attachType=PDF&id=5802 http://www.iyonly.com/xbbjb.iyonly.com/CN/article/downloadArticleFile.do?attachType=PDF&id=5802 http://www.iyonly.com/xbbjb.iyonly.com/CN/article/downloadArticleFile.do?attachType=PDF&id=5802 https://doi.org/10.1016/j.iref.2015.08.009 https://doi.org/10.1016/j.iref.2015.08.009 https://doi.org/10.19592/j.cnki.scje.2015.07.001 https://doi.org/10.19592/j.cnki.scje.2015.07.001 https://doi.org/10.1016/0304-3932(88)90168-7 https://doi.org/10.1016/0304-3932(88)90168-7 https://www.cnki.com.cn/Article/CJFDTotal-LDJJ201404009.htm https://www.cnki.com.cn/Article/CJFDTotal-LDJJ201404009.htm https://doi.org/10.1016/S0922-1425(96)00236-8 https://doi.org/10.1016/S0922-1425(96)00236-8 https://ideas.repec.org/p/cwl/cwldpp/189.html https://ideas.repec.org/p/cwl/cwldpp/189.html https://doi.org/10.1080/0964529042000239140 https://doi.org/10.1080/0964529042000239140 https://doi.org/10.19343/j.cnki.11-1302/c.2012.08.013 https://doi.org/10.19343/j.cnki.11-1302/c.2012.08.013 https://www.journals.uchicago.edu/doi/abs/10.1086/261725 https://www.journals.uchicago.edu/doi/abs/10.1086/261725 https://www.jstor.org/stable/1818907?seq=1 https://www.jstor.org/stable/1818907?seq=1 https://doi.org/10.19337/j.cnki.34-1093/f.2016.01.002 https://doi.org/10.19337/j.cnki.34-1093/f.2016.01.002 https://www.2002n.com/d/img/upfiles_6/200849153137197.pdf https://www.2002n.com/d/img/upfiles_6/200849153137197.pdf Zhao & Du. Higher Education, Human Capital Quality and “Local-Neighborhood” Economy. Vol.8, No. 1, 2021 1057 Received: 10 April 2021 Revised: 20 April 2021 Accepted: 09 May 2021 The Chinese version of this article has been published in Higher Education Research, 2020, 41(8):52- 62. The English version has been authorized for being publication in BECE by the author(s) and the Chinese journal. 赵冉, 杜育红. (2020). 高等教育, 人力资本质量对“本地-邻地”经济增长的影响. 高等教育研究, 2020, 41(8):52-62. Article-RanZhao-BECE_title_May2021 Article-RanZhao-BECE_MaintextMay2021