1.Theoretical basis
China's National Railway Administration has defined bullet trains (new trains running at 250km/h and at least 200km/h in the initial operation) as high-speed trains.In this paper, 41 cities at prefecture-level and above in China's Yangtze River Delta from 2003 to 2018 were selected as the research and analysis samples. The prefecture-level city panel data were obtained from The Economic Management Home, EPS data platform, Guotaian database and China City Statistical Yearbook.
$$Pollutio{n}_{jt}={a}_{0}+{a}_{1}*HS{R}_{jt}+{a}_{2}*Con\text{tr}ol{s}_{jt}+{\beta }_{j}+{\gamma }_{t}+{e}_{jt}$$
1
In the formula, represents the environmental pollution level of region J in year T, is a dummy variable, and means whether high-speed railway is opened in region J in year T. 0 is the year before the opening, and 1 is the year after the opening.\(Pollutio{n}_{jt}HS{R}_{jt}\)a1It is used to measure the impact of the opening of high-speed rail on the level of urban industrial agglomeration.\(Con\text{tr}ol{s}_{jt}\)Represents a series of control variables, and is the index that affects the degree of industrial agglomeration.Beta.jUrban fixation, gammatRefers to the time-fixed effect. The function of these two variables is to control the individual characteristics of j region and the unobservable factors in time.ejtIs a random disturbance term.
2.Data and variable description
(1)Data selection
Data selection mainly includes: environmental pollution level (environmental pollution comprehensive index), economic development level, urbanization rate, degree of opening-up, secondary industrial structure, infrastructure level and human capital scale.The opening time of high-speed railway is confirmed by 12306 and "China Railway Yearbook" on the opening time and station of each line.
(2)Explained variables
The comprehensive environmental Pollution index (Pollution) is the explained variable. By selecting the industrial wastewater discharge (Water), industrial waste Gas discharge (Gas) and industrial Dust discharge (Dust) of the sample cities in China's Yangtze River Delta, the comprehensive environmental Pollution index is calculated by factor analysis method.Improving the accuracy of explained variables is the function of constructing composite indexes.The new principal component analysis formula rewritten from the calculated component matrix is:
$$Pollution=0.233*Water+0.341*Gas+0.427*Dust$$
2
(3)Explanatory variables
In Stata analysis, HSR is represented by the lowercase letter HSR, which is essentially a crossing term between the grouping code TREAT (the experimental group is 1, the control group is 0) and the time code year of the opening of the high-speed railway (1 if the high-speed railway has been opened this year, 0 otherwise). HSR = 1 if it is the year of the experimental group and the opening of the high-speed railway, HSR = 0 otherwise.
(4)Selection of control variables
Economic development level: GDP per capita is used to measure regional economic development. Scholar believes that when human productivity is rapidly liberated and developed and the economic growth rate is significantly increased, the ecological environment is under great pressure, which leads to various ecological and environmental problems, such as acid rain, blue fog, water pollution and air pollution (Wang Yali 2021) .
Urbanization rate (urban): Based on the urban population/total population, generally speaking, the higher the urbanization level is, the industry of the city will be relatively reduced and the high-tech industry will be relatively increased. Research found through empirical research that with the continuous improvement of China's urbanization level, industrial pollution will experience an inverted U-shaped path of first rise and then decline, but urbanization is ultimately conducive to the improvement of urban environmental quality (Ma Lei 2010) .
Openness: The measurement index is the ratio of the amount of foreign investment actually utilized to GDP. People believe that technology and management methods brought by foreign investment are conducive to decoupling economic development from ecological consumption and gradually realizing green economy (Gaimei et al. 2021).
Secondary industry structure (second) : The ratio of the added value of the secondary industry in the current year to GDP is used to calculate the structure of the secondary industry.Generally speaking, the greater the proportion of secondary industry, the more serious the pollution.The reason is that there are many heavy pollution industries in the secondary industry, such as thermal power, steel, cement, chemical industry and so on. It adopted the dynamic spatial panel model analysis method and found that: every 10% reduction in the proportion of the secondary industry will reduce SO in local and adjacent areas respectively20.03% and 0.09% of the emission level (Ge Xiangyu et al. 2020).
Infrastructure: Per capita road area can control economic and environmental differences caused by differences in infrastructure conditions in different cities, so per capita road area is used to measure regional infrastructure level.
Human capital scale (labor) : The improvement of labor quality, technical level and working ability means more human capital, which also means stronger consumer demand and higher urban attraction, which is conducive to urban industrial development.Therefore, the measurement index is the proportion of the number of college students in the total population.
The descriptive statistical results of the above variables are shown in Table 1.
Table 1
Descriptive statistics of main variables
variable | define | The average | The standard deviation | The minimum value | The maximum |
pollution | Local GDP per capita | 0.29 | 0.35 | 0 | 1.99 |
hsr | Whether the high-speed rail opened that year | 0.37 | 0.48 | 0 | 1 |
level | | 4.64 | 3.40 | 0.26 | 17.43 |
urban | Urban population to total population | 0.44 | 0.19 | 0.10 | 0.90 |
opening | The proportion of actually utilized foreign capital to GDP | 0.03 | 0.02 | 0 | 0.17 |
second | The ratio of added value of local secondary industry to local GDP | 49.24 | 8.50 | 23.23 | 74.73 |
infrastructure | Road area per capita | 4.96 | 4.04 | 0 | 22.82 |
labor | Proportion of the number of students in colleges and universities in the total population | 0.02 | 0.02 | 0 | 0.13 |