High-speed rail to prosperity? Assessing the role of transportation improvement in the urban economy

Figures & data

Figure 1. Chinese HSR network before 2015.

Source: Authors.

Table 1. Descriptive statistics.

Variable	Obs.	Mean	Std.Dev.
Real GDP/10000 Yuan	1988	14.8	1.156
HSR access	2016	0.375	0.484
Fixed-asset investment/10000 Yuan	1980	14.39	1.095
Total retail sales of consumer goods/10000 Yuan	1985	13.86	1.201
Foreign direct investment/10000 dollar	1785	9.3	1.939
Total population/10000 persons	1990	4.585	0.778
Students enrolled in universities/person	1928	10.38	1.322
Government size/%^a	1983	15.93	8.931
Infrastructure development^b/%	1986	0.012	0.015
Innovation index	2016	7.182	33.61
Density of population/ (people/square kilometers)	2000	6.472	0.953
Secondary industry as Percentage to GDP	2000	51.19	12.19
Tertiary industry as Percentage to GDP	2000	42.04	11.2
Marketization level/%^c	1518	0.475	0.456
The least cost tree	1757	0.315	0.465
Real GDP per capita/Yuan	1971	10.16	0.601
Railway passenger volume/10000 persons	1749	5.788	1.269
Highway passenger volume/10000 persons	1982	8.697	0.976
Average nighttime light brightness	1698	7.122	8.468
Persons employed in manufacturing/person	1977	19.65	1.499
Persons employed in consumer service/person	1973	18.70	1.233
Persons employed in knowledge economy/person	1982	19.06	1.090

Note: The cross-city standard deviation of $Y$ is the standard deviation of ($Y_{st}-\widetilde{Y_s}$), where $\widetilde{Y_s}$ is the average value of Y in city s over the sample period.

^a Government size is measured by the ratio of government expenditure to GDP.

^b Infrastructure development refers to the value area of paved roads divided by total land area of city.

^c Marketization level is expressed by the proportion of foreign investment to GDP.

Source: Authors.

Figure 2. The trend of the average local GDP of cities.

Source: Authors.

Figure 3. Histograms of average local GDP (1999–2006).

Source: Authors.

Figure 4. The logarithm of GDP for the treatment group and control group.

Source: Authors.

Figure 5. The difference in GDP between the control group and treatment group.

Source: Authors.

Table 2. DID regressions.

Ln(GDP)	Full-Sample		Matched sample
	Model (1)	Model (2)	Model (3)	Model (4)
HSR	0.021	0.023*	0.114***	0.033***
	(0.027)	(0.013)	(0.040)	(0.015)
Control	No	Yes	No	Yes
Cons	14.379***	5.996***	13.936***	5.910***
	(0.010)	(0.735)	(0.0129)	(0.807)
Year Fixed Effect	Yes	Yes	Yes	Yes
City Fixed Effect	Yes	Yes	Yes	Yes
Observations	2271	1654	1017	767
R-square	0.779	0.924	0.768	0.930

Note: Standard errors are adjusted for city level clustering and appear in parentheses. *, **, and *** indicate statistical significance levels at the 10%, 5%, and 1% levels, respectively. Covariates variables include fixed-asset investment, total retail sales of consumer goods, foreign direct investment, total population, students enrolled in universities, government size, infrastructure development, innovation index, density of population, secondary industry as percentage to GDP, tertiary industry as percentage to GDP, marketization level. Same hereinafter.

Source: Authors.

Figure 6. Dynamics of the economic effects of the HSR network.

Source: Authors.

Figure 7. Logarithm of the number of industrial enterprises.

Source: Authors.

Figure 8. Logarithm of gross industrial output.

Source: Authors.

Figure 9. Logarithm of the number of people employed in manufacturing.

Source: Authors.

Figure 10. The ratio of industrial output to local GDP.

Source: Authors.

Figure 11. Logarithm of the number of people employed in the service sector.

Source: Authors.

Figure 12. Logarithm of the number of people employed in finance.

Source: Authors.

Figure 13. Logarithm of the number of people employed in information transmission, computer services and software.

Source: Authors.

Figure 14. Logarithm of the number of people employed in wholesale and retail trade.

Source: Authors.

Figure 15. Logarithm of the number of people employed in hotels and catering services.

Source: Authors.

Figure 16. The index of service sector specialization.

Source: Authors.

Figure 17. The index of service sector diversification.

Source: Authors.

Table 3. IV Regression.

First-stage probit regression		Second-stage regression
HSR	Model (5)	Ln(GDP)	Model (6)
Potential HSR Route	0.847***	HSR	0.126***
	(0.125)		(0.053)
Population	0.440***
	(0.105)
GDP Per Capita	0.442***
	(0.110)
Train Passengers	0.975***
	0.076	Control	Yes
Constant	Yes	Constant	Yes
LR Chi2	334.95	Year Fixed Effect	Yes
Year Fixed Effect	Yes	City Fixed Effect	Yes
Pseudo R2	0.205	R-square	0.920
Observations	1691	Observations	1880

Source: Authors.

Table 4. PSM-DID Regressions.

Ln(GDP)	Model (7)	Model (8)
HSR	0.064***	0.039***
	(0.033)	(0.013)
Controls	No	Yes
Cons	14.204***	6.403***
	(0.011)	(0.802)
Year Fixed Effect	Yes	Yes
City Fixed Effect	Yes	Yes
Observations	1324	1069
R-square	0.795	0.926

Source: Authors.

Figure 18. Placebo tests.

Source: Authors.

Table 5. Sensitivity tests.

	170 samples		1999–2014		Nighttime light data
	Ln(GDP)	Ln(GDP)	Ln(GDP)	Ln(GDP)	Dnvalue	Dnvalue
HSR	0.093***	0.040***	0.089**	0.043**	0.166**	0.159*
	(0.033)	(0.013)	(0.037)	(0.018)	(0.080)	()0.088
Control	No	Yes	No	Yes	No	Yes
Cons	14.163***	6.024***	13.151***	6.168***	4.504***	1.542
	(0.011)	(0.734)	(0.021)	(0.959)	(0.033)	(3.971)
Year Fixed Effect	Yes	Yes	Yes	Yes	Yes	Yes
City Fixed Effect	Yes	Yes	Yes	Yes	Yes	Yes
Observations	1353	1077	1903	1336	756	666
R-square	0.775	0.926	0.882	0.947	0.551	0.572

Source: Authors.