Abstract
During the Olympic Games in 2008, a driving restriction based on vehicle license plate numbers was implemented in Beijing to mitigate air pollution and traffic congestion. Following the Games, the restriction was modified several times. This paper investigates the effects of two policy changes: a weakening policy change due to a shorter restricted time period, and a strengthening policy change due to a higher penalty for violators and the complementary car purchasing restriction. By employing a regression discontinuity design in a Tobit model, I find that the weakening policy change led to more pollution and the strengthening policy change improved air quality in restricted areas. Several robustness checks confirm the results. I also provide suggestive evidence that driving restrictions increased the use of public transportation and alleviated traffic congestion.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
For a review of epidemiological studies on the respiratory effects of air pollution, please refer to Lebowitz (1996). For a review of epidemiological studies on the relationship between particulate matter and heart disease, please refer to Peters (2005). For a review of epidemiological studies on the relationship between air pollution and lung cancer, please refer to Cohen and Pope (1995).
From WHO website: http://www.who.int/mediacentre/factsheets/fs313/en/.
Industrial plants and power plants could have long-distance effects and lag effects, but vehicle emission is at least one of the main existing contributors of air pollution in Beijing. Westerdahl et al. (2009) suggest that traffic is a major source of ultrafine particles. A 4-day traffic control experiment conducted by the Beijing Government as a pilot to test the effectiveness of proposed controls was found to be effective in reducing extreme concentrations that occurred at both on-road and ambient environments.
From China Daily (2012), June 11, 2012. http://www.chinadaily.com.cn/regional/2012-06/11/content_16133037.htm.
Ring roads in Beijing are the main roads that surround the center of the city. The fifth ring is regarded as the threshold between the main city and suburbs, within which is the main city, while outside are suburbs.
Peak weekdays were defined as 5 a.m.–10 p.m. in weekdays, and nonpeak weekdays were defined as 10 p.m.–5 a.m. in weekdays.
The 2011 monthly average income in Beijing was RMB 4672, and those who have cars would have much higher incomes. The average cost to maintain a car is RMB 20–25 thousand per year in Beijing, and the cost varies a lot by different types of cars.
The maximum pollutant is not reported when API is below 50, i.e. when the day is “excellent”.
For the data used in the following analysis, the aggregate API data are calculated from station-level API data by the formula: \({\rm API}_{t}=\frac{1}{8}\sum _{s=1}^{8}{\rm API}_{st}\), where \(s\) = 1–8 if station s is within the 5th ring areas, \(s\) = 9–27 if station s is outside the 5th ring areas. For the data used to compare with the aggregate API data directly from MEP in this section, the aggregate API data are calculated from station-level API data by the formula: \({\rm API}_{t}=\frac{1}{27}\sum _{s=1}^{27}{\rm API}_{st}\), since the aggregate API data directly from MEP include data from all stations.
For more details, please refer to Andrews (2008).
The accurate percent change of API change because of policy change \(i\) is given by the formula: \(100(e^{\alpha _{i}}-1)\,\%\). Since \(\alpha _{i}=\log ({\rm API}_{{\rm TP}_{i}=1})-\log ({\rm API}_{{\rm TP}_{i}=0})=\log (\frac{{\rm API}_{{\rm TP}_{i}=1}}{{\rm API}_{{\rm TP}_{i}=0}})\), we can get \(\frac{{\rm API}_{{\rm TP}_{i}=1}}{{\rm API}_{{\rm TP}_{i}=0}}=e^{\alpha _{i}}\) by taking exponential on each side. By further transformation, \(\frac{{\rm API}_{{\rm TP}_{i}=1}-{\rm API}_{{\rm TP}_{i}=0}}{{\rm API}_{{\rm TP}_{i}=0}}=e^{\alpha _{i}}-1\). The left side is the exact percent change of API due to policy change \(i\).
For more details, please refer to Lee and Lemieux (2010).
The sharp RD design is a concept relative to a fuzzy RD design, which requires that the identification of causal effects hinges on the crucial assumption that there is indeed a sharp cutoff, around which there is a discontinuity in the probability of assignment from 0 to 1. In contrast to the sharp RD design, a fuzzy RD does not require a sharp discontinuity in the probability of assignment but is applicable as long as the probability of assignment is different. See Imbens and Lemieux (2007) for more details.
The “rd” command is contributed by Austin Nichols (2007). The revised version I use is modified based on the original “rd” command to allow for Tobit model.
For particulate matters, the smaller the particle, the longer it can remain suspended in the air before settling. PM2.5 can stay in the air from hours to weeks and travel very long distances because it is smaller and lighter. PM10 can stay in the air for minutes to hours and can travel shorter distances from hundreds of yards to many miles because it is larger and heavier.
6 months is the largest window available for the first policy change, while 20 months is the largest symmetric window available for the second policy change.
Figures in this section use station-level API data, as there are more observations. But aggregate API data give similar shapes.
Here, I interpret it with the more accurate percent changes using the formula \(100(e^{\alpha _{{i}}}-1)\,\%\). Same for the following interpretation.
The percent changes above and below are calculated from the formula in footnote 12.
I use month dummies instead of season dummies because they could better describe the data. There are enough observations for the station-level data, so adding a few more independent variables would not lose much degree of freedom.
The long-run effects may also be overestimated since many people who participated the lottery would not buy a car urgently if there was not the lottery. Since the winner rate is quite low, they just want to have an option whether or not to buy a car.
References
Andrews SQ (2008) Inconsistencies in air quality metrics: ‘Blue Sky’ days and PM10 concentrations in Beijing. Environ Res Lett 3(2008):1–14
China Daily (2012) http://www.chinadaily.com.cn/regional/2012-06/11/content_16133037.htm. Accessed 11 June 2012
Chen Y, Jin GZ, Kumar N, Shi G (2011) The promise of Beijing: evaluating the impact of the 2008 Olympic Games on air quality. NBER Working Paper #16907
Cohen AJ, Pope CA III (1995) Lung cancer and air pollution. Environ Health Perspect 103(Suppl 8):219–224
Davis LW (2008) The effect of driving restrictions on air quality in Mexico City. J Polit Econ 116:38–81
Economist Intelligence Unit (2012) Best cities ranking and report: a special report from the Economist Intelligence Unit
Eskeland GS, Feyzioglu T (1997) Rationing can backfire: the ‘Day without a Car’ in Mexico City. World Bank Policy Research Working Paper #1554
Gallego F, Montero JP, Salas C (2012) The effect of transport policies on car use: theory and evidence from Latin American Cities
Greenstone M, Hanna R (2011) Environmental regulations, air and water pollution, and infant mortality in India. HKS Faculty Research Working Paper Series #RWP11-034
Imbens G, Lemieux T (2007) Regression discontinuity designs: a guide to practice. NBER Working Paper 13039
Imbens G, Kalyanaraman K (2009) Optimal bandwidth choice for the regression discontinuity estimator. NBER WP 14726
Lebowitz MD (1996) Epidemiological studies of the respiratory effects of air pollution. Eur Respir J 1996(9):1029–1054
Lee DS, Lemieux T (2010) Regression discontinuity designs in economics. J Econ Lit 48:281–355
Li S (2014) Better lucky than rich? Welfare analysis of automobile license allocations in Beijing and Shanghai. Welfare Analysis of Automobile License Allocations in Beijing and Shanghai (March 1, 2014)
Lin CYC, Zhang W, Umanskaya VI(2011) The effects of driving restrictions on air quality: São Paulo, Bogotá, Beijing, and Tianjin
Nichols A (2007) Causal inference with observational data. Stat J 7(4):507–541
Peters A (2005) Particulate matter and heart disease: evidence from epidemiological studies. Toxicol Appl Pharmacol 207(2005):S477–S482
Tanaka S (2010) Environmental regulations in China and their impact on air pollution and infant mortality
Viard VB, Fu S (2011) The effect of Beijing’s driving restrictions on pollution and economic activity. MPRA Paper 33009, University Library of Munich, Germany
Wang W et al (2009) Atmospheric particulate matter pollution during the 2008 Beijing olympics. Environ Sci Technol 43(2009):5314–5320
Wang X et al (2009) Evaluating the air quality impacts of the 2008 Beijing olympic games: on-road emission factors and black carbon profiles. Atmos Environ 43(2009):4535–4543
Westerdahl D, Wang X, Pan X, Max ZK (2009) Characterization of on-road vehicle emission factors and microenvironmental air quality in Beijing, China. Atmos Environ 43(3):697–705
World Health Organization (2011). http://www.who.int/mediacentre/factsheets/fs313/en/. Accessed Sep 2011
Author information
Authors and Affiliations
Corresponding author
Additional information
I thank Kelsey Jack, Jeffrey Zabel and Richard Carson for useful advice, and thank TIE fellowship and Henken scholarship for financial support. All errors are my own.
About this article
Cite this article
Lu, X. Effectiveness of government enforcement in driving restrictions: a case in Beijing, China. Environ Econ Policy Stud 18, 63–92 (2016). https://doi.org/10.1007/s10018-015-0112-7
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10018-015-0112-7