Discussion Paper

No. 2013-30 | May 08, 2013
Environmental Impact of the 2008 Beijing Olympic Games

Abstract

Beijing organized the 2008 Summer Olympic Games, and the main goal of the Chinese government regarding this event was to hold a "Green Olympics". A difference-in-differences approach was used to estimate the environmental impact the Olympic Games on air quality improvement in Beijing, compared to improvements in other areas in China. The results indicate that compared to other regions, air quality in Beijing improved for a short period of time. These improvements were largely due to the implementation of several temporary measures, including factory closures and traffic control. However, there is no evidence indicating that the Olympic Games reduced the concentration of sulfur dioxide in Beijing.

Data Set

Data sets for articles published in "Economics" are available at Dataverse. Please have a look at our repository.

The data set for this article can be found at: http://hdl.handle.net/1902.1/21407

JEL Classification

Q51 Q53 G14 L83

Cite As

Cao Huijuan, Hidemichi Fujii, and Shunsuke Managi (2013). Environmental Impact of the 2008 Beijing Olympic Games. Economics Discussion Papers, No 2013-30, Kiel Institute for the World Economy. http://www.economics-ejournal.org/economics/discussionpapers/2013-30

Assessment



Comments and Questions


Anonymous - Referee report 1
June 06, 2013 - 10:59

see attached file


Huijuan Cao - Reply to the reader comment
July 05, 2013 - 08:39

We will address your comments in the revised draft of the paper once
we collect all the readers'/reviewers' comments and suggestions.


Anonymous - Referee report 2
June 20, 2013 - 11:31

This is a well-written, concise paper on whether the actions taken by the Chinese government around the 2008 Olypmic games resulted in reduced pollution levels in Beijing. There is some promise in this paper, but I believe that in its current state the paper is coming at the question from ...[more]

... the wrong perspective.

The way the authors put it, I find the question not as appealing given the previous work in this area. As the authors themselves state, their goal is to evaluate whether “air quality improved in Beijing compared to other areas in China.” Other carefully done studies cited by the authors already find that during the Olympics air quality improved in Beijing. Hence, what does this “relative” comparison bring us? I think what it can really tell is whether there was some pollution substitution from Beijing to other areas of China. Now to me that is an extremely interesting question and any evidence towards that would be a really great paper. I would recast the tone of the paper to think about whether other areas in China increased their pollution levels relative to Beijing during this period. This might be the case since surely power generation must have increased due to greater demand for power in Beijing. Traffic laws would not result in increased pollution in neighboring areas (and in fact the authors could use these specific types of pollution kicked up by traffic as sort of a “placebo” in that these types of pollution should not increase in neighboring areas), but I think the fact that a lot of the movement in pollution seems to come from power plants there is scope here to write a really good paper about substitution effects. I think using distance to Beijing as a measure of “treatment” would be interesting if the authors can provide some sense of whether power transfers over longer distances are less likely/more expensive, so power plants closest to Beijing increased production in response. In some sense, this could already be happening, and its entirely possible that the relative decreases the authors find are precisely due to surrounding areas increasing their pollution levels.

Even as it stands, I have some comments:

1. The authors should provide many pictures in these types of difference in differences design that can assure the readers that parallel trends assumptions are met. Moreover, having figures will tell us where the DinD estimates comes from – whether its from the decreases in the treatment group, or from increases in the control group or both. Given my general comments on how the authors might want to approach the paper, I think providing clear graphs would be very important. At a minimum a discussion about potential threats to validity should be present in the paper.
2. It is not clear how the authors are clustering standard errors. Standard methods for dealing with clustering with small samples should be used.
3. The tables are a bit hard to read. If the authors can replace the table headings with more interpretable groupings than “Group B” and such it would be very helpful.


Huijuan Cao - Reply to the reader comment
July 05, 2013 - 08:39

We will address your comments in the revised draft of the paper once
we collect all the readers'/reviewers' comments and suggestions.


Anonymous - comments
July 02, 2013 - 16:06

This paper analyzed the effect of Beijing Olympic Games on air pollution. The feature is using the difference in difference approach and compares the air quality in terms of not only time periods but also locations. Some interpretations are a little confusing. However, the results are interesting.

1. ...[more]

... Briefly explaining the features of difference in difference approach compared to the conventional econometrics approach will help readers better understand the paper.

2. Explaining the difference in methodology compared to the previous studies will show the contribution of this paper clearly.

3. Check again the following confusing interpretations.
・[On page 6, 8th line from the bottom] “GroupB-2007 also exhibited an interaction effect for the NOx concentration that was statistically insignificant”
It seems significant.

・[On page 7, 3rd paragraph, 2nd line] “However, there was no significant effect in the long term compared with that in other areas”
What is the basis of this sentence? Please show the corresponding figure in the Table.


Huijuan Cao - Reply to the reader comment
July 05, 2013 - 08:41

We will address your comments in the revised draft of the paper once we collect all the readers'/reviewers' comments and suggestions.