How Public Opinion Affects Air Quality: A Chinese Case Study

China has achieved notable economic development in recent decades. However, such achievement comes with serious environmental and social costs. The issue of environmental deterioration, particularly air pollution, has drawn extensive concern from the public as well as policymakers and scholars.

In early 2012, the Chinese government released documents requiring the adoption of a new standard for Air Quality Index (AQI), which addresses the concentrations of air pollutants including PM_2.5, PM₁₀, ozone (O₃), sulfur dioxide (SO₂), etc., and their constant updates and publication on media. However, frequent high levels of AQI have become a grievance of the public and drawn media critiques. Even though China is an authoritarian country with strict media control, with the rapid development and application of information and internet technologies, it is increasingly difficult to limit public opinion. Therefore, public opinion can put great pressure on the government and force it to take actions to regulate air pollution.

We intended to examine whether and how public opinion on air quality, specifically smog, affects air quality. This is the first study to quantitatively examine the influence of public opinion on air quality in China. We addressed the questions from three aspects: 1. to examine the connection between the change in air quality and the formation of public opinion; 2. to examine the extent to and the pace at which public opinion affects air quality once it is formed; and 3. to examine the details of the influence of public opinion on the concentrations of individual air pollutants.

We utilized monthly data from 109 prefecture-level cities in China from November 2013 to October 2016. Air quality data included AQI, concentrations indexes of PM_2.5, PM₁₀, O₃, SO₂, carbon monoxide (CO), and nitrogen dioxide (NO₂), and the numbers of days with rain or snow and wind of grade 3 or above. The dataset of public opinion consisted of the numbers of online mainstream media commentaries related to smog issue and were collected from national and local media sources. We then calculated the monthly intensity of public opinion that is the weighted average numbers of the media reports from national, provincial, and local media sources. In addition, we grouped the cities for comparison and contrast based on the proportion of secondary industrial output values, per-capita GDP, the size of the population, and geographical location.

First, we analyzed the formation of public opinion and found that the deterioration of air quality can cause the outbreak of public opinion. In addition, subsample regression results showed that in cities with higher values for secondary industrial output, higher per-capita GDP, larger population, and that are located in Northern China, public opinion reacted more sensitively to the level of AQI. Furthermore, it is statistically confirmed that if the city is a provincial capital, or if it is in Middle or Western China, public opinion may be formed more quickly. It is also worth noting that in October, November, and December, public opinion occurred more frequently.

To address the second question, we analyzed the temporal difference of the impact of public opinion and concluded that public opinion that has been formed for a long period may have limited impact, while a recent one is more likely to induce governmental regulatory actions. Therefore, we emphasized the recent public opinion and conducted statistic estimations to examine to what extent and at what pace public opinion affects air quality, which we call the ameliorating effect and decelerating effect, respectively.

Two models were utilized to examine these effects after proper adjustments. Regression results based on both models showed that the one- and two-period lags of public opinion exhibited significant negative effects on air quality indicators, which implies that up to two months after its formation, public opinion does reduce the level of air pollution, or it prevents air quality from worsening.

Interestingly, the three-period lag of public opinion had a positive effect on air quality, although the results are mostly not statistically significant. Such may be caused by the retaliatory rebound of air pollution near and after the relief of the pressure from public opinion. The regression results from contrasting groups show that both effects are relatively distinct in cities with a larger proportion of secondary industrial output value, higher GDP per capita, less population, and that are in Northern China.

Last but not least, we applied the models to examine both ameliorating and decelerating effects on each individual air pollutant to capture the effectiveness of public opinion on controlling individual air pollutants. What we mainly found from the estimations was that public opinion functions well in controlling PM_2.5 and PM₁₀, while it was not necessarily effective in controlling the others. Considering the fact that the majority of PM_2.5 and PM₁₀ emission comes from industrial production and construction activities, they are relatively manageable for governments through short-term regulations. SO₂ and O₃ are both mainly the byproducts of coal-fired power plants and coal boilers or furnaces, which are somewhat controllable. However, controlling the sources of emissions of CO and NO₂ is more difficult since the major sources of them are automobile exhaust, and government cannot strictly regulate the use of automobiles.

In summary, we examined the effect of public opinion on air quality in terms of both “extent” and “pace” using monthly data of 109 prefecture-level cities in China from November 2013 to October 2016. The statistical results show that public opinion has both ameliorating and decelerating effects within two months after its formation. In addition, both effects exhibit variations with respect to the industrial structure, stage of economic development, population, and geological location of cities. Furthermore, examinations on individual air pollutants show that public opinion is effective in controlling PM_2.5 and PM₁₀ but less effective for other pollutants.

Our findings suggest that firstly, the government should address and respond to public opinion by regulating the sources of air pollution. Secondly, public opinion can be a criterion in evaluating local policymakers’ performance in order to supervise leaders’ environmental efforts. Thirdly, public opinion may serve as an indicator to help governments to identify the source, composition, and controllability of air pollution.

It is necessary to notice that the study in this paper still requires future development, such as the expansion of sampling cities and periods and utilizing other methods to address the possible correlation between economic development and the effectiveness of public opinion on air quality.

These findings are described in the article entitled Does public opinion affect air quality? Evidence based on the monthly data of 109 prefecture-level cities in China, recently published in the journal Energy Policy. This work was conducted by Shengling Zhang, Yue Li, and Yipeng Zhang from Beijing Normal University, and Yu Hao from the Beijing Institute of Technology, Sustainable Development Research Institute for Economy and Society of Beijing, and Beijing Key Lab of Energy Economics and Environmental Management.