Correlation between PM2.5 pollution and its public concern in China: Evidence from Baidu Index

Since 2013, the severe haze has occurred in China and caused great harm; thus, public concern regarding PM2.5 pollution has been aroused. Evidence has proven that public concern over the haze influences the decision-making of the government; however, a comprehensive analysis has not been conducted of the correlation between the public concern over haze and PM2.5 concentration. Based on the daily data of Baidu Index and daily PM2.5 concentration, this paper investigates the correlation between haze and its public concern from two dimensions: keyword and temporal. The methodology of this paper consists of Spearman rank correlation, the time-lag correlation, and the regression analysis method. Various observation keywords are mined through a quantitative selection process to reflect different aspects of public concern over haze. The daily, monthly, seasonal, and annual characteristics of the public concern and PM2.5 pollution are examined to investigate the evolutionary trends over the past 5 years. The regression analysis is conducted to model the relationship between PM2.5 pollution and its public concern, and the regional differences and the trends of related respiratory diseases are further discussed. The results show that most of the observation keywords have a strong positive correlation (ρ > 0.6, P < 0.001) with PM2.5 concentration, and the public is more concerned about the real-time PM2.5 index and less about pollution prevention. The findings also tell that the decrease in PM2.5 concentration in summer is more significant than in winter, and the decrease in the time lag between PM2.5 concentration and public concern, from 2 days in 2014 to 0 days in 2018, may indicate that the public has become increasingly sensitive to the haze and the public tends to be more vigilant over haze in heavily polluted seasons. The relationship between PM2.5 and public concern can be well portrayed by the cubic model. The discussions present that there exist significant differences in the distribution of PM2.5 pollution and its public concern among 31 provinces and PM2.5 pollution has a significant negative effect on public health. Finally, some policy implications are provided. This paper could improve the understanding of public concern about pollution by enabling policy-makers to formulate a set of more effective pollution control measures.