Keynote Speaker: Sun Wenkai
Discussants: Wang Xianghong, Fan Yafeng, Yu Yihua, Xu Yiqing, Zhao Wenzhe

Dr. Sun Wenkai reported his findings on the issue of the residents' willing to express. He took the following 4 aspects in this quantitative analysis: individual characteristics, social characteristics, institutional environment and historical and geographical factors. By simple descriptive statistics, Dr. Sun found, everyone was trying to express their views in private, but not in public. The correlation between the explanatory variables average and the explained variables average in 30 cities showed: the higher the corruption was, the lower expression freedom was; the less government communicated, the lower expression freedom was; the government violent enforcement would reduce expression freedom; government information disclosure would enhance expression freedom; the higher news attention and trust were, the higher expression freedom was.

Subsequently, Dr. Sun Wenkai used various econometric analyses, controlling other variables in the case, and then he found that: the level of corruption had no effect on the degree of expression freedom.

Prof. Wang Xianghong suggested adding specific scenario description to the questionnaire, linking the survey questions to the actual city when analyzing, and adding objective variables to model to increase the amount of information.

Prof. Fan Yafeng introduced the status of quantitative study in the constitution science. He proposed to optimize the analytical framework. He pointed out that the reason why corruption and expression freedom was not a significant association may be the result that the civic virtue level and the attitudes on corruption had been changed.

Prof. Yu Yihua said, Ordered Logit Model had strong implicit assumption, so Dr. Sun could try Generalized Ordered Logit Model and Multinomial Logit model and do Wald test.

Xu Yiqing suggested that democracy may not be directly equivalent to one vote. Prof. Frank Knight had a very profound idea: the one which could promote public discussion was a democratic society.

Zhao Wenzhe thought that the trust in government was an explanatory variable instead of a stock, taking it directly into the regression would lead to an uneasily explanation. He also suggested using simultaneous equation estimation when quantitative analysis.

Mar. 26th, 2010