In 2004, media reports on incidents of domestic social unrest in China were frequent and widespread. Beginning in early 2005, official Chinese sources began to report significant declines in the numbers of both mass incidents and citizen petitions. At the same time, media reports regarding incidents of social unrest tapered off.
Official Chinese sources attribute these developments to official success in resolving citizen grievances. But there is another possible explanation: official directives issued during this period that ban media coverage of mass incidents.
For example, article 21 of the joint 2005 Shenzhen Party and
regarding the implementation of the local response measures to mass incidents
clearly states "mass incidents are in principle not to be openly reported
on." Similarly, article 3.4(1) of the 2005 Jiujiang municipal
response plan for large-scale mass incidents similarly states that "large-scale mass
incidents are generally not to be reported on."
Now, tracking back the source for many of these local response plans, one quickly hits upon a 2003 central Party document – "Notice of the General Offices of the Communist Party Central Committee and State Council Regarding Improving and Strengthening the Work of Media Reporting on Emergency Incidents," [中共中央办公厅国务院办公厅关于进一步改进和加强国内突发事件新闻报道工作的通知（中办发22号)]. I haven't been able to find the text of this document. But various local plans seem to link the prohibition on media reporting with this central document, suggesting it may include a specific ban.
Indeed, Central Chinese authorities made efforts to enact a
prohibition on such reporting into national law. The original draft of the "Emergency Response
Law" （突发事件应对法）introduced in 2006
included a clause imposing fines of up to 100,000 yuan on news media outlets
that "independently reported" on public emergencies such as natural disasters,
industrial accidents, health emergencies, and public order crises (i.e. mass
incidents). At least partially as a result of popular
pressure, this clause was removed in the final version of the law passed on
August 30, 2007.
So, at roughly the same time as one sees a flurry of media reports regarding social unrest, you also see a central Party directive issued regarding media reporting on such incidents, followed in the following year or two by local authorities adopting "emergency response plans" that bar reporting on them. And then media reports dry up. It's at least somewhat suggestive of a cause-and-effect relationship . . . .
However, I'm still a little uncertain about concluding that a ban in the 2003 central government notice is the sole explanation for a decline in reports of social unrest starting in 2005. Would it actually take a year and a half for a central Party directive restricting media reporting on social unrest to take effect? On other issues (such as shifts in the reporting of SARS), central directives seem to go into effect much faster. I would definitely be interested in the thoughts of others on this issue.