China: Statistics of Mass Incidents

Friday, 24. November 2006, 10:22:08

By Roland Soong
November 15, 2006

Since I am a statistician by profession, I get very sensitive and sensitized to numbers and their exact meanings. The number that has been bothering me for some time is the number of "mass incidents" in China. This particular number is one of the most frequently cited numbers for China (well, not as often as the total population of 1.3 billion people, about the same as the 123 million Internet users and more than the US-China trade deficit/surplus). The reason for the frequent citations is that it is favored for certain types of discussions, such as the "Coming Collapse of China" theory. For example, it is frequently cited that there were 87,000 "mass incidents", which then gets spun into (365 days) x (24 hours per day) x (60 minutes per hour) / (87,000 incidents) = 6 minutes per incident -- every six minutes, another mass resistance against human rights violation occurs in China! How shocking! And how could a nation stay together at this rate!

But I am not comfortable with some of these characterizations. Here is an illustration:

(Henan Business News via 6Park). In Xinzheng City, Henan Province, several tens of thousands of emotionally worked-up people showed up spontaneously together by word of mouth and surrounded the police.

The reason why these people were on the streets was due to an incident on the night of June 25, 2005. An elderly couple named Li ran a family small enterprise in which they sold agricultural equipment accessories. On this night, they were robbed and died after being stabbed more than 110 times. The robber stole a few cartons of cigarettes and some bicycle tire tubes from the shop. The entire city was outraged. On September 15, news came that the police had apprehended the suspect in Suzhou and was bringing him back to stand trial. Twenty thousand citizens poured into the streets. They were banging on gongs and drums and thanking the police for solving the case. The relatives of the victims were crying their eyes out too in appreciation.

This improvised public assembly was unauthorized, it was massive and unprecedented and it had a huge social impact in the city. Thus, it qualifies as a 'mass incident.'

So was this a 'mass incident' in support of the "Coming Collapse of China" theory?

I have my doubts.

Recently, a new number on the subject was published.

(Reuters) China says protests, riots down a fifth this year. November 7, 2006.

The number of protests and riots by discontented Chinese citizens fell by more than a fifth in the first nine months of 2006, a senior official was quoted as saying in reports seen on Tuesday. Chinese police dealt with 17,900 "mass incidents" from January to September this year, the vice minister of China's Ministry of Public Security, Liu Jinguo, told a police meeting on Monday, according to the official Xinhua news agency. This was a drop of 22.1 percent on the number of protests, riots, mass petitions and other "mass incidents" in the corresponding months of last year, Liu said.

(Los Angeles Times) China says it's calmed down. By Mark Magnier. November 8, 2006.

The number of "mass incidents" in China, a reference to protests, riots and other forms of social unrest, fell by one-fifth in the first nine months of 2006, according to Chinese government statistics released Tuesday.

The official New China News Agency, quoting Liu Jinguo, a vice minister of the nation's Public Security Ministry, reported that police dealt with 17,900 disturbances from January through September, a drop of 22.1%. At the same time, Liu warned that unapproved religious groups gained in number and clout.

Government statistics in China have long been viewed with skepticism by those who say they tend to be inaccurate and engineered for political purposes. With President Hu Jintao and Premier Wen Jiabao having made social stability a cornerstone of their administration, some analysts wonder whether the statistics are geared toward showing progress on that front.

"The government has never defined what 'mass incidents' refer to, so it's hard to tell if we're comparing apples and oranges," said Robin Munro, research director of the China Labor Bulletin, a Hong Kong-based activist group that monitors labor conditions and worker complaints in China. "I'm instinctively suspicious of official Chinese statistics, which tend not to be reliable, especially when they're dealing with social instability," he said.

17,900 is the number of "mass incidents" for the first nine months of 2006. If the same rate is maintained, then the total number of "mass incidents" for the entire year 2006 will be 17,900 x 12 / 9 = 23,900.

If 17,900 represented a 22.1% drop, then the number of "mass incidents" for the first nine months of 2005 is 17,900 x 100 / 77.9 = 23,000. If the same rate was maintained, then the total number of "mass incidents" for the entire year 2005 was 23,000 x 12 / 9 = 30,700.

This recent number is at odds with the previous "numbers" for "mass incidents" (via (sourced to China Ministry of Public Security; Murray Scot Tanner, "China Rethinks Unrest," The Washington Quarterly 27, No. 3 (2004):137-156; US State Department):

( Data show social unrest on the rise in China. By Richard McGregor. January 19, 2006.

Anti-social and mob violence in China rose sharply last year, according to official statistics released on Thursday by the Public Security Bureau, confirming anecdotal evidence of a growing willingness of citizens to take their grievances to the street.
'Public order disturbances' increased by 6.6 per cent to 87,000 in 2005 as a whole, but mob violence rose more quickly, by 13 per cent, the bureau said in an announcement posted on its website. The bureau counts four different kinds of incidents under the overarching classification of 'public order disturbances' but did not define them in any detail in Thursday's release.
The figures on 'disturbances' are consistent with a previous statement by Zhou Yongkang, the public security minister, who has said the number of 'mass incidents', or protests, rose by nearly 30 per cent in 2004 from 2003 to 74,000.

The first observation is that the 87,000 refers to 'public order disturbances' but it was labeled in the chart as 'mass incidents.' A 'public order disturbance' may or may not be the same thing as a 'mass incident.' If 'public order disturbances' increased by 6.6 percent to 87,000 in 2005, then the number of 'public order disturbances' in 2004 was 87,000 / 1.066 = 81,600. This may be 'consistent' ('in the same ballpark') with Zhou Yongkang's 74,000, but it is not the same number. So the above chart looks like it has 'apples and oranges' ('mass incidents' and 'public order disturbances').

( Beijing reports decline in protests. By Richard McGregor. November 8, 2006.

Liu Jinguo, a vice-minister at the Public Security Bureau, said police had dealt with 17,900 "mass incidents" in the first three quarters, down 22.1 per cent on the same period of last year. "Mass incidents" are defined more narrowly than "public order disturbances", of which there were 87,000 last year, up 6 per cent on 2004, according to government figures.

HRIC points out that 'public order disturbance' (????????) includes (but is not limited to) 'provocation/troublemaking, gambling, obstruction of official business and mob fighting' (???????????????????). This is different from the recent Reuters report (11/07/2006) about 'mass incidents' being "protests, riots, mass petitions and other 'mass incidents'."

So what we have is a lot of confusion about the terms 'mass incidents' and 'public order disturbances.' At this point, let me track back to examine some primary documents about the definitions of 'mass incidents.'

The first appearance of the term 'mass incident' (?????) was apparently given by Minister of Public Security Zhou Youkang. This was originally published in Ta Kung Pao on June 5, 2005.

[in translation]

In speaking about mass incidents (?????) arising from conflict among the people, Zhou Yongkang said that mass incidents are an outstanding problem affecting social stability, with five features that require attention.

1. The number has obviously increased and the scope has expanded. From the 10,000+ mass incidents in 1994 to the 74,000+ mass incidents in 2004, the increase has been more than sixfold. The number of participants has increased from the 730,000 persons in 1994 to the 3,760,000 persons in 2004, for more than a fourfold increase;

2. The scope has expanded. Mass incidents occur in cities, rural villages, enterprises, governments, schools and various domains and sectors and they occur in all the provinces, autonomous regions and municipalities;

3. The main participants in mass incidents are more diversified, and include dismissed workers, farmers, urban dwellers, enterprise owners, teachers and people from various social strata;

4. The methods are extreme, including laying siege and attacking party and government offices, blockading public roads, stopping trains and other situations;

5. The tendency is towards greater organizing. There are sometimes even spontaneously rising organizations with certain leaders.

This does not help much for the purpose here, because Zhou Yongkang really did not explain what a 'mass incident' is. We confirmed that Zhou Yongkang used the term "mass incidents" (?????). But we still don't how Zhou's 'mass incidents' are different from the most recently mentioned 'mass incidents.'

Two days later at the next State Council press conference on July 7, 2005 (XinhuaNet), the term was brought up and highlighted. Here is the translated transcript from the Congressional Executive Commission on China:

At a July 7, 2005, State Council Information Office press conference, a Reuters journalist asked Li Jingtian, then the Deputy Director of the Party's Organization Department:

In recent months, China's countryside has witnessed a number of riots. What method does the Chinese Communist Party use to deal with riots? ...

Deputy Director Li responded:

We term the incidents in China's rural areas "mass incidents (?????)" and not riots.

Li Jingtian specifically rectified the term "riot" used by foreign correspondents and used the term "mass incident" instead. The Xinhua article then proceeds to cite that between 1993 to 2003, the number 'mass incidents' had increased from ~10,000 to ~60,000. These are the numbers that appeared in the chart of "mass incidents" from 1993 to 2003.

From the Congressional Executive Commission on China,

According to a March 2, 2006 Beijing News article and a transcript of the press conference appearing on the Chinese government's Web site, at a March 1 State Council Information Office press conference, an Agence-France Presse reporter asked Ouyang Song, Deputy Director of the Party's Organization Department and Deputy Director of the Leading Group for the Party's "advanced education" campaign:

What we have heard regarding villages is not like [what you have described] . . . Every month, we hear at least twice about rural farmers engaging in mass petitions, demonstrations, or riots because of land requisitions or the abuse of power. These are the cases we have heard of, there are many others we haven't heard of. Official statistics also show that more and more incidents of social discontent are taking place, more than 87,000 last year.

Ouyang responded that:

First, I want to correct two of your statements: I have heard of nowhere in China that has experienced riots, and mass incidents are not increasing.

China is the most stable country in the world. This point has already been recognized by the world. As to the fact that a few particular areas have experienced some mass incidents, in a country as big as China, in an era experiencing rapid development, this should not be considered unusual . . . as to those few areas with mass incidents, the Party and the government are highly attentive and concerned.

Ouyang also said that mass incidents constitute only a small portion of the 87,000 public order disturbances in 2005 reported in public security statistics in January 2006.

So now we get the idea that mass incidents are a subset of public order disturbances, but still we do not get an actual number of mass incidents. How small is that portion? We do not know.

If you come down to it, we need to know how mass incidents are defined. I have not been able to find a full definition but I have found some bits and pieces. Here is how the Ningxia Autonomous Region Governmentdefines an 'incident' as being one of the following six types for the purpose of reporting upwards to the State Council:

1. Charging and laying siege to a county- or higher-level party or government department, political-legal department, military, armed police, news and other critical departments, thereby causing bad influence; attacking, vandalizing, looting and committing arson against a town- or higher-level party or government office with serious impact on social stability;

2. Armed clash between groups of people causing injuries and deaths;

3. Terrorist activities, violent incidents, illegal organization and assembly by hostile forces and separatists;

4. Causing the disruption of railroad and state highway traffic, or the blockading of major cities, traffic hubs and urban transportation;

5. A group petition at the county, city or autonomous region level with:
- a likelihood of proceeding to Beijing;
- more than 100 persons involved;
- violent tendency or the possibility of becoming violent;
- seriously affecting social stability, sensitive locations or business operations;

6. More than 60 people involved in assembly, troublemaking, work strike, business strike, school class strike, etc with relatively high impact.

Just when you think you've got it, here is another set of definitions for the Jiangsu provincial government (via

Mass Incidents

Extraordinarily important mass incidents include:

1. An incident involving more than 5,000 participants with serious impact on social stability;

2. An incident involving either laying siege and charging county-or higher-level party, government or military departments or other critical departments, or the attacking, vandalizing, looting and/or committing of arson against town- or higher-level party, government or military departments;

3. An incident in which the participants were particularly antagonistic and engaged in large-scale attacking, vandalizing, looting, arson and other criminal activities;

4. An incident interrupting major railroad arteries, state highways, expressways, major traffic hubs or urban transportation for more than 8 hours, or interrupting and/or preventing work at major state construction projects for more than 24 hours;

5. An incident causing more than ten deaths and/or more than 30 injuries with serious impact on social stability;

6. An incident at a university by which students engage in large-scale marches, assembly, hunger strike, sit-in and petition outside of the school without permission, thereby leading to chain reactions in other regions with serious impact on social stability;

7. An incident in which more than 500 people clashed with weapons and resulting in serious injuries;

8. An incident in which more than 10 people engage in a prison riot;

9. An incident in which the impact on social stability extends beyond the province through interactive chain reactions;

10. An incident not covered by the above but must be treated as an extraordinarily important mass incident.

Important mass incidents include:

1. An incident involving more than 1,000 persons but fewer than 5,000 persons in illegal assembly, petitioning, troublemaking, strike (business/school) and so on; or an incident involving fewer people in illegal assembly and petitioning with wide impact including the possibility of going to Beijing;

2. An incident with more than 3 but not more than 10 deaths, or more than 10 but not more than 30 injuries;

3. An incident in which information first appeared on university networks to establish ties, incite and mislead in order to form a joint action across universities that seriously disrupt or even paralyze normal educational activities, or the leaking of test questions in the university entrance tests;

4. An incident in which more than 200 but fewer than 500 people clashed with each other using weapons and causing serious injuries;

5. An incident involving national and international religious ethnic religious issues that seriously affect national unity;

6. An incident that was caused by property rights violation, pollution or destruction of land, mines, water supply, forests, water surfaces and marine space;

7. An incident in which the impact on social stability extends beyond the province through chain reaction, or an incident that has already caused serious damages and losses with the prospect of expanding and escalating;

8. An incident not covered by the above but must be treated as an important mass incident.

Notice that they make a distinction between 'extraordinarily important mass incidents' and 'important mass incidents' in Jiangsu. At this moment, I will say that I still don't know what exactly a 'mass incident' is or its difference from 'public order disturbance.' And I won't know until there comes an explicit statement as to what the definition for each statistic is.

Next we examine the primary documents with respect to "public order disturbances." Here are some partial excerpts from the People's Republic of China Code of Criminal Law (via Xinhua):

Chapter 6. Crimes that damage the admiistration of social order

Section 1. Crimes that disrupt social disorder

Article 277. (Obstruction of public business)

Using violence or threats to prevent state government workers to carry out their duties in according to the law. Three or fewer years in jail, detention or supervision or fines.

During a natural disaster or a suddenly breaking incident, using violence or threats to prevent Red Cross personnel to carry out their duties. Same penalty as above.

Deliberately obstructing the national security and public security bureau to carry out their national security duties to cause major consequences evern though no violence or threats were used. Same penalty as above.

Article 278. (Incitement to use violence to resist law enforcement)

Inciting the masses to resist the enforcement of nationl laws and administrative regulations. Three years or fewer in prison, detention or supervision, or deprivation of political rights; if the consequences were major, more than three years but less than seven years in prison.

... Article 280.

... Forging or modifying resident ID cards. Three years or fewer in prison, detention or administration, or deprivation of political rights; if the consequences were major, more than three years but less than seven years in prison.

Article 281. (Illegal manufacturing and trading in police equpiment) ...

Article 282. (Illegally obtaining state secrets) ...

Article 283. (Illegal manufacturing and training in espionage equipment) ...

Article 284. (Illegal use of surveillance and snooping equipment) ...

Article 285. (Illegal intrusion into computer systems) ...

Article 286. (Satoage of computer information systems) ...

Article 287. Using computers for financial fraud, theft, corruption, embezzlement of public funds, stealing state secrets and other crimes. ...

Article 288. (Interruping with wireless telecommunication systems) ...

Article 289. Forming mobs to assault, vandalize and loot to cause death and injuries. ...

Article 290. Forming mobs to disrupt social disorder. Forming mobs to attack state organizations. ...

Article 291. Forming mobs to disrupt order in public locations (such as train stations, piers, airports, commercial malls, parks, movie houses, exposition halls, sports arenas and others) and transportation. ...

Article 292. Forming mobs for armed fights. In a mob armed fight, the leaders and other active participants sentenced to three years or fewer in prison, detention or supervision. The penalty increases to more than three years but less than ten years for any of the following: (1) participation on multiple occasions; (2) the number of persons in the mob is large and causes bad influence on society; (3) forming a mob at public places or transportation hubs, causing serious disruption in social order; (4) fight was doncuted with weapons ...

Article 293. Provoking and seeking trouble. ...

Article 294. (Organizing, leading and participating in underworld criminal organizations) ...

Article 295. (Teaching the techniques of committing crimes) ...

Article 296. (Illegal assembly, demonstration and protest) ...

Article 297. Carrying weapons, restricted knives or explosives to participate in assembly, demonstration and protest. ...

Article 298. (Discrupting assemblies, demonstrations and protests) ...

Article 299. (Insulting the national flag or insignia) ...

Article 300. (Organzing and using churches and evil cult sects to stop law enforcement)

Article 301. (Organizing group sex) ...

Article 302. (Stealing and defiling corpses) ...

Article 303. (Gambling) For the profit purposes, organizing group gambling, establishing gambling dens or otherwise working in gambling ...

Article 304. (Deliverately delaying of the delivery of mail) ...

Here is my guess as to what has happened (remembering that statistics are never ever totally objective but they are necessarily socio-politico-economic artefacts). I believe that there have been three series of numbers.

The first series was labeled 'mass incidents' and ran from 1993 to 2004. In 1993, the number was 8,700; in 2004, the number was 74,000. This series has been discontinued since. The corresponding 2005 number has never been released.

The second series was labeled 'public order disturbances' and began in 2004. In 2004, the number was 81,600; in 2005, the number was 87,000. It is not comparable to the first series, but appears to be a substitute. This count purportedly covers: provocation/troublemaking, gambling, running underworld criminal organizations, obstruction of official business, mob fighting, delaying the delivery of mail, holding mass orgies, computer hacking, making and selling fake police uniforms, forging ID cards, burning national flags and corpse desecration among other things.

The third series was labeled 'mass incidents' and began in 2004. All we know at this point is that the January-September 2005 number was 23,000 and the January-September 2006 number was 17,900. This count purportedly covers: protests, riots, mass petitions and other "mass incidents."

Why should there be multiple time seies of data with different meanings? In a way, this is understandable -- you produce a time series of data that contain all manners of incidents (e.g. disco brawls, gambling den raids, protests petitions, sit-ins, riots, etc) but the western media and overseas hostile forces prefer to position this as "public riots against human rights violations." At this point, it becomes understandable if you would rather split your data stream into 'mass incidents only' and the more generalized 'public order disturbances.'

There is also a huge difference between the subjects in 'mass incidents' and 'public order disturbances.' A 'mass incident' refers to the incident which usually involves large numbers of people (e.g. 10,000 people rioting at a university over their diplomas). A 'public order disturbance' is an individual crime and the number of people do not come into it. For example, if I was caught selling fake police uniforms, then I am the sole criminal. This is a 'public order disturbance' (Article 281) because my action caused social mistrust of people in uniforms. The number of actual people who lost their trust is not known to any precision. As another example, if a group of five hackers went and crippled the People's Daily website, they would be guilty of "disturbing the public order" (Article 285 and Article 286 of the PRC Code of Criminal Law). This is one incident with five criminals. How would that be counted in the statistics? One or five? But the number of people affected is not known with any precision. Neither of these two examples may be considered "mass incidents."

If you don't like statistics such as these, you can refuse to use them, or you can use them and state your qualifications (e.g. lack of transparency, dubious quality, inconsistency, suspicion of manipulation, etc). But it does not mean that you can re-interpret them to suit your own political needs.

I do not enjoy being put in this position. I attribute all my grief and discontent to certain Chinese bureaucrats thinking that manipulation and obsfucation is the best approach. It isn't. And I promise that I will remind you of this fact every time that another update is issued.

Meanwhile, I know that I will continue to read about "every six minutes, another mass protest against human rights violation occurs in China" while knowing full well that we may be talking about disco brawls or gambling den raids.

More about Mass Incidents In China by Roland.

Share this

Rethinking ‘Capitalist Restoration' in China