Social unrest and social stability in China

(1)

0

Master Thesis

Political Science: International Relations

Political Elites & the Economy

Social Unrest and Social Stability in China

Robyn Kelly-Meyrick

11246359

Supervisor: Gijs Schumacher

Second Reader: Aofei Lu

(2)

1

This paper examines which of the most commonly contested issues in China (land, labour and environmental grievances) is most threatening to social stability, and potentially to regime legitimacy. Based on a framework set out by Yongshun Cai, the research collects information on protest events across the three most common causes of discontent and analyses the ‘forcefulness’ of each protest as measured by the core indicators of size and violence, in relation to the government’s response in order to test whether one type can be said to be more threatening to social stability in China than others. The wider implications of the analysis are then discussed.

Keywords: China, Protest, Environmental Degradation, Land Disputes, Labour Strikes, Social Stability, Regime Legitimacy, Discontent, Economic Growth.

(4)

3

Introduction

The Chinese are no strangers to protest. Although the one party system of governance has maintained tight control over China’s social and economic development for decades, protests are and always have been a part of the societal fabric. Whilst some of the most prominent public revolts in Chinese history have been met with extreme force – the events which unfolded on Tiananmen Square in 1989 being the prime example – the social and economic grievances resulting from a complex, rapid and on-going developmental process continue to give rise to widespread protests of varying scale and magnitude across the country.

The Chinese Communist Party (CCP) can be seen to have kept an iron grip over civilian uprisings through internet censorship, (often aggressive) suppression and punishment. The authorities will often detain protest leaders and those who clash with the police in riots, as well as individuals who seek to publish information pertaining the outbreak of protest – in 2016, journalist Lu Yuyu was detained for his protest-tracking activities on a personal blog entitled ‘Not News’ which provided detailed reporting and statistics on mass

demonstrations across China (Lin, 2017). Yet despite the common and not ill-founded perception of heavy-handed government approaches to popular resistance as the norm, large-scale protests in China are not always violently repressed or shutdown.

Indeed, the very fact that so many large-scale protests reach news outlets worldwide and that NGOs like the Chinese Labour Bulletin document thousands of protests through strike maps suggests that at least to some degree, protest is a tolerated expression of discontent (although it is technically illegal). In the first half 2016, strikes and protests rose by almost 20%, and 1,456 instances were recorded by the end of June (Financial Times 2017). Whilst many instances of protest reported are suppressed by police and government efforts, there are also numerous examples involving large numbers of people where police have shown considerable constraint, and in some cases concessions have been made by government officials: in 2012 for instance, a protest which gathered crowds as large as 50, 000

(5)

4

local coastline, with the mayor announcing that the project was permanently cancelled following the uprising (Wade 2017). In the latter example however, riots between police and protestors left many injured and raised the question as to whether the instance would spark further environmental protests and how government response may alter in the future. As Cai states; ‘depending on their past experiences, local governments may become more suppressive or more tolerant’ (Cai 2010, p15).

Over the past decade, research publications have reported common reasons for protest to include land disputes, environmental degradation, labour conflicts and ethnic strife, however conclusions on which of these issues produces the highest number of protests have varied. In 2010, research by Tong & Lei placed labour disputes as the most frequent cause of mass incidents, constituting 45% of all incidents. In 2012, a report produced by the Europe China Research and Advice Network stated that land disputes account for 65% of all cases of social unrest in China (Gobel & Ong 2012, p11) whilst in 2016, a report by

Tanigawa-Lau & Headley stated ‘although previous studies of protests in China cite land disputes as the leading cause of protests on the mainland, our data shows that at least in 2015, this is changing. That year, around 68% of protests were wage-related’ (Tanigawa-Lau & Headley 2016). The latter report is based on data collected from the Chinese Labour Bulletin and the Google Database of Events, Language and Tone, which cites social security as the second most common reason for protests and environmental concerns as the third. Due to the difficulty in obtaining transparent data on protest in China – as Knight states: ‘the more authoritarian the state, the more difficult it is for researchers to obtain the relevant information’ (Knight 2012, p4) - statistics on the most common reasons for discontent can only be regarded as rough estimates, however the existence of NGOs and increasingly sophisticated data collection methods is allowing for more precise monitoring.

The relationship between protest and social instability is often discussed in the discourse of discontent in China, for instance a report by Reuters based on research conducted by the Chinese Urgent Action Working Group on land-based protest stated that ‘China risks

growing social instability and even violence if the government does not take effective action to address rising public anger about forced evictions and demolitions’ (Reuters 2010). In 2012, the China Growth Centre of the University of Oxford reported that the number of civil

(6)

5

incidents had increased rapidly in recent years, ‘suggesting that social instability – albeit localised and uncoordinated – is on the rise’ (Knight 2012). The latter research did not find that social instability has thus far had a seismic effect on political stability; however it references a number of works which found a correlation between economic growth

stagnation and social instability which could ultimately lead to political instability. Yongshun Cai’s extensive body of work on discontent in China also gives attention to social stability and regime threat or political instability which will be discussed in greater detail in the theoretical framework of this paper. Despite the existence of a wealth of studies focusing on protest and social and/or political stability however, research which directly connects the type of issue which causes protest with social and political instability is sparser.

The aim of this research is to establish a causal relationship between the most common types of protest in China and the threat posed to social instability in order to examine which reason for discontent is giving rise to the most cases of social instability which could

ultimately give rise to political instability and regime threat. The research question is formulated as follows:

Which of the most commonly contested issues in China reported are most threatening to social stability?

An important sub consideration to this primary research intent is to examine how social instability interacts with political instability in China. The framework set out by Yongshun Cai (which is also utilized in Tong & Lei’s work on mass incidents and government response in China) is explained and operationalized in the subsequent sections of this paper in order to produce a methodology suitable for answering the above question. Whilst research on the most common causes of protest have varied over the last decade, land related grievances, labour disputes and environmental concerns are frequently referred to as the top three issues, and will therefore constitute the focus of the analysis in this research.

(7)

6

Theoretical Framework

The Chinese Communist Party and Chinese Capitalism

Since protest in China cannot be understood outside of China’s developmental path and style of governance, a brief discussion on the idiosyncrasies of Chinese development is relevant to the theoretical foundation of this research. The Chinese variation of capitalism is one of contradictions; whilst the neoliberal notion of capitalist development is heavily entwined with democracy, China balances capitalist accumulation with continued party-state hegemony (McNally 2012, p757). Sino-capitalism’, as McNally terms it, functions by reliance on informal business networks as opposed to legal codes and transparent rules, and assigns the Chinese government a leading role in fostering and guiding capitalist

accumulation. The structure of Sino-capitalism therefore combines top-down state-led development with bottom-up entrepreneurial private capital accumulation (McNally 2012, p744).

Sino-capitalism has strongly characterised China’s rise into the global economic system, and the opportunities and threats which face the country today are resultant of this two-fold developmental path which has seen the Chinese Communist Party retain control whilst rapidly accelerating economic growth. Indeed, the concept of Guanxi politics - defined as ‘a patriarchic structure of family ownership and control in network of reciprocity’ (McNally 2012, p750) - has been integral to China’s development and has had widespread effects on economic entrepreneurship, allowing for ‘Chinese entrepreneurs to link up with state officials, access finance, overcome government indifference, build trust and compensate for institutional uncertainty’ (McNally 2012, p752). One such effect of Guanxi politics

materialised as the reform leadership of 1978 gained traction and gave policy priority to rapid economic development in order to restore political legitimacy, which had been damaged by the economic stagnation resulting from the Great Leap Forward and the Cultural Revolution (Knight 2012, p17). Under Deng Xiaoping’s leadership, the central government delegated many responsibilities and powers to already existing autonomies at the local level, giving cadres further political autonomy to circumvent central rules

(8)

7

2012, p752). Cadres appointed by higher level officials therefore had new incentives to improve local economic performance, fuelled by the promise of promotions or bonuses for carrying out the state’s redistributive objectives (Knight 2012, p17).

The reforms proceeding 1978 allowed for the central government to control finance and the commanding heights of industry through the dominance of state owned enterprises (SOEs) whilst simultaneously benefiting from private or hybrid ownership firms (McNally 2012, p751). Sino-capitalism’s development has certainly illustrated the resilience of the political structure headed by the Chinese Communist Party, yet the system of patronage has, as Knight states, ‘produced a breeding ground for rent-seeking and the corruption of bureaucrats and party officials, especially at local levels’ as cadres compete for rewards delivered from further up the hierarchy (Knight 2012, p13). The knock-on effect of the latter structural issue is that social instability can often occur as a direct result. In a survey by the China Household Income Project in 2002, corruption was reported by 21% of respondents to be the most serious social problem after unemployment and layoff - which can sometimes arise as a consequence of corruption (Knight 2012, p13). As corruption is most prominent at the local level, the outbreak of social unrest is often due to the self-seeking actions of local government officials which include neglect of pollution, unfair compensation for land

compensation, excessive taxation and bribery (Knight 2012, p15). Whilst up until 2005, rapid economic growth ‘was viewed as the overriding means of protecting against social

instability’ a second objective has since been added which links the central government’s goal of a ‘harmonious society’ with the avoidance of social instability (Knight 2012, p16). Since the implementation of the second objective meant high rewards and conversely reprimands for local authorities regarding the maintenance of social stability, it is hardly surprising that a survey of local officials in 2005 found that ‘upholding social order’ was the perceived as the most critical task faced by them(Minzer 2005, p19). As the outbreak of social unrest is usually localised, and the responsibility for maintaining social stability lies for the most part with local officials, the central government will usually take a back seat when it comes to intervention, unless the social instability is so severe that it threatens the regime’s legitimacy (Cai 2010, p5-6).

As a result of China’s developmental path, it is hardly surprising that protests arising over land disputes, labour rights and environmental concerns rank among the top causes of

(9)

8

protests. Land requisition has been particularly prolific in rapid developing areas, or places where there are mining discoveries, with plans to convert hundreds of millions more farmers into city dwellers: “In land acquisitions and forced demolitions, for example, many officials often overlook public interest,” Shan Guangnai of the Chinese Academy of Social Sciences told Bloomberg (Bloomberg 2014). Labour strikes have proliferated as the drive to maintain productivity and maximize profit has intensified: as reported by Headley &

Tanigawa-Lau; in 2014, the National Bureau of Statistics of China wrote that per capita, disposable income amounted to $3,300. According to the 2015 U.S. Congressional Committee Report on China, the average wage growth in China has slowed from 22% in 2011 to 10% in 2015. The report further found that minimum wages in many Chinese cities were “insufficient to cover basic living expenses” (Headley & Tanigawa-Lau 2016). Protests tend to be concentrated in coastal regions – the heartlands of China’s industrialisation, and since Xi Jinping took office in 2013, most protests have erupted in Guangdong province, in the major factory cities of Shenzhen and Dongguan (Headley & Tanigawa-Lau 2016). A further fallout of economic expansion is of course issues relating to the environment – heavy pollution and the creation of chemical plants, waste incinerator projects and coal-fired power plant expansions cause widespread mass protests across the countries, as farmers, villagers and city dwellers object to toxic health concerns.

Government Response & Protest in China

Assessing the relationship between the outbreak of protest and government response in China is crucial to producing an understanding of which type of protest may be causing the most severe threat to social instability, as the features of the protest and the government’s actions towards such an event provide an indication of the level of threat posed. As briefly mentioned, the hierarchical system of governance in China results in differing priorities for the central and local governments in handling social unrest: perceptions of the costs and benefits at different levels, as Cai postulates, therefore shape the official’s choice of the mode of response. Whilst the 2008 China Survey found that the most common pattern of government trust is ‘hierarchical trust’ – of 3989 respondents, 44.8% expressed strong or modest confidence in provincial and/or county leaders (Li 2016, p106) – an assessment of five national surveys by Li (including the 2008 survey) finds that although the results suggest

(10)

9

stronger trust in central government and weaker trust in local government, the implicit pattern is one of ‘partial trust in the central government in that manifest distrust of local authorities reflects latent doubts about central leaders’ (Li 2016, p113). Existing studies on hierarchical trust thus express a mixture of ‘stronger confidence in the central government’s policy intent and weaker confidence in its capability to monitor and discipline local deputies’ (Li 2016, p115). These findings are consistent with empirical data (i.e the data collected by Cai, and by numerous reports of protest which were gathered for this research) on protest which suggests that citizen’s resistance in China is mostly directed at local governments, or at least it begins with the targeting of local governments as resentment and distrust over the handling of grievances often manifests at the local level.

Due to the central government’s concern over regime legitimacy, it is often more tolerant of non-regime threatening social unrest than local governments: of the 261 cases of protest collected by Yongshun Cai between 1994 and 2007, the central government intervened in only 13 cases, and 10 were successful. Central government intervention usually occurs when the need to strengthen and maintain regime legitimacy is severe, and is often driven by media exposure and casualties, which can spark widespread public outrage. If the central government receives criticism in media coverage, by offering concessions to protestors and punishing local government officials, it can often enhance its governing mandate. Indeed, repressing social unrest when the claims made are legitimate is damaging to the regime’s legitimacy, as often protests are the result of a breach of law by local officials or village leaders (Cai 2010, p6). The pressure on local government officials to maintain social stability is consequentially heavy, as their political careers are often determined by their ability to fulfil this responsibility as delegated by the central government.

In Cai’s framework, which was designed to assess the reasons for success or failure of

protests in China, government response to collective resistance depends on the forcefulness of resistance and the costliness of demands: forcefulness, he states, ‘not only threatens the social order, but it also enhances the salience of an issue’ (Cai 2010, p15). Cai’s research defines an action as forceful ‘if it has more than 500 participants, involves casualties, or receives media coverage’ (Cai 2010, p44). According to the (unspecified) police department, an action with more than 500 participants is regarded as a large-scale action, and actions

(11)

10

with more than 1000 participants are regarded as especially large in scale (Cai 2010, p48). Cai further notes that since incidents that involve deaths or media coverage are limited, the size of the protest group is the most important indicator of the chance of success (Cai 2010, p187). In variation with Cai’s report, Gobel & Ong state that ‘according to the Jiangsu provincial government, important mass incidents involve between 1000 and 5000 people, and extraordinary mass incidents involve more than 5000 participants’ (Gobel & Ong 2012, p22). However, Tong & Lei, whom analyse government responses to 248 cases of social unrest from 2003 to 2009 also define a large-scale incident as one involving more than 500 participants. The latter research documents that in 20 of the 248 cases local officials were disciplined by higher-level governments and that each of those situations involves one or more of the following: a large number of protest participants, protestor assault of

government institutions, occurrence of deaths or injuries, the incident had coincided with an important public event such as the Olympic Games, or the protest had evolved into a mass riot (Tong & Lei 2012). Cai, Tong & Lei seem to converge somewhat therefore, on what constitutes a mass incident, and as to the criteria which produce a more ‘forceful’ protest, i.e one which has the capacity to threaten social stability, or at the extreme level, political stability or regime legitimacy (Cai 2010, p44).

Yongshun Cai’s work references four modes of government response: (1) concessions (i. e., citizens’ demands are met); (2) concessions with discipline (i. e., citizens’ demands are met, but some or all participants are punished); (3) tolerance (i. e., citizens’ demands are ignored, but the government tolerates their resistance and refrains from using force); and (4)

repression (i. e., citizens’ demands are ignored, and some or all participants are punished) (Cai 2010, p5). These responses are rank ordered in the sense that forms of concessions convey some level of successful resistance, whilst tolerance and repression express a more negative or dismissive outcome of resistance. Cai’s analysis of 257 cases revealed that ‘forceful’ action increases the odds of success, especially in cases where the cost of citizen’s demands was low. As mentioned, a ‘forceful’ protest is threatening to social stability, consists of a large number of people and can result in violence, casualties and occasionally widespread media coverage.

(12)

11

Whilst Cai’s research assesses successful protest based on ‘forcefulness’ and government response, with concessions being deemed as a successful outcome, the aim of this research paper is to assess which type of protest is most threatening to social instability. Social instability is often a somewhat loose concept in research, as often it is not clearly defined, however Knight’s research for the China Growth Centre directly links the term to the occurrence of ‘mass incidents’ in the form of protests, as a result of complaints that are often linked to negligent local government actions, and more widely, economic and environmental grievances. Cai makes a comparison between Kowalewski & Shumaker’s research on protest outcomes in the former Soviet Union, which suggests that ‘militant groups can increase the probability of favourable outcomes regardless of size’ (Kowalewski & Shumaker 1981), and protest outcomes in China, stating that ‘in China, size does matter because large-scale resistance poses a more serious threat to social stability, while small-scale disruptive action may be suppressed’ (Cai 2010, p112). In Guangdong for instance, the criterion for a three-party body’s intervention into labour rights (consisting of the

construction sector government agency, the trade union and the association of construction enterprises) is primarily based on the number of workers involved – actions involving more than 1000 participants or drastic actions involving 500 – 10, 000 participants will fill the criteria for intervention (Cai 2010, p112). Based on previous research, therefore, the threat posed by a protest to social stability predominantly concerns the size of the protest (and to a lesser extent, casualties and media coverage), and the government response – in cases where the government has intervened and offered concessions, or concession with discipline (often local governments will discipline selective participants such as protest leaders or activists), it is likely that the threat to social instability was more intense and required action to maintain social stability or in extreme cases regime legitimacy (Cai 2010).

The aim of this thesis is to gain an insight into which type of protest – environmental, land, or labour – is posing the greatest threat to social instability using the framework set out above to collect relevant data about protest events across the three types. The size of the protest and the government’s response will be of paramount importance in measuring social instability as caused by the different types of protest. Although Cai references

‘casualties’ as an indicator of the forcefulness of protest, casualties will not be a central part of my methodology. This is because casualties are evidently a result of some kind of

(13)

12

confrontation, most usually between protestors and police forces deployed by the local government. Therefore, casualties are in fact most often part of the government response of repression, where force and detention is used to stifle the issue. Whilst I will still record violence – violence is indicative of injuries or ‘casualties’ - in my dataset, I expect there to be a strong collinearity between this variable and the dependent variable of government response.

Cai’s research found that of 90 cases where the protests were deemed forceful, the

government made concessions in 57 instances, repressed in 26 instances, and tolerated in 7 cases. Conversely where the protests were not forceful out of 171 cases the government made concessions in 21 cases, repressed in 131 cases and tolerated in 19 cases. As a result of these findings, I expect that the same could hold true for my dataset, and therefore pose the following hypothesis:

H1: As the size of the protest increases, the government response will decrease (from repression towards concession).

I subsequently hypothesise that one type of protest will show a larger decrease in

government response than the other types as the size of the protest increases. Thus one of the following three hypotheses will be accepted, whilst the others are rejected.

H2: If the protest type is land related, the government response will show a larger decrease than the other types of protest.

H3: If the protest type is labour related, the government response will show a larger decrease than the other types of protest.

H4: If the protest type is environmental, the government response will show a larger decrease than the other types of protest.

By confirming one of the above hypotheses, it should be possible to ascertain which of the three types of protest poses the biggest threat to social stability based on Yongshun Cai’s framework which holds that the more forceful a protest, the more threatening it is to social stability, and the more likely it is that the government will make concessions in order to

(14)

13

uphold social order.

Case Selection and Methodology

Case Selection

Previous research has tended to focus largely on physical protests, as opposed to petitions or legal action. Tong & Lei focused specifically on mass incidents (over 500 participants) because they are hard for local authorities to cover up, and therefore ‘it is more

manageable to collect information on large-scale mass incidents’ since they are reported by the media (Tong & Lei 2012, p489). Cai, Tong & Lei rely on media reports, personal blogs from activists, reports from the China Labour Bulletin (a Hong Kong based NGO) and

personal fieldwork (however only 9 of Cai’s cases are from fieldwork). Since the reporting of protest outbreak in China is often repressed, international media outlets and

non-governmental organisations are often the richest sources of information, and will therefore be utilized to gather databases of protest information for each protest type. Cases of physical protest: marches and disruptive actions such as strikes, sit-ins and road blockages will be taken into account. Since the protest events recorded will be separate instances in time, the research will take the form of a latitudinal study, and due to data limitations, it will reflect trends and an overall indication of which type of protest may pose the most threat to social stability in China at this time based on available information.

Data Collection

Collecting complete and verified data on protest events in China is nearly impossible at present: firstly, media outlets are often restricted from reporting on protest, and those that do rely on reports from eye witnesses, activists or first-hand observations. The exact size of a protest is difficult to confirm, for instance, and sometimes whilst it is reported that there has been violence, injuries or arrests, the precise figures are unclear. As a non-Chinese speaker, it has not been possible to use social media platforms such as Weibo (the Chinese alternative to Whatsapp) for data collection. There are, however, a number of English-Chinese sources and free press organisations which exist to provide extensive information on protests and strikes specifically. These include the Chinese Labour Bulletin (CLB), an

(15)

14

extensive organisation which ‘provides practical help and encouragement to workers in China engaged in collective struggle’ (Chinese Labour Bulletin, 2017), Radio Free Asia (RFA), a private non-profit international broadcasting corporation which provides ‘news and information to listeners in Asian countries where full, accurate, and timely news reports are unavailable’ (Radio Free Asia, 2017), and China Digital Times (CDT), ‘an independent,

bilingual media organization that brings uncensored news and online voices from China to the world’ (China Digital Times (CDT) 2017).

For the purpose of my research, I chose to focus on a relatively short, recent time frame from which to gather protest data, as I am interested in trying to gage which type of protest is most threatening to social stability at this time in China. I subsequently chose to focus on events which occurred between 2011 to present day. This coincides with China’s Twelfth Five Year Plan, which ran from 2011 to 2015, and also encompasses the Thirteenth Five Year Plan which ran from 2016 and will continue until 2020. China’s five year plans are a series of economic and social initiatives, the first of which commenced in 1953. Both the Twelfth and Thirteenth plans have sought to intensify anti-corruption efforts, as well as conserving farmland, improving rural production and living standards, and expanding the channels for farmer’s income (China Direct, 2017). The Thirteenth plan places specific emphasis on the ‘clean production’, the promotion of ‘orderly transition of land operation rights’, and the assurance of ‘legal land use rights of farmers’ thus it may be possible to observe some changes in land-related protest in particular over the last 7 years (Xinhuanet, 2017). In 2012, a report by Mingyu Dong in the China Digital Times stated that, according to a report by Amnesty International, the number of land grabs had dropped, which was ‘in part due to a shift in official focus away from economic growth and also a less confrontational approach to resolving social disputes’ (Dong 2017). Information from the Five Year Plans, and the government’s changing response to social disputes in recent years is important to keep in mind, as the socio-political context of the last 7 years which are being taken into account for this research will of course be expected to have an impact on the results, since protest is an inherent part of changing social and economic dynamics.

Having selected a time period, my general strategy has been to gather every single instance of protest which fits my criteria (i.e. a physical protest of over 500 participants on the topic

(16)

15

of land, labour or environmental concerns) from 2011 to 2017, from the resources available. Whilst searching for the sites which would provide me with the most abundant information, I consulted the South China Morning Post (SCMP), Hong Kong Free Press (HKFP), RFA, the blog of the now detained Lu Yuyu, CDT, The Guardian, Reuters, and the CLB, as well as conducting extensive Google searches to find relevant sources. For each resource, I scoured archives between 2011 and 2017 to find details on protest in China. What I found, was that Radio Free Asia offered by far the most frequent reporting on the outbreak of protests across the three types of my focus. The source tended to cover the same events which I was able to find from other sites, and many more. Whilst I used all of the above mentioned sources for data collection, Radio Free Asia became my main source for data collection, as the reports generally offered me the information I was looking for in terms of protest size, government response and in many cases provided details on casualties, violence, detentions and injuries. Every single case of protest from 2011 to 2017 matching my criteria from the source archives has been recorded.

The existence of the China Labour Bulletin’s strike map meant that for labour protests, I had access to an extensive database of recorded strikes, however even the extensive efforts of an organisation with a mission to provide information on the collective struggles still cannot provide a complete picture. As stated on the CLB website, ‘strike map figures, in the absence of government statistics have always been a sample of the broad spectrum and an indicator of trends in worker’s actions were never meant to be interpreted as a tally of the total number of strikes and protests in China’ (Chinese Labour Bulletin 2017). By using the database available, I was able to extract cases of protest which exceeded 500 participants from the period of 2011 to the end of 2016 (hereafter, the CLB stopped providing 24/7 monitoring and now aims to release reports every six months). I ended up with a database of 409 cases of labour protest detailing the estimate of size for each event, as well as location. Beyond this, it was more difficult to effectively code the government response from this data as although there was input for this, many of the cells were blank or vague, listing negotiation, government mediation, police, police arrests, and police beatings as the responses. As my scale of government response runs from 1: concession through to 4: repression, I had to decide how to code the CLB’s response terms. I chose to code ‘negotiation’ and ‘government mediation’ as either 1: concession or 2: concession with

(17)

16

discipline (depending on whether or not arrests were detailed), ‘police’ was coded as 3: tolerance since the information suggested a police presence but did not document violence or arrests, and ‘police: arrests’ and ‘police: beaten’ were coded as 4: repression. Due to the vagueness of this data and the importance of government response in my research, I decided to also collect my own data from the RFA and available sources for labour protests, so as to be able to code more effectively for government response. Moreover, since my datasets for land and environmental protests were far smaller, I needed to have a labour database that was more similar in size to the other types so as to be able to run a

meaningful regression. The data I collected from the CLB will therefore be used only to compare and support the results of the smaller labour database which I built myself. In the time selected, the sources I referred to reported 36 environmental protests, 25 labour protests, and 26 land protests that had over 500 participants (those under 500 participants were excluded as this research focuses on ‘forceful’ protests which, as suggested by

previous research, are 500 participants and above).

Variables

Dependent Variable

The dependent variable in my research is the government response, which will be analysed in relation to the ‘forcefulness’ of the protest in order to measure social instability. As stated, it is coded as 1: concession, 2: concession with discipline, 3: tolerance and containment, and 4: repression. This ranked scale of response allows me to analyse increases or decreases in relation to the size of the protest. As it happens, all cases of protest collected in my dataset were responded to by local governments, and not the central government (which already suggests something about the relationship between social instability and regime), thus all responses are recorded from the local level. Concessions are usually in the form of compensation in accordance with the protestor’s demands – often in land protests, that means the delivery of a sufficient sum of money to compensate for a forced eviction or the loss of farmland, for instance. In environmental cases, a concession can be the decision of officials (usually at the local level) to halt plans to build a new chemical plant, or compensation for polluted crops. In labour protests, it may be the granting of higher wages, or the payment of arrears owed. Tolerance is a somewhat ‘neutral’ response from officials where there is usually a police presence, but the protest is

(18)

17

not actively repressed, nor demands met. Repression is the forceful shutdown of a protest event, usually by police forces deployed by the local government. Repression often involves violence, injuries and arrests. Figure 1 below provides details of the frequencies of

government response in the data collected. One can see that in the majority of cases – 59 - the government is still repressing protests, with far fewer being tolerated or met with concessions. This is because protest is ultimately still illegal in China, and as Gobel & Ong state, ‘unfortunately, repression has proven to be more expedient than intervention or co-operation’ (Gobel & Ong 2012, p16)

(19)

18

Independent Variables

The independent variables are the aspects of a protest which measure the ‘forcefulness’ of a protest under the assumption that the more forceful the protest, the more likely a threat is posed to social stability.

V1: Size: the key indicator of the forcefulness of the protest.

V2: Violence: the existence of violence or no violence is coded as a dummy variable. Violence replaces ‘casualties’ in Cai’s framework (Cai 2010).

V3: Dummy variable indicating land protest, or not. V4: Dummy variable indicating labour protest, or not.

(Environment will be held as the reference variable for the purpose of the regression).

(20)

19

Figure 3: Histogram showing size frequencies and mean size for labour protests.

(21)

20

Although the range in protest size across the three types is large, a high percentage of protests in each type had under 2000 participants: 74% of land protests, 71 % of environment protests, and 66% of labour protests had fewer than 2000 participants. Protests with over 2000 participants are therefore far less common according to the data collected. Regarding violence, in 65 cases, violence is reported, in 22 cases, it is not reported. This relates to the fact that the majority of cases are repressed.

Regression

In order to test the dependent variable of government response with the independent variables of size, violence and the variables which distinguish each type of protest, I have chosen to run an ordinary least squares regression . Although the dependent variable is ordinal, a linear regression provides a straightforward mechanism of analysis. The variables of size, land protest and labour protest will be run as independent variables with

environment as the reference group, in order to see in which protest type size has the largest effect on government response.

Statistical Analysis

The following results provide the outcome of running an ordinary least squares regression analysis: Constant 3.454*** (.136) Size -2.216E-5* (.000) Land .424* (.187) Labour .240 (.189)

(22)

21

Figure 5: This table presents the regression of IVs size, land, and labour on the DV government response. The unstandardized coefficients and the standard errors are presented.

*Significant at the 0.1 level ***Significant at the 0.01 level

Since the coefficient for the size variable is negative, the results show that as size increases across all three types of protest, the government response decreases. This finding is

significant at the .10 level and therefore confirms H1: As the size of the protest increases, the government response will decrease (from repression towards concession). Regarding land protests, the results illustrate that compared with the reference category –

environmental protests – the government response is higher (it increases). This finding is also significant at the .10 level and means that H2: If the protest type is land related, the government response will show a larger decrease than the other types of protest, can be rejected, since the results indicate the opposite. The coefficient for labour protest is not significant, and therefore, H3: If the protest type is labour related, the government response will show a larger decrease than the other types of protest can also be rejected. The results indicate that the government response if the protest is land related increases more than for the reference category, environment, and that there is an insignificant relationship between labour protests and government response. If the reference category is swapped to land, the following results ensue:

Constant 3.728 *** (.141) Size -2.307E-5* (.000) Environment -.321* (.189) Labour .119 (.204)

Figure 6: This table presents the regression of IVs size, environment, and labour on the DV government response. The unstandardized coefficients and the standard errors are presented.

(23)

22

When land is changed to the reference category, the regression shows that when a protest is environmental, the government response decreases by .321 – a significant finding. Labour protests remain insignificant. Analysis of the two regressions allow for H4: If the protest type is environmental, the government response will show a larger decrease than the other types of protest, to be accepted. The results of my data collection and analysis therefore show firstly that as the size of a protest increases, the government response decreases, and secondly that environmental protests decrease the government response compared with land and labour protests.

Figure 7: Descriptive statistics showing government response tallies for environmental, labour and land protests.

The above tables are worth interjecting at this point to support the regression outcome – they illustrate the more even spread of government response across environmental protests compared with land protests, and in fact, the government was not reported to have made a single concession in the land cases collected. The fact that more mass environmental

protests were reported than the other types of protest in the time-frame selected also suggests that environmental concerns may be posing the biggest threat to social stability.

If violence is added to the regression with environment as the reference category, it is clear to see from the strength of the coefficient that there is a very strong relationship between violence and increasing government response, and this is significant at the 0.01 level as illustrated by the figure below:

(24)

23 Constant 2.917*** (.176) Size -2.065E-5* Land .379* (.184) Labour .262 (.177) Violence .619*** (.172)

Figure 8: This table presents the regression of IVs size, land, labour and violence on the DV government response. The unstandardized coefficients and the standard errors are presented.

If there is violence in a protest, it suggests that casualties (in the form of injured people) are likelihood. In the framework laid out by Yongshun Cai, casualties are a part of the

‘forcefulness’ of a protest along with size, and as discussed, his analysis of the 266 protest events collected finds that the more forceful the protest, the more likely the government is to make concessions (Cai 2010). What my results show however, is that whilst an increase in size decreases the government response, the presence of violence (an indicator of

casualties) increases government response. This is most likely the case because violence is either a direct effect of a repressive government response, or because violence initiated by protestors provokes a more repressive response. In my database, I coded for ‘violence initiator’ – this was not a part of my measurement of a ‘forceful’ protest, and therefore was not mentioned as an original independent variable, however ‘violence initiator’ can help to predict how much a part of ‘government response’ the outbreak of violence is – if it is started by the police, i.e. through use of tear-gas to disperse crowds, then it is likely a part of repression. I coded violence initiator as follows: -1: missing data, 0: unclear 1: protestor initiated 2: police initiated.

(25)

24 Constant 2.952*** (160) Violence .345 (.269) Violence Initiator .211* (.124)

Figure 9: This table presents the regression of IVs violence, and violence initiator on the DV government response. The unstandardized coefficients and the standard errors are presented.

The above ordinary least squares regression shows that as the violence initiator increases (from unclear through to police initiated), the level of government response increases. This finding is significant and confirms that violence is often a part of government response and cannot be treated independently. The same is most likely true for ‘casualties’. This is a difference between Cai’s framework and my own, and illustrates the value of applying a more quantitative approach to analysing protests and government response, albeit to study what makes a protest successful or in this case to try to discover which type of protest may be posing the biggest threat to social stability.

Discussion of Results

According to the data and analysis, built around Yongshun Cai’s notion of a ‘forceful’ protest and the government’s response, it appears that environmental protests may pose the largest threat to social stability at present. The mean environmental protest size was much larger than for land and labour, and in ten of the protests, participants exceeded ten

thousand. A discussion regarding some of the large-scale environmental protests included in the database which prompted a less harsh government response - or indeed a concession - will help to shed light on why environmental issues are currently proving most disruptive to social stability when reports on mass incidents are analysed. The biggest protest recorded across all protest types, was one of 50,000 people in the Jiangsu city of Qidong which occurred in July of 2012. In this instance, locals wanted to prevent of a waste pipeline which they feared would pollute the coastline. According to eye witness reports, the protest was

(26)

25

violent; “There are people everywhere, on walls, cars, rooftops, in streets (…) the air is filled with the smell of alcohol, and there are sounds of breaking glass” (Wade 2017). The violence in this case stemmed from the protesters, and another witness stated that police reacted with considerable constraint on the mayor’s orders: ‘the mayor didn’t call for tear gas or tanks to crack down the protesters. It’s an improvement. This mayor deserves some applause. Even better, he immediately announced that the project was permanently

cancelled’ (Wade 2017). The outcome of this very large-scale protest, which posed a severe threat to social stability, prompted a complete concession from the local government, which called for police restraint and ultimately cancelled the pipeline project. In this circumstance, it seems that the sheer size and force of the protest drove the government to make a concession. According to the report by China Digital Times, a resident of Qidong city said in spite of the concession that ‘we won’t believe the government until they make clear what measures they will take to stop pollution from harming our health’ (Wade 2017),

highlighting a suspicious attitude towards local level government.

Whilst the latter protest event was not only large-scale but violent from the side of the protesters, most of the environmental protests recorded which exceeded 10, 000 participants were reportedly peaceful. In July of 2013, over 10,000 people marched

peacefully in Guangzhou over plans to build a waste plant. In this instance, the government did not make a concession but tolerated the protest, and when 300 of the protesters entered a government building to express their views, ‘local officials told them that

environmental experts were visiting the area on Friday, and would explain the projects more carefully to local people’ (RFA 2013). This protest was a result of another large-scale

environmental protest in Jiangmen which resulted in a concession from the local

government, who said they would cancel plans to build a nuclear fuel processing plant. This illustrates that a government act of concession in one case can act as a catalyst for further protest and demands. A local resident stated "If the government doesn't cancel the project, there will be more protests (…) and I think even more people will go" (RFA 2013), suggesting that the reluctance of local officials to make a concession can manifest in even more

(27)

26

Another very large and peaceful protest was reported in June of 2015 in Shanghai, when 40, 000 civilians protested a planned PX plant. No clashes were observed and as a result of the protest, the municipal government ordered the Shanghai industrial parks committee to ‘terminate its current strategy and to edit the environmental impact assessment’ which can be viewed as some form of concession (RFA 2015). Protester’s attitudes towards the

government was again sceptical, however, and the consensus was that the government’s attempt to quell popular anger had largely failed: ‘the government keeps saying that there is no PX project, but the general public are no fools’(RFA 2015). Further still, in 2016, 10,000 people marched in Shandong against a planned petrochemical plant, which resulted in a statement from the authorities which claimed to have halted the plans. A witness stated that the protests had been rational and non-violent, but that ‘there is no trust whatsoever in the government right now, because we think they are trying to fob us off’ (RFA 2016). In this case, authorities branded the uprising as ‘illegal’ and ‘disruptive of public order’ (RFA 2016).

The above four protests were some of the largest in the environment dataset and the

government tended to show constraint, and was tolerant or made concessions in all cases. A common theme throughout the protests seems to be a collective distrust towards the local government, which links back to the survey analysis by Li which shows a pattern of

hierarchical government trust whereby citizens tend to have ‘stronger confidence in the central government’s policy intent and weaker confidence in its capability to monitor and discipline local deputies’ (Li 2016, p115). The case studies illustrate how a mass incident of very large proportions can put enough pressure on the local government in order to achieve demands, and indeed the continued mistrust of local governments often sparks a chain of protests which can increase in size and raise the salience of the issue.

One point which is interesting to consider when dealing with environmental protests however, is that studies have found that the relationship between pollution and protest is not linear, and that serious pollution does not always motivate protest, whilst

environmental protests may occur without any real pollutions (Deng & Yang 2013, p322).This can be a result of framing processes, and on the flipside; ‘in environmental contention, protesters aggrieved by pollution may piggyback on issues unrelated to pollution, thereby allowing more opportunities to maximize the vulnerability of their

(28)

27

opponents: protests aiming to redress environmental grievances do not necessarily proceed with environmental claims’ (Deng & Yang 2013, p322). In 2005 for instance, villagers in Huaxi were able to force 11 polluting factories off their land by framing their claims mostly in terms of violations of land related regulations, linking protests to issues of illegal land requisition. This approach was taken because the villagers understood that it would be too difficult to prove the relationship between health defects and pollution, thus issue linkage secured the result that they desired (Deng & Yang 2013, p328). This interjection on framing is relevant to illustrate the overlap which can sometimes occur between two types of protest – environmental and land protests are often interlinked.

When analysed with a linear regression, the data collected illustrates that if a protest is land related, the government response is increases, and indeed from the descriptive statistics, we can see that the government response either repressed or tolerated the protests in all cases, and did not make a concession. This result is quite interesting given that China’s Twelfth and Thirteenth Five Year Plans have placed an emphasis on ensuring legal land use rights of farmers, on promoting the orderly transition of land operation rights and on intensifying anti-corruption levels. Whilst there were fewer reported mass land-related incidences, the government still appears to use repression as the main tool of response. On a basic level of analysis, taking the framework of this research into account, this could have to do with the size of the protests: the mean size of land protests in the database is much smaller than that of the other types at 2066, whilst for labour protests it is 4256.32 and for environmental protests it is 5802.63. Since H1: As the size of the protest increases, the government response will decrease (from repression towards concession) was confirmed in the regression analysis with a negative, significant size coefficient against the dependent variable, one can thus deduce that as the size of the protest decreases, the government response will increase. This provides a logical explanation as to why the government is still more repressive towards land protests – they are smaller.

To understand the wider implications of this result, empirical evidence can be observed which explains why local officials are reluctant or unable to make concessions in land-related disputes. In Guangdong, the village of Wukan gained attention in 2011 for a series of protests against anti-corruption and land acquisition, which were sparked when ‘village

(29)

28

officials sold land to property developers without properly compensating the villagers’ (Knight 2012, p15). These protest commenced a long standing dispute between local residents and authorities, which escalated again in 2016 when village leader Lin Zulian was sentenced to jail; the clashes were suppressed (South China Morning Post 2017). As Knight states, ‘underlying the Wukan incident was the fiscal pressure on local governments’, an issue which deepened 50% of the tax on local enterprise profits was appropriated by the central government in 2002, and agricultural tax was abolished in 2006, which meant that self-serving officials ‘increasingly raised funds by selling land at market prices while normally compensating the land-holding community with only a small fraction of the proceeds’ (Kung et al. 2012). Persistent tensions like these between the aims of central and local

governments are likely another reason as to why land protests are often suppressed. Moreover, land protests usually occur after evictions, demolitions and land sales have already taken place, which therefore makes it more complex for local governments to meet protester demands – as Cai’s research shows, when demands are harder to meet,

repression is a more likely outcome (Cai 2010, p45).

In 2013, research conducted by Bo Zhao on land expropriation, protest and impunity in rural China argued that conflicts over land expropriation pose a significant threat to the country’s social stability and the sustainability of its economic development, and further that such conflicts are inevitable under China’s current political and legal system (Zhao 2013, p97). Zhao points to the lack of fundamental reform to China’s party politics as a barrier in preventing conflicts, and states the following:

‘Though Beijing wants to fight for control over local officials on land issues, it cannot achieve the aim because land law violations at the local level are hardly punished. The main reasons for this are strong local protectionism and the absence of an independent judiciary’ (Zhao 2013, p100).

The consequence of this unsatisfactory political system, therefore, is that farmers are forced to stage public protests and riots in order to gain attention from the media and call for their grievances to be addressed (Zhao 2013, p101).

(30)

29

What has further become more clear in recent years under the Thirteenth Five Year Plan, is that Xi Jinping’s anti-corruption campaign may be backfiring: the campaign’s stated aim is to hold all levels of Chinese officials accountable for abuses of power, however Fu Hualing, a Law professor at the University of Hong Kong made the following statement for the Financial Times:

“I don’t see any clear political will to seriously punish corrupt officials at the grassroots level (…) maybe they understand that is probably very destructive if China does that in every county, every district…the whole country would probably be in chaos.” (Financial Times 2016).

Moreover, since the central government has taken responsibility for anti-corruption, new studies have found that the higher the number of reported graft cases in a prefecture, the more people in the area perceive Beijing as being more corrupt than their local government (Financial Times 2017). This again is reflected in Li’s research which finds that hierarchical trust holders are not as trusting in the central government as it may appear from surveys (Li 2016).

Whilst a clear relationship between the ‘forcefulness’ of labour protests and an increase or decrease of government response was not established by this research, this certainly does not imply that these protests are not affecting social stability. Indeed, workers do not have the formal right to organise , to bargain or to strike, however ‘the emerging shortage of labour and the rise of internet and mobile communication have made workers more

demanding’ (Knight 2010, p 22). Another reason which affects the labour result is that many labour strikes are smaller and more internal – of 8041 instances of labour protest recorded by the Chinese Labour Bulletin between the beginning of 2011 and the end of 2016, only 409 were over 500 participants, which constitutes a ‘mass incident’ (Chinese Labour Bulletin 2017). Labour protests involve not only workers and local officials, but also the employers and management related to each protest, and therefore the response often comes from multiple sources which could complicate the relationship between labour protests and government response. Labour grievances remain one of the key reasons for discontent, and as Headley & Tanigawa-Lau state, in 2015 ‘around 68% of protests were wage-related’ (Headley & Tanigawa-Lau 2016). Whilst labour disputes may not have the same level of ‘forcefulness’ (as utilized in this research) which tends to cause government concessions in

(31)

30

environmental protests, their frequency puts pressure the central government to solve wage-related issues and poor working conditions in an era where economic growth is slowing.

Social Stability and Regime Legitimacy

A report by the Europe China Research & Advice Network in 2012 makes a comparison between social unrest in China and the Arab Spring, stating the following:

‘By contrast to the Arab Spring, social unrest in China is not a sign of an impending regime change or even of a major legitimacy crisis. It should be seen instead as a form of

participation. The increase of social unrest does not necessarily stem from a falling quality of life of the citizenry. Rather, it is probably the result of the inadequacy of formal channels of communication between the citizenry and authorities’ (Gobel & Ong 2012, p9).

Knight’s research echoes the above statement regarding the inadequacy of formal channels of communication:

‘Citizens lack both independent political institutions to participate in the decisions that affect their lives and independent legal institutions to protect their rights. As a result, people’s grievances have been channelled into mass protests and petitions’ (Knight 2012, p14).

According to the research conductedby the Europe China Research & Advice Network, social unrest does not appear to have a negative impact on the legitimacy of the Chinese Communist Party’s one party rule, and despite rising unrest, the death toll of such activities is not significantly increasing (Gobel & Ong 2012, p9). In spite of this conclusion, the report still finds that social unrest ‘bears risks that could severely disrupt China’s social stability’ (Gobel & Ong 2012, p9). Social stability is thus separated from political stability or regime legitimacy and has differing repercussions for society. The report further postulates that the increase in social unrest is not the result of an economic downturn: ‘rising unrest is not directly correlated with economic growth rates. It arises from immediate reasons such as land disputes, environmental degradation, labour conflicts and ethnic strife’ (Gobel & Ong 2012, p11). Whilst rising unrest may not be directly correlated with economic growth rates,

(32)

31

it seems somewhat simplistic to separate ‘immediate reasons’ like land, labour and environmental issues from economic growth, as they are heavily intertwined. Social instability can be seen as an adverse shock to economic growth, as issues related to discontent are often an effect of the drive to industrialise and maintain the growth rate: a higher growth rate entails a growing industry which can worsen pollution levels, cause land to be seized for development projects, and result in poor labour standards as wages are lowered to keep profits high and working hours are extended.

As economic reform took hold in China in the 1970s, rapid growth was deemed paramount to securing political legitimacy and avoiding social instability, however the transformation of China’s economy and society as a result of such growth has manifested new threats (Knight 2010, p4). The fact that China’s semi-marketised economy has tended cultivate corruption – particularly at local levels – and that corruption is in turn a serious source of discontent is illustrative of the causal relationship between the economy and social unrest, especially as corruption often manifests in excessive taxation, the confiscation of land or property without adequate compensation and the neglect of pollution (Knight 2012, p15). Research on the topic of social unrest in China generally seems to concur that whilst the outbreak of protest is not currently affecting economic growth, it certainly could if the discontent becomes prolonged and widespread: as Knight states:

‘localised and uncoordinated cases of social instability do not pose a threat to expectations of continued rapid economic growth, however should they become more widespread and coordinated, investor confidence might be shaken and aggregate investment might fall. That might set in motion a cumulative train of events. For instance, the external imbalances of the Chinese economy make it prone to asset bubbles through their effect on liquidity’ (Knight 2012, p21).

As a result of the unstable relationship between social instability and regime legitimacy, with social unrests deemed as ‘the expression of intense social discontent and barometers for regime stability, the study of social protest in contemporary China has become a rapid-growing industry among social scientists’ (Tong & Lei 2010, p488).

(33)

32

Despite the threat that social instability could potentially pose to economic growth, protests which are driven by economic grievances in some sense – which is often the case for land and labour protests – can actually enhance the legitimacy of the government by requiring them to ‘enforce justice’ and satisfy an economic demand which has been denied at the local level (Tong & Lei 2010, p499). If the state has the financial resources to respond to a demand effectively, it further consolidates its legitimacy. Again however, if the central government takes responsibility for corruption occurring at the local level, as can be seen with Xi Jinping’s anti-corruption campaign, it risks the chance of weakening legitimacy as blame is transferred from the local level to the state level.

What can be taken from the proceeding discussion, then, is that whilst social instability does not indicate the collapse of the one-party regime, it does have the potential to slow down economic growth and decrease the life satisfaction of citizens, which could ultimately result in an erosion of regime legitimacy if the central government does not respond effectively to issues which cause mounting social discontent. This research paper has found that from the data collected, environmental issues are currently causing the outbreak of the largest mass incidents, which are forcing the government to make concessions. This is clearly a result of deepening environmental degradation and soaring pollution levels, which manifests in social instability. ‘Clean production’ is listed as an aim of the Thirteenth Five Year Plan, and

certainly it should be a priority if the Chinese government – both on a national and a local level – wishes to avoid an escalation of social instability (Xinhuanet 2017).

Research Limitations

Whilst the available resources for data collection were closely scrutinized, and the results offer an analysis of all mass incidents reported in the chosen sources (Radio Free Asia, China Digital Times, South China Morning Post, Lu Yuyu’s blog, Hong Kong Free press, The

Guardian, Reuters and the Chinese Labour Bulletin) between 2011 and 2017, the research suffers from the data restrictions which arise in trying to obtain the relevant information on protest events in an authoritarian state, given that reporting is heavily repressed and can lead to punishment in China. As a non-Chinese speaker, it was necessary to rely on the reporting of secondary English-Chinese resources, and whilst a range of outlets were consulted, the validity of each report cannot be confirmed, and for the purpose of the

(34)

33

research, the information was taken at face value. Oftentimes, the reports provided the information necessary to build the databases for the differing types of protest (where they did not, they were not used), and all of the events collected detailed the size of the protest, and the government response. The size of the protest can only be estimated; as usually, this information comes from eyewitnesses or reporting agencies on the ground – there is no easy way to know the exact number of people involved in a mass incident, therefore

eyewitness accounts are often the best indicator of size. Government response is also often understood through eye-witness accounts or through reports from police or officials.

Moreover, it is important to keep in mind the information bias of each report: although the main source – Radio Free Asia – is an independent broadcasting corporation, it is

headquartered in Washington, and whilst many of its journalists are Chinese citizens, there will always be an information bias: the framing of the report, the time and place in which the details were gathered, eyewitnesses whom divulged information on the event, and the style of each individual reporter are all subjective features which ultimately build the narrative on which the data collection is based. Although the analysis of this paper is quantitative, the data stems from qualitative observations of events. As such, this research cannot provide an unbiased, objective, and complete scrutiny of protest in China and it is based on what has been reported. From collecting as many protest events as possible for the three types of discontent however, it has been possible to observe trends via means of statistical analysis which are indicative of which type of protest poses the largest threat to social stability at this time.

Because of the second-hand nature of the data collection, it was not possible to build more accurate datasets based on the information available. This means, for instance, that it was not possible to record which precise government body intervened in the protest to either offer concession or deploy police forces. If this were feasible, it would allow for a more stringent government response scale which detailed a hierarchy of response within local level government response. For a scholar with more means and resources, this would be advisable. Moreover, due to the variance of information within reports, missing data was an issue and prevented accurate use of variables such as number of injured, and number of detained, which were originally prepared to be a part of the regressions. The data collection

(35)

34

would also be far less limited if it was larger in size, and therefore access to a greater number of events across all three protest types would allow for a stronger analysis with a larger explanatory power.

Conclusion

The aim of this research paper has been to determine which of the most commonly contested issues in China – taking into consideration land, labour and environmental grievances – is most threatening to social stability, and potentially to regime legitimacy. Based on a framework introduced by Yongshun Cai, which originally aimed to decipher what makes collective resistance successful, the method employed to answer the posed research question entailed gathering data on protests which were deemed as ‘forceful’ mass

incidents. A consensus across existing research and police reports pointed to an action with 500 participants or more as being perceived as a large-scale ‘forceful’ action, thus protests with a participant figure below 500 were excluded from this research. The government response was recorded in line with Cai’s scale of response on a scale from 1-4: (1) concession, (2) concession with discipline, (3) tolerance, and (4) repression.

According to Cai’s research findings, the more forceful a protest; the more threatening it becomes to social stability; and the more likely the government is to make concessions. The results of my data collection collaborate Cai’s findings, and confirmed H1 As the size of the protest increases, the government response will decrease (from repression towards

concession). Through running an ordinary least squares regression, I was also able to deduce that environmental protests triggered the most concessions from the government,

confirming H4: If the protest type is environmental, the government response will show a larger decrease than the other types of protest. From the data I collected therefore,

environmental protests are most threatening to social stability, as they are larger in size and prompt a bigger decrease in government response than the other types of protest do. The research further found that the presence of violence in a protest strongly increased the government response (since violence if often a feature of repression).

The analysis of research on social stability and regime legitimacy, suggests that at this present time, social instability is not posing a direct threat to regime legitimacy, however if

(36)

35

social unrest were to become more widespread and coordinated, it could start to have a great effect on economic growth, and potentially begin to erode the legitimacy of the Chinese Communist Party. With this in mind, further research into which type of protest is most threatening to social stability is of interest and value, especially to a scholar with greater resources. This research has illustrated the potential value of applying a quantitative method to analysing the relationship between types of protest in China and social stability, and with a larger dataset, the results of this kind of analysis could be very useful in