1. Introduction
Autocratic China has experienced rapid economic development and gradual institutional change, where the latter can be attributed to economic decentralization initiated by the central government, adaptive policy experiments conducted by local governments, and informal institutional innovation from the bottom up (Ang, Reference Ang2016; Coase and Wang, Reference Coase and Wang2016; Nee and Opper, Reference Nee and Opper2012; Tsai, Reference Tsai2006). However, as the theory of the predatory state illustrates, China's economic growth has been accompanied by government confiscation of private property because the political constraints of public participation and power sharing are lacking and because citizens have failed to prevent state capture (Barzel, Reference Barzel2002; Boettke and Candela, Reference Boettke and Candela2020; Murtazashvili and Murtazashvili, Reference Murtazashvili and Murtazashvili2020). As Vahabi (Reference Vahabi2016) claims, because land is immobile and appropriable, it is a captive (not fugitive) asset and vulnerable to state expropriation.
Eminent domain often benefits government, and its legitimacy is often questioned (Paul, Reference Paul2017). Even in democratic regimes (such as the United States) in which property rights are well protected and formal rules stipulate that land taking should be in the public interest and landowners should be justly compensated, local governments sometimes abuse their eminent domain power to seek rents, often in the name of economic development, and offer less-than-full compensation because their budgets are limited (Benson and Brown, Reference Benson, Brown and Bruce2010; Schäfer and Singh, Reference Schäfer and Singh2018). But the injustice and unequal application of eminent domain is much more serious in China, where the bargaining power of local governments and residents is highly imbalanced. Local governments have monopolistic power over land use and exchange, and they take land – without paying sufficient compensation – to pocket land-lease fees (the major source of their extrabudgetary revenues) (Fan, Qiu, and Sun, Reference Fan, Qiu and Sun2020; Hsing, Reference Hsing2010). Residents often lack legal means to bargain with the authorities about expropriation proposals (Cai et al., Reference Cai, Murtazashvili and Murtazashvili2020a, Reference Cai, Murtazashvili, Murtazashvili and Wang2020b; Whiting, Reference Whiting2011).
However, this does not mean they have no opportunity to defend their land. Residents in many places and times have resisted predatory land taking through extralegal bargaining. Studies have examined how resisters overcome collective-action problems and collectively resist through petition, violence, and networked protest (Chen, Reference Chen2020; Lei, Reference Lei2019; Sargeson, Reference Sargeson2013; Zhang and Zhao, Reference Zhang and Zhao2014). Collective resistance triggers social unrest and attracts the attention of the public and higher-level governments. It puts the question of the legitimacy of local governments’ eminent domain power under the spotlight and thereby tends to force the governments to compromise and the central government to respond to the unfair formal rules (Cai, Reference Cai2010; Heurlin, Reference Heurlin2016; O'Brien and Li, Reference O'Brien and Li2006). Nonetheless, residents must overcome the free-rider problem and find strong leadership, both of which are especially difficult when residents vary in their vulnerability to predation (Tullock, Reference Tullock1971).
This paper contributes to our understanding of resistance and institutional change in autocracies by shedding light on individuals’ resistance strategies. Individual resisters, or resistant entrepreneurs, are alert to the circumstances of time and place and can develop novel ideas and moral appeals to garner public sympathy and defend against government predation (Chen, Reference Chen2012). Specifically, confronting eminent domain, China's resistant entrepreneurs defend their land by exploiting local authorities’ vulnerability to social unrest and public criticism and by portraying as illegitimate the land takings, thereby credibly signaling their capacity to lead the authorities to be punished by their higher-level counterparts and forcing the authorities to make concessions (Cai, Reference Cai2014; Heurlin, Reference Heurlin2016). By showing how residents ex ante resist land taking by holding out, using violence, harming themselves, or gaining media exposure, I explore the conditions for effectively pressuring local authorities. Such resistance ex ante can be seen as an effort at synthesizing resistance signaling and extralegal bargaining (Chen, Reference Chen2009; Wilson, Reference Wilson1961).
The paper investigates the role of moral appeals in extralegal bargaining with local authorities in autocracies. It contributes to the literature on legitimate resistance (Moore, Reference Moore1993; Scott, Reference Scott1977; Tufekci, Reference Tufekci2017). As defined by Levi et al. (Reference Levi, Sacks and Tyler2009) and Beetham (Reference Beetham2013), legitimacy derives from citizens’ shared beliefs about the normative appropriateness of governmental power and law enforcement. Lack of legitimacy increases the transaction costs of governing by increasing a government's reliance on coercion (Tyler, Reference Tyler2006). I find that if resisters can show how local governments’ eminent domain is predatory and worsens their living conditions, thereby making it illegitimate, they can garner public sympathy and support. Rather than opposing the central government, they justify their actions to it by framing their appeals for protection in terms of the central government's ideology and statutes. This in turn allows them to threaten the local authorities: if local governments rebuff or suppress their appeals, the resisters publicize the actions of those governments, which are highly likely to be punished by a higher-level government (O'Brien and Li, Reference O'Brien and Li2006). As Rithmire (Reference Rithmire2015) explains, in China, where the process of reallocating resources is uncertain, property rights enforcement is accompanied by political bargaining, ideological argumentation, and moral suasion.
However, not all residents can afford to credibly signal their willingness to resist and make these moral appeals. Only resistant entrepreneurs, who can craft ideas for garnering sympathy and can endure the resulting hardship, can bargain extralegally. To show local authorities that they are determined to defend their property and able to attract public attention and disrupt the social order, resistant entrepreneurs have to employ strategies that are difficult for others to mimic. This finding adds to the extensive work on signaling (Gambetta, Reference Gambetta, Hedström, Bearman and Bearman2009; Gintis et al., Reference Gintis, Smith and Bowles2011; Leeson, Reference Leeson2008; Lohmann, Reference Lohmann1993; Schelling, Reference Schelling1960). It also contributes to the new institutional economics tradition by enriching our understanding of behaviors – such as prisoners’ self-harm, tribal communities’ human sacrifice, and dueling – that are apparently at odds with the maxim ‘More is preferred to less’ (Allen, Reference Allen2011; Gambetta, Reference Gambetta, Hedström, Bearman and Bearman2009; Leeson, Reference Leeson2014: 139). It substantiates Leeson's (Reference Leeson2020: 145) claim that ‘every observed institution is efficient’, meaning that individuals attempt to maximize their profit given their constraints when it comes to enforcement of property rights. Notably, the claim does not imply that individuals always manage to profit-maximize; they might make mistakes, and they vary in their alertness to profit opportunities and ability to develop strategies for seizing them.
This paper also enriches our empirical understanding of the political and moral economy of land disputes and institutional change in China. To understand how individuals make moral appeals to deter land takings, I draw on a broad range of evidence concerning ‘nail-house strategies’ – supplementary strategies (such as violence, self-burning, and media exposure) adopted by residents who refuse to vacate their land even after it has been targeted for eminent domain. These residents are as stubborn as nails and hence are called ‘nail-residents’. They garner public sympathy and embarrass local authorities; this makes them more likely to maintain or improve on the status quo than those who do not protest unfair land expropriation. By comparing cases by type of resistance strategy and by conducting process tracing, I investigate how nail-residents can send credible resistance signals to threaten local governments and why some types of strategies are more effective than others. The goal is to understand how resisters make moral appeals and why it can be difficult. I seek to understand the causal mechanism, rather than the causal effect (Skarbek, Reference Skarbek2020).
Examining these cases also allows me to explore the spillover effect of resistance signals on institutional change in autocracies. Instead of demanding changes to formal rules, resisters aim at deterring eminent domain or getting higher compensation. Nonetheless, when successful resistance signals are copied by others, they weaken the legitimacy of governments’ expropriation, increase the governments’ enforcement costs, and provide central authorities constructive criticism (Sargeson, Reference Sargeson2013; Tsai, Reference Tsai2015). Central and local governments respond by revising laws or experimenting with new policies. This represents a typical path of institutional change in autocratic China, one overlooked in the literature on China's development miracle (Ang, Reference Ang2016; Coase and Wang, Reference Coase and Wang2016; Nee and Opper, Reference Nee and Opper2012; Tsai, Reference Tsai2006).
Section 2 presents a theoretical framework to analyze how resisters under autocracy are able to extralegally bargain with local authorities by making moral appeals and seeking public sympathy and how they make the appeals by sending resistance signals. Section 3 presents empirical evidence on nail-house strategies. It illustrates why such strategies are difficult to mimic and can serve as credible resistance signals that embarrass local authorities. Section 4 discusses the spillover effect of resistance signals on institutions. Section 5 concludes.
2. Resistance signals: making moral appeals in land bargaining under autocracy
In autocracies, local governments’ eminent domain power can be arbitrary, and residents are often in weak legal positions. Consider China, where land property rights are deliberately ambiguous (Ho, Reference Ho2005). Legally, local governments can expropriate land in the public interest. But the scope of the public interest is ambiguous and subject to political discretion because such political constraints as elections and public participation are absent (Boettke and Candela, Reference Boettke and Candela2020; Cai et al., Reference Cai, Murtazashvili and Murtazashvili2020; Heurlin, Reference Heurlin2016). Under the tax-sharing system in place since 1994, local governments account for more than 65% of the nation's total fiscal expenditure, while their on-budget revenue is only approximately 45% of the nation's total fiscal revenue (Qun et al., Reference Qun, Li and Yan2015). To acquire the land, pocket land-lease fees as off-budget revenues (accounting for fully 36% of local-government revenue from 2007 to 2016), and relieve financial pressure, local governments are keen to expropriate the land on which rural and urban housing is built and to pay residents too little for it. They often justify the land taking in the name of urban (re)development (Fan et al., Reference Fan, Qiu and Sun2020; Hsing, Reference Hsing2010).
The compensation criteria for expropriated land and houses are formulated by local governments without residents’ input (Whiting, Reference Whiting2011). Residents are typically unsatisfied with the compensation. A 17-province survey found that over 70% of households affected by land taking were ‘very unhappy’ with it (Zhu et al., Reference Zhu, Prosterman, Ye, Li, Riedinger and Ouyang2005). Though residents can appeal to courts, their claims are often rejected, not because formal rules protect local governments’ monopolistic power to take land but because courts lack judicial independence: the Chinese Communist Party dominates the legal-political system, so the courts are subordinate to political and administrative powers (Li, Reference Li2016; Mao and Qiao, Reference Mao and Qiao2021).
However, local governments’ monopolistic power over land taking does not imply that they are invulnerable. Residents can resist predatory local authorities for two reasons. First, because land property rights are ambiguous, the moral content and persuasive power of objections are significant for local governments to justify land taking (Rithmire, Reference Rithmire2015). When eminent domain significantly lowers residents’ quality of life or even threatens their subsistence, the public often believes that, rather than serving the public interest, it is predatory and unjust and that resistance is defensive. This belief was widespread in the rebellions in Southeast Asia in the early 20th century and in Europe in the 19th century and is common in networked protests today (Moore, Reference Moore1993; Scott, Reference Scott1977; Tufekci, Reference Tufekci2017).
Second, unlike the legitimacy of democratic regimes, which rests on elections, the legitimacy of autocracies, which lack preference-revelation mechanisms, often rests on mass mobilization and is indicated by social stability (Beetham, Reference Beetham, Nash and Scott2001; Li, Reference Li2018; Shue, Reference Shue, Gries and Rosen2004). If the public and the media observe resistance and are led to criticize the local government, citizens might come to believe that the local government is untrustworthy or incompetent, which can undermine the regime's legitimacy (Cai, Reference Cai2010; Chen et al., Reference Chen, Pan and Xu2016). Given that in a hierarchical bureaucratic system the appointment and promotion of officials in subordinate governments are determined by higher-level governments, local officials may be criticized by their superior-government counterparts or even removed from office if citizens resist and threaten social instability (Chen et al., Reference Chen, Pan and Xu2016; Whiting, Reference Whiting2006; Xu, Reference Xu2011). As Cai (Reference Cai2014: 11) points out, ‘When failing to discipline erring agents is believed to damage the regime's legitimacy, the state authority comes under pressure to mete out punishment’. Conspicuous resistance can detract from local governments’ achievements in other areas, such as spurring economic growth (Fu, Reference Fu2018).
In sum, local governments have incentives to take land without provoking conspicuous and rightful resistance. Note that local governments’ concessions are selective and conditional: resistance that fails to attract public attention or call into question the governments’ moral and political legitimacy may be suppressed (Cai, Reference Cai2010; Li, Reference Li2018). Therefore, in an autocratic system, residents, despite lacking effective formal channels to resolve land disputes, can strengthen their extralegal bargaining power by making moral appeals. More specifically, the success of residents’ resistance depends on their ability to publicize their unjust treatment to elicit public sympathy and criticism and thereby call into question the legitimacy of local governments. Residents force the hand of local governments by provoking (or credibly threatening to provoke) higher-level governments to impose discipline.
Resistance against land taking can be collective or individual, ex ante or ex post. Though residents might conduct collective protests (such as organized violence or marches) to deter eminent domain, such protests might fail because of a free-rider problem, especially when residents’ incentives and capacity to resist are heterogeneous (Chen, Reference Chen2020). Resistance ex post, particularly through petitioning, is widespread, but given that land taking is ongoing and visible, purely ex post resistance is less frequent (Heurlin, Reference Heurlin2016). Individual resistance ex ante has been a prevalent strategy adopted by residents, who enhance their extralegal bargaining power by sending costly resistance signals in order to elicit public sympathy. They signal their determination to fight against unfair land takings and display their capacity to question and publicize the immorality and illegitimacy of local authorities.
Not every resident can succeed in sending credible resistance signals to elicit sympathy and threaten local governments. In fact, only a few – the resistant entrepreneurs – can. They have two features. First, they are institutional entrepreneursFootnote 1 who can draw on their local knowledge about their institutional environment to recognize an opportunity and develop creative ideas to protect property rights (Anderson and Hill, Reference Anderson and Hill2004; Leeson and Boettke, Reference Leeson and Boettke2009). They seek extralegal solutions to protect land property rights from government predation, whether by extracting more compensation or preventing the act of eminent domain. To call local governments’ legitimacy into question and improve an unfavorable situation, they have to develop strategies to tell narratives about their grievance in a way that elicits public discussion and sympathy. Resistant entrepreneurship – that is, alertness to resistance opportunities and the capacity to craft resistance strategies – can be seen as a key endowment and endogenous feature of resisters that is not available to everyone. It is present in different types of resistance strategies, which I discuss below.
Second, resistant entrepreneurs can afford (or have access to) resistance signals, while other residents cannot (or do not). Besides relying on resistant entrepreneurs’ perception of them for them to function properly, resistance signals, whether violent or nonviolent, often are costly in the sense that they often require signal senders to waste resources, as they entail dedicating time, energy, and resources, suffering bodily or property damage, or risking arrest. It is similar to how extremist activists are more motivated to participate in costly political actions, such as protesting and lobbying, than apathetic moderates in democratic regimes (Lohmann, Reference Lohmann1993). The reasons why resistant entrepreneurs are more willing to bear such costs are objective or subjective. It may be that they have more valuable properties (e.g. houses in better locations or properties in commercial use) than those who do not resist (Hsing, Reference Hsing2010). Or it may be that they harbor a strong sense of fairness concerning the distributive process and outcome of a dispute (Gibson, Reference Gibson2008; Whiting, Reference Whiting2011). In some scenarios, some tactics – such as media exposure, which relies on residents’ ability to tell a narrative or their personal ties to journalists or public elites – are nearly costless for resistant entrepreneurs but unavailable to others (Gambetta, Reference Gambetta, Hedström, Bearman and Bearman2009).Footnote 2
Successful resistance signals have to meet some conditions regarding their intention, context, and timing. First, considering that local governments are very likely to ignore or even suppress uninfluential resistance but make concessions in response to conspicuous resistance, the resistance signals must be strong enough that resistant entrepreneurs signal a credible commitment to further resist the local government if the government rebuffs their appeals. Rather than physical power, what signifies resisters’ determination to fight against predatory land taking and their capacity to publicize their grievances and convince the public of the immorality of local governments is their resoluteness. An effective signal must show resoluteness to local governments. It credibly threatens that if governments rebuff their appeals, they will resist even more fiercely and publicize their stories even more, which can impose on the governments substantial risks and costs (Chen, Reference Chen2009; Fu, Reference Fu2018). Local officials not only have to dedicate administrative resources to restore social order, but might be seriously disciplined (e.g. expelled) by their superior-government counterparts.
Second, since the power of the resistance signals lies in the public's criticism and the potential punishment by higher-level governments, their purpose is to morally and politically pressure local governments. The signals should be observable by the public and higher-level governments; sensational resistance thus has more impact. Moreover, the resistance should be defensive, rightful, and self-limiting; otherwise, it may render resisters’ moral appeals ineffectual. Resisters who frame their appeals to the higher-level governments in terms of such governments’ ideologies and rules may be seen as engaging with the regime and are likely to be backed by it; but attacking the legitimacy of the regime might be regarded as threatening the regime and might invite suppression (Li, Reference Li2018; O'Brien and Li, Reference O'Brien and Li2006). Similarly, resisters who engage in excessive violence or make excessive demands for compensation risk losing the public's sympathy. The fact that effective resistance signals must meet certain conditions regarding their intention and context, in turn, explains why only a few residents extralegally bargain with local governments.Footnote 3
Third, given that the goal of the resistance signals is to threaten local governments and create extralegal bargaining opportunities, they must be preemptive; that is, they must be sent before the land is taken. By acting preemptively, resistant entrepreneurs attempt to reveal their resoluteness and sway local governments in a favorable direction (Fu, Reference Fu2018).
3. Nail-house case study
To understand how resistant entrepreneurs resist predatory land taking, I examine the nail-house phenomenon in China, in which residents (colloquially called nail-residents) created opportunities to bargain with local authorities by signaling their resoluteness. They held out after their properties were slated for eminent domain and deployed the tactics of violence, self-burning, or media exposure to embarrass local governments and thereby extract higher compensation or prevent the land takings. I collect as many nail-house cases from media reports and scholarly studies as possible. To ensure that the accounts are reliable, I carefully compare the details of the cases and exclude contradictory statements. I rely mostly on mainstream media with relatively high credibility such as Infzm (Nanfang Zhoumo), Nanfang Metropolis Daily, and Phoenix Weekly. The cases span just over a decade (2003–2013) and cover nine provinces (see Table 1). The nail-house phenomenon also occurred in many other provinces, but because the data are incomplete there, I cannot examine them. I try to identify what the causal mechanisms of specific signaling strategies in the land bargaining were, how they posed credible threats to the local governments, and why some of them were more effective than others.
Table 1. Nail-house cases
Note: Asterisks indicate that data are unavailable.
3.1 Holding out
The basic strategy the nail-residents executed is holding out: holding on like nails to the houses to be expropriated, regardless of how tough the living conditions were. Because the local governments generally disregarded the problem of nail-residents until they expropriated all other residents’ houses, nail-residents, whose houses are covered in sand, soil, and bricks, had to endure a harsh physical environment. Worse still, to drive nail-residents away, some local governments even cut off water and electricity in the name of constructing infrastructure.
Table 1 shows that nearly all the nail-residents clung to their houses for three to eight years. Hsing (Reference Hsing2010: 201) writes, ‘Through occupying the place with boundary markers and buildings’, they made explicit claims to their houses and land and established territorial control. Their holding out handicapped the eminent domain projects and signaled their determination and ability to bear harsh living conditions and to strive for more bargaining opportunities. As a neighbor of a nail-resident said, it is reasonable for nail-residents to get higher compensation because they sacrifice a lot. Portillo (Reference Portillo2018) finds that in negotiation games with relatively equal bargaining power, holding out tends to increase final transaction prices. But in the context of land expropriation in China, where bargaining power is unequal, it is often insufficient to force local governments to pay higher compensation. In other words, simply holding out is often not enough to convince the public of the injustice of eminent domain and to credibly threaten local governments, which accounts for why supplementary approaches have emerged to strengthen the effect of holding out.
3.2 Limited violence
Violence is one of the supplementary strategies to enhance the credibility of the resistance signals. To be an effective resistance signal, violence must be preemptive and tactically self-limiting. It often serves as a powerful weapon to allow nail-residents to keep their houses from being expropriated or to get higher compensation (see Table 1). The struggle of Baochen Chen and his family in Jingouzi Village, Pingdu County, Shangdong Province, is a typical case. The Pingdu government expropriated collective-owned land to pursue a village redevelopment project. According to the expropriation plan, Jingouzi villagers would be compensated for their removal with resettlement apartments as large as their lots. However, Chen and other villagers doubted the legality of the expropriation plan and were especially dissatisfied with the property rights to the land on which the resettlement apartments were built and with the apartments themselves (Li, Reference Li2015). More specifically, the housing consisted of neither commercial apartments (built on state-owned land and saleable) nor villager houses (built on collective-owned lots with permanent use rights). In other words, their property rights were incomplete, so the villagers did not have rights to sell or rebuild their apartments.
In 2013, to prevent their houses from being demolished and to express their anger and determination to resist, Chen and his family responded fiercely when a crane operator said that he would tear down their houses. They detained themselves and the operator in the crane, armed with a homemade weapon (an object with an ax on one side and a knife on the other) and bottles of gasoline. They splashed gasoline on the driver and threatened to burn him and themselves. They were confronted with 90 policemen for 25 hours until the driver was rescued. They were arrested on suspicion of unlawfully detaining the driver. Chen was imprisoned for seven months. However, in the meantime, the Pingdu government announced that it would give up on expropriating Chen's house and involving them in the village redevelopment project. Chen and his family defended their land at the cost of his imprisonment (Su, Reference Su2013).
The use of violence as a resistance signal requires resistant entrepreneurship. As Collins (Reference Collins2009) argues, people might resent the established laws and rules, but it might not give them the initiative, courage, and ability to resist fiercely. Instead, only a small proportion of people can afford to engage in violence and achieve their aims through it. Nonetheless, the effectiveness of violence as a resistance signal aiming to garner public sympathy does not depend on resisters’ physical power. Rather, it depends on how well they can identify the appropriate time to resist, control their power to make it appear that their violence is defensive, and convince the public that they mean to prevent government predation and not to harm anyone.
In Chen's case, he did not mean to injure the operator, although the operator got sunburned after prolonged exposure to the sun. Additionally, he reported himself to the police after he detained the driver, suggesting that he believed, or at least wanted to convince the public, that his violence was justifiable and defensive. Likewise, Anfang Zhang, a nail-resident struggling with a developer and the Bengbu government for eight years, kept fierce animals as a defense and equipped himself with four cameras, a spear, and a flame gun, making himself appear determined and prepared to defend against predatory land taking. When the Bengbu government attempted to conduct forced demolition, he smashed the excavator and threw bricks to deter it from approaching, but he refrained from attacking the operators. He was given a penalty of a five-month administrative detention for the offense of vandalism, but his resistant investment paid off when he was given six apartments for his family members (Yan, Reference Yan2011).
In contrast, excessive violence against others (such as homicide) can significantly undermine the legitimacy of nail-residents’ resistance, which can lead them to lose public support and prevent them from swaying the local authorities. Xueming Ma, a nail-resident, killed an agent of the property developer and an official of the Suzhou government by smashing a chair over their heads when he found they had not only refused to offer higher compensation but even reduced the original offer to punish him for delaying the demolition. Ma was sentenced to death. Engaging in violence left Ma little opportunity to negotiate with government officials, and it placed him in a morally unflattering position. Ma said despairingly that ‘they hounded me to death’ before he committed the homicide (Cao, Reference Cao2008). It might be arbitrary to say that Ma was irrational, because his violence was vengeful and may have satisfied his sense of fairness (Elster, Reference Elster2015). But compared with Chen's resistance, Ma's was much less tactical and failed despite its huge costs.
3.3 Self-burning
Self-burning is another drastic way for nail-residents to elicit public attention and sympathy. The Zhongs (Ruqin Zhong, Zhifeng Luo, and Zhongcheng Ye) found the compensation proposal of the Yihuang County government – either the amount of monetary compensation or the floor space of alternative housing – to be unacceptable. Observing that their neighbors’ houses were slated for coercive demolition, when about a hundred officials of the Yihuang government went to their home to persuade them to accept the compensation proposal, Ruqin, Zhifeng, and Zhongcheng doused gasoline on their windows and roof and started a fire; the fire spread rapidly to themselves. Unfortunately, Ruqin and Zhifeng were severely burned, and Zhongcheng burned to death. The unexpected and tragic incident attracted tremendous attention. Considering that the Yihuang government was highly likely to be punished by a higher-level government for failing to maintain social stability, and considering that sacrificing lives for land would undermine its trustworthiness and legitimacy, the Yihuang government decided not to expropriate the Zhongs’ house, paid 300,000 yuan as compensation for the death of Zhongcheng, and promised to pay all of the medical expenses and compensate the family for emotional damages to Runqin and Zhifeng (Wang and Wu, Reference Wang and Wu2010).
Self-burning is a salient symbol of grievance that can provoke the public's sympathy and even ‘stir the crowd's emotion against [the] local state’ (Fu, Reference Fu2018: 186). It is meant to communicate that ‘even death is preferable to living under unacceptable political conditions’, and therefore it erodes the moral foundation of local governments’ eminent domain (Lee and Kleinman, Reference Lee, Kleinman, Perry and Selden2003: 301). Self-burning, as a suicide attempt, ‘may cause physical damage and even death [and] can get the message across’ and can therefore effectively morally and politically corner local governments (Becker and Posner, Reference Becker and Posner2004: 15; see also Rosenthal, Reference Rosenthal1993). Empirical evidence shows that local governments are very likely to be punished when residents burn themselves (see Table 1). For instance, the Yihuang officials responsible for the Zhong case were dismissed from their posts. They might have faced a more serious penalty if they had failed to settle the dispute. By burning themselves, nail-residents credibly inform local governments of their desperation and ‘madness’ (Gambetta, Reference Gambetta2011). The costs and risks are so high that only a few residents can copy them, which can separate the residents who are truly desperate from those who pretend to be; in turn, local governments are put on higher alert to the threat.
The threat or act of self-burning became a focal point in residents’ set of strategies of self-harm for two reasons. First, it often takes a long time to go from threatening self-burning – by, say, demonstrating that one has gasoline – to setting oneself on fire. The time gives the residents opportunities to bargain with the local governments in a way that is not possible under the formal rules. This explains why some residents succeed at getting higher compensation without bringing on tragedy. It also explains why some local governments have torn down the houses of residents who doused themselves with gasoline: like merely holding out, merely threatening to burn oneself cannot function as a credible signal. Second, compared with other methods of self-harm, self-burning is more of an arresting image. It allows the residents to communicate how unfair the act of eminent domain is and how miserable they are. It elicits publicity sufficient to communicate news of the governmental predation even to those who do not observe it directly (Leeson, Reference Leeson2014). In summary, self-burning does not entail incurring considerable costs but does instantiate the art of resistant entrepreneurship.
Table 1 suggests that five of the seven nail-residents resorting to self-burning were nearly or over 60 years old and were grandparents. We might expect that, compared with younger generations, their future contribution to their households’ economic well-being was less. But they improved the material welfare of younger generations by sacrificing themselves to exert considerable pressure on local governments (Yu, Reference Yu2010). Death by fire was certainly not their favored outcome, but that does not imply they were irrational, given the potential for sharply increased compensation. The behavior may have been reinforced by the Confucian reciprocity norm that family members are obligated to assist (by providing financial support, say) and take care of each other (Pistrui et al., Reference Pistrui, Huang, Oksoy, Jing and Welsch2001; Zhang, Reference Zhang2017).
3.4 Media exposure
Reporting one's situation to the media is another way to pose a credible threat to a local government. Violence or self-burning was a common means of getting media exposure in the nail-house cases. For instance, Chen and the Zhongs broadcast their violent defense and self-burning, respectively, by texts and photos on Weixin (the Chinese version of Twitter) (Wang and Wu, Reference Wang and Wu2010). The sensational news and outrage quickly traveled on Weixin, which generated substantial pressure on local governments and reinforced the effect of the resistance signal. In addition, Chen was a journalist at Caixin (a mass media group), and he gained the support of many influential jurists and public intellectuals. They published an open letter to the Pingdu government, arguing that Chen's violence was a justifiable defense. The letter circulated online and attracted the signatures of 140 elites, which significantly challenged the legitimacy of the Pingdu government's eminent domain.
More remarkably, some nail-residents, though neither engaging in violence or self-harm nor having strong personal ties to the media to help spread their narratives, can still credibly showcase their resoluteness by seizing opportunities to shape narratives that capture the attention of the media and the public to challenge the legitimacy of local governments’ land-taking power. The famous case of Ping Wu and her husband Wu Yang in Chongqing is powerful evidence. The year 2006 was the fourth year after Wu and Yang nailed themselves to their house and also the year when the Property Law was passed. Article 42 of the milestone law declared, ‘The property of the state, the collective, the individual and other obligees is protected by law, and no units or individuals may infringe upon it’, thereby offering equal protection under the law to owners of state and nonstate property.Footnote 4 It was the first time that formal rules highlighted that it is necessary to safeguard the legitimate rights and interests of the expropriated owners of real estate. Though it did not specify operational rules, it implied that the central government realized the importance of protecting the property rights of ordinary people, which had become a high-profile topic all over the country (Chen, Reference Chen2010).
Prior to the passing of the Property Law, the Jiulongpo District government in the municipality of Chongqing had applied to the Jiulongpo court for permission to forcibly demolish Ping Wu's house, and the court granted it. However, Wu successfully turned the tide by wrapping her and her husband's resistance in the rhetoric of the statute and tapping into popular moral sentiment (Hess, Reference Hess2010). Wu draped over the building a scroll quoting Article 13 of the constitution: ‘The citizens’ legal private property rights must not be violated!’Footnote 5 More dramatically, Wu Yang equipped himself with water, food, and a liquid-gas storage tank, climbed over the barriers surrounding his building, raised the national flag of the People's Republic of China, and prepared to live on the roof until the couple obtained satisfactory compensation. By framing it as the first case tested under the landmark Property Law, their resistance had high narrative capacity. Their actions caught the eye of media outlets such as Southern Metropolis Daily. Simultaneously, by waving the constitution and national flag in front of journalists, they got the central government's backing, which put substantial pressure on the Jiulongpo government.
Both the public and elites showed considerable concern about the Jiulongpo government's response to Wu's claim. Although the Jiulongpo government was not as reckless and radical as the Yihuang government in the Zhongs’ case, the public was interested enough to warn the Chongqing government to deal with the issue cautiously. The mayor of Chongqing put it at the top of the agenda and led the negotiation with Wu in person. In the end, Wu and Yang acquired an alternative house and monetary compensation in the amount of one million yuan beyond what was in the original compensation proposal.
Similarly, Zhuxiang Cai, whose house faced expropriation in Shenzhen, also succeeded in eliciting media attention. Besides claiming that the expropriation would violate the Property Law, Cai hit the Shenzhen government in its soft spot by comparing it with the Hong Kong government. Cai had emigrated to Hong Kong (very close to Shenzhen) and obtained citizenship there. He noticed that the eminent domain institution in Hong Kong allows the territory's residents to participate in the negotiations. As the administrative authority of a special economic zone, the Shenzhen government boasts inclusive governance, a crucial comparative advantage to attract talent and external investment (Moberg, Reference Moberg2015). After Cai presented his experience – lack of bargaining power in the land-expropriation project in the village of Caiwuwei – on Oeeee, a popular online forum focusing on public affairs in Shenzhen, many journalists from Hong Kong flocked to Shenzhen to observe the event and compare the relevant rules in Hong Kong and Shenzhen. The Shenzhen government was concerned about criticism. To signal its commitment to protecting property rights, which Hong Kong's experience demonstrated is valuable to investors, the Shenzhen government sacrificed part of the land revenue and made a compromise (Farber, Reference Farber2002). It paid Cai monetary compensation of 1,200,000 yuan (2.5 times the initial offer) (Fu, Reference Fu2007).
Describing one's experience or expressing opinions on media platforms appears easy, but it is challenging to quickly attract media coverage and public attention, which requires resistant entrepreneurship – more specifically, novel ideas and strong narrative capacity. Wu seized the moment when the Property Law was passed, and Cai took advantage of the facts that Shenzhen is near Hong Kong and that the Shenzhen government highly valued its reputation and is sensitive about being compared unfavorably with Hong Kong. They were able to construct arguments for private property protection based less on the letter of the law and more on popular and elite discourse (Erie, Reference Erie2012). They triggered heated discussion on the legitimacy of local governments’ eminent domain power among the public.
It bears emphasis that the ways the nail-residents framed the injustice of the eminent domain power are difficult to repeat, as the residents benefited from voluminous media discussion of a time-sensitive matter. The media and public do not continue to concentrate on the same or similar stories. For nail-residents to stay in the spotlight, their followers must be able to create new stories and frame the residents’ unjust treatment in a way that will continue to elicit public sympathy (Mertha, Reference Mertha2009). As one follower, Shiguang Wang, said, ‘Not everyone can afford to be as visible as Wu Ping's case’ (Erie, Reference Erie2012: 44). Nail-residents’ ideas for garnering public attention might not have been costly for them; but not everyone could generate ideas for making a news sensation, and therefore such behavior was unavailable for them (Tufekci, Reference Tufekci2017).
We can see from the above cases that by holding out and by resorting to violence, self-burning, or media disclosure as supplementary and extralegal bargaining approaches, nail-residents were often effective in eliciting public sympathy and inducing higher-level governments to criticize local governments, which imposed considerable political and moral pressure on local governments and thereby deterred eminent domain or induced higher compensation. The actions credibly signaled that if local governments did not abandon the unsatisfactory proposals, the residents were determined and capable of further engaging in violent or conspicuous resistance and spreading their narratives. The resistance signals should credibly communicate the resoluteness of the nail-residents, which hinges on the residents’ resistant entrepreneurship: they discover legitimate resistance opportunities and seize them by engaging in costly or risky actions, such as violence, self-burning, or attracting media coverage and public attention with strong narratives. It therefore accounts for why simply holding out was often less effective.
4. The spillover effects of resistance signaling on institutions
Resistance signaling is essentially the actions of individuals or households, in the role of resistant entrepreneur, struggling for personal gains rather than demanding systematic change of formal rules. However, the signaling strategies can be copied and developed by others and therefore might have unintended spillover effects on institutions, as the effect of the nail-house strategies on China's formal land-expropriation institutions shows.
Strategies of resistance signaling can spread in two ways. First, a resistance signal serves as an important means of creating common knowledge through social and mass media's ‘gradual effect’, which refers to the media's ability to cause marginal institutional changes by gradually informing others and influencing their ideas (Coyne and Leeson, Reference Coyne and Leeson2009). News about the successful nail-residents shows others that it is possible to get higher compensation, what types of strategies provide credible resistance signals, and how the strategies work, which encourages other residents acting as resistant entrepreneurs to bargain with local governments. The nail-house strategy becomes a focal point. As Kuran (Reference Kuran1995) points out, one person's small act of dissent can encourage others to dissent as well. When Wu and Yang protested on their roof, for example, many residents flooded into Chongqing to express support for them and learn resistance tactics from them. Cai also acknowledged that getting media exposure is an art and that he learned a lot from Wu.
Second and similarly, we can see a neighborhood effect. Cai's success directly stopped the redevelopment of an urban village in Gangxia, Shenzhen, that is quite close to the village of Caiwuwei. The percentage of villagers who accepted their original compensatory arrangements dropped from 85 to 50%, as the news that Cai had acquired high compensation significantly increased the compensation they expected for themselves (Zhang, Reference Zhang2011). This is consistent with the finding of Alexander and Skyrms (Reference Alexander and Skyrms1999) that individuals tend to imitate the most successful people in the neighborhood.
With resistance signals being copied by more and more followers, conflicts have occurred more frequently. Amnesty International (2012: 56–63) documents nine deaths and 41 individuals’ self-burning as a result of coercive demolition between 2009 and 2011. This has increased the costs of enforcing land property rights, and it has challenged the legitimacy of the eminent domain power and decreased the trustworthiness of local authorities. Though relying on land-lease fees, governments have realized it is necessary to adjust some rules to mitigate cumulative resentment, either by revising top-down statutes or by undertaking policy experiments.
Specifically, rules governing state-owned land expropriation were revised in 2011Footnote 6 to increase the bargaining power of residents and the fairness of procedures. The changes introduced third parties as institutions that evaluate the market price of the houses on the land to be expropriated, and they stipulated that compensation would be decided by all or the majority of residents through consultation. As Table 1 shows, resistance against state-owned land expropriation has fallen since then.
Additionally, though top-down rules governing expropriation of collective-owned land are still absent, some large cities, such as Shenzhen and Guangzhou, whose local governments are less dependent on revenue from land leases, have developed a revolutionary set of rules on eminent domain in their urban-village redevelopment program as a policy experiment (Lai et al., Reference Lai, Wang and Lok2017; Lin, Reference Lin2015). The program allows villagers to cooperate with developers to reconstruct the villagers’ houses. In Shenzhen, the houses cannot be torn down until all villagers agree with the redevelopment-and-compensation proposal.Footnote 7 In Guangzhou, 80% must agree.Footnote 8 This creates a decentralized democratic mechanism for the villagers as a group to autonomously weigh the value of the land and negotiate with developers.
In sum, the cumulative effect of the nail-house phenomenon, along with such other forms of land disputes as collective protest, petition, and lawsuits, gradually increased the enforcement costs, demonstrated the injustice of the formal rules, and finally activated the ‘tipping points’ of change of formal institutions (Coyne and Leeson, Reference Coyne and Leeson2009: 16). To sustain the legitimacy of and trust in the autocratic regime, central and some local authorities are motivated to improve public policies. It reflects a significant pathway of institutional change in autocracies: rightful resistance against predatory or incompetent local authorities by individuals who seek to protect their own properties and interests can serve as an extralegal channel to inform central or other high-level authorities of principal-agent problems and provide them with constructive criticism, which contributes to the resilience of autocracies.
5. Conclusion
This paper examined how resistant entrepreneurs, who confront predatory land taking by a government under autocracy and lack effective formal channels to protect their property rights, protect their assets by sending resistance signals. The resistance signals attempt to expose the immorality of local governments’ acts of eminent domain, garner public sympathy and support, induce punishment by the higher-level governments, and thereby effectively force the hand of local authorities. The nature of resistance signals is twofold: resistance ex ante against eminent domain and extralegal bargaining efforts with governments. Substantiated by the evidence of nail-house strategies, the paper illustrates the conditions of effective resistance signals, which have to be able to challenge the legitimacy of governments’ eminent domain authority and have to be credible enough to force governments to make concessions. Credible resistance signals are often costly or unavailable, and only resistant entrepreneurs can afford or have access to them. The resistant entrepreneurs can increase the costs of enforcing the unjust formal rules and might unintentionally precipitate institutional change. This paper's contribution is not limited to the topic of resistance against land taking. It attempts to contribute to rational-choice theory by accounting for seemingly odd and sacrificial behaviors. It also enriches our empirical understanding of the operation of autocracies and the possibility of resistance at the individual and local level within them.
Nonetheless, some questions about individual resistance remain unanswered and call for further research. Why are some people more motivated to resist predation? Do personal attributes (such as age, gender, family status, or education level) or preference for fairness explain it? Interdisciplinary research might help to solve these puzzles. Insights from sociology and psychology that have been overlooked by political economists might help us to understand whether and how resisters’ motivations are rational. Exploring these questions requires multiple approaches. Qualitative work can clue researchers in to the causal effects of certain variables, while quantitative analysis can help to test them. Data from the field are useful but might be difficult to collect, so a quasi-experimental approach might complement such data.
Acknowledgements
I thank the editor, Geoffrey Hodgson, and four anonymous reviewers for useful comments. For stimulating discussions and comments, I would like to thank Li Ding, David Skarbek, Javier Portillo, Ilia Murtazashvili, Mark Pennington, Peter Leeson, Emily Skarbek, Peng Wang, John Meadowcroft, and Shitong Qiao.
Appendix 1: A simple model of resistance signaling
A signaling-game model synthesizing the ex ante resistance of resistant entrepreneurs and their bargaining with local governments over compensation can help us understand the conditions determining their resistance (or bargaining) strategies. The game goes as follows (Figure A1):
(1) Nature assigns the types of residents: resistant entrepreneur (RE) with probability p (0 < p < 1) or not a resistant entrepreneur (${{\rm R}{\tilde {\rm E}}}$
) with probability (1 − p).(2) Residents decide whether to send (S) or not send (NS) a resistance signal. The signaling costs of resistance ex ante to RE and ${{\rm R}{\tilde {\rm E}}}$
agents are c and $\tilde{c}$, respectively ( $\tilde{c}$ > c > 0).(3) Local governments decide whether to insist on paying residents low compensation (L) (costing them l) or instead offer high compensation (H) (costing them h). Local governments acquire t from taking the land (t) (t > h> l > 0).
(4) If they get low compensation, residents decide either to further resist fiercely ex post (R) to get additional compensation (a) or to not resist (NR). The costs of resistance ex post for RE and ${{\rm R}{\tilde {\rm E}}}$
agents are c* and $\tilde{c}_\ast$, respectively ( $\tilde{c}_\ast$ > a > c* > 0), indicating that in the subgame, RE agents resist ex post while ${{\rm R}{\tilde {\rm E}}}$ agents do not. If residents resist, local governments incur a substantial cost s (l + s > h). For simplicity, I assume the percentage of RE agents (i.e. p) is relatively low, such that (t − h) < (t − l − s)p + (t − l)(1 − p) (i.e. h − l > ps) holds.
Figure A1. Resistance-signaling game.
I deploy the perfect-Bayesian equilibrium concept to solve the game. Given that l + s > h > l, it benefits local governments to offer RE and ${{\rm R}{\tilde {\rm E}}}$
high and low compensation, respectively. But whether they are able to make the distinction depends on residents’ signaling costs (c and $\tilde{c}$
). There are three potential equilibria.
Equilibrium 1: when h − c > l + a − c* and h − $\tilde{c}$
< l, there is a separating equilibrium: RE agents choose S and ${{\rm R}{\tilde {\rm E}}}$
agents choose NS; governments offer H when they observe a resistance signal but L when they observe no signal.
This equilibrium implies that to ensure the effectiveness of the resistance signals, the signals must be costly or inaccessible so that only RE agents can send one and credibly reveal their type
Equilibrium 2: when h − c > l + a − c* and h − $\tilde{c}$
> l, there is a semi-pooling equilibrium: RE sends the signal, ${{\rm R}{\tilde {\rm E}}}$
adopts a mixed strategy (choose S with probability x and NS with probability 1 − x), and local governments also adopt a mixed strategy (offer L with probability y and H with probability 1 − y). According to Bayes’ rule, x must satisfy 0 < x < 1 and
(such that x = ((p(l + s − h))/((1 − p)(h − l)))). Likewise, y must satisfy 0 < y < 1 and (1 − y)(h − $\tilde{c}$
) + y(l − $\tilde{c}$
) = l (such that $y = 1\;-( \tilde{c}/( h-l) ) \;$
).
This equilibrium implies that when the resistance signal is not credible enough to make the residents’ types fully distinguishable, ${{\rm R}{\tilde {\rm E}}}$
agents might be able to mimic RE agents and acquire high compensation and RE agents will resist ex post if they receive low compensation. This accounts for why simply holding out or relying on self-burning sometimes is less effective at cornering local governments than threatening self-burning or combining holding out with other strategies is.
Equilibrium 3: when h − c < l + a − c* and h − $\tilde{c}$
< l, there is a pooling equilibrium: both RE and RE~ agents choose NS, and governments offer L to both types of resident.
In this equilibrium, RE agents strictly prefer resistance ex post to resistance ex ante. It may be because they are more familiar with the ex post tactics than ex ante tactics. However, given that land taking is an ongoing and open process, it is rare in reality that RE agents do not resist if they receive unjust compensation.
The analysis in the main text mainly concerns the first two equilibria.
