Complicity in democratic engagement with autocratic systems

ABSTRACT Autocratic control of civil society, including academia, can be extended to democratic societies and institutions in ways that pose threats to liberal-democratic values, such as academic freedom, for example through mechanisms and practices that lead to academic self-censorship. Engaging critically with the literature on ‘sharp power’ and ‘authoritarian influencing’ addressing this phenomenon, this paper argues that democratic actors who, without sharing the repressive goals of autocracies, contribute to their success in settings of international collaboration and exchange can become structurally complicit with such wrongs. Recognizing the risk of complicity is a necessary first step towards addressing the political responsibilities resulting from it.

academic institutions' international collaboration and exchange; to sign "democracy pledges" to protect against anti-democratic funders' influence; and to adopt a principle of "reciprocity" when countering authoritarian civil society repression (Benner 2018;Watch 2019a; Working Group on China's Influence in the USA 2018) The background for these calls is that countries with autocratic systems of governance control people, coercively or by inducements, without the rights protections that underpin the domestic institutional frameworks of liberal democracies. As political communities and systems interact more intensely at a civil society level in today's globalized world, the control exercised by autocrats can be extended to democratic societies and institutions in various ways, potentially undermining the protections of liberal-democratic orders.
This paper engages with the problems raised by such cases, mainly drawing on examples of transnational academic exchange and research and critically engaging with some of the concepts and arguments that have shaped the debate of these issues so far. Using a scenario that amalgamates publicly discussed cases and issues, it argues that democratic actors and their systems are often not mere victims of 'sharp power' exercise but, rather, participate in or contribute to the wrongs of autocracies expanding their influence control beyond their national borders. However, as I illustrate through a selective discussion of liberal-democratic legal orders, legal liability and sanctions mechanisms are of limited use in addressing democratic actors' participation in autocratic wrongs. I argue that some forms of democratic participation in autocratic wrongs, understood as forms of complicity, engender responsibilities even in cases where there are no legally liable or individually culpable participants. Institutionalized responses to transnational autocratic wrongs discussed here must take responsibilities resulting from complicity into account.

A scenario
To illustrate problems of transnational participation in autocratic wrongs, let us imagine the following scenario. 1 Dr Zhang is a researcher and human rights activist in Beijing. After years of surveillance, forced disappearance and torture, as well as censorship of his writings, he leaves China for a one-year visiting researcher position at a university in London ('the University), funded by an NGO ('the NGO').
During Zhang's year in London, his public talks and guest appearances in academic classrooms appear to be reported back to the 'relevant authorities' in China in minute detail; they lead to interrogation and criminal detention of fellow advocates and family members, held in order to pressure Zhang into silence. 2 One of the persons reporting on Zhang may be current University student Yang, who has attended almost all events with Zhang and posted critical comments on Zhang via her Chinese microblog account. However, whether there is a human informant and who it might be remains speculative. 3 It is possible that Zhang is being subjected to electronic surveillance, especially since the University, like other universities, is using more and more technology that facilitates monitoring by the Party-State and its dependent corporate enforcers (Association of Asian Studies 2020).
Zhang's persecuted fellow advocates in Beijing report that domestic security police division officer Sun has been directing and supervising their continued persecution. The Chinese security apparatus is thought also to be responsible for anonymous death threats sent to Zhang's London home address. 4 Sun is known to visit Europe once a year with his wife on a tourist visa; they have opened bank accounts and are planning to buy properties in the UK and Germany.
Zhang struggles with his traumatic memories of persecution in China, as well as the continued attempts to intimidate and control him remotely. He knows that he would be criminally detained directly if he returned to China. He also feels anxious and exposed when communicating by phone or email. He knows that all China-based social media, search and other online services are integrated into China's intelligence and law enforcements systems (Sigal 2018). Yet avoiding these services entirely means that he will not only be cut off from friends and family in China, but also cut off from Chinese news and useful information sources. On the other hand, he has no confidence that his UK phone is safe from being hacked by a China based entity.
Zhang tries to obtain a research fellowship funded by the University but is told that he is unlikely to get it, partly because his research proposal has academic weaknesses, and partly due to anxieties on the part of China Studies experts serving on the University fellowship selection committee. The University's China studies department has several large joint research programmes, funded by various Chinese entities, with prestigious universities in China, vulnerable academic partners in China, and Chinese students paying tuition fees at the 'overseas' rate. One or two of the University's academics are also thought to hold lucrative visiting researcher positions in China, but the details of these positions are not known. The University is also considering the establishment of a 'Confucius Institute' with Chinese government funding. 5 The China studies experts are also aware that scholars critical of China have reportedly suffered mysterious attacks on their homes, 6 and that in response to 'Magnitsky' sanctions by the EU, the UK, and other countries, China has inter alia banned critical Uyghur studies scholars and their families (sic) from the country. 7 If he fails to get the fellowship, Zhang is unlikely to obtain an extension of his current scholar visa in the UK, and although he could probably get asylum, he hesitates to take this step, which would make it near impossible for him ever to return to his country and see his ageing parents.
The NGO has meanwhile indicated, reluctantly, that it will not be able to provide continued financial support, possibly because it is currently seeking registration under the strict rules of China's new Foreign NGO Management Law and concerned about the impact of supporting such a prominent critic as Zhang. 8 Zhang was in hopes of augmenting his income by publishing a book-length account of his struggles. After initial indications of strong interest, he is disappointed to learn that an academic publisher ('the Publisher') based in Germany will not, after all, offer him a contract. The Publisher says this is because the book is commercially nonviable, but Zhang thinks that it is a reaction to threats that the Publisher might be sued for defamation of Chinese officials, and that the book might lose the Publisher Chinese business. 9 Yang, meanwhile, is luckier. Having achieved excellent grades, she transitions from her master's to a PhD programme at the University, on the topic of 'The Concept of Tianxia (天下) and the Future of Global Governance.' 10 To get funding for her doctoral project, Yang carefully prepares an application that will raise no concerns of 'political sensitivity' with the China Scholarship Council (CSC), a government-directed entity. When Yang obtains a CSC scholarship, she, her supervisor, and the University receiving her tuition fees are delighted.
But ten months into her programme, Yang changes her mind about tianxia governance; she decides to write her dissertation instead on the more 'sensitive' topic of the Party's governance of ethno-religious minorities, with a focus on crimes against humanity in Xinjiang (Human Rights Watch 2021). Yang's friends advise her that this change could result in her losing the scholarship and that it might get her, and perhaps also her family and former teachers in China, into trouble. Yang feels that she and her supervisors have little choice but to falsify her progress reports to the CSC, hoping to deceive it about her topic change to avoid persecution affecting herself or others.

Democratic participation in autocratic wrongs: the limitations of 'sharp power' discourse
In the above scenario, the power exercise and influence produced by actors of or under the influence of the Chinese Party-State vary across a broad range, at one end of which is (material) inducement, and at the other end of which are physical coercion and fear (Etzioni 1968). These kinds of influencing cannot always be neatly separated. Money, in the case of autocratic actors engaging in transnational exchange, does not so often 'talk' as it silences. Self-censorship induced by financial dependency -the attractiveness of the funding for a conference, a research centre, or a research students' tuition -can easily be heightened by fears for colleagues in partner institutions, responsibility for research students at risk of repression, for example. Inducement and coercion can combine to become powerful motives -and potentially, moral justifications -for staying silent or engaging in deceit.
Political scientists have introduced a variety of concepts to grapple with these phenomena (Walker 2018;Benner et al. 2018;Godement and Vasselier 2017). The starting point for these analyses is the belief that China's power exercise is not to 'make us want what the Party-state wants' (soft power), but rather to render us no longer able to get what we want -liberal democracy and associated values (Nye 2018). According to Ludwig and Walker, China makes use of 'sharp power' by exploiting the asymmetry between open liberal-democratic systems and closed authoritarian systems, rendering democracies' 'spheres of culture, academia, media and publishing' more vulnerable. In what they see as 'the new competition' between autocracies and democracies, autocracies' 'sharp power' techniques 'should be seen as the tip of [autocracies'] dagger -or indeed as their syringe.' (Ludwig and Walker 2018, 6) The metaphor of sharp power does help capture some of the aspects of transnational repression discussed earlier on: the detention of an academic from a non-Chinese institution in China, for example (Feng 2017) or the raiding of a non-Chinese academic's office (Scholars At Risk 2018 b), or a travel ban imposed on them (Solidarity statement on behalf of scholars sanctioned for their work on China/学界联署声明:声 援因从事中国研究而被制裁之学者2021), or the hacking into confidential information held by a university (Liao 2019), or threats sent to an academic in exile, or the use of an autocratically controlled judicial system to bring vexatious litigation against a government critic abroad (Clarke 2021). In all these cases, individuals or organizations operating within the parameters of a liberal-democratic system are victimized in ways calculated to produce effects well beyond the individual target, exploiting the openness of the democratic system. However, the assault metaphor of 'sharp power' does not help much in addressing the fact that democratic actors' agency can compound the wrongs of autocratic repression -for example, when publishers in democratic societies decide to selfcensor or when universities withhold support from the victims of transnational academic repression, as in the scenario above. The image of the sharp, moving object used (actively) to perforate, damage, wound, destroy or kill the (animate, but passive and static) fabric of the liberal-democratic system invites us to focus on those wielding the dagger or syringe as agents. Because it casts the targets of sharp power as passive, the image does not invite us to consider the ethical complexity of acting for sharp-power-related reasons. Yet, if we accept that the exercise of such power gives its targets reasons to act that do not necessarily constitute justification, an evaluation of the conduct of actors in transnational repression scenarios must take the availability of moral choices on both (autocratic and democratic) sides into consideration, even though 'sharp power' exercise may limit available choices, or render them unduly hard. 11 Because it does not seem to commit to a view on the free will or the agency of sharp power targets, sharp power theory also limits our understanding of targeted democratic actors' responsibilities. And yet, in the context of academic exchange, for example, authoritarian and liberal-democratic systems interact in complex ways. An analysis of such complex situations that aims at an evaluative assessment of actions and decisions -and I would suggest that discussions of autocracy and democracy tend to be implicitly, if not explicitly evaluative -would have to address this complexity. Also, at least if the goal of analysis is an ascription of culpability, the analysis requires some engagement with the state of mind of relevant actors (e.g. willing, mindless, reluctant, or coerced) with regard to any specific wrongs of repression. 12 Because it relies on contrasting autocratic with democratic systems, the sharp power metaphor might also invite an exaggerated focus on regime types. It might suggest, selfservingly, that liberal democracies use soft power by definition, whereas autocracies use sharp power. Yet as Glasius and others have argued, actors in liberal democracies may themselves be engaging in authoritarian practices or use authoritarian conceptions of governance (Glasius 2018;Tsimonis and Rogelja 2019). Such actors are not entirely defined by the political system they are part of. For example, the governance structure of a university in a liberal democracy may be undemocratic. 13 A democratically elected government may yet actively work to undermine the democratic principles that produced it (Huq and Ginsburg 2018;Scheppele 2018). These observations urge the conclusion that autocratic-minded actors within democracies may be users of 'sharp power' mechanisms.
Conversely, for an assessment of the ethics of complicity with autocratic repression, one must also recognize the complexity of actors within autocratic systems. Sharp power discourses appear to have difficulties distinguishing actors and motives within an autocratic state, because it relies so heavily on a dichotomy of democratic and autocratic actors, but such distinctions cannot be neglected in an assessment of culpability or if, for example, one wishes to determine who ought to bear the burden of sanctions in response to transnational repression.
In sum, although the 'sharp power' argument has made an important contribution to our understanding of autocracies' global reach, we should question the implicit activepassive paradigm associated with it and we should also remain alert to the possibility that governance practices within systems categorized as democratic or autocratic can shift, change and vary along an autocratic-democratic spectrum. This interim conclusion is borne out by the scenario introduced earlier. While there can be a point in describing some incidents in the scenario as sharp power attacks, this approach is not helpful for understanding the scenario in its entirety. Rather, the scenario calls for a normative assessment of the legal and moral responsibilities of specific agents or actors within the affected political system, including the degree to which liability or culpability can be ascribed in these complex situations.

The limitations of legal responses to participation in autocratic wrongs
As discussed in the previous section, a 'sharp power' based analysis of the scenario is limited not only by the dichotomies of active and passive, victim and perpetrator, but also focused on a descriptive, rather than a normative understanding. To overcome these limitations, a normative analysis will need to address the scenario from legal and moral perspectives.
As this section sets out, in some cases, democratic participation in autocratic wrongs can trigger legal liability, for example, under tort law or criminal law. Tort or criminal liability often, but not always, depends on culpability. 14 The availability of criminal or tort law mechanisms varies across jurisdictions. However, the principles underpinning these mechanisms in all liberal-democratic orders also supply grounds for defences against liability. 15 The frameworks of tort and criminal law in different domestic orders therefore only provide limited bases for 'going after' actors participating in autocratic wrongs.
A few examples taken from the scenario help to illustrate the complex limitations of any legal responses to the issue of participation in autocratic wrongs at domestic (national), international or transnational levels. The discussion offered here draws on examples from English (or UK) and German law and shows that in both systems, the law has established high thresholds for individual liability for participation in criminal wrongs, not only in terms of standards of evidence and problems with ascertaining the facts, but also in terms of what constitutes criminal acts and criminal culpability. It also suggests that public law constraints of institutional decision-making hinge on the institutional design of the relevant democratic actors.
Both English and German law in principle criminalize some of the actions taken by Chinese government officials to silence Zhang while abroad, such as for example the death threat against him. 16 Similarly, in the case of the scholar targeted by vexatious litigation, the laws of liberal democracies would likely provide protection, because no state respecting academic freedom and freedom of speech ought to be recruited for the purpose of persecuting critical scholarship in this manner.
The German system, additionally, contains a provision on a crime of 'duress' or 'coercing another' (Nötigung), that might conceivably be applied to acts pressurizing Zhang if he were in Germany. 17 However, any liability for the crimes of assault or 'coercing another' in the present scenario can only be attributed to actors far removed from the jurisdiction these crimes are part of. Were such actors ever to come within reach of UK or German criminal justice, moreover, they might be able to defend themselves by pointing to the pressures they themselves were under. Additionally, there would be evidence issues. The likelihood of a prosecution -say, of Sun on one of his trips abroad -would appear to be minimal.
Regarding Yang, the student suspected of reporting back on Zhang, merely providing factually truthful information to the government would probably not trigger criminal liability under German or UK defamation law (which in Germany can trigger criminal and civil liability whereas in the UK it can only trigger civil liability). German criminal law contains a highly specific crime on exposing another to the risk of political 14 In legal terms, liability encompasses 'strict liability' scenarios that do not require a culpability test; however the present discussion does not deal with strict liability scenarios. 15 Potential liability under autocratic legal systems is not discussed here. Inside China, actors such as Sun operate with evident impunity while the system remains what might be described as 'dual state.' (Fraenkel 1941;Meierhenrich 2018). 16 Under English law, this could be an offence of a 'threat to kill' s 16 of the Offences Against the Person Act 1861; see also s. 3 (affray) of the (UK Public Order Act 1986). Under German Law it could be an offence of threatening the commission of a felony (s 241 of the German Criminal Code). 17 Section 240 of the German Criminal Code. This provision dates back to Germany's Nazi era and it remains a constitutionally contested provision, partly due to its vagueness.
persecution by 'informing' on them (s 241a of the German Criminal Code). This provision might conceivably apply to a person such as Yang.
Both the English and the German legal system might raise questions of criminal liability for fraud committed against the CSC as funder of Yang's project, if Yang and the University go ahead and provide false reports, but Yang and the University might well be able to argue that they did nothing wrong in criminal law terms, or at least that they are not culpable, for example because they acted under duress.
Although liability under tort law rules leads to different consequences for anyone found liable, compared to criminal liability, and although the process for determining liability differs, there are broad structural similarities between how the law deals with criminal and civil wrongs. Tort rules might trigger civil liability of those held criminally liable in the cases just mentioned. It is also possible that rules and principles of vicarious liability -for instance, if the suspicions towards student Yang were proved correct and if her actions could be ascribed to the University -might expand liability, but such an expansion would not apply to other kinds of relationship between primary wrongdoers and other actors in the scenario.
Beyond legal rules about liability, the law might constrain universities in their decision-making, for example, by prohibiting discriminatory or corrupt decisions or in the case of public authorities, by ruling out arbitrary decisions. However, in the scenario, it is hard to see how these broad and principled constraints could be used to stop the scenario's democratic actors from making the decisions they did.
In sum, the law of a liberal democracy, which rightly imposes inherent constraints upon the coercive power of the state, does not stop the NGO, the University, and the Publisher, from making decisions not to host or fund or offer a book contract to Zhang, or the University from deciding to accept Yang as a CSC funded scholar. Although these decisions result in harm and suffering, the scenario's democratic actors have not at this point done anything that would trigger legal liability of any sort. Liability under both criminal and tort law rules would be substantively limited, hard to establish, and difficult to litigate. So far as Zhang is concerned, they have not committed civil or criminal wrongs or participated in any acts that would make the principal wrongdoers liable, as their actions and decisions are not primary or sufficiently proximate causes of the situation Zhang is in. 18 In Yang's case, the only potential liability might arise in the context of vicarious liability for her potential fraud.
As the scenario further suggests, in real-life cases, there could also be problems with establishing some crucially relevant facts, such as who issued the threats against Zhang and on whose orders, and whether Yang did engage in reporting on Zhang. Lastly, substantive liability rules would have to reflect factors limiting culpability, such as for example well-founded fear of triggering repressive or punitive acts against one's project partners in China. These interim conclusions are borne out by the fact that in the scenario, the democratic actors' decisions are taken for a variety of reasons that would prima facie seem legally innocuous -for example, the Publisher's pursuit of commercial gain and the University's pursuit of income or funding for research. Even so far as they are deplorable, they seem understandable -for example, the NGO and University's decision to distance themselves from Dr Zhang, apparently for fear of bringing trouble to others or jeopardizing by themselves valuable projects. Yet, the democratic actors' actions and decisions have promoted repressive causes in various ways that could make them in complicit in a sense indicating not legal liability, but, rather, (as discussed in the next section).
On the basis of this analysis, we might see the University and actors within it, the Publisher, and the NGO, or indeed the liberal-democratic societies of which they are part, as victims of transnational repression. While they may have neglected duties towards primary victims such as Dr Zhang and Ms Yang, these actors become involved in repressive practices they are unlikely to have wanted. They might also, for example, face reputational damage as a result of what happens in the scenario, if their connection (however loose) with acts of repression and the fact that they benefit (even if indirectly, even in pursuit of broadly good or acceptable goals) from support by the repressive state in question becomes more widely known. In such a situation, they might point to the valuable goals they pursue and argue that they must not be reproached for their implication in the wrongs that befell Zhang and Yang and that in any event, the focus of attention should be on the agents of autocracy as primary wrongdoers.
Legislation such as the 2017 U.S. Global Magnitsky Human Rights Accountability Act, which authorizes asset freezes and visa bans targeting foreigners responsible for gross human rights violations or complicit in corruption, does indeed focus on what one might call primary wrongdoers in this way (Human Rights Watch 2017). The UK and EU have adopted such legislation. (Miadzvetskaya 2021) Government sanctions such as travel bans may be available in cases of egregious civil society repression -such as, for example, the murder of journalist Jamal Kashoggi in the Saudi Embassy in Turkey, or the imposition of repressive national security legislation on the Hong Kong Special Administrative Region (Salem 2018). In the present scenario, Mr Sun or his superiors might face a travel ban or asset freezes under Magnitsky legislation, or criminal charges if their actions reached the very high threshold of liability under international criminal law. 19 However, it is important to realize that the effort to identify and sanction wrongdoers abroad relies on what with Iris Marion Young, we might call a liability-focused conception of autocratic wrongs (Young 2013, chapter 3). In doing so, it also favours a 'sharp power' conception of these wrongs -using the image of the syringe or the dagger, Magnitsky sanctions can be used to go after whoever is thought to be yielding these instruments. Yet, such sanctions can lead to the problem, typical of tools of 'militant democracy,' that they might jeopardize what they seek to protect: the open society that accommodates, indeed celebrates, a plurality of views within it, and the protection from abuse of public power (Loewenstein 1937;Müller 2015;Wagrandl 2018;Accetti 2017). Imposing sanctions on supposed perpetrators of abuses, or on contributions to the democratic process considered illegitimate and unlawful, therefore requires powerful justifications. For example, it would be important to respect due process rights in the case of the revocation of media licences or of imposition of personal property sanctions by a host country; and in all cases, the human targets of such sanctions should not be dehumanized, e.g. by seeing them merely as representatives of an enemy group or nation.
Even if primary wrongdoers can be identified and sanctioned, moreover, it remains important to understand the responsibilities of the democratic actors concerned in the scenario, and as has been argued here, law in liberal-democratic systems rightly sets narrow limits to liability of complicit actors -such narrow limits are appropriate, because they serve the limitation of the power of the state as law enforcer. 20 To understand complicit actors' responsibilities fully, we therefore need to move beyond the question of whether they are legally liable or whether the law could stop them from taking decisions that help to 'transnationalise' autocratic repression. As Lepora and Goodin have argued, (2013, 94), the mere fact that someone has been excused does not mean they had no 'case to answer.' As Young argues, involvement in injustice can be 'structural.' It is these insights we need to consider further when assessing democratic actors' involvement in autocratic wrongs.

Different forms of complicity with autocratic wrongs
The previous section analysed the kinds of legal responsibility that different actors in the scenario might bear for the wrongs suffered by other actors in it. It concluded that while the law and the duties, responsibilities and liabilities it establishes are relevant to some aspects of the scenario, legal liability is -rightly -limited to instances of individual wrong actions, that is, it is usually limited to cases where someone culpably and attributably caused a specific harm to another. In the transnational scenario considered, those who are most evidently legally liable for specific wrongs tend to be actors within autocracies unlikely ever to be held to account within their own systems, and the bases for attributing legal liability to other actors are limited, due to the causation and culpability (or attributability) considerations set out. This conclusion urges us to consider alternative, broader conceptions of complicity with the autocratic wrongs under consideration, to follow up on the intuition that some of the democratic actors have nonetheless made morally significant contributions to these wrongs.
Young has argued that claims of injustice 'often concern structure.' In her essay on structural injustice and her later Responsibility for Justice, she presents a case of socioeconomic injustice affecting a woman who faces homelessness as a result of structural domination and oppression; her difficulties, Young says, are not 'sheer bad luck' but, rather, they are 'socially caused.' (Young 2003) They result from structures, understood as the 'relation of social positions that condition the opportunities and life prospects of the persons located in those positions' (Young 2003) Actions (necessarily) taking place within these structures have unintended consequences for people's opportunities and constraints, she argues, and they tend to consolidate the structures in question (Young 2003) On this basis, Young defines structural injustices as 'harms that come to people as a result of structural processes in which many people participate' (Young 2003, 6) and as [existing] 'when social processes put large groups of persons under systematic threat of domination or deprivation of the means to develop and exercise their capacities,' making it 'a kind of moral wrong distinct from the wrongful action of an individual agent or the repressive policies of a state.' She adds that 'social structures do not constrain in the form of the direct coercion of some individuals over others; they constrain more indirectly and cumulatively as blocking possibilities.' In her book (Young 2013), Young expands her analysis to global structural injustices, using the illuminating example of harms suffered by women working in sweatshops for the global garment industry in countries of the Global South.
Young's analysis is pertinent to some actors in the scenario who live in interconnected and (to varying degrees) nonideal societies that have landed them in situations over which they seem to have no or little control. Zhang and Yang, in particular, have been born into a persecuting society and happen to be citizens of a country that oppresses them, especially if, like Zhang, they hold and express dissenting views that have triggered severe persecution and forced him into an exile that still offers him no safe refuge. Yet their situations result neither from what Young would call 'sheer bad luck' nor from entirely free choices on their part (Although in some domestic discourses, there may be a tendency to argue that Zhang's difficulties were deserved (in Chinese, huogai 活该), because predictable and avoidable, from the perspective of this analysis, he must not be blamed for using speech to criticize an immoral system. 21 ) The NGO, the University, and the Publisher are in some ways also victims of China's repressive actions. They face challenges resulting from power positions, relations and processes that export autocratic lawlessness and oppression into liberal-democratic societies. However, each of these actors' connection with autocratic wrongs described in the scenario is different and is more or less a matter of choice, which complicates an account of their responsibilities.
However, comparing the scenario at hand and the structural injustice examples Young discusses, there are some important dissimilarities that need to be addressed. For one thing, we cannot really say that the scenario illustrates a case of oppression or domination without identifiable agents engaging in domination and oppression. From a public international law perspective, as human rights NGOs and other governments, as well as people like Zhang keep pointing out, the Chinese government is responsible for its violations of rights to freedom of expression, integrity and liberty of person, and so on. Due to China's internal political structure, the ultimate agent of oppression in the scenario is the Chinese Communist Party, currently headed by its General Secretary and PRC President Xi Jinping, who, along with others, would bear responsibility for the wrongs discussed in the fictitious scenario. In other words, there are some clearly identifiable agents; the problem is just that they occupy a distant and -at least for the purposes considered here -lawless space.
For another, while both in Young's domestic scenario (concerning Sandy) and the transnational one (concerning sweatshops), the primary injustices considered concern the challenges arising from poverty and dispossession, the injustices of the present scenario primarily concern civil and political rights violations (such as arbitrary detention, infringement of expression and personality rights, etc.).
But as I discuss below, Young's account is nevertheless relevant to the scenario, because Young rejects the 'liability model' of responsibility and instead proposes a 'social connection' model of responsibility. It is also not categorically limited to socio-economic injustice, as Young's abstract account of the model clarifies (my emphases): 'The social connection model of responsibility says that individuals bear responsibility for structural injustice because they contribute by their actions to the processes that produce unjust outcomes. Our responsibility derives from belonging together with others in a system of interdependent processes of cooperation and competition through which we seek benefits and aim to realize projects. Within these processes, each of us expects justice toward ourselves, and others can legitimately make claims of justice on us. All who dwell within the structures must take responsibility for remedying injustices they cause, though none is specifically liable for the harm in a legal sense. Responsibility in relation to injustice thus derives not from living under a common constitution, but rather from participating in the diverse institutional processes that produce structural injustice.' (Young 2013, 105) The political and social reality of our 'globalised world' is one of greater interconnectedness and closer associations, not only in business relations such as global supply chains, but also among civil society actors who increasingly 'dwell within structures' of transnational cooperation and interaction, as illustrated by the scenario used in this paper. Conversely, transnational repression and practices of lawlessness are bound to create more and more connections with autocratic wrongs in other jurisdictions or societies -connections relevant to the moral assessment of who is to take responsibility for these transnational autocratic wrongs, and how.
Drawing on analyses advanced by Young and others, and taking into account the fact that democratic actors' legal liability for autocratic wrongs is limited, we should consider the possibility of democratic actors becoming complicit in autocratic wrongs without intending to do so, and potentially under circumstances where their individual culpability is greatly reduced or where indeed, they are not culpable, because they are both victims and perpetrators: they have agency, but they are also under undue pressure, pressure that may be so great that we cannot individually blame them for succumbing to it.
The situation of some actors in the scenario is well captured by this description. While in the scenario, the country's political leadership and officials are responsible for many wrongs, other wrongs are the result of, in Young's terms, 'harms that come to people as a result of structural processes in which many people participate.' (Young 2003, 6) The practice of censorship and self-censorship in academic life in repressive countries could be described as a structural process in Young's terms, for example. The participants in this practice are in many ways both victims and perpetrators. They have not necessarily done the right thing, but it may be hard to blame them or hold them liable if, for example, they have acted out of fear for themselves or others.
Drawing on Young and some of the authors who have critically engaged with her work, we can sketch out an account of democratic actors' involvement in autocratic wrongs. I propose that we can understand some instances of such complicity as 'remote,' 'mitigated,' and 'structural complicity.' The reason why I propose these differentiated conceptions of complicity is that we can observe different causal relationships and degrees of moral attributability at work in the contributions democratic actors make to the kinds of autocratic wrong discussed in the scenario.
I would suggest that in cases of remote complicity, the causation of harm more directly caused by a specific autocratic wrong is weakened by distance (in a sense similar to 'remote causation' as used in tort law); that complicity is mitigated where the an actor was under undue pressure to contribute to the causation of a harm more directly caused by an autocratic wrong, and acted out of fear for themselves or others in a way that made their action less or not at all blameworthy; and that complicity is structural in cases where an actor, to use Young's phrase, merely 'contribute[s] by their actions to the processes that produce unjust outcomes.' This account follows Young in observing that '[t]o judge a circumstance unjust implies that we understand it at least partly as humanly caused' (Young 2013, 95). It departs from Young, and follows Aragon and Jaggar, in suggesting that an extended, structural conception of complicitynot necessarily reliant on what Young calls the liability model -is appropriate to capture certain types of contribution to or participation in injustice. (Aragon and Jaggar 2018) 22 This account also draws on the much more complex and comprehensive work of Goodin and Lepora on the concept of complicity. Goodin and Lepora, in their nuanced discussion of contributions to others' wrongdoing, capture the fact that there can be justifications for complicity. With Lepora and Goodin, we therefore could acknowledge not only that the democratic actors in the scenario do not share the repressive purposes of the Chinese Party-State and that some of them may even qualify as 'moral agents trying to do the best they can in a world full of bad.' 23 In some situations, such as that of the NGO in the scenario, 'being complicit with the wrongdoing of another, doing something that potentially contributes causally to that wrongdoing, may be the best thing to do in a bad situation.' 24 Whereas Goodin and Lepora treat fear as a reason for action -for example, in a bank robbery scenario in which victims act on the basis of 'reasonable fear' 25 -, however, complicity can be mitigated also by fear that has clouded reason. Fieldwork experience suggests that living under constant threat of violence can lead to erratic judgment (in both directions: Pils 2014, chapter 6). Scholars of liberalism such as Shklar have reminded us of the ways in which fear 'created by arbitrary, unnecessary, and unlicensed acts of force and by habitual and pervasive acts of cruelty' can undermine liberalism, as a system in which '[e]very adult should be able to make as many effective decisions without fear or favor about as many aspects of his or her life as is compatible with the like freedom of every adult.' (Shklar 1989, 21) In cases such as that of the NGO, their decision not to extend Zhang's funding, seen as a contribution to academic repression, is mitigated, at the least, by the duress-like situation they are under with regard to their own staff in China, and by the very proper 22 Young's argument that complicity requires intent is problematic, considering that some legal orders recognize the possibility of negligent participation in a wrong primarily perpetrated by another. See, for example, s 25 of the German criminal code. 23 Lepora and Goodin, Chapter 1 (Introduction) p. 8. 24 Lepora and Goodin, p 171. 25 Lepora and Goodin,p. 63. sense of responsibility the NGO's directors have towards their own staff. A university colleague of Zhang's in China may express support for the actions against Zhang out of fear for themselves, or a colleague abroad may have failed to protest them out of fear of making everybody's situation worse. These colleagues may have allowed their judgement to be clouded by fear; but their fear is relevant to the way in which they have become complicit.
In the case of the University, the decision to reject Zhang's application could be said to contribute causally to Zhang's immigration status predicament, but absent a prior connection with Zhang, this decision can be called a remote cause, in a sense drawing on how 'remoteness' is used in law, where a claim in negligence can only succeed if the defendant owed the claimants a duty of care and the damage caused was foreseeable as a result of neglecting that duty: if it was not too remote. 26 In that same case, a University-employed academic who has no participation in or even knowledge of the decision to reject Zhang might yet be said to participate in practices of academic research and teaching that overall, as a matter of structure, contribute to Zhang's predicament, although we cannot trace a causal connection between what the academic does and the harm to Zhang. Such an academic, from the perspective of this analysis, would be structurally complicit.
The point of suggesting that complicity with autocratic wrongs can take the different forms suggested here is not to propose a conclusive typology. Rather, it is to account for democratic actors' agency in a way that overcomes the implausible passive-active dichotomy of the 'sharp power' metaphor; moves beyond (as I have argued, inherently limited) analyses of legal liability; and takes on board that even where they have actively contributed to an autocratic wrong, democratic actors are not necessarily to blame. Having tried to establish this point, it remains to consider whether and how democratic actors should respond to the problems of remote, mitigated, or structural complicity.

Democratic actors' political responsibilities
In the previous section, I have argued that democratic actors may have agency in contributing to transnational autocratic wrongs, even though the kinds of harm considered here can be attributed to wrongdoing by autocratic actors. This conclusion raises the further question of whether the democratic agency discussed has any further moral implications. I suggest that we can also draw on Young for a more accurate account of the political responsibilities triggered by democratic actors' complicity in autocratic wrongs -establishing such responsibilities is, in a sense, the point of the social connection model. In Young's account, there is a clear difference between personal, individual guilt and blame as legal-moral concepts related to the causation of wrongs, and political responsibility as an aspect of the fact that, '[b]ecause we dwell on the stage of history, and not simply in our houses, we cannot avoid the imperative to have a relationship with actions and events performed by institutions of our society, often in our name, and with our passive or active support.' (Young 2013, 88) Similarly, we might argue, democratic actors who cannot be blamed or held liable for causally contributing to transnational civil society repression must not use the fact that others are to blame as a shield. Their position does not give them licence to act as though the autocratic wrongs produced through their engagement were not their responsibility at all. 27 Where, on the other hand, blame does attach to the actions of democratic actors through their engagement with autocratic counterparts, there is at least as strong an argument that they must act responsibly in future: 'watching these institutions, monitoring their effects to make sure that they are not grossly harmful, and maintaining organized public spaces where such watching and monitoring can occur and citizens can speak publicly and support one another in their efforts to prevent suffering.' (Young 2013, 88) The contours of the political responsibility to prevent and address transnational repression risks can be outlined on this basis: this responsibility is shared with other actors contributing to structurally unjust processes. It recognizes and opposes the tendency of complicit actors to engage in 'strategies of denial' such as 'culpable ignorance, self-deception, [or] . . . conformity out of fear' -strategies that ring true for some of the actors in the fictitious scenario. (Aragon and Jaggar 2018, 250) An explicit recognition of responsibility for autocratic wrongs is a minimal prerequisite for addressing' democratic actors' potential complicity in these wrongs. In the scenario, such recognition can accommodate the sense that the University, the NGO and the Publisher are structurally or remotely complicit in different ways that affect not only the moral assessment of their past actions, but also their present responsibilities.
In assessing the actors of the scenario, we need to recognize not only that complicity can be proximate and direct or indirect and remote in accordance with degrees of culpable involvement. Fearing for the safety and liberty of a colleague, friend or relative in China, for example, is generally a better reason for self-censorship than the desire to hold on to a 'visiting researcher' position that pays for one's own expensive academic trips and research stays. We would also need to acknowledge that the motivations for acts of complicity can be mixed and entangled: for example, academic employment insecurity can make individuals more justifiably susceptible to financial incentives that lead to self-censorship and (even if in a mitigated way) contribute to others' repression.
Most importantly, recognizing the problem of complicity with academic wrongs would clarify the basis on which democratic societies and institutions should address their institutional, collective, and/or political responsibilities vis-à-vis the autocratic wrongs that occur within them. For actors such as the university and the NGO, we can acknowledge that they are not in a position to prevent the injustices that affect Zhang, Yang, the NGO's Chinese partners and (in generally less dramatic ways) researchers affiliated with the University's China Studies Centre as well as the University's partners in China. Yet we must insist that the University and the NGO have an obligation to investigate, understand and recognize their own connection with these wrongs. Following Young, further, we should realize that it is 'up to us to take specific actions, coordinated with others,' to address those autocratic injustices with which we have relevant connections. (Young 2013, 170) We can try to outline what this means for universities as major democratic actors in engagement with autocracies by way of example. In a better institutional environment for academic engagement and exchange, universities would sign up to codes of conduct governing their transnational engagements. These codes should stipulate clear commitments to academic freedom, as understood under UN frameworks such as UNESCO and the ICESCR, and in self-governance codifications such as including, for example, Magna Carta Universitatum, a charter which stipulates that the university 'must be morally and intellectually independent of all political authority and economic power.' (UNESCO Recommendation concerning the Status of Higher-Education Teaching Personnel 1997; Magna Carta Universitatum) and implemented by committees set up under robust principles of academic self-governance. Publicly available, such codes of conduct would not by themselves necessarily mitigate the risk of complicity, but they would serve as standards that members of the university as well as the wider public could draw on to call for appropriate responses to autocratic wrongs in their midst. 28 Further, more detailed obligations arguably include, for example, transparency about any funding the University takes from autocratic sources, in recognition of the influences such funding creates. 29 There would also need to be transparency requirements regarding the University's terms of engagement, such as for example memoranda of understanding, with other universities in autocratic countries, as a way of minimizing the risk of universities in democratic countries signing away control of their hiring and curricula, for example. (Moody 2020) University-internal committees, ideally strengthened by principles of democratic academic self-governance, would scrutinize autocratic funding sources, such as those for the University's proposed new Confucius Institute, and ensure that commitments to academic freedom were written into the terms of engagement abroad.
As the University has a political (institutional) responsibility to defend academic freedom within its own institution as well as in exchange with others, and to support resistance to academic repression, the University would also have to acknowledge its responsibility to assist victims of academic repression. It could do so, for example, by running 'scholars at risk' programmes with dedicated funding support, a type of programme that could alleviate Zhang's predicaments. 30 The University would also be prepared to be a vocal institutional actor advancing academic freedom, for example, by signing on to statements protesting the persecution of colleagues abroad.
Given their intensifying institutional associations, universities also have special responsibilities and duties of care towards members such as Yang, the PhD student on whom vague and unsubstantiated suspicions have fallen. The recognition that generalized suspicion towards members from autocratic countries would be wrong does not only entail an obligation to avoid discriminatory or hostile language stamping 28 The author has been involved in one such initiative: the UK-based academic freedom and internationalization working group (https://hrc.sas.ac.uk/networks/academic-freedom-and-internationalization-working-group/aboutafiwg). 29 The University may even, as the GPPi has suggested, even have a political responsibility to stop taking such funding. (Benner 2018) 30 Such as, for example, the NGOs Council for At-Risk Academics (CARA, https://www.cara.ngo/) and Scholars At Risk (SAR, https://www.scholarsatrisk.org/). members of the university as (potential) enemies just by virtue of where they are from. Additionally, the University would have to recognize that members exposed to autocratic pressure may be in need of assistance, for instance through counselling or the provision of immigration advice. (Watch 2019a) Lastly, the recognition of an inherently political and collective responsibility on the part of democratic actors to address autocratic wrongs urges us not only to acknowledge that the defence of democratic values cannot come at the expense of democratic rights. For actors such as Sun in the scenario, for example, the availability and use of Magnitsky-like sanctions against him would have to be limited by considerations tracking due process rights and the right to a fair trial. (Paul Tsai Center 2018) We also need to consider the positive responsibilities of the polity towards universities (and similar actors). For example, the problems of autocratic funding result in part from the marketization of higher education and research, especially in countries such as the UK. The University should be made more resilient in the fact of autocratic influencing. One of the responses to the problems discussed here is therefore to provide more public funding to universities and their members, including students, as a way of enabling academia to honour its commitments to academic freedom and thereby also defending the values of a liberal democracy.

Conclusion
Thinking about transnational contributions to autocratic wrongdoing in terms of remote, mitigated and structural complicity can help us address recent proposals and debates about how to deal with practices which the literature has described as 'authoritarian influencing' and 'sharp power' exercise and which include threats and attacks targeting individuals in democratic societies, as well as practices such as selfcensorship to promote autocratic aims. This paper has argued that in order to understand our responsibilities vis-à-vis these wrongs, we need to assess and address the moral position of democratic actors who, without sharing the aims of autocracies, yet contribute to their success in settings of collaboration and exchange with autocracies, thereby becoming complicit, albeit perhaps only in remote, mitigated, or structural ways.
The adoption of a conceptual framework of political responsibility for autocratic wrongs helps accommodate the fact that, due to civil society interconnectedness, repressive practices can more easily become a transnational or global problem than they used to. Understanding the problem of different forms of complicity offers a way out of a binary distinction between victims and perpetrators, which does not work well in repressive environments that rely heavily on fear to produce reasons for action. It also provides a path towards discussing the political responsibilities generated by implication in autocratic wrongs as matters of collective and imperfect moral obligation arising from the need to address deteriorating governance practices in democracies.
If one is able to accept political responsibility for some autocratic wrongs along these lines, one should be particularly wary of invoking binary friend-enemy distinctions and adopting the vocabulary of warfare, for example, by characterizing particular autocratic actors as 'enemies' to be fought or their democratic partners as traitors to be punished. Instead, the argument advanced here favours a cosmopolitan outlook on responsibility for these wrongs and urges the conclusion that even if we regarded certain actors as part of an enemy force, we would still be required to honour our political responsibility towards them.