To trust or not to trust? Structures, practices and discourses of transboundary trust around the Swedish nuclear power plant Barsebäck near Copenhagen

Abstract This article examines the role of transboundary trust throughout the evolution of relations regarding the Swedish Barsebäck nuclear power plant, located opposite Copenhagen on the narrow straits separating Denmark and Sweden, which created cross-border tensions until its closure in 1999/2005. The article examines changes in transboundary trust in and between the actors relevant for decision-making regarding the plant, during the three phases of the conflict. It focuses on three aspects, which shaped the development of transboundary trust: first, the role played by pre-existing structures of cross-border institutional, interpersonal and ideological trust, secondly, practices of trust-building that various actors applied to maintain, enhance or rebuild transboundary trust, and thirdly, the growing discourses of mistrust and even distrust in the institutions advocating the use of nuclear power – regardless on which side of the border they were located. The article argues that changes in contemporary views on nuclear risk and the concomitant politicisation of the issue of nuclear power broadened the range of relevant actors, and contributed to changing trust relations. It suggests a mutually reinforcing interaction between interpersonal, institutional and ideological trust, and finds few differences in discourses about institutional trust between transboundary and domestic authorities. Its findings highlight the importance of productive mistrust for nuclear safety, and thus demonstrate that both trust and mistrust are important for living with, but also for limiting (nuclear) risk.


Introduction
'Is Barseb€ ack in safe hands with the Swedes?', a Danish booklet asked in 1993. Could 'the Swedes' be trusted to safely operate the nuclear power plant located 20 kilometres away from Copenhagen across the € Oresund, the narrow strait separating Denmark and Sweden? Moreover, could they be trusted to honestly communicate any safety problems of the plant, given Sweden's economically important and politically influential nuclear sector? From a Danish perspective Barseb€ ack's safety was a pressing issue: a nuclear meltdown under unfavourable weather conditions would mean a catastrophe for the Danish capitalhome to a large part of the Danish populationand thus put the entire Danish state at risk (Frederiksen 1993, 7). Who can be trusted regarding nuclear power? Is trust more problematic across national borders than within a nation state? What makes nuclear operators and regulators appear trustworthy? What can they do to build trust? And how does this change over time, along with changing (perceptions of) nuclear risks? These are some of the key questions this article addresses based on a close examination of the changing role of trust in the history of the Swedish Barseb€ ack nuclear power plant. The article will explain why the plant was built near the border, why it became so contentious after the two units went online in 1975 and 1977, and finally why it was the first Swedish nuclear plant to be closed down in 1999/2005. At the core of my explanation are three factors: first, the changing (awareness of the) risks of nuclear power, which, secondly, impacted on trust between relevant actors. Thirdly, what also changed was which actors were considered to be relevant, as the nuclear issue politicised: until the early 1970s, decision-making was in the hands of technocratic elites, who trusted each other and in nuclear power. However, with the rise of anti-nuclear movements, civil society's risk perception and trust in nuclear authorities became relevant for decision-making. Thus, its border location, which had been viewed favourably from an economic and technocratic perspective, and perfectly safe, when the siting decision was made, turned the plant into a "transboundary issue" (Kaijser and Meyer 2018a, 10). This eventually soured intergovernmental relations between the two traditionally friendly neighbours.
Barseb€ ack provides an excellent case study regarding trust across borders. While nuclear installations at national borders have tested transboundary trust and neighbourly relations at various national borders in Europe (M€ uller 2013;Kaijser, Meyer, and forthcoming 2021;Pohl 2019;Rubio-Varas, Carvalho, and Torre 2018), Barseb€ ack is unique because it resulted in the actual closure of the plant at the border. This article examines changes in transboundary trust in and between relevant actors, including nuclear industry, utilities, regulatory agencies, governments and politicians, as well as anti-nuclear movements. It focuses on three aspects, which shaped the development of transboundary trust: first, the role of pre-existing structures of crossborder institutional, interpersonal and ideological trust, secondly, practices of trust building that various actors applied to maintain, enhance or rebuild transboundary trust, and thirdly, the discourses of mistrust and even distrust in the institutions advocating the use of nuclear powerregardless on which side of the border they were located. The article also highlights examples of productive mistrust, and thus demonstrates that both trust and mistrust are important for living with, but also for reducing risk.
What do we already know about trust and transboundary trust in particular regarding Barseb€ ack? In recent years, there has been some incipient research on the cross-border conflict about this nuclear power plant, as an example of a 'nuclear installation at the border' (Kaijser, Meyer, and forthcoming 2021, Kaijser and Meyer 2018b, Kaijser and Meyer 2018a. Also its afterlife as a 'post-industrial landscape scar' and the memory of the conflict have been explored (Storm 2014(Storm , 2010Borg and Sannerstedt 2006). However, none of these studies focussed on trust in a systematic manner. In the mid-1990s, however, Ragnar L€ ofstedt (1996bRagnar L€ ofstedt ( , 1996a analysed the conflict over Barseb€ ack with a view to 'fairness across borders' and 'crisis communication' in the aftermath of an incident that led to the temporary closure of the plant in 1992. Using surveys, expert interviews and media analyses, L€ ofstedt assessed trust on both sides of the border vis-a-vis the 'Swedish industry' (including the nuclear sector), nuclear regulators, and the Swedish policymakers, also compared to Danish authorities.
L€ ofstedt's findings suggest that perceptions of institutions' trustworthiness greatly depended on trust-inspiring, reliable behaviour. Trust could be earned: for instance the Swedish regulator's SKI (Statens K€ arnkraftinspektion) behaved reliably and openly provided information, and thus appeared trustworthy even in the eyes of its fiercest Danish critics, the Danish anti-nuclear organization (Organisationen til Oplysning om Atomkraft, Organisation for Nuclear Information, OOA). While the border apparently made a difference for instance in the perception of the trustworthiness of Swedish industry, actors on the other side were not automatically perceived as less trustworthy than domestic institutions. L€ ofstedt draws on evidence at a certain moment in the late phase of the Barseb€ ack conflict. This article, by contrast, will examine the development of trust over time and assess, whether L€ ofstedt's observations hold more generally.
The article is organised as follows. After a brief discussion of the concept of trust and the methodology used, I will examine trust relations during three phases in the development of the Barseb€ ack case against the backdrop of the changes in the politics of nuclear energy in both countries.

Conceptualising and operationalising trust across borders
Theorists of modernity have highlighted that trust is an important prerequisite for modern societies to function (Giddens 1990). It acts as a lubricant for social, economic and political relations by reducing complexity (Luhmann 2014). Trust can be defined as the stance whereby an individualthe trustoranticipates the cooperative and mutually beneficial behaviour of the trustees, of those being trusted (Deutsch 1960;Skinner, Dietz, and Weibel 2014, 208). Trust means to 'believe without knowing' (Lehtonen and De Carlo 2019, 204). Ultimately, trust relies on good faith (Gambetta 1988, 219).
Nevertheless, trustees' apparent trustworthiness depends on previous experience and perceptions. Different 'indicators' have been suggested (Earle and Siegrist 2006, 385): The influential model by Mayer, Davis, and Schoorman (1995, 717-20) specifies three relevant 'factors of perceived trustworthiness', namely, 'ability', i.e. skills and competences, 'benevolence', namely the intention 'to do good' to the trustor, and 'integrity', i.e. that the trustee is viewed to act according to normatively acceptable principles (Mayer, Davis, and Schoorman 1995, 717-20). With a view to nuclear power, we can thus assume that actors are expected not only to be technologically competent, but would also do the utmost to minimise risks and be truthful, transparent and reliable. Research suggests that 'moral' indicators (benevolence, integrity) are most important and impact on the perception of 'performance' indicators (ability) (Earle and Siegrist 2006, 385).
We can analytically distinguish between interpersonal, institutional and ideological trust, with a view to the trustees and objects of trust (Tait 2011, 158;Lehtonen and De Carlo 2019, 206). First, interpersonal or 'personalised' trust relates to people (Chryssochoidis, Strada, and Krystallis 2009, 137). It comes in two shapes and guises: 'particular/ised' trust (Delhey, Newton, and Welzel 2011, 787) relates to concrete persons and groups, who are usually known from face-to-face interactions. By contrast, 'general/ised' trust refers to unspecified others (e.g. 'most people'), or even certain categories of others (Siegrist, Gutscher, and Earle 2005, 147;Lehtonen and De Carlo 2019, 206). Such general trust may not only differ in its 'level' (high or low), but also in 'radius' of how large the circle of trustworthy strangers may be, even across national borders, thus perhaps also to 'the Swedes' (Delhey, Newton, and Welzel 2011, 787;Brewer, Gross, and Vercellotti 2018, 659;Inglehart 1991).
Secondly, institutional trust, 'trust in organizations' (Earle and Siegrist 2006, 383) relates to public or private bodies, authorities, or institutions, such as nuclear regulators or operators in the context of this study. Institutional trust may be diffusewith a view to institutions at a general level, such as 'government', or specific, with a view to the performance of a concrete institution (Lehtonen and De Carlo 2019).
Thirdly, ideological trust refers to larger frameworks of ideas, such as the belief in the legitimacy and reliability of 'meta-institutions' like the state or the market (Tait 2011, 158-9;Lehtonen et al. 2021), or with a view to nuclear energy, 'Promethean' ideas about 'technological fixes' or 'socio-technical imaginaries' (Dryzek 2005, 50;Oelschlaeger 1979;Jasanoff and Kim 2013).
As outlined in the introduction to this special issue, we can distinguish trust from its rejection or absence in two ways: as mistrust or distrust. Trust assumes the predictability of others' actions. It thus allows for the delegation of authority or licence to act on behalf of those who trust. By contrast, mistrust involves uncertainty, but nevertheless some hope for cooperative, well-intentioned behaviour regarding the trustee's course of action. This at least allows for conditional or limited delegation, even if only with certain controls (Lehtonen et al. 2021). Mistrust can thus be productive, in the sense that it encourages 'civic vigilance' and the establishment of checks and balances (Lehtonen and De Carlo 2019, 205). Such mistrust forces experts, technocrats and policy makers or regulators to justify or at least make transparent their actions. With a view to a riskprone technology such as nuclear energy, this should normally contribute to measures to reduce risks (Gambetta 1988;Lenard 2008;Skinner, Dietz, and Weibel 2014;Sumpf 2014;Weisker 2003).
Distrust, however, is much more fundamentally negative, with a 'certainty of no hope' regarding cooperation, good intentions or responsible behaviour. Distrust thus does not allow for any delegation. Instead, distrust actually implies that delegation would be a very dangerous strategy to pursue. Distrust thus leads to resistance, hardened opposition and incredulity (Lenard 2008).
Trust does not emerge in a vacuum. This article will focus on three factors that that shape trust relations: First, there are pre-existing structures or 'antecedents of trust'resulting from long-existing relations of cooperation, as well as 'social relations, ingroup membership, morality, benevolence, integrity, inferred traits and intentions, fairness, and caring' (Earle and Siegrist 2006, 386), which provide a sediment of experiences, expectations that have shaped societal perceptions and discourses for a long time (Tanaka and Kawamata 2014, 204), and eventually lead to a 'belief that others have a particular character and that they are inherently trustworthy'even others across national borders (Rathbun 2018, 693).
Secondly, trustees are usually aware that their behaviour is being evaluated by others. It is rational for them try to maintain or build trust via certain practices, for instance, by close, mutually beneficial cooperation, or by seeking to demonstrate their competence, reliability and responsibility. However, trust is more easily lost than gained, and it requires both sides to engage in a 'reinforcing spiral of cooperation', which eventually will be 'transformative', turning an 'other' 'to partner, to friend' (Rathbun 2018, 692).
Thirdly, whether an actor's behaviour will be viewed as trustworthy remains in the eyes of the beholder, and these perceptions are shaped by public discourses (Jacobs, Craig, and Pelsmaekers 2014). It is through such discourses that societies negotiate trust, also vis-a-vis other nations (Brugger 2015). Hence this study examines contemporary sources regarding the claims made about the trustworthiness of actors, within and across the border.
Methodologically, this article takes a qualitative historical approach, drawing on a diversity of published and unpublished written sources from Denmark and Sweden, including archival materials from the Danish Risø nuclear research establishment and the Danish anti-nuclear organisation Organisationen til Oplysning om Atomkraft (literally: Organisation for Enlightenment about Atomic Power, OOA), deposited at the Danish national archive in Copenhagen, the OOA's magazine 'Atomkraft' and some oral history testimonies (Sommer and Quinlan 2009). Analytically, I will distinguish between (1) structures of trustreflected in pre-existing long-term patterns of cooperation, which form the basis on which actors operate; (2) actors' practices and strategies intended to gain trust; and (3) discourses, in which a variety of actors question or affirm trust. Furthermore, in the course of the analysis I will engage with two hypotheses from the debate in the literature: First, I will examine whether there is a 'positive impact' of mistrust (Lehtonen and De Carlo 2019)-with a view a view to reducing risk. Secondly, I will seek to assess the impact of national borders on trust in (nuclear) institutions. L€ ofstedt's (1996b, 1996a) evidence on this issue remained inconclusive with a view to whether trust in institutions was higher within or beyond the border. While research on national identity and trust suggests that trust within national borders is particularly high (Lenard, Miller, and and 2018;Gustavsson and Stendahl 2020), trust seems more fragile in cross-border cooperation. However, shared group membership strengthens cross-border trust (Efthymiadou 2018). Evidence from cross-border nuclear conflicts in Europe equally points in that direction (Rubio-Varas, Carvalho, and Torre 2018), indicating that national sentiments can be mobilised to undermine trust (M€ uller 2013).

Why Barseb€ ack? Nuclear energy in Denmark and Sweden
The history of the Barseb€ ack power plant is deeply embedded in the policies, public debates and attitudes towards nuclear energy in Denmark and Sweden. Until the early 1970s, policies and debates developed very similarly, but increasingly diverged thereafter. In a first phase, until the early 1970s, in both countries, social democratic governments supported nuclear research as a force of economic modernisation (Kaijser 2019, 240f.;Meyer 2019, 78f.). Swedish and Danish nuclear researchers collaborated closely during the 1960s (Nielsen et al. 1999). However, Sweden moved ahead more quickly than Denmark, and build up its own nuclear industry.
In Sweden, a first commercial nuclear plant was opened in May 1972 (Kaijser 2019, 246f.) The relatively small Danish utilities agreed to purchase electricity from the Swedish utility Sydkraft which started to build the two reactors at Barseb€ ack in 1971 and1973, with the explicit goal not only to provide electricity for the Malm€ o-Lund-Landskrona conurbation but also for export to Denmark (L€ ofstedt 1996a, 137;Nielsen et al. 1998, 285). Danish utilities made public their plans for nuclear power only in the wake of the oil crisis, in late 1973 (Nielsen et al. 1998, 285-9).
In a second phase, from the early 1970s until the mid-1980s, nuclear energy was an issue of major societal conflict in both countries. In Sweden, the critique of nuclear emerged around 1972 in the context of the United Nations Conference on the Human Environment (Ecologist/FoE 1972). Nobel prize winner Hannes Alfv en was the most prominent critic and, quickly found support by the Centre Party, Sweden's second largest party, which won the 1976 elections (Kaijser 2019, 247f). However, the Swedish opposition to nuclear power was not able to stop the well-advanced nuclear programme. They narrowly lost the 1980 referendum.
A Danish anti-nuclear movement only emerged in 1973/74, but it arrived well before nuclear reactors had been built. In January 1974, the new anti-nuclear movement organisation OOA managed to delay the first application for a licence for a nuclear power plant and convince the Danish government to hold a public information campaign instead. The unresolved issue of nuclear waste induced the government on 10 th August 1976 to postpone the decision on nuclear power indefinitely (Nielsen et al. 1998, 284, 9-93). With the Danish decision on hold, Barseb€ ack became the primary target of Danish anti-nuclear activism, involving cross-border cooperation among anti-nuclear activists (Buns 2017; Kaiser and Meyer 2018b). Three days before the Danish government decision, the first 'Nordic March against Nuclear Power' took place in Barseb€ ack on 7 th August 1976, the largest anti-nuclear demonstration in Scandinavia until then.
During the third phase, from roughly the mid-1980s, Sweden and Denmark increasingly diverged with a view to nuclear power. In Denmark, in 1985, the parliament decided to give up any plans for nuclear power. Subsequently, opposition to Barseb€ ack increasingly turned into a consensual issue in Danish society, and a bone of contention with Sweden. After Chernobyl (1986) the Swedish government intended to accelerate the nuclear phase-out, including an early closure of Barseb€ ack, but these plans were rescinded in the early 1990s. Protest by the OOA and the Danish government continued, and eventually contributed to the decision to start the Swedish phase-out at Barseb€ ack (Kaijser 2019).

Transboundary trust: Structures, practices and discourses
Building a power plant on transboundary trust Before nuclear power became an issue of contention, decision-making remained in the hands of technocratic elites from the utilities, nuclear research bodies and government administrations, with governments supportive of nuclear energy and wider populations largely ignorant of nuclear risks (Josephson et al. 2021). Neighbourly relations between the decision-makers were characterised by traditional structures of strong institutional trust, which were backed up by trust-building practices andoccasionally reinforced by productive mistrust.

Building on traditional structures of transboundary trust among utilities
The decision to locate Barseb€ ack on the Swedish shore of the € Oresund right next to border between Sweden and Denmark is a result of traditional structures of institutional trust between Danish and Swedish authorities, built on wider frameworks. Both Sweden and Denmark are hightrust societies, characterised by high levels of trust in institutions, as well as high levels of interpersonal trust (Lenard, Miller, and and 2018, 63). Neighbourly relations between Sweden and Denmark have equally been characterised by trustful relations, particularly in the border region around the € Oresund. Denmark and Sweden are linguistically and culturally close. Shared religious traditions of Protestantism and political traditions with dominant social democratic parties account for similarities in political culture. Both countries are members of the Nordic Council, a forum for informal cooperation and discussion of issues of common concern, including nuclear issues (Holm 1980). All this facilitated trustful relations.
Excellent neighbourly relations between Sweden and Denmark included business and utilities. Since 1915, Swedish and Danish electricity networks have been linked with a first undersea cable. Thus the Danes were able to import inexpensive hydropower from Swedish Sydkraft, whenever necessary, the Danes could supply electricity from their coal-fired power plants to Sweden (Kaijser and Meyer 2018b, 76). Such 'power relations' were not only economically beneficial. Interdependence also strengthened utilities' mutual trust.
Trustful relations between the utilities were regularly reaffirmed at an interpersonal level through social practices. High-level managers from both sides met at least once a year for mutual exchange of information. Such meetings also involved ample opportunities to socialise informally. Managers and engineers were part of the same transnational 'epistemic community' (Haas 1992) of the energy sector. Having gone through the same kind of technical training, they shared worldviews and professional attitudes (Kaijser 1997). All this created high levels of mutual trust across the border. Transboundary institutional trust was backed up with interpersonal trust through socialisation and ideological trust based on shared worldviews, including the expectation that nuclear energy was indispensable in the face of rising electricity demand (Nielsen et al. 1998, 285-9).
Such structures of transboundary trust between Danish and Swedish utilities informed the planning for nuclear power. In 1965 Sydkraft offered its Danish partners a co-ownership of a new plant in southern Sweden, but theDanes only committed to buying 100 MW of Barseb€ ack's electricity output and financing a new sea cable (Nielsen et al. 1998, 285). Other Danish utilities signed import deals with Sydkraft, too (L€ ofstedt 1996a, 137). Hence it made commercial and technical sense to build the plant near the Danish border.
Trustful relations between the Danish and Swedish utilities equally related to sharing of sensitive technological knowledge. When building up its team of engineers from 1968 onwards to prepare for the construction of nuclear plants, the Danish utility Elsam sent the engineers to observe the construction of the first nuclear power plants in Sweden (Nielsen et al. 1998, 287). Exchanging engineers and researchers in nuclear industry between countries was not unusual at the time (Allgeier 2016). Industry used such 'training' for sales promotion, too. Nevertheless, it is suggestive of the close relations with its Swedish counterparts that Elsam sent its engineers to Sweden, rather than to the US and Britainthe technological leaders at the time.
Against the backdrop of these traditions of trustful transboundary cooperation and inter-institutional trust between the Danish and Swedish utilities, conflict about Barseb€ ack seemed initially unlikely to them.

Practices of transboundary Trust-Building: Utilities and regulators
When the decision about the location of the Barseb€ ack plant was taken, the Swedish regulators actively engaged in practices transnational trust-building, communicating the plans to their relevant counterparts in Denmark.
In the late 1960s, not only in Sweden, siting decisions were routinely taken in a technocratic 'decideannouncedefend' manner, without popular participation (Holmberg and Hedberg 2017, 239f.). When selecting Barseb€ ack, the utility acted according to the criteria in use in the late 1960s. With regard to safety, the absence of inhabitants within a radius of five kilometres was deemed sufficient, and the seaside location offered plentiful cooling water (Kaijser and Meyer 2018b, 77). Airplane crashes were discussed, but considered unlikely (Nilsson 1974, 1).
Vis-a-vis the Danish authorities, the Swedes engaged in what can be read as a strategy of openness, clearly intended to build trust, avoid mistrust, and pre-empt diplomatic problems in the future. As early as May 1968, the Delegation for Atomic Issues (Dfa), the nuclear regulator at the time, sent two officials to Copenhagen to consult with the Danish Atomic Energy Commission and the Danish Health authority and informed them of Sydkraft's application for a permission for a nuclear plant at Barseb€ ack. Even though the Danish participants were aware that this would have been the moment when objections could have been raised, they did not. They also did not inform the Danish Ministry of Research at this stage. The minister only learned about these plans in 1969, almost a year after the formal submission of the application. This is suggestive of the high level of cross-border inter-institutional, interpersonal and ideological trust among members of the relevant epistemic communities in regulation and nuclear energy. It also reflects their technocratic attitude that this was a decision for experts to take, rather than politicians (Nielsen 2000).
Trustful relations between nuclear and energy experts were reinforced by a trail of information regarding the application sent by the Swedish authorities towards the Danish authorities. A number of meetings were held between the Danish and Swedish authorities (Nilsson 1974, 2). The Danish government discussed Barseb€ ack briefly in October 1969. However, at no point in time did the Danes object to the location (Kaijser and Meyer 2018b, 78f.;Nilsson 2002). Trustful interpersonal and institutional transboundary relations among experts and authorities and the apparent de-politicisation of the decision might indeed have contributed to avoiding mistrust vis-a-vis the siting decision.

Productive transboundary mistrust among regulators and researchers
However, when receiving the plans for the reactor design submitted by ASEA-Atom, the Danish Atomic Energy Commission, as the relevant regulator at the time, did not blindly trust in the technical expertise of the Swedish. They ordered their own experts at the Risø nuclear research centre to check the technology. Risø researchers found the emergency cooling system to be obsolete as compared to American and West German technologies, and concluded: 'We can at present not say that the Barseb€ ack reactor can satisfy the safety requirements as outlined in this report.' However, they opened the door to improvements: 'Further information may change this conclusion.' (Andersen et al. 1972, 18).
This mistrust was strictly limited to technical details, and surely did not involve ideological distrust vis-a-vis nuclear power. The Danish researchers believed in their Swedish partners' capacity to improve and master the technology. This limited mistrust proved productive for the Swedish side. The Swedes took the critique very seriously, and decided to change the emergency cooling and also the design of the fuel rods (Kaijser and Meyer 2018b, 79f.). Both sides regularly stayed in touch and delivered additional information, which further fostered interpersonal as well as institutional trust across the border. Ahead of Barseb€ ack's going critical, from 1974 onwards they cooperated on local 'preparedness planning' for a potential accident on either side of the border (Forskningscenter Risø 1972, 1976.

Politicising Barseb€ ack: new discourses of mistrust and distrust
From the mid-1970s onwards, in both Denmark and Sweden, societal actors politicised nuclear energy, in the face of a growing awareness of nuclear risks, once the plants at Barseb€ ack had gone online. Critics challenged technocratic authorities' monopoly in nuclear decision making. Ideological trust in nuclear power eroded, as more critical information became available to a broader public. Mistrust and distrust were voiced vis-a-vis the Swedish authorities responsible for the Barseb€ ack plant, as well as vis-a-vis Danish authorities procuring safety, whichat timesresponded by trying to rebuild trust. At the same time, anti-nuclear activists trustfully cooperated across the border. They strongly relied on interpersonal trust, on a basis of shared ideological distrust of nuclear power.

Discourses of institutional mistrust and distrust towards the authorities at home and across the border
Ideological mistrust and growing distrust vis-a-vis nuclear power as such characterised the early critique of the Danish anti-nuclear movement. OOA's early slogan asked citizens: 'Do you feel safe and comfortable with nuclear power?' (OOA 1975), until the organisation started to publish its stickers with the more famous slogan 'Nuclear Power. No, thanks'. In the anti-nuclear discourse, institutional mistrust and increasingly distrust played an important role.
Already before the first Barseb€ ack reactor had gone online, OOA's journal Atomkraft criticised the absence of any comprehensive preparedness planning for a potential nuclear accident, on either side of the border. OOA activists engaged in a discursive strategy of institutional mistrust, demanding action from Danish and the Swedish authorities, whom they accused of 'irresponsible' behaviour (Knudsen and Nielsen 1975, 7). Comparing Swedish and Danish plans, they highlighted the blind spots. The Swedish authorities' plan only covered the area within five kilometres of the reactor, which was largely uninhabited. The Danish preparedness plan, which included the Copenhagen conurbation, apparently just foresaw informing the local population, whichthey claimedso far had been left totally unprepared. No evacuation plans were foreseen. In the eyes of OOA these plans were hardly trust-inspiring, already at the basic level of 'ability', because they contradicted the risk assessments not only of critical experts like Arthur Tamplin and Bj€ orn Gilberg (1975), but even the latest official Swedish and U.S. studies regarding nuclear safety (Industridepartementet 1974;Rasmussen 1975). The OOA activists equally challenged the authorities' 'integrity', and accused them of 'blind trust in "experts", who instinctively distrust any critique of nuclear power'. Indeed, researchers from Denmark's Risø nuclear research centre had apparently attacked Tamplin's and Gilberg's calculations immediately (Knudsen and Nielsen 1975, 7).
On 3 rd December 1976, shortly before Barseb€ ack's second reactor went critical, OOA organised a demonstration and took its critique to Copenhagen city hall and the Swedish embassy. Protesters submitted a resolution to stop the already operating first Barseb€ ack reactor and to cancel the imminent start of the second one. This uncompromising demand showed a fundamental distrust towards nuclear power and its operatorsboth at the ideological and the institutional level, as OOA's four points of critique illustrated: the high population density around the plant, the problem of waste heat discharged into the maritime ecosystem of the € Oresund and the unknown impact on fisheries of such thermal pollution (Walker 1989), the unresolved issue of nuclear waste, and the apparent opposition to nuclear power by a 'majority of the people' in Denmark and Sweden (Nielsen and Haaland 1976, 1). The social democratic member of the Danish Parliament Mogens Camrean early critic of nuclear power within his partysupported in OOA's demands towards the Swedish authorities. Speaking in front of the city hall, Camre contrasted the Danish decision not to move towards nuclear but to 'safer alternatives' instead, with Sweden's nuclear path, and criticised that this contradicted the promises made during the 1976 Swedish election campaign. Camre, nevertheless appealed to the Swedish authorities, highlighting traditions of transnational trust, Danish friendship with and admiration for Sweden, and emphasised that opposition to nuclear power was shared on both sides of the € Oresund (Camre 1976). Camre's statement reflects a certain transboundary mistrust towards the Swedish authorities, with a view to their integrity and benevolence, but some hope for a change in behaviour. By contrast, OOA statements give evidence of more fundamental distrusttowards both Danish and Swedish authorities.

Discourses of domestic distrust: regulators and the Risø research establishment
OOA activists repeatedly challenged institutional trust in two domestic institutions: first, the Danish Ministry of the Environment (Miljøministeriet) and the Environmental Protection Agency (Miljøstyrelsen) as the country's nuclear regulator, and secondly, the Risø Research Establishment as the most important nuclear research institution and a major promoter of nuclear power.
Since 1976 Miljøstyrelsen was responsible for nuclear regulation and radiation protection. Their regulatory experts were based on the premises of Risø, which undermined institutional trust in their independence (Henningsen 2017;Nielsen et al. 1998, 308). Furthermore, OOA criticised the various revisions of the preparedness planning produced by Miljøstyrelsen, which seemed 'politically' motivated. OOA activists questioned the authorities' scientific ability and integrity: apparently, preparedness planners redefined their assumptions, whenever international reports suggested greater risks. OOA also questioned authorities' benevolence, claiming authorities systematically played down risks, notably the so-called 'maximum credible accident' (Radkau 2013) (Sørensen 1982, 12-4). This reflects the OOA's growing institutional distrust towards Danish regulatory authorities.
Institutional distrust was also increasing vis-a-vis the Risø research establishment. Critics highlighted its apparent pro-nuclear bias: Risø representatives had downplayed the problems of nuclear waste (Camre 1976, 3), the risks of Barseb€ ack (Sørensen 1982, 8) and routinely assailed any nuclear critics (Knudsen and Nielsen 1975, 7). Since 1972, Risø's director Allan Roy MacIntosh encouraged Risø researchers to attack anti-nuclear activists, until his successor Erik Ib Schmidt stopped this practice in 1975, and later concluded that by taking a pro-active political stance, Risø had lost much institutional trustfor its integrity and abilityas a scientific institution (Nielsen et al. 1998, 308-10).
Nevertheless, a number of Risø researchers continued to confront OOA, and founded the pronuclear association REO (Real Energy Information) in 1976-1989. They defended Risø's 1977 report on the potential consequences of a Barseb€ ack accident against a critical uptake in Danish newspapers, and suggested that concerns about a Barseb€ ack accident were exaggerated (Korsbech 1983(Korsbech , 1982.

Practices of fostering institutional trust: Health and environmental authorities
Danish public authorities also engaged in attempts of establishing institutional trust. For instance, officials from the health and environmental administration, and from Risø, followed the invitation by OOA's Barseb€ ack-group and participated in public hearings on Barseb€ ack in January 1980(OOA 1980. In a comprehensive attempt to regain institutional trust, after Chernobyl, in 1986, Miljøstyrelsen distributed a leaflet entitled 'On Barseb€ ack-Readiness', thoroughly informing the citizens of the Copenhagen conurbation (Miljøstyrelsen 1986).

Transboundary distrust towards the Swedish authorities
Transboundary institutional distrust towards Swedish authorities was growing among the Danish anti-nuclear activists, as the nuclear accidents of Three Mile Island (TMI) 1979 and Chernobyl 1986 demonstrated that their fears regarding the maximum credible accident could become reality. They demanded the immediate closure of Barseb€ ack with large-scale petition campaigns, collecting some 312,000 signatures after TMI and about 160,000 after Chernobyl. Transboundary institutional distrust hardened in the early 1990s, when the Swedish government reversed its promises concerning an accelerated phase-out of nuclear power (Kaijser and Meyer 2018b, 86-90).

Practices of trustful cooperation in cross-border protest
Since the first common Nordic demonstration in August 1976, OOA's Barseb€ ack group cooperated trustfully with Swedish anti-nuclear activists, organising joint common marches almost every year, including creative protest events (Buns 2017)such as an 'Armada' of sailing boat steering towards Barseb€ ack in 1982, for example. Even where conflicts existed between the Danish and the Swedish side, such as in the late 1970s regarding whether to prioritise the phase-out of Barseb€ ack or prevent further plants from being built, the archival records suggest close and trustful cooperation from the 1970s until the 1990s. Activists shared confidential information, and Danish activists supported the Swedish People's Campaign against Nuclear Power during the referendum campaign in 1980. They encouraged Danes to write letters to the editor of Swedish newspapers, or to public authorities (Mortensen, Christiansen, and Carlsson 1980). Cooperation at the local, transboundary level in the joint opposition to Barseb€ ack was sustained even after the referendum defeat, and exchanges were revived after Chernobyl well into the early 1990s: In 1996, for instance, the Swedish and Danish nature protection associations jointly demanded Barseb€ ack's closure (Naturskyddsf€ oreningen and Danmarks Naturfredningsforening 1996). Transnational cooperation was based on shared ideological distrust towards nuclear power, but also resulted from interpersonal trust built up through continuous exchange.

Closing Barseb€ ack: Diplomatic trust-building
During the third phase, from the 1980s onwards, the conflict about Barseb€ ack increasingly turned into an diplomatic issue between the governments. In order to avoid harming traditional transboundary trust, the Danish and Swedish governments undertook various attempts at trust-building.

Practices of transboundary Trust-Building among governments
The Barseb€ ack issue strained the traditionally close and friendly, trustful diplomatic relations between Denmark and Sweden. After TMI, and OOA's signature campaign against Barseb€ ack, the Danish government contacted the Swedish government concerning the safety of the two reactors. The Swedish government responded with a series of trust-building efforts, setting up an expert committee, which found that the Swedish safety assessments proved sufficient (Reaktors€ akerhetsutredningen 1979). To further enhance the safety of the plant, this committee suggested installing filter chambers at Barseb€ ack. Filled with rubble, they would be able to absorb most radioactivity of gases that could be released during a certain type of nuclear accidents. In fact, in 1981, the Swedish government made the installation of such filter chambers a condition for the continued operation of Barseb€ ack after 1985. This so-far untried technology was clearly intended to build trust vis-a-vis the Danish side and domestically, and may also be viewed as an example of the benefits of mistrust for nuclear safety (Borg and Sannerstedt 2006).
In 1982 the Danish Minister of the Environment expressed his concern about the location of Barseb€ ack. Another committee was set up as a trust-building exercise, consisting of Swedish and Danish members, chaired by the former Norwegian Minister of the Environment. The committee concluded in 1985 that the risk of an accident was low, but that the unfavourable location of the plant should be taken into account in future phase-out plans (Barseb€ ackkommitt en 1985).
Such diplomatic trust-building practice proved less persuasive in the face of Denmark's growing anti-nuclear consensus after the Danish parliament had finally decided to exclude nuclear power from its energy planning in 1985. Already a few days before Chernobyl, the Danish social democratic party brought in a motion for a resolution to formally ask Sweden to close Barseb€ ack. The motion received large-scale support.
Chernobyl reignited the Swedish nuclear debate, and for a short period of time led to a convergence of official views on nuclear issues between the two countries. The Swedish social democratic government enacted legislation in 1988 to start the phase-out of nuclear power in the mid-1990s, including the closure of one of the two Barseb€ ack reactors. However, when the Swedish government rescinded these plans in the early 1990s, Danish trust in the reliability of Swedish policy makers was badly harmed. This was reflected in L€ ofstedt's (1996b) research, and also in the booklet quoted at the outset of this article.
In the context of the Nordic ministers' meetings in the 1990s, the Danish government representatives exerted substantial pressure on their Swedish partners concerning Barseb€ ack. They explicitly demanded early phase-out, because of its problematic location (Weiss 1994). The Danish government trustfully cooperated with OOA, sharing diplomatic documents with them. In 1994, a new Swedish government finally advanced phase-out legislation prioritising Barseb€ ack's closure. In 1997, the Swedish government agreed on the end of Barseb€ ack. Re-establishing trustful neighbourly relations clearly motivated this decision (Kaijser and Meyer 2018b, 86-92).

Conclusions
This article traced the evolution of trust relations among those groups relevant for nuclear decision-making during the history of the Barseb€ ack power plant, in the face of changing risks and risk perceptions, which coincided with the growing politicisation of nuclear power since the mid-1970s. Originally, the plant had been planned and built directly at the border for economic and technical reasons, in order to serve both sides. Taking this decision, without involving the public, Danish and Swedish utilities and nuclear experts cooperated closely, based on preexisting transnational structures of inter-institutional trust, based on transnationally shared ideological trust in nuclear power, and interpersonal trust. Such a high-trust environment probably preempted potential concerns and productive mistrust regarding the location of the plant. At the same time, the siting decision was well in line with the standards of the late 1960s, when it was still considered acceptable to place nuclear plants relatively close to centres of population (IAEA 1967). Nevertheless transnational mistrust regarding specific aspects of the plant's technology served to upgrade the design of the reactors.
From the early 1970s onwards, societal risk assessments changed. Nuclear power was increasingly considered a high-risk technology, and civil society demanded a say in decision-making. This challenged the earlier consensus of ideological trust in nuclear among governments, utilities and nuclear research. With the arrival of anti-nuclear movements, Barseb€ ack's location on the border was increasingly viewed to be posing unacceptable risks to citizens in the neighbouring conurbations. In public discourse, citizens and social movements expressed mistrust and increasingly distrust vis-a-vis government authorities and nuclear researchers, who appeared to downplay nuclear risks. Only slowly did these institutions start engaging in more credible trustbuilding efforts. At the same time, shared anti-nuclear ideological trust, cross-border interpersonal and institutional trust strengthened the transnational cooperation of antinuclear movements.
From the mid-1980s onwards, views concerning nuclear power and Barseb€ ack differed strongly between Denmark and Sweden, as Denmark increasingly became an anti-nuclear nation. Trust-building efforts by the Swedish government largely failed, as ideological trust regarding nuclear power diverged. Moreover, not only the author quoted at the outset of this article questioned Swedish reliability and integrity regarding Barseb€ ack, given the various U-turns on the issue in Swedish politics Also L€ ofstedt (1996b) found this to have harmed cross-border institutional trust.
The findings illustrate how mistrust proved productive to enhance safety: Risø researchers' critique of Barseb€ ack's technology was taken up and helped upgrading the reactor design. The various efforts at improving the safety of Barseb€ ack, including the installation of a filter chamber, were equally a result of the productive power of mistrust and Swedish efforts to rebuild trust. Forced by outside pressure, the Swedish authorities had a close eye on Barseb€ ack's safety, and eventually accepted closing it, for the risk that it posed and that the Danish side found unacceptable, given the low ideological trust in nuclear on the Danish side. Barseb€ ack thus became the first example in Europe, where protest across borders led to such a closurethe controversial French Fessenheim plant on the Rhine on the Franco-German border was only shut down in 2020, however, primarily because of its age (Kaijser, Meyer, and forthcoming 2021).
Finally, the article suggests that there was little difference between domestic and transboundary or transnational institutional trust. Anti-nuclear activists challenged domestic and Swedish institutions alike, for their supposed attempts to downplay nuclear risks. When the Danish authorities took a clearer anti-nuclear and anti-Barseb€ ack stance in the third phase of the conflict, domestic institutional trust improved. This again is suggestive of the importance of shared ideological trust. Apparently, shared views about how actors assess nuclear risks at a certain point in time strongly influence trust and solidaritywithin and beyond borders.