What is at stake in the information sphere? Anxieties about malign information influence among ordinary Swedes

ABSTRACT

Scholars, states and organisations have warned that authoritarian regimes and other hostile actors are projecting information to inflict harm upon others. Yet, there is little agreement on the nature of this threat. This is mirrored in the plethora of labels in use, ranging from “disinformation” to “sharp power” and “information warfare”. In order to investigate this menace further, we turn our focus to ordinary people’s anxieties, since a better understanding of threat perceptions will also provide a better understanding of the problem. We conducted a comprehensive case study comprising focus group discussions (n: 97) and an extensive survey (n: 2046) among Swedish citizens. We asked: To what extent do people worry about information influence and why? What can this tell us about the nature of this problem or threat? The empirical results suggest that respondents were first and foremost worried about societal cohesion and democracy. They also identified a risk that information influence can undermine trust in societal institutions and the EU. Based on our findings, we suggest that “malign information influence” is an appropriate label to be used in future research. Finally, we propose directions for future systematic research on how malign information infuence is received and processed in different national contexts.

KEYWORDS:

• Information warfare
• Sweden
• malign information influence
• democracy
• national security
• polarisation

Scholars, states and organisations such as the EU and NATO have warned that authoritarian regimes and other hostile actors are using information as a weapon to inflict harm upon others. At the same time, there is little agreement on the nature of this purportedly new threat. In order to investigate this complex menace further, we turn our focus to ordinary people’s anxieties. We argue that a better understanding of threat perceptions will also provide a better understanding of the problem. This article provides empirical results from focus groups (n: 97) and a survey (n: 2046) that explored how ordinary Swedish people perceive information influence activities. We asked the following research questions: To what extent do ordinary people worry about information influence activities, and why? What can this tell us about the nature of the problem? The findings support previous research in that three key societal goods are imperilled: democracy, societal cohesion and national security. Based on these findings, we provide directions for future research and propose that the problem of information influence should be characterised in an inclusive manner to reflect its complexity. The label “malign information influence” should enable comprehensive future research endeavours within disciplines ranging from international relations and security studies, to media and communication research.

Given our focus on the Swedish case, we applied the Swedish Civil Contingencies Agency’s (2019) term “information influence activities” when addressing respondents (see further below for information on the Swedish definition and approach). Nevertheless, our point of departure is the problematic absence of an agreed upon definition and common understanding of the problem among practitioners and scholars alike. While media and communication scholars warn of the undermining of the free flow of information and public sphere (Corner 2017, Bennett and Livingston 2018), scholars from disciplines such as international relations, security studies and strategic studies tend to define the problem in terms of information warfare or hybrid warfare/non-linear warfare and apply a militarised vocabulary, for example referring to the “weaponization” of information (Jonsson and Seely 2015, pp. 1–6, Waltzman 2017, Ramsay and Robertshaw 2018). Yet, the problem is complex and cannot be confined within one disciplinary boundary. Aiming to grasp the problem in its entirety, this article draws upon insights from different disciplines in the pursuit of a better understanding and conceptualisation. Above all, we argue that a bottom-up approach is fruitful in order to get a better view of what is at stake.

The article proceeds as follows. The next section presents the problem from the point of view of states, organizations and scholars and explains the value of exploring citizens’ perceptions. The subsequent section rationalises the choice of Sweden as the case study and provides an overview of the official Swedish approach to the problem. The following section explains the methods used, after which the empirical results are presented, analysed and placed in relation to previous research.

How to grasp the problem with projection of harmful information across borders

The organised and systematic spread of harmful information in the digital sphere is an increasing problem. The Computational Propaganda Research Project have noted a sharp rise in organised social media manipulation campaigns directed by political parties or governmental agencies; observed in 28 countries in 2017, rising to 70 in 2019 (Bradshaw and Howard 2019, p. i, 4). The European Union acknowledged the problem of information influence in 2015 and has since then undertaken a range of countermeasures (European Council 2015, European Commission 2018). NATO and the EU have encouraged member states to participate in the “European Centre for Countering Hybrid Threats” established in Helsinki on the basis of a joint EU/NATO declaration (NATO 2016).

Yet, scholars have so far failed to provide a widely accepted definition of the threat. An arsenal of labels have been used, including strategic communication (Hallahan et al. 2007, Archetti 2019); disinformation (Bennett and Livingston 2018, Ramsay and Robertshaw 2018, Tucker et al. 2018); fake news (Khaldarova and Pantti 2016, Allcott and Gentzkow 2017) and propaganda (Watanabe 2018, Bjola 2019). Christopher Walker’s “sharp power” constitutes yet another attempt at grasping the problem (Walker 2016, 2018). Whereas hard power relies on force and soft power on attraction (Nye 1990), sharp power constitutes an aggressive form of communication that aims to undermine and harm other actors by targeting key institutions (culture, media, academia and publishing), which are particularly vulnerable due to the open nature of democratic societies (Walker 2018, p. 13).

All these attempts at labelling and characterising the problem are however flawed in different ways and do not capture the problem in its entirety. To begin with, information can be heavily skewed and systematically projected with the intent of inflicting harm even when void of fake news or disinformation. In turn, strategic communication, i.e. “ … the purposeful use of communication by an organization to fulfill its mission” (Hallahan et al. 2007, p. 27) is not inherently illegitimate, but can be vital to the functioning of modern society (Wilbur 2017, p. 3). Moreover, modern information influence activities differ from traditional propaganda (Lasswell 1927, p. 627). First, actors no longer rely primarily on secret practices such as infiltration and espionage but can exploit new channels and normal media consumption patterns to reach citizens in other societies, thus potentially harming other states from within. They can use open sources, hire native journalists and even make citizens in other states re-tell their stories through social media (Mejias and Vokuev 2017). Second, social media makes it possible to reach larger audiences and tailor messages to specific audiences in new ways. Third, the projection of narratives has become more central, and narratives can evoke strong feelings, which may weaken an audience’s ability to make a sound judgment on the content of a story (Maan 2018). Finally, sharp power is primarily about the undermining of democracy, which is a narrow reading of the problem.

In lack of an appropriate definition of the problem with the spread of information for antagonistic purposes, it is vital to unpack this problem further. Previous research has sought to provide a clearer picture by studying the threat, focusing directly on practices: cases of disinformation, (antagonistic) strategic narratives and fake news (e.g. Mejias and Vokuev 2017, Ziegler 2018), or the practice of “information warfare” (Baumann 2020) or “hybrid warfare” (Jonson and Seeley 2015, Renz 2016). We instead propose a bottom-up approach, focusing on threat perceptions in a democratic society. We thus attempt to disentangle the problem through scrutinising people’s anxieties. Although we do not know precisely how much substance there is to people’s concerns, they often provide a basis for policymaking. Studying threat perceptions is thus a fruitful way to expose what is at stake in a community in view of a new hazard, and a good way for refining the characterisation of the threat and find a proper definition. The empirical investigation will thus provide a clearer picture of the problem at hand. This, in turn, will enable a more informed discussion on appropriate labels in the concluding discussion.

Furthermore, ordinary people’s reasoning on information influence is also of interest because they are the ones that many countermeasures seek to protect and prepare. There are no standard remedies to this problem and states resort to different, sometimes controversial, solutions that might infringe upon democracy, freedom and security (Hellman and Wagnsson 2017). Inviting ordinary people to express their thoughts on this kind of problem without presenting them with simple solutions yields knowledge that can facilitate an inclusive, nuanced and informed public debate. Awan et al. (2011, p. 104) take radicalisation and climate change as an example of this kind of complex political problems that require the involvement of ordinary people:

Our concern is not simply with whether or not citizens are scared of terrorism or trust government counter-terrorism policies. More important is the way citizens think about issues like radicalization – issues where there is little concrete reliable data in the public domain, issues that may involve thinking about a complex set of further problems (immigration, religion, multiculturalism, policing), and issues where should a problem be established then any response would involve trade-offs of political values like liberty, equality, and security.

In sum, a focus on ordinary people’s anxieties is fruitful to expose what the problem might be. This method will also provide results that can enable and stimulate public deliberation and debate on the nature of the threat and on suitable approaches and countermeasures.

The Swedish case

The investigation is carried out as an exploratory case study. Media cultures and societal settings differ substantially among democracies and European states, and whereas information influence activities from terrorist organisations may be the major concern of some states, disinformation from authoritarian states may cause more alarm elsewhere. Still, European states, including Sweden, have acknowledged a shared challenge and have joined forces through the EU and NATO to combat information influence activities. Thus, the Swedish case can be studied in order to unearth basic concerns that are likely to reoccur in other democratic states. This information can be used to provide focus and structure to future research.

Moreover, Sweden is particularly suitable for a case study since citizens in Nordic countries can be expected to be comparatively skilled in elaborating on the issue. The Nordic states have been presented as particularly apt at resisting information influence activities (US Senate 2018, p. 628) and Nordic governments and authorities have taken the problem seriously and initiated countermeasures (Deverell 2019). The problem has thus received considerable attention and ordinary people are likely to be rather well informed and capable of providing their own views on the complex subject. As stated above, we use the Swedish label “information influence activities” in our focus group discussions and when constructing the survey, since people are likely to be familiar with this term. According to the official definition, this label denotes “[a]ctivities carried out by a foreign power, or antagonistic actors, with the aim of influencing the perceptions, behaviors and decision-making of different target groups” (author’s own translation, Swedish Civil Contingency’s Agency 2019). The official Swedish definition states that information influence activities target a society's vulnerabilities and can threaten the “ … population's lives and health, the functioning of society, our fundamental values such as democracy, the rule of law and human rights and freedoms or other national interests” (Swedish Civil Contingencies Agency 2019).

After having discontinued the agency responsible for psychological defense in 2009, Sweden re-initiated the build-up of a system of defense against information influence activities in 2015 (Deverell 2019, p. 36). This was provoked by Russian use of information operations in connection with the war in Georgia 2008, the annexation of Crimea in 2014 and the Russian involvement in the American election in 2016; but also by the threat from influence from radicalisation/Islamic extremism (Deverell 2019, p. 45). The official approach has been to work broadly with all kinds of problems related to cognitive influence with a bearing upon security (Deverell 2019, p. 46). At the most basic level, this work is about the protection of democracy, yet it also concerns the safeguarding of societal security, state sovereignty and national interests (Deverell 2019, p. 37–38). The authorities undertake both direct measures – such as monitoring Russian strategic narratives – and long-term preventive measures, including education of civil servants and the public. The authorities have also developed cooperation with Finland (Deverell 2019, p. 38). Sweden moreover participates in international cooperative institutions dealing with the problem, including NATO’s StratCom Center of Excellence in Riga and the EU’s European External Action Service, and the Hybrid Centre of Excellence, Europe Centre of Excellence for Countering Hybrid Threats in Helsinki.

Russian information influence activities constitute a key Swedish focus. Sweden’s national security strategy explicitly mentions Russian information influence activities as a threat to national security (Government Offices of Sweden 2017, p. 12). Extremist Islamic propaganda has been another topical issue in the Swedish setting. Information influence activities aiming at radicalisation and extremism have been acknowledged among scholars (e.g. Mahood and Rane 2017) and the recruitment of Swedish youths to ISIS and other terrorist organisations has received wide coverage in Swedish media. Following Donald Trump’s famous reference to disorder in Sweden (“You look at what’s happening last night in Sweden … ”) in a political speech, concerns regarding US information influence activities have also been debated in Swedish media (BBC 2017).

Method

This study is based on a mixed-methods approach. In the initial exploratory stage, focus groups were conducted. Based on the results, a survey was constructed and conducted in order to analyse public knowledge and views about information influence activities. In the initial stage, we wanted to uncover how people reason regarding information influence activities. Examining a similarly complicated phenomenon, Awan et al. (2011, p. 104) argue “[b]y asking not just what citizens think but how they think about radicalization, we open up the processes of deliberation and negotiation–with themselves, and with others– through which citizens engage with complex political problems.” Focus groups provide knowledge on “range not distribution”, thus uncovering a broad spectrum of opinions (range) but not on how frequent they are in societal at large (distribution) (Morrison 1998, pp. 198–203). They can be used in preparation for quantitative studies, to generate hypotheses or to formulate questions for surveys (Morrison 1998, p. 124).

In order to attract participants spread geographically across the country and to reach respondents that are otherwise difficult to recruit (e.g. young people, disabled people), we chose to conduct online focus groups. Participants were recruited by the research agency Novus, based on a representative sample of the Swedish population. In total, 97 people participated (52 women, 45 men). They were evenly spread over the ages 18–75, and the groups were divided into six categories; young (18–29 years), middle-aged (30–49 years) and old (50+) men and young (18–29 years), middle-aged (30–49 years) and old (50+) women. The youngest participant was 18 and the oldest 75 years old. We conducted two series of online focus groups in 2019: six in April, and another six in August. The online focus groups were conducted in Swedish, with questions and replies translated by the authors.

Through providing an informal forum where participants are not communicating face-to-face with each other nor the interviewer, online focus groups also have the advantage of making participants feel more comfortable to speak freely on political topics (Stewart and Shamdasani 2017, p. 49). We asked a small number of open-ended questions, thus offering respondents the opportunity to articulate/formulate what they find problematic in their own words. We allocated plenty of time for people to express their own views with very little guidance, asking questions such as “What do you think of when you hear the expression “information influence activities”; “Do you think that information influence activities is a serious problem? Why?” Interviewees were only presented with the official definition of information influence activities used by the Swedish Agency for Civil Contingencies if it was deemed necessary, for example in case interviewees got stuck in discussing non-antagonistic manners of influencing opinions, such as advertisement and political campaigning.

A content analysis of the results of the first set of interviews indicated the prevalence of three main referent objects among respondents: information influence threatening democracy, societal cohesion and national security. In the second set of interviews, while keeping to the strategy of encouraging an open discussion, we devoted some time to encourage in-depth discussions on these three topics in particular. The questions used in the two sets of interviews are listed in the appendix.

Thanks to the large number of respondents included, we had sufficient material for unearthing a wide range of potential problems connected with information influence activities. The focus groups also provided us with ideas, and questions, which we then used a survey to analyse. In the survey, we asked questions on media habits, trust in different public institutions and news sources, skills regarding evaluating information, knowledge and worries regarding information influence activities, as well as behaviours when it comes to evaluating and sharing information. We also controlled for background variables such as age, gender, political views (what party they voted for in the Swedish 2018 general election) and income. The questions were largely developed out of the findings in the focus groups.

As with the focus groups, the participants in the survey were recruited by the research agency Novus and consisted of a representative sample of the Swedish population. The response rate was 52%, with a total number of 2046 people participating, aged between 18 and 79. The survey was conducted online, with participants receiving a link to the survey by email. There was a slight overrepresentation of younger people, with 56% of those completing the survey aged between 18 and 49, while 44% of respondents were aged between 50 and 79. According to data from Statistics Sweden (2018), 52% of Swedes aged above 18 belonged to the age span of 18 and 49 in 2018, while around 48% were aged 50 and above (Statistics Sweden 2018). University-educated people were also overrepresented in the survey (36%, compared to 28% of the general population according to Statistics Sweden 2019).

Results

Analysis of the results

Many interviewees expressed deep concerns about the possible occurrence of information influence activities as well as about the long-term consequences that such activities can pose to society as a whole. Even though the respondents discussed a wide range of problems that they associated with information influence activities, most of their responses can be grouped into three broad categories centred on a specific reference object, or vulnerability. A close reading suggested that the respondents view information influence activities as threatening:

1. Democracy (by eroding peoples’ beliefs in the existence of a “truth” and their trust in the media, and the possibilities for a decent public debate, and by influencing peoples’ opinion including voting behaviour);

2. Societal cohesion (since information influence activities can generate polarisation and undermine trust in institutions);

3. National security.

These three concerns occurred among respondents in all twelve focus groups, although to varying degrees. The statistical evidence validates that there is a substantial degree of worry about information influence, with a focus on these three domains. First, the results of the survey indicate a high awareness of the label “information influence”: 72% claimed to have heard of the concept. When asked about in what manners information influence activities can constitute a societal problem, 72% of respondents answered that it polarises the population; 69% replied that it weakens our democracy; 68% replied that information influence activities reduce trust in news media; 61% replied that it weakens trust in politicians and/or political institutions; 51% replied that it threatens Swedish security, and 48% replied that it divides/polarizes the EU. Thus, respondents were most of all worried about societal cohesion and democracy, yet they were also concerned about other referent-objects such as trust in institutions, national security and the EU.

A threat to democracy

Many respondents in the focus groups brought up the effects that skewed, biased or false information can have on political institutions and the political system, often making a connection between general elections and information influence activities. One young man, for example, stated that “[i]nformation influence activities can affect the Swedish populace to vote against our own interests if we do not have an accurate depiction of reality” (participant 2, focus group 8), while one older woman argued that “if you are affected by disinformation you will not vote for the party you actually agree with” (participant 1, focus group 11).

A couple of people also told stories about how themselves or someone they know had been affected by information influence activities spread online. One man, for example, described how two acquaintances had “completely changed political views in response to conspiracy theories and submerging themselves in the internet” (participant 3, focus group 10). One young woman told that “[w]hen I take a good look at myself, then probably social media, primarily Facebook, affected for example how I voted in the general election” (participant 1, focus group 7). In the survey, 34% of all respondents stated that they think that they could be affected by information influence activities, while 59% thought that someone close to them could be. This suggests a “third person effect”, with respondents perceiving it to be more likely for other people to be affected (Jang and Kim 2017).

When probed directly about whether information influence activities can affect democracy, a group of young men in focus group 8 noted that while information influence activities are not currently a problem in Sweden, it has undermined democracy elsewhere. However, they argued that a similar situation could, or will, happen in Sweden:

Interviewer: Can information influence affect democracy?

• Eventually (participant 3)

• We see that people are losing confidence in authorities for example. This can lead to fewer people voting. And then the electoral result is less representative, which reduces the confidence in democracy as a system (participant 4)

• It has already weakened democracy. Maybe not as much in Sweden as it has in other countries, but it’s only a question about time before we catch up. (participant 6)

• Absolutely. It is already happening in countries like Russia, China and North Korea, et cetera (participant 8)

• Absolutely! (participant 5)

Moreover, the focus groups also indicated that worries regarding information influence activities are intertwined with concerns regarding inaccurate, or biased reporting. One group of young women (aged 18-29) from focus group 1 voiced such concerns:

• Uninformed people tend to share “news” and provide space for biased media (participant 1);

• That you can no longer know what information to trust (participant 2);

• Even established media can share lies by mistake because the pace is so high today (participant 3);

• I want to believe in a democratic society. Information influence activities direct our opinions and thoughts and influences our values I think (participant 4);

• People stop thinking and lose their own opinion, democracy is put at stake (participant 5);

• As a voter you become biased in a certain way unwillingly without even reflecting upon it (participant 2);

• Manipulation, that’s what it is! (participant 4).

Other respondents also raised some concerns regarding biased, or inaccurate, reporting. One older man, for example, stated that as a “long-time public service-listener, it has occurred to me that Swedish Radio [the public service radio channels], through a systematic choice of perspectives and topics also deals with information influence activities” (participant 1, focus group 12). Thus, problems regarding foreign and antagonistic information influence activities are often intertwined with problems emerging from a distorted domestic public debate. Corner’s review of research (2017, pp. 1102–1103) exposes some of the complexities involved, directing attention to problems of bad journalism, including over-coverage on some topics and under-coverage on other. Feeling constantly manipulated, there is therefore a risk that people might not know whom and what to trust, having lost their capacity to orient themselves and take part of political life. One older woman even stated being “skeptical to most things, and I think that everything is made up” (participant 2, focus group 5).

Overall, these findings mirror media scholars’ warnings of information influence as a threat to free expression briefly accounted for in the introduction. Nonetheless, the survey showed that the general trust in the media remains high in Sweden, which other studies have also confirmed (Andersson et al. 2017, p. 19). In the survey, 53% of respondents declared a large trust in Swedish newspapers, while 69% had large trust in the public service channels. However, among supporters of the right-wing populist party the Sweden Democrats, trust in Swedish media was significantly lower. Among this group, only 29% trusted Swedish newspapers, while 39% had little trust in Swedish newspapers. Moreover, only 33% trusted the public service channels, while almost half (48%) had little trust in the public service channels, with 28% answering that they had “very” little trust. This echoes findings from previous research (Andersson et al. 2017, p. 21).

A threat to societal cohesion

In the focus groups, several respondents argued that information influence activities have the ability to jeopardise the cohesion of society. In the survey, 72% answered information influence activities can constitute a societal problem by polarising the population, making it the largest concern of survey respondents. Polarisation on intense issues complicates policy-making by increasing the possibility for compromise (DiMaggio et al. 1996, p. 693, Archetti 2019, p. 10), which may harm governability. Polarised groups may cultivate ideas that are undesirable from a democratic viewpoint, and polarisation can spur practices of “othering” and societal division that harm possibilities for a decent public debate on timely political issues. Polarisation is used as part of information influence activities, for example as “trolls” are paid to spur divergences online to muster polarisation (Tucker et al. 2018). Fridman (2019 p. 75) suggests that polarisation is a strategy used in Russian hybrid warfare, with the goal to “ … destroy the political cohesion of an adversary from the inside by employing a carefully crafted hybrid of nonmilitary means and methods that amplify political, ideological, economic, and other social polarizations within the adversary’s society, thus leading to its political collapse”. There is evidence that actors, including Russia and the international far right, have tried to fuel polarisation in Europe and the US at large, using (the bad) Sweden as a bad example with the aim of spurring right-wing, anti-liberal, anti-migrant feelings (Colliver et al. 2018).

Among the respondents in the focus groups, there appeared to be a relatively widespread concern that polarisation has increased during the last couple of years and that information influence activities could exacerbate these divisions. One young woman, for example, stated that

I feel like the debate atmosphere has become extreme, you can barely have an opinion before you are called a racist or a leftist. The hatred and the threats increase. Instead of having a healthy dialogue and discuss the issues at hand you are immediately categorized and put into a box. (participant 3, focus group 1)

While most respondents did not necessarily blame information influencing activities for the perceived polarisation in Sweden, a couple of respondents communicated a belief that information influence activities had sowed distrust. One young man was adamant that information influence activities have already polarised people:

Yes, it already has. Look at the issue of climate change. There are many that now believe in/vote for climate change deniers because they “read” that Greta Thunberg is traveling across the Atlantic to make money. And now we are in a position where some want to save the planet and others don’t. (participant 4, focus group 8)

Other respondents worried about the consequences that polarisation could have on society, including worrying about increased antagonism between people and an undermined trust in societal institutions. One older woman stated that “[m]isleading information can easily break down people’s trust for other groups of people as well as for authorities which can create antagonism that others can use to their advantage” (participant 6, focus group 11). One young man moreover argued that there could be an escalated effect, with information influence activities “[d]ividing the people and putting groups against group … which can make people more receptive toward disinformation about the other group” (participant 1, focus group 8). One young man also stated that “polarization means that it, for example, might be more difficult to get a job as the “camps” get increasingly hostile toward each other” (participant 1, focus group 2). Three men aged in their 30s-40s in focus group 10 discussed the potential effects that information influence activities might have on society and their lives:

• I feel worried that it may have consequences on me and my family’s living conditions. Then I do not mean that it will entail a direct conflict in my immediate surroundings, but that the insecurity can affect the economic situation, job opportunities, et cetera (participant 8);

• I feel worried that my son will encounter a hostile atmosphere because false rumours are being spread (participant 1);

• That we do not succeed in dealing with the actual problems in our society and on our planet (participant 3).

It should be observed that Sweden is likely vulnerable to messaging that aims to polarise society and to undermine trust in societal institutions. On issues related to migration, researchers have noted an increase in polarisation in the last ten years. While supporters of left-wing parties have become more in favour of accepting refugees to the country, supporters of right-wing parties have become more opposed (Oscarsson 2017, Bergström et al. 2017). Polarisation regarding migration also seems to be linked with cultural values and perceptions regarding globalisation (Oscarsson 2017). Moreover, while still high, trust in societal institutions has declined since the 1980s (Holmberg and Weibull 2017, p. 48).

A threat to national security

In discussing possible risks and consequences connected to information influence activities, the respondents also brought up and reflected on the possibility of foreign powers trying to weaken Sweden, even eroding national self-determination to facilitate a military invasion. Taking a closer look at the survey results, 75% believed that it was likely that other countries are conducting information influence activities targeting Sweden; while 57% perceived information influence activities to be a major societal issue (only 10% saw it as a small societal issue). Asked about their worries, 43% claimed to be very or rather worried about information influence from other states, whereas 22% stated a low degree of worry or no worry at all. When asked about specific states, great powers such as Russia, China and the US were the most commonly mentioned. Given the large number that believes that other states actively directs information influence campaign at Sweden, this indicates that a majority understand the problem as not only connected to internal societal dynamics, but also to national security.

The following discussion between a group of young women from focus group 1 is a typical example of that a range of actors were brought into discussion:

• I believe China does –. And Russia. Probably the US. And other extremist groups (ISIS) (participant 4);

• Definitely, I immediately think of Trump who said “look at what’s happening in Sweden” and meant immigrants creating an unstable society without having a clue about what is happening (participant 5);

• I’m sure there are. Russia might have an interest in a weakened Sweden, a Sweden that is not part of the EU and the community of the West. Or because instability weakens our ability to vote, and our economy (participant 9).

Although respondents mentioned many different “emitters” of information influence activities, among several respondents, Russia was clearly pointed out as the originator of information influence activities. A middle-aged man for example said:

Russia – is what I feel. How they actively try to sow discord in other countries by inflating conflicts between groups, in order to make one or more EU-countries paralyzed, and by extension make the entire EU paralyzed in the conflict against Russia (participant 1, focus group 10)

Different respondents moreover presented varying motivations regarding why an actor might conduct an information influence campaign versus Sweden and/or other countries. For example, one young man suggested that “Russia [conducts information influence activities] to strengthen their position within world politics” (participant 6, focus group 8), while another man argued that Russia might, among other reasons, conduct information influence operations against Sweden in order to “prevent Sweden from approaching NATO” (participant 1, focus group 10). Other respondents focused on other countries and reasons, such as one young man arguing that “China also does it to strengthen their position” (participant 5, focus group 8). In another group, a middle-aged woman argued that “Sweden has an important geographic location, which makes it important for the USA and other countries” (participant 2, focus group 9).

These responses reflect a widespread scholarly concern with Russia’s use of information in the military domain. Indeed, Russian policy-makers do view information as central to security policy and “hybrid warfare” or “full-spectrum conflict”, i.e. the use of a wide range of means ranging from conventional military ones to “ … clandestine special forces and intelligence operatives, to economic threats, political influence, online and offline information battles, as well as “traditional” subversion” (Jonsson and Seely 2015, p. 6). The Russian Information Security Doctrine of 2016 lists a number of threats in the information sphere, such as the increased use of information “ … for geopolitical goals, goals of a military-political nature contravening international law or for terrorist, extremist, criminal and other unlawful ends detrimental for international security and strategic stability”, and states a range of measures needed to strengthen Russian national interests in the information sphere (Russian Government 2016) and the Military Doctrine of 2014 establishes information as part of modern conflict (Russian Government 2014). Moreover, Swedish citizens’ worries should be seen against the background of state-sponsored Russian media platforms broadcasting about Sweden, and the country’s proximity. Previous research demonstrates that the Russian government supports media projection of negative information about other states (Yablokov 2015, Watanabe 2018, p. 139; Ramsay and Robertshaw 2019). RT and Sputnik are the major state-sponsored media outlets for the projection of influence abroad (Ramsay and Robertshaw 2019, p. 11, Groll 2014). A framing analysis of RT and Sputnik coverage on the UK, US, France, Germany, Sweden, Italy and Ukraine exposed that 81.7% of the articles included a frame relating to political dysfunction (Ramsay and Robertshaw 2018, p. 7). The coverage of Sweden was the narrowest, 88% of the articles including the political dysfunction frame (Ramsay and Robertshaw 2018, p. 80).

Other respondents in the focus group discussions argue that the aim of information influence activities is to destabilise Sweden and other societies in general and that this destabilisation could make the country more vulnerable and less able to resist an armed attack. One young man, for example, argued that “[w]e as a nation can also feel weaker, without the same unity, the same trust in one another. A weaker democracy and a weaker society is easier to attack” (participant 4, focus group 8). The same participant later stated that “if we stop trusting, for example, the Swedish Civil Contingencies Agency or other state agencies that deal with the nation’s security, then we will have difficulties with resisting a military attack”. When asked if information influence activities can be used to divide the population, one older man even responded that “[o]ur willingness to defend [the country] will for example decrease” (participant 1, focus group 12). Still, one middle-aged man was hesitant, arguing that “I have a hard time seeing that information influence activities would be so effective that they would lower the resilience against foreign invasions” (participant 1, focus group 10).

Finally, one clear tendency among several respondents was to acknowledge that while foreign powers and antagonistic actors such as terrorist groups are, or probably are, waging information influence campaigns against Sweden, they could not say who, nor why. The discussion held between three young women in focus group 1 is a good example of this sense of being targeted by an unknown enemy:

• I think it would be extremely naive to believe that we are outside of this, but I do not have any concrete examples right now (participant 7);

• I cannot see why there SHOULD NOT be, but I don’t know about any specific right now (participant 8);

• Difficult to know who would benefit from it but I’m sure there are (participant 6).

Concluding discussion

This article aimed to take a comprehensive grasp of the problem with information influence since a narrow focus can lead to an incomplete understanding and an underestimation of the various kinds of harms that information influence activities can inflict on democratic societies.

Our findings indicate a high level of anxiety among Swedish citizens. Above all, respondents worry about the exacerbation of polarisation within society and the undermining of democracy. It is also notable that 48% acknowledged that information influence divides/polarises the EU.

At the same time, respondents are less prone to isolate causal effects, by identifying who, or what, is causing the problems. Interestingly, respondents are thus remarkably worried despite the lack of a well-defined antagonist. If people become unsure and bewildered about the reliability of information in general, this can harm the efficacy of governmental messaging and undermine the potential for an inclusive, democratic public debate. A general anxiety in society in view of public messaging may result in undesirable adaptions of behaviour; people may shy away from public debate or become distrustful of public service or governmental messaging. This mirrors research positing that disinformation and antagonistic messaging can erode trust in democratic institutions, primarily the press and politics but also potentially in other civil society institutions and the educational system (Hoskins and O’Loughlin 2015, p. 1332, Corner 2017, Sismondo 2017, Bennett and Livingston 2018, pp. 126–127). Corner (2017, pp. 1102–1103) exposes some of the complexities involved, directing attention to problems of bad journalism, including over-coverage on some topics and under-coverage on others. Feeling constantly manipulated, in the end, people might not know whom and what to trust, having lost their capacity to orient themselves and take part in political life. Sismondo (2017) describes how people are seen less as political agents and more as objects for manipulation:

Whether in the echo chambers of social or older media, we might be just as concerned with the (truth-era) power to direct attention as we are with fake news. Both play into instrumental and behaviorist approaches to politics (and other arenas), treating voters as people to be manipulated rather than as people to be convinced.

The results of the empirical analysis have provided valuable insights that are helpful when tackling the difficulties with regard to conceptualisation that were presented above. Although a substantial share of respondents are alert to the possibility that national security might be imperilled, respondents are more concerned with democracy and societal cohesion being at stake. This suggests that scholars should avoid narrowly militarised conceptions such as “information warfare”, which links the problem to traditional notions of security, connote military warfare and a territorialised understanding of security. On the other hand, labels that focus exclusively on the media environment or on the defense of democracy, such as “sharp power”, exclude threats to national security. There is an urgent need to avoid labels that coin the problem either as a military one, or one linked to the undermining of free media and/or democracy. Our focus group discussions suggest that “information influence” might work since it enabled free and unconstrained deliberation. Nevertheless, it remains somewhat unspecified. Like strategic communication, information influence is problematic only when there is an intent to inflict damage in some way, for example by seeking to weaken, undermine or harm someone. We thus propose the addition of “harmful”, “malign” or, as a minimum, “manipulative”. In conclusion, “malign information influence” seems like a functional and appropriate label, which also constitutes our favoured suggestion for use in future research that aims to take a comprehensive grasp of the problem.

In sum, this article has uncovered that ordinary people in Sweden identify three main referent objects as particularly vulnerable to information influence: democracy, societal cohesion and national security. We suggest that further research should aim to expose how widespread and general these concerns are across democratic societies, among political leaders as well as in civil society. There is also a need for more research on how much substance there is to these anxieties. Knowledge on the effects of malign information influence is scant, whereas available data is inconclusive (e.g. Hayes and Guardino 2011; Della Vigna et al. 2014; Tucker et al. 2018; Bail et al. 2019; Lanoszka 2019). We thus need better knowledge about the seriousness of the perceived danger and also about what kinds of damage it can do. How vulnerable are societies, states and organisations such as the EU and NATO, and in what ways?

One suggestion for future research is to monitor media habits and attitudes to expose whether messages spread through antagonistic messaging, bots, sponsored advertising are taking root in society, affecting national identity and interests in the longer term (c.f. Szostek 2017, p. 575). Some important work has been done. Scholars have performed qualitative studies using audio-diaries, interviews and focus groups (Szostek 2016, 2018, Mickiewicz 2017, Crilley et al. 2019) and textual analysis of audience’s comments on social media (Crilley and Chatterje-Doody 2020). Scholars have also exploited quantitative methods including network analysis (Siegel and Tucker 2018, Stukal et al. 2019) and analysis of peoples’ sharing activity on social media sites (Guess et al. 2019). The Center for Social Media and Politics’ (CSMaP) has as one of its objectives to develop new methods for the study of social media data.

This is definitely a burgeoning and very important field of study. Additional studies are required and there is a particular need for systematized research, and in particular, research on how information influence is received and processed in different national contexts and with a larger number of cases. One possible way forward is experimental studies (see for example Munger et al. 2020). There is a need to build on research from psychology, marketing and propaganda studies, in order to understand the mechanisms of how, and to what extent, information influence activities function. Pennycook, Cannon and Rand’s study on prior exposure is one example. Their 2018 study showed that the more times respondents were exposed to a fake news headline (which had featured on Facebook in reality) the more accurate they perceived the headline to be, raising concerns that social media infrastructure facilitates the believability of dis- and misinformation (Pennycook et al. 2018, p. 1874).

Future research should also explore how citizens’ view countermeasures. Although this was not the primary focus of this study, our focus groups provided some clues. Above all, the participants perceived the counteracting of information influence to be a shared task, or a societal effort, and asked for a major focus on education. Information influence activities were not perceived as something that can be easily countered through regulation, with several respondents expressing deep concerns regarding potential side effects, and many viewing free speech as a basic right that cannot be encroached upon. Interestingly, this is largely in line with the traditional approach to psychological defense in Sweden. The traditional approach has been identified as “[r]aising the threshold approach”, which seeks to discourage “information influence activities by, for example, establishing resilient government structures with high legitimacy in society, actively pursuing and punishing the perpetrators, as well as strengthening the population’s vigilance and will to resist” (Pamment et al. 2018, p. 84). Since countering information influence activities is a delicate matter, potentially infringing on the free flow of information and the reputation of democratic states, the pros and cons of countermeasures need to be problematised and analysed more thoroughly and systematically. Scholars for instance warn of exploiting independent international broadcasters, institutions within the cultural and public diplomacy sector and other civil society actors, in (defense against) information warfare, since this can undermine their independence and credibility (O’Loughlin 2015, p. 170, Ördén 2020, p. 431). Future research thus needs to closely monitor and analyse the roles of different actors in dealing with the problem of information influence.

Acknowledgments

The author would like to thank M.A. Isabella Nilsen for valuable research assistance. The author is also very grateful for immensely valuable comments from an anonymous reviewer.

Disclosure statement

The author disclosed receipt of the following financial support for the research of this article: Parts of this research were funded by the Swedish Civil Contingencies Agency as part of the project “Information influence activities: Vulnerabilities and Effects”.

Notes on contributor

Charlotte Wagnsson is Professor of Political Science at the Swedish Defence University. Her research interests include European security, political communication in the security sphere and strategic narratives. She has published her work at Routledge and Manchester University Press and in journals such as Journal of Common Market Studies, New Media and Society, Media, War and Conflict, European Security, Co-operation and Conflict and Journal of European Public Policy.

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Appendix

Questions used in focus group interviews 1–6

Today we talk a lot about information influence, information warfare and disinformation.

What is your spontaneous reaction when you hear about that kind of information influence?

Have you heard about the terms? In what contexts?

What do you feel when you hear about information influence? Why?

Do you believe that there are foreign actors – countries or groups – that conduct information influence campaigns against Sweden and Swedes? Who? Why do you think that they do that?

Do you think that you have been subjected to information influence (activities)?

Why do you think that you have been subjected to information influence (activities)? Can you give an example? Why not?

Have you heard of someone who has suspected that they have been subjected to information influence (activities)?

Can you give an example?

Do you think that you could be affected?

Tell me more about why you believe that/do not believe that you could be affected?

Do you think that you could be affected without knowing?

Tell me more!

Do you have any examples of TV-channels, newspapers or other media channels or forums that you think are used/have been used for information influence activities?

Tell me more, when was that, it connection to what?

Do you view information influence (activities) as a serious societal problem?

Why is it a problem? Why not?

What is it that makes information influence (activities) a problem? Is it the content (true or false, biased or objective) or the sender (evil intentions)? Or something else?

Or is it part of democracy that different groupings strive to influence us citizens with their views?

Is it possible to distinguish information influence (activities) from free, democratic, formation of opinions?

Questions used in focus group interviews 7–12

Today we talk a lot about information influence, information warfare and disinformation.What is your spontaneous reaction when you hear about that kind of information influence?

What do you feel when you hear about information influence? Why?

In what contexts have you experienced information influence (activities)?

Do you think that you, or someone you know, have been subjected to information influence (activities)?

Can information influence (activities) weaken democracy?

Do you view information influence (activities) as a serious societal problem?

Can information influence (activities) split, or polarise, the population?

Do you believe that there are foreign actors – countries or groups – that conduct information influence campaigns against Sweden and Swedes? Who? Why do you think that they do that?

Do you think that information influence (activities) can constitute a military threat to national security? If that is the case, in what manner(s)?

Some people argue that information influence (activities) can be used in preparation for an armed attack, to reduce the resistance within the society. What do you believe? Why?