Free speech versus defence of the nation? The media as sources of national insecurity in Ukraine

ABSTRACT

The media can cause concern in the context of national security: they are described as potential tools and targets of attack that can be “weaponised”, and as a space where “information war” is waged. Governments may try to block media that are deemed a security threat, but the rationale for taking such an action deserves careful consideration, given the tension between media restrictions and the democratic principle of free speech. This article scrutinises the security rationale for restrictions imposed by Ukraine on Russian and “pro-Russian” media from 2014. When justifying restrictions, Ukrainian officials highlighted the threat of media content both distorting perceptions of reality and weakening the foundations of Ukrainian nationhood. We, therefore, analyse survey data to investigate whether the use of the banned media was associated with variation in Ukrainian citizens’ perceptions of truth and national values. We find that the use of the banned media was linked to mistaken beliefs about the veracity of news headlines, both true and false; it was also associated with lower support for democracy in Ukraine (a key national constitutional value). This evidence from the Ukrainian case informs our discussion about the media’s impact on national (in)security and rationales for media restrictions in democratic contexts more broadly.

KEYWORDS:

• Ukraine
• Russia
• media
• security
• information war
• propaganda

Introduction

Recent years have seen a growing awareness of national security threats associated with the media (Hellman and Wagnsson 2017, Clack and Johnson 2021). In contemporary media environments, there is intense – sometimes underhand – competition to influence public opinion and shape understandings of reality, competition which traverses national borders. States experience insecurity when hostile actors are observed to deliberately target their domestic audiences with information deemed harmful to national interests, and/or when such information spreads organically to become salient in public discourse.

What makes a media outlet a national security threat? This is an important but contentious question, given that a source of information which some people regard as a threat may be defended by others as a legitimate contributor to political debate (Yanchenko 2023). Democracies must tread a careful line between dealing with media security threats and ensuring freedom of speech. Some media clearly violate laws and regulations (for example, by concealing unlawful funding or deliberately spreading falsehoods) and can be legitimately sanctioned on those grounds. But other media may generate security concerns without obviously breaking any laws if they seem to be fuelling disruptive trends in public opinion or undermining the values of society. Addressing the security concerns associated with such sources presents dilemmas.

Ukraine has been dealing with media-related security concerns for a long time (Szostek 2014, Doroshenko and Lukito 2021). After Russia’s seizure of Crimea and (by proxies) of parts of Donbas in 2014, which was accompanied by large volumes of disinformation in the Russian media, Ukraine began to treat Russian “information war” as an existential threat (Doctrine of Information Security 2017). It restricted access to Russian broadcasters and websites and media content of Russian origin; some domestic Ukrainian news sources were also sanctioned. These measures helped Ukraine to substantially reduce the use of Russian media among its population (USAID-Internews 2020) although insecurity about Russian media influence on Ukrainian audiences remained salient throughout the years leading up to Russia’s full-scale invasion of Ukraine in 2022.

This article investigates whether the use of media banned by Ukraine was associated with (a) lower support for values fundamental to Ukraine’s constitutional order and distinct national identity; and (b) factually inaccurate perceptions of truth/reality. By addressing this research question, our study aims to advance wider debates about the nature of threats which the media can pose to national security and the logic of media restrictions as a countermeasure. We start by reviewing the existing literature on these debates, before outlining Ukraine’s experience of and response to media-related threats between 2014 and 2022. In our empirical analysis that follows, we use survey data to show that regular use of banned media in Ukraine was associated (inter alia) with mistaken beliefs about the veracity of news headlines, both true and false, and lower support for democracy (a key national constitutional value). We conclude by discussing how the Ukrainian case can inform assessments of whether media restrictions make sense as a measure to strengthen the security of democracies.

How can the media threaten national security, and should censorship be deployed in defence? A review of the literature

Research about threats to security arising from the media can be found in the literature of several academic disciplines, as well as the publications of think tanks, governments and non-governmental organisations. As Wagnsson (2020) points out, there is no single approach to identifying or defining media-related threats, and researchers working in this area have used different vocabularies to describe the general problem. Some authors write about “information warfare” in which the media have become “weapons” or “weaponized” (Clack and Johnson 2021). Many of these authors focus on foreign-sponsored disinformation as a threat (Golovchenko et al. 2018), along with covert attempts to manipulate public opinion and electoral choices in the interests of a hostile power (Lukito 2019, Golovchenko et al. 2020). Russian disinformation campaigns and covert information operations have been studied extensively since they became global news in connection with the seizure of Crimea in 2014 and the election of Donald Trump as US president in 2016 (Mejias and Vokuev 2017, Jamieson 2020, Erlich and Garner 2021). Such externally sponsored disinformation and covert, deceptive influence operations are clearly transgressive: they violate the norm of honesty in communication, thus undermining deliberative democracy (McKay and Tenove 2021), as well as the norm of national sovereignty (elections, in particular, are expected to be sovereign affairs with results that reflect domestic, not foreign, interests).

However, it is important to recognise that the media do not have to operate deceitfully to be considered a national security threat. Threats are also associated with strategic narratives (Hellman and Wagnsson 2017) and the promotion of incongruent values and identities. The idea of strategic narratives has become a popular conceptual tool for explaining how states pursue and achieve influence via communication in international relations (Miskimmon et al. 2013). Governments project strategic narratives to international audiences, mostly openly and legitimately, via official statements and the tools of public diplomacy, to shape understandings of the world and the behaviour of other actors in international politics (Miskimmon et al. 2017). Strategic narratives may, to varying extents, contain disinformation, but even factually accurate narratives can be deemed threatening if the values and/or identities which they promote are incongruent with those of states at the “receiving end”. The promotion of values and identities internationally was for a long time studied through the lens of “soft power”, a concept which has connotations of harmlessness. Yet, as Walker (2016) writes, soft power sounds like an ill-fitting label for efforts to discredit and supplant liberal democratic values, efforts which are apparent in narratives promoted internationally by some authoritarian states. He observes that “a real competition over norms has emerged” (Walker 2016, p. 52), with Russia and China publicly promoting civilisational diversity and the preservation of “national traditions” as norms which help to justify their human rights violations. The United States understands these competitor-states to be using the media to attack the “values and institutions that underpin free societies”, and has treated this as a security matter that is far from soft or harmless (United States of America 2017). Walker (2018) and others suggest that the concept of “sharp power” could be used instead of “soft power” to describe the efforts of authoritarian states to exert international influence via the media. But in fact, whenever one state – democratic or authoritarian – tries to change or challenge values and national self-images that are established in the society of another state –democratic or authoritarian – such activity is likely to be perceived as threatening (not “soft”) by those in the target state who wish to preserve the status quo.

The media are thus considered a security threat, firstly, when external actors use them in a dishonest way that undermines the integrity of political discussion, and secondly, when they convey narratives which challenge core established values and national identities that underlie the political status quo. A third type of security threat associated with the media is the risk of negative behavioural and psychological responses among audiences. “Psychological operations’ conducted via the media are said to induce effects among the public including fear and disgust (which potentially deepen polarisation), cognitive exhaustion, low motivation and anxiety (Nisbet and Kamenchuk 2019). Mölder and Shiraev (2021, p. 13) write that “global information warriors” in the domain of psychological warfare strive to generate uncertainty, fear, and irrational responses in the minds of adversaries, proceeding from the principle that “if you affect the minds of the people in an adversarial country, the country should become a weakened opponent”.

The literature about psychological operations/warfare and their effects quite often refers to Soviet theories of influence, centred on the concepts of “active measures” (активные мероприятия) and “reflexive control” (рефлексивное управление). The theory of “reflexive control”, developed in Soviet military literature, postulates that “one side can impose its will on the enemy” and cause enemy decision-makers to make disadvantageous decisions through carefully tailored communications, deceptions and other means (Thomas 2004). “Active measures” is a term that was used by the Soviet KGB to describe covert influence activities directed against international adversaries, often involving forged content, doctored sources and/or inauthentic agents (Rid 2020, p. 9). Russian mastery of reflexive control and active measures has been credited with preventing direct Western intervention during the illegal annexation of Crimea (Doroshenko and Lukito 2021); it is argued that decisive Western action was derailed by Kremlin disinformation which presented Russian servicemen (the “little green men” without insignia) as Crimean locals. Others doubt whether such techniques could have much impact on policymaking (Lanoszka 2019). However, Soviet-era terminology and theories give the impression that Russia has an established method for “controlling” the actions of its adversaries via (dis)information, and this adds to reasons for viewing the media as a potential national security threat – moreover, a threat to the nation’s physical security (its capacity to defend lives and territory), not only to its values, identity and sovereign democratic politics.

The literature thus describes a diverse set of interconnected national security threats linked to the media. The authors writing about “information warfare” tend to conflate these threats, without differentiating systematically between the different aspects of media content that generate insecurity, or the different aspects of national security (“referent objects” in the language of securitisation theory) which the media can threaten. The above discussion has attempted to illustrate that media content may be deemed a threat based on (1) a lack of honesty (disinformation and deceptive communications being the most prominent focus of concern in recent debates), and also (2) incongruent values, identities and national self-images conveyed within narratives and (3) emotive content which produces negative psychological/behavioural effects. Moreover, the discussion has shown that the media are believed to be capable of threatening (a) the integrity of national political debate which democracy needs to function well; (b) national sovereignty in collective decision-making such as elections; (c) shared national values and identities that are foundational to the legitimacy of regimes and institutions (democratic or authoritarian); (d) public morale and (e) the capacity of elites to take decisions which are optimal for the national interest.

What role, if any, should censorship play in countering these complex threats to national security? In an article asking how states can counter Russian information warfare, Hellman and Wagnsson (2017) identify “blocking” as one of four possible policy responses to hostile Russian narratives, a response that was adopted by Lithuania and Latvia, where Russian state broadcasters were banned after 2014. However, the authors note that media blocking as a security measure is likely to face criticism on the grounds that it is “undemocratic” and that hostile senders can find ways to bypass blocking, rendering it ineffective. To date, arguments in the academic literature have tended to emphasise the importance of democracies preserving “normative power” (Wagnsson and Hellman 2018) and “moral authority” (Bjola 2018) when responding to media-related threats. Censorship is not inherently at odds with this principle, but might be considered so if it is disproportionate to the harm it is intended to tackle, or if it is imposed without a sufficient degree of accountability and public scrutiny (Bjola 2018). In general, policy-focused discussions about how democracies should approach media-related threats mostly downplayed censorship relative to other available options, such as “debunking” falsehoods, exposing inauthentic sources and projecting counter-narratives (Bjola and Pamment 2016, Bjola and Pamment 2019). The scale of censorship adopted by Ukraine, described in the next section, was thus somewhat controversial.

Media, national security and censorship in Ukraine

The Russian Federation under President Vladimir Putin poses an existential threat to Ukraine and its people, territory and sovereignty. This has been clear since the invasions of 2014–2015, and blindingly so since the larger invasion of February 2022. In essence, this is rooted in the fact that the Russian leadership does not regard the internationally recognised borders of Ukraine as valid, nor, since 2014, has it regarded the elected government in Kyiv as legitimate. Putin describes the administration of Ukrainian President Volodymyr Zelenskyy as a “gang of drug addicts and neo-Nazis … that have taken the Ukrainian people hostage” (Putin 2022b); he describes Ukraine as “the territories adjacent to us … our [Russia’s] own historical territories”, and his claim at the start of the 2022 invasion that Russian plans “do not include the occupation of Ukrainian territories” simply reflects his view that parts of Ukraine occupied (at the time of writing) by Russian forces were never really Ukrainian in the first place (Putin 2022a).

The media-related threats which Ukraine has faced must thus be understood in the context of Ukraine’s more general situation of insecurity that stemmed from the Moscow regime’s active opposition to Ukraine’s full independence within internationally recognised territorial boundaries. Moscow’s approach to achieving the subordination of Ukraine encompassed military and political elements. On the military side, its intervention in Donbas gave it leverage to make demands about how Ukraine should conduct its internal and external affairs (Åtland 2020). On the political side, the Russian state fostered ties with politicians, political parties and other organisations in Ukraine which promoted aspects of Russia’s agenda (Hurak and D’Anieri 2022). Despite alienating ordinary Ukrainians by its actions in Crimea and Donbas, the Moscow regime continued to wager that “pro-Russian” forces within Ukrainian society might eventually regain political dominance over the “pro-Western”, anti-Russian incumbents. The media were integral to Moscow’s efforts to strengthen the former and weaken the latter.

Russian media outlets based within Russia itself constitute one category of media that the Ukrainian authorities regarded as a major threat from 2014. Until 2014, Russia-based media held a strong presence in Ukraine’s media environment. Most Russia-based media – particularly Russian federal TV channels – were conduits for Russia’s strategic narrative because the Kremlin had, by 2014, consolidated control over the most influential Russian news sources (Vendil Pallin 2017, Knobel 2020). And Russia’s strategic narrative, projected via the Russian media, contained all the features that can (as outlined above) threaten a nation-state’s security: it included large amounts of disinformation (Khaldarova and Pantti 2016); it promoted values and understandings of Ukrainian identity that clashed with the values and national identity promoted by the post-2014 authorities in Kyiv (Mahda and Vodotyka 2021); it was highly emotive (Adamova 2017), conveying endless negativity about Ukraine’s situation and prospects. Russia’s strategic narrative was and remains particularly pernicious in relation to Ukraine’s national identity because it not only presents a “different” discourse about Ukrainian national identity, it discursively erases Ukraine as a distinct nation by trying to reduce the idea of Ukraine to “a set of Slavic folklore markers” (Mahda and Vodotyka 2021, p. 358). The administration of President Viktor Yanukovych (2010–2014) was fairly accommodating towards the idea of Ukrainian identity as subordinate to (not distinct from) a wider Russian identity; it did not fundamentally threaten the preferences of Yanukovych or his circle, who, therefore, accepted the Russian media’s presence in Ukraine. However, Russian denials of Ukraine’s distinct political identity were anathema to the administrations of Presidents Poroshenko (2014–2019) and Zelenskyy (2019–present), whose policies were firmly oriented towards shifting Ukraine out of Russia’s political orbit and into European and trans-Atlantic alliances.

Russia’s domestic media were not, however, the only category of media deemed threatening to Ukraine. There were also concerns about media based in Ukraine, which named Ukrainian but “pro-Russian” owners and were rumoured to benefit from Russian funding. In 2018–2019 Taras Kozak, a member of parliament from Ukraine’s Opposition Platform – For Life (OP-ZZh) party, became the official owner of three news TV channels, 112 Ukraine, NewsOne and ZIK (LB 2019). Kozak was an ally of Viktor Medvedchuk, “one of Russia’s most consistent and most powerful supporters in Ukraine” (Hurak and D’Anieri 2022, p. 126). The Kozak channels helped to promote OP-ZZh prior to the 2019 elections and echoed talking points from the Russian media (Burkovskyy 2019). Similar security concerns were associated with NASH, a news TV channel owned by Ukrainian politician Yevhen Murayev. Murayev had been a member of the For Life party until 2018, when he formed a new party, Nashi; the NASH news channel was launched around the same time as the party. Murayev was no ally to Medvedchuk – more of a rival in fact, targeting the same voters – but NASH featured some similar guests and similar anti-Western talking points as the Kozak channels (Rybak and Kravchenko 2021).

However, the complicating factor with these “pro-Russian” Ukrainian media outlets is that they could also be described as “opposition” media. The voices to which they gave a platform were often “pro-Russian” voices from Ukraine’s parliamentary opposition, for whom millions of Ukrainian citizens had voted. The fact that sections of Ukraine’s population voted for “pro-Russian” politicians and sympathised with views expressed in the Russian and “pro-Russian” media is an important part of the explanation for why those media threatened Ukraine’s national security, but it is also the knottiest part of the issue. Statistically, people in certain regions of Ukraine (the south and east) have been more likely than average to express political views which align in some ways with views expressed by the Russian leadership (Barrington 2022). Such people are often described as “pro-Russian”, but the “pro-Russian” label (which we place in scare quotes throughout this article) can be analytically unhelpful because it is reductive: it obscures the different forms that “pro-Russian” attitudes and sentiment can take in Ukraine. For example, a Ukrainian might prefer the continued use of the Russian language in schools (opposing recent Ukrainian language policies) without admiring Putin; or they might believe it necessary to preserve a close economic relationship between Russia and Ukraine, without wanting to see Ukraine subordinate to Kremlin rule. “Pro-Russian” views in post-2014 Ukraine ranged from outright pro-Putinism, which was rare, to a relatively common pragmatic wish for better bilateral relations and/or a desire to accept aspects of Russian culture (the Russian language, Russian Orthodoxy, Russian interpretations of history) as non-threatening to Ukraine.

Both Russian and “pro-Russian” Ukrainian media found audiences among Ukrainians with these kinds of views. Our own survey data, as well as previous research (Szostek 2018) suggest that few Ukrainians relied exclusively on Russian or “pro-Russian” media for news – in other words, pro-Russian “echo chambers” which excluded the Ukrainian mainstream media were rare. It was more common for Ukrainians who felt disillusioned with the government in Kyiv and sympathised to some extent with Russia’s narrative to get news from “both sides” in the belief that neither side could be trusted and the truth had to be located “in between”. Yet, any engagement with Russian and “pro-Russian” media was regarded as a security problem after the events of 2014 when the complicity – or at least passivity – of some “pro-Russian” locals facilitated Russia’s seizure of Ukrainian territory.

In this context, the Ukrainian leadership moved to restrict access to Russian and “pro-Russian” sources. The main Russian state TV channels were banned from Ukrainian cable networks in March 2014, soon after Russia’s seizure of Crimea, in a series of pronouncements by Ukraine’s National Security and Defence Council (NSDCU 2014) and media regulator, the National Council of Television and Radio Broadcasting (Natsrada 2014, ZN 2014). Dozens of other Russian channels (state- and privately owned) were banned subsequently after Ukraine amended its law “On television and radio broadcasting” (2018). The amendment stipulated that TV channels originating in non-EU states, or states that had not ratified the European Convention on Transfrontier Television, could only broadcast in Ukraine following inclusion in a list compiled by Natsrada which certified their compliance with Ukrainian legislation; most Russian channels were deemed non-compliant (Decision 2019).

Steps to block access to Russian online media came later, in 2017, when sanctions were imposed on major Russian social media platforms and online services, including VK, Odnoklassniki, Yandex and Mail.ru (Decree 2017). In 2018, more Russian online media, as well as those of the so-called “people’s republics” in Donbas, were blocked in a similar manner (Decree 2018). Altogether 189 websites were blocked in the first round of sanctions that affected online media (Dvorovyy 2021). This policy was sustained after the election of Zelenskyy as Ukrainian president in 2019. By 2021, over 600 websites were blocked in Ukraine (Dvorovyy 2021).

The sanctions mechanism that had initially been created to address threats from external actors was further employed under Zelenskyy to restrict Ukrainian-owned media (Rybak 2021). In 2021, following a recommendation from the National Security and Defence Council, Zelenskyy imposed sanctions via decree on MP Taras Kozak, whose TV channels were forced off the air (Decree 2021). Several more Ukrainian-owned media associated with “pro-Russian” content, including TV channels NASH, Pershy Nezalezhnyi, UkrLive and news website strana.ua (Detektor Media 2021a, Detektor Media 2021b, BBC Ukrayina 2022), were subsequently banned as well.

What arguments did the Ukrainian authorities use to justify these measures? Yanchenko (2023) tackles this question in relation to the Ukrainian-owned channels that were sanctioned, finding that the channels were seen as overstepping the boundaries of legitimate debate on matters including Ukraine’s sovereignty. Our own review of official statements which accompanied the media restrictions found an evolving range of security-based arguments. Initial media restrictions were justified principally with reference to content that violated Ukrainian laws. For example, Natsrada said it had monitored Russian channels disseminating untruthful information about the situation in Ukraine, calls for the violent overthrow of the constitutional order, propaganda of war and incitement of hatred; such content was declared to violate Articles 2 and 28 of the Law “On information” (which prohibits calls to overthrow the constitutional order), Article 6 of the Law “On television and radio” (which also prohibits inter alia calls to overthrow the constitutional order) and article 7 of the European Convention (which requires broadcasters to ensure that news “fairly presents facts and events”) (Detektor Media 2014). In a statement to international journalist organisations, Ukraine’s parliament similarly highlighted examples of “untruthful, incomplete and biased information about Ukraine” spread by the Russian media, and it raised concerns about specific negative effects: the statement asserted that such content was aimed at “demoralizing the population of Ukraine and the personnel of the Ukrainian Armed Forces, inciting them to state treason and siding with the adversary, and giving citizens of Ukraine and Russia a distorted view of events” (Appeal 2014). The law “On sanctions” which the parliament adopted in 2014 reflects similar concerns for the country’s physical security (it permits sanctions, inter alia, on those who offer “informational assistance” to armed aggression against Ukraine). But notably, it also reflects concerns about Ukraine’s ontological security – its capacity to maintain a stable “self or identity, the subjective sense of who one is” (Mitzen and Larson 2017). The law’s clause on sanctionable “informational assistance” includes inciting hatred towards Ukraine’s people, culture, state language and national identity; and distorting the idea of the Ukrainian people’s distinctiveness (самобутність) (Law of Ukraine 2022). Elsewhere, the need for online media restrictions was also justified by the need to minimise cyber risks, and to protect Ukrainian personal data from access by the Russian security services (Detektor Media 2020).

Often missing from official Ukrainian statements was an evidence-based explanation for why particular media and online services had been banned (Burdyha 2021). The lack of transparency and procedural safeguards around this issue attracted criticism: some international organisations argued that Ukraine’s approach “endangered media freedom” (OSCE 2021), and denied the Ukrainian people’s “right to information and freedom of expression” (RSF 2017). However, Ukraine’s leadership continued to depict the media restrictions as a wartime necessity. After sanctioning Kozak, President Zelenskyy was quoted saying that his three banned TV channels had maintained “talking armies”, which “professionally lied to and zombified” the Ukrainian public (Chernysh 2022). The risks to Ukraine of failing to block such media were deemed greater than any possible weakening of Ukrainian democracy that might ensue from blocking them.

How reasonable was this position? As the above discussion has shown, the restrictions were intended to counteract threats to multiple, overlapping aspects of Ukraine’s security (physical, ontological and cyber) that arose from multiple feared media effects (demoralising, misleading, diluting national identity and allegiance) among different audiences (civilian and military). The task of evaluating the security rationale for media restrictions is complicated by the multidimensional nature of the threat, the difficulty of estimating the various feared effects on different audiences, plus the difficulty of tracing responses to bans themselves (which may include “backlash”). However, recent research provides some relevant evidence about the pieces of this puzzle. One notable study found that Russian TV channels were capable of increasing electoral support for pro-Russian political candidates and parties, by strengthening the attitudes of voters who already had pro-Russian views (Peisakhin and Rozenas 2018). Another study investigated the extent to which pro-Kremlin disinformation was believed in Ukraine, finding that while most Ukrainians could distinguish between true news and disinformation, they had trouble recognising some economic disinformation as untrue (Erlich and Garner 2021). Watching Russian television was found to be a strong predictor of refusing to acknowledge the responsibility of Russia or the Donbas militants for the MH17 plane tragedy (Toal and O’Loughlin 2018). Studies conducted outside Ukraine suggest that Russian propaganda is capable of fuelling negative perceptions of Ukraine (Fisher 2020), and that users of Russian media outside Russia tend towards disillusionment with the state of affairs in their country (Wagnsson 2022). Thus, existing research gives grounds to believe that Russian media were indeed likely to be having some of the feared effects on public opinion in Ukraine – misleading citizens on matters of fact, and demoralising them about the state of Ukraine. Fewer studies have been conducted into the effect of bans as a countermeasure, but Golovchenko’s (2022) work suggests that Ukraine’s ban on Vkontakte was successful at substantially reducing activity on the platform, among “pro-Russian” and “pro-Ukrainian” users alike.

In our empirical analysis that follows, we extend the available evidence regarding media-related threats to Ukraine. Whereas most previous studies have investigated the use of Russian TV channels as a predictor of problematic beliefs, we also investigate the use of Russian social media platforms and the use of the “pro-Russian” Ukrainian TV channels that were sanctioned on security grounds. Moreover, while previous studies have predominantly focused on belief in disinformation as the outcome of concern, we look additionally at support for values that are relevant to Ukrainian national cohesion. Based on the theories and Ukrainian political discourse reviewed above, we expect an association to exist between the use of the banned media and (1) lower support for values fundamental to Ukraine’s constitutional order and distinct national identity; as well as (2) factually inaccurate perceptions of truth/reality. These expectations are tested in the following section.

Values and perceptions of truth among users of Ukraine’s banned media

In our analysis, we use data from a survey conducted in September–October 2021. The survey was commissioned from Ukrainian company InfoSapiens as part of a larger project comparing the understanding and practice of “good citizenship” across Ukrainian regions (see Szostek and Orlova 2022). Rather than gathering responses across the whole of Ukraine, the survey focused on four regions in different parts of the country: Odesa Region in the south, Sumy Region in the northeast, Zakarpattia Region in the west and Kyiv Region together with the city of Kyiv – the political centre. The total survey sample size was 4160, with 1040 respondents interviewed in each region (note that Kyiv and Kyiv Region are treated as a single region throughout this analysis). Survey interviews were conducted face-to-face, by the CAPI (computer-assisted personal interview) method. The overall response rate for the sample of 4160 was calculated as 51%. This research received approval from the University of Glasgow Research Ethics Committee on 30 November 2020, application number 400200038.

We study 11 dependent variables: five relating to values, and six relating to truth perceptions. The first value we look at is support for democracy. Our survey respondents were asked: “Does Ukraine need democracy, in your view?” to which they could answer: “Yes”, “No” or “Hard to say”. We derive the other four value-variables from a question asking: “In your view, what should children be taught to become good citizens of Ukraine?” The list of possible answers included, inter alia, “Good knowledge of Ukraine’s history and culture”; “Good knowledge of the Ukrainian language”; “Good knowledge of the Russian language”; “Patriotism and pride in one’s country”; respondents rated each of these as “necessary”, “desirable” or “unnecessary”. This question about “good citizenship” was developed following focus groups in which the concept of “good citizenship” was discussed in depth, and reflected ideas voiced by focus group participants (Szostek and Orlova 2022). Our choice of values to study as dependent variables here is limited and to some extent determined by the questions included in our survey. But we believe the value variables we study here are all relevant to Ukraine’s ontological and physical security. Support for democracy is fundamental to Ukraine’s constitutional order (Article 1 of the Constitution establishes that Ukraine is a democratic state). Support for instilling children with knowledge of Ukraine’s history and culture, the Ukrainian language and patriotism indicates support for distinct Ukrainian national identity and statehood. Meanwhile, support for teaching children good knowledge of the Russian language as part of good Ukrainian citizenship can be interpreted as challenging the “de-Russified” vision of national identity promoted by Ukraine’s authorities in recent years. Of course, millions of Ukrainians still speak Russian every day. The Constitution (Article 10) commits the state to guarantee free development and use of Russian, as the language of one of Ukraine’s “ethnic minorities”, alongside a commitment to ensure the development of Ukrainian as the state language in all spheres of public life. However, the status of the Russian language in Ukraine has long been a wedge issue and heavily securitised (Maksimovtsova 2020). For the post-2014 Ukrainian leadership, boosting the use of Ukrainian over Russian in education was considered an essential defence against Russia’s efforts to erase Ukraine’s distinctive identity, while Russian state propaganda portrayed those same efforts as “discrimination against Russian-speakers” (and ultimately as an excuse “legitimizing” Russian support for Donbas separatism and military intervention). We study the value placed on Russian-language knowledge within Ukrainian citizenship to understand whether the use of the banned media was linked to polarisation on this contentious issue.

As for truth perceptions, we derive our dependent variables from a question about news headlines, which our survey respondents were asked to rate as “true”, “probably true”, “probably false” or “false”. Three of the headlines were, in fact, true: (1) “Ukraine has risen in the UN human development rankings”; (2) “Poland has suffered a major cyberattack from Russia” and (3) “Medics in Kharkiv Region are protesting about wage delays”. Three other headlines were untrue (and can be considered typical examples of disinformation found in Russian media): (1) “Ukrainian homosexuals are attacking Donbas”; (2) “Soros is planning to stir up colour revolution in Hungary” and (3) “In Kharkiv Region they want to sack a pregnant woman for criticising Ukrainianization”. All these headlines were taken verbatim from Ukrainian websites.

The main independent variables in our analysis relate to the use of banned media. Our media use variables derive from two open-ended survey questions: “On which specific TV channels do you watch the news in a typical week?” and “From which social networks and messaging apps do you get news in a typical week?” From these questions, we derive three variables. Two are dichotomous: use/non-use of Russian federal TV channels, and use/non-use of Russian social networks (VK and Odnoklassniki). The other variable is continuous on a 0–1 scale: it is calculated as the proportion of all Ukrainian TV channels mentioned by the respondent that had been banned at the time of the survey (the five banned channels are 112, NewsOne, ZIK and PNK, linked to Kozak/Medvedchuk and NASH linked to Murayev). The fact that the banned media were still being used by some people, even after being banned, reflects the fact that most media restrictions can be circumvented one way or another: for example, Russian federal channels were available via satellite, while VPNs allowed access to blocked websites.

In addition to the media use variables, we include the following independent variables as controls: sex (male/female); age (in years); region (Odesa/Sumy/Zakarpattia/Kyiv); education level (secondary/secondary-specialised/higher-graduate/higher-postgraduate); Russian natsionalnist (dichotomous); Ukrainian natsionalnist (dichotomous);¹ fluency in Ukrainian (0-–12 scale constructed from a series of questions about speaking, listening and reading abilities) and fluency in Russian (0–12 scale constructed from a series of questions about speaking, listening and reading abilities).

We fit proportional odds logistic regression models to the data, treating each of our dependent variables as ordinal. These models estimate the probability of being in one category (or lower) versus being in categories above it. The coefficients in such models can be hard to interpret because they are on the log-odds scale. We, therefore, follow the convention of converting the coefficients to proportional odds ratios, which we report with their 95% confidence intervals; if the confidence intervals around the odds ratio for a particular variable do not encompass 1 (i.e. even odds) we can interpret that as evidence that the variable has a statistically significant effect on the outcome. We have 11 models altogether, one for each of our 11 dependent variables, and these models are presented in full in the supplemental file.

Given the number of models and space constraints, we are selective in the effects we visualise. We summarise the results of each model in turn below, but provide visualisations only to illustrate significant effects (where they exist) of our media use variables upon the outcome variables. We visualise the effects by creating stacked area plots with the “effects” package in R. Each plot shows the predicted probability of different responses when media use varies, but all other independent variables in the model are held constant at their mean.

Model 1 estimates the relationship between banned media use and support for democracy in Ukraine. The use of Russian federal TV channels was found to reduce the probability of answering “yes” to this question (see Figure 1). However, the use of Russian social media and the proportion of banned Ukrainian TV channels mentioned among weekly news sources did not have significant statistical effects.

Figure 1. Russian federal TV use and support for democracy.

Model 2 estimates the relationship between banned media use and support for teaching “Good knowledge of Ukraine’s history and culture” as an element of good citizenship. None of the media use variables were found to have significant statistical effects on responses to this question, and support for this citizenship value was, in fact, almost universal among respondents.

Model 3 estimates the relationship between banned media use and support for teaching “Good knowledge of the Ukrainian language” as an element of good citizenship. Again, none of the media use variables were found to have significant statistical effects on responses to this question, and support for this citizenship value was again almost universal.

Model 4 estimates the relationship between banned media use and support for teaching “Good knowledge of the Russian language” as an element of good citizenship. There was a much more substantial split in opinion on this question than on the other values-related questions, and two of the three media use variables were found to have significant statistical effects: using Russian federal TV channels (see Figure 2) and using a higher proportion of the banned Ukrainian TV channels (see Figure 3) were associated with a substantially higher probability of answering “essential” to this question.

Figure 2. Russian federal TV use and support for teaching good knowledge of the Russian language as an element of good citizenship.

Figure 3. Use of banned Ukrainian TV channels and support for teaching good knowledge of the Russian language as an element of good citizenship.

Model 5 estimates the relationship between banned media use and support for teaching “patriotism and pride in one’s country” as an element of good citizenship. In this model, the proportion of banned Ukrainian TV channels used had a statistically significant negative effect on the probability of answering “essential” (see Figure 4). The other two media use variables were not significant.

Figure 4. Use of banned Ukrainian TV channels and support for teaching patriotism and pride in one's country as an element of good citizenship.

The remaining models relate to the credibility of news headlines. Model 6 estimates the relationship between banned media use and belief in a false headline, “Ukrainian homosexuals are attacking Donbas”. Belief in this disinformation was very low overall, but two of the three media use variables did have statistically significant effects: the use of the banned Ukrainian TV channels (Figure 5) and the use of Russian social media (Figure 6) both increased the probability of a respondent thinking the headline was “true” or “probably true”.

Figure 5. Use of banned Ukrainian TV channels and belief in the false headline “Ukrainian homosexuals are attacking Donbas”.

Figure 6. Use of Russian social media and belief in the false headline “Ukrainian homosexuals are attacking Donbas”.

Model 7 estimates the relationship between banned media use and belief in another false headline, “Soros is planning to stir up a colour revolution in Hungary”. Again, two of the three media use variables had statistically significant effects, and again, the use of the banned Ukrainian TV channels (Figure 7) and the use of Russian social media (Figure 8) were associated with a higher probability of believing the headline was true or probably true.

Figure 7. Use of banned Ukrainian TV channels and belief in the false headline “Soros is planning to stir up colour revolution in Hungary”.

Figure 8. Use of Russian social media and belief in the false headline “Soros is planning to stir up colour revolution in Hungary”.

Model 8 estimates the relationship between banned media use and belief in a third false headline, “In Kharkiv Region they want to sack a pregnant woman for criticising Ukrainianization”. In this model, none of the media use variables had significant statistical effects on belief in the headline.

Model 9 estimates the relationship between banned media use and belief in a true headline, “Poland has suffered a major cyberattack from Russia”. In this model, using a higher proportion of banned Ukrainian TV channels was associated with a higher probability of thinking this true headline was false (see Figure 9). However, the use of Russian federal TV channels and Russian social media had no significant statistical effects.

Figure 9. Use of banned Ukrainian TV channels and belief in the true headline “Poland has suffered a major cyberattack from Russia”.

Model 10 estimates the relationship between banned media use and belief in another true headline, “Ukraine has risen in the UN development rankings”. Using Russian federal TV channels (Figure 10) and a higher proportion of banned Ukrainian TV channels (Figure 11) were both associated with a higher probability of rejecting the truth of this headline, although the use of Russian social media had no significant statistical effect.

Figure 10. Use of Russian federal TV channels and belief in the true headline “Ukraine has risen in the UN development rankings”.

Figure 11. Use of banned Ukrainian TV channels and belief in the true headline “Ukraine has risen in the UN development rankings”.

Model 11 estimates the relationship between banned media use and belief in a third true headline, “Medics in Kharkiv Region are protesting about wage delays”. In this model, the media use variables had no significant statistical effects.

Summing up, we find the use of various banned media to be associated with lower support for democracy in Ukraine, lower support for teaching patriotism and pride in one’s country and higher support for teaching knowledge of the Russian language as part of Ukrainian citizenship. We find no relationship between banned media use and support for teaching Ukrainian history and culture or the Ukrainian language. We find the use of banned media to be associated with greater belief in two out of three false headlines (about “gays attacking Donbas” and liberal billionaire-philanthropist George Soros “stirring up a colour revolution”) and greater disbelief in two out of three true headlines (about a Ukrainian success and a Russian attack). The implications of these results are discussed in the next section.

Discussion

The analysis above has limitations: most notably, it does not allow us to draw firm conclusions about causal direction. For example, we cannot be certain whether watching Russian TV was causing Ukrainians to doubt the value of democracy, or whether Ukrainians already inclined to doubt the value of democracy for different reasons were more inclined than others to watch Russian TV. However, our results do indicate that the use of the media banned by Ukraine was associated with variations in values and truth perceptions among the public that were problematic for state security.

Ukraine defines itself, in the first article of its constitution, as a democracy, so the scepticism about the need for democracy in Ukraine associated with watching Russian TV can be read as an ontological threat to the country’s stable sense of self. The weaker support for teaching patriotism associated with watching “pro-Russian” Ukrainian channels can similarly be considered problematic for the country’s ontological security, as it suggests a lower commitment to the Ukrainian national idea. Support for teaching the Russian language as part of Ukrainian citizenship, which was strongly associated with watching Russian and “pro-Russian” TV channels, represents a challenge to the conception of Ukraine enshrined in legislation after 2014, wherein the Ukrainian language was declared a “state-creating factor of the Ukrainian nation” (“державотворчого чинника української нації”) and the linguistic Ukrainianisation (effectively, de-russification) of public life was, therefore, deemed necessary for Ukraine’s consolidation as a sovereign nation-state (Law of Ukraine 2019). It is noteworthy, however, that despite fears of the banned media “inciting hatred towards Ukraine’s people, culture, state language and national identity” (Law of Ukraine 2022), our results show no relationship between the use of banned media and support for teaching knowledge of Ukraine’s history, culture or the Ukrainian language. Indeed, the strong and apparently media-resilient consensus that we observed on these questions is worth noting in the context of Russia’s attempts to destroy Ukrainian identity through propaganda and enforced russification in the territories it has occupied.

As for truth perceptions, our results support (without confirming causation) the Ukrainian authorities’ fear that the banned media were sustaining a distorted view of events among citizens. Using Russian social media and the “pro-Russian” Ukrainian TV channels was associated with higher belief in homophobic and anti-liberal disinformation. Using Russian federal TV channels and the “pro-Russian” Ukrainian channels was associated with lower belief in true news about Ukraine rising in the development index. Using the banned “pro-Russian” Ukrainian channels was associated with lower belief in true news about Russia cyberattacking Poland. Inaccurate views on such issues are detrimental to democracy in the first instance (because they impede rational deliberation) but they also have negative implications for Ukraine’s physical security. For example, scepticism about Ukraine rising in the development index suggests weak national morale, while scepticism about the cyberattack on Poland might indicate a reluctance to believe in (and thus respond rationally to) Russian aggression. It should be noted, of course, that our respondents were only presented with headlines, not detailed evidence, and some may have responded differently had the full news stories been presented alongside the headlines.

Our results align with previous research which found that users of Russian media outside Russia tend towards disillusionment with the state of affairs in their country and reluctance to recognise threats from Russia (Toal and O’Loughlin 2018 Wagnsson 2022). We go beyond previous research by finding that users of Russian social media platforms (not just TV channels) are more likely than non-users to believe certain examples of disinformation and that users of both Russian TV channels and “pro-Russian” domestically owned channels tend to express lower support for important values, such as democracy and patriotism.

In light of this evidence, is it reasonable to say that Ukraine’s media restrictions were logical from a security viewpoint and likely to strengthen national security? This question is hugely complex. On the one hand, the use of the banned media did reportedly fall substantially following the imposition of restrictions (USAID-Internews 2020). If we assume that the statistical effects described above were causal effects, then it is plausible that removing the banned media from Ukrainian media diets might (sooner or later) lead to a reduction in the problematic effects on perceptions of truth and national values, which could, in turn, potentially boost national resilience, morale and resistance to Russia’s invasion. Much of this logical chain is, however, as speculative much as it is plausible.

On the other hand, there are still important unknowns which complicate the assessment of the security logic. We do not know the extent to which the restrictive measures antagonised “pro-Russian” minorities. More critically, we do not know precisely how the media restrictions affected decisions in Moscow. The Moscow regime’s visibly diminishing ability to influence Ukrainian politics and society via the mass media may have contributed to its decision to ramp up military aggression (Zhegulev 2023). Putin said in May 2021 that Russia would respond “properly and in a timely fashion” to the media restrictions, without elaborating further (AFP 2021). He associated the media restrictions with the notion that Ukraine was becoming a threatening “anti-Russia”, an idea that has featured prominently in his public statements to justify the full-scale invasion of 2022.

Essentially, Ukraine in recent years faced an irresolvable dilemma: the more it tried to defend its ontological security by restricting media that threatened Ukrainian democracy, national pride and distinct national identity, the more it found its physical security (the integrity of its borders and lives of its people) threatened by the regime in Moscow that was unwilling to tolerate the prospect of a politically, culturally and linguistically “de-Russified” Ukraine. Perhaps Ukraine could have preserved the integrity of its borders by accepting Russian political and cultural influence (via the media and other channels), as it did during the Yanukovych presidency. But in accepting such influence, vital elements of its national and constitutional identity – as a sovereign democracy, with a distinct and proud history and culture, rooted in the Ukrainian language – would have risked erosion.

What can our study of Ukraine contribute to the wider debate about when and whether democracies should impose media restrictions for security purposes? It is important to acknowledge that Ukraine’s security situation is more precarious than most, and overgeneralising is probably inadvisable. Few other democracies face the same level of existential threat, physical and ontological, from a powerfully armed autocratic neighbour; therefore, the type of intractable security dilemma described in the previous paragraph is not common (some have suggested parallels with Taiwan which might be worth exploring, but that part of the world lies beyond the scope of our own expertise). The Ukrainian case is important, however, for the debate on how to deal with media-related security threats, because it shows that the relationship between security, democracy and media restrictions is not necessarily one of straightforward “trade-offs”, whereby any boost to security from tighter media restrictions comes at the cost of undermining democracy. Previous research has emphasised tensions between countering Russian propaganda and staying true to democratic ideals (Wagnsson and Hellman 2018). But in Ukraine’s case, Russia posed and continues to pose an existential threat not only to Ukraine as a sovereign nation-state, but also to Ukraine as an (albeit imperfect) democracy. Russia, a repressive autocracy, has been striving to incorporate Ukrainian lands and people into its own political system by military force, with Russian and “pro-Russian” media supporting these efforts. The analysis here (while not conclusive) suggests multiple mechanisms via which such media were threatening the continued development of democracy in Ukraine: via eroding public support for democracy; via distorting perceptions of truth and thereby hindering rational debate and via weakening the morale needed to fuel resistance and defence of the democratic state in the case of physical attack. A democratic rationale, therefore, exists for the media restrictions, alongside a security rationale. While neither is totally robust against criticism, it is questionable whether Ukraine would have been “more democratic” over the long term had the media restrictions not been introduced.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

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

Additional information

Funding

This work was supported by the British Academy under grant IC4/100133.

Notes

1 Respondents were asked: “До якої національності чи національностей Ви себе відносите?” (“What do you consider to be your nationality or nationalities?”). The word for nationality can also be translated as ethnicity. Respondents could choose multiple answers (Ukrainian, Russian, Hungarian, Slovak, etc.).