Time, Space and Information

This report examines how governments exploit legal and institutional mechanisms to shrink civic space, particularly targeting civil society organizations (CSOs) and the media. These actors—critical forms of diagonal accountability—not only scrutinize government power but also connect citizens to democratic processes beyond the ballot box. When captured or silenced, governments undermine both institutional checks and the emotional life of the electorate, easing manipulation and entrenchment of power.

Drawing from comparative case studies and the insights of experts convened by International IDEA and the Bonavero Institute of Human Rights, the report maps the legal-institutional toolkit of democratic backsliding and distils lessons for strengthening resilience.

Backsliding tactics

The report identifies seven key strategies used globally to constrict civic space:

1. Capturing oversight bodies. Governments pack or restructure media regulators and non-governmental oversight institutions to ensure partisan control, as in Ecuador’s media councils or Hungary’s institutional mergers.
2. Co-optation of outlets and organizations. Regimes acquire media through proxy ownership, resource starvation or legal disputes, while CSOs are enticed or pressured into collaboration, as seen in Cambodia and Nicaragua.
3. Obstructive regulations. Byzantine registration systems, ‘foreign agent’ laws and technical requirements drain resources and expose groups to punitive enforcement.
4. Activity restrictions. Vague laws criminalize dissent, restrict protest or ban content, from El Salvador’s laws curbing reporting on gangs to Türkiye’s prohibitions on ‘praising criminals’.
5. Abuse of exigency laws. Emergency and anti-terrorism frameworks justify curbs on expression and association, as in El Salvador’s rolling state of exception or in Türkiye following the 2016 coup.
6. Lawfare. Legal harassment through defamation suits, selective audits or trivial prosecutions (‘rule by law’) burdens opponents while preserving democratic façades.
7. Manipulating funding. Governments withhold or redirect advertising and subsidies to punish critical outlets and reward allies, as documented in Hungary, India, Poland and beyond.

Collectively, these tactics erode independence, silence dissent and tilt the information environment in favour of incumbents.

Challenges of resistance and reversal

Resisting backsliding is notoriously difficult once regimes capture the courts, the media and the electoral rules. Successful resistance depends on intact institutions, such as Brazil’s independent judiciary or India’s robust access-to-information regime. Reversing democratic erosion after illiberal regimes fall is even harder: institutional distortions are persistent, authoritarian holdovers obstruct reform, and new governments face perverse incentives to retain expanded powers. Nearly 90 per cent of democracies experiencing a turnaround since 1994 have failed to regain their prior levels of democracy within five years (Bianchi, Cheeseman and Cyr 2025).

Building resilience by design

The report stresses the importance of proactive institutional design, emphasizing resilience as the system’s ability to absorb shocks, adapt and recover. Protecting civic space requires attention to three dimensions:

1. Space. Safeguarding fundamental freedoms of speech, assembly and association; ensuring accessible, nonpartisan registration regimes; and curbing abuses of discretionary funding.
2. Time. Preventing governments from overwhelming CSOs with litigation or accelerating legislative processes to preclude scrutiny—for example, through cross-partisan approval of urgency procedures or opposition ‘pause buttons’.
3. Information. Guaranteeing access to government information, protecting media pluralism and insulating regulators from political capture.

At the system level, resilience requires strong, independent judiciaries, diverse institutions that cannot all be captured at once, diffusion of power and safeguards such as cross-partisan appointments.

The comparative evidence demonstrates that attacks on civic space are central to the playbook of backsliding governments. Because resistance and reversal are so challenging once erosion is underway, the imperative lies in proactively embedding resilience—through robust rights protections; institutional design that defends space, time and information; and system-wide resilience. Strengthening civic space is not only about protecting CSOs and the media; it is about fortifying democracy itself.

The effects of democratic backsliding have been felt across the pillars of democracy, from core constitutional checks and balances to social support for democratic norms. It has had no less an impact on the civic space—the realm in which people and organizations outside government organize to ‘take collective action in pursuit of the public good’ (Alagappa 2004).

This report examines those impacts, with particular attention given to the media and civil society organizations¹ (CSOs)—though the civic space generally can be thought of in broader terms to include mass membership organizations (MMOs) like trade unions or religious groups, grassroots movements and social organizations, which themselves play an important role in a healthy democracy and are also harmed by the closing of civic space.

Partly based on a discussion among experts and practitioners hosted by the International Institute for Democracy and Electoral Assistance (International IDEA) and the Bonavero Institute of Human Rights at the University of Oxford, the report explores the legal-institutional mechanisms that have been used to undermine CSOs and the media, and draws lessons on how to strengthen their resilience against future backsliding through better institutional design. These reflections can inform efforts such as those recently undertaken in Germany, Norway and Sweden to proactively reinforce democratic institutions against potential illiberal, anti-democratic attacks (Oancea and Bisarya 2025).

The media and CSOs have been characterized as institutions of diagonal accountability, a crucial part of a larger ecosystem alongside institutions of horizontal (checks and balances) and vertical (the electorate) accountability. CSOs, alongside the media, seek to hold government accountable through promoting transparency and accountability.² Indeed, a free and independent media, for example, is correlated with higher levels of democratization (Norris 2008), economic development (Stiglitz 2002; Roll and Talbott 2003), reduced corruption (Djankov et al. 2002) and increased political participation (Leeson 2008; Monetti 2024: 1640). These institutions have come under strain due to ‘notable declines in Freedom of Expression, Freedom of Association and Assembly, and Freedom of the Press ... [in] every region of the world’ (International IDEA 2023: 30).

Equally important to the role of the media and CSOs in promoting accountability is their relationship to the populace. By opening channels of communication and mobilization, the media and CSOs enable and encourage the public to go beyond promoting vertical accountability through the franchise and to play a more proactive role in holding the government accountable through civic engagement and advocacy (Anheier 2004; Diani 2015; Gerő et al. 2023).

Attacking the media and CSOs thus presents a double opportunity for democratic backsliding. In addition to undermining diagonal checks, it enables the capture of a major channel to the emotional life of the electorate, allowing more direct manipulation and charging of emotions, which are key drivers of backsliding (Koncewicz and Strother 2019). At the same time, ‘[b]oth autocracy research and social movement studies suggest that repression can either stifle civil society or promote anti-incumbent mobilisation, making repression a double-edged sword for autocratic regimes’ (Lorch 2023: 398; see also Bischof and Fink 2015; Sombatpoonsiri 2021).

Many strategies used against the media and civil society are political in nature. One popular tool is rhetoric, which is used to paint outlets and organizations as (a) unreliable; (b) obsolete; or (c) an enemy of the people (Monetti 2024).

While political strategies and culture are important, this report limits its focus to legal-institutional channels of democratic backsliding. To that end, it adopts the definition of ‘backsliding’ laid out in International IDEA’s report Designing Resistance: Democratic Institutions and the Threat of Backsliding (Bisarya and Rogers 2023), following Bermeo (2016), Choudhry (2018), and Ginsburg and Huq (2018).

This definition contains three components. First, backsliding is carried out by governments which were elected through democratic elections (whether free and fair or not). Second, it is carried out through legal means, ‘in the sense of both (a) acting within the formal constraints of the law and (b) instrumentalizing the law to further self-serving partisan ends of insulating the regime from constraints and competition’ (Bisarya and Rogers 2023: 20). And third, it undermines the core institutions of democracy, including elections and the fundamental rights necessary to make those elections meaningful, in order to aggrandize and entrench power and tilt the electoral playing field.

The media, CSOs and civil society at large may collectively be regarded as one core institution of democracy. Backsliding strategies have been used to great effect in undermining this institution in recent years. Since 2013, for example, media freedom has decreased in 15 out of 18 Latin American countries, concerns about media pluralism in Europe have climbed steeply, and CSOs in both regions have found their funding restricted through government redirections or ‘foreign agent’ laws (Reporters Without Borders n.d.; Bleyer-Simon et al. 2024). Legal-institutional changes have even been successful in producing social stigma strong enough to prompt self-censorship. A survey of televised debates in India, for example, found not a single instance of criticism of the government of Prime Minister Narendra Modi in one three-month period (Jaffrelot and Jumle 2020). ‘It’s ridiculous’, said one Turkish professor about such chilling effects in Türkiye, ‘but it sows fear’ (Champion 2011).

Backsliding strategies used to undermine the media and civil society sectors include direct attacks on organizations and outlets, capture of the government institutions which regulate them, and the closing or shrinking of civic space generally (Carothers 2016), such that ‘the legal and political environment for civil society organisations … is increasingly hostile’ (Pospieszna and Pietrzyk‐Reeves 2022; Gerő et al. 2023: 16).

Chapter 1 of this report lays out the most prevalent of the legal-institutional tactics used to carry out these strategies, and Chapter 2 discusses takeaways for better protecting these institutions in the future.

Backsliding encompasses a broad suite of strategies to undermine checks on the power of the political majority and to tilt the electoral playing field. International IDEA’s 2023 report Designing Resistance catalogues several tactics that fall into 12 general strategic threads (Bisarya and Rogers 2023: 34–53). While many of these strategies can be used to undermine the media and civil society, this report focuses on seven that are the most prominent and most directly applicable: (a) capturing independent regulatory and oversight bodies; (b) co-optation; (c) obstructive regulations; (d) legal restrictions; (e) abuse of exigency laws; (f) lawfare; and (g) withholding and redirecting funding. Each of these strategies, and the tactics used to implement them, is discussed below.

1.1. Capturing independent regulatory and oversight bodies

Facing a media campaign and growing civic action that threatened to stymie his intention to implement new oil drilling, President Rafael Correa of Ecuador unveiled a new media oversight system that changed the face of Ecuadoran media. The Organic Communications Law established two oversight bodies for the media, and another law redefined media as a ‘public service’, enabling the government to regulate it in the same way it could regulate electricity or water (CPJ 2018). Members of these oversight bodies were appointed by the political majority, resulting in government control over media regulation. These bodies were then used to carry out partisan attacks, often including levying multimillion-dollar fines for violations such as failing to report on an event or providing what the authorities deemed ‘unbalanced coverage’ (Conaghan 2015).

This story is an example of institutional capture which targets bodies governing the public media, institutions regulating the private media and bodies responsible for oversight of civil society organizations such as those that determine how much public funding will be distributed to non-governmental organizations. This kind of institutional capture may be achieved by installing loyal actors or making adjustments to an institution’s architecture to bring it under government control. Often, governments mask changes to institutions in a democratic shell, creating the appearance of proactivity and making institutions nominally more democratic—that is, subject to majoritarian control.

One common way to capture what are meant to be independent institutions is by packing them with partisan actors who will prioritize their loyalty to the regime over the professional standards of the office. This strategy also negates the effect of slow turnover, which would result in a body comprised of members appointed by different administrations and would gravitate towards professionalism and moderation, thereby encouraging impartial oversight. Packing is most expediently achieved by creating a new body which can then be completely filled by the sitting majority, as in Ecuador (Haggard and Kaufman 2021:79). Other options include removing current members and replacing them (often creatively, such as by lowering statutory retirement ages), expanding the number of posts to dilute the power of existing office holders and changing appointment procedures to facilitate partisan appointments.

A second tactic of institutional capture is to realign chains of accountability and control in order to create a pipeline for political power to flow into previously independent bodies. This sort of realignment may be carried out, for example, by bringing an office into the remit of a trusted cabinet minister, as in Sri Lanka, where the National Secretariat of Non-Governmental Organizations was moved from the Ministry of Health and Social Services to the auspices of the Ministry of Defence and then of Public Security, in line with the ‘securitization’ of the civic sector (Surie, Saluja and Nixon 2023: 5), or, as in Hungary, by merging previously independent offices into another institution that is subject to closer political control (Haggard and Kaufman 2021: 109; Gerő et al. 2023: 19). Changing the rules regarding termination and tenure can effect a similar type of control.

Those seeking to weaken or capture the civic space can also attain more control without changing the formal institutional structure, by leaving the institution itself formally untouched and then shifting the powers, resources or jurisdiction of the body to another, less independent body (an existing body or one specially created for this purpose). This tactic was used in Poland to sidestep the constitutionally protected National Broadcasting Council: Parliament established a new National Media Council that duplicated much of the National Broadcasting Council’s mandate. The five-member National Media Council was appointed by the president and was given the power, among other things, ‘to hire and fire personnel for state television and radio’ (Rice 2017: 162; Sadurski 2019: 139).

Of course, where they have the political capital to do so, backsliders may replace an institution altogether, which allows them to pack it by appointing officials from scratch and tailor the institutional structure to their liking. Often, this move is justified with plausible, even laudable, goals, like making institutions more ‘democratic’ (i.e. majoritarian) or ‘accountable’ (i.e. subject to political control). For example, Slovakia’s Parliament recently approved a reform replacing the Slovak public broadcaster (RTVS) with a new one (STVR) which is managed by a board of directors directly appointed by the government (Čuroš 2024).

1.2. Co-optation of media outlets and civil society organizations

The administration of President Daniel Ortega in Nicaragua has taken steps to reduce plurality in media ownership, with a curious result. The administration has used a variety of channels to bring the media under their control. In some cases, investors linked to the regime bought private stations. In another, the administration leveraged a legal issue—a dispute over a private media channel’s ownership structure—to take over an outlet. Meanwhile, the administration redirected state advertising funds (discussed further in 1.7: Withholding and redirecting funds) significantly enough to secure complete control over an outlet. Using these various methods of direct or proxy acquisition, President Ortega’s children—Camila, Carlos Enrique, Juan Carlos, Luciana and Maurice Ortega—were installed as new owners and directors (Colburn and Cruz 2012: 116; State Media Monitor 2025).

This series of events in Nicaragua illustrates some of the various ways in which backsliding governments can move beyond capturing government oversight bodies and seek to capture or co-opt media outlets and CSOs themselves.

With regard to the media, capture or co-optation usually entails putting the media market under direct government control or influence³ by expanding the market share of the public media (as in Bolivia, where a statute reserved one third of the market for public media) or acquiring private media, whether by direct acquisition or crony ownership (Dunham, Nelson and Aghekyan 2015). In order to take control of the media market, a government can change the rules regarding media ownership to restrict the market, in particular by restricting the share of foreign ownership. In Ecuador, the government held a significant market share directly due to the failure to divest company holdings that reverted to the government via bankruptcy (Karlekar and Dunham 2013). In Türkiye, meanwhile, more than two thirds of private media were controlled by government-affiliated groups as of 2024 (Bleyer-Simon et al. 2024: 110).

A common mechanism for acquiring media outlets is to drain private media of resources and then scoop them up in bankruptcy proceedings. The administration of President Evo Morales in Bolivia used a silent strategy in which ‘several major media outlets [were] sold following targeted tax audits and corruption investigations that press groups claimed drained them of resources and readied them to be bought out by new, pro-government ownership’ (Dunham, Nelson and Aghekyan 2015).

An enabling factor for co-optation is the lack of protection for market pluralism or weak conflict-of-interest standards. In Serbia, a provision was introduced enabling state-owned companies to establish and purchase media outlets; in Montenegro, no legal barriers were available to prevent all commercial television stations with a national frequency from coming under foreign, pro-Serbian ownership; and in Cyprus, ‘almost every barrier against ownership control was lifted’, allowing anyone, including politicians, to own media companies (Bleyer-Simon et al. 2024).

Maintaining control over public and private media simultaneously silences dissent while acting as an affirmative tool to further democratic erosion. Control allows governments to dominate the airwaves with their presence and their narrative, and to block the opposition out of the minds of the electorate, such as state-run television in Bolivia being used to disadvantage opposition candidates by, for example, broadcasting a football match in place of a presidential debate that then-President Morales did not take part in (Dunham, Nelson and Aghekyan 2015).

Co-optation of civil society often involves more political and non-institutional tactics than capturing the media. Cambodia, albeit not a democratic context, demonstrates how the same legal tactics that are used as sticks to repress critical CSOs can also be used as carrots to entice CSOs into acquiescing to co-optation, termed ‘abeyance’ (Taylor 2013). Facing a choice between dissolution or absorption into a government- and Chinese-funded coalition of co-opted CSOs, many Cambodian CSOs chose the latter in order to be able to continue their activities, such as service provision (Lorch 2023).

Another notable target of co-optation in the civic sphere is the religious establishment (Halmai 2018). Dissolution of the separation of church and state reciprocally increases the influence of government in religion, and of religion in government. In Poland, for example, the Law and Justice party (PiS) co-opted religious groups into the government, including as part of their media capture strategy: once PiS took power, ‘some two hundred journalists were purged from public TV and radio, and replaced mainly with journalists coming from ... the “media empire” of Fr. Tadeusz Rydzyk’, the founder of two influential fundamentalist Catholic TV and radio stations, to run the public media (Sadurski 2019: 138).

1.3. Obstructive regulations

In Ecuador, the Correa administration sought to tame civil society by forcing organizations to navigate a byzantine registration system. This system, enacted by executive decree, required ‘all organizations-whether social movements, advocacy groups, [labour] unions, or chambers of commerce-to follow complex reporting rules regarding their activities or risk losing legal status’ (Conaghan 2016: 115). It further prohibited groups from engaging in ‘partisan activity’, ‘interfering’ with public policy or threatening ‘internal security’ (these substantive restrictions are discussed in more detail in 1.4: Restrictions on activities). This web of registration, reporting and other technical requirements simply overwhelmed many groups, especially smaller and social groups like workers’ organizations (Conaghan 2016: 115).

This system in Ecuador exemplifies the use of legal requirements to impose hurdles, making it difficult for media outlets and civil society organizations to operate, draining them of resources, and setting them up for technical infractions which can be used as grounds for fines or even dissolution. Such a system runs counter to the international standard on freedom of association to create an enabling regulatory environment (see, for example, United Nations General Assembly 2013).

Qualification restrictions play a similar gatekeeping role, such as in Ecuador and El Salvador, which have set specific qualification requirements on who can become a journalist (Dunham, Nelson and Aghekyan 2015; Free Press Unlimited 2024). Short turnaround times to register can exacerbate the logistical burden on journalists: in one instance, journalists in Uganda were given seven days in which to register themselves as journalists or else forfeit their right to cover the presidential election (Gloppen, Gerzso and van de Walle 2022: 66).

The most notorious type of obstructive regulations is the increasingly common ‘foreign agent’ law. From Georgia to Peru, such laws have become a mainstay of backsliding and have proven effective in chilling civic activity both by shutting down organizations (by imposing high taxes on foreign aid or interfering with their ability to receive foreign funds) and by labelling them as ‘representing foreign interests’. The Venezuelan Supreme Court went so far as to rule that CSOs receiving funding from foreign governments or whose leaders who were not Venezuelan were by definition not a part of ‘civil society’. The Court held that such organizations were no longer allowed to carry out certain activities, including representing citizens in court (Haggard and Kaufman 2021: 264). At the time of writing this report, the Hungarian ruling party was considering legislation which would significantly broaden the government’s scope to monitor, penalize and ban civil society organizations and media outlets which are deemed to be using foreign funds to threaten the country’s sovereignty.

Finally, complex and onerous regulations create minefields, making it difficult for organizations to avoid even inadvertent infractions. Violations of subtle and obscure provisions can carry heavy penalties or prompt dissolution. One critical newspaper in Ecuador that was already in the process of closing its doors faced a USD 57,000 fine for failing, in that process, to publish the number of copies in circulation over the previous two months (Dunham, Nelson and Aghekyan 2015). Such regulations set the stage for arbitrary and abusive enforcement (see 1.6: Lawfare).

1.4. Restrictions on activities

A legal provision in Türkiye prohibits ‘praising a crime or criminals’, and in Ecuador organizations face dissolution for ‘pursuing partisan activity which is reserved for parties or interfering in public policies in a way that contravenes internal or external security of the state or disturbs the public peace’ (Conaghan 2015: 18; United States Department of State 2018). Similarly, in 2021 the Constitutional Court of Guatemala upheld a law permitting the government to dissolve CSOs for any activity that might ‘alter the public order’ (USAID 2024).

As these examples illustrate, often the most direct and attractive way for a would-be autocrat to control the messages put into the public sphere is to place direct substantive restrictions on the content that media can cover or activities in which CSOs can engage.

Content restrictions are most often connected to matters of security or public order and use vague wording to allow for broad prosecutorial discretion. However, some subject-matter restrictions seek to prevent organizations from speaking out on particular human rights issues. A media law in this vein, passed in 2025 in Georgia, explicitly prohibits broadcasters from disseminating content that ‘promotes identification with a gender other than one’s biological sex or relationships between individuals of the same biological sex based on sexual orientation’ (Media Freedom Rapid Response 2024). Restrictions of this kind undermine the civic space by seeking to cut entire swaths of the population—usually minorities which do not fit into a regime’s nationalist-populist version of ‘the people’—out of it (see, for example Suteu 2025). Content control can also be used in the opposite direction, however, to force outlets to cover certain events and topics or CSOs to promote a certain message. In Venezuela, media outlets were required to broadcast all of Hugo Chávez’s speeches, and in Ecuador outlets have faced multimillion-dollar fines for failing to report on specific events (CPJ 2018).

Backsliding governments are also increasingly using criminal law to silence civil society actors. One prominent manifestation of this tactic is the weaponization of criminal defamation laws (see 1.6: Lawfare). Criminal laws have also been used to deter or punish participation in public protests. Authorities frequently invoke broad or vaguely worded criminal charges such as ‘incitement to mass disorder’ or ‘calling for unauthorized protests’ to prevent activists from attending demonstrations. Activists may also be arrested during protests and face criminal charges for offences such as ‘disobeying lawful police orders’ or ‘participating in an unauthorized mass gathering’ (Human Rights Watch 2023; Amnesty International 2024). In Kazakhstan, authorities have arrested and detained activists who expressed their intention to take part in protests or shared related content on social media, at times subjecting them to prolonged interrogation. They have also repeatedly relied on criminal charges, on grounds such as ‘mass rioting’ or ‘acts of terrorism’, to arrest and detain protest participants (Amnesty International 2016; Human Rights Watch 2022).

Another manifestation of restrictions on activities is the broadening of privacy protections to prohibit the unauthorized dissemination of personal information even for issues of public interest (Conaghan 2016, discussing Ecuador). Such restrictions have also increasingly involved the criminalization of specific messages or behaviours in the civic sector. In April 2022, for example, the Legislative Assembly of El Salvador approved a law criminalizing the publication of material replicating gang messages.⁴ The prominent opposition outlet El Faro alleged that this criminalization was intended, in part, to prevent journalists from reporting on secret negotiations between President Nayib Bukele and the country’s main gangs (El Faro Editorial Board 2023).

1.5. Abuse of exigency laws

Since the 1990s, a group of women (and others) called the ‘Saturday Mothers’ in Türkiye have met weekly to hold vigils to ‘commemorate the disappearances of relatives following their detention by Turkish security forces in the 1980s and 1990s and call for accountability’ (United States Department of State 2018). In 2018 the government shut down this long-standing practice under the guise of a state of emergency that had been imposed in the wake of a coup attempt two years prior, in 2016, claiming that ‘the group was exploiting the concept of motherhood to mask support for terrorism’. In the previous year, the Human Rights Joint Platform (2018: 43) reported that the government had shut down nearly 1,500 non-governmental associations or foundations for alleged threats to national security.

Laws governing emergencies, security and anti-terrorism have proven valuable for aspiring autocrats in restricting freedoms and justifying government interference in the media and CSOs. Such laws have primarily been used to justify restrictions on certain rights, particularly of expression, assembly and association, in order to quash protests and silence dissent. That said, this strategy has also been used to target other rights as well. For example, the state of exception in El Salvador—which has been in place since 2022 and renewed 35 times—has been used to justify restrictions on access to government information by journalists and citizens (Reporteros Sin Fronteras 2024). And in many places—including El Salvador, Israel and Zimbabwe—states of exception have provided grounds for selectively targeting critical organizations through surveillance, raids and other intrusions.

Often, exigencies also provide cover for the restriction of fundamental rights, which are a necessary predicate to the exercise of diagonal accountability. A coup attempt in Türkiye in 2016 provided grounds for declaring a state of emergency that lasted for two years. While it was in place, the government expanded its powers to restrict freedom of assembly and other rights (Sinclair-Webb 2016). Similarly, the Covid-19 pandemic handed executives powers that could be abused to similar effect. In Sri Lanka, for example, a curfew imposed with no clear legal basis served as grounds for thousands of arrests (Fonseka, Ganeshathasan and Welikala 2021). Those detained for curfew violations also found themselves bereft of due process, as the pandemic had prompted the courts to close (Fonseka, Ganeshathasan and Welikala 2021).

1.6. Lawfare

A satirical social media post uploaded in Türkiye in 2014 featured an image containing three pictures of President Recep Erdoğan next to three pictures of the character Gollum, from The Lord of the Rings, with similar facial expressions. The image led to a series of prosecutions known as the ‘Erdoğan–Gollum comparison trials’. Rifat Çetin, who made the post, was prosecuted for insulting the President, a process which included expert testimony surrounding whether Gollum was good or evil (Varol 2018: 344). Upon being found guilty, Çetin received a one-year suspended sentence and lost his parental custody rights.

In a similar case, Bilgin Çiftçi, a physician who shared a meme comparing Erdoğan and Gollum, lost his job and was also put on trial (Orucoglu 2015). During the trial, Peter Jackson, director of the Lord of the Rings films, commented on the case, arguing that ‘the character depicted [in the meme shared by Çiftçi] was Smeagol, the “good” part of Gollum’s personality’ (Shaheen 2015).

Jackson’s comment also served as the defence for a third defendant who was brought to trial, the fashion designer and LGBTQ+ activist Barbaros Şansal, who said the following: ‘A couple of days before tweeting, I heard in a court’s verdict that Gollum is a helpful and good character. Thereupon I felt the urge to retweet the tweet in question.’ Şansal was later sentenced to 14 months in prison (Bianet 2023).

Lawfare refers to the weaponization of extant legal rules to attack an opponent (à la the apocryphal ‘for my enemies, the law’) (Cheung 2018: 4). Often, this tactic involves adhering strictly to the letter of the law while undermining its spirit—rule by law instead of rule of law (Sajó 2021). Taking opponents to court, even where a favourable verdict is unlikely, can still be used to occupy and discredit them, as well as drain resources, sap morale, foster stigma and encourage self-censorship (Gloppen, Gerzso and van de Walle 2022: 68). There may even be an incentive for governments not to succeed in such suits lest they take on the internationally frowned-upon reputation of someone who jails their opponents. Indeed, in Bolivia, President Evo Morales on more than one occasion actually pardoned journalists and opposition members when criminal charges were successful (Dunham, Nelson and Aghekyan 2015).

Strategic lawsuits against public participation (SLAPPs) are legal actions filed primarily to intimidate, silence or burden critics by dragging them into costly and time-consuming litigation. The use of SLAPPs by governments has been a growing concern throughout the world. One particularly common form of SLAPPs are defamation suits—both civil and criminal—which have been used in virtually all recent episodes of democratic regression. Criminal defamation, where it still exists, is often a little-used colonial holdover that has been resurrected by backsliders and, as a result, is being used more often than ever before (Ahuja 2018). Examples from Türkiye—in addition to the Erdoğan–Gollum trials mentioned above—include sentencing a 14-year-old boy to five months in prison for insulting the President, and convicting a former Miss Türkiye for sharing a satirical poem on social media (BBC 2016a, 2016b; Bleyer-Simon et al. 2024). The United States has also increasingly become a paradigm in this regard. President Donald Trump has long relied on lawsuits of all varieties to pressure opponents in business dealings as well as in politics. As President, civil defamation suits have become especially effective in silencing unfavourable press coverage. Even major media companies have capitulated to the pressure of the threat of litigation, choosing to settle cases with little or no legal merit (Grynbaum and Feuer 2024). Conversely, President Trump has sought to punish those helping bring defamation suits against outlets he favours. In April 2025 he signed an executive order against Susman Godfrey, the firm that represented Dominion Voting Systems in a suit settled by Fox News, seeking ‘to harm the firm by limiting its attorneys from accessing government buildings, revoking security clearances and essentially making it impossible for it to represent anyone who has business before the federal government’ (Levine 2025).

A second major form of lawfare is selective prosecution, whereby critics are charged for crimes that either are minor and not usually prosecuted (such as zoning or minor tax violations) or are not enforced against the government’s supporters or associates (e.g. anti-corruption violations). Selective investigations, fines and prosecutions against critical media outlets or CSOs represent the other side of the coin to obstructive regulations and legal restrictions: while the examples of regulations in the preceding two sections are largely neutral on the surface, it is prosecutorial discretion that enforces them selectively against criticism and dissent. In the run-up to the 2016 legislative elections in Zambia, for example, a tax audit was launched against the opposition newspaper The Post, which was forced to close as a result of unpaid taxes. While both the tax law and the audit were ostensibly legal, it was notable that each of the state-run newspapers owed more in unpaid taxes than The Post did (Hinfelaar, Rakner and van de Walle 2022: 196).

Similar to the selective enforcement of regulations in ways that technically adhere to the law but are abusive in practice is the use of ‘formally correct structures of the [rule of law to] serve efficient, disciplined arbitrariness’ (Sajó 2021: 239). This arbitrariness is often manifested in the minuteness or almost frivolous triviality of regulatory violations. In North Macedonia, for example, an independent outlet was fined for failure to devote sufficient airtime to folk music (Haggard and Kaufman 2021: 126).

In South Korea, one news station faced disciplinary action when, in reporting a Level 1 dust concentration on Seoul’s air-quality index, it displayed a large blue number 1. This display, the Election Broadcasting Deliberation Committee found, demonstrated bias in favour of the Democratic Party of Korea, which is associated with the colour blue and which used the number 1 in its election symbol (The Korea Times 2024).

Here again the vagueness of the statutory wording becomes potent. A requirement in Hungary for reporting to be ‘thorough’ and ‘responsible’, for instance, poses almost no limit on the law’s scope (Keleman 2017: 222). Similarly, enforcement often involves an especially broad interpretation of statutory terms in order to portray opposition behaviour in line with that interpretation. For instance, an Argentine law criminalizing ‘terrorist acts’ was interpreted to cover published material that ‘terrorizes’ the public, though the government stated that the law was not intended for use against the media. This posture changed, however, when the editor of an online outlet was charged with ‘terrorizing the population’ for his coverage of the brutal arrest of a police officer who had led a campaign for higher salaries (Martínez 2014; Dunham, Nelson and Aghekyan 2015).

In addition to vagueness, laws and regulations giving government actors unduly wide swaths of discretion can similarly be weaponized. Appointing allies to positions holding wide discretion enables what Gloppen, Gerzso and van de Walle (2022) term ‘bureaucratic lawfare’, which ‘seeks to change the nature of administrative institutions and practices by exploiting the discretionary space inherent in the law’ (Gloppen and Rakner 2024: 4).

The inverse of selective prosecution is selective enforcement of rights and protections, such as failing to protect protesters or minority groups from violence. This tactic allows those in power to punish selected groups by omission while having robust protective laws on the books to create the outward appearance of democratic liberalism. For example, the Justice and Development Party in Türkiye initially signalled movement towards increased human rights protections as part of a potential path to European Union accession, including more robust legal and constitutional protections for ethnic and religious minorities, women and children, and a new national human rights institution and ombudsperson to investigate complaints against the government (Human Rights Watch 2013). However, the government broadly declined to enforce these protections (Varol 2018: 346). Human Rights Watch noted in a 2013 report that ‘[v]iolence in the home remained endemic, with police and courts regularly failing to protect women who have applied for protection orders’ in spite of the new Law on the Protection of the Family and Prevention of Violence against Women; that prolonged pre-trial detention periods were being used to incarcerate members of the opposition and prevent them from taking their seats in parliament in spite of a reform introducing and encouraging alternatives to incarceration pending trial; and that Türkiye ‘failed to prevent, conduct effective investigations into, or provide a remedy for torture’ in spite of three separate rulings from the European Court of Human Rights ordering the government to do so (Human Rights Watch 2013).

1.7. Withholding and redirecting funds

Media and civil society are in a structurally vulnerable position in that part of their raison d’être is to hold government to account, and yet they are often dependent to at least some degree on that very government for their ability to exist. It is unsurprising therefore that exploiting public funding as a carrot or a stick, whether by withholding access to public funds or using discretionary spending powers to siphon support away from critical voices, is integral to the backslider’s playbook.

In the media, this carrot-and-stick approach largely manifests in redirecting government advertising spending. In Hungary, India, Nicaragua, North Macedonia, Poland, South Korea and elsewhere, governments have used the discretionary allocation of advertising spending to punish critical outlets. In Serbia, for example, the party in power enacted a suite of independence-enhancing laws in order to facilitate EU accession negotiations, while continuing to exert control through the less direct channel of advertising spending (Haggard and Kaufman 2021: 187).

The same tactic has been used to reward or co-opt outlets in other contexts. According to the media watchdog Newslaundry, advertising revenue for the Indian newspaper Dainik Jagran in 2017 amounted to more than the total the outlet had received in the three years before Narendra Modi came to power (Tiwari and Kumar 2021). This period also saw the paper’s editor-in-chief added to the board of India’s public broadcaster by the Ministry of Information and Broadcasting (Newslaundry 2020; ThePrint 2020).

Regarding civil society, discretionary funding schemes have generally enabled governments to withhold funds from specific CSOs. Since 2012, for example, funding from the Hungarian National Cooperation Fund has been allocated ‘increasingly selectively, limiting funding [given] to the government critical parts of civil society’ (Rohlfing 2021).

The same logic can be used to gather and fortify government-friendly CSOs. In Poland, for example, priorities have even been codified in law, privileging CSOs that espouse ‘Christian values’. This approach has been coupled with the withholding of funds from organizations on ostensibly ‘liberal’ grounds—such as denying funding to groups combating violence against women on the grounds that doing so is discriminatory, as it disregards violence against men (Sadurski 2019: 145).

This collection of strategic pathways is non-exhaustive, and backsliders continue to adapt to changing circumstances in finding new and creative ways to contract the civic space. Countering democratic backsliding, therefore, requires similarly creative and adaptive strategies. The next chapter explores some possible responses, looking at both the present and the future.

Countering backsliding can take three different forms: (a) resisting backsliding as it occurs; (b) reversing democratic erosion when pro-democracy actors return to power in the wake of a backsliding episode; and (c) building resilience into the system to prevent democratic backsliding in the future. These forms may be applied separately or simultaneously, given the stage of autocratization faced in a particular country or other contextual factors.

However, recent cases illustrate that both resistance and reversal are exceedingly difficult tasks, and both depend greatly on the amount of institutional resilience built into the system from the outset.

First, resistance. Where populist anti-democrats have taken power, supported by large political majorities and able to maintain broad popular support (e.g. El Salvador, Hungary), even the most determined and organized resistance efforts are likely to prove unable to prevent the closing of civic space and democratic regression more generally. The limited effectiveness of resistance efforts is in part due to the fact that backsliding regimes tend to target the very institutional sites that offer means of resistance first. Generally, a democratic system enables legal challenges to the government’s actions in the courts, and political challenges at the ballot box. By capturing the judiciary first, as tends to be the case, and tilting the electoral playing field, backsliding regimes undermine the ability to resist all future moves. Regarding the civic space specifically, backsliding efforts seek to make the very tools that the media and civil society can employ ineffectual. Thus, for example, as the media’s primary tool is the ability to broadcast to the masses, if a backsliding regime restricts broadcasting, the media itself cannot effectively protest because it would ordinarily do so through broadcasting.

However, in those places where efforts have been successful in stemming, or at least slowing down, backsliding, it is in large part because there have been critical elements of the architecture of constitutional democracy that have remained intact and able to act as tools or sites of resistance. In Brazil, for example, litigation was effective in resisting attempts by President Jair Bolsonaro to ignore Covid-19 measures instituted by parliament and to neglect constitutionally protected Indigenous rights during the pandemic, because the courts had managed to remain sufficiently independent (Supremo Tribunal Federal 2022; De Oliviera 2023). In India, on the other hand, a robust access-to-information framework has enabled civil society to slow down and block some efforts at backsliding through freedom-of-information requests (Roberts 2010). In Poland, the bicameral system proved important during the second half of the PiS administration, as the opposition won a majority in the Senate, which gave them a forum to enable civil society organizations to voice their opposition to various backsliding initiatives.

Second, as recent experiences in Poland, Zambia and elsewhere illustrate, where a pro-democracy government has managed to oust an illiberal regime, reversing democratic degradation presents even more of a challenge. Indeed, a recent study shows that ‘of all countries that experienced a democratic turnaround—that is, to have gone from being democratic to authoritarian and back again—since 1994 reveals that nearly 90 percent failed to sustain the level of democracy they returned to for at least five years following the end of the turnaround’ (Bianchi, Cheeseman and Cyr 2025). Further, because ‘many of the institutional distortions introduced by backsliding leaders are particular to the institutional landscape of the country in question,’ reformers ‘cannot easily draw upon solutions from other countries’ (Carothers and Carrier 2025).

The obstacles facing a government hoping to reverse democratic damage are vast. Often, for example, the previous administration will have used ‘harpooning’ tactics (Bisarya and Rogers 2023), wherein a majority designs changes such that they are much more difficult to reverse than they were to implement. Such tactics were used in Hungary, where Fidesz, the ruling party, was able to entrench changes by placing them in fundamental laws (requiring a supermajority to amend) and using electoral manipulation to then prevent the opposition from being able obtain such a supermajority in the future. Packing the courts, where judges’ long terms may mean that turnover could take decades using normal appointment procedures, is another means of protecting backsliding reforms against reversal in the future.⁵

Other challenges include ‘authoritarian holdovers’ (Bianchi, Cheeseman and Cyr 2025)—that is, veto players from the previous regime. In Guatemala, the holdover attorney general used a variety of tactics to attempt to obstruct then-incoming President Bernardo Arévalo, including requesting the suspension of his party, initiating frequent aggressive investigations and asking a court to lift the President’s immunity on at least six occasions (The Associated Press 2024; Human Rights Watch 2024). Similarly, judges appointed by PiS to Poland’s Constitutional Tribunal have blocked reform efforts by the subsequent government, including regarding depoliticization of the media (Bianchi, Cheeseman and Cyr 2025).

Facing obstacles of this kind, incoming governments will often end up with a dilemma: they can try to undo the corruption of the system from within the system, or they can act extralegally in a strategy of bending the rule of law in order to save the rule of law. The former may be impossible, given the rules and structures erected by the previous regime, and the latter runs a major political risk of inviting the criticism that the government is resorting to the same types of tactics as its predecessors and potentially inviting a game of tit for tat.

Such an incoming government also faces an array of perverse incentives that may hinder reversal. For instance, term lengths mean that it is incentivized to prioritize actions that will boost its popularity in time for the next election, rather than focusing on the long and arduous task of institutional reform. Similarly, the incoming administration inherits an enticing array of powers or reduced checks it may be loath to give up, and in some cases where such governments are made of loose coalitions, as in Poland, they may find it difficult to find and maintain consensus on their legislative agenda.

Finally, unlike backsliding, where leaders learn quickly from each other the tools and tactics they can use to undermine institutions, the same is not necessarily true with reversal. While the tools and tactics of backsliding are transferable, the effects they have in different legal systems will be contextual and complex. Thus, as ‘many of the institutional distortions introduced by backsliding leaders are particular to the institutional landscape of the country in question … reformers cannot easily draw upon solutions from other countries’, leaving each pro-democratic administration to invent the wheel for itself (Carothers and Carrier 2025).

The difficulties of resisting and reversing backsliding underscore the importance of proactively designing institutions to be more resilient against backsliding in the first place. Resilience is the ‘property of a social system to cope, survive and recover from complex challenges and crises. The characteristics of a resilient social system include flexibility, recovery, adaptation and innovation’ (International IDEA 2017).⁶

To reiterate the point that resistance, reversal and resilience are interconnected, democratic resilience in the institutional sense speaks to the built-in institutional capacity not only to prevent but also to resist and reverse backsliding. Of course, institutional design is but one component of democratic resilience, alongside things like strong norms and civic education. However, strong institutions make backsliding more difficult at the margins by heightening the political cost of backsliding measures, creating hurdles to slow them down and providing focal points for mobilization (Bisarya and Rogers 2023). Further, much as Churchill observed that ‘we shape our buildings; thereafter they shape us’,⁷ institutions shape the environment in which democratic norms are fostered and democratic practices entrenched.

It is important to note that civic space resilience has two major aspects: (a) design choices geared specifically towards protecting the media and civil society; and (b) the overall resilience of a liberal-democratic constitutional system, which ultimately protects these sectors from backsliding.

Building resilience into a system such that it is able to prevent backsliders from coming to power in the first place, or from undermining the democratic integrity of the system if they do, is the primary bulwark against attacks on the media and CSOs. Thus, while this chapter focuses primarily on how specifically to protect the civic space (2.1: Protecting the civic space), it is also important to discuss considerations for system-wide resilience (2.2: System-wide resilience).

Section 2.1 also contains a number of questions to guide the user in an interrogation of the legal architecture for civic space in their own country, along with several references to international guidelines on particular issues and themes. Many of these are region-specific, including several from European supranational institutions. These are provided here not as normative standards to be complied with but rather as guidance or inspiration for reviews of national legal frameworks which would need to be conducted in a context-sensitive manner by national actors.

2.1. Protecting the civic space

In the various cases of democratic backsliding examined, three critical elements of a civic space ecosystem emerge—space, time and information. Space refers to the critical civil and political freedoms which allow civil society to function—including freedoms of association, assembly and expression—as well as spaces in which to contest government action, including online spaces, courts and ‘the street’. But in order to utilize this space—to organize, mobilize, advocate, discuss and deliberate—time is also an important resource and has been regularly stolen by backsliding governments. Third, information is critical: where the government dominates the information space, civil society and an independent media cannot communicate with the citizenry at large to voice dissenting opinions, criticize government action and counter the government’s narrative. Thus, in considering ways in which institutional design can be strengthened to build more democratic resilience, it is helpful to think about these three elements, or goals, of democratic institutions.

2.1.1. Space

Space refers to both physical and digital space and essentially amounts to the ability of outlets and organizations to access and compete for the attention of the public. As illustrated in the previous chapter, backsliding governments limit this space or organizations’ ability to use it by (a) taking away their independence; (b) taking away their platforms; or (c) diverting resources (including time, energy and financial resources) from their work.

Backsliding regimes go to great lengths to exercise control over the civic space. Backsliding tools can be used to shape, limit, occupy, build fences around and shut players out of the civic arena. Protecting space is rooted, first and foremost, in fundamental freedoms. Freedoms of speech, assembly, association and other such basic rights are necessary preconditions that enable the media and CSOs to do their work. Legal frameworks that could be weaponized to restrict these freedoms should be reviewed, clarified and reinforced. Emergency powers provisions, for example, are less open to abusive restrictions on assembly when they are subject to clear and objective standards that are subject to legal (courts) and political (legislative) review.

A second way in which the government can attack platform space is to act as a gatekeeper and prevent access to the civic arena. Burdensome registration requirements, for example, prevent many grassroots or small-scale organizations from being able to form in the first place, and the requirements for registering as a journalist (see 1.3: Obstructive regulations) excludes individuals from the press on grounds determined by the government.

Lastly, a major tool for shaping the civic space is the use by backsliding governments of carrots and sticks to influence outlets’ and organizations’ behaviour as well as to shape the space by determining which organizations and media outlets can grow and which ones will be forced to shrink to close. The use of carrots and sticks largely comes down to the discretionary allocation of resources, such as through government advertising spending, or the allocation of public funds to non-governmental organizations. Both types of funding allocation present challenges. In terms of advertising spending, the government has a legitimate interest in communicating to the public via the media, such as through public service announcements. The allocation of public funds to CSOs is less than straightforward, in that setting priorities and allocating public funds based on those priorities is precisely what governments are elected to do, and they theoretically act in a representative capacity in choosing which public organizations to support. Limiting only the ability to abuse these discretionary spending mechanisms is the aim and may be supported in part with clearer and more objective standards in laws regarding conflicts of interest in government funding of the non-profit sector and political party advertising.

On the stick side, selective enforcement and other forms of lawfare—which will primarily be discussed in the context of harassment below (see 2.1.2: Time)—also create a chilling effect, silencing media outlets and forcing organizations out of the civic space.

It should be noted that any concerns regarding basic freedoms, access and the discretionary allocation of resources also apply in specific ways to the media, and as such are discussed below (see 2.1.3: Information).

2.1.2. Time

An important corollary of space is time. Even where the space to act is maintained, without the time to organize, mobilize, provide scrutiny and communicate to domestic and international audiences, the potential power of civil society and the media to resist backsliding is significantly diminished.

Time, as a resource for media and civil society, comes in multiple forms. This section looks first at the time of organizations in the sense of manpower hours and, second, at the ability of backsliding governments to speed up the political process to effectively eliminate scrutiny on the part of the opposition, the media and civil society.

With regard to the former, many of the backsliding strategies surveyed in the preceding section have the effect of constricting the time available to organizations. Lawfare strategies like defamation suits and prosecutions and selective audits; burdensome requirements that take time and resources to comply with; and restrictions on activities imposed under the guise of exigency, anti-terrorism or ‘foreign agent’ laws all serve to use up the manpower resources of organizations at the expense of their substantive work. Reforming statutory frameworks around defamation is one self-evident necessity in this regard. In addition to clarifying concepts and closing loopholes, thought can be given to redressing mechanisms to ensure the right to challenge decisions and receive an adequate remedy. The very fact that remedial mechanisms are available may reduce the incentive and thus the likelihood that backsliding regimes will rely on these tools in the first place. Interpretive guidance can be added to guide decision makers and courts in deciding borderline cases. And even the act of amendment can signal to a judge discerning legislative intent that the previous abusive applications of the law are meant to be avoided. Similarly, reference to international standards may be useful because those standards are accompanied by a corpus of interpretation that might not yet exist in domestic case law (and to pre-empt radical interpretations by a future packed court).

Second, time is a critical element of the political process. The media and civil society rely on political processes taking a certain amount of time so that they themselves have time to scrutinize and evaluate changes, build objections and counterarguments, communicate with the public, organize and mobilize a response. At the same time, those who control political time control a precious resource that can be allocated and redirected to the advantage and disadvantage of various groups.

Backsliding has regularly featured the abuse of fast-track legislative procedures, including urgency provisions and the disingenuous use of feeding government bills to individual members to be introduced as private members’ bills. These procedures often skip steps designed to allow for public oversight and input and rush through the process in order to undermine the opposition and civil society’s ability to organize, mobilize and effectively respond. For example, private members’ bills are usually not subject to the same comment periods and impact assessments as government bills. While such procedures are necessary—in particular to respond to urgent needs—backsliding episodes have been littered with fast-tracking of legislation for non-urgent subject matters and often for laws of critical importance for safeguarding the separation of powers and free and fair elections.

The abuse of fast-track legislative procedures has been facilitated by the fact that urgency procedures are often triggered by a simple majority or even by the presiding officer. Criteria for what constitutes ‘urgent’ may be vague or non-existent, and even where such criteria exist, whether they have been met is not usually a question that is subject to review. Requiring a non-partisan or cross-partisan body to approve a motion for processing an initiative under urgency procedures—Slovenia, for example, uses a cross-partisan committee—clarifying urgency with clear, less subjective criteria; and allowing for review can all help protect against this kind of acceleration.

A second method for protecting legislative time for oversight and scrutiny (both for the opposition and for civil society actors) is to empower the opposition with ‘pause buttons’ in certain, carefully delineated circumstances. In Sweden, for example, 10 parliamentarians can pause legislation that concerns fundamental rights, preventing legislation that undermines basic freedoms from being pushed through before the public can protect its interests. In Poland, meanwhile, the Senate—which has no real power of amendment—has a 30-day window in which to debate legislation it receives from the lower house. When PiS lost the Senate in 2019, this period became a precious window for civic space, where civil society could be brought into the Senate to engage with the process, and the opposition could build a coalition capable of winning the subsequent legislative elections.

Third, in addition to protecting the time for civil society and the media to act, consideration should also be given to proactively boosting public participation in the legislative process. International IDEA’s Inter Pares project notes five components of parliamentary public engagement: (a) information (‘providing updates on parliamentary business using tools such as websites, social media, radio or materials disseminated to communities’); (b) communication (‘enabling interaction between people and parliaments, building on information channels to enable two-way engagement’); (c) education (‘improving public understanding of parliament and its work, usually with a focus on young people’); (d) consultation (‘facilitating input into parliamentary business from the general public or specific groups’); and (e) participation (‘directly involving individuals or groups in parliamentary processes or decisions’) (International IDEA 2025). Outreach programmes, such as in Trinidad and Tobago, where representatives travel to different communities and pass the results of their consultations on to the relevant committees; programmes to proactively seek responses to public consultations, rather than passive comment periods; and the incorporation of more direct democracy tools, such as citizens’ assemblies, are all increasingly common tools that can serve the aim of public participation.

2.1.3. Information

Finally, the civic ecosystem depends on information in order to be effective. The media and civil society organizations need to be able to acquire information from the state, and, in turn, the media and CSOs play a critical role in facilitating the flow of information to the larger civic space. Common vulnerabilities in the legal and institutional framework for the information ecosystem in cases of democratic backsliding have involved freedom-of-information rights and procedures, media independence, media pluralism and prosecutions for defamation (see 2.1.2: Time).

A robust freedom-of-information (FOI) structure is crucial. In India, for example, civil society has a strong tradition of using FOI requests to further a variety of social causes (Roberts 2010; Calland and Bently 2013). Furthermore, research on the Indian case has found that collaborations between journalists, social activists and CSOs were a key factor of success in building a right-to-information agenda (Relly and Pakanati 2021). In Albania, in contrast, where oversight mechanisms like FOI systems are subject to political influence, FOI requests have often been ignored or approved but unfulfilled (European Commission 2024), and in El Salvador, the state of exception has been used to dissolve the right to government information (Human Rights Watch 2020).

Second, the populace in general relies on the media and civil society organizations for information. Two areas which have been particularly salient in backsliding cases are independence in media regulation and pluralism in media ownership. With regard to independence in media regulation, governments have used their control over appointments to and funding of public media bodies, registration of media outlets and investigation into complaints against the media to clamp down on critical media, while boosting media supportive of the government. In Poland, for example, PiS used its majority control over the National Broadcasting Council as well as the establishment of a new body, the National Media Council, to ensure that government-friendly news dominated the airwaves (Wójcik 2023). In general, regulation of the media is an area which has been less commonly included in constitutions; however, there are exceptions—such as the 1992 Constitution of Ghana, which contains a chapter dedicated to media regulation—which offer examples to consider.

Lastly, concentrated media ownership reduces the diversity and viability of independent media sources and endows a small number of people with a large amount of power over what content gets produced, creating a fertile ground for corruption and crony ownership, which are particularly prevalent where there is a lack of transparency of media ownership, partisan influence over the allocation of licences and a lack of regulation over monopolies in the media.

The European Media Freedom Act (EMFA) encourages (but does not require) EU member states to (a) introduce requirements that media service providers disclose relevant information on their ownership and (b) ‘entrust national regulatory authorities or bodies, or other competent authorities and bodies, with developing media ownership databases’ (European Union 2024b: articles 32–33). These transparency measures serve the goal laid out in the EMFA that ‘recipients of media services know with certainty who owns and is behind the media so that they can identify and understand potential conflicts of interest’, which ‘is a prerequisite for forming well-informed opinions and, consequently, for actively participating in a democracy’ (European Union 2024b: article 32). Conflict-of-interest rules will also often require that public officials declare any financial interests in media companies.

2.2. System-wide resilience

While crucial, many of the considerations above speak to changes that might be made at the statutory level. Although having good laws on an issue is better than having no law, or bad laws, anti-democratic governments can use—and have used—their control of the legislature to change or annul these laws one by one. Therefore, safeguarding the media and civil society requires broader resilience of the system as a whole, to prevent backsliding forces from gaining power or sufficiently constraining their ability to undermine institutions if they do. While there are potentially endless issues to consider in this regard, and a high degree of context dependency is involved, a few key considerations are worth highlighting.

First, it must be noted that in near misses—those cases in which backsliding has been successfully stemmed—the courts have been the primary bulwark. In Brazil, for example, fortified judicial independence prevented then-President Bolsonaro from being able to disable the judiciary as a check, and in Colombia the courts’ ability to exercise broad judicial review, such as the ability to strike down laws on procedural grounds, was a major factor in grinding the backsliding efforts of Álvaro Uribe when he was President to a halt (Gamboa 2022). Stronger protections for the judiciary, from structural considerations such as constitutionalizing the size of a court, to independence protections such as removal processes (including preventing manipulation of the retirement age), are likely a necessary, if not sufficient, condition of resilience.

Second, other institutions which can protect media and civil society sectors—ombudspersons, media commissions, the bureaucracy, human rights commissions and the opposition, to name a few—must be treated as part of an ecosystem with resilience that is greater than that of each component part. Bednar (2016) identifies three interlocking characteristics of resilient systems—diversity, modularity and redundancy. Introducing diverse institutions (diverse in mandate and in structure) allows for more sources of experimentation and also means that institutions cannot all be captured by the same means. A modular system, meanwhile, is composed of semi-independent subunits (modules) that interact with one another but retain a degree of autonomy. The use of systems allows for better damage containment: when one module is captured, other modules within the system will be more insulated from that assault and better able to continue functioning on their own. Finally, building redundancy into the system means that, ‘ [s]hould one element of the system fail, a redundant pathway, with identical functionality, can play the same role’ (Bednar 2016: 178).

Daly (2025), meanwhile, proposes considerations for resilience structured around the so-called six Bs: (a) ‘broadening the office’ (diffusing powers); (b) ‘blindfolding the backslider’ (creating a veil of ignorance where possible, such as by random rotation of judges on panels); (c) ‘balancing imbalances’ (between the executive and legislature, government and oversight institutions, or elsewhere); (d) ‘blocking bad laws’ (such as by repealing colonial holdovers like criminal defamation); (e) ‘bolstering the bureaucracy’ (protecting the impartiality of public administration); and (f) ‘building (and rebuilding) holistically’.

A system built according to these precepts will also result in a diffusion of powers, allocated between separate and occasionally overlapping institutions. Diffusing power in this way may make the system as a whole more difficult to capture, counterbalance tendencies towards aggrandizement and increase the probability that at least one institution will be able to act as an ‘island of integrity’ in the event that an illiberal regime regains power (Galvin and Shogan 2004; Carrigan and Poole 2015; Bernauer and Vatter 2019).

Finally, one notable consideration regarding resilience and the media and civil society is that a substantial number of the backsliding tactics described in Chapter 1 are carried out through sub-legislative acts or executive orders. Many decisions pertaining to regulatory fines, for example, are taken in administrative agencies; police and prosecution services are the front end of investigation and prosecution (or lack thereof); tax audits are generally pursued by tax-specific offices.

A certain amount of insulation of these bodies from partisan politics limits the politicization of these regulatory areas and reduces the possibility of political capture. As Yesilkagit et al. (2024) conclude, achieving sufficient independence may necessitate both structural independence and imbuing the bureaucracy with respect for professional norms. Although ‘a prolonged rule of populists may ultimately place the bureaucracy under full control of illiberal populists’, a bureaucracy populated by public servants who are dedicated to preserving the rule of law ‘will significantly increase the transaction costs of populists gaining such a degree of control’ (Yesilkagit et al. 2024: 418).

Building on this insight, Yesilkagit et al. (2024: 420) recommend a strategy for insulating the civil service from backsliding that includes ‘(1) the training and socialization of civil servants, especially bureaucratic leaders, who are deeply immersed in the norms and values of the liberal democratic state; (2) granting autonomy to the civil service and state agencies; and (3) reinforcing internal monitoring and regulatory mechanisms to elevate the bureaucracy as a fourth branch of government.’

In addition, recruitment to the civil service ‘should be defined by equality of access to public posts and selection based on merit, fair and open competition and an absence of discrimination’, and hiring systems should be transparent and their standards clear (Council of Europe 2000).

An increasing number of countries constitutionalize a public services commission tasked with safeguarding the independence and functioning of the public service. The Public Service Commissions of Kenya, Nepal and Papua New Guinea, for instance, are charged with the creation of, and appointment to, offices in the public service, as well disciplinary control, and their members generally receive heightened protection from removal. It is important to ensure that the governing majority does not control a majority of the members of these commissions (Independent State of Papua New Guinea 1975: article 190; Republic of Kenya 2010: article 233; Federal Democratic Republic of Nepal 2015: article 242).

The debate surrounding building system-wide resilience is wide-ranging and complex, and the above discussion offers a few examples of proposals and relevant considerations. For further considerations and suggestions, see International IDEA’s report Designing Resistance: Democratic Institutions and the Threat of Backsliding (Bisarya and Rogers 2023).

The underlying point here is that systemwide resilience, rather than targeted protections, may be the most crucial component of protecting media and civil society from future backsliding.

Considerations for strong institutional design are useful only to the extent that they can be implemented. As discussed in Chapter 2, once an illiberal, anti-democratic leader has come to power, it is often too late. Because backsliders can use rather than simply break the law, legal means of resistance have fewer hooks on which to gain traction. Even if backsliders fail to entrench themselves fully, reversing the damage once they are out of office can present the paradox of having to fix bad legal structures from within those same structures. Further, resistance to backsliding depends on a pre-existing robust legal and institutional architecture for civic space.

Much like the conventional wisdom that the best time to fix the roof is before it rains, the time to think about strengthening institutions is before backsliders have an opportunity to enter office. In an era of rising populist attitudes (often correlated with backsliding tendencies) and disillusionment with and even waning public support for democracy, governments must act proactively to safeguard their democracies while pro-democratic actors are at the helm. Inspiration might be taken from countries such as Germany and Norway, which have recently launched projects to proactively address vulnerabilities in their constitutional frameworks in light of lessons learned from backsliding elsewhere and adopted constitutional amendments to address these vulnerabilities (Oancea and Bisarya 2025).

This report offers considerations for strengthening institutional design in one key aspect of the democratic structure: the twin diagonal accountability institutions of the media and civil society. As democracy’s watchdogs, they must be protected and equipped to aid in holding illiberal, would-be backsliders to account should the need ever arise.