Abstract
Introduction
Singapore is a well-known illiberal democracy, ruled by one party, the People's Action Party (PAP), since 1959. Despite holding more than 12 multi-party general elections, the hegemonic PAP governs with ‘soft authoritarianism’ and resists embracing greater political pluralism (Fukuyama, 1992; Means, 1996). Faced with declining vote shares, disinformation challenges and a pandemic outbreak in the foreign workers’ dormitories, the ruling elites appear to have returned to what Silverstein (2008) terms ‘rule by law’ to suppress dissent. In 2021, the US Democratic Party refused to recognize Singapore as a democracy and excluded it from a 110-nation democracy summit. Critics now warn that the small city-state is backsliding or becoming more autocratic (Abdullah, 2020; Croissant and Lorenz, 2018; Tan, 2020). Is Singapore's democracy in fact in decline?
Based on global democratic indicators from agencies such as Freedom House and the Economist's Index of Democracy, Singapore's level of democracy is ranked somewhere in the middle of the pack (Freedom House, 2020; Economist, 2021a). Regionally, the country's level of democraticness does not appear to be in trouble compared to its neighbours such as the Philippines or Indonesia, based on Varieties of Democracy (V-Dem)'s conceptualization of democracy. 1 In fact, V-Dem's 2020 Liberal Democracy Index shows that Singapore's quality of democracy has experienced little change in the last decade (Coppedge et al., 2020).
Yet, democratic backsliding may manifest in different ways, and in both democracies and autocracies. As a well-known adage goes, democracy dies in darkness. The decline in quality of democratic institutions may occur slowly and weaken covertly over time (Bermeo, 2016; Croissant and Haynes, 2021; Diamond, 2021). Our article engages in this debate to investigate to what extent Singapore's hybrid regime is under threat. We do so by disaggregating the measures of democracy in the international indices and tracing
Focusing on the last two decades, our study finds that while most indicators of vertical and horizontal accountability remain stable in Singapore, diagonal accountability – especially media and online freedom – has suffered significantly in the last few years. While incidences of democratic backsliding have occurred before 2000, our study focuses on events after 2000, the year that coincides with the global ‘third wave of autocratization’. 2 Focusing on the past two decades will allow us to identify the significant incidences or critical junctures that led to democratic backsliding over time (Capoccia and Kelemen, 2007). The PAP government has returned to using the law to muzzle critics and change public opinion. While the formal institutions associated with vertical and horizontal accountability through the judiciary and legislature are largely stable, the decline in media freedom using new forms of digital repression represents a growing intolerance and defensiveness to criticism. This is especially the case during the COVID-19 pandemic in the last two years.
Our findings dovetail with other structural, institutional or crisis-focused explanations of democratic decline in Singapore. For example, Abdullah (2020: 1134–1135) posits that the lack of reformers within the PAP and the absence of a strong opposition within the legislature have exposed the authoritarian tendencies in the current leadership. While quantifiable evidence on self-censorship is elusive, studies find that punitive actions against academics and activists to curb media freedom have a ‘chilling effect’ on a small city-state where the government is a major employer (Abdullah, 2020; Fong, 2018; George et al., 2021). While some argue that the PAP's adoption of social media marketing strategies and arbitrary media laws has dampened opposition support (Tan, 2020), the government's heavy-handed ‘top-down’ pandemic management approach has drawn criticism for constraining mobility, information and assembly (Woo, 2021; Yuen et al., 2021).
Singapore is not politically troubling or in a cusp of regime change like other cases discussed in this Special Issue such as Malaysia, Thailand or Myanmar. Yet, it remains timely to focus on it for several reasons. First, Singapore highlights the perennial methodological problem of studying the quality of democracy in hybrid regimes. Instead of focusing on the overarching outcome or regime type, we address this problem by disaggregating the components of a democracy and identify areas of decline and growth. Second, Singapore is an ideal example to showcase how an illiberal, technocratic regime behaves in a crisis such as the COVID-19 pandemic, when state power may be quickly expanded without opposition oversight. As part of the wider pandemic policy implementation, the PAP government, like most governments in Asia, has rapidly increased its capacity for online surveillance and censorship through new cyber and technology laws (Mahapatra and Sombatpoonsiri, 2021). The long-term implications of these new laws remain unstudied. Third, Singapore offers a comparative case to see if it has hindered flow of information, expanded surveillance powers or imposed internet regulations like its regional neighbours (Shahbaz and Funk, 2020). While Singapore has long experienced constraints to press freedom, it remains useful to investigate if pandemic disinformation laws present new threats such as constraining citizens to make informed health decisions or undermining public confidence in governments.
Our arguments are organized as follows. The first section examines the concept of democratic accountability and backsliding, especially in hybrid regimes such as Singapore. The second section identifies the indicators of democratic accountability and backsliding in Singapore especially in the last two General Elections (GEs) and during the management of the pandemic. We conclude with implications for democratic backsliding and areas for future research for hybrid regimes.
Democratic backsliding and accountability in hybrid regimes
Scholars often face problems classifying Singapore, which reflects methodological issues with studying hybrid regimes (Bogaards, 2009; Collier and Levitsky, 1997). As a hybrid regime with both democratic and authoritarian traits, labels from both ends of the spectrum have been used.
3
For example, based on V-Dem v12 data set, Singapore was scored as an “electoral autocracy” as it only earned a score of 1 (from 0–3) on V-Dem's Regimes of the World index (Coppedge et al., 2022). However, Singapore was classified as a “flawed Democracy” by the Economist's Democracy Index (EIU, 2021). The lack of consensus in classifying Singapore reflects divisions in the democratization literature on the criteria to measure the quality of democracy in intermediate categories (Bogaards, 2009; Morlino, 2009).
4
We address this methodological challenge by moving away from the futile dichotomous measures of political regimes and focus on the degrees of participation, contestation and accountability (Dahl, 1971; Munck, 2006). This means paying attention not to the regime outcome but to the components of the regime type and
Following the position taken by this Special Issue, we view the quality of regime in a continuum, or in a ‘more’ or ‘less’ democratic lens. 5 Viewed this way, an electoral autocracy that holds multi-party elections such as Singapore could also be democratically backsliding, or on the path to autocratization – two sides of the same coin. Broadly, democratic backsliding refers to ‘a decrease in competitiveness of the electoral playing field due to increasing concentration of power in the hands of the incumbent executive, relative to other actors’ (Dresden and Howard, 2016, p. 1123). The terms democratic backsliding, decline or regression are thus used inter-changeably, referring to the regime losing its democratic qualities or attributes.
Ensuring accountability between the state and citizens is central to democracy (Dahl, 1989; Przeworski et al., 1999; Schmitter and Karl, 1991). 6 Accountability ensures that institutions and politicians act in the public interest (Mechkova and Pernes, 2019). It also means that the power of elected officials and government is limited or constrained by electoral and other provisions (Moncrieffe, 2001). Conceptually, backsliding studies often focus on formal institutions, through the lens of vertical and horizontal forms of accountability. Analysis is typically on the erosion of free and fair elections or lack of parliamentary and judicial constraints on the executive (Andersen, 2019). But as democracy is a multifaceted concept, efforts have moved beyond the dichotomy and included diagonal accountability, which considers less formal institutions and the role of non-state actors in holding government accountable (Lührmann et al., 2020). Newer conceptualizations now include formal obligations and innovative mechanisms associated with civil society, media and non-state actors (Bovens, 2007; Goetz and Jenkins, 2005).
Accountability matters in hybrid regimes, as there is a minimal democratic norm or expectation for the citizens to hold their government to account (Przeworski et al., 1999). Regardless of the degree of electoral competitiveness or participation, constraints are placed on democratic institutions or the executive, no matter how small (Ekman, 2009; Levitsky and Way, 2010; Wigell, 2008). This is because hybrid regimes rely on elections, the act of parliamentary debates and public opinion to maintain their clout of legitimacy and push through legislation (Magaloni, 2008). Unlike closed autocracies or single-party states, hybrid regimes have to practise some form of self-restraint and appear restrained from below to maintain electoral legitimacy. However, due to a penchant for repression, many hybrid regimes remain vulnerable to democratic backsliding. The forms of backsliding in this regime type may include expanding formal and informal executive powers (i.e. abolishing term limits; pushing through draconian bills) or restricting institutions and actors who can challenge the executive (e.g. jailing of opposition, media censorship, etc). Additionally, backsliding can also occur gradually, weakening democratic processes and institutions such as by holding unfair elections, violating civil liberties or ensuring ineffective constraints on the executive (Abdullah, 2020; Dresden and Howard, 2016).
Method and data
To examine the quality of democracy in Singapore, we draw on Luhrmann et al.'s (2020) work and V-Dem's democratic indicators 7 to investigate the key forms of accountability, namely: 1) vertical accountability, or the extent to which governments are accountable to citizens through elections and political parties; 2) horizontal accountability, or the extent to which governments are accountable to other state institutions including the legislature and judiciary; and 3) diagonal accountability, the extent to which governments are accountable to the media and civil society (Lührmann et al., 2020). 8 The success of diagonal accountability is correlated to and dependent on other accountability institutions. For example, civil society mobilization is necessary for other accountability mechanisms to be effective (Lührmann et al., 2020; Mainwaring and Welna, 2003).
To assess if the quality of democracy changed overtime, we use data from international indices such as V-Dem, Freedom House and the Economist's Intelligence Unit's Democracy Index (Coppedge et al., 2021; Economist Intelligence Unit, 2020; Freedom House, 2021b), along with other qualitative data sources from news articles and reports. We disaggregate and highlight specific indicators that measure vertical, horizontal and diagonal accountability. For example for vertical accountability, or the existence and quality of elections, we use V-Dem's Clean Elections Index for ‘free and fair elections’ (Coppedge et al., 2021).
To measure horizontal accountability or judicial and legislative oversight, we use V-Dem's indicators for judicial constraints and the legislative constraints on executive (Coppedge et al., 2021). For diagonal accountability, V-Dem indicators related to media freedom and data from V-Dem's Digital Society Project (DSP), including indicators on social media monitoring, social media censorship and other forms of internet surveillance, are used (Mechkova et al., 2022). We also use V-Dem's Pandemic Backsliding Index, especially indicators related to media freedom and freedom of expression (Edgell et al., 2021). We supplement these country indicators with news reports and academic literature to assess changes in Singapore's quality of democracy over the last two decades. This timeframe coincides with the ongoing debate on the ‘third wave of autocratization’ that occurs after 2000 and allows us to analyse the impact of the COVID-19 pandemic on its democratic governance since 2020.
Weak vertical accountability
Since 2000, Singapore has held five GEs, with an average of nine political parties competing in each. While the PAP enjoyed a high of 75% vote share in 2001, its popular vote share dipped to 61% in the last 2020 GE. The PAP's vote share is on a declining trend, despite the exceptional blip of support it enjoyed in the 2015 GE, largely attributed to the outpouring of sympathy and nostalgia due to the death of its former PM Lee Kuan Yew (Weiss, 2015).
Vertical accountability refers to the extent to which governments are accountable to citizens through elections and political parties. The PAP's declining popular support and high seat shares offer two key points on this dimension. First, elections do matter. Singapore's electorate holds the government to account and shows its disaffection or support through the ballot box. Second, the PAP's legislative supermajority remains exaggerated by its unusual electoral system. Singapore's electoral rules that involve multimember districts (MMDs) elected through plurality bloc voting for party slates (Group Representative Constituency scheme) facilitates gerrymandering and district magnitude changes and magnifies electoral disproportionality (Tan and Grofman, 2018). This means the vote shares earned by the respective parties do not translate proportionately to seat shares. Hence, while all the opposition parties earn around 30 per cent vote share in each election, they only get seven to nine seats in each election due to the electoral system's strong majoritarian effect.
Singapore's elections are notoriously free but not fair, as they are systematically disadvantageous to the opposition parties (EBRC, 2015; Morgenbesser, 2017; Oliver and Ostwald, 2018). Existing literature finds that electoral malpractices concerning the short campaign period, gerrymandering and arbitrary changes in sizes and number of constituencies negatively impact opposition parties (Lee and Tan, 2015; Tan, 2015). The lack of specific data on the electorate and voting behaviour at the national and district level also continues to hamper voter accountability. 9
Based on V-Dem's democracy indicators, electoral fairness has dipped slightly in Singapore. As Figure 1 shows, V-Dem's indicator for ‘free and fair elections’ is at its lowest ebb, post 2015 GE. This indicator takes all aspects of the pre-election period, election day and post-election period into account (Coppedge et al., 2020). The drop indicates concern over irregularities in the election or in the quality of competition, for example the release of the electoral boundaries report by the Electoral Boundaries Review Committee (EBRC) making significant changes to the boundaries, number and sizes of constituencies, only a month before the 2015 election.

Vertical accountability indicators in Singapore (2010–2021).
Similarly, in the 2020 GE, the electoral boundaries were changed with very short notice on electoral rule changes and restrictive campaign rules (Tan, 2021). Worse, the election was held in the middle of the pandemic, despite widespread disapproval (Tan and Lee, 2021). The 2020 election imposed strict conditions on the voters, parties and voting process. 10 It was also the first time when campaigns were permitted on mainstream and online media, under the new controversial Protection from Online Falsehoods and Manipulation Act (POFMA) (Government of Singapore, 2019) and Personal Data Protection Act (PDPA) media laws (PDPC, 2012, 2017). Physical rallies were banned and no assembly of more than five people for walkabout visits was permitted, which posed challenges to resource-poor parties. 11 While the PAP expected to gain from holding a pandemic election, it lost a group constituency and a cabinet minister to the opposition Worker's Party (WP). The WP won a total of 10 seats, the most any opposition party has secured since 1963. Given the opposition's support, one wonders if the PAP would have lost more seats if the election had been conducted under less-restrictive conditions. Regardless, given the highly tilted playing field as discussed earlier, formal institutions such as elections and parties remain weak vehicles of vertical accountability to check the government.
Dip in horizontal accountability
The PAP's horizontal accountability to major state institutions such as the legislature and judiciary remains somewhat the same. This is despite a slight improvement in V-Dem's 2020 score for Singapore's legislative constraint on the executive due to an increase in opposition-held seats in Parliament. The PAP holds 83 out of 93 elected legislative seats after 2020 GE. The V-Dem score for Singapore's legislative constraint on the executive increased slightly from 0.44 in 2010 to 0.54 in 2017–2020 (Coppedge et al., 2021). 12 Based on Freedom House scores for its legislative checks on the executive, Singapore scores 2 (out of 4), citing parliament's inability to serve as an effective check due to the PAP's strict party discipline and institutional dominance (Freedom House, 2021a). Taken together, the PAP's legislative supermajority, use of party whip and threat of law suits against the opposition indicate that the legislature is a weak check on the government.
Singapore's V-Dem score on judicial constraints on the executive has declined slightly from 0.743 in 2017–2019 to 0.709 in 2020 (Coppedge et al., 2021). 13 While Singapore ranks higher than most countries in Southeast Asia on judicial constraints on the executive, judicial independence remains a cause for concern. In particular, the PAP's ability to push through controversial laws while sustaining the ‘government's standing as a “rule of law” regime’ continues to receive criticism (Rajah, 2012a: 13). In particular, the use of libel, ‘contempt of court’ or defamation suits to intimidate or suppress opposition parties and critics remains in practice. Although countries are moving towards making defamation a civil instead of criminal offence, Singapore still recognizes defamation as a criminal offence to impose limitations on opposition voices in parliament (and outside of parliament) (Kenyon et al., 2014). In these cases, judicial review is extremely rare (Chua and Haynie, 2016).
Given the PAP's sheer hegemonic size, it is also difficult for the elected opposition leaders or even its backbenchers to challenge or hold the PAP leaders to account in parliament. To begin with, the PAP's numerical presence and ability to field questions are proportionately more than all the opposition leaders combined. To date, the 10 elected opposition members from the WP are the largest cohort the Parliament has seen in post-independent Singapore.
Second, the PAP has a strong party whip that ensures party discipline during voting of bills. 14 Unless the party whip is lifted, very few PAP MPs will vote against the party. As legal scholar Tan (1989: 165) notes, ‘the institution of the party whip and the culture of obedience to one's party leaders that constrains parliamentarians means that the cabinet controls parliament rather than vice versa’. There are very few occasions where the whip is lifted, usually when a bill concerns religious, moral or ethical issues. For example, in 1969, the whip was lifted for the passing of the Termination of Pregnancy Act to legalize abortion. 15 More recently, the whip was lifted in 2002 for a debate related to junior college and upper secondary education, and again in 2009 for the passing of amendments to the Human Organ Transplant Act. The whip was also lifted in 2017 during PM Lee's Ministerial Statement to address his family's very public dispute over the legacy of their family home (Cheong, 2017). While the lifting of the party whip in this instance allowed for all parliamentarians to openly discuss the issue, there was no vote taken since a Bill was not introduced. In other words, lifting the whip was more a symbolic gesture than any substantive action to allow for greater opposition.
Table 1 highlights strict party discipline for PAP members, showing the voting results at the second reading for four selected legislations introduced in the last decade. For all four bills introduced, no PAP MPs abstained from voting or opposed the Bill from passing to the next stage. All opposition votes came from the opposition MPs and one non-constituency MP (NCMP), while nominated MPs (NMP) chose either to vote with the PAP representatives or to abstain (except for two NMPs voting against the Criminal Law Bill). Despite an improvement in opposition seat shares in the 2020 GE, the PAP still controls the parliamentary agenda with a legislative supermajority and steam-rolls unpopular policies without much opposition challenge. Most bills get passed very quickly without rigorous debate.
Selection of voting records in Singapore parliament (2015–2020).
MP = Member of Parliament; NMP = Nominated MP; NCMP = Non-Constituency MP; Opp. MP = Opposition MP
Source: Extracted from Singapore Parliament Reports.
Even with the introduction of NMP and NCMP schemes in 1988 to improve the quality of debate in Parliament, opposition leaders are observed to ‘shy away from challenging the ruling party’ for fear of litigation and intimidation (BTI, 2021). This is unsurprising given the negative media coverage over a parliamentary investigation into an opposition WP member, Raeesah Khan, for lying in Parliament (Goh, 2021). Further, WP chair Sylvia Lim also chose not to pursue a police investigation following suspicions that her phone had been hacked by state-sponsored attackers (Tay, 2022). In-depth interviews with opposition MPs and non-elected MPs will help to determine the extent of restraint from challenging the dominant PAP members.
Third, Singapore's electoral system that allows non-elected members such as those appointed through the NMP and NCMP to sit in the Parliament has not significantly increased opposition challenge in practice. Presently, there are two ‘best-loser’ opposition members from the Progress Singapore Party (PSP), appointed through the NCMP scheme, and another nine non-partisan MPs who sit in the House through the NMP scheme. While the non-elected MPs can participate in debates and introduce bills (which rarely occurs in practice), they are not allowed to vote on constitutional amendments, money and supply bills, no-confidence votes or the removal of the president from office (Tan, 2013). In fact, Rodan argues that the scheme is ‘not intended to harness electoral politics and opposition parties to authoritarian reproduction, but is another element of the broader project of actively fostering alternatives to such competitive politics based on democratic representation’ (Rodan, 2009: 441).
Critics such as current PSP leader Tan Cheng Bock insist that the NMP scheme undermines the representativeness of elected MPs and diminishes accountability to constituencies. As he said, the issue of accountability is a ‘very important’ one and ‘when you go to Parliament to speak, you speak with conviction, representing a sector of the population who voted for you … If you don't have that responsibility … you’ve got no mandate’ (Romero, 2020). Other critics cite issues associated with the non-transparent nomination process (Lim, 2021). While NMPs have been praised for contributing to debates in parliament and for raising issues deemed too ‘sensitive’ for elected representatives (Abdullah, 2016), their efforts rarely result in substantial policy outcomes. In 2019, for example, three NMPs proposed amendments to the controversial POFMA Bill, but they were all rejected (Lim, 2019). As seen from Table 1, NMPs either voted with the PAP or abstained, without challenging the PAP-proposed legislation. This thus raises the role and utility of the Parliament as a key institution of horizontal accountability.
Despite the PAP's pride in the judiciary's reputation for independence and efficiency (Tan, 2009), its judicial review process has been criticized. Singapore is well known internationally for its strong rule of law, if defined by the extent to which agents have confidence in and abide by the rules of society, especially in areas of contract enforcement, property rights, the police and the courts, as well as the likelihood of crime and violence. Indeed, it often scores well on the World Bank's Rule of Law Index, receiving the highest ranking from 2010–2020, and was 12th out of 128 countries in the 2020 World Justice Project's Rule of Law Index (World Justice Project, 2020). Yet, given the strong role of the government and communitarian ‘green-light’ style of judicial review, the government's accountability to the judiciary remains rather opaque ‘where the courts are not the first line of defence against administrative abuses of power’ (Lee and Neo, 2021: 135). In fact, Freedom House (2021a) allocates a score of 1 (out of 4) to Singapore for judicial independence, highlighting the issue of judicial appointments and the close relationship between the courts and the executive.
Additionally, the family dispute over Lee Kuan Yew's former residence at Oxley Road also demonstrates the practical inability to hold the executive to account, despite formal provisions to do so (Vadaketh, 2022). While the investigation into the dispute could have been conducted solely by the courts, it was instead led predominantly by the executive political machinery. Further, the appointment of individuals closely related to the Lee family or the PAP to judicial mechanisms could also hinder horizontal accountability. The appointment of Lucien Wong, PM Lee Hsien Loong's former personal lawyer, to Attorney General is an example of a potential lack of independent scrutiny and oversight in the appointments that raises questions about judicial independence and executive accountability (Ng, 2017).
Singapore's judiciary is widely perceived to act professionally and impartially in business-related cases. Yet, it is less so when it comes to civil-political cases, when the PAP government or leaders launch defamation and lawsuits against the opposition leaders and dissidents (Freedom House, 2021a; Thio, 2012; Rajah, 2012b). Further, its process of judicial review rarely, if ever, finds any legislation to be unconstitutional (Chua and Haynie, 2016). Since the onset of the COVID-19 pandemic, there has only been one reported instance of judicial review of government action, when the courts sided with the PAP government (Lee, 2020). 16
The PAP has a track record of suing opposition leaders and international media outlets that are critical of its government (Han, 2021; Human Rights Watch, 2018; Sim, 2011). Domestically, many opposition leaders have been sued for defamation, bankrupted, jailed and barred from competing in elections. 17 The PAP has been known to pursue litigation until the opposition member is declared bankrupt and therefore cannot run for political election or hold political office. Prominent targets of PAP-driven lawsuits against opposition members have included WP leader Joshua Benjamin Jeyaretnam in 2001 and SDP leader Chee Soon Juan in 2011 (Andrews, 2015), along with more recent examples against People's Voice party members Leong Sze Hian and Lim Tean in 2021 and 2022 (Alkhatib, 2022; Jaipragas, 2021).
The repression of opposition leaders and dissidents appears to have been exacerbated by the introduction of the Administration of Justice Act in 2016, which criminalizes criticisms of the judiciary as not impartial, 18 and the Protection from Online Falsehoods and Manipulation Act (POFMA) in 2019, which is intended to tackle fake news but has been used to stifle dissent online (Abdullah, 2020). In 2018, an SDP opposition leader, John Tan, was charged with contempt of court under the Administration of Justice Act, by ‘scandalizing the judiciary in posts … on Facebook’ (Lum, 2018). Tan was fined S$5000 and disqualified from competing in the 2020 GE (Tang, 2019).
Apart from fighting disinformation, POFMA has also been used against opposition members, often requiring them to post correction notices to their Facebook and YouTube posts while campaigning (Geddie, 2020). While the PAP government has denied suggestions that POFMA is used to stifle dissent and criticism, opposition members such as Lim Tean, who has received several correction direction notices prior to the 2020 GE, claim that the notices are ‘intimidation tactics’ to distract the opposition during elections (Low and Wong, 2020).
Thus far, the PAP leaders have not lost a defamation suit against opposition figures and critics. This track record has fostered a climate of self-censorship in political speeches and online media (Gomez, 2000; Ong, 2021). More recently, significant lawsuits against opposition members include the S$33.7 million lawsuit against opposition WP leaders Low Thia Khiang, Sylvia Lim and Pritam Singh, for breaching their fiduciary duties in mismanaging the town council in Aljunied in October 2019 (Today, 2019). As discussed below, the PAP government also uses defamation and libel suits to stifle dissenting voices among the media and civil society, including prominent bloggers and online journalists.
Backsliding in diagonal accountability
Singapore's diagonal accountability has suffered most since 2010. Measures of diagonal accountability include indicators associated with civil society; media freedom, including censorship efforts and journalist harassment; freedom of expression, including cultural and academic freedom; and government repression. While Singapore is notorious for constraining its media and less than vibrant civil society, the repression over media and academic freedom 19 has worsened in the last few years (Freedom House, 2021b; George, 2011). Figure 2, based on V-Dem data, shows Singapore's most serious decline is in ‘government censorship efforts-media’, which has experienced a significant drop since 2019. This drop is consistent with other measures and reports on media freedom in Singapore (Freedom House, 2021a; RSF, 2021). The reliance on legislation such as POFMA and defamation lawsuits is expected to foster self-censorship among journalists and activists. While it is difficult to verify the degree of media self-censorship, Singapore's mainstream media is traditionally seen as muzzled and sterilized due to political pressure, media control by the PAP and state-friendly organizations (Fong, 2018; George, 2019; Gomez, 2000; Han, 2020).

Diagonal accountability indicators (Singapore).
It is unsurprising that Singapore fares poorly in international media indices. What is concerning is that the rankings have worsened in recent years. For example, V-Dem data and RSF's Press Freedom Index indicate a decline in media freedom and an expansion of self-censorship among journalists and media in the last few years (Coppedge et al., 2021; RSF, 2021). In 2020, Singapore was classified as ‘very bad’ for press freedom and dropped its ranking to 160th out of 180 countries (RSF, 2021). Freedom House has consistently given Singapore a score of 2 (out of 4) for a free and independent media, citing issues of state-friendly reporting and self-censorship (Freedom House, 2021a). International indices highlight the strong government monopoly over mainstream media and, more recently, the introduction of vaguely worded media laws to regulate online content. 20
There are concerns over the rise of digital repression in Singapore. Digital technologies, including social media, have provided new opportunities in the past decade for citizens to participate in civil society. While early studies viewed social media as a liberalizing tool, especially for its ability to increase political participation and discourse, solve collective action problems and spur mobilization (Diamond and Plattner, 2012; Howard, 2001), recent work shows that digital technologies have been used as tools of repression (Feldstein, 2019; Morozov, 2012).
In Singapore, the PAP government has adapted digital marketing strategies in innovative ways. Studies find it has used digital media consultants, ‘internet brigade’ (IB) and influencers to engage in ‘guerrilla-type activism’ to suppress opposition and online activism (Tan, 2020: 4). Data from V-Dem's DSP also indicate that the PAP government engages in a range of concerning digital repression activities, including social media monitoring, the censorship of social media (in practice) and filtering of the internet (in practice). For example, Figure 3 shows a concerning trend since 2000 for indicators related to: 1) the surveillance of political content on social media; 2) government censorship of political information online by filtering; 3) government censorship of political content on social media; and 4) the potential for arrest if citizens post online content that is counter to the government and its policies.

Digital society project: Digital repression (Singapore).
Since 2000, Singapore has raised levels of social media monitoring and censorship (Mechkova et al., 2022). Specifically, efforts have been made to monitor and censor online speech that is critical of the government (Ang and Nadarajan, 1996; Ong, 2021; Seng, 2008). In 2015, for example, PM Lee won a defamation case against an activist and blogger, Roy Ngerng, for blogging about the PM misappropriating funds (Columbia Global Freedom of Expression, 2014). In 2016, Law and Home Affairs Minister K Shanmugam accused Terry Xu and the Online Citizen of orchestrating a smear campaign against the police that some have equated to harassment (Lee, 2016). Increases in digital repression and decreases in media freedom are attributed to: the introduction of two new media legislations (POFMA and the Foreign Interference (Countermeasures) Act); and the increase in targeted defamation and other lawsuits against bloggers and online citizens such as the alternative news outlet the Online Citizen 21 and political dissident Alex Tan.
In 2019, Singapore passed an anti-fake news law (POFMA) aimed at countering false and misleading information online. Since then, POFMA has been used regularly to remove and censor information and opinions on a variety of topics that are considered false, misleading or undesirable by the government (Abdullah, 2020; Asia Centre, 2020). As Table 2 highlights, POFMA has been used approximately 18 times against civil society groups or individual activists since it was established, and at least 19 times against media or news outlets (Singapore Internet Watch, 2021). Many of these instances are correction direction notices, which do not involve fines or criminal sanctions but require the recipient to publish a correction notice in their post (Singapore Legal Advice, 2022). Regardless, POFMA has been used to stifle online opposition voices and diminishes mechanisms of diagonal accountability. For example, controversial activist and journalist Alex Tan has been targeted numerous times and has had multiple Facebook pages blocked under POFMA (Singapore Internet Watch, 2021).
POFMA prosecutions/notices (2019–2021).
‘Civil society group or figure’ includes individual journalists and activists such as Alex Tan.
Source: Compiled by authors from Singapore Internet Watch (2021).
The introduction of the Foreign Interference (Countermeasures) Act (FICA) in 2021 also strengthened government power over media regulation, including the ability to compel ISPs and social media platforms to provide user information, block content and remove applications used to spread content deemed hostile (Han, 2020; Plannet, 2021). The FICA limits media freedoms due to its broad language and lack of clarity on the extent of its powers, intentions and application (Liang, 2021). While the frequency of prosecution of critical speech is lower in Singapore compared to in Thailand or Vietnam, 22 the Singapore government is believed to ‘deliberately target only a small number of key activists for legal prosecution’ that are heavily publicized, thereby fostering a culture of self-censorship (Freedom House, 2021b).
The Administration of Justice Act, FICA and the POFMA have set precedents to prosecute online activists. In 2021, PM Lee won a high-profile defamation suit against Leong Sze Hian, a blogger who shared an article on his Facebook page in 2018 about the PM being involved in ongoing investigations in the 1MDB scandal in Malaysia (Han, 2021). In September 2021, Terry Xu from the Online Citizen and blogger Rubaashini Shunmuganathan were sued for defamation by PM Lee for their article on Lee's family disagreement over property. Both bloggers were ordered to pay a combined sum of S$210,000 in damages (Lin, 2021).
In addition to these lawsuits, the PAP government has also suspended high-profile critical news sites. In September 2021, the Infocomm Media Development Authority of Singapore suspended the licence of the news website the Online Citizen, citing their failure to declare funding sources (International Press Institute, 2021). The suspension means that the Online Citizen will not be able to publish new articles or post on its social media channels, and is prohibited from operating under any other broadcasting service (International Press Institute, 2021). These cases demonstrate the difficulty for media and civil society in questioning the government online.
Beyond the digital sphere, examples of pressure on and harassment of civil society and media actors in Singapore abound. These include the arrest and imprisonment of activists such as Jolovan Wham for minor acts of protest (Jha, 2020); of LGBTQ + activists for a minor protest in front of the Ministry of Education in 2021 regarding discriminatory treatment in schools (Low, 2021); and of climate activists in 2020 advocating for action against climate change (Mahmud, 2020). In 2022, police interrogated human rights activists Kirsten Han and Rocky Howe for potential offences under the Public Order Act (2009) related to their work advocating against the death penalty in Singapore (Human Rights Watch, 2022). Ultimately, the PAP government has proven adept at devising new laws, navigating the digital landscape to its benefit (Tan, 2020) and stifling opposition voices in order to further its hold on power and diminish mechanisms of diagonal accountability.
Pandemic backsliding
V-Dem's Pandemic Backsliding Index shows that all nine countries in Southeast Asia faced restrictions in media freedom since early 2020, the start of the pandemic. 23 For example, anti-fake news legislation, such as POFMA in Singapore, was used in the region to combat the spread of misinformation (prior to and throughout the pandemic) but also to stifle dissenting voices among activists, journalists and those critical of the government especially in countries such as Thailand, Cambodia, Malaysia and Indonesia (DigitalReach, 2021; Edgell et al., 2021; Kurlantzick, 2020; Mahapatra and Sombatpoonsiri, 2021). The pandemic has fuelled digital repression and threatened diagonal accountability. While the region has experienced issues related to press freedom and media censorship, the pandemic has exacerbated these and contributed to democratic backsliding.
Singapore is no exception. Since the onset of the COVID-19 pandemic in 2020, two aspects of its democracy have suffered setbacks, especially from March 2020 to June 2021. As V-Dem's pandemic backsliding data shows, Singapore experienced two major violations in its democratic standards for emergency provisions 24 in response to the pandemic, in Type 1: Discriminatory measures and Type 7: Restrictions of media freedom. 25 See Figure 4. Issues associated with Type 1 include government-imposed excessive restrictions on movement and the detention of hundreds of migrants in unsanitary or crowded detention camps (Amnesty International, 2020; Young, 2021). Its government's response to the pandemic has resulted in an increase in cyber surveillance and using health data collected from COVID-19 contact-tracing applications for non-health reasons. The government acknowledged the data collected through its TraceTogether app had been accessed by law enforcement for criminal investigations, despite earlier promises stating otherwise.

V-Dem's pandemic backsliding index (Singapore).
Violations related to media freedom correspond with the use of POFMA and pandemic legislation to stifle dissent (Chua, 2019). Since POFMA was introduced in 2019, it has been used at least 30 times to combat false information related to COVID-19 (Singapore Internet Watch, 2021). While many of these instances can be viewed as protecting the public interest from false information about the virus, other instances have been judged as ways to curb government criticism during the pandemic. In 2020, for example, POFMA was invoked against five online media platforms for using statements by an infectious diseases expert and opposition leader, Dr Paul Tambyah, who claimed that the ministerial committee in charge of tackling COVID-19 was prioritizing the 2020 GE instead of public health concerns (Edgell et al., 2021). The use of POFMA by the government during the pandemic serves as an example of how dissenting voices among political parties, the media and civil society are stifled, further undermining diagonal accountability. The government's use of digital surveillance through COVID-19 contact-tracing apps is also concerning, as it remains unclear how data collected during the pandemic will be used. The use of digital surveillance and lack of regulations over health data could threaten civil liberties and contribute to democratic decline.
Conclusion
Regionally, the quality of democracy in Singapore appears stable, unlike the volatility seen in its regional neighbours such as Thailand or the Philippines. However, the rise of disinformation, leadership succession crisis and COVID-19 pandemic have posed challenges to the PAP's technocratic, soft-authoritarian governance. Disaggregating and examining Singapore's component of democracy reveals that mechanisms of diagonal accountability related to media and civil society have especially declined. Vertical and horizontal accountability remains weak, as expected in a hybrid regime such as Singapore.
Decline of diagonal accountability may lead to a decline in horizontal and vertical accountability. Singapore's democratic institutions were strained by the government's handling of the COVID-19 pandemic, especially related to media freedoms. Notably, the return to using lawsuits to silence dissent or stamp out pandemic disinformation had also contributed to a decline in media freedom. Similar to the regional governments, Singapore has expanded its government powers, increasing its capacity to surveil, to limit mobility and to silence dissent. More scrutiny is needed over the application and enforcement of new laws the Administration of Justice Act, FICA and POFMA, beyond the pandemic.
Singapore's case corresponds to the larger global expectations of backsliding in hybrid regimes. When the ruling elite faces stronger opposition challenge, it is more likely to use repression to crush threats (Dresden and Howard, 2016). Given the close relationship between the executive and judiciary in Singapore, more work is needed on its judicial independence, especially the role of the judiciary in political processes and litigations brought forth by the political leaders. An indicator of the erosion of horizontal and vertical accountability is if the PAP government pursues charges against the WP leader Pritam Singh for Raeesah Khan's act of lying in Parliament resulting in his loss of seat in parliament. If democratic institutions continue to be undermined, Singapore's civil society will need to mobilize the electorate to bring the PAP to account in the next elections.
