Skip to main content


Front. Polit. Sci., 09 January 2023
Sec. Peace and Democracy
This article is part of the Research Topic Politics, Nationalism and Identity: A World in Flux? View all 4 articles

The democratic backsliding paradigm in enlarged European Union countries: In-depth analysis of V-Dem indicators

  • Department of Policy Studies, The University of Kitakyushu, Kitakyushu, Japan

The phenomena of democratic backsliding in enlarged European Union countries of central, eastern, and southeastern Europe tend to be lumped together. These paradigms fail to capture differences across countries and time, and the differences in backsliding among the subcomponents of liberal democracy. This study analyzes this issue from several indicators provided by V-Dem for 18 countries. The results indicate that democratic backsliding has been occurring in Poland, Hungary, Serbia, and Turkey, while not as much in other countries. What has been observed in these four countries is that the violation of the liberal component of democracy, the weakening of judicial control and restrictions on freedom of expression, comes before the violation of the cleanliness of democracy and the full breakdown of democracy. Through a comparative weighing of multiple indicators, this study clarifies the nuanced reality of democratic regression in the region.

1. Introduction

1.1. Democracy in enlarged Europe

The consolidation of democracy in central, eastern and southeastern European countries has long been of interest to both academics and practitioners (Stepan and Skach, 1993; Diamond, 1994; Zielonka, 2001). Now, many observations focus on the phenomenon in countries such as Hungary and Poland, which had completed their accession to the European Union and were freed from the shackles of membership requirements (Krastev, 2007; Sedelmeier, 2014; Kelemen, 2017). It has become widely discussed through various incisions such as “democratic backsliding,” “democratic regression,” “autocratization,” and so on (Diamond, 2015; Levitsky and Way, 2015; Agh, 2016; Hanley and Vachudova, 2018; Lührmann and Lindberg, 2019; Cianetti et al., 2020). These are more or less the same phenomenon described by different terms (Przeworski, 2019; Diamond, 2021). While both literature and media not limited to Hungarian or Polish cases argue that democratic backsliding (or recession, regression, dismantling, etc.) are taking place in Serbia (Castaldo, 2020), Turkey (San and Akca, 2021), Bulgaria (Dainov, 2020), Latvia (Greskovits, 2015), and Czech (Hanley and Vachudova, 2018), each case substantially varies in its aspects (as discussed later), which calls for a systematic comparison of the commonalities in and differences of these backslidings.

Whether democratic backsliding in central eastern and southeastern European countries is a region-wide trend is debatable [cf. featured in a special issue of the Journal of Democracy (Dawson and Hanley, 2016; Krastev, 2016)]. This tendency has prompted some to caution. “Viewing the entire region through the lens of ‘backsliding' may obscure more than it reveals” (Cianetti and Hanley, 2020). Even among countries that have gone through the same accession process or are in the process of negotiating accession to the EU, there are countries that are showing signs of democratic backsliding and others that are not. While democratic backsliding occurs in some areas, democratic consolidation occurs in others (Cianetti and Hanley, 2021). Based on this interest and similar attempts through data analysis (Schimmelfennig and Sedelmeier, 2019; Bochsler and Juon, 2020; Kapidzic, 2020), this study makes the uses V-Dem data to answer the following two questions for enlarged EU countries and candidate countries for EU membership: (1) to clarify “where” and “to what extent” democratic backsliding (or autocratization) are taking place and (2) to clarify “in which field” the backsliding are taking place in the countries where it happens. V-Dem—compared with other democratic indicators—has a wide and diverse range of sub-item variations and is sensitive to minute changes. Thus, it serves the purpose of the present objective.

This study broadly considers 18 countries in the Europe region that include the new member states of the EU (5th enlargement 2004-7; Bulgaria, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovakia, and Slovenia),1 Croatia, candidate states (Albania, Montenegro, North Macedonia, Serbia, Turkey), and the possible candidate states with the Stabilization and Association Agreement as of 2020 (Kosovo, and Bosnia and Herzegovina).

1.2. Democratic backsliding and EU expansion

Democratic backsliding is “the state-led debilitation of democratic institution sustaining democracies” (Bermeo, 2016). This phenomenon occurs as a rather endogenous, gradual, and partial processes of autocratization among democracies. While some democracies are toppled by forces emanating from outside the government, such as the military, others are more commonly corroded from inside by the democratically elected government's leaders (Maeda, 2010; Levitsky and Ziblatt, 2018). The process of democratic backsliding is also incremental, and “it is very unlikely that a country experiences democratic backsliding in any particular year” (Kyle and Mounk, 2018, p. 16). This aspect of backsliding makes it a “slow democratic erosion” that Huq and Ginsburg (2018a) distinguish from “fast authoritarian collapse.” In this sense, democratic erosion and backsliding are the same phenomenon; in fact, Przeworski (2019) and Diamond (2021) suggest that the terms “democratic backsliding,” “erosion,” “recession,” “retrogression,” and “regression” are different only in their preference.2 One possible exception is when the term “democratic deconsolidation” is used, which implies an element of public opinion and political culture, such as a retreat in attitudes in favor of democracy and an increase in attitudes in favor of authoritarianism (see Foa and Mounk, 2017; Norris, 2017).

Democratic backsliding may not always entail the quick collapse of democracies. Instead, countries may “lose democratic traits to varying degrees without fully, and long before breaking down” (Lührmann and Lindberg, 2019, p. 1098). These processes are quantitative by degree, and not qualitative dichotomies, such as whether they have or have not occurred. In terms of analysis, preoccupation with qualitative variation leads to discussing the presence or absence of democratic backsliding solely on the basis of distinct changes, such as institutional ones, and may thus not be a good indicator of backsliding or good practice in general (Vachudova, 2020).

Democratic backsliding typically involves a decay in (1) competitive free and fair elections, (2) freedom of speech and association, and (3) the rule of law (Huq and Ginsburg, 2018b; Przeworski, 2019, p. 172). Importantly, the various civil rights and institutional mechanisms that allow these liberal democracies to operate do not simultaneously weaken (or strengthen) in parallel. For example, it is generally believed that institutional constraints—exemplified by the “checks and balances” — on power through the legislative and judicial branches of government work to protect liberalism and the human rights of political minorities from the electoral winner. Przeworski (2019, p. 177) remains skeptical of the effect of such horizontal institutional checks and balances on contemporary democratic backsliding on the grounds that they are meaningless if they are occupied by the same parties. Frantz (2018, p. 136) similarly argues that the advances of liberalization in one area could lead to the strengthening of authoritarianism in another. These contradictions make it necessary to understand the sub-indicators of backsliding and analyze how this phenomenon actually occurs in reality. Particularly, analyzing the violation of the liberal rights of citizens separately from the violation of the liberal institutional mechanisms (horizontal equilibrium) may be more illuminating. This approach is suitable when we consider that the V-Dem has different indicators monitoring the current status of civil liberties and the institutional mechanisms that protect these liberties.

To examine democratic backsliding in Europe, it is essential to consider the EU as an actor. In general, the EU has required its member states to establish the rule of law and the protection of freedoms and human rights as a condition of membership (Kelley, 2004; Vachudova, 2005). Candidate governments are thus incentivized to maintain the institutions and freedoms necessary for democracy; however, these incentives disappear once a country has joined the EU and the European Court of Justice has limited influence on democratic backsliding (Blauberger and Kelemen, 2016). Krastev (2007) makes a conservative argument that democratic backsliding is the result of antipathy toward the imposition of a liberal consensus by the EU.

Democratic backsliding is not uniformly advanced in existing member countries and suppressed in candidate countries. For national governments, the critical issue is the balance between the credible threat of exclusion and the merit of membership against the domestic political cost of complying with democratic conditionality (Schimmelfennig and Sedelmeier, 2005, p. 213). Importantly, the EU's conditions for inclusion for central eastern and southeastern European countries continue to evolve, especially in their determinism and concreteness (Schimmelfennig and Sedelmeier, 2019). Compliance mostly depends now on the domestic political condition than on whether the country is a pre- or post-EU member. Therefore, we must observe and specify the variety of patterns of democratic backsliding in this region. As is already known and will be seen in detail herein, democratic backsliding occurs in different ways in the region, depending on the context of each country experiencing this phenomenon. If such patterns of backsliding are indeed different, it further reasserts the importance of specific regional studies. If, on the other hand, we find common patterns in a small number of countries experiencing democratic backsliding, we may be able to ascertain a common mechanism of democratic backsliding. Thus, it is essential to examine in what countries, in what areas, in what order, and by what small nuances of change infringement occurs.

To fulfill this aim, in the following section, we analyze the V-Dem indicators, including their sub-indicators, to identify trends and conditions in new EU members and candidate countries. It will show that there is no evidence of democratic backsliding in eastern Europe as a whole, but rather some countries (regions) are making progress in democratization. The democratic backsliding is clearly seen in Turkey, Serbia, and Hungary, and to a lesser extent in Poland. In the case of Turkey and Serbia, democratic backsliding goes further and reaches the democratic breakdown.

2. Material and methods

2.1. Overview of V-Dem indicators, electoral democracy index, and liberal democracy index

Among several indicators in the V-Dem score, our study examines the Electoral Democracy Index (EDI),3 which focuses on the minimal definition of democracy as the holding of competitive elections, and the Liberal Democracy Index (LDI), which takes into account multiple indicators in a more comprehensive manner.

The EDI consists of five indicators (freedom of expression and alternative sources of information index, freedom of association index, share of population with suffrage, clean elections index, and elected officials index).4 For example, its score for Sweden is 0.91; the US is 0.81; Malaysia is 0.42; Russia is 0.26; and North Korea is 0.09 (as of 2020). Its core value is freedom of electoral contestation (Coppedge et al., 2020a, p. 29). This indicator covers some aspect of possible democratic backsliding, but not all (cf. Huq and Ginsburg, 2018b; Przeworski, 2019).

Possible decay in the rule of law—or more broadly, institutional checks and balances—are detected by the indicator LDI. The LDI is a combination of the aforementioned EDI and the Liberal Component Index, which consists of three indicators: equality before the law and individual liberty index, judicial constraints on the executive index, and legislative constrain5 on the executive index [for details, see the Codebook (Coppedge et al., 2020b, p. 356–57)]. Although the EDI score tends to correlate with LDI, it is not completely equivalent to it. For example, Mexico's LDI 0.65 and EDI 0.41 differ (as of 2020), indicating both strong concentration of executive power and decay of rule of law under Andrés M. López Obrador, the incumbent President of Mexico. The EDI does not cover such decay of democratic institutional systems, which requires us to consider both EDI and LDI indicators.

This very wide range of sub-indicators is advantageous and why the study focuses on the V-Dem; however, it does not necessarily capture all events. The V-Dem team also recognized this point as “constrained by the limitations of extant measures” (Coppedge et al., 2017, p. 8). Interpreting the indicators requires us to occasionally refer to actual political development. We analyze the period from 2000 onward and changes thereafter. The membership negotiations for the 5th enlargement began in 1997–1999, which is close to our observation period. In 2000, EU countries imposed sanctions against Austria for the first time on the grounds of actions contrary to common European values. This year is a milestone in reinforcing the tenet that the protection of law and democracy set forth in the Copenhagen Criteria was not a nominal topic but an credible standard, which included sanctions.

In the following two sections, we first review changes over time in LDI and EDI, respectively. The data used are V-Dem version 11 files; since the V-Dem data are quantitative data, they do not quantitatively specify where the democratic breakdown occurred. Although, the V-Dem official uses an EDI score of 0.5 as one criterion for convenience (Lührmann et al., 2018). This is merely an intermediate number and has no theoretical or empirical basis. Hence, the criteria must also be taken into account in light of reality. As one attempt to do so, Kasuya and Mori (2022) proposed an appropriate cutoff point by comparing it with previous studies that classified qualitatively from theoretical democracy criteria (Cheibub et al., 2010; Boix et al., 2013; Geddes et al., 2014), the EDI score of 0.39 is the most discriminative and consistent level (0.42 when matched only with Boix et al., 2013). In other words, if we use the cutoff point around 0.39–0.42 as a qualitative criterion for whether a country is a democracy, we can make roughly the same judgment as the criteria used in previous studies on these democratization theories. This chapter also reports the Kasuya-Mori criterion as the main reference point. We then take countries where democratic backsliding is clearly occurring and examine which elements of the EDI and LDI subcategories show a numerical decline for those countries.

3. Results

3.1. Post-accession EU member countries

Among the 10 new EU member countries, both LDI and EDI scores show that democratic backsliding clearly has happened in Hungary and Poland (Figure 1). Hungary's score has been declining since 2010 and Poland's since 2015. Fidesz and PiS won these respective years. These are well-known findings from earlier studies (Cianetti et al., 2018; Grzymala-Busse, 2019; Vachudova, 2020). Bulgaria has also experienced a somewhat slow but slight downward trend (from about 0.70 in 2010 to about 0.62 in 2020 at EDI score) over the medium term under the government of Borisov, who took office as prime minister in 2009. Romania, in the midst of an upward trend from 0.62 in 2005 to 0.70 2016, showed a marked regression in the second half of the 2010s, especially in the LDI indicator, but has since been recovering over 0.75 in 2020. During that period, while the narrowly democratic part was maintained, problems with the balance between the judiciary and parliament (Mijatovic, 2019) probably led to an abrogation of the liberty factor indicator. Among the new EU member countries, Hungary and Poland are the two countries that are clearly in retreat from liberal democracy. However, there exist distinctions in the timing and extent of these differences.


Figure 1. LDI and EDI indicators in post-accession EU states.

Hungary and Poland have experienced varying degrees of decline. Poland has reached a score level of around 0.50–0.75 after 2016, which is equivalent to Romania in the early 2000s. The democratic backsliding that is commonly referred to in Poland is only a democratic backsliding and has not reached a democratic breakdown yet (as of 2020). However, Hungary's score is below 4.0 on the LDI index and 5.0 on the EDI index in 2018, when Fidesz reassured a strong grip on media once it controlled the vast majority of seats in the general election, after the subsequent declines since 2010 when the Fidesz won government power. The V-Dem official's report on the latter point, which declared Hungary as the only non-democratic regime in the EU, was received with some shock (Lührmann et al., 2020). However, it has not fallen below the Kasuya-Mori standard of EDI = 3.9 and still meets the minimum definition of democracy. It may be possible to consider that this does not reach the democratic breakdown (but even in that case, the LDI score is low, indicating that the liberal element has been considerably violated). Poland's democratic backsliding began in 2015, during the European refugee crisis. For Hungary, it was after the so-called Great Recession. In any case, the difference in the degree of democratic backsliding between Poland and Hungary has been quantitatively confirmed.

Nevertheless, it should not be overlooked that even though several of the new EU member states are experiencing a democratic backsliding, the majority of the countries are maintaining the same level of democracy as before (e.g., Estonia, Lithuania, and Slovakia), or experiencing smaller decline without any significant changes (e.g., Bulgaria and Czech Republic) compared with Hungary of Poland. While Hungary and Poland are certainly large central and eastern European countries, their democratic backsliding is not a prediction of backsliding in other new member countries. Their situation should be considered a political phenomenon occurring under specific circumstances, and paradigms treating this area as a whole in a structural phase of democratic backsliding are exaggerations. To clarify its implications in relation to the purpose of this study, democratic backsliding is seen as a country-specific phenomenon than a systematic phenomenon arising due to lack of control after the completion of EU accession. Although the exploring the cause of their democratic backsliding requires another examination (cf. Vachudova, 2020; Bernhard, 2021), some have focused on the more domestic political logic (cf. Pytlas, 2016). We will discuss this point later.

3.2. Pre-accession EU member countries

Next, we will review the situation in (former) pre-accession EU member countries. Here we look at the status of democracy indicators in the Western Balkans and Turkey, both of which are candidates for membership. In terms of EDI scores, Serbia and Turkey clearly show democratic backsliding (Figure 2). These scores of Serbia (in 2017) and Turkey (in 2015) are also below the Kasuya-Mori score of 3.9 on the EDI and can be recognized as cases that democratic backsliding has clearly reached the point of breakdown of democracy. Regarding Serbia, although democratic backsliding in the government of Aleksandar Vučić (Serbian Progressive Party), who took over as prime minister in 2014, is often noted, the score began to drop critically in 2013, the year before (Castaldo, 2020; Haggard and Kaufman, 2021). In Turkey, scores began to drop around 2007, when the second administration of the then Prime Minister Recep Tayyip Erdogan began and mass arrests of journalists began. Turkey's EDI score fell below 0.5 in 2014, when Erdogan remained president, and also below the Kasuya-Mori criterion of 0.39 the following year.


Figure 2. LDI and EDI indicators in EU candidate states.

Croatia became a potential EU candidate state in 2000, signed Stabilization and Association Agreements in 2001, became a candidate state in 2004, and finished the official negotiation process in 2011 (European Commission, 2022). During this process, its score of EDI and LDI changed little. Croatia's accession to the EU did not necessarily have a significant effect on improving the long-identified problem of lack of rule of law (Basheska, 2022). The score finally rose in 2010, when the tumultuous presidential elections were properly conducted and corruption-related influential politicians were ousted. During this period, the influence of EU accession conditions was marginal. The country's score had been on a gradual downward trend since it achieved EU accession in 2013, until 2018, when ruling party Croatian Democratic Union attempted to weaken judicial constraints (Cepo, 2020). The score stopped at around 0.75 points and has been slightly reversing in trend in recent years. This suggests that the trend after EU accession has not been uniform.

Bosnia and Herzegovina have also been on a long-term trend of gradual decline in scores, from 0.65 around 2003 and approaching 0.50 in 2019 on the EDI (although it will recover slightly in 2020 as well), although long-lasting corruption and clientelism partially undermine democratic norms (Bertelsmann Stiftung BTI, 2022). For this country, the LDI score has been significantly lower than the EDI score from the beginning. This gap in score implies that, although competitive electoral democracy has been largely achieved in this country, there is a lack of formal restraint through the legislature and the judiciary. In this case, it is not clear whether it is a subject that can be discussed as democratic backsliding. The same can be said for Albania and Montenegro, where the issue is whether they were liberal democracies in the first place before discussing democratic backsliding. If we take a restrained interpretation, it would be more appropriate to consider these countries as being or as areas of gray zone (between democracies and autocracies) from the beginning rather than as experiencing a democratic backsliding.

Notably, both EDI/LDI scores for Kosovo and North Macedonia (from 2015) have been clearly increasing since around 2013, indicating progress toward liberal democracy. It might be difficult to be optimistic that this democratic progress is an effect of the EU accession negotiations. This change reflects these two countries' improvement in democratic process, including electoral management (Freedom House, 2018). These two countries are increasingly distrustful of the European integration process and are struggling to reconcile their views on statehood with those of the EU (Gola and Bloom, 2022; Koneska et al., 2022). Therefore, if progress toward liberal democracy is to be seen, it would be more appropriate to see it as an internal process.

3.3. In-depth analysis of four backsliding countries

We have examined regime changes in new and candidate member countries with respect to European integration by looking at changes in the broad LDI and EDI indicators. Clearly, democratic backsliding observed in Poland, Hungary, Serbia, and Turkey as those V-Dem scores significantly decreased. In fact, these four countries are also among the top five countries in the “top 10 authoritarian states” in the report released by V-Dem in 2021 (Alizada et al., 2021) (the other country is Brazil). Poland's score remains in the democratic category, while Turkey and Serbia show a full breakdown of democracy. The judgment on the status of Hungary is divided based on various criteria.

For the other 14 countries, there is no clear signal of democratic backdown. Kosovo and North Macedonia in the 2010s even show the progress of democratization. Some cases have belonged as gray-zone fluctuations since the early 2000s and are not easily subject to the question of democratic backsliding in the first place. Overall, as Cianetti and Hanley (2020) point out, the argument that democratic backsliding is developing in the whole of former communist Europe and Eastern Europe seems to be an argument that is too drawn to a few cases. It cannot be said that whole enlarged European countries are experiencing democratic backsliding because of the structural commonality of their opposition and backlash to European integration. Rather, it is better to see it as a phenomenon peculiar to certain countries. Next section examines more detailed variations in the V-Dem scores of these four countries to get a closer look at what is really happening.

3.4. Sub-analysis of EDI scores (freedom of expression, freedom of association, clean elections)

Among the five indicators that make up the EDI, “freedom of expression and alternative sources of information,” “freedom of association,” and “clean election index” show fluctuations.6 Particularly large fluctuations are seen in the freedom of expression and clean election indicators.

In particular, there is a significant drop in scores for freedom of expression and alternative information in all countries (Figure 3, left). Since it is essential for the minimal democratic component of the conduct of competitive elections, it is not surprising that this indicator is included within the electoral democratic component. Regarding the fluctuations in clean election indicators (Figure 3, right), Hungary's, Serbia's, and Turkey's have experienced clear declines, while Poland's has remained high (except for a decline in 2020), indicating differences even within the four countries in this respect.


Figure 3. EDI sub-indicators of democratic backsliding countries.

As a common trend, the decline in freedom of expression scores preceded the decline in clean elections scores in all four countries. In Poland, violations of the former can be seen immediately after the PiS government in 2015, but it was not until 2020 that electoral transparency deteriorated. Similarly, in Hungary, the Fidesz government of 2010 has been characterized by tight media regulation; clientelism and manipulations of its elections arose around 2014 (Agh, 2016; Kovarek, 2022). Turkey also experienced a slight decline in both indicators in 2008 when it reformed several acts influencing writers and publishers, followed by a continued decline in the freedom of expression score, and then a large decline in the clean elections indicator from 2014 to 2015. In Serbia, the indicators show similar trends. Interestingly, the same political parties were responsible for declines in Poland, Hungary, and Turkey, respectively, whereas different political forces contributed to Serbia's backsliding. For instance, Ivica Dačić's Socialist Party government regulated the free press by implementing the 2013 Penal Code amendment (Human Rights Watch, 2015, p. 33), whereas Vučić's Progressive party, which has been in power since 2014, subsequently violated the transparency of the elections. It may be that media restrictions imposed by one ruling dispensation encouraged authoritarian behavior by rival parties.

Media restrictions could occur in any democracy. Thus, one should not infer from this direction alone the cause of strong democratic backsliding in four countries. Under normal circumstances, if an elected government tries to impose restrictions on the media, different political powers are expected to block such illiberal transitions. This is why we must consider institutional constraints as well.

3.5. Sub-analysis of LDI scores (judicial and legislative constraints)

One of the roles of the judiciary and the legislature are to exercise a mechanistic restraint on the actions of the democratically elected chief executive. The LDI indicator measures whether this form of balance of power and mutual restraint makes liberal democracies (or constitutional democracies) more robust. Among the sub-indicators of the LDI (the liberty component indicator on which the LDI is based), the “equality before the law and individual liberty” indicator differs significantly between Turkey and the rest of the world (Figure 4, left). In Turkey, bare violations of personal liberties with equality of law by authority have already occurred. Especially, torture has become a common tool as a means of tacking the Islamic State and clashes with Kurdistan Workers' Party since 2016–2017, although its score slightly recovered because, despite the response to Islamic State terror (Freedom House, 2019), violations of legal procedure are still few.


Figure 4. LDI sub-indicators of democratic backsliding countries.

The lack of horizontal accountability for political power won through competitive elections is widespread in all four countries. The violation of judicial constraints of executive power is conspicuous in Poland and Turkey (Figure 4). Although there is a downward trend in Hungary, the extent and pace of the decline are slower than in the other two countries. It is not clear whether this means that Hungary has a weaker rule of law than Poland. In practice, the European Commission initiated procedures based on the Article seven toward both countries. Interventions of judicial office by elected officials have been widespread in both countries. The only difference between the two nations is that Poland adopted an amendment to law in 2018 that allows government to intervene in lower and local courts. This is an institutional difference that may be over-sensitized by the V-dem expert coders. Although, similar suspicions exist of interventions in judgements of lower-level judges in Hungary as well, it tends to be taken less into account than institutional differences (This is not a problem specific to the V-Dem. Turkey and Poland are considered worse than Hungary in terms of independence of the judiciary, even by Freedom House).

Serbia has always had an inadequate level of judicial constraint on the executive. The level of control over the administration by non-judicial organs, including the legislature, has declined significantly in Hungary, Serbia, and Turkey, and is also declining in Poland. In Turkey, where presidential power has always been strong and parliamentary control of the executive has tended to be weak, this trend has been further weakened. Notably, although the V-dem score on this aspect shows signs of “improvement” from 2018 to 2020, it should be interpreted as a blur within the category of measurement error. Indeed, Turkey introduced constitutional reforms in 2017, where it shifted from a parliamentary system to a presidential one, and subsequently strengthened executive power against the parliament.

We must also consider timing. Poland began weakening judicial and legislative constraints in 2015, at the same time media regulations were put into effect. Hungary also charted a similar path of weakening institutional constraints in 2010 while tightening its grip on media. Turkey also began diluting media freedom around 2010 before it began curbing institutional constraints and declining in electoral transparency. In Serbia, we see the reverse trend—a regulated media that followed institutional decline around 2010–12. But in any case, they may have in common the fact that they are occurring in close proximity to each other.

This coincidence in timing may be one of the keys to the cause of strong democratic backsliding in the four countries. Cases in which an attack on the media and the disempowerment of the judiciary/legislative occur simultaneously are the common denominators of strong democratic backslidings in other cases. Sometimes this occurs after EU accession (e.g., Poland and Hungary); and sometimes it occurs while accession negotiations are underway (e.g., Serbia and Turkey) and is a consequence of the arbitrary actions of domestic political actors rather than being defined by some structural cause.

4. Discussion and conclusion

Based on an observational analysis of the V-Dem LDI and EDI and their sub-indicators, among the 18 new EU member and candidate countries, only four countries have strong empirical evidence of democratic regression or collapse. The rest of the countries are either robustly maintaining liberal democracy, are in a gray zone that is not suitable to be considered as a “democratic retreat” in the first place or are making progress in democratization. The cases of Poland, Hungary, Serbia, and Turkey indicate violations of liberal principles and weakening of institutional constrains, that is, an abrogation of the operations and institutions that constrain the political powers of electoral winners in a competitive democracy. Whether Hungary has reached the point of democratic collapse is controversial when considering the criterions. Serbia and Turkey can more clearly be judged to be in a worse state of democratic backsliding, leading to the collapse of democracy, and the Turkish case in particular is a serious one. The four countries experiencing “democratic backsliding” were not dissimilar in their internal conditions when we checked the variation in their sub-indicators. What they all had in common was the violation of freedom of expression and association, and the failure of institutional controls, such as the judiciary and legislature on the executive. Violations of electoral fairness occurred much later. The characteristics of backsliding and how it occurs have been noted before (Frantz, 2018, p. 94–7; Huq and Ginsburg, 2018b; Przeworski, 2019, p. 172). Our analysis, however, shows that violations of freedom of expression and the constraining of judicial and legislative checks and balances come before election rigging and violations of equality before the law. The implication can be drawn here that those who wish to preserve democracy should keep a particularly close eye on violations of freedom of expression and changes in institutional constraints, among other “telltale signs” (Frantz, 2018, p. 94) of democratic backsliding and authoritarianization. To the question of why democratic backsliding clearly occurred in only four countries (Poland, Hungary, Turkey, and Serbia), a simple answer might be that the elected forces in power in these four countries adopted a strategy of simultaneously and efficiently weakening these two aspects.

Beyond that, however, the situation differed in each country. In Poland, the fairness of the elections has not been seriously impaired, and the election of a democratic government is fair, but it is difficult to form and express objections to the government formed there and to the democratic winner, and it is also difficult to exercise institutional control over the government. The situation is truly one of illiberal democracy. The situation is similar in Hungary, but the overall degree of backsliding is stronger, especially in terms of violations of electoral fairness, even more so than in Poland. It may be that the two patterns are not qualitatively different but simply represent a gradual change. Depending on the future situation in Poland, this could be the case if the fairness of elections becomes more compromised. In Turkey, in addition to the process described above, even the modern rule of law, equality before the law, has been violated.7 The V-Dem official's perception is that their democratic backsliding has already reached the point of collapse of democracy. Serbia is uniquely positioned here because backsliding occurred across mutually competing political parties unlike the cases of Poland, Hungary, and Turkey. The violation of institutional judicial and legislative constraints may have caused democratic backsliding across party lines in this country.

Arguably, the backsliding in these countries may be considered to have been brought about by “democratic forces” during regime change. Indeed, the PiS was in the vein of the Solidarity movement, which promoted democracy in Poland, and Fidesz in Hungary was itself a democratic force. Turkey's accession negotiations to the European Union began and democratization progressed under the rule of the Justice and Development Party. These forces, being the winners of elections, have labeled dissent opinions and forces as undemocratic or reactionary, and have undermined the judiciary and other institutional constraints on their “democratically” elected actions and these are also the typical authoritarian processes of the day (Frantz, 2018, p. 94–7).

However, it is not clear whether these past circumstances are a decisive factor. Above all, there exist democratization forces in other countries where democratic backsliding has not occurred. Merely the fact that democratically elected forces seek to shut down dissent will not be enough to cause democratic backsliding to occur. Theoretically, there are various ways to set back democracy, but the analysis in this paper shows that the simultaneous occurrence of attacks on the media and institutional constraints is the most efficient way to accelerate democratic backsliding. What, then, efficiently produces such simultaneous attacks? That is a question that requires further investigation; although, it would seem that there must be an ideological basis to justify simultaneous attacks on the other side. On this point, it may be useful to broadly extend existing discourse analysis (e.g., Dawson and Hanley, 2019; Sata and Karolewski, 2019; Agarin, 2020). These issues cannot be understood from data alone, which represents a limitation of indicator-based comparative studies such as ours. It suggests the possibility or agendas for further comparative research.

Data availability statement

Publicly available datasets were analyzed in this study. This data can be found here:

Author contributions

RN contributed to the design and implementation of the research, the analysis of the results, and the writing of the manuscript.


This work was financially supported by the JSPS Grants-in-aids #18H00821.


An earlier version of this work was presented at the 2020 JACP annual conference. The author appreciates the comments and help from Yuko Kasuya, Takeshi Kawanaka, Kohei Imai, Adam Szymanski and the reviewers.

Conflict of interest

The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.


1. ^Croatia, which had already joined the EU in 2013, is grouped with the southeastern European countries, as per Vachudova (2014) and Schimmelfennig and Sedelmeier (2019). This group includes Kosovo, which had already signed the Stabilization and Association Agreements.

2. ^Lührmann and Lindberg (2019) argue that “recession” is more appropriate than “backsliding,” as these phenomena do not necessarily imply a “return” or “turning back” to the past and they are also not an unintentional “slide”.

3. ^As can be seen from this definition, this definition corresponds to Dahl's polyarchy concept—with competitive free and comprehensive elections—and is given the variable name “polyarchy” as it is in the V-Dem dataset.

4. ^Precisely speaking, EDI is made from two indices, the additive polyarchy index, which is the sum of these five variables, and the multiplicative polyarchy index, which is the multiplication of the five variables.

5. ^Although the variable name is “legislative,” the coding includes the government actors such as the Comptroller General, the Public Prosecutor, and the Ombudsman, so it should actually be taken as an indicator of horizontal accountability by non-judicial institutions to the government, not limited legislatures alone.

6. ^There was no decline at all in the scores for the “ratio of electoral vote holders” and the score for the “election of public office holders,” which continued to show the highest value.

7. ^However, as noted above, Turkey scored lower on the same indicator from the beginning. Therefore, careful consideration is needed as to whether we can equate them as part of this phase of democratic backsliding.


Agarin, T. (2020). The (not so) surprising longevity of identity politics: contemporary challenges of the state-society compact in central eastern Europe. East. Eur. Politics. 36, 147–166. doi: 10.1080/21599165.2020.1714598

CrossRef Full Text | Google Scholar

Agh, A. (2016). The decline of democracy in east-central europe: hungary as the worst-case scenario. Probl. Post Communism. 63, 277–287. doi: 10.1080/10758216.2015.1113383

CrossRef Full Text | Google Scholar

Alizada, N., Cole, R., Gastaldi, L., Grahn, S., Hellmeier, S., Kolvani, P., et al. (2021). Autocratization Turns Viral: Democracy Report 2021. Gothenburg: Varieties of Democracy Institute (V-Dem).

Google Scholar

Basheska, E. (2022). EU enlargement in disregard of the rule of law: a way forward following the unsuccessful dispute settlement between croatia and slovenia and the name change of macedonia. Hague J. Rule Law. 14, 221–256. doi: 10.1007/s40803-022-00169-7

CrossRef Full Text | Google Scholar

Bermeo, N. (2016). On democratic backsliding. J Democr. 27, 5–19. doi: 10.1353/jod.2016.0012

CrossRef Full Text | Google Scholar

Bernhard, M. (2021). Democratic Backsliding in Poland and Hungary. Slavic Rev. 80, 585–607. doi: 10.1017/slr.2021.145

CrossRef Full Text | Google Scholar

Bertelsmann Stiftung BTI. (2022). Country Report — Bosnia and Herzegovina. Gütersloh: Bertelsmann Stiftung.

Google Scholar

Blauberger, M., and Kelemen, D. (2016). Can courts rescue national democracy? Judicial safeguards against democratic backsliding in the EU. J. Eur. Public Policy. 24, 321–336. doi: 10.1080/13501763.2016.1229357

CrossRef Full Text | Google Scholar

Bochsler, D., and Juon, A. (2020). Authoritarian footprints in Central and Eastern Europe. East Eur. Politics. 36, 167–187. doi: 10.1080/21599165.2019.1698420

CrossRef Full Text | Google Scholar

Boix, C., Michael, M., and Rosato, S. (2013). A complete data set of Political regimes, 1800–2007. Comp. Polit. Stud. 46, 1523–1554. doi: 10.1177/0010414012463905

CrossRef Full Text | Google Scholar

Castaldo, A. (2020). Back to competitive authoritarianism? democratic backsliding in Vučić's Serbia. Eur. Asia Stud. 72, 1617–1638. doi: 10.1080/09668136.2020.1817860

CrossRef Full Text | Google Scholar

Cepo, D. (2020). Structural weaknesses and the role of the dominant political party: democratic backsliding in Croatia since EU accession. Southeast Eur. Black Sea Stud. 20, 141–159. doi: 10.1080/14683857.2020.1709721

CrossRef Full Text | Google Scholar

Cheibub, A., Gandhi, J., and Vreeland, J. R. (2010). Democracy and dictatorship revisited. Public Choice. 143, 67–101. doi: 10.1007/s11127-009-9491-2

CrossRef Full Text | Google Scholar

Cianetti, L., Dawson, J., and Hanley, S. (2018). Rethinking “democratic backsliding” in Central and Eastern Europe – looking beyond Hungary and Poland. East Eur. Politics. 34, 243–256. doi: 10.1080/21599165.2018.1491401

CrossRef Full Text | Google Scholar

Cianetti, L., Dawson, J., and Hanley, S. (2020). Rethinking “Democratic Backsliding” in Central and Eastern Europe. London: Routledge.

Google Scholar

Cianetti, L., and Hanley, S. (2020). “We must go beyond the ‘backsliding paradigm' to understand what's happening to democracy in Central and Eastern Europe.” LSE EUROPP Blog. Available online at: doi: 10.1201/9780429264795-1 (accessed April 1, 2022).

CrossRef Full Text | Google Scholar

Cianetti, L., and Hanley, S. (2021). The end of the backsliding paradigm. J. Democr. 32, 66–80. doi: 10.1353/jod.2021.0001

CrossRef Full Text | Google Scholar

Coppedge, M., Gerring, J., Glynn, A., Knutsen, C., Lindberg, S., Pemstein, D., et al. (2020b). Varieties of Democracy: Measuring Two Centuries of Political Change. Cambridge: Cambridge University Press. doi: 10.1017/9781108347860

CrossRef Full Text | Google Scholar

Coppedge, M., Gerring, J., Knutsen, C., Lindberg, S., Teorell, J., Altman, D., et al. (2020a). V-Dem Codebook v10. Gothenburg: Varieties of Democracy (V-Dem) Project.

Google Scholar

Coppedge, M., Gerring, J., Lindberg, S. I., Skaaning, S. -E., and Teorell, J. (2017). V-Dem comparisons and contrasts with other measurement project. Working Paper Series, 2017, 45. doi: 10.2139/ssrn.2951014

CrossRef Full Text | Google Scholar

Dainov, E. (2020). How to Dismantle a Democracy: The Case of Bulgaria. Open Democr.

Google Scholar

Dawson, J., and Hanley, S. (2016). What's wrong with east-central Europe? the fading mirage of the “Liberal Consensus”. J. Democr. 27, 20–34. doi: 10.1353/jod.2016.0015

CrossRef Full Text | Google Scholar

Dawson, J., and Hanley, S. (2019). Foreground Liberalism, Background Nationalism: A Discursive? Institutionalist Account of EU Leverage and ‘Democratic Backsliding' in East Central Europe. J. Common Mark. Stud. 57, 710–728. doi: 10.1111/jcms.12845

CrossRef Full Text | Google Scholar

Diamond, L. (1994). Rethinking civil society: toward democratic consolidation. J. Democr. 5, 4–17. doi: 10.1353/jod.1994.0041

CrossRef Full Text | Google Scholar

Diamond, L. (2015). Facing up to the democratic recession. J. Democr. 26, 141–155. doi: 10.1353/jod.2015.0009

CrossRef Full Text | Google Scholar

Diamond, L. (2021). Democratic regression in comparative perspective: scope, methods, and causes. Democratization. 28, 22–42. doi: 10.1080/13510347.2020.1807517

CrossRef Full Text | Google Scholar

European Commission. (2022). Croatia, European Neighbourhood Policy and Enlargement Negotiations. Available online at:

Google Scholar

Foa, S., and Mounk, Y. (2017). The signs of deconsolidation. J. Democr. 28, 5–15. doi: 10.1353/jod.2017.0000

CrossRef Full Text | Google Scholar

Frantz, E. (2018). Authoritarianism: What Everyone Needs to Know. New York, NY: Oxford University Press. doi: 10.1093/wentk/9780190880194.001.0001

CrossRef Full Text | Google Scholar

Freedom House. (2018). Freedom in the World 2018. North Macedonia. Available online at:

Google Scholar

Freedom House. (2019). Freedom in the World 2019. Turkey. Available online at:

Google Scholar

Geddes, B., Wright, J., and Frantz, E. (2014). Autocratic breakdown and regime transitions: a new data set. Perspect. Politics. 12, 313–331. doi: 10.1017/S1537592714000851

CrossRef Full Text | Google Scholar

Gola, G., and Bloom, E. (2022). Democracy in the Western Balkans: A Long and Winding “European Path”. International IDEA Blog. Available online at:“european-path” (accessed July 26, 2022).

Google Scholar

Greskovits, B. (2015). The hollowing and backsliding of democracy in east central Europe. Global Policy. 6, 28–37. doi: 10.1111/1758-5899.12225

CrossRef Full Text | Google Scholar

Grzymala-Busse, A. (2019). How populists rule: the consequences for democratic governance. Polity. 51, 707–717. doi: 10.1086/705570

CrossRef Full Text | Google Scholar

Haggard, S., and Kaufman, R. (2021). The anatomy of democratic backsliding. J. Democr. 32, 27–41. doi: 10.1353/jod.2021.0050

CrossRef Full Text | Google Scholar

Hanley, S., and Vachudova, M. (2018). Understanding the illiberal turn: democratic backsliding in the Czech Republic. East Eur. Politics. 34, 276–296. doi: 10.1080/21599165.2018.1493457

CrossRef Full Text | Google Scholar

Human Rights Watch. (2015). A Difficult Profession: Media Freedom Under Attack in the Western Balkans. Human Rights Watch. Available online at:

Google Scholar

Huq, A., and Ginsburg, T. (2018a). How to Save a Constitutional Democracy. Chicago: The University of Chicago Press. doi: 10.7208/chicago/9780226564418.001.0001

CrossRef Full Text | Google Scholar

Huq, A., and Ginsburg, T. (2018b). How to lose a constitutional democracy. UCLA Law Rev. 65, 78–169. doi: 10.2139/ssrn.2901776

CrossRef Full Text | Google Scholar

Kapidzic, D. (2020). The rise of illiberal politics in Southeast Europe. J. Southeast Eur. Black Sea Stud. 20, 1–17. doi: 10.1080/14683857.2020.1709701

CrossRef Full Text | Google Scholar

Kasuya, Y., and Mori, K. (2022). Re-examining thresholds of continuous democracy measures, Contemp. Politics. 28, 365–385. doi: 10.1080/13569775.2021.1993564

CrossRef Full Text | Google Scholar

Kelemen, D. (2017). Europe's other democratic deficit: national authoritarianism in Europe's democratic union. Gov. Oppos. 52, 211–238. doi: 10.1017/gov.2016.41

CrossRef Full Text | Google Scholar

Kelley, J. (2004). Ethnic Politics in Europe: The Power of norms and incentives. Princeton and Oxford: Princeton University Press.

Google Scholar

Koneska, C., Huskic, A., and Krasniqi, G. (2022). Macedonia, bosnia and kosovo: contested statehood and the EU. J. Interven. Statebuilding. online first. doi: 10.1080/17502977.2022.2086390

CrossRef Full Text | Google Scholar

Kovarek, D. (2022). Distributive politics as behavioral localism: evidence from a vignette experiment in Hungary. Res. Politics. 9, 1–7. doi: 10.1177/20531680221122279

CrossRef Full Text | Google Scholar

Krastev, I. (2007). East-Central Europe Backsliding? The Strange Death of the Liberal Consensus. J. Democr. 18, 56–63.

Google Scholar

Krastev, I. (2016). What's wrong with east-central Europe? liberalism's failure to deliver. J. Democr. 27, 35–38. doi: 10.1353/jod.2016.0001

CrossRef Full Text | Google Scholar

Kyle, J., and Mounk, Y. (2018). The Populist Harm to Democracy: An Empirical Assessment. London: Tony Blair Institute for Global Change.

Google Scholar

Levitsky, S., and Way, L. (2015). The myth of democratic recession. J. Democr. 26, 45–58. doi: 10.1353/jod.2015.0007

CrossRef Full Text | Google Scholar

Levitsky, S., and Ziblatt, D. (2018). How Democracies Die. New York, NY: Crown.

Google Scholar

Lührmann, A., and Lindberg, S. I. (2019). A third wave of autocratization is here: what is new about it? Democratization, 26, 1095–1113. doi: 10.1080/13510347.2019.1582029

CrossRef Full Text | Google Scholar

Lührmann, A., Maerz, S., Grahn, S., Alizada, N., Gastaldi, L., Hellmeier, S., et al. (2020). Autocratization Surges – Resistance Grows. Democracy Report 2020. Gothenburg: Varieties of Democracy Institute (V-Dem).

Google Scholar

Lührmann, A., Tannenberg, M., and Lindberg, S. (2018). Regimes of the world (RoW): opening new avenues for the comparative study of political regimes. Politics Gov. 6, 60–77. doi: 10.17645/pag.v6i1.1214

CrossRef Full Text | Google Scholar

Maeda, K. (2010). Two modes of democratic breakdown: a competing risks analysis of democratic durability. J. Politics. 72, 1129–1143. doi: 10.1017/S0022381610000575

CrossRef Full Text | Google Scholar

Mijatovic, D. (2019). “The independence of judges and the judiciary under threat.” Human Rights Comments, Commissioner for Human Right, Council of Europe. Available online at: (accessed April 10, 2022).

Google Scholar

Norris, P. (2017). Is Western democracy backsliding? diagnosing the risks. J. Democr. 28, 1–24. doi: 10.2139/ssrn.2933655

CrossRef Full Text | Google Scholar

Przeworski, A. (2019). Crises of Democracy. Cambridge: Cambridge University Press. doi: 10.1017/9781108671019

CrossRef Full Text | Google Scholar

Pytlas, B. (2016). Radical Right Parties in Central and Eastern Europe: Mainstream Party Competition and Electoral Fortune. Abingdon and New York, NY: Routledge. doi: 10.4324/9781315712741

CrossRef Full Text | Google Scholar

San, S., and Akca, D. (2021). How Turkey's democratic backsliding compromises the international dimension of democratization. Digest Middle East Stud. 30, 34–52. doi: 10.1111/dome.12223

CrossRef Full Text | Google Scholar

Sata, R., and Karolewski, I. P. (2019). Caesarean politics in Hungary and Poland. East Eur. Politics. 26, 206–225. doi: 10.1080/21599165.2019.1703694

CrossRef Full Text | Google Scholar

Schimmelfennig, F., and Sedelmeier, U. (2005). The Europeanization of Central and Eastern Europe. Ithaca, NY: Cornell University Press.

Google Scholar

Schimmelfennig, F., and Sedelmeier, U. (2019). The Europeanization of Eastern Europe: the external incentives model revisited. J. Eur. Public Policy. 27, 1–20. doi: 10.1080/13501763.2019.1617333

CrossRef Full Text | Google Scholar

Sedelmeier, U. (2014). Anchoring democracy from above? the European Union and democratic backsliding in Hungary and Romania after accession. JCMS. 52, 105–121. doi: 10.1111/jcms.12082

CrossRef Full Text | Google Scholar

Stepan, A., and Skach, C. (1993). Constitutional frameworks and democratic consolidation: parliamentarianism versus presidentialism. World Polit. 46, 1–22. doi: 10.2307/2950664

CrossRef Full Text | Google Scholar

Vachudova, M. (2005). Europe Undivided: Democracy, Leverage, and Integration after Communism. Oxford: Oxford University Press. doi: 10.1093/0199241198.001.0001

CrossRef Full Text | Google Scholar

Vachudova, M. (2014). EU leverage and national interests in the Balkans: the puzzles of enlargement ten years on. J. Common Market Stud. 52, 122–138. doi: 10.1111/jcms.12081

CrossRef Full Text | Google Scholar

Vachudova, M. (2020). Ethnopopulism and democratic backsliding in Central Europe. East Eur. Politics. 36, 318–340. doi: 10.1080/21599165.2020.1787163

CrossRef Full Text | Google Scholar

Zielonka, J. (2001). Democratic Consolidation in Eastern Europe Volume 1: Institutional Engineering. Oxford: Oxford University Press. doi: 10.1093/0199244081.001.0001

CrossRef Full Text | Google Scholar

Keywords: democratic backsliding, V-Dem, southeast Europe countries, central and eastern Europe (CEE) countries, European Union (EU), autocratization

Citation: Nakai R (2023) The democratic backsliding paradigm in enlarged European Union countries: In-depth analysis of V-Dem indicators. Front. Polit. Sci. 4:966472. doi: 10.3389/fpos.2022.966472

Received: 11 June 2022; Accepted: 16 December 2022;
Published: 09 January 2023.

Edited by:

Adam Szymanski, University of Warsaw, Poland

Reviewed by:

Sorina Soare, University of Florence, Italy
Matteo Boldrini, Guido Carli Free International University for Social Studies, Italy
Mattia Collini, University of Florence, Italy

Copyright © 2023 Nakai. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*Correspondence: Ryo Nakai, yes

Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.