Tag: Freedom House

Democracy in the Eye of the Beholder?

20230529 Democracy in the Eye of the Beholder Image 01

There is growing concern that democratic institutions in Eastern Europe are fragile. This brief compares two perspectives on the state of democracy: expert assessments and surveys of the general population. We show that while experts’ perception of some countries’ institutions has worsened in recent years, voters are increasingly satisfied with their own democracies. This trend is broad-based, encompassing almost all new EU member states and all age groups. We provide evidence that over time, survey respondents’ assessment of democracy has become more closely tied to the outcome of elections rather than actual institutional change. Where governments have imposed restrictions on media freedom or judicial independence, their supporters continue to report high levels of satisfaction with the way democracy works.

“Across the world, democracy is backsliding”

UN’s Secretary-General António Guterres, 2022

In recent years, the prevailing narrative around democracy in Eastern Europe has been negative. The reform momentum that propelled countries towards EU membership has not been sustained after accession. Discussions of global democratic backsliding frequently cite countries from the region as examples (Grillo and Prato, 2023; Chiopris et al., 2021; Mechkova et al., 2017). Following restrictions on judicial independence and media freedom, some new EU members have seen their ratings slide on indices that measure the quality of democratic institutions based on expert opinions. This brief contrasts these expert assessments with a different perspective on the state of democracy: that of the voters themselves.

Data from Eurobarometer surveys show that satisfaction with ‘the way democracy works in our country’ has been increasing in the new EU member states. This upward trend is visible for all age groups and in almost all countries – including those where experts’ assessment of democracy has worsened. We document patterns in the data that may help to explain this divergence. Survey responses increasingly reflect an instrumentalist view of democracy; respondents who are aligned politically with the winning party are more likely to feel that democracy is working well. This trend can be observed across the EU, but it is most pronounced in the new EU member states where the governing parties are right-of-centre.

Perceptions of Democracy

Expert Assessments

The quality of democracy is hard to measure. A range of indices classify countries by regime type or provide numerical ratings of institutional quality (the Polity, V-Dem, and Freedom House measures are among the most prominent). These indices have somewhat different objectives and methodologies, but they all rely on subjective judgements by expert coders.

Some academic research casts doubt on the prevailing narrative of a global phenomenon of democratic backsliding. For instance, Treisman (2023) and Lueders and Lust (2018) show that there is little consensus across indices, both in terms of individual countries and the global trend. A recent paper by Little and Meng (2023) contrasts subjective indices with more objective indicators of democratic health (e.g. the rate at which incumbents lose elections). The authors find no evidence for global democratic backsliding using the objective measures and suggest that the pessimistic narratives around democracy may have biased coders’ assessment.

There is less disagreement about the development of democracy in Eastern Europe. Treisman (2023) cites Hungary as the only example of a country that has recently been downgraded both from the status of “liberal democracy” by V-Dem and “free state“ by Freedom House. Little and Meng (2023) highlight three cases where both objective and subjective measures indicate backsliding: Hungary and Poland (as well as Venezuela). Further, Becker (2019) shows that downgrades to V-Dem democracy scores in Bulgaria, Czechia, Hungary, Poland, and Romania are relatively broad-based, driven by declines across multiple sub-categories including freedom of expression and constraints on the executive.

Surveys of Public Opinion

We use individual-level data from the Eurobarometer – a survey of public opinion in the EU Member States and candidate countries conducted by the European Commission. The surveys are conducted at an approximately monthly frequency and comprise of a representative sample (about 1000 face-to-face interviews) for each state. We combine data from 42 surveys, spanning 20 years (2002 to 2022), with a total of 1.3 million respondents. The main question we are interested in is: “On the whole are you very satisfied, rather satisfied, not very satisfied or not at all satisfied with the way democracy works in [our country]?”

At the beginning of the sample period in 2002, around a third of respondents in Eastern European EU countries were satisfied with their respective democracies compared to close to twice as many respondents in Western Europe (Figure 1). Over the past 20 years, the share of Eastern Europeans satisfied with their democracy has grown to around 50 percent, narrowing the gap with Western Europe. Figure 2 shows that this pattern is broad based across age groups. All cohorts of Eastern Europeans are more satisfied with democracy than earlier generations and among the youngest respondents, satisfaction is almost as high as in Western Europe.

Figure 1. Satisfaction with Democracy vs V-Dem Score.

Source: Eurobarometer, V-Dem and authors’ calculations.

Figure 2. Satisfaction with Democracy by Age Group.

Notes: Each point shows the sample mean for a single year cohort. 95 percent confidence intervals in grey.
Source: Eurobarometer, authors’ calculations.

Figure 1 also shows a stark divergence in expert assessment of the state of democracy in Eastern Europe compared to public opinion in the same countries. While the V-Dem democracy scores for Eastern Europe have declined rapidly since the mid-2010s, average satisfaction with the own country’s democracy has increased. A much smaller gap between these two measures has also started to open up in Western Europe over the past couple of years.

In Figure 3, we show the same patterns of satisfaction with democracy and expert opinions for individual countries. Satisfaction with one’s own democracy has increased in almost all Eastern European countries, including in Poland and Hungary which at the same time showed the largest declines in democracy scores.

Figure 3. Satisfaction with Democracy vs V-Dem Score by Country.

Source: Eurobarometer, V-Dem and authors’ calculations.

This divergence in individual survey responses and expert assessments is not altogether surprising. First, the Eurobarometer surveys a sample of the population in each country, while V-Dem (and most other similar democracy indices) relies on country experts. Another likely explanation for the difference is the interpretation of the question. Democracy ratings tend to emphasise institutional aspects of a democracy, for instance, the V-Dem liberal democracy index is designed to capture rule of law and checks on executive power (see, e.g., Becker, 2019). In contrast, the survey responses are likely to reflect both satisfaction with the state of democracy in a country, as well as the outcomes of that democracy.

Satisfaction with Democracy and Political Alignment

In this section, we investigate whether stated satisfaction with democracy depends on the outcomes of elections and the political ideology of the respondents. A common way of measuring political ideology is the placement on a right-left scale, where the right favours a free-market economy and traditional values while the left favours economic redistribution and socially progressive policies. We compare the right-left placement of each country’s governing party as coded by the Chapel Hill Expert Survey (CHES), with the self-identified right-left placement of Eurobarometer respondents. We calculate the ideological distance from the government as the absolute difference between these two scores.

Figure 4. Relationship Between Ideological Distance from Government and Satisfaction with Democracy.

Source: Eurobarometer, Döring, Huber and Manow (2022) and authors’ calculations.
Notes: Each point shows the coefficient from a separate regression of satisfaction with democracy on ideological distance from government, age, gender, year fixed effects and country fixed effects. 95 percent confidence intervals in grey. Coefficient estimates for Eastern and Western EU overlap on the chart for 2019 and 2020.

We find that people are on average less satisfied with their country’s democracy when they are ideologically further from the parties in government (Figure 4). This is consistent with prior evidence (Anderson and Guillory, 1997; Ezrow and Xezonakis, 2011). The alignment effect has become stronger over time – even when taking into account average satisfaction levels for each country and demographic characteristics of the respondents, such as their age and gender. In the past three years, political alignment with the government has become a major factor in explaining satisfaction with democracy, especially in Eastern Europe. Svolik (2019) suggests that voters trade off democratic principles and partisan interests. As political polarisation increases, voters become more willing to accept a government that undermines democratic institutions, as long as it is on ‘their side’ ideologically.

Figure 5. Satisfaction with Democracy and Political Ideology. Western Europe in the Left Panel and Eastern Europe in the Right Panel.

Source: Eurobarometer, Döring, Huber and Manow (2022) and authors’ calculations.
Notes: Respondents with the most left-leaning ideology are at the extreme left of the x-axis, those with the most right-leaning ideology are at the extreme right. The sample covers the period 2002 to 2022 and excludes observations where the government is coded as centrist (scores of 5-6 in the CHES data). 95 percent confidence intervals in grey.

In Figure 5, we break down the effect of political alignment on satisfaction with democracy according to individuals’ political leanings. On the x-axis is the respondents’ left-right placement and on the y-axis there are two series of dots showing satisfaction with democracy depending on whether the government is left of centre (lighter coloured dots) or right of centre (darker coloured dots). As before, being politically aligned with the government increases satisfaction, that is, to the left of the chart, the lighter coloured dots are placed higher than the darker coloured dots and vice versa for the right of the chart. The further from centre a person’s political leanings, the less satisfied they are with a government of the opposite ideology. There is also some evidence of asymmetry across the political spectrum in Eastern Europe, with respondents on the political right reporting much higher levels of satisfaction with right-wing governments compared to voters on the left under a left-wing government.


Over the past decade, there has been increasing concern over democratic backsliding in some of the Eastern European countries that are members of the EU. This is reflected in commonly used democracy indices whose country experts note the worrying trends in countries’ institutions – such as the reduction of freedom of expression, the strengthening of rule of law and constraints on the executive, all hallmarks of a liberal democracy. In this policy brief, we investigate whether this erosion of institutional safeguards affects people’s stated satisfaction with democracy in one’s respective country. We find a broad-based increase in satisfaction with democracy in the Eastern European EU countries, including in the countries that have seen some of the largest declines in liberal democracy ratings. We show that stated satisfaction with democracy reflects less the institutional changes in countries, but more the outcome of democratic elections. Voters who are politically aligned with their government are systematically more likely to report that they are satisfied with the state of democracy in their country. And this effect has become stronger in the most recent years, particularly in the Eastern European EU countries. We also find that this effect is not symmetric across the political spectrum. In the Eastern European EU countries, respondents on the political right are more satisfied with right-wing governments than those on the left are with left-wing governments.

The descriptive patterns outlined in this policy brief illustrate a worrying disconnect in the minds of many voters between institutions and outcomes of the democratic process. The threat of democratic backsliding in Europe and across the globe is predominantly not due to electoral democracies being replaced by autocratic regimes. Rather, genuinely popular (and often populist) governments are democratically elected and, once in power, proceed to undermine and dismantle liberal democratic institutions, such as a free press, an independent judiciary, and a fair electoral system. This process in turn makes it more difficult for opposition parties to win future elections, further cementing the power of the rulers of these illiberal democracies. While the electorate might support these governments now, voters need to be aware that these liberal institutions are designed to safeguard their democratic future.


  • Anderson, C. J. and Guillory, C. A. (1997). Political institutions and satisfaction with democracy: A cross-national analysis of consensus and majoritarian systems. American Political Science Review, 91(1), pp.66-81.
  • Becker, T. (2019). Liberal Democracy in Transition – The First 30 Years. FREE Policy Brief.
  • Chiopris, C., Nalepa, M. and Vanberg, G. (2021). A wolf in sheep’s clothing: Citizen uncertainty and democratic backsliding. Working Paper.
  • Döring, H., Huber, C. and Manow, P. (2022). ParlGov 2022 Release. Harvard Dataverse. https://doi.org/10.7910/DVN/UKILBE
  • Eurobarometer (multiple waves: 2002-2022), European Commission. Brussels
  • Ezrow, L. and Xezonakis, G. (2011). Citizen satisfaction with democracy and parties’ policy offerings. Comparative Political Studies, 44(9), pp.1152-1178.
  • Grillo, E. and Prato, C. (2023). Reference points and democratic backsliding. American Journal of Political Science, 67(1), pp.71-88.
  • Little, A. and Meng, A. (2023). Subjective and Objective Measurement of Democratic Backsliding. Available at SSRN 4327307.
  • Lueders, H. and Lust, E. (2018). Multiple measurements, elusive agreement, and unstable outcomes in the study of regime change. The Journal of Politics, 80(2), pp.736-741.
  • Mechkova, V., Luhrmann, A. and Lindberg, S. I. (2017). How much democratic backsliding?. Journal of. Democracy, 28, pp.162-169.
  • Svolik, M. W. (2019). Polarization versus democracy. Journal of Democracy, 30(3), pp.20-32.
  • Treisman, D. (2023). How great is the current danger to democracy? assessing the risk with historical data. Comparative Political Studies, https://doi.org/10.1177/00104140231168363.

Disclaimer: Opinions expressed in policy briefs and other publications are those of the authors; they do not necessarily reflect those of the FREE Network and its research institutes.

Liberal Democracy in Transition – The First 30 Years

20191027 Liberal Democracy in Transition FREE Network Policy Brief Image 07

This year marks 30 years since the first post-communist election in Poland and the fall of the Berlin Wall. Key events that started a dramatic transition process from totalitarian regimes towards liberal democracy in many countries. This brief presents stylized facts from this process together with some thoughts on how to get this process back on a positive track. In general, the transition countries that joined the EU are still far ahead of the other transition countries in terms of democratic development.

The recent decline in democratic indicators in some EU countries should be taken seriously as they involve reducing freedom of expression and removing constraints on the executive, but should also be discussed in light of the significant progress transition countries entering the EU have shown during the first 30 years of transition. The brief shows that changes in a democracy can happen fast and most often happen around elections, so getting voters engaged in the democratic process is crucially important. This requires politicians that engage the electorate and have an interest in preserving democratic institutions. An important question in the region is what the EU can do to promote this, given its overloaded political agenda. Perhaps it is time for a Greta for democracy to wake up the young and shake up the old.

This brief provides an overview of political developments in transition countries since the first post-communist elections in Poland and the fall of the Berlin Wall 30 years ago. It focuses on establishing stylized facts based on quantitative indices of democracy for a large set of transition countries rather than providing in-depth studies of a small number of countries. The aim of the brief is thus to find common patterns across countries that can inform today’s policy discussion on democracy in the region and inspire future studies of the forces driving democracy in individual transition countries.

The first issue to address is what data to use to establish stylized facts of democratic development in the region. By now, there are several interesting indicators that describe various aspects of democratic development, which are produced by different organizations, academic institutions and private data providers. In this brief, three commonly used and well-respected data providers will be compared in the initial section before we zoom in on more specific factors that make up one of these indices.

The big picture

The three indicators that we look at first are: political rights produced by Freedom House; polity 2 produced by the Polity IV project; and the liberal democracy index produced by the V-Dem project. Figures 1-3 show the unweighted average of these indicators for two groups of countries. The EU10 are the transition countries that became EU members in 2004 and 2007 and include Bulgaria, the Czech Republic, Estonia, Hungary, Lithuania, Latvia, Poland, Romania, Slovakia, and Slovenia. The second group, FSU12, are the 12 countries that came out of the Soviet Union minus the three Baltic countries in the EU10 group, so the FSU12 group consists of Armenia, Azerbaijan, Belarus, Georgia, Kazakhstan, Kyrgyzstan, Moldova, Russia, Tajikistan, Turkmenistan, Ukraine, and Uzbekistan.

Figure 1. Freedom House

Source: Freedom House and author’s calculations
Note: Scale inverted, 1 is best and 7 worst score

Figure 2. Polity IV project

Source: Polity IV project and author’s calculations
Note: Scale from -10 (fully autocratic) to 10 (fully democratic)

Figure 3. V-Dem

Source: V-Dem project and author’s calculations
Note: Scale from 0 to 1 where higher is more democratic

All three indicators convey the message that the democratic transformation in the EU10 group was very rapid in the early years of transition and the indicators have remained at high levels since the mid-90s only to show some decline in the most recent years for two of the three indicators. The FSU12 set of countries have made much less progress in terms of democratic development and remain far behind the EU10 countries in this regard. Overall, there is little evidence at the aggregate level that the democratic gap between the EU10 and FSU12 groups is closing. While the average EU10 country is more or less a full-fledged democracy, the average FSU12 country is at the lower end of the spectrum for all three democracy measures.

The average indicators in Figures 1-3 obviously hide some interesting developments in individual countries and in the following analysis, we will take a closer look at the liberal democracy index at the country level. We will then investigate what sub-indices contribute to changes in the aggregate index in the countries that have experienced significant declines in their liberal democracy scores.

For the first part of the analysis, it is useful to break down the democratic development in two phases. The first phase is from the onset of transition (1989, 1991 or 1993 depending on the specific country) to the time of the global financial crisis in 2009 and the second phase is from 2009 to 2018 (the last data point).

Figure 4. Liberal democracy, the first phase

Source: V-Dem project and author’s calculations

Figures 4 and 5 compare how the liberal democracy indicator changes from the first year of the period (measured on the horizontal axis) to the last year of the period (on the vertical axis). The smaller blue dots are the individual countries that make up the EU10 group while the red dots are the FSU12 countries. The 45-degree line indicates when there is no change between start and end years, while observations that lie below (above) the line indicate a deterioration (improvement) of the liberal democracy index in a specific country.

In the first phase of transition (Figure 4), all of the EU10 countries increased their liberal democracy scores and the average increase for the group was almost 0.5, going from 0.26 to 0.74. This was a result of many of the countries in the group making significant improvements without any countries deteriorating. The FSU12 group had a very different development with the average not changing at all since the few countries that improved (Georgia and Ukraine) were counterbalanced by a significant decline in Belarus and a more modest decline in Armenia.

Figure 5. Liberal democracy, the second phase

Source: V-Dem project and author’s calculations

The very rapid improvement in the liberal democracy index in the EU10 countries in the first phase of transition came to a halt and also reversed in several countries in the second phase of transition. Of course, as they had improved so much in the first period, there was less room for further positive developments, but the rapid decline in some of the countries was still negative news. However, it does point towards that reform momentum was very strong in the EU accession process, but once a country had entered the union, the pressure for liberal democratic reforms has faded.

Overall, the EU10 average fell by 0.1 from 2009 to 2018. This was a result of declining scores in several countries. The particularly large declines in this period have been seen in Hungary (-0.28), Poland (-0.27), Bulgaria (-0.14), the Czech Republic (-0.14), and Romania (-0.12). Again, the average FSU12 score did not change much, although Ukraine (-0.2) put its early success in reverse and lost as much in this period as it had gained earlier.

Country developments

Since much of the current discussion centers on how democracy is being under attack, the figures name the countries that have seen significant declines in the liberal democracy score in the first or second phase of transition. Figures 6 and 7 show the time-series of the liberal democracy index in the countries with significant drops at some stage of the transition process.

Figure 6. FSU12 decliners

Source: V-Dem project and author’s calculations

In many countries, the drop comes suddenly and sharply, with the first and most prominent example being Belarus. There, it only took three years to go from one of the highest ranked FSU12 countries to fall to one of the lowest liberal democracy scores. In Poland, Romania, Bulgaria and Armenia, the process was also very rapid and significant changes happened in 2-3 years.

Figure 7. EU10 decliners

Source: V-Dem project and author’s calculations

In the Czech Republic and Hungary, the period of decline was much longer and in the case of Hungary, the drop was the most significant in the EU10 group. Ukraine stands out as more of an exception with a roller-coaster development in its liberal democracy score that first took it up the list and then back down to where it started. For those familiar with politics in these countries, it is easy to identify the elections and change in government that have occurred at the times the index has started to fall in all of these countries. In other words, the democratic declines have not started with coups but followed election outcomes where in most cases the incumbent leaders have been replaced by a new person or party.

How democracy came under attack

We will now take a closer look at what has been behind the instances of decline in the aggregate index by investigating how the sub-indices have developed in these countries. The sub-indices that build up the liberal democracy index are: freedom of expression and alternative sources of information; freedom of association; share of population with suffrage; clean elections; elected officials; equality before the law and individual liberty; judicial constraints on the executive; and legislative constraints on the executive (the structure is a bit more complex with mid-level indices, see V-Dem 2019a).

Table 1 shows how these indicators have changed in the time period the liberal democracy indicator has fallen significantly (with shorter versions of the longer names listed above but in the same order). The heat map of decline indicated by the different colours is constructed such that positive changes are marked with green, smaller declines are without colour, declines greater that 0.1 but smaller than 0.2 are in yellow and larger declines in red. Note that the liberal democracy index is not an average of the sub-indices but based on a more sophisticated aggregation technique (see V-Dem 2019b). Therefore, the Czech Republic and Bulgaria can have a greater fall in top-level liberal democracy index that what is indicated by the sub-indices.

Table 1. Changes in liberal democracy indicators at times of democratic decline

Source: V-Dem project and author’s calculations

For the countries with the largest changes in the liberal democracy index, it is clear that both freedom of expression and alternative sources of information have come under attack together with reduced judicial and legislative constraints on the executive. Among the EU10 countries, Hungary and Poland stand out in terms of reducing freedom of expression, while Romania has seen most of the decline coming from reducing constraints on the executive. Not surprisingly, Belarus stands out in terms of the overall decline in liberal democracy coming from reducing both freedom of expression and constraints on the executive in the most significant way.

On a more general level, the attack on democracy does differ between the countries, but in the cases where serious declines can be seen, the attack has been particularly focused on information aspects and constraints on the executive. At the same time, all countries let all people vote (suffrage always at 1) and let the one with the most votes get the job (elected officials).

Policy conclusions

This brief has provided some stylized facts on the first 30 years of liberal democracy in transition and some details on how democracy has come under attack in individual countries. It leaves open many questions that require further studies and some of these are indeed ongoing in this project and will be presented in future briefs and policy papers here.

Some observations have already been made here that can inform policy discussions on liberal democratic developments in the region. The first is that changes can happen very rapidly, both in terms of improvements but also in terms of dismantling important democratic institutions, including those that provide constraints on the executive or media that provides unbiased coverage before and after elections. What is also noteworthy is that these changes have almost always happened after an election where a new person or party has come to power, so the democratic system is used to introduce less democracy in this sense.

It is also interesting that in all of the countries, the most easily observed indicators of democracy such as suffrage and having the chief executive or legislature being appointed by elections are given the highest possible scores. In other words, even the most autocratic regime wants to look like a democracy; but as the old saying goes, “it is not who votes that is important, it is who counts”.

The regime changes at election times that have led to declining liberal democracy scores have also in many cases come as a result of the incumbents not doing a great job or voters not turning up to vote. It was enough for Lukashenko in Belarus to promise to deal with corruption and rampant inflation that was a result of the old guard’s mismanagement to turn Belarus into an autocracy. In Hungary, the change of regime came after the Socialist leader was caught on tape saying he had been lying to voters. While in Romania, only 39% voted in the 2016 election. And in Bulgaria, around half of the voters stayed at home in the presidential election the same year.

In sum, both incompetent and corrupt past leaders and disengaged or disillusioned voters are part of the decline in a liberal democracy that we have seen in recent years. It is clearly time for policy makers that are interested in preserving liberal democracy in the region and elsewhere to think hard about how democracy can be saved from illiberal democrats. Part of the answer clearly will have to do with how voters can be engaged in the democratic process and take part in elections. It also involves defending free independent media and the thinkers and doers that contribute to the liberal democracy that we cherish. The question is if the young generation will find a Greta for democracy that can kick-start a new transition to liberal democracy in the region and around the world.

For those readers that want to participate more actively in this discussion and have a chance to be in Stockholm on November 12, SITE is organizing a conference on this theme which is open to the public. For more information on the conference, please visit SITE’s website (see here).


  • Freedom house data downloaded on Oct 4, 2019, from https://freedomhouse.org/content/freedom-world-data-and-resources
  • Freedom house methodological note available at https://freedomhouse.org/report/methodology-freedom-world-2018
  • Polity IV project data downloaded on Oct 4, 2019, from http://www.systemicpeace.org/inscrdata.html
  • Polity IV project manual available at http://www.systemicpeace.org/inscr/p4manualv2018.pdf
  • V-Dem project data downloaded on Sept 24, 2019, from https://www.v-dem.net/en/data/data-version-9/
  • Coppedge, Michael, John Gerring, Carl Henrik Knutsen, Staffan I. Lindberg, Jan Teorell, David Altman, Michael Bernhard, M. Steven Fish, Adam Glynn, Allen Hicken, Anna Lührmann, Kyle L. Marquardt, Kelly McMann, Pamela Paxton, Daniel Pemstein, Brigitte Seim, Rachel Sigman, Svend-Erik Skaaning, Jeffrey Staton, Steven Wilson, Agnes Cornell, Lisa Gastaldi, Haakon Gjerløw, Nina Ilchenko, Joshua Krusell, Laura Maxwell, Valeriya Mechkova, Juraj Medzihorsky, Josefine Pernes, Johannes von Römer, Natalia Stepanova, Aksel Sundström, Eitan Tzelgov, Yi-ting Wang, Tore Wig, and Daniel Ziblatt. 2019a. “V-Dem [Country-Year/Country-Date] Dataset v9”, Varieties of Democracy (V-Dem)
  • Pemstein, Daniel, Kyle L. Marquardt, Eitan Tzelgov, Yi-ting Wang, Juraj Medzihorsky, Joshua Krusell, Farhad Miri, and Johannes von Römer. 2019b. “The V-Dem Measurement Model: Latent Variable Analysis for Cross-National and Cross-Temporal Expert-Coded Data”, V-Dem Working Paper No. 21. 4th edition. University of Gothenburg: Varieties of Democracy Institute.

On the Timing of Turkey’s Authoritarian Turn

20160404 FREE Network Policy Brief featured image 01

This policy brief examines the timing of Turkey’s authoritarian turn using raw data measuring freedoms from the Freedom House (FH). It shows that Turkey’s authoritarian turn under the ruling AKP is not a recent phenomenon. Instead, the country’s institutional erosion – especially in terms of freedoms of expression and political pluralism – in fact began much earlier, and the losses in the earlier periods so far tend to dwarf those occurring later.


A growing field in economics emphasizes the importance of the media in shaping economic outcomes. Whereas the role of the media in informing voters is well established (Strömberg, 2015), recent research also points to the link between the rise of illiberal democracies (Mukand and Rodrik, 2016) and authoritarian control over media (Guriev and Treisman, 2015).

In few countries is the link between authoritarianism and restrictions on freedom of expression as pervasive as in Turkey. The Turkish government under the Erdoğan-led Justice and Development Party AKP has gutted the judiciary of most of its independence, and set it loose to crack down on critical media. These dire circumstances, however, contrast markedly with descriptions of Turkey from of a few years ago. Quite recently many analysts still deemed Turkey a “vibrant democracy”, and as late as in 2013 the foreign minister of Sweden proclaimed: “Erdoğan’s Turkey is on the right path.” This media shift raises concerns over the public portrayal of Turkey’s institutional development up until the last few years as well as the extent to which analysts may have misinterpreted Turkey’s institutional development over the past decade.

Measures of Freedoms

To this date, the most prominent source of measuring freedoms in the world is the Freedom House’s annual Freedom of the World reports (Freedom House, 2015), which designates countries into one of three statuses: Free, Partly Free, and Not Free, in ascending order in freedoms. In constructing these statuses, FH uses subscores for 7 subcategories, aggregated into 2 categories — Political Rights (hereby PR), with a range of 0 to 40 increasing in freedoms, and Civil Liberties (hereby CL), with a range 0 to 60 increasing in freedoms – for which each then gets its own 1-7 (with a low value indicating more freedoms and vice versa), and in turn these are used to classify a country as having a particular Freedom status. From the FH’s methodology section:

A country or territory is awarded 0 to 4 points for each of 10 political rights indicators and 15 civil liberties indicators, which take the form of questions; a score of 0 represents the smallest degree of freedom and 4 the greatest degree of freedom. The political rights questions are grouped into three subcategories: Electoral Process (3 questions), Political Pluralism and Participation (4), and Functioning of Government (3). The civil liberties questions are grouped into four subcategories: Freedom of Expression and Belief (4 questions), Associational and Organizational Rights (3), Rule of Law (4), and Personal Autonomy and Individual Rights (4).

The PR category thus focuses more on rights pertaining to politics, and is used also to construct an indicator variable for whether a country constitutes an electoral democracy or not. CL, as the name indicates, is more focused on liberties, including the right to freedom of expression, including media freedoms.

Somewhat curiously, Turkey’s ratings have barely budged over the last decade, and have been consistently classified as a Partly Free country. In 2005, FH assigned Turkey a 3 in both PR and CL, and the only change since then was a one-point drop in 2012’s CL rating down to 4.

The PR and CL, as well as their subcategory, scores are available on FH’s website. Using these, the graphs below plot the evolution of the PR and CL total scores for Turkey between 2005-2015 (which is the period for which FH provides this data), as well as the 25th, 50th, and 75th global percentiles from the annual world distribution of the respective scores. The latter allows gauging not just the absolute performance of Turkey but also that relative to the rest of the world.

Figure 1. Freedom House Category Scores

fig1Note: The red lines in the upper (lower) graph indicate the Political Rights (Civil Liberties) score. The dashed gray lines indicate the 25th, 50th, and 75th percentiles.

Turkey’s PR score (upper graph) is relatively flat over the decade with a 4-point drop after 2013, while the CL score (lower graph) has been falling since 2010 and was stagnant in the period before. Whereas Turkey started the period with a CL score just below the median country, it ended closer to the 25th percentile. None of the years show any indication of expanding freedoms during AKP’s rule.

The disaggregation of these scores into subcategories sheds further light on which of the latter have driven the former. For Turkey, especially relevant subcategories are the political pluralism and freedom of expression. The former refers to the degree to which people can organize in political groupings, credible opportunities for a political opposition, freedom from outside interference, and political rights for ethnic minorities. (This has clear links with the political barriers to entry such as the ten-percent threshold, the banning of political parties, and the persecution of Kurdish political activists in Turkey.) The latter subcategory refers to freedoms related to media, culture, religion, and academics, as well as the degree of freedom from government surveillance. Turkey’s subscores are plotted below, grouped by PR and CL:

Figure 2. Freedom House Subcategory Scores


In Figure 2, there is rather striking fall in both the political pluralism (red in upper graph) and freedom of expression (blue in lower graph).  The other subscores tend to be more stagnant over time, and the only subcategory that exhibited any significant positive momentum during the period is Electoral Process, although by 2015 it had reverted to its 2005 value. The falls in political pluralism and freedom of expression are large in magnitude, from 12 to 9 in the former, and 12 to 8 in the latter. Taking the latter decrease at face value would imply that Turkish citizens in 2015 have two-thirds of the freedoms of expression that they enjoyed in 2005. Moreover, the clear majority of the falls in both these variables occured before 2013 – the post-2013 fall in freedom of expression accounts for only one fourth of the total fall observed since 2007.

Given the degree to which civil liberties have eroded faster than that of political rights in Turkey, this begs the question how its inherent degree of illiberalism has evolved relative to other democracies. For this purpose, I plot the Freedom of Expression subcategory (a part of the CL score) against the PR score for the 2015. This latter score is used by FH to classify countries into whether they can be called ‘electoral democracies’ or not (CL subcategories have no bearing on this classification):

“An ‘electoral democracy’ designation requires a score of 7 or better in subcategory A (Electoral Process) and an overall Political Rights score of 20 or better.”

FH thus classifies Turkey as an electoral democracy – its Electoral Process is consistently above 7 (see upper graph in Figure 2) and PR score above 20 (see Figure 1). But are there many FH-classified electoral democracies with similar levels of freedom of expression? Figure 3 answers this question:
Figure 3. Freedom House Subcategory Scores

fig3Note: Green = Autocracy, blue = Democracy, Red = Turkey. The dashed green/blue lines indicate the median values for autocracies/democracies.

In 2015, Turkey is much closer to the median autocracy than the median democracy in terms of freedom of expression. Numerous autocracies have higher levels of freedom of expression than Turkey, and only two other electoral democracies have lower values of this variable: Bangladesh and Pakistan.

Consequently, Turkey today is a clear outlier due to its freedom of expression deficit among the countries classified as electoral democracies by FH. Yet the slide in Turkeys’ freedoms did not start in the last couple of years but has remained a pervasive feature of its institutional trajectory during the last decade.

Much of Turkey’s erosion of freedoms would not be as visible using only the most aggregate seven-point scaled ratings, as these obfuscate important changes within the rating scores. A shift in analysts’ focus to more disaggregated data may thus be useful in order to detect warning signs in a more timely fashion.

Finally, the severe deficit in freedoms in countries nonetheless classified by Freedom House as electoral democracies raises the issue of whether there should be a lower freedom bound to inclusion into the group of democracies in the world. And if so, what should determine such a lower bound?


  • Freedom House, 2016, “Freedom of the World 2016.”
  • Guriev, Sergei, and Daniel Treisman, 2015, “How Modern Dictators Survive: An Informational Theory of the New Authoritarianism”, NBER Working Paper No 21136.
  • Meyersson, Erik, and Dani Rodrik, “Erdogan’s Coup”, Foreign Affairs, May 26th, 2014,
  • Mukand, Sharun, and Dani Rodrik, 2016, “The Political Economy of Liberal Democracy”,working paper.
  • Rodrik, Dani, 2014, “A General’s Coup”, mimeo.
  • Strömberg, David, 2015, “Media and Politics”, Annual Review of Economics, 7: 173-205.