Tag: Political economics

Democratic Backsliding and Electoral Autocracies: Research Shared at the 2023 FROMDEE Conference

20231203 Democratic Backsliding Image 021

On October 13th, 2023, the Stockholm Institute of Transition Economics (SITE) and the Forum for Research on Media and Democracy in Eastern Europe (FROMDEE) hosted an academic conference on “Democratic Backsliding and Electoral Autocracies”. This brief provides a short summary of the keynote lecture and research presentations featured at the conference.

The most recent report by the V-Dem Institute concludes that “72 percent of the world’s population […] live in autocracies by 2022” and “the level of democracy enjoyed by the average global citizen in 2022 is down to 1986 levels” (V-Dem Institute, 2023). In Europe, these declines have manifested in the previous Polish government undermining judicial independence, in tightened political repression in Belarus, and most prominently in Putin’s invasion of Ukraine. But the rise of electoral autocracies and democratic backsliding are not confined to Europe – their strategies of propaganda, corruption, electoral manipulation, as well as attacks on judicial and media independence are a global phenomenon. The October 13th FROMDEE Conference gathered researchers from economics, law and political science to bring insight into why and where reversals are taking place, and what measures are needed to reverse the negative trend. This policy brief provides an overview of the research shared at the conference.

Populism and Autocracy – the Case of Russia

In the keynote lecture, Arturas Rozenas (New York University) focused on the roots of populism, challenging the common view of illiberal democracies as a mix between democracies and dictatorships. Rather, dictatorships evolve into populist dictatorships that then take one of two paths: either the path to democracy, or the path towards electoral autocracy, illiberal democracy, or totalitarianism. In this framework, populist dictatorships have historically made use of populist elements we recognize from modern times, i.e., democratic-seeming institutions misused for the ruler’s purposes.

In a populist dictatorship, Rozenas continued, there is a monopoly of power. Institutions such as elections and parliamentary representation, serve not to allocate power but to legitimise it. The transition from passive to active dictatorships coincided with a move away from the common notion of a king or similar leader deriving rightful power from God to rule the masses, to a reality built on the idea that the ruler’s legitimacy stems from the masses. This historic transformation should however not be interpreted as a transition to democracy. In fact, Rozenas showed that for most of recent history, the majority of elections and expansions of suffrage took place in dictatorships rather than in democracies. These seemingly populist institutions serve not only to legitimise governments, but also to coopt the population in a public display of the ruler’s strength. Rozenas argued, that in an active populist dictatorship, the ruler creates a setting which suppresses dissent and expectations of dissent, through institutionalised expressions of support (in the form of political participation, elections, large rallies etc.).

Turning to the Russian setting, the first thing to notice is the deep tradition of autocracy – from tsarism to Stalinism. In Russia, the words “society” and “the people” briefly blossomed during past revolutions or uprisings but have largely been absent in the Russian language and are once again on the decline under the rule of Putin. Further, the Russian population has time and again been exploited by its rulers during succession crises for displays of power and dominance. Examples of this are the mandatory elections held under Stalin two weeks after the invasion of the Baltic States in 1939 and more recently under Putin in the occupied territories of Luhansk and Donetsk in Ukraine in 2021. Such populist autocratic strategies are nothing new in Russia, concluded Rozenas – rather they derive from the internal logic of dictatorship that has played out throughout Russian history.

Continuing the notion of the “absent” Russian society, Olha Zadorozhna (Kozminski University) began her presentation by explaining that protests are infrequent in Russia and have surprisingly few attendees given the country’s large population. While there were mass protests in the run-up to the collapse of communism in the 1980’s and protests took place against corruption in 2017-2018, and in relation to the arrest of Alexey Navalny in 2021, protests in Russia are typically not motivated by an overarching ideology or broader political questions. Rallies in favor of authoritarianism and ethno-nationalism are a more common occurrence. Moreover, there are few indications that the invasion of Ukraine, sanctions and subsequent economic downturn have negatively affected the Russian population’s support for the regime. Still, literature has shown that war-related deaths can mobilize opposition against war participation (e.g., the U.S. participation in the Vietnam War). Considering this, Zadorozhna evaluates whether the deaths of Russian soldiers provoke a reaction among the Russian population. By combining social media data on fallen soldiers with protest activity for the first four months of the Russian invasion in 2022, the study find that casualties lead to an increase in protest activity, indicating that deaths can in fact mobilize public opposition in Russia.

Other populist strategies to ensure support for Putin in Russia relate to political participation and the judiciary. Nicholas James (University of Oxford) analysed electoral rule changes in the Russian Duma – from mixed member majoritarianism to proportional representation (PR) – by measuring their effect on floor participation. Applying a difference-in-differences framework, James found that deputies experiencing a change from PR included less words in their speeches following the switch (about 15-20 percent of an average speech). This effect should be understood in the political context of the ruling party’s (United Russia) increased influence during this time period (2010s). In fact, James concluded, the results point in the direction of the regime tampering with the Duma in an impromptu and reactionary manner with the overall goal of obtaining closer control and the appearance of support for the regime.

Yulia Khalikova’s (University of Hamburg) presentation gave further insight into how ostensibly democratic institutions can be exploited to make an authoritarian regime appear legitimate. In her work, Khalikova considers judicial references to international law that may be employed strategically, without necessarily adhering to the spirit or content of the law. Looking specifically at international law citations in 601 judgements made in the Russian Constitutional Court (RCC) between 2000 and 2021, Khalikova find evidence that the RCC has increasingly cited international courts when making judgements on topics related to politics and physical rights, indicating that state policy influences citation patterns. The change in citation patterns also points to the RCC currently using international law to support the regime and uphold its legitimacy, meaning that international law – adopted with the ambition of enhancing democratic values and ensuring human rights – is misused for undemocratic and repressive purposes.

Censorship and Propaganda

Information control is an important feature of autocratic regimes. Philine Widmer (ETH Zurich) considers the Chinese setting – where the regime controls the amount of foreign information available on the internet via a countrywide firewall. Research has shown that autocracies make use of censorship strategies to control their citizens, but these are associated with high reputational costs and can be overcome by tech-savvy citizens. Using a machine learning algorithm, Widmer first predicts a newspaper article’s alignment with the Chinese regime before comparing the placement of more/less aligned articles on news websites. Her results show that front-page news stories in Chinese newspapers are more aligned with the regime’s stance than other content. Front-page placement in turn matters for information uptake. Widmer ended the presentation by comparing the additional cost of finding less aligned articles to the technological costs required to access outside media (e.g. VPNs). For an autocracy to achieve its information control objectives, independent news may just need to be relatively harder to access. It does not need to make it impossible to access for all citizens.

Censorship is typically accompanied by, and complementary with, propaganda. Restricting other narratives allows autocratic regimes to spread their own. While propaganda is a common feature within autocracies, Jaakko Meriläinen’s (Stockholm School of Economics) presentation evaluated the effect of autocratic propaganda in a democratic setting.

Meriläinen’s study focuses on a rogue experiment in which some Finnish children in the 1970’s were taught history and social sciences using material from the Soviet Union – material which was in essence Soviet propaganda. By exploiting geographical and cohort variation, Meriläinen use a difference-in-differences approach to compare the 213 exposed children to children taught the regular Finnish curriculum. The long-term outcomes show that exposed children had lower incomes in adulthood, worked fewer months per year and were engaged in more left-leaning and publicly beneficial occupations (such as, nurses and firefighters).

Information and Accountability

The use of technological innovations to access otherwise restricted information was central to Arieda Muço’s (Central European University) presentation. She studies the spread of the Xerox photocopying machine in communist Hungary in the 1980’s – a setting characterised by limited freedom of speech and restrictions on the media. She reported that areas with early placement of Xerox machines are found to exhibit higher shares of pro-democratic voting. Muço ascribes these outcomes to the fact that the machines allowed for the spread of information and eased coordination of the opposition, suggesting that new technologies and information can act as key facilitators in the fall of autocratic regimes.

Providing citizens with information was also a key feature in Enrique Seira Bejarano’s (Michigan State University) presentation. He began by discussing two potentially related trends: in Latin America recent years have seen (i) increased levels of corruption and (ii) increased dissatisfaction with democracy among citizens. The number of corruption-related news articles have increased threefold in Spanish and doubled in English and the share of people perceiving corruption to be the greatest challenge to their country has doubled in the last decade. The study uses two empirical strategies to identify the effect of corruption on democratic values. Firstly, Seira Bejarano described an observational study, in which data on major corruption scandals were combined with Latinobarometer data on support for democracy. The authors find that corruption scandals increase corruption perceptions while decreasing stated support for democracy. Secondly, Seira Bejarano reported the results of a randomized controlled trial in which some respondents were shown videos of a politician accepting bribes. This had a negative effect on preferences for democracy and on trust more broadly. Both studies show that revelations of corruptions decrease the support for democracy, suggesting a potential tradeoff between the public’s belief in democratic institutions and increased transparency which is important for accountability but can also expose corruption.

Right-wing Populism

Yet another threat to democracy is the rise of right-wing populism – currently a reality in many well-established democracies across Europe. In Germany, the far-right party Alternative für Deutschland (AfD) enjoys around 21 percent of voters’ support according to recent polls. To understand their rise in popularity, Navid Sabet (Goethe University Frankfurt) builds on previous literature on cultural conflict as a driver for right-wing party support. The paper he presented examines the role of violent conflict in the form of terrorist acts. It evaluates whether acts of terror can alter the political landscape and shift support to the far-right. To avoid selection problems, the authors compare successful terror attacks to attacks that failed. Sabet reported that successful small-scale attacks (predominantly targeting migrants) increase AfD’s vote share by about 6 percentage points in state elections (in the time period 2013-2021). The acts of terror were found to increase voter turnout, by mobilizing otherwise idle voters, but also by gaining votes at the cost of other parties. Exploring the mechanisms behind these results, the authors study the language used by political parties and the way successful attacks were covered in the media. Relative to coverage of unsuccessful attacks, media coverage used a more negative tone, more words related to Islam and terror and fewer words related to right-wing populism. This suggests that media plays an important role in shaping the public’s response to acts of terror and that far right parties are able to exploit this dynamic.

Concluding Remarks

The 2023 FROMDEE Conference brought together academics from different fields to shed light on some of the main challenges to democracy today. In part, the research presented supported the prevailing narrative that democracies are backsliding in many parts of the world. However, by analysing how autocracies and populist leaders operate, the presenters also highlighted the vulnerability of dictatorships.

Arturas Rozenas cited the example of a rally in Bucharest in 1989, which was organised to display support for Ceauşescu’s regime and descended into an anti-government protest. Dictatorships can benefit by coopting the populist elements of democracy but, in doing so, they risk creating a vehicle for genuine democratic expression.

The audience learned about autocracies’ efforts to control the flow of information but also about citizens’ ability to circumvent restrictions whether in 1980s Hungary or present-day China. Several presentations focused on the extent of autocratic control in Russia but even in this setting, the death of soldiers in Ukraine motivates citizens to participate in protests.

Recent trends suggest that democratic institutions should not be taken for granted in any country. Societies can become more resilient to the threat of democratic backsliding, in part by better understanding how both democracies and autocracies operate and what makes them vulnerable. Researchers around the world are using innovative methods to expand our knowledge in this area, as reflected in the presentations at the 2023 FROMDEE Conference.


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.

Does the Russian Stock Market Care About Navalny?

Moscow citi in the sunset representing Russian stock market and Navalny

Alexei Navalny is the most prominent opposition leader in Russia today. During 2020, he entered not only the domestic Russian news flows, but was a major news story around the world following his horrific Novichok poisoning in August. This brief investigates the response in the Russian stock market to news about Navalny. For many significant Navalny news stories, the stock market experienced large negative returns that are not explained by the regular factors that move the market. Although the causality and permanency of these negative excess returns in the stock market are difficult to pin down completely, a first look at the numbers suggests that the short-run drops in the stock market on the days with most significant news regarding Navalny translates into several billion dollars in lost market value on the Russian stock market. In other words, for people that care about their stock market investments and the health of the Russian economy more generally, it makes a lot of sense to care about the health of Navalny.


Alexei Navalny has become the leading political opponent to the current regime in Russia. His visibility (and possibly support) has been growing as he has endured poisoning, recovery in hospital, and court rulings that have imposed a harsh prison term. At the same time, Navalny and his team have posted new material online to make his case that both the president and other Russian leaders are seriously corrupt.

The question addressed in this brief is whether the news regarding Navalny affected the Russian stock market. The reasons for such a response may vary between different investors but could include a fear of international sanctions against Russia; an aversion to keeping investments in a country that put a nerve agent in the underwear of a leading opposition leader; or that news of a national security service poisoning one of its own citizens could trigger domestic protests that create instability.

This brief only investigates if Navalny-related news or events are taken into account at the macro level in the stock market and if so, how important the news seem to be relative to other news as drivers of the stock market index. However, there is a long list of related questions that are subjects for upcoming briefs that include differential effects across sectors and companies as well as identifying what dimensions of the news stories investors responded to.

Navalny in the News

Since August 2020, news regarding Alexei Navalny’s health and his role as the most important opposition leader in Russia have featured prominently in media around the world. There are different ways to analyze the significance of Navalny in the news and here the readily available measure provided by Google trends will be used. Figure 1 shows a global search on the keyword “Navalny” over the period July 1, 2020 to March 13, 2021 relative to total searches, where the maximum level in the period is normalized to 100 and other values are scaled to this. While the numbers on the graph are just relative measures, not telling much about the actual popularity, or market relevance of searches, the spikes in Figure 1 have very clear connections to major news stories as will be detailed below.

Figure 1. Google trends on Navalny

Source: Google trends, global search on Navalny on 2021-03-18

Four episodes stand out in Figure 1 and are marked by red numbers:

1 (August 20-25, 2020) is associated with Navalny falling ill on the flight from Tomsk to Moscow which led to an emergency landing in Omsk and then going to Germany for specialist treatment where it was stated that he had been poisoned.

2 (September 2-3, 2020) is when the German government said that the poison Navalny was exposed to was Novichok, which was also confirmed by laboratories in Sweden and France.

3 (January 17-25, 2021) is an extended period covering the arrest of Navalny as he returned to Russia on January 17; the publication of the YouTube video on “Putin’s palace”; and the street protests that followed.

4 (January 31-February 5) is a period covering a new weekend of public protests and then on February 2, Navalny being sentenced to prison for not complying with parole rules when he was in a coma in Germany. At the tail end of this period, Navalny’s chief of staff announced that street protests will be suspended due to thousands of arrests and police beatings.

Russian Stock Market Reactions

Using stock markets to investigate the value of political news is not new; for example, Fisman (2001) looks at how news regarding Suharto’s health differentially impacted firms that were connected to Suharto versus those that were not. On a topic more closely related to this brief, Enikolopov, Petrova, and Sonin (2018), show that Navalny’s blog posts on corruption negatively affect share prices for the exposed state-controlled companies. Looking at the overall stock market index rather than individual shares in Russia, Becker (2019) analyzes stock market reactions to Russia invading Crimea.

To get a stock market valuation effect of Navalny news that is as clean as possible, we need to filter out other factors that are known to be important drivers of the stock market. In the case of Russia’s dollar denominated stock market index RTS (short for Russia Trading System), we know from Becker (2019) that it is sensitive to movements in global stock markets and international oil prices. The former factor is in line with other stock markets around the world and the oil dependence of the Russian economy makes oil prices a natural second factor (see Becker, 2016).

Figure 2 shows how the RTS index moves with the global markets (proxied by S&P 500 index) and (Brent) oil prices in this period. The correlations of returns are around 0.4 between the RTS and both S&P500 and oil prices respectively. This figure is also the answer to the obvious argument that the stock market was doing very well in the time period of Navalny in the news, so he could not be a major concern to investors. As we will show below, this argument goes away when the effects of the exogenous factors are removed.

To filter out these exogenous factors, we follow the approach in Becker (2019) and regress daily returns on the RTS on daily returns of the exogenous variables. We then compute the residuals from the estimation to arrive at the excess returns that are utilized in the subsequent analysis. For more details on this, see Becker (2020). Since the estimated model provides the foundation for the subsequent analysis, it is important to note that all of the coefficients are statistically significant, and that results are robust to changes in the estimation period and exclusion of lagged values of the exogenous variables.

Figure 2. RTS and exogenous factors

Source: Data on RTS from the Moscow Exchange (MOEX), S&P500 from Nasdaq, and Brent oil prices from the US energy information administration.

With a time-series of excess returns for the Russian stock market, we can look at the stock market reactions to the four Navalny episodes identified in Figure 1. These periods cover some days for which we cannot compute excess returns since there are days when there is no trading, but all dates in the period are shown in Figure 3 to provide a full account of what stock market data we have for the events. In addition to excess returns during the events that are shown in blue, the day before and the day after the events are shown in light grey. In the first three episodes, the cumulative returns during the events windows were minus 6.2, minus 2.4, minus 6.0 percent, while in the fourth event window it was plus 0.8 (although in this period, the day after Navalny was sentenced to jail, the excess return was minus 1.7).

The correlations between news and excess returns in this brief are based on daily data. Since many things can happen during a day, the analysis is not as precise as in the paper by Enikolopov, Petrova, and Sonin (2018), where the authors claim that causality is proven by the minute by minute data. Although we have to be more modest in claiming that we have identified a causal relationship going from Navalny news to negative stock market returns, the daily data used here provides enough evidence to claim that there is a strong association pointing in this direction. If we take all four events and translate the cumulative excess returns in percent (which is 14) into dollars by using the market capitalization on the RTS at the time of the events (on average around 200 billion dollars), this amounts to a combined loss in market value of over 27 billion dollars.

Figure 3. Excess returns and Navalny news

Source: Excess returns from author’s calculations based on data from the Moscow Exchange (MOEX), Nasdaq, and the US energy information administration. The chart indicates days for which we cannot compute excess returns since not all days are trading days.

We may think that excess returns of this magnitude are common and that what we pick up for the four Navalny episodes are regular events in the market. To investigate this and other potential factors that have been important to explain excess returns in this time period, Table 1 provides a list of all the days when the excess return in the market was minus 2 percent or worse. Between August 2020 and mid-March 2021, there were eight such days. The table also shows what could be an associated Navalny event on or close to those dates as well as other competing factors or news that could explain the large negative returns on these days.

Out of 8 days with strong negative returns, the first three days are very clearly associated with major news regarding the poisoning of Navalny. The fourth day is close to Navalny’s release from the hospital but also when there are discussions about U.S. views on Iran and Ukraine. Two of the days are in the time period of the protests following Navalny’s video on “Putin’s palace” and two more days are related to important international institutions speaking out regarding first the poisoning with Novichok and then about the prison term of Navalny.

Although we would need a more fine-grained look at market data to make a final judgment on the most important drivers of the excess returns of a specific day, the fact that every single day with large negative excess returns is on or close to a Navalny news story is again pointing in the direction of a stock market that reacts to news about Navalny. Furthermore, the most significant drops with less competing news are associated with events that have a direct connection to Navalny’s health and how his life was put in danger. In the list of competing news are Nord Stream, Biden affecting the oil and gas industry, and a law regarding the taxation of digital currencies. They are likely to be of at least some relevance for stock market valuations and could account for certain days or shares of poor performance of the RTS, but it is hard to ignore the general impression of Navalny being important for the stock market in this period.

Table 1. Days with RTS excess returns of minus 2% or worse (August 1, 2020 to March 12, 2021)

Source: Excess returns from author’s calculations based on data from the Moscow Exchange (MOEX), Nasdaq, and the US energy information administration. News comes from internet searches on Navalny and relevant dates.


Although it is difficult to prove causality and rule out all competing explanations, this investigation has shown a strong association between major news regarding Navalny and very poor performance of the Russian stock market. Every day since August 2020 that had excess returns of minus 2 percent or worse is more or less closely associated with significant news on Navalny. More than that, almost all days with significant Navalny news during this period, – as captured by high search intensity of Navalny on Google, – are associated with a poorly performing stock market. In particular, this holds for the day of his poisoning and the following days with comments by international doctors, politicians, and institutions regarding the use of Novichok to this end.

It could be noted that a 1 percent decline in the RTS equates to a loss in monetary terms of around 2 billion USD in this time period since the market capitalization of the RTS index was on average around 200 billion USD. The combined decline in the events shown in Figure 3 is 14 percent and for the days listed in Table 1, it is 21 percent, i.e., corresponding to market losses of somewhere between 28 and 42 billion USD. Even if only a fraction of this would be directly associated with news on Navalny, it adds up to very significant sums that some investors have lost. One may argue that the losses are only temporary and recovered within a short time period (which would still need to be proven), but for the investors that sold assets on those particular days, this is of little comfort. At a minimum, events like these contribute to increased volatility in the market that in turn has a negative effect on capital flows, investments, and ultimately economic growth (Becker, 2019 and 2020). For anyone caring about the health of their own investments or the Russian economy, it makes sense to care about the health of Navalny.


  • Becker, Torbjörn, 2016. “Russia and Oil — Out of Control”, FREE policy brief.
  • Becker, Torbjörn, 2019. “Russia’s Real Cost of Crimean Uncertainty”, FREE policy brief, June 10.
  • Becker, Torbjörn, 2020. “Russia’s macroeconomy—a closer look at growth, investment, and uncertainty”, Ch 2 in Putin’s Russia: Economy, Defence And Foreign Policy, ed. Steven Rosefielde, Scientific Press: Singapore.
  • Enikolopov, Ruben, Maria Petrova, and Konstantin Sonin, 2018, “Social Media and Corruption”, American Economic Journals: Applied Economics, 10(1): 150-174.
  • Fisman, Raymond, 2001, “Estimating the Value of Political Connections.” American Economic Review, 91 (4): 1095-1102.
  • Google trends data.
  • Moscow Exchange (MOEX), RTS index data.
  • Nasdaq, S&P 500 data.
  • U.S. Energy Information Administration, 2021, data on Brent oil prices.

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.

Media Freedom in Eastern Europe

Photography of Magazine representing Media Freedom Eastern Europe

In recent years, press freedom in many Eastern European countries has increasingly come under threat. This policy brief provides an overview of the importance of a free press for democracy and the challenges to media freedom in these European transition economies.


Freedom of expression – which encompasses media freedom – is a fundamental human right enshrined in most countries’ constitutions. Yet for many of their citizens, it is more of an aspiration than a reality. Following the dissolution of the Soviet Union, a number of countries in Eastern Europe embarked on a process of democratisation and accession to the European Union – for which one of the prerequisites is a free press.

Figure 1 shows a measure of press freedom for the eight Eastern European countries that joined the EU in 2004. These countries saw a general improvement in press freedom from the early 1990s to the early 2000s. But since then, experiences have diverged and in 2017 only Estonia and the Czech Republic showed better scores on press freedom than when they first joined the EU. This pattern of backsliding is not confined to the media, but is also evident in other measures of democracy.

Figure 1. Media Freedom in Eastern Europe

Media and Democracy

A free press and a strong democracy are mutually reinforcing. Research, from mainly Western democracies, shows that the media plays an important role in informing the electorate and holding politicians accountable. For example, Snyder and Strömberg (2010) find that U.S. voters are less informed about their Congressmen when they are covered less in the local press. This is ultimately damaging for voters, as these politicians work less for their constituency and these constituencies also receive less federal funding.

Investigative journalism can play an important role in uncovering corruption and other forms of wrongdoing by politicians. For instance, using the Panama Papers and other leaked documents, journalists uncovered 11,562 offshore entities linked to Russia, 2943 linked to Latvia, and 103 linked to Sweden (see: Offshore Leaks Database). While there are legitimate uses for these offshore entities, the lack of transparency surrounding offshore finance also facilitates tax evasion and money laundering. The revelations of offshore holdings became an embarrassment to many politicians, with some forced to resign. In Russian media, the allegations that the leaks document suspected money laundering by President Putin were characterised as US propaganda (Hoskins and Shchelin, 2018).

Figure 2 shows the relationship between the length of time a country’s leader has been in office and its press freedom score in 2020. While there is no systematic relationship between leader tenure length and press freedom in Western Europe (in blue), across Eastern Europe (in red), countries whose leader has been in power for longer tend to have less media freedom. This correlation is likely to reflect three factors: 1) media coverage can affect a government’s chances of staying in power; 2) a longer-lived government might be more able to control the media and 3) a host of other factors, such as the public’s political engagement and the strength of democratic institutions, could influence both freedom of the press and the longevity of governments.

Figure 2. Media Freedom and Leader Tenure

Source: Freedom of the Press, Freedom House. This figure shows the Freedom of the Press total score for the 8 central and eastern European countries that joined the EU in 2004 from 1993 to 2017. A country with a score between 0 and 30 (31 and 60) is designated as having a free (partly free) press.

Electoral Effects of the Media

A number of papers show the causal effects of (biased) media coverage in shaping support for political parties. For instance, watching Fox News increases voting for the Republican party in the US (DellaVigna and Kaplan, 2007; Martin and Yurukoglu, 2017).

Enikolopov, Petrova, and Zhuravskaya (2011) investigate the influence of NTV (the only national TV channel that was at the time independent of the government) on voting in the 1999 parliamentary election in Russia. They find that areas with greater access to NTV were significantly less likely to vote for the government party and more likely to vote for opposition parties.

Biased media can also be used as a foreign policy tool. Peisakhin and Rozenas (2018) find that Ukrainian areas that received Russian TV had on average greater support for pro-Russian parties and candidates in the 2014 elections.

The media landscape in many CEE countries is highly polarised and politicised. Kostadinova (2015) cites research showing that in some former communist countries many journalists still rely on government officials as news sources. In other countries, media in opposition to the communist regimes emerged at the end of the 1980s, such as in Poland where the Gazeta Wyborcza became one of the leading daily newspapers.

Government Control of the Media

Governments have many ways of controlling the media in their country. At the extreme, governments can own and run media outlets, dictate their contents, and censor any dissenting voices. While political and media systems across CEE are diverse, they share some common experiences that might explain their current fragility.

Transitions in Media Ownership

In the Eastern Bloc, the mass media was owned and tightly controlled by the state and used as a tool for propaganda. After the fall of communism, many state-owned media were privatised – along with other state-owned enterprises.  Foreign (mostly western European) media conglomerates purchased a significant fraction of media outlets in a number of countries.

While private and foreign ownership of the media can reduce the government’s ability to influence media content, the experience of CEE was not entirely positive.  Stetka (2012) argues that while foreign owners brought capital and technology, they were less concerned with transplanting Western journalistic and professional standards.  Dobek-Ostrowska (2015) claims that this focus on profit led to the tabloidisation of news across the CEE.

Following the global financial crisis in 2007/2008, foreign investors started to pull out of the CEE media markets and are being replaced by local owners who often have strong links with the government. This is evident in Hungary, where businessmen close to the government have been buying up independent media outlets, including its largest news website, one of two national commercial TV channels, and all regional newspapers (Bede, 2018). The Polish government also aims to “re-nationalise” its media. Plans by a state-run oil company to buy one of the country’s largest media publishers from its German owners were recently approved.

Elsewhere, domestically owned and previously independent media outlets are also being bought by new pro-government owners. In Russia, the formerly independent NTV from the above example was taken over by a state-owned company in 2001 and started to cover the ruling party in the run-up to the following elections in a similarly favourable way to state-controlled TV channels. Gehlbach (2010) argues that Putin’s media strategy is to exert tight control over the news coverage of these three main national television networks, while allowing media outlets with less reach to operate more independently.

In some countries of the region, there is limited information about the ultimate owner of media outlets. Within the EU, Latvia, Hungary, the Czech Republic, Slovakia and Cyprus, are assessed as high risk in terms of transparency of media ownership (Brogi et al. 2020). In 2009, the Swedish company Bonnier sold Diena – one of Latvia’s largest newspapers – to an initially undisclosed investor. A year later, a Latvian businessman acquired a controlling stake in the paper.

Government Advertising

Around the world, traditional news media is facing increased competition from digital platforms and becoming highly dependent on advertising revenue, including advertising from the government and pro-government businesses According to the Centre for Media Pluralism and Media Freedom, there are no clear and fair criteria for the distribution of state advertising to the media in the majority of EU countries – especially those in Eastern Europe (with the exception of Estonia).

Szeidl and Szucs (2021) document how the Hungarian government targeted advertising to friendly media outlets and how these media in turn covered the government more positively.  They also present suggestive evidence that a similar favour exchange between government and the media occurs in nine other Eastern European countries, including Poland.

Two weeks ago, many private Polish media outlets coordinated a media blackout to protest government plans to tax advertising revenues. The media companies complained that the tax would cost them $270m a year, while public media received twice as much from taxpayers.

Public Service Media

The establishment of public service media forms an integral part of the EU’s agenda for promoting press freedom. While public service media are an important and trusted source of unbiased information in many western European countries, they generally play a smaller role in the  Eastern European media markets. Furthermore, no laws are guaranteeing the independence of public service media from the government in eastern EU countries, with the exception of the Baltic states and Slovenia  (see Centre for Media Pluralism and Media Freedom).

Intimidation of Journalists

Governments can also ensure positive coverage by intimidating editors and journalists. Since 1992, 91 journalists were killed, imprisoned, or went missing in Russia, 18 in Ukraine, 15 in Belarus, and 8 in Georgia (data by the Committee to Protect Journalists). While not all of these cases reflect government action, several recent examples illustrate how the judicial system may be used against journalists. For instance, according to the CPJ, ten journalists were imprisoned in November 2020 for covering protests against President Lukashenko in Belarus and one journalist was charged with high treason and espionage in Russia in July 2020.

There are also fears that governments can use defamation laws to deter and punish unwelcome media reports. For instance, the head of Poland’s ruling party filed a libel charge against two journalists from the Gazeta Wyborcza for reporting about his alleged involvement in a real estate project (see, e.g. Council of Europe media freedom alert).


The media plays a vital role in shaping the public debate and holding those in power accountable to the wider population. This power of the media also increases the risk that governments attempt to influence media content.

In recent years, many countries in CEE have seen press freedom come increasingly under threat, undermining some of the progress made since the dissolution of the Soviet Union. Part of the present fragility of media freedom in Eastern Europe may be due to their historical experience. During the transition from communism, many formerly state-owned media companies were sold to private and often foreign owners. In the past decade, local business interests with strong ties to the government started to buy up large shares of the media market in a number of Eastern European countries. Meanwhile, public service media have been less successful at establishing themselves as important and unbiased sources of information across Eastern Europe compared to Western Europe.  To ensure positive media coverage, many governments adopt a carrot and stick approach: state advertising revenues and intimidation of individual journalists.

Article 19 of the Universal Declaration of Human Rights states that “everyone has the right to freedom of opinion and expression; this right includes freedom to hold opinions without interference and to seek, receive and impart information and ideas through any media and regardless of frontiers”. To ensure these fundamental rights, there need to be transparent and fair rules governing the ownership, management, and financing of media outlets and safeguards for individual journalists.


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.

Political Implications of the Rise of Mobile Broadband Internet

Image of 5g broadband tower representing implications mobile internet

In the last ten years, the world has experienced the dramatic rise of mobile broadband internet brought by third-generation (3G) and fourth-generation (4G) mobile networks. This has resulted in major political changes – reduced confidence in governments around the world, lower voting shares of incumbent political parties, and the rise of populists. The empirical evidence is consistent with both the optimistic view of 3G internet (the “Liberation Technology”) and the pessimistic one (the “Disinformation Technology”). 3G internet helps to expose actual corruption; however, it also contributes to electoral successes of populist opposition.

The Spectacular Rise of 3G

Communication technologies have undergone a dramatic change in the last 10-15 years. According to the International Telecommunications Union (ITU), there were only 4 active mobile broadband subscriptions per hundred people in the world in 2007, while this number reached 75 per hundred in 2020. The growth of mobile broadband internet – provided by the third and fourth generation of mobile networks (3G and 4G, respectively) – was the main driver of growth in broadband access. The number of fixed broadband subscriptions per hundred people has only increased from 5 to 15 percent in the same period of time.

Relative to the previous generations of mobile technology, 3G provides a qualitatively different way of using the internet. First, it is broadband access on the go, available wherever the user is rather than at a fixed point at home or in the office. Second, it allows for downloading and uploading photos and videos. Before 3G, mobile technology only allowed exchanging text messages along with limited and slow access to the web. Third, it is the technology that is best suited for social media. While social networks started before 3G and were initially accessed on fixed broadband, today most Facebook, Twitter and YouTube users are mobile.

Liberation Technology or Disinformation Technology?

What are the political implications of the spread of this new technology around the world? Initially, political scientists were excited about the internet as a “Liberation Technology”, especially after it played an important role in the Arab Spring. Internet – and in particular mobile internet –helped pro-democracy activists in autocratic states to disseminate critical information about the government, expose corruption, and coordinate protests.

Later on, however, it became clear that social media also provided a platform for the dissemination of false news and hate speech – thus supporting the rise of populists. This led to a rethinking of the role of mobile internet – and rechristening it into a “Disinformation Technology.”

Which view, the optimistic or the pessimistic one, is correct? In Guriev et al. (2021), we study the impact of the expansion of 3G around the world on attitudes to government and electoral outcomes.

Exposing Actual Corruption

In order to explore the effects on confidence in government, we use data from Gallup World Poll surveys of 840,537 individuals from 2,232 subnational regions in 116 countries from 2008 to 2017. In each region and year we calculate the population-weighted average access to mobile broadband relying on the network coverage data from Collins Bartholomew’s Mobile Coverage Explorer.

First, we find that increased access to 3G internet causes lower confidence in government, judiciary, honesty of elections, and a lower belief that the government is not corrupt. As shown in Figure 1, the magnitudes are substantial. In our paper, we show that a decade-long 3G expansion has the same effect on government approval as a 2.2 percentage-point rise in the national unemployment rate.

Figure 1. Mobile Broadband Access and Government Approval.

Source: Guriev et al. (2021), Table 1, authors’ calculations.

This effect is only present when there is no online censorship and stronger when traditional media are not free. Furthermore, the spread of 3G makes people think that the government is corrupt when the actual corruption is high. In the cleanest countries of the world, the effect is actually positive – better access to information may help citizens to understand that other countries are much more corrupt relative to their own. 

This positive impact is, however, limited to about 10% of the world’s countries. On average, the effect of 3G on the perception that government is clean is negative (see Figure 1). There are two potential explanations. First, as suggested by Gurriv (2018), before the arrival of the fast internet, the elites controlled the media and, as a result, the public was not fully aware of the elites’ corruption. 3G helped to expose this corruption and corrected the pre-3G positive bias. The second explanation is related to the negative bias of social media where critical messages spread faster and deeper (see the references in Guriev et al. 2021).

Another potential explanation is that social media promote overall negative and pessimistic attitudes. We show that this conjecture is not consistent with the evidence: the spread of 3G does not reduce life satisfaction or expected future life satisfaction.

Helping European Populists

The evidence above is consistent with the view that mobile broadband internet and social media help to expose misgovernance and corruption. These findings are in line with the optimistic view of mobile broadband internet as a “Liberation Technology.” However, it turns out that the pessimistic view of “Disinformation Technology” may also be correct.

We examine the impact of 3G expansion on the outcomes of 102 parliamentary elections in 33 European democracies between 2007 and 2018. Using subnational data, we show that the spread of 3G, not surprisingly, decreases the vote share of incumbents substantially (see Figure 2).

 Figure 2. The impact of 3G expansion on incumbent vote share in Europe.

Source: Guriev et al. (2021), Figure VIII.

Figure 3. The impact of 3G expansion on opposition vote share in Europe.

Source: Guriev et al. (2021), Figure IX.

If incumbents lose votes, who picks them up? We show that the main beneficiaries of 3G expansion are the populist opposition parties, both on the left and right (Figure 3). The non-populist opposition does not gain.

Why do populists benefit from the spread of mobile broadband and social media? One explanation is that social media is decentralized and has no entry barriers. It is not the first time in history that populist politicians have relied on new communication technology to circumvent mainstream media controlled by the elites (e.g. the US late 19thcentury populists used telegraph and railroads, the Nazis in Germany used radio). It may also be the case that populist messages may be simpler, and thus, better suited for a short and catchy communication on social media. For example, another pan-European family of anti-system parties, the Greens, do not benefit from the spread of the 3G internet at all (see Figure 3): their narrative is more complex, asking voters to take responsibility for the planet.

Fact-Checking Alternative Facts

Many populist politicians point to actual corruption of the incumbent elites, but some also spread false narratives or “alternative facts.” (It was Donald Trump’s Counselor Kellyanne who, in January 2017, when asked to comment on false statements by Trump’s Press-Secretary about his inauguration, famously said that these were not falsehoods but “alternative facts.”) What can be done to stop the dissemination of these falsehoods on social media? Can fact-checking by mainstream media and independent organizations help?

In two studies, Barrera et al. (2020) and Henry et al. (2021), we carry out two randomized online experiments to identify the causal effects of alternative facts spread by populist politicians and their fact-checking. The findings are as follows: (i) alternative facts are highly persuasive; (ii) fact-checking helps to correct factual beliefs – but do not change voting intentions; even though the voters understand that the populists misrepresent the facts, they still support their agenda; (iii) fact-checking, however, substantially reduces sharing of alternative facts on social media; (iv) the impact of fact-checking on sharing is equally strong regardless of whether the users are forced to view the fact-checking information or are simply given an option to click on a fact-checking link; (v) asking users to re-confirm their intention to share alternative facts with an additional click greatly reduces sharing.

Our results suggest that fact-checking may not be as effective as fact-checkers themselves hope, but can help slow down the dissemination of falsehoods on social media. Furthermore, our analysis delivers clear policy implications – both providing fact-checking (even in the form of accompanying alternative facts with fact-checking links) and requiring additional clicks before sharing can be very effective.


The findings from our analysis of the worldwide spread of mobile broadband internet in the last decade are consistent with both optimistic and pessimistic views. On the one hand, 3G internet does help expose actual corruption. On the other hand, it helps populist opposition to gain votes. Likely, the latter result is eventually due to the populists’ abuse of online platforms for spreading disinformation. We show that the propagation of falsehoods on social media can be at least partially slowed down by fact-checking.


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.

The Economic Track Record of Pious Populists – Evidence from Turkey

FREE Network Policy Brief | A Case Study of Economic Development in Turkey under AKP

In this policy brief, I summarize recent research on the economic track record of the Justice and Development Party (AKP) in Turkey. The central finding is that Turkey under AKP grew no faster in terms of GDP per capita when compared with a counterpart constructed using the Synthetic Control Method (SCM). Expanding the outcome set to health and education reveals large positive differences in both infant and maternal mortality as well as university enrollment, consistent with stated AKP policies to improve access to health and education sectors for the relatively poorer segments of the population. Yet, even though these improvements benefited women to a large extent, there are no commensurate gains in female labor force participation, and female unemployment has increased under AKP’s watch. Of further concern is the degree to which the SCM method applied to institutional measures fail to find any meaningful early improvements along this dimension, and more often than not reveals adverse institutional trajectories.

The Turkish political economy represents something of a puzzle. After a traumatic financial crisis in 2001, a series of political and economic reforms brought higher economic growth and a promise of EU membership. An authoritarian political elite, spearheaded by a military with a troubled past of controversial coups ousting democratically-elected governments, looked set to give way to a new cadre of political and economic elites who, despite a recent past as radical Islamists, seemed to favor free markets as well as democratic reform.

News media, as well as several international organizations, heaped praise on the Turkish government. In some cases, these represented optimistic interpretations of events, whereas in some cases they inadvertently served to spread a misleading picture of the strength of the Turkish economy. A recent World Bank report described Turkey’s economic success as “a source of inspiration for a number of developing countries, particularly, but not only, in the Muslim world” (World Bank, 2014).

Today, the state of Turkey’s political economy is represented very differently. Several international rankings of political institutions (Meyersson, 2016b) and human rights show Turkey spiraling ever lower, following years of stifling freedom of speech, recurring political witch hunts, and escalating internal violence. Lower GDP growth rates, falling debt ratings and exchange rates are evidence less of a rising new economic giant than a stagnating middle income country under increasingly illiberal rule. A recent IMF staff report (IMF, 2016) noted how Turkey remains “vulnerable to external shocks” and a labor market “marred by rapidly increasing labor costs, stagnant productivity, and a low employment rate, especially among women.”

What has been the AKP’s track record on economic growth in Turkey? While some has described it as an economic success (as noted above), others have pointed out that Turkey’s economic development has not been much more than middling (Rodrik, 2015).

Evaluating the economic track record of the AKP faces numerous challenges. The rise to power of the AKP government came in the wake of one of the worst financial crises in modern history and following a number of substantial economic and political reforms. Finding a candidate for the counterfactual, a Turkey without AKP rule, is challenging and looking solely at time series of Turkish development omits significant trends that likely shape its trajectory.

The focus of my new paper (Meyersson, 2016a) is thus to examine the economic and institutional effects of the AKP in a comparative case study framework. Using the Synthetic Control Method (SCM), developed by Abadie et al. (2010, 2015), I estimate the impact of the AKP on Turkey’s GDP per capita by comparing it to a weighted average of control units, similar in pre-intervention period observables. The construction of such a “synthetic control” avoids the difficulty of selecting a single (or a few) comparable country, and instead allows for a data-driven approach to find the best candidate as a combination of many other countries. This avoids ambiguity about how comparison units should be chosen, especially when done on the basis of subjective measures of affinity between treated and untreated units. The method further complements more qualitative research with a research design that specifically incorporates pre-treatment dynamics, which due to the financial crisis preceding the election of AKP to power, is essential. Similar to a difference-in-differences strategy, SCM compares differences in treated and untreated units before and after the event of interest. But in contrast to such a strategy design, SCM allocates different weights to different untreated units based on a set of covariates.

Figure 1. Results for Turkey’s GDP per capita

fig1Note: Upper graph shows Turkey’s GDP per capita compared to a synthetic counterpart. The middle graph shows the difference between the former and the latter (black line) as well as placebo differences for untreated units (gray lines). The lowest graph plots the weights assigned to countries that constitute the synthetic control for Turkey. See Meyersson (2016a) for details.

As shown in Figure 1, I find that GDP per capita under the AKP in Turkey has not grown faster than its synthetic control. A “synthetic Turkey” (upper graph in Figure 1), which went through similar pre-2003 dynamics in its GDP per capita, also experienced an economic rebound very similar to that of Turkey.

This is robust to a range of specifications that in different ways account for the pre-AKP GDP dynamics. Restricting the set of control units to Muslim countries only, reveals Turkey to have actually grown significantly slower than the weighted combination of the Muslim counterparts. Moreover, a comparison of severe financial crises using SCM shows Turkey’s post-crisis trajectory in GDP per capita to be no faster than its synthetic control. The focus on post-crisis recoveries allows estimation of the composite effect, including both the financial crisis of 2001 as well as the election of AKP and, under the assumption that post-crisis – and pre-AKP – reforms were indeed growth enhancing, provides an upper bound for the effect of the AKP.

These results, however, hide some of the more transformative aspects of how the Turkish economy has changed during the AKP’s reign. Focusing on education outcomes, I instead find large positive effects on university enrollment for both men and women. These improvements are mirrored for key health variables such as maternal and infant mortality, and are likely responses to large-scale policy changes implemented by the AKP that are discussed in Meyersson (2016a). The policy changes include the extensive Health Transformation Program (HTP) implemented by the AKP government (Atun et al 2013), as well as mushrooming of provincial universities from 2006 and onward (Çelik and Gür, 2013).

As such, to the extent that the AKP has engaged in populism from a macroeconomic perspective, it has nonetheless also experienced a significant degree of social mobility, especially among the poorer segments of society. An exaggerated focus on economic output risks obfuscating the structural changes in key factor endowments that could very well prove beneficial in the long run. Still, the improved access to these areas has not been followed by improved outcomes in the labor markets, especially for women. The period under AKP has seen significant reductions in both female labor force participation as well as higher female unemployment. This raises concerns over to what extent the Turkish government has been able to put a valuable talent reserve to productive use, as well as allowing women meaningful labor market returns to education.

Figure 2. Results for Turkey’s gross enrollment in tertiary education

fig2Note: Upper graph shows Turkey’s gross enrollment in tertiary education compared to a synthetic counterpart. The middle graph shows the difference between the former and the latter (black line) as well as placebo differences for untreated units (gray lines). The lowest graph plots the weights assigned to countries that constitute the synthetic control for Turkey. See Meyersson (2016a) for details.

An evaluation of the AKP’s institutional effect using multiple institutional indicators, measuring various aspects ranging from institutionalized authority, liberal democracy, and human rights results in a failure to find any durable early positive effects during AKP’s tenure. In the longer run, for all outcomes the overall effect seems to have been clearly negative. Finally, the significant reduction in military rents, whether measured in terms of expenditure or personnel, is illustrative of the degree to which the military’s political power diminished relatively early on, and posits concerns over lower economic rents as another source of friction between the civil and military loci of power in the country.

Overall, the results point to Turkey undergoing a transformative period during the AKP, socioeconomically as well as politically. Even though the initial years of higher GDP per capita growth under the AKP, in absolute terms, dwindle significantly in comparison to a synthetic counterpart, increased access to health and education provide reasons for political support of a government that has extended a socioeconomic franchise to a larger segment.


  • Abadie, Alberto, Alexis Diamond, and Jens Hainmueller, “Synthetic Control Methods for Comparative Case Studies: Estimating the Effects of California’s Tobacco Control Program,” Journal of the American Statistical Association, 105 (2010), 493-505.
  • Abadie, Alberto, Alexis Diamond, and Jens Hainmueller, “Comparative Politics and the Synthetic Control Method,” American Journal of Political Science, 2015, 59 (2), 495-510.
  • Atun, Rifat, Sabahattin Aydin, Sarbani Chakraborty, Safir Sümer, Meltem Aran, Ipek Gürol, Serpil Nazlıoğlu, Şenay Özğülcü, Ülger Aydoğan, Banu Ayar, Uğur Dilmen, Recep Akdağ, “Universal health coverage in Turkey: enhancement of equity,” The Lancet, Vol 382 July 6, 2013.
  • Çelik, Zafer and Bekir Gür, “Turkey’s Education Policy During the AKP Party Era (2002-2013),” Insight Turkey, Vol. 15, No. 4, 2013, pp. 151-176
  • International Monetary Fund, “Staff Report for the 2016 Article IV Consultation: Turkey,” IMF Country Report No. 16/104
  • Meyersson, Erik, 2016a, “’Pious Populists at the Gate’ – A Case Study of Economic Development in Turkey under AKP”, working paper.
  • Meyersson, Erik, 2016b, “On the Timing of Turkey’s Authoritarian Turn”, Free Policy Brief, http://freepolicybriefs.org/2016/04/04/timing-turkeys-authoritarian-turn/
  • Rodrik, Dani, 2015, “Turkish Economic Myths”, http://rodrik.typepad.com/dani_rodriks_weblog/2015/04/turkish-economic-myths.html
  • “The World Bank, Turkey’s Transitions: Integration, Inclusion, Institutions.” Country Economic Memorandum (2014, December).

Electoral Competition and Political Selection

20160118 FREE Network Policy Brief Featured Image

Research in political economy has recently rediscovered that individuals, and not only institutions, matter for economic outcomes: not all politicians are of the same quality. Competition in the political market may have another virtue than its traditional disciplining role: it may favor the selection of good politicians. In a recent paper, Marijn Verschelde and I study the (time-varying) relationship between political competition and the quality of French deputies.

Note: this column draws heavily on a paper written with Marijn Verschelde, which is available here. The figures are taken from it.

How to design the political selection mechanism in order to favor the emergence of “good” politicians? The traditional public-choice literature emphasized the role of institutions in shaping the behavior of politicians, but have paid little attention, if any, to the individual decision maker. From this point of view, it is purely illusive to seek for “morally superior agents who will use their powers in some public interest” (Buchanan, 1989): all politicians are purely self-interested agents who have to be incentivized to act in accordance with voters’ preferences.

However, a recent stream of empirical papers provides evidence that individual politicians matter for economic outcomes (Besley et al., 2011; Jones and Olken, 2005). In other words, as suggested by common sense, they are not all cut from the same cloth. This emphasizes the need not only to shape institutions in order to discipline incumbent politicians, but also to select “good” decision makers.

Several determinants have been shown to affect political selection: the wage of politicians (Besley, 2004; Messner and Polborn, 2004), the transparency of politics (Mattozzi and Merlo, 2007), the institutional flexibility (Acemoglu et al., 2010), and reservation quotas (Besley et al., 2005; Besley et al., 2013). In addition to a well-known disciplining effect (for instance Stigler, 1972, and Becker, 1983), electoral competition is also likely to enhance this selection process as well (Galasso and Nannicini, 2011): competition pushes parties to select candidates of good quality in order to seduce neutral voters, who, contrary to partisans, are sensitive to the quality of the candidates. As recruiting experts are difficult, hence costly, parties will allocate them in the most contested districts only.

In a recent paper, Marijn Verschelde and I investigate the relationship between quality and competition in the case of the French National Assembly. Measuring the quality of politicians is not straightforward. A wide theoretical definition of quality is a combination of competence and motivation. Such a broad definition is challenging to operationalize in empirical work. Some papers approximate quality by variables like level of schooling and experience of politicians (Baltrunaite et al., 2014; Besley et al., 2011; Kotakorpi and Poutvaara, 2011), while others use their absenteeism rate or vote attendance (Galasso and Nannicini, 2011; Gagliarducci and Nannicini, 2013). We innovate by proposing a measure of quality based on productivity, i.e., on what deputies actually do. We gathered from the Assemblée Nationale archive all the information that has been systematically collected over the years for each deputy and for each year from 1958 to 2012: (i) propositions of law, (ii) oral questions, (iii) reports, and (iv) debates in which the deputy has been involved in. From these four aspects of parliamentary work, we compute a composite indicator of productivity (bounded between 0 and 1, higher values indicating higher performance). To measure political competition, we use a measure based on a Herfindahl index of the vote shares at the first round (bounded between 0 and 1, larger values indicating stiffer competition).

Figure 1. Estimated Productivity Conditional on Competition


We document a strong effect of competition on deputies’ productivity, in accordance with the theoretical framework. Figure 1 displays the estimated level of productivity as a function of the degree of competition, controlling for personal (age, schooling, etc.) and political (party, committee, and the likes) variables. Everything else equal, a deputy elected in the a priori most contested district is estimated to reach a productivity close to 0.8, while if she is elected in a safe district her productivity is estimated to be at most 0.6. This means that the productivity of deputies can vary by up to 30% depending on the degree of contestability. This relation holds even after minimizing problems of potential reversed causality and controlling for reelection incentives, indicating that we actually observe a selection effect.

Considering the time span of our dataset, it is interesting to observe how the positive relationship between electoral competition and political selection evolved over time. The most noticeable evolution over the second part of the XXth century in France is certainly the ideological convergence after the 80’s, marked with the reconciliation of the Socialist Party with the market and the rise of the Pensée unique (“one track thinking”, Knapp and Wright, 2001). Such an evolution should have decreased ideological voting and produced more competence-based elections, hence increasing the effect of competition on quality.

Figure 2. Effect of Competition over Time


However, we uncover a clear inverse-U shaped evolution. Figure 2 plots the effect of competition conditional on the year: the effect of electoral competition sharply increased till the 1980’s, but it has continuously decreased since then. How to explain this pattern? A possible explanation is the increased voters volatility, i.e., a higher unpredictability in voters’ behavior. If at first sight this phenomenon indicates an increase in the share of neutral voters, this is however not necessarily the case. Instead of moving from partisanship to neutrality, it is possible that supporters of a party A at election t turn into supporters of party B at election t+1. Partisans disappointed by the behavior of their party while in office might provide an unconditional support to the competing party at the next election. This is consistent with the fact that not any party succeeded to win two national elections in a row since the 80’s.


Our analysis point out a clear positive relationship between electoral competition and the quality of the French deputies, measured by a composite indicator of productivity. Overall, deputies elected in contested districts tend to perform better than others. From a public policy perspective, this result suggests that reforms enhancing political competition (for instance by redistricting jurisdiction in a proper way or by ensuring a fair access to media to competing parties) should be supported. However, we observe that the impact of competition increased till the 80’s, but continuously decreases since then. This opens the door for a vast research agenda, as it indicate that drivers of an efficient selection mechanism are not necessarily stable over time.



  • Acemoglu, D., Egorov, G., Sonin, K., 2010. Political selection and persistence of bad government. Quarterly Journal of Economics 125 (4), 1511-1575.
  • Becker, G.S., 1983. A theory of competition among pressure groups for political influence. Quarterly Journal of Economics 97, 371-400.
  • Baltrunaite, A., Bello, P., Casarico, A., Profeta, P., 2014. Gender quotas and the quality of politicians. Journal of Public Economics 118, 62–74.
  • Besley, T., Montalvo, J., Reynal-Querol, M., 2011. Do educated leaders matter? Economic Journal 121, 205-227.
  • Besley, T., Pande, R., Rao, V., 2005. Political selection and the quality of government: Evidence from south india. PEPP/8, Suntory and Toyota International Centres for Economics and Related Disciplines, London School of Economics and Political Science, London, UK.
  • Besley, T., Folke, O., Persson, T., Rickne, J., 2013. Gender quotas and the crisis of the mediocre man: Theory and evidence from sweden. IFN Working Paper 985.
  • Buchanan, J., 1989. Constitutional Economics, Explorations into Constutional Economics. College Station, TX: Texas A & M University Press.
  • Gagliarducci, S., Nannicini, T., 2013. Do better paid politicians perform better? disentangling incentives from selection. Journal of the European Economic Association 11 (2), 369-398.
  • Galasso, V., Nannicini, T., 2011. Competing on good politicians. American Political Science Review 105 (1), 79-99.
  • Gavoille, N., Verschelde, M., 2015. Electoral competition and political selection: An analysis of the productivity of French politicians, 1958-2012. Condorcet Center working paper series.
  • Jones, B., Olken, B., 2005. Do leaders matter? National leadership and growth since World War II. Quarterly Journal of Economics 120 (3), 835-864.
  • Knapp, A., Wright, V., 2001. The governmentand politics of France. Routledge.
  • Kotakorpi, K., Poutvaara, P., 2011. Pay for politicians and candidate selection: An empirical analysis. Journal of Public Economics 95, 877–885.
  • Messner, M., Polborn, M., 2004. Paying politicians. Journal of Public Economics 88, 2423-2445.
  • Mattozzi, A., Merlo, A., 2007. The transparency of politics and the quality of politicians. American Economic Review 97 (2), 311-315.
  • Stigler, G., 1972. Economic competition and political competition. Public Choice 13, 91–106.




Evaluating the Political Man on Horseback – Coups and Economic Development

Image of a military man standing in the middle of the street representing coups and economic development

In a new paper (Meyersson, 2015) I examine the development effects of military coups. Coups overthrowing democratically elected leaders imply a very different kind of event than those overthrowing autocratic leaders, and these differences relate to the implementation of authoritarian institutions following a coup in a democracy. Although coups taking place in already autocratic countries show imprecise and sometimes positive effects on economic growth, in democracies their effects are distinctly detrimental to growth. Moreover, when coups overthrow democratic leaders, they fail to promote economic reforms, stop the occurrence of economic crises and political instability, as well as have substantial negative effects across a number of standard growth-related outcomes including health, education, and investment.  

Do military coups matter for economic development? After all, successful coups – i.e. where the military or state elites have unseated an incumbent leader – have occurred 232 times in 94 states since 1950 (see Figure 1). Moreover, around a quarter of these overthrew democratically elected governments (Powell and Thyne, 2012). The prevalence of military coups has not been lost on researchers, yet despite an abundance of research aiming to explain the occurrence of coups (see for example Acemoglu and Robinson, 2001; Collier and Hoeffler, 2006 & 2007; Leon, 2014; Svolik, 2012) much less research has focused on its economic effects (two exceptions are the papers on covert US operations during the Cold War by Dube, Kaplan, and Naidu, 2011 and Berger, Easterly, Nunn, and Satyanath, 2013). Olsen (1963), for example, claimed that coups “often bring no changes in policy.” Londregan and Poole (1990), in their panel-data analysis, find no effects of coups on income.

By now, there is mostly a consensus that significant military influence in politics is detrimental for democracy (Dahl, 1971; Huntington, 1965; Linz and Stepan, 1996). Nonetheless, military coups overthrowing democratically elected governments are often met with ambiguity. Western governments have a long history of tacit support for military coups overthrowing democratic governments, be it left-leaning governments in Latin America or Islamist governments in the Middle East and North Africa (Schmitz 2006). Commentators expressing support for coups often do so invoking extreme outcomes to represent the counterfactual to the military coup; if Pinochet had not overthrown President Allende, the latter would have created a Castro-style regime in Chile; if the Algerian army hadn’t annulled the elections in 1992, the Islamist FIS would have turned Algeria into an Islamist dictatorship in the Maghreb, and so on (Los Angeles Times 2006, Open Democracy 2013). Similarly, the fault for the coup and preceding problems fall invariably upon the ousted leader, with the coup constituting an unfortunate, but necessary, means to rid the country of an incompetent, if not dangerous, leader (Foreign Policy, 2013).

Other commentators have pointed out the risks of allowing a military to intervene and dictate post-coup institutions to their advantage; a “Faustian” bargain likely to bring regime stability but no solution to the real underlying problems behind the conflict in the first place. Yet others lament the human rights abuses following coups, and the inherent ineptitude of military leaders in running the economy (NYT, 2013; New Republic, 2013; Washington Post, 2013).

Figure 1. Successful and Failed Coup Attempts by Country and Year

fig1Notes: The graph shows successful (solid circles) and failed coup attempts (hollow circles) by country and year, and aggregated by country (right graph) as well as by year (top graph). A circle in blue means the political regime was classified by Cheibub et al 2010 as a democracy in the year before the attempt and a red circle means they classified the regime as an autocracy.

Military coups tend to be endogenous events, and establishing a causal relation between coups and development is therefore a challenge. The unobservable likelihood of a coup – often referred to as coup risk (Collier and Hoeffler, 2006 & 2007; Londregan and Poole, 1990; Belkin and Schofer, 2003) – may be driven by many factors also affecting a country’s development potential, such as weak institutions, the military’s political power, social conflict, and economic crises etc.

In order to address this problem, I employ several empirical strategies including comparing successful versus failed coup attempts, matching methods, as well as panel data techniques, using a dataset of coup attempts during the post-World War II era. These methods facilitate, in different ways, comparisons of development consequences of coups in situations with arguably more similar degrees of coup risk.

Of significant importance is distinguishing coups when they occur in clearly autocratic settings from those where they overthrow democratically elected governments. I show that a military coup overthrowing a regime in a country like Chad may have very different consequences than a military leader overthrowing a democratically elected president in a country like Chile. In the former, a coup appears to constitute the manner in which autocracies change leaders. In the latter, coups typically imply deeper institutional changes with long-run development consequences.

I find that, conditional on a coup-attempt taking place, the effect of coup success depends on the pre-intervention level of democratic institutions. In countries that were more democratic, a successful coup lowered growth in income per capita by as much as 1-1.3 percent per year over a decade. In more autocratic countries, I find smaller and more imprecisely estimated positive effects. This effect is robust to splitting the sample by alternative institutional measures, as well as to a range of controls relating to factors such as leader characteristics, wars, coup history, and natural resources. As Figure 2 illustrates, the economic effect of coups tend to worsen over time. Extending the analysis to matching and panel-data methods reveal these results to be highly robust.

Figure 2. Relationship between a Successful Coup and Growth in GDP per capita

fig2Notes: The three graphs represent the coefficient on a successful coups on growth in GDP per capita (PPP) between year t-1 and t+s with s given by the x-axis for all regimes(left), autocracies (middle), and democracies (right). Controls include period t-1 values of log GDP per capita, annual growth, log population, PolityIV index, annual change in the PolityIV index military expenditures as a share of GDP, annual change in military exp/GDP, military personnel as a share of population, years since the last coup, total number of previous coups, social unrest, leader tenure, as well as continent and year dummies respectively. See Meyersson (2015) for details.

A commonly held view is that coups overthrowing democratically elected leaders often provide an opportunity for engaging in unpopular but much needed economic reforms. Not only do I show that coups fail at this, but also that they tend to reverse important economic reforms, especially in the financial sector, while also leading to increased indebtedness and an overall deteriorating net external financial position, and an increased propensity to suffer severe economic crises. A documented reduction in social spending suggests a shift in economic priorities away from the masses to the benefit of political and economic elites.

Whereas coups occur mostly in dire situations, their prescriptions, as shown, rarely constitute adequate remedies to the underlying problems, as the institutional changes brought by these events show clear detrimental development consequences. Any short-lived benefit of regime stability a coup brings, comes at a steep economic, political, and human cost in the longer run.


  • Acemoglu, Daron and James A. Robinson, “A Theory of Political Transitions,” The American Economic Review, Vol. 91, No. 4 (Sep., 2001), pp. 938-963
  • Berger, Daniel, William Easterly, Nathan Nunn, and Shanker Satyanath. 2013. ”Commercial Imperialism? Political Influence and Trade during the Cold War.” American Economic Review, 103(2): 863-96.
  • Belkin, Aaron, and Evan Schofer, 2003,“Toward a Structural Understanding of Coup Risk”, Journal of Conflict Resolution, Vol. 47 No. 5, October 2003 594-620
  • Cheibub, Jos ́e Antonio, Jennifer Gandhi, and James Raymond Vreeland, 2010, “Democracy and dictatorship revisited,” Public Choice (2010) 143: 67-101.
  • Collier, Paul and Anke Hoeffler, 2006, “Grand Extortion: Coup Risk and the Military as a Protection Racket,” working paper
  • Collier, Paul and Anke Hoeffler, 2007, “Military Spending and the Risks of Coups d’ ́etat,” working paper.
  • Dahl, Robert A., Polyarchy: Participation and Opposition, Yale University Press 1971.
  • Dube, Arindrajit, Ethan Kaplan, and Suresh Naidu, “Coups, Corporations, and Classified Infor- mation”, Quarterly Journal of Economics, Quarterly Journal of Economics, 2011 (Vol. 126, Issue 3)
  • Foreign Policy, “Blame Morsy,” Michael Hanna, July 10 2013,
  • Huntington, Samuel P., 1965, “Political Development and Political Decay,” World Politics, 386- 429
  • Leon, Gabriel, 2014, “Loyalty for Sale? Military Spending and Coups d’Etat,” Public Choice 159, 363-383
  • Linz, Juan, and Alfred Stepan, Problems of Democratic Transition and Consolidation: Southern Europe, South America, and Post-Communist Europe, Johns Hopkins University 1996
  • Los Angeles Times, “Iraq needs a Pinochet”, Jonah Goldberg, December 14, 2006
  • Londregan, John B and Kenneth T. Poole, “The Coup Trap, and the Seizure of Executive Power,” World Politics, Vol. 42, No. 2 (Jan., 1990), pp. 151-183
  • Meyersson, Erik, 2015, Political Man on Horseback – Military Coups and Development, working paper, http://erikmeyersson.com/research/
  • Olsen, Mancur, “Rapid Growth as a Destabilizing Force,” The Journal of Economic History, Vol. 23, No. 4 (Dec., 1963), pp. 529-552
  • Open Democracy, February 11 2013, https://www.opendemocracy.net/arab-awakening/hicham-yezza/how-to-be-different-together-algerian-lessons-for-tunisian-crisis.
  • Powell, Jonathan M, and Clayton L Thyne, 2012, “Global instances of coups from 1950 to 2010: A new dataset,” Journal of Peace Research 48(2) 249-259
  • Schmitz, David F. “The United States and Right-Wing Dictatorships”, Cambridge University Press 2006
  • Svolik, Milan W., The Politics of Authoritarian Rule, Cambridge University Press 2012.
  • The New Republic, “Egypt Officially Declares What Is and Isn’t Important”, Nathan J. Brown, July 9 2013, http://www.newrepublic.com/article/113792/egypt-president-adli-mansour-makes-constitutional-declaration.
  • The New York Times, “A Faustian Pact: Generals as Democrats”, Steven A. Cook, July 5, 2013

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.

The Political Economy of the Latvian State Since 1991: Some Reflections on the Role of External Anchors

This brief discusses the role of external anchors or goals such as WTO accession, NATO and EU accession in Latvia’s development strategy since 1991. On the one hand the external goals ‘depoliticised’ many potentially contentious areas of Latvian life. On the other hand, some developments would not have happened or would not have happened as fast without the constraints imposed by the external goals. For example liberalisation of the citizenship laws was prompted by NATO accession and the balance was tipped when the rejection of Latvia from fast-track EU accession talks in December 1997 led Latvia to abandon its quota or ‘windows’ naturalisation system. Most recently, Eurozone accession was an externally defined exit strategy from the austerity episode induced by the economic and financial crisis. Today there are no big external goals left to guide policy making. Home grown problems such as inequality require home grown solutions. But even now an external dependency persists. For example a long needed reform of the financing model of higher education has had to wait for a World Bank report published in September 2014 for action to be taken.

On January 1st, 2015 Latvia assumed the Presidency of the European Union. This milestone represents a certain level of maturity of the Latvian state and offers an opportunity for reflection on some aspects of how politics and political economy have evolved in Latvia between 1991 and today.

After Latvia regained independence in 1991, it faced (at least) two political economy challenges: one was to disentangle the economy from the Soviet system in which it had been deeply integrated, and the second, perhaps more difficult challenge, was to create an independent nation state. At a formal level, the solution to the latter challenge appeared straightforward – assume continuity of the Latvian state. Effectively this meant reinstating the pre-war constitution, which was indeed done for the most part. Symbolically this continuity was signalled by, for example, calling the first post-Soviet parliamentary elections held in June 1993 the elections for the 5th Saeima (parliament). The elections for the 4th Saeima had taken place more than 60 years earlier in October 1931.

At a practical level the challenges were more complex – Latvia had had no practical experience of statehood for nearly fifty years and mistakes were made. For example, Latvia initially diplomatically recognised Taiwan rather than the Peoples Republic of China.

There was a presumption that newly independent Latvia should become a market economy but little consensus on how this should be achieved. This is in contrast to Estonia where a group of ‘young market economy Turks’ were able to implement a kind of zero option i.e. zero tariffs, fast privatisation, etc. In Latvia there were strong protectionist sentiments and the initial privatisation was a muddled process.

Advice and advisers were abundant in post-independence Latvia. In the early 1990s, Latvia was awash with international advisers: the IMF and the World Bank were both present, the Germans were advising on a constitution for the Bank of Latvia, the British were active in public administration reform, the Danish advised on research and higher education and so on. Advice was often conflicting with different advisers promoting their own visions of structures as models that Latvia should adopt e.g. on legal and education systems. Today, we see something akin to this in the Eastern Partnership countries such as Moldova and Ukraine.

There was a general sense of the desirability of a ‘return to Europe’ but no plan or strategy. Nevertheless, even without a conscious plan a strategy emerged – namely a strategy of external anchors.

The external goals or anchors that emerged included the following:

  • World Trade Organisation, 1998
  • NATO, 29 March 2004
  • European Union, 1 May 2004
  • Eurozone, 1 January 2014

The most important effect of the external anchors was that they ‘depoliticised’ many potentially contentious areas of Latvian life. This has been particularly important given the fragmentation that has historically dominated Latvian politics. Thus, in the interwar period, no less than 32 different political parties were represented in the Saeima. In the early post-Soviet parliaments, similar tendencies were observed with newly created parties being the winners in terms of the number of seats in the first four elections. The election of 2006 was the first in which the previously largest party returned as the largest party. Between the first post-Soviet election in 1993 and the 2014 election, there have been no less than 17 governments which mostly have been uneasy coalitions of 3 or 4 partners with divergent views and interests. In this context the benefit of external anchors is self-evident.

The external anchors each contributed in different ways: WTO accession contributed to modify the protectionist sentiments that were rife in the early years of independence. Rather curiously, Estonia, which adopted a radical free trade policy right from the first days of independence, had more difficulties in achieving their WTO membership than ‘protectionist’ Latvia. Estonia was obliged to implement additional economic regulations in order to conform to the rules of the WTO and the EU (to which it was committed to join as its WTO application proceeded), and as a consequence, Estonian WTO accession was delayed to 1999. The WTO accession process also gave Latvia’s fledgling Foreign Ministry invaluable experience of multi-lateral negotiation.

Apart from the obvious security benefit, NATO membership was conditional on the creation of the Latvian anti-corruption Bureau (KNAB) and on the liberalisation of citizenship legislation, the latter because NATO was concerned about the prospect of a member state with a large number of non-citizen residents.

EU accession represents the biggest and most significant anchor. The requirement of candidate countries to accept the EU acquis communautaire took huge swathes of economic and social legislation out of the political arena. While the economic criteria for accession presented few difficulties of principle for Latvia – most people were in favour of a market economy – the requirement of respect for and protection of minorities presented problems for many Latvian politicians and liberalisation of the citizenship law was resisted until after 1997 when the rejection of Latvia from fast-track EU accession talks in December 1997 prompted a rethinking of Latvia’s intransigent position on the quota or ‘windows system’.

It is hard to over-estimate the impact of EU accession on Latvia. What would Latvia be like today if it were not a member state of the EU? There are sufficient tendencies even now in Latvia to suggest we would observe something like a tax-haven, off-shore economy, probably with weak democratic institutions. EU accession has saved the Latvian people from something like such a fate.

Even later in Latvia’s largely self-inflicted financial and economic crisis of 2008-10 it was the ‘Holy Grail’ of accession to the Eurozone that politically anchored Latvia’s famous austerity programme.

What of today? The ‘big’ external anchors are used up, and Latvia today:

  • Is the fourth poorest country in the EU with GDP per capita in 2013 at 67% of the EU average (only Croatia, Romania and Bulgaria are poorer);
  • Is a particularly unequal society – Latvia has some of the worst poverty and inequality indicators in the EU;
  • Has a shadow economy at 23.8% of GDP (data on 2013; Putniņš and Sauka (2014)); and
  • Has an internationally uncompetitive higher education system.

These and other problematic aspects of Latvian life and society are home grown and it is hard to imagine external anchors that can improve poverty or inequality, that can reduce the size of the shadow economy, or which can improve the quality of the Latvian higher education system.

Nevertheless, Latvian policy makers seem to be addicted to the external anchor concept and often find difficult to progress without it. The recent experience of reform of the financing of higher education illustrates. Latvia has historically had a funding mechanism for universities and other higher education institutions based entirely on student numbers. The lack of a link between funding and quality has resulted in a Latvian higher education system that is strong on enrolment but low on quality e.g. as measured by peer-reviewed publications. At some level this has been understood and there has been much talk of reform. Although various reports and evaluations have been published, there has been little progress on concrete reform until the Ministry of Education commissioned the World Bank in December 2013 to produce a report on funding models for Latvia. The final report was delivered in September 2014 and action has now been taken to adopt the World Bank recommended three-pillar model where the funding criteria will now include performance and innovation.

Of course, the new model will not solve all the problems of Latvian higher education – far from it – but it illustrates the pervasive nature of policy makers seeming dependency on external anchors.


  • Putniņš, Tālis & Arnis Sauka (2014). “Shadow Economy Index for the Baltic Countries. 2009-2013,” The Centre for Sustainable Business at SSE Riga, May 2014.

The Arab Spring Logic of the Ukrainian Revolution

20140331 The Arab spring logic of the Ukrainian revolution Image 01

Motivated by the unusual patterns and dynamics of the Arab Spring, we construct a model explaining the vulnerability of the newly established incumbent to popular unrest. Using this model for the case of similar protests in Ukraine, we find that the current combination of availability of information, military capacity of the incumbent and his radicalization, together with the opportunity costs of participation in a protest, are likely to result in the formation of new government that is also vulnerable to popular protests. The persistence of the protests after the formation of a temporal government in Ukraine supports this hypothesis. Additionally, as the policy position of Viktor Yanukovych was relatively mild, his potential successor might be more radical. Exponential growth of social media users, reduction of military capacity, relatively high unemployment and the possible radicalization of the Ukrainian President might put the country into an “instability zone” with recurrent protests.

On the night of 21 November 2013 spontaneous protests erupted in Kiev, the capital of Ukraine, after the Ukrainian government suspended preparations for signing an Association Agreement and a Free Trade Agreement with the European Union in favor of agreements with Russia. The movement concentrated on Independence Square (Maidan Naseljenosti) soon took on the name “Euromaidan”. Soon the protest spread to other cities in the country. The initial agenda of closer relations with the EU was soon encompassed in the wider protest against Viktor Yanukovych, elected President in 2010. He fled the country on February 21 under the pressure of popular protests, exacerbating the leadership crisis. Temporary leadership was taken up by the Speaker of the Supreme Rada – Oleksander Turchinov, while new elections were scheduled to take place on May 25.

Despite the successful removal of Victor Yanukovich from power and a promise of new elections in May, the protests on Maidan did not cease. The major factor of uncertainty comes from the very nature of the protests. For many months, it ran without organizers or formal leadership so that the future course of action remains unclear. It is hard to comply with the demands of Maidan, as no clear set of demands are formulated. Though five figures of Maidan: Tymoshenko, Klitschko, Tyagnybok, Yatsenuk and Yarosh remain the most visible, none of them has sufficient support of Maidan. Whichever course prevails – resumed Eurointegration or an alliance with Russia (which became a less likely option)– the number of people who oppose the new course is likely to be enough to fill a new Maidan.

The swift happenings in Maidan are highly reminiscent of the events of the Arab Spring at its crux: it also was a leaderless protest, coordinated mainly with social media, and encompasses people of vastly different socio-economic, political and demographic characteristics.

Using social media technologies, Euromaidan has created an interactive map of logistics (http://maydanneeds.com/) that provides detailed information on and locations of where to eat, makeshift hospitals, information booths, and the barricades. Clicking on the icons of the map, one discovers not only the locations of the facilities but also their needs, which enables coordination of protesters’ efforts to contribute to the common cause. However, just as in case of Tahrir Square or the Tunisian unrest, the common cause is poorly defined: aside from dissatisfaction with Viktor Yanukovych, the protesters exhibited very different preferences for the future course of action, and the three most prominent figures of the protest – Klitschko, Tyagnybok and Yatsenuk – were shunned as they spoke about the common agenda.

The aftermath of the Arab Spring remains unclear for both protesters and the world. The Syrian social unrest has resulted in ongoing violent conflict, while Libyan society still experiences serious problems with the formation of a new government after the murder of Kaddafi and the end of civil war. Tunisia and Egypt were able to choose new Presidents and form new governments. The latter were themselves dismissed soon after they came to power: the first elected post-Mubarak government collapsed in mid-2013 after a year of almost uninterrupted protests. These two cases are especially interesting as constitutional exits of leaders who were in autocratic office for less than one year were generally caused by coups and not protests between 1945 and 2002 (Svolik, 2009).

Nevertheless, we can apply the knowledge acquired there to the new Ukrainian protest we observed on Maidan and try to predict its development by the means of stylized models suggested in Dagaev, Lamberova, Sobolev and Sonin (2013).

Our approach relies on four simple parameters that drive the dynamics of the protests. First, we consider the costs of collective action – the opportunity costs of spending time on Maidan. The second parameter is the military capacity of the incumbent that can be devoted to the suspension of the protest. The higher it is, the more numerous should the protest be to succeed. The third parameter we use is the degree of the radicalization of the incumbent (the difference between his position and the preferred policy of the majority of the population). Finally, we use an information availability parameter (how many people are aware of the place and time of the occurrence of the protest).

With the electric telegraph, a communication tool of the 19th century, information availability was low and many of those who would have been glad to pay the costs of collective actions to replace the incumbent stay at home as they are not aware of the protest taking place. With Facebook and Twitter, the availability of information is much higher. According to our findings, the crucial role in dynamics of contemporary mass actions is played by the ratio of military capacity to the information availability rather than their values per se.

Our framework assumes that each citizen’s decision of whether to participate in the protest or not is based on the difference between her position and the preferred policy of the incumbent. According to this decision, all citizens can be classified into two groups – those who participate in a protest against the incumbent, and those who do not. We define a person, who has the median position among the protesters, as the expected new incumbent. So if the elections were held among the protesters, he would receive the widest support. If the number of citizens participating in the protest is sufficient to overcome the military capacity of the current incumbent, the protest becomes successful, and the expected new incumbent of the protest becomes the new incumbent. The combination of military capacity, opportunity costs and costs of coordination determine the size of the stability zone – a segment of policy space where the incumbent is not vulnerable to mass protest.

The model allows us to predict the dynamics of the protests that is generated by different combinations of the parameters. For illustrative purposes, the availability of information about the protest is proxied by data on Facebook penetration and military capacity is described by the number of military personnel per capita in 2009 collected by the International Institute for Strategic Studies (Hackett, 2010). The incumbent policy position is proxied by the Legitimacy Index from Polity IV, where a higher index corresponds to lower legitimacy of the incumbent and his regime. Finally, the costs of participation in a protest are proxied by the employment rate. A recent study by Campante and Chor stresses unemployment as important determinants of opportunity costs of taking to the streets during Arab Spring (Campante & Chor, 2012). As unemployed individuals have fewer options of how to spend their time, one should expect that a substantial number of unemployed people corresponds to a relative ease of sparking unrest.

Using these parameters, we can explain success or failure of the protest, and predict some proprieties of its aftermath.

For example, high military capacity, opportunity costs and costs of coordination generate a broad stability zone, so that even a radical incumbent would not face a threat of revolution. The decline of any of three parameters can narrow the zone of stability and make the autocrat vulnerable to mass protest. As the incumbent is highly radical, a significant part of the population takes to the streets. As a result, the new incumbent’s position is sufficiently close to the one of the median voter and is, thus, inside the stability zone. An example of such a scenario is the overturn of Slobodan Milosevic after the fall of communism, when there were eight failed and one successful attempts to form a wide coalition of opposition parties (Spoerri, 2008). The process of finding a common ground started in 1990 with the emergence of the coalition of six parties, the Associated Opposition of Serbia, which broke shortly after a series of power struggles, policy disagreements, and personality clashes. It was only ten years later that the protest which facilitated Milosevic’s downfall took place, as the leader of the united opposition, the Democratic Opposition of Serbia, was able to ensure the non-involvement of the crucial military unit on behalf of Milosevic (Bujosevic & Radanovic, 2003).

In contrast, the events of the Arab Spring had different political dynamics. Low military capacity and high unemployment of Egypt and Tunisia determined a narrow stability zone ex ante. The absence of protests of these long-lived regimes can be explained by the relatively moderate position of the incumbent. In 2010, the Egyptian and Tunisian regimes had scores of 5 and 4 (out of 12) of Political Legitimacy from Polity IV, respectively. However the emergence of social media that enabled the users to coordinate their actions easily narrowed the relatively small stability zone. As the incumbent was less radical, fewer citizens benefit from its replacement and take to the streets. Thus, the new incumbent is defined by protesters, her policy position is more radical and now is out of the stability zone. The new incumbent immediately faces the new social unrest.

Table 1 presents the stylized results of our study. Locating the combination of parameters of the country in the table allows us to make the prediction about the dynamics of the protest. There are several possible courses of events: the protest can be weak and die out soon, with the incumbent staying in place; it can be significant, but yet not large enough to overthrow the incumbent; it can lead to the replacement of the incumbent, followed by the period of stability; and, finally, it can result in the replacement of the incumbent, but not cessation of the protest.

Table 1. Protest Outcome as a Function of Parameters

What do our findings tell us about Ukraine? The previously used proxy for the information index there is not a good choice, as the majority of users prefer the Russian version of Facebook – Vkontakte – as the major social network of the country. Thus, we rely on the Vkontakte penetration data (as of 2013, 18.5 million of people in Ukraine were using the network, constituting 40.6% of the population, see report of Ukranian IT-news agency AIN.UA: http://ain.ua/2013/11/28/503853).

The military parameter, that reduces the likelihood of successful protest, is low, compared to the countries of Arab Spring and constitutes 2.8 active military per 1000 people (see the Ukrainian law “Armed Forces of Ukraine for 2013”, http://w1.c1.rada.gov.ua/pls/zweb2/webproc34?id=&pf3511=49126&pf35401=283834), which is half of the one in Egypt at the beginning of the protests. The unemployment parameter fell to 8.6 during the incumbency of Victor Yanukovych, which corresponds to the pre-protest unemployment in Egypt in 2010.

The legitimacy measure presents a difficulty for comparison with the Arab Spring cases, as the Legitimacy Index has not yet been updated. However, the harsh actions of Victor Yanukovych during the 2013-2014 protests (including the suppression of the protest and the passage of laws denying freedom of assembly and freedom of expression, as well as the refusal to repeal his earlier changes of the constitution towards more presidential form of government) suggest that his legitimacy level fell by approximately 50% (and was about 20% right before he was ousted from power), which is corroborated by polls (“Ukraine’s future in peril under President Yanukovych”, The Washington Post, 2 December, 2013). Thus, the conservative estimate is that his legitimacy index shifted from 3 to 4 or 5, as shown on Figure 1.

For the purpose of comparison, we plot the Arab Spring countries and Ukraine in the space of our variables in Figure 1. The X-axis shows the incumbent’s departure from the median population-preferred policy (proxied by the Legitimacy Index from Polity IV). We use the value of the index in the year prior to the start of unrest, and the index value in 2010 for countries with no protests (Morocco, Oman, Djibouti). The Y-axis corresponds to the employment level. The size of the bubble corresponds to the ratio of military capacity and Facebook (or Vkontakte) penetration (data from the Arab Spring Social Media Report).

The shading of the bubble reflects the type that country belongs to: striped (no significant protest), light gray (continuing protest), and dark grey (multiple protests). Syria is excluded from the classification and is marked white, because of the civil war and international intervention.

Figure 1. Legitimacy index (X-axis), Employment (Y-axis), Military capacity / Social media penetration (size of the bubble) in Arab Spring countries and UkraineFigure 1

Figure 1 illustrates that countries appear in tight clusters in line with our theoretical predictions. The countries with continuing protests that did not lead to the downfall of the incumbent are divided into two groups. The first group (Kuwait, Jordan, Lebanon, and Bahrain) has relatively a moderate incumbent policy position and extremely high level of development of new media. The reasons why these countries are not «striped» is high military capacity of the government that could be employed against protesters, combined with high opportunity costs of protesting (low unemployment rates).

The second group of countries (Syria, Algeria, Iraq, Yemen, and Mauritania) has more radical incumbents and higher unemployment rates (so the incentives to protests are higher there), but is poor in terms of IT development. The ratio of military capacity and Facebook (Vkontakte) penetration is high, which is reflected by the size of bubbles that are much larger than in the first groups. That is why in a country with a small military capacity (such as Yemen), the protests did not lead to the incumbent’s replacement.

Two Arab Spring countries belong to the “multiple protests” group. Both Egypt and Tunisia had relatively mild incumbents in the pre-protest era, with Tunisia’s Bashar al-Assad being the milder of the two. Both countries had relatively high unemployment rates and wide Facebook coverage, both factors alleviating the problem of organizing a collective action. Despite the fact that before the start of the protests Facebook coverage in Egypt had close to average values among the countries of Arab Spring, they grew at exponential rates and Egypt attained leading positions in the region in usage of new media several months later. Moreover, low rates of military capacity made protest activity less risky in both Tunisia and Egypt. The remaining differences in Facebook coverage and employment rates in Egypt and Tunisia account for the different structure of recurrent protests, predicted by our model.

Comparison of the Arab Spring countries data with the Ukrainian case shows that high military capacity, new media penetration and unemployment generate even more narrow stability zones than one observes in the cases of Egypt and Tunisia. The reason why the incumbent was not vulnerable to mass protests can be explained by the political legitimacy of Yanukovych as the winner of relatively free elections in 2010.

But if the Arab Spring protests were triggered by rapid growth of new (cheap) communication technologies, the successful protest against Yanukovych can be explained by his radicalization. The radicalization took the form of parliamentary acts that put significant constraints on political rights and civil liberties and violent suppression of dissident actions.

Employing the proposed approach and contrasting the Ukrainian case with the countries of the Arab Spring allows us to draw several conclusions.

Firstly, the current combination of availability of information, military capacity of the incumbent and his radicalization, together with the opportunity costs of staying on Maidan, are likely to result in successful and recurrent protest. The persistence of the protests after the formation of a temporal government supports this hypothesis.
Secondly, it is worthwhile to note that as the policy position of Viktor Yanukovych was relatively mild, his potential successor might be more radical.

Thirdly, the exponential growth of social media, the reduction of military capacity and relatively high unemployment puts Ukraine into an “instability zone”. This implies that the 2004 scenario of the Orange Revolution is unlikely to repeat. The protest of 2004 resulted in a general election, and the elected president Viktor Yushchenko served his term without interruption. The protests of 2013 are more likely to result in a rapid change of incumbents and a period of instability.

One factor can strengthen the possible incumbent’s vulnerability. The external pressure of the Russian government reduces costs of collective resistance to the new Ukrainian authorities among pro-Russian citizens, while the promise of Western countries to support fast EU integration can incentivize politicians to accelerate reforms opposed by significant parts of the population.


  • Bujosevic, D., & Radanovic, I. (2003). The fall of Milosevic: the October 5th revolution. Palgrave Macmillan.
  • Campante, F. R., & Chor, D. (2012). Why was the Arab world poised for revolution? Schooling, economic opportunities, and the Arab Spring. The Journal of Economic Perspectives, 167–187.
  • Dagaev, D., Lamberova, N., Sobolev, A., & Sonin, K. (2013). Technological Foundations of Political Instability. Centre for Economic Policy Research Working Paper Series
  • Hackett, J. (2010). The Military Balance 2010: The Annual Assessment of Global Military Capabilities and Defense E conomics. London: The International Institute for Strategic Studies.
  • Spoerri, M. (2008). Uniting the opposition in the run – up to electoral revolution – Lessons from Serbia 1990 – 2000. Totalitarismus Und Demokratie, 5(2005), 67–85.
  • Svolik, M. (2009). Power sharing and leadership dynamics in authoritarian regimes. American Journal of Political Science, 53(2), 477–494

Political Instability in Fragile Democracies: Political Cycles Kyrgyz Style

Political Cycles Kyrgyz Style Image

Democratization is rarely a straight and predictable process. Freedom House data from the Central and Eastern European Countries (CEEC) and the countries of the Commonwealth of Independent States (CIS) since 1991 reveals two distinct patterns. In one set of countries, democratization took root quite quickly and the transformation of political institutions seems quite deep and sustainable. In the other countries, the road to democratization, if ever started, has been much more partial and full of reversals. Among the CIS countries, none is regarded as free by Freedom House in 2012, four are regarded as partly free (Armenia, Kyrgyz Republic, Moldova and Ukraine), while the remaining seven countries (Azerbaijan, Belarus, Kazakhstan, Russia, Tajikistan, Turkmenistan, and Uzbekistan) are regarded as non-free. There has also been volatility over time within countries. Russia and Belarus have seen their score steadily deteriorating, while countries on the Balkan and south-east Europe have seen gradual improvements. With the lack of consolidated democratic institutions has also typically followed much political instability. Frequent changes in power, civil unrest, popular revolutions and military conflicts have pervaded countries like Ukraine, Georgia, and the Kyrgyz Republic. In other nations, repressive leaders have put a lid on visible instability, but at the cost of political rights and a fair judiciary system. In both cases, the economy has suffered as instability has deterred investors looking for a predictable environment guided by transparent rules of the game implemented equally for all. Corruption has flourished and political connections and nepotism has determined the opportunities for economic success.